i' JT « •^ 7^^ c < « > f 5 ^^^\'A± 5C ^ ^ -3 m. «r r«. i ''-iT -^ Mr. C. Tate Regan, M.A., F.Z.S. Exhibition of lantern-slides of a Flatfish (Vlatophrys iwdas) showing its power of changing its colour and markings - 556 Dr. R. W. SiiuFELDT, C.M.Z.S. Exhibition of sMns of two Virginia Opossums (Bidelphis virginiana) 556 Mr. D. Seth-Smith, F.Z.S. Exhibition of lantern-slides of the display of the male Pea- cock Pheasant {Tolyplectron chinqui^. Also of photographs of the young Cariama cristata hatched and reared in the Gardens. (PI. LXIX.) 557 The Secketaky. Exhibition of a liviug specimen of a young female Dorsal Hyrax {Dend^roliyrax dorsalis) , 671 The Secretary. Exhibition of photographs of an Elephant Kraal in Siam 671 Mr, R. I. PococK, F.R.S., F.L.S., F.Z.S. Exhibition of a skin and a living specimen of a fawn variety of the Brown Rat {Ejoimys iiorvegicus) .... , 671 Mr. D. Setii-Smith, F.Z.S. Exhibition of epidermal sheaths shed from the base of the lower mandible of the King Penguin {Aptenodytes pennanti) 671 Dr. Francis Ward, F.Z.S. Demonstration, illustrated by lantern-slides and the cine- matograph, of a method of observation of fishes, birds, and mammals under water. 672 PAPERS. 26, Observations on some Alcyonaria from Singapore, with a brief Discussion on the Classification of the Family Nephthyidse. By Edward W. Siiann, B.Sc, Demonstrator and Assistant Lecturer in Zoology in the Victoria University of Manchester, (Pis. LXI.-LXIII.) 505 27. Some early Fossil Cirripedes of the Genus Scalpellum. By Thomas H. Withers, F.G.S. (Text-figs. 64 & 65.) 528 Contents continued on page 3 of Wrapper. ■^-.:^; P.Z.S. 1912. PI. LXI. Qrout Intaglio et Imp. ALCYONARIA FROM SINGAPORE. P.Z.S. 1912. PI. LXII. Grout Intaglio et imp. ALCYONARIA FROM SINGAPORE. P.Z.S. 1912. PI. LXIII. Grout Intaglio et Imp. ALCYONARIA FROM SINGAPORE. PROCEEDINGS GENERAL MEETINGS FOR SCIENTIFIC BUSINESS ZOOLOGICAL SOCIETY OF LONDON. PAPEES. 26. Observations on some Alcjonaria from Singapore, with a brief Discussion on the Classification of the Family Nephthyidse. By Edwaed W. Shann, B.Sc, Demon- strator and Assistant Lecturer in Zoology in the Victoria University of Manchester *. [Received January 17, 1912 : Read March 19, 1912.] (Plates LXI.-LXIII.t) Index. Systematic : Pao-e Scleropliytum pinnulatt»n,^\}. n 507 Nephtliya hedfordi, sp. n 514 Stereoneplithya hitea, sp. n 521 WrighteUa rohusfa, sp. n 525 Nephihi/a : Historical Summary of 510 DendronepJitJiffa do. 518 Stereonephthya do. 520 A collection of Alcyonaria, made by Mr. W. F. Lanchester and the late Mr. F. P. Bedfoi'd during theii' residence in Singa- pore, was sent to the Zoological Depai-tment at the Victoria University of Manchester. Professor S. J. Hickson kindly * Communicated by Prof. S. J. Hickson, D.Sc, P.R.S., F.Z.S. t For explanation of the Plates see p. 527. Peoc. Zool. Soc— 1912, No. XXXIV. 34 506 MR. E. W. SnANN ON entrusted to me the task of identifying the specimens in the collection. The collecting was carried on in shallow water around the coast of Singapore ; the precise localities will be cited under the description of the several species. Several examples of Nephthyidte occur, and this necessitated facing the vexed ques- tion of the classification of that family ; I shall state below my reasons for adopting the scheme of classification advanced by Kiikenthal. I wish to express my indebtedness to Prof. Hickson for his constant advice as to my procedure and help in the pursuit of literature bearing upon the subject. In the last-named branch of the work I had frequent recourse to an exceedingly useful catalogue of the Alcyonaria prejDared by Dr. J. Stuart Thomson. My thanks are also due to Mr. J. T. Wadsworth for prepai-ing the excellent photographs from which Plate LXII. figs. 7, 8, & 9 were made. In all eleven species are described, of which four are new. The systematic positions of the several species are as follows : — Order ALCYONAOEA Verrill {pro 2xirte). Family ALCYONiiDiE. Sclero^jhytum 2nnn7 datum., sp. n. Family Tblestid^. Telesto ricpicola F. Milller. Famil}'^ JSTephthyid^. Nephthya hedfordi, sp. n. Dendronei^lithya diHciformis Kiikenth. Stereonephthya hctea, sp. n. Farasjyongodes crassa Kiikenth. Family Siphonogorgtid/E. Si2yhonogorgia variabilis Hickson. Order GORGONACEA. Suborder Pseudaxonia, Family ScLEROGORGIIDiE. Suherogorgia suberosa Pallas, Family Melitodidje. Melitodes cdbitincta Ridley. Psilacabaria yraciUima Ridley. Wrightella robusta, sp. n. ALCTONARIA FROM SINGAPORE. 507 Older ALOYONACEA. Family Alcyoniid^. Genus Scleeophytum. So. PiNNULATUM, sp. n. (PI. LXII. fig. 7 ; PI. LXIII. fig. 10.) A single complete colony was taken just below low-tide mark from Blakang Mati. In its form and mode of branching the colony closely resembles Sc. ]xdmatiLm Pratt (1903). As in the last-named species, the colony (fig. 7) is erect, branched, and shows a marked lateral compression. The total height is 88 mm., of which the stalk comprises 45 mm. The diameter of the stalk is 31x11 mm., and that of the capitulum is 60 x 35 mm. The stalk divides at its distal end into two almost equal primary branches. From these branches spring numerous secondary lobes which are arranged in no definite order. The secondary lobes frequently show lateral compression in their proximal regions, but their termi- nations are usually bluntly conical, while a few are reniform. The length of these lobes is 12-20 mm., and their diameter 6-9 mm. The colour of this specimen in alcohol is slate -grey. The consistency is tough ; the stem is hard and brittle, but the secondary lobes are soft and fleshy. The autozoids are of a brown colour and are uniformly scattered over the surface of the secondary lobes. The average distance between two adjacent polyps is "6 mm. Autozoids are also found on the primary branches almost down to the level of the stalk, but in this region the distance between them is very much greater, viz. 3-4 mm. Many of the autozoids ai-e more or less expanded, but none of them are situated on raised rounded areas such as occur in Sc. palmaUmi. The anthocodiee are of medium size when compared with those of the genus Sclero2)h7jUim as a whole ; the diameter of an expanded crown is 1 mm., but the polyj)-heads apjDear only half that size when contracted. The tentacles (fig. 10) form the distinctive feature of this species. They are '55 mm. in length, compressed, and expanded at the distal end. There is a single row of free well- developed pinnules on each side of a tentacle. The pinnules are "06 mm. in average length, but the larger ones attain a length of -10 mm. and a breadth of -035 mm. The larger pinnules are those which are placed third, fourth, and fifth from the distal end ; they become shorter as they approach the base of the tentacle. There are about twelve pinnules in each row. The stomodseum is long and convoluted. The mesenteries and mesenterial filaments are well marked. No sexual organs were observed. The siphonozoids, if present, are extremely degenei-ate. A very few minute diverticula of the superficial canals, which point in the direction of the surface, but never open to the exterior, 34* 508 MR, E. W. SnANN ON may represent rudimentary siplionozoids. Tlie superficial canals form a dense netwoi'k just beneath the surface of the colony ; their diameter is •044--120 mm. The vessels of the internal system are clearly defined and circular in cross-section, and are fairly numerous. Zoochlorell?e are numerous in the superficial canal-system, and occur in less abundance in the interior. They are also found in the endoderm of the poljqis, and extend into the tentacles, occupying even the lumen of the pinnules. The spicules are quite characteiistic of the genus Sclerophytum, and indeed only differ from those of Sc. pcdmattim in almost insignificant detail. The little knotted clubs average "1 6 X "034 mm., and the spindles are sometimes only '058 x '015 mm. These minute spicules only occur immediately beneath the surface. The spicules of the coenenchyme are fairly numerous and are all of the tuberculate warted type ; they vary considerably, however, in shape and size. The majority are spindles which narrow rapidly towards their ends to rather acute points ; their measiu^e- ments range from •4x'll mm. to 3"0x45 mm. Irregularly branched forms are by no means uncommon. The following considerations are advanced as an apology for the creation of a new species of Sclerophytum based on the examination of a single specimen. This colony had been assigned in a preliminary investigation to the species 8g. pahnattmn Pratt (1903), and at first sight this diagnosis appeared perfectly accurate. The external appearance, and indeecl the actual measurements, agree closely with the description of the type specimen. The characters and distribution of the spicules, the well-marked mesenteries and mesenterial filaments, the riidimentary condition of the siphonozoids, the orientation of the canal-systems, and the distribution of the zoochlorellsB, all tend to enhance the resemblance between the two species. It is only when the autozoids are examined tha,t the true specific difference is realized ; those of So. pcdmatimi are distinctly larger than those of Sc. pinnulatum. The characters of the tentacles which foi'm the essential divergence of the two species are tabulated below : — Sc. pcdmatimi Pratt. Sc. pinmdattmi, sp. n. •7 mm. in length. '55 mm. in length. Almost of uniform length. Expanded at distal end. Possess a double row of rudi- A single row of free well-deve- mentary pinnules down loped pinnules down each each side. side. Sc. pmoiulal'um could be confounded with no other species of Sclerophytum. At the same time, it is of interest to note that a single row of free pinnules has been recorded on either side of the tentacles of Sc. viride by Thomson and Henderson (1906). The possession of free pinnules, now recorded for two species of ALCYONARIA FROM SINGAPORE. 509 Sderophytutn, tends to strengthen the relationship of the genera Sclero'phytum and Xenia which has been suggested by Pratt (1903). Order ALCYONACEA. Family Telestid^. Genus Telbsto. T. RUPicoLA F. Midler. Carijoa rupicola F. MUller, Arch. Naturg., Jg. 33, 1867, p. 33, tab. 9. figs. 56 & 57. Telesto {Carijoa) rupicola Wright & Studer, ' Challenger ' Reports, Zool. vol. xxxi. 1889, p. 262. Telesto rupioola May, Jena Zeitschr. Naturvv. vol. xxvi. 1900, p. 58. Telesto rupicola Hickson & Hiles, Willey's Zoolog. Res. 1900, p. 496, tab. 50. figs. 1 & 2. Telesto rtipicola Thomson & Henderson, Marine Fauna of Zanzibar, 1906, p. 434. Telesto rujncola H. Laackmann, Zoolog. Jahrb. Supp. 11, Heft 1, 1908, p. 81, Taf. 2. figs. 1, 2 ; Taf, 3. fig. 3. Although the only species of Telesto previously recorded from Singapore is T. prolifera v. Koch, the numerous small colonies in this collection appear to bear a closer resemblance to T. rtcpicola, and are therefore described under that name. The largest of these colonies is of a grey hue. Its longest axial polyp measures 85 mm. in length, it is 3 mm. in diameter at the base and 1*25 mm. at the top. It bears six lateral polyps, the longest of which measures 24 mm. Anthocodite arise at frequent, but pretty regular intervals, both from the axial and lateral j)olyps; their average length is 3 mm., and breadth 1*5 mm. The majority of the other specimens are pale yellow and of smaller size. All these forms were obtained in shallow water near Singapore. Some of the exact localities read as follows : Pulo Brani, 6 fms. ; Pulo Brani, 5-10 fms. ; Blakang Mati, below low tides ; Tanjong Pagar, 10 fms. Previously recorded from Rio de Janeiro (F. Midler); Brazilian coast (Munich Museum) ; Blanche Bay, New Britain (Hickson & Hiles) ; Zanzibar (Thomson & Henderson) ; Bahia (' Chal- lenger '). The spicules show a very wide variety of forms. ISTot only is this the case, but a diflferent selection of spicules was found in each of the five specimens examined. One specimen contained spicules very much resembling those of T. riisei. Indeed, the two species are probably very closely related, for Laackman (op. cit. pp. 72 & 82) is at some pains to distinguish between them. It should be remembered, however, that 1'. riisei has not yet been recorded from the Old World. A very small specimen of a pale yellow colour was examined 510 MB. E. W, SIIANN ON and found to hear many points of resemManco with T. proUfera ; the spicules, for instaiice, agree well with those ilgnred hy Lsiackmann for that species (oji). clt. p. 87), and the stem walls are very thin. At the same time, it was taken from the same locality (Pulo Bra-ni) as one of the larger specimens which ini- donbtedly belongs to the species T. rwpicola. Moreover, another specimen was intermediate in every particular between the above exceptional form and the larger examples, Avhich had been assigned without difficulty to T. rupicola. As this small specimen was evidently very young, one would not wish to attach too deep a significance to the observations made or to draw any hasty conclusions from them. At the same time, it is well within the range of possibility that the accumulation of such knowledge may lead eventually to a reduction in the number of species of Telesto. An Historical Summary of the Genern. JVephthj/a, DendronejyJdhya, and Stereo)iephthi/a, with reasons for i'eta,ining the definitions of these genera of the family Nephthyidje as enumerated by Kiikenthal. Genus Nephthya Savigny. 1817. Nepldhee Savigny, Descr. de I'Egypte, Hist. Nat. Suppl. i. Atlas, Polypes, tab. 2. fig. 5. 1828. Neplhihea Audouin, Ex})lication sommaii'e dcs Planches de polypes de I'Egypte et de la Syi'ie, ])ubliees par Jvdes- Cesa.r Savigny dans : Description de I'Egypte, vol. xxiii. Paris. 1834. Nephthya Ehrenberg, Die Corallenthiere des Eothen Meeres, p. 284. 1846. Nephthi/a Dana, 'Zoophytes,' Philadelphia, p. 610. 1857. Nephthya Milne-Edwards, Histoire Naturelle des Coral- laii-es, vol. i. p. 127. 1877. Nephthya Klunzinger, Die Korallthiere des Ilothen Meeres, Th. l,p. 33. 1887. Nephthya Studei-, in Arch. Naturg., Jg. 53, vol. i. pp. 19, 20. 1887. Nephthya (paivs), Danielssen, in Norskc Nordhavs Exp. vol. v. p. 82. 1889, Nephthya and /.V^^ov^c/OfZes (pars) Wright & Studer, 'Chal- lenger' lleports. Zoology, vol. xxxi. p. xxv, 1895. S'pomjodes (pars) Holm, in Zool. Jalirb., Bd. 8, p. 24. 1895. ISpongodes (pars) Kukeuthal, in Zool. Anz., Jg. 18, ]). 428. 1896. Nephthya Kiikenthal, in Abh. Senckenb. Ges. Frankfurt, vol. xxiii. pp. 89-91. 1899. Nephthya May, in Jena Zeitschr. Naturw. vol. xxxiii. p. 156. 1903. Ne2>hthya Kiikenthal, in Zool. Jahrb., Bd. 19, p. 141. In 1817 Savigny described two genera of Nephthyida) to which he gave the names Amwothee and Nephthee. Audouin (1828), ALCYONAUIA FROM SlJIQArOHE. 511^ to wliom fell the task of describing Savigny's plates, believed that Savigny's Tab. 1. fig. 8 represented Ammothee, and tliiit Tab. 2. figs. 5 & 6 represented Nephthee. The genus Nephthea, as the author wrote it, was recognised by Ehrenberg (1834) ; but n,t the same time he disputed the interpretation of Havigny's pliites, maintaining that Audouin had given the name Nephthea cordlerii to the form, represented in Tab, 2. fig. 6, which Havigny had intended to call Ammothee. Ehrenberg's view has been accepted by all subsecpient authors, and it is now generally agreed tliat Tab. 2. fig. 5, correctly designated by Audouin Nephthea chahrolii, represents the ty|)e of Savigny's genus Nephthee, while Tab. 2. fig. G represents his type of the genus Amviothee, namely A.vlresce'ihS. There can be no reasonable doubt with regard to the authenticity of origin of the genus Nephthi/a, that is to say, that the genus was based on the description of the species N. chahrolii, which is figured in Savigny's Tab. 2. fig. 5. In the case of Ammothee, or Aminothea as the genus was known for many years, the name was changed to Lithop)hytum by Kukenthal (1903), since that author found that Savigny's type species, A. virescens, is identical with a form described forty-two years previously by Forskiil under the name Lithophyton arboreum ; in deference to the International Rules of Zoological Nomenclature the older name must he retained. Thus it is of little moment whether or not Ehrenberg was justified in disputing Audouin's interpretation of Ta,b. 2, fig. 6. Copies of Savigny's plates are extremely scarce, so that it is not always possible for the research worker to examine the original figures ; many have probably been compelled to content th(!mselves with descriptions by other authors. With this (liHiculty in view. Professor Bourne very kindly had photogra,phs taken for Professor Hickson from Savigny's plates, Tab, 2. figs. 5 (fe 6, in the Radcliflfe Library at Oxford. Prof. Hickson lias given me permission to publish these figures in this paper, so that they may be readily accessible to all workers on the Nephthyidte. They are reproduced in PI. LXI. figs. 1-5 and PI, LXII. fig, 6. Ehrenberg (1834) distinguished Nepihthya from Ammothea by the prominence of the polyp-spicules in the former genus, for he says of Nephthya : — " poly pis in verrucas inermes retractilibus," We see, then, that Ehrenberg recognised the distinction between the genera. Nej^hthya a.nd LiiJurphytimi {Ammothea) which obtains at the pre-sent (la,y, namely the jire.sence and absence of armed polyps (polyps with " Stiitzbiindel ") in these genera respectively. Ehrenberg's definititm of Nephthya was recognised by Dana (1846), Milne-Edwards (1857), Klunzinger (1877), Studer (1887), and l)a.niel.sseu (1887). The numerous new species described during this period were distinguished by their authors, on the one hand from Ammothea by tlie presence of a,rmed verruca), and on the other from Hpongodes Less, by the comparatively slight development of the spicules which formed the armament of 512 MR. E. W. SHANN ON the verrucse. Wright and Studer (1889) described a specimen under the name Spongodes nejjhthyceformis, concerning which they observe : — " The entire habit of the colony recalls much more that of Nephihya than that of Spongodes, and this im- pression is strengthened by the slight development of the spicules surmounting the little heads, whence the colony does not appear so prickly as other species .... The species must be referred to the genus Spongodes, because the polyps are placed sideways within a bundle of spicules, although these only project slightly." That is to say, Wright and Studer recognised on the anthocodife of Sp. nephthyceformis the presence of what is now known as a " Stiitzbiindel," and here we have the starting-point of the difficulty of discriminating between the genera Spongodes and Neplithya. Holm (1895) faced the problem of reconstructing the genus Spongodes in the light of the knowledge which had accumulated since Lesson first described the genus in 1834. This author pointed out that Spongodes nephihywformis W. h St. is identical with Neplithya chctbrolii Audouin, and added that iV. chabrolii differs from Spongodes in many characters, such as the branching of the colony and the arrangement of the polyps : on these characters one can establish two geneia, but it is necessary then to add to the genus NepMhya many species hitherto included in the genus Spongodes, including the type Sp. celosia. Though Holm shrank from submitting a well-known type like Sp>ongodes celosia to such treatment, he proceeded fearlessly to include all the species of Nephthya, including the type N. chabrolii, within the genus Spongodes ; he retained Nephthya, however, as a subgeneric title. Here was a step in the direction of elucidation ; Nephthya and Sp>ongodes,?i& hitherto defined, were shown to be synonymous; but Holm's solution of the problem threw too great a burden on Spongodes. Kiikenthal (1895), writing during the year in which Holm's paper was published, accepted the genus Spongodes in its new distended form ; but the term Spongodes had become so obviously cumber- some that this author (Kiikenthal), in a later paper *, reinstated Nephthya with full generic honours. In this paper Kiikenthal gave a summary of the family ISTephthyidse, and divided the various genera into two groups as they possessed or lacked a " Stiitz- biindel " ; he summed up his remarks as follows : — " Innerhalb der Familie der Nephthyiden ist als wichtigstes Merkmal zu betrachten, ob die Polypenkopfchen terminal auf ihrem unteren Telle, dem Stiele, sitzen oder seitlich davon. Letzterer Fall tritt stets dann ein, wenn sich auf einer Seite, der oberen, ein Biindel Spicula besonders stark entwickelt : das Stiitzbiindel." The genera possessing a " Stiitzbiindel " were distinguished from one another by the disposition of the polyps on the colony, as Holm (1895) had already suggested; the xi&xq.b Nephthya was applied to forms resembling the original type N. chabrolii Audouin, in which the polyps are collected on branchlets, the latter being arranged in catkins or lappets, and the name Spongodes was * Abb. Senckenb. Ges. Frankfurt, vol. xxiii. p. 88 (1896). ALCYONARIA PROM SINGAPORE. 513 retained for forms in which the polyps are disposed sporadically or in bundles. The net i-esult of this reformation was the inclusion in the genus Nephthya of all foi-ms hithei'to included in the " Spicatfe " group of the genus Spongodes. In adopting this means of classification it became necessary to include Sp. celosia, Lesson's type of the genus /Spongodes, in the emended genus Nephthya ; nevertheless, the name Spongodes Avas retained by Kiikenthal for his other emended genus, since the latter included a large number of forms which during many years had been described under Spongodes. The definition of the genus Nephthya given in the paper under consideiation runs thus : — " Nephthyiden mit ' Stiitzblindel.' Die Kolonie ist buschig veriistelt, die meist Kurzen unci nur vereinzelt sterilen Stammteile sind durch einen abgeflachten, oft membranosen Basalteil verbunden. Die Polypen stehen in grosser Zahl und zeimlich gleichmjissig verteilt auf den Steinzweigen, die dadurch die Form von jihrenformigen Lappen oder 'Katzchen' erhalten. Hervorragende Spicula der Polypenkopf chen fehlen." May (1899) accepted the foregoing definition of the genus Nephthya, and in his ' Revision of the ISTephthyidas ' (1903) Kiikenthal has not had occasion to modify it. A '' Stutzbi'mdel." The crucial point in Kiikentlml's classification of the Neph- thyidfe lies in the definition of the term " Stiitzbiindel." Much of the opposition to the above classification has arisen through the different interpretations which various authors have attached to the term. Kiikenthal, to whom we owe the word, uses it in an extremely comprehensive sense, the range of which can be circumscribed, however, by the following limits : — A " Stiitzbiindel " is an aggregation of spicules disposed along the abaxial aspect of an anthocodia and lying approximately parallel to its axis. The spicules are usually spindle-shaped ; they are not infrequently lai'ger than those from any other portion of a given specimen, and one or two of them commonlj^^ but not invariably, reach from the apex of the polyp-stalk into the substance of the colony. A few of the spicules in a characteristic example protrude beyond the polyp-head, but such a condition is not essential. A definite "Stiitzbiindel" may not be recognisable in every polyp of a given colony ; but if such is present it will appear most obvious in the younger polyps near the distal ends of the branches. A specimen in which a " Stiitzbiindel " is demonstrable, whether or not in eve^y polyp, must be classified as possessing a " Stiitzbiindel." In the genus Lithojohytitm, which closely resembles Nephthya in external appearance, the a,nthocodise being massed on small terminal lobes or lappets, there is no " Stiitzbiindel." The anthocodije contain very few spicules, some of which are loosely arranged en chevron along the abaxial surface. Both the small size and oblique position of the spicules so arranged prevent their being described as foi'ming a " Stiitzbiindel." 514 MR. E. W. SHANN ON NEPnTnYA BEDFORDi, sp. 11. (PI. LXII. fig. 8 ; PL LXIII. figs. 11, 12.) Two specimens which, while conforming with the characters described above for the genus Nephthya, fail to ngree in detail with any of the large number of forms hitherto described, have necessitated the creation of a new species of this gemxs. The colony (fig. 8) is bilaterally compressed, the growth is bushy, and the mnjor diameter of the capitulum is approximately equal to the total height, including the stem. The consistency is tough and leathery. The stem, which shows signs of bilateral compression, is short, and gives rise at its distal extremity to a variable number of main branches. These main branches are again divided into uneqiial secondary branches. From both main and secondary branches spring the short terminal lobes. The latter are conical in shape, but rounded ; on them the anthocodijs are tolerably evenly distributed. As the polyps are situated very close together, and the terminal lobes are ex- ceedingly numerous, the whole capitulum appears to be covered with anthocodife. The polyp-heads when at rest make an acute angle with their stalks. The latter scarcely protrude from the colony, with the result that the polyp-heads are very closely apposed to the surface from which they arise. Colour in alcohol cinder-grey, polyps brown. Locality : below low-tide mark, Blakang Mati. Detailed measvirements : — Hegion measured. Specimen T. Specimen II. Total height 60 mm. 42 mm. Height of stem 17 mm. 14 mm. Diameter of stem 18 x 10 mm. 12 X 10 mm. Height of capitulum 43 mm. 28 mm. Breadth of capitulum 52 X 20 mm. 45 x 20 mm. In both specimens the length of the terminal lobes is about 4 mm., but the variation in this respect is between 3 and 6 mm. The diameter of the terminal lobes is from 3 to 4 mm. The edges of adjacent polyps are seldom more than "175 mm. apart. The polyp- heads have an average length of I'l mm. and diameter of '6 mm. The spicules show a wide range in size and shape. The smallest forms are found in the tentacles, where they are arranged in no veiy definite order, but for the most part lie at an acute angle with the axis of the tentacle. The tentacle- spicules are minute spiny spindles, some of which are straight, others crescentic. The polyp-spicules (fig. 11,/) fii'e very brittle, longitudinally striated spiiidles. They are straight or curved, and usually smooth, but sometimes bear a few minute spines. It is diificult in most cases to discern an arrangement of the polyp-spicules en chevron, but some of the polyps show such a condition more clearlv than others. ALCYONARIA FROM SINGAPORE. 515 The " Sfciitzbiindel," though unrlonbtedly present, is very ill- defined. From four to six spindles can usually be observed along the abaxial surface of each anthocodia, but in no instance do these protrude beyond the polyp-head. It is exceedingly difficult to dissect out a single polyp with its supporting spicules intact; some preparations were obtained, however, after soaking portions of the branches for twelve hours or more in oil of cloves. A more satisfactory method of observing the " Stiitz- blindel" is obtained by placing one of the terminal lobes, previously cleared in oil of cloves, under a low-power binocular microscope. By this means a stereoscopic view of the polyp and its supporting spicules is obtained, such as is represented in fig. 12. The cortex of the capitulum contains numerous hori- zontally disposed spindles, some of which are among the largest spicules found in this species ; they resemble in type those described from the polyps. In the outer wall of the stem are found small spicules with broad rays (fig. 11, a) which are usually numerous, but vary considerably in size and concentration. These forms are covered with warts -015- -036 mm. long, and interlock with one another, thus accounting for the tough consistency of the stem. Among the forms described above there are also in the outer wall of the stem some larger spindles with remarkable spines (fig. 11, h); the latter, which have an average height of -044 mm. and basal breadth of -02 mm., are often larger and sharper on one side of the spindle than on the opposite side, and, since the spines all lie approximately in one plane, give the spicule a comb-like appearance. The spicules of the _ canal-walls (fig. 11, c, d) are not very plentiful, but are distinguished by their stout appearance. They may be described as very thick, longitudinally-striated spindles, somewhat flattened and bearing low rounded warts. Their ends are either rounded or bluntly pointed, the body of the spicule is straight or slightly curved. Among these regular forms there are found a few branched spicules. The latter resemble the regular forms in structure, but are either triradiate or show short irregular processes springing from the central region of a typical spindle. The following are characteristic measurements of the spicules, length by breadth, in millimetres : — (a) Polyp (outer) -3 X '025 (b) Polyp (inner) -17 X "02 -112 X -023 •07 X -02 (e) "Stutzbuiidel" -75 X -05 "6 X "Oo •4 X -06 (c?) Cortex of capitulum TO X -075 "9 X "08 (e) Cortex of stem, radiate forms -115 X 'ISl '058 X -054 Thickness of branches "OS-'Olo Do. spindles (including spines) ... So X '125 '22 X "125 (/) Canal-walls, spindles Tl X -19, -85, '2, -ex-ll, •4 X -08 Do. triradiate forms -41 x '4, thickness of ray 'l (measured by taking two ter- minal points as a base line and the third point as the extremity). "22 X -18 thicknessof ray -05 516 MR. E. W. SHANN ON On applying the key to the species of Nej)liihyaj given by Kiikenthal (1903, p. 145) to the case of N. hedfordi, it is seen that this species falls within the group chai'acteiised by forms in which the terminal lobes are conical but rounded. The inner polyp-spicules are smaller than the outer ; moreover, they have the form of smooth rods. These characters taken together indicate that N. hedfordi resembles closely N. pacifica and N. alhida ; and, since the spicules do not form a ring around the base of the polyp, it may be infei'red that the new species is most nearly allied to N. albida Holm, so far as the key to the species can be relied upon. Excepting in the character of the inner polyp-spicules, however, iV. hedfordi appears to be more nearly related to K. elongata Kiikenthal than to iV. alhida. This supposition is supported by the fact that N. elongata is reported from Ternate, while N. alhida is a Red Sea form. The points of difierence and of resemblance between N. hedfordi and its two nearest allies may be seen at a glance in the appended table : — Character. N. elongata. N. hedfordi. iV. alhida. Terminallobes 1 .^j°V^^ -• 5 3 6-3 4-3 9 7 rolyp-heads{5^°^|^ ";;;; •8 •9--6 11 •6 1 •7 Angle between Polyp and Stalk 45° •27X — •13X— (Spiny spindles.) Acute. •3X-025 •112X-023 (Smooth rods.) Eight. •3X^03 •4X-015 (Smooth rods.) Po,„..n.ic«L,{£S :;; " Stiitzbiindel " Projects slightly. 1-5 Does not project. •6 •05 Projects slightly. 1-2 •12 Spicules of " Stiitzbiindel " Long Thick ... Spicules of Upper Cortex : Long Thick ... •8 1 •075 1^4 •12 Spicules of Lower Cortex : Long Thick ... •8 (Compact.) •35-^22 •125 (Comb-like and radiate forms.) •85 •22 Spicules of Canal-walls : Long Thick ... 1-2 (Scattered.) l^l--4 •2-^08 (Spindle and tri- radiate forms.) 1 •2 Colour (in alcohol) Greyish yellow. Cinder-grey. Greyish white. Locality Ternate, 5 fms. Singapore. Ived Sea. All measurements are in millimetres. Lest the use of an artificial key to the identification of species should have led me to overlook a member of the genus not included in the group to which the Singapore specimen apparently ALCYONARIA FROM SINGAPORE. 517 belongs, but at the same time exhibiting a close resemblance to it in structural detail, I have read descriptions of all the species of Nephthya which might possibly have proved to be identical with N. hedfordi. In no instance is there a greater resemblance between N. heclfordi and another species of Nei^hthya than exists between any two established species of that genus. Indeed, the exceedingly feeble development of the " Stiitzbiindel " and the presence in the stem-cortex of exceptionally small, though curiously comb-shaped, spicules are characters which serve cleai'ly to distinguish N. hedfordi from any species previously recorded. Among some unpublished notes by Miss Coward, which Professor Hickson has kindly placed at my disposal, the following paragraph occurs : — ■ "In his work on the family Nephthyidfe (1903) Kiikenthal names a specimen Nephthya chabrolii. In doing this he refers to Hickson and Hiles' (1900) description of N. chabrolii. These writers, however, state that the spicules of their specimen are just as described by Klunzinger (1877) — that is to say, they do not form a ' Stiitzbiindel.' In the description of his specimen Kiikenthal says the spicules only rarely project beyond the polyp-heads — and yet his diagnosis is the presence of a ' Stiitzbiindel.' " This observation came to hand after the account of the new species JV. hedfordi had been written and the " Stiitzbiindel " described. As stated above, the " Stiitzbiindel" was very poorly developed. I then thought it advisable to examine JSf. chabrolii to satisfy myself as to the nature of the " Stiitzbiindel " in that species. The specimen at my disposal was the identical one described by Hickson and Hiles (1900), in Willey's collection. Preparations, which have been cleared in oil of cloves and examined under the binocular microscope, reveal the presence of distinct bundles of spindles supporting the polyps. One or two of these spicules not infrequently project beyond the polyp- heads. One must admit that N'. chabrolii is characterised by the presence of a small but clearly defined " Stiitzbiindel." It seemed possible that the " Stiitzbiindel " was so degenerate in N. hedfordi that the polyps might be described as being without this characteristic. Were this the case, the specimen in question would come under KukenthaUs definition of the genus Litho- phytttm Forsk. I accordingly made and examined preparations of Lithophytum arboreum ; but in this case there was no ti'ace of a bundle of spicules on the abaxial surface of the anthocodia? in the least degree comparable with the condition which I have described in N. hedfordi. The genus Nephthya^ as at the present time accepted, contains a large series of species showing every givadation in the develop- ment of a " Stiitzbiindel." At one end of the series is found such a form as N. celosia Lesson, in which one or two of the " Stiitzbiindel " spicules reach 2 mm. beyond the polyp-heads ; in the middle iY. chabrolii, in which they only project slightly ; and at the other end A', hedfordi, in which they do not project. 518 MR. E, W. SHANN ON Genus Dendkonephthya Kiikenthal. 1791-97. Alcyonium Esper, Pflanzenthiere, pp. 49, 50, tab. 16. 1834. Nee Sj^ongodes Lesson, Illustrations de Zoologie, vol. ii. part 2, p. 89. 1834. Nephthya (pars) Elirenberg. 1846. Spoggodia (pars) Dana, p. 625. 1857. SjMggodes (pars) Milne-Edwards, p. 127. 1862. Spoggodes (pars), Morchellcma Gray, in Proc. Zool. Soe. London, p. 27. 1877. Spongodes (pars) Klunzinger, pp. 34, 35. 1889. Spongodes (pars) Wright & Studer. 1895. Spongodes (pars) Holm, p. 16. 1896. Spongodes (pars) Kiikenthal, p. 97. 1899. Spongodes (pars) May. 1905. Dendronephthya Kiikenthal, Zool. Jahrb. xxi. p. 526. The genus Spongodes was founded by Lesson (1834) on the type Sp. celosia, and for many years workers in the field of the Nephthyidpe found no difficulty in discriminating between SjJongodes Lesson and Neplitliya Savigny. I have endeavoured to show in my section on the genus Neplitliya that as the number of species of the above-mentioned genera increased it became more and more difficult to draw a hard-and-fast line between them ; further, that this fact became manifest when, in 1889, "Wright and Studer described a specimen as Spongodes nepliihycB- formis, which was subsequently relegated to the genus Neplitliya^ and, indeed, shown to be identical with N, chahrolii, the type of that genus. A new distinctive feature was required : such a feature was discovered by Holm and applied in a practical form by Kiikenthal. The feature in question is the arrangement of the polyps upon the stem, and its application has already been mentioned above. Kiikenthal 's (1896) definition of the emended genus Spongodes is as follows : — '* Polypenstock baumartig veras- telt, unterer Stammteil nackt. Die Polypen sind in Biindeln vereinigt oder stehen vereinzelt." Under the genus Sj^ongodes thus defined, Kiikenthal included as subgenera Spongodia and Spongodes ; he further subdivided the latter into three grouj^s, namely, Glomeratee, Umbellatse, and Divaricatse. May (1899) and others adopted this classification ; and so the matter stood until the year 1905, when Kiikenthal published the second part of his " Revision of the Neph thy idee." In this paper Avas introduced the division of the time-honoured genus Sj^ongodes Less, into the two new geiiersi, Dendronephthya and Stereonephthya, which has provoked such a stonn of criticism. It must be remembered, however, that the group Spooigodes " Spicatse," and with it the original type /Sjo. celosia, had already been relegated to the genus Nej^lithya ; so that the genus Spongodes, as then accepted, no longer retained its ancient prestige ; moreover, since the genera Neplitliya and Spongodes were shown to be synony- mous, Kiikenthal Avas justified by the International Rules of Zoological Komenclatiu-e (Art. 28) in retaining the older name — ■ ALCYONARIA FROM SINGAPORE. 519 Nephihya — to designate the merged genei-a. The genus Dendro- nephthya is ahiiost synonymous with Spongodes (Glomeratte, Umbellatse, and Divaricatse), as defined by Kiikenthal in 1896, and Stereonephthya has a like relation with his Sp)ongodes (Spongodia). There is no doubt that here, again, Kiikenthal was acting in accordance with the International Rules of Nomenclature. Had these Rules been extant at the time Kiikenthal published his classification of the Nephthyidfe in 1896, the name Spongodes would doubtless have been discarded by him then, for in the second section of Art. 30 it is written: "In no case, however, can the name of the original genus be transferred to a group containing none of the species originally included in the genus." Thus Spongodes, as used by Kiikenthal in 1896, would not be considered as a valid generic name by those who accept the International Rules of Zoological Nomenclature, which appeared in 1905. One cannot but feel the loss of a name which for nearly a century has been familiar to students of the Alcyonaria, and regret the inconvenience caused in Museums and Zoological Laboratories throughout the Avorld by its suppression. Nomen- clature is a matter of convenience and cannot be regulated absolutely by arbitrary laws. Generally speaking, of two synonyms, that which is most familiar to the majority is to be preferred ; a more practical reason should be required than a mere regulation before supplanting an old familiar generic name by a new one. If we are required to obey literally the Inter- national Rules of Zoological Nomenclature, such everyday names as Astacus, Holothuria, Actinia, Modrepora, and many others will be forfeited, as has been Sp)ongodes. While regretting that Kiikenthal did not exercise his authority in retaining the name Spongodes, in deference to his extensive knowledge of the Neph- thyidse and his able reorganisation of that family, I have adopted the term Dendronephthya for a genus of Nephthyidee defined as follows by Kukenthal (1905) :— " Nephthyiden von baumformig verzvveigtem Aufbau, deren Polypen stets in Biindeln vereinigt sind, Polypen mit ' Stiitz- biindeln.' " Dendronephthta disciformis Kiikenthal 1905. (PI. LXIII fig. 13.) Three specimens in the collection agree so nearly in all essential features with the description given by Kiikenthal (1905) for B. disciformis that there is little doubt that they may be assigned to that species. Kiikenthal's specimen came from the China Sea. The largest of the three colonies is 8 cm. high, 7 cm. broad, and 3"5 cm. thick. It was taken ofi" Pulo Brani in 5 fms. of water. Two other specimens were taken in 7 fms. of water between Pulo Hautu and Blakang Mati. Their measurements, height by breadth by thickness, are 4x4x2-5 cm., and 3-5 x 2-5 x 1-5 cm. respectively. 520 MR. E. W. SnANN ON It may be noted that the two last-mentioned specimens are much smaller than any recoi'ded by Kiikenthal. The colonr of the upper branches and " Stiitzbiindel " is pink, whereas the typical colour for this region is deep red. Moreover, no radial spicules were observed in the lower portion of the stem. The exact sj'stematic position of a fourth specimen is rather doubtful ; but, until furbher evidence is forthcoming, it has been deemed advisable to describe the specimen under the specific title D. disciformis. This specimen shows externally the typical " disciformis " features. It is 9*5 cm. in height and 8'5 x3"0 cm. in breadth, thus exceeding somewhat in dimensions the largest specimen described above. The characters and distribution of the spicules are identical with those of typical examples, ordy excepting those from the lower portion of the stem. The spicules found in the last-mentioned area are spiny clubs (fig. 13), among which are scattered a few of the typical radiate forms. The clubs vary in length from •70--27 mm., but the majority are -45 mm. ; the thickness is almost the same even in forms of different length, the average width of the handle of the club is "05 mm., that of the head "12 mm., excluding the spines, which sometimes reach -044 mm. in height. The colour of the branches and " Stiitzbiindel " is a dull yellowish brown, and there are no red spicules such as are found in these areas in all the other recorded specimens. Kiikenthal (1905) described a colony whose branches were pale brown and whose polyps were white, but even in this specimen red spicules were observed. Doubtless one might be led to base a new variety of Dendro- nephthya on the characters of the colony described above, a colony possessing a very characteristic form of spicule which has not been observed, either from the same region or from any other region, in any previously recorded specimen of D. disci- formis. Such a course might have been adopted had not this aberrant form been taken at the same time and from the same place as the two smaller of the three colonies described above, namely, between Pulo Hautu and Blakang Mati, 8/xii/98, in 7 fms. It seems probable, then, that the species D. disciformis is liable to considerable variation in the colour of the terminal branches and " Stiitzbiindel," and in the character of the spicules wdiich are found in the lower region of the stem. The latter phenomenon is very remai'kable, and, it is hoped, may act as a check to the frequency with which new species are created to separate two or more specimens of Alcyonaria which only differ from one another in the shape of their spicules. Genus Stereonephthya Kiikenthal. 1869. SjJoggodia (pars) Gray, in Ann. & Mag. Nat. Hist. (4) vol. iii. p. 128. 1877. Spoggodia (pars) Klunzinger. 1889. " Divaricatse " {Spongodes) Wright & Studer. ALCYONARIA FROM SINGAPORE. 521 1895. Spongodia (pars) Holm, p. 25. 1896, Spongodia (pars) Kiikenthal. 1904. Spongodia (pars) Holm, Weiteres liber Xephthya und Spongodes. 1905. Stereonephthja Kiikenthal, p. 694. " Spongodes "-like forms in which the polyps were scattered irregularly over the stem were recognised at an early date, and were classified sometimes as a distinct genus, at others as a sub- genus of Spongodes, under the name Spoggodia or Spongodia. Wright and Studer (1889) recognised this group, but included its members in the genus Spongodes under the designation " Divaricatje." Later writers, such as Holm (1895) and Kiiken- thal (1896), agreed that Spongodes and Spongodia were synony- mous. Hence Spongodes dragged Spongodia with it in its downfall in 1905. For similar reasons to those stated in the .section on Dendronephthija I have adopted the new genus Stereo- nephthya which Kiikenthal (1905) defined thus: " Sehr starre Ne^jhthyiden, deren Polypen weder in Liippchen noch in Biindeln angeordnet sind, sondern einzeln oder in kleinen Gruppen direkt vom Stamm wie den nicht oder wenig verzweigten Hauptasten entsjaringen. Polypen mit ' Stiitzbiindeln.' " Stereonephthya lutea, sp. n. (PI. LXIII, fig, 14.) The specimen, on the obsei^vation of which this new species is based, agrees in every respect with the diagnosis given by Kiikenthal (1905) for the genus Stereonephtjiya, At the same time it possesses characters which serve clearly to distinguish it from other species hitherto described. The colony is upright and branched. ISTumerous tapering branches leave the insignificant stalk either singly or in groups, and spread in all directions, but with a distinct tendency to assume an upright position. When they arise in groups, the members forming a group are frequently united for a considerable distance by a common cortex. The primaries sometimes, but not invariably, bear a few stationary branches, which also show a tendency to grow in a vertical direction. From both the primary and secondary branches short lateral lobes are given off at irregular intervals. Measurements : — Height of colony 7"5 cm. Breadth at base 1-5 ,, Maximum breadth of capitulum 5 "5 ,, Length of primary branches 3"5-5'0 ,, Breadth of primary branches ■{ , . ■'.' ;; ? " ^ •' [ at extremity 0"5 ,, Length by breadth of secondary branches 1*5 x 0-5 Length by breadth of lateral lobes 0-5 x 0-5 Proc. Zool. Soc— 1912, No. XXXV. 35 522 MR. E. W. SHANN OJ^ The polyps arise sporadically from all parts of the colony. They ai'e closely aggregated on the lateral lobes, and are most scarce on the basal portion of the primary lobes. The anthocodise are supported by a " Stiitzbiindel '' (fig. 14), three horny spindles of which project •2--4 mm. beyond the polyp-head. The polyp- heads make an acute angle with their stalks ; the stalk is 1*2 mm. high, the head is 1 mm. in length and 'T-'SS mm. in maximiim breadth. The polyps contain thorny spindles of various sizes and a few clubs. These are sometimes placed horizontally, but are more usually oblique, and in the distal area are ari'anged en chevron. These spindles measure -45 x "035, -^T X '04, -2 x •025 mm. Among the above and extending into the tentacles are numerous little spicules which are nearly rectangular in outline; they measure -110 x '036, '095 x '015, -050 x "015 mm. The " Stiitzibundel " is composed of a number of thorny spindles of a bright yellow oolovir (they resemble at first sight crystals of sulphur). These spicules are ten in number round the base of the polyp-stalk, but there are only three at the distal end. The four spicules which underlie the point of origin of the polyp- head are semi-lunar in shape, the concave sides facing the polyp-head. The " Stiitzbiindel" spicules reach 1*8 mm, in length and •12 mm. in breadth. The cortex of the colony derives its characteristic colour from the enormous number of slim yellow spicules which it harbours. There are also thorny spindles ; they lie pai-allel to and very near the sui>face, and vary very much in size and shape. Some are twice, others only half, the size of the " Stiitzbiindel" spicules. Few ai^e straight, most are semi-lunar or S-shaped, and veiy exceptionally a triradiate form occurs. The ends of the spindles are usually bluntly pointed, but occasionally forms were observed with their ends frayed out into a number of points. The largest measure 3'0 x "28, 2*7 X '30, 2'1 X -15, the smallest -67 x '07, -67 x '04, -55 x '05 mm., but every conceivable intermediate stage can be found, A remai-kable feature of this species is the entire absence of spicules from the canal-walls. The colour of the colony in alcohol is pale yellow, the polyps are cream -coloui-ed. The specimen was taken in Imbiab Bay, where it was left exposed at the lowest tides. Two other specimens taken in the littoral region at Tehik Ayer have also been assigned to this species. At first sight they do not resemble closely the type specimen, for they are shorter, broader in proportion, and more bushy in appearance than the colony described above. One of these colonies is 4*3 cm. in height, its base is "8 cm., and the maximum diameter of the capitulum is 3"1 cm. ; the measurements of the other, taken in the above order, are 3 x '7 cm. : 4*5 cm. The apparent difference in the branching from that of the type is due to the fact that the secondary branches are no larger than the lateral lobes, and that both these kinds of offshoot are relatively numerous. The polyps are arranged much more densely than in the type, ALCYOXARIA PROM STKGAPORE. 523 especially at the tips of the branches. The colour of these specimens, moreover, is of a duller yellow. These slight differences appear quite insigniiScant when one considers that these two specimens resemble the type of Stereonephthya lutea in the following important characters : — The size and shape of the anthocodise. The size, distribution, and remarkable sulphur-yellow colour of the spicules ; the dis- tribution of spicules agrees even to the disposition of those forming the " Stutzbiindel " ; the large internal canals, with thin walls devoid of spicules, Genus Parasppngode^ Kiikenthal. P. crassa Kiikenthal, 1896. This species was first described by Kiikenthp^ in 1896, and further reference to it i§ made by the same author in his 'Revision der Alcyonarien ' (1907), where it is described under " Species incerti generis." Two smajil colonies of a pale brown colour are placed without hesitation in this species. One colony was taken in 5-9 fms, of water fpom ]bhe New Harbour. It measures 20 mm. in hejglit ; the djamete).' of the capitulum is 13 mm., thafc of the stem 7*5 mm. The other colony was taken in 5 fms, of wa|:^r, ^he precise locality is not recorded. Jt measures 20 mm. in height; the diameter of the capitulum is 11 mm., and the stem is broken. The only important character in whfch these specimens differ from the type is the size of the spicules in the stem^cortex. Measurements, length by breadth in millimetres, for thi?ee of the largest spicules from this region 9,re J'2 x '\^, 1-0 x -15, -95 X '22. Kiikenthal found spicules measupng as much as l-*8 x "24 mm. in the stem-cortex. The spicules from the Singapore specimens show, however, the typical warts (-'OS mjii. high). Previously recorded from Ternate, at a depth of 30 fms, Family Siphonqgorgiid^, Genus SlPHPNOGORQIA. S. VARIABILIS Hickson (olim Chironephthya vm'iahilis Hickson). Two beautiful little specimens taken in JO fms, of water south of Blakang Mati are referred to this very variable species, Each possesses a very slender stem, from the apex of which two main branches are given off, so that the appearance of the colony is roughly Y-shaped. These primary branches give rise to secondary branches, which in one or two cases are again divided. The polyps occur most frequently on the tips of the brai^ches, but a few are scattered on all portiojjs of the colony except the steip. Measurements : — Specimen A. S2:>ecimen B. mm, njin. Height of stem 18 11 Diameter of stem 6 3 Length of primary branches 23 and 23 24 and 18 35* 524 MR. E. W. SHANN ON The spicules are of the usual shape and of medium size, and assume the colour of the part of the colony in which they are embedded. The large warty spindles of the ccenenchym are tinted with mauve and do not exceed 3-0 x "3 mm. In the polyp- area the spindles are more frequent than the clubs ; all the polyp- spicules are more or less bent, they are either bright red or bright yellow in colour, and the larger ones measure '6 X "05 mm., •47 X "065 mm. (the red spicules mainly compose the " crown," the yellow ones the " points "). The colour of the colony in alcohol is a delicate pinkish mauve. The polyps are deep red, and the tentacles bright yellow. It is difficult to conceive how the colouring could have been more vivid or the blended tints more pleasing even before these little colonies Avere removed from their natural habitat. For a list of the colour-variations to which this species is liable to I'un, see Hickson's " Alcyonaria of the Maldives," Part I. p. 488 (under Chirovephthi/a ixcriabilis), and Thomson and Simpson (1909, p. 125). Previous!}^ I'ecorded from Mahlos Atoll, S. Nilandu, Persian Gulf, Andamans, and the Arakan Coast. Order GORGONACEA. Suborder PSBUDAXOSIA. Family Sc lerogorgiid.'E. Genus SuBEROCxORGIA. S. suBEROSA Pallas. Gorgonia suherosa Pallas, Elench. Zooph. p. 191. iSaherogorgia suherosa Gray, Proc. Zool. Soc. 1857, p. 159. iScIerocto7'gia suherosa Kolliker, Icon. Histiol. p. 142, pi. xix. fig. "13 (2). Sclerogorgia suherosa Studer, Monatsber. K. Akad. AViss. Berlin, 1878, p. 666. Siiherogorgla suherosa Ridley, Journal Linn, Soc. Zoology, vol. xxi. p. 243. Sxiberogorgia suherosa Wright it Studer, ' Challenger ' Re- ports, vol. xxxi. p. 166. Suherogorgki stiherosa Brundin, " Alcyonarien aus dem Samm- lung der Zool. Mus. in Upsala," Svenska Vet. -Akad. Handl. xxii. pt. iv. There is in the collection a bottle which contains a number of fragments, appaiently the component parts of a single colony of y^. suherosa. The basal portion is present and has a diameter of 6 mm. ; the diameter of the terminal branches is 2*5 mm. The specimen was taken in about 14 fms. of water oft" Pasir Panjang. ALCYONARIA FROM SINGAPORE. 525 Previously recorded from the coasts of Western Africa and the West Indies (Pallas and Esper) ; the Mermaid Straits, Dampier Archipelago, and IST.W. Coast of Australia (Studer) ; Port Denison, Queensland, and Torres Straits (' Alert ' Coll.) ; Mauritius (Coll. Brit. Mus.) ; Admiralty Islands (' Challenger ' Coll.) ; and Sumatra (Brundin). Famil}^ Melitodid^. Genus Melitodes. M. ALBiTiNCTA Ridley. This species is represented by numerous fragments, but none of these represents an entire colony. The specimens were taken in 15 fms. of water from Blakang Mati. Previously recorded from Port Molle, Queensland. Genus PSILACABARIA. P. GRACiLLiMA Ridley. Psilamharia gracillima, n. gen. et sp., Ridley, Rep. Zool Coll H.M.S. ' Alert,' Alcyonaria, p. 363. After a lapse of nearly three decades this delicate little species is to be recorded once more. It is represented in the collection by four portions of colonies, which show the mode of branching and length of the internodes, and by sundry fragments. None of the colonies are complete, but the largest intact portion measures 45 mm. in height. The internodes vary in length, the majority, however, fall within the range given by Ridley, viz. 12-16 mm. The branches are much more delicate than those of the type, measuring only 1-2 mm. in diameter, as compared with 3-7 mm. In the mode of branching, size of polyps, and details of spicule-distribution the Singapore specimens resemble the type very closely. The spicules show the typical shapes; but they are in each instance somewhat smaller than their prototypes. The colour of the coenosarc and polyps is fawn, that of the axis white. Locality. Salat Sinki, in 4-5 fms. Previously recorded from Port Molle, Queensland, 12-20 fms. ; Port Darwin, 8-12 fms. ; E. Australia, 42 fms. (Ridley). Genus Wrightella. W. ROBUSTA, sp. n. (PI. LXII. fig. 9 ; PI. LXIII. fig. 15.) This species has been formed to include a single well-developed colony, which agrees more closely with Wrightella Gray (1870) than with any of the other genera of the family Melifcodidaj. The form of_ the colony differs, in its stouter dimensions and in its erect position, from the members of the four existing species of Wrightella ; but it was thought expedient to associate it with 526 MR. E. W. SHANN OK these, rather than to create a new genus from the observation of a single specimen. Wright and Stnder (1889, p. xxxvi) give the following de- finition of the genus Wrightella : — " The branches and twigs are compressed ; the projecting polyp calyces occur especially on the sides. In the cortex there are foliaceous clubs. There are no nutritive canals in the axis." The specimen fulfils all these conditions. The single main stem arises abruptly from a strong reticulate base (fig. 9). The base, which is broken at the edges, measures 15 mm. in diameter, and the gaps in its meshes average 1 mm. in diameter. The colony is 120 mm. in height and 45 mm. in breadth (some of the lateral branches are broken, so it is probable that the true breadth exceeded 45 mm.). The branching of the colony is dichotomous, and takes place at the swollen nodes. The nodes are less prominent in the distal branches. All the branchiug takes place in one plane; the terminal twigs ai-e markedly flattened in the plane of branching. Anastomosis of the upper branches takes place at infrequent intervals. The nodes near the base are globular, having a diameter of 4 mm. ; the internodes in this region are circular in section, their diameter is 2-5 mm., which is slightly exceeded by their length. The internodes beyond the lowest three become more elongated and show an average length of 10 mm. The terminal twigs are only 1 mm. wide. The varrucse, which measure '75 x "75 mm., are not densely crowded, and show a tendency to arrange themselves on the lateral aspects even of many of the Idwer branches, but more especially in the terminal twigs. Both cortex and vertucse are yellow in colour. The axis is Avhite, and is not traversed by nutrient canals. The precise locality is hot recorded) but, like the other speci- mens in this collection, it was taken in shallow water near Singapore. The spicules attaiil all manner of shapes (fig. 15); they are quite colourlessj The foliaceouS clubs characteristic of the genus Wrightella are present in large tiumbers, and show the following range in measurement : length by breadth -27 x '12 mm., •15 X 'OG mm. ; handlfes of clubs '02 mm. in diameter. Numerous spindles occur, some are foliaceotis and me&,sure '27 X "07 mm., •20 X "08 mm., -12 ?< -06 mm.; others are spiny, the spines fre- quently being confined to the central region, and measure •22X-05 mm*, '17 k -03 mm;, •12x-03 mm. A few stellate forms are found which have a diameter of *l0-'15 mm. Minute scales abound ; they tneastire '05 X "025, "036 X "OoO mm., •028 X -014 mln. lyiteftttti:r'e: 1870. GftAY, J. El— dat. Lithophytes Brit. Mus. p. 31. 1900: HlCKSON, S. Ji, atid Hiles, Isa L. — "'The Stolonifera and Alcyonacea collected by Dr. Willey in New Britain, etc." Willey's Zed. Reslilts^ pt; iv. pp^ 493-508. ALf'YONARIA FROM SINGAPORE. 527 1877. Klunzinger, C. B.— Die Korallthiere des Rothen Meeres. Part I. p. 33, tab. 2. fig. 5. 1903. KtJKENTHAL, VV.— "Veisuch einer Revision der Alcyo- navien." IT. Die Familie der Neplithyiden : 1 Theil. Zool. Jahrb. Syst. xix. 1905. Do. do. 2. Theil. Op. cit. xxi. 1907. Do. do. 3. Theil. Op. cit. xxiv. 1903. Pratt, Edith M.— •' The Alcyonaria of the Maldives. Part II. Sarcoiihyium, Lobophytimi, Sclerojohytum, and Alcyonium." Fauna and Geog. of the Maldive and Laccadive Ai-chipelagoes. (J. Stanley Gardiner), VoL ii. Part I. 1906. Thomson, J. Arthur, and Henderson, W. D. — " Alcyo- naria of Zanzibar and British East Africa." Proc. Zool. Soc. pp. 393-443. 1909. Thomson, J. Arthur, and Simpson, J. J. — " An Account of the Alcyonarians collected by the Royal Indian Marine Survey Ship 'Investigator* in the Indian Ocean." Part II. 1889. Wright, E. P., and Studer, Th.— " Report on the Scientific Results of the Voyage of H.M.S. ' Ohallenger.' " Vol. xxxi. Alcyonaria^ EXPLANATION OF THE PLATES. Plate LXI. *Pig. 1. AmmotJiea vh-escens Sav. {^Lithopligtum arhot'eum PorsL), entire colony : Savigny, Tab. 2. tig. 1. 1. /> j . *Pig. 2. Nephthyd- chabtolii Audouiii, terminal lobe magnified ; Savigny, Tab. 2. fig. 5. 2. *Pig. 3. N.chahfolii, single polyp magnified, lateral aspect; SAvigny, Tab, 2. fig. 5. 3. *Fig. 4. Same, abaxial axis; Savigny, Tab. 2. fig. 5. 4. *Fig. 5. Same, more highly magnified ; Savigny, Tab. 2. fig. 5, 5. Plate LXII. *Fig. 6. N-. cliahroUi, entire colony ; Savigny, Tab. 2. fig. 5. l. Fig. 7. Sclerophytuin, pinnnlatunii sp. n.^ entire colony; from a photo"-raph slightly reduced. ° ' Fig. 8. Nephthya, hedfordi, sp. n., entire colojiy ; from a photograph, natural size. Fig. 9. WrighMla robmfa, sp. n., entire colony; from a photograph, X |. Plate LXIII. Fig. 10. Scleropliytum pinnvlaium, sp. n., tentacle bearing pinnules, X 245. Pig. 11. Nephthya hedfotdi, sp. n., spicules, X 72. Fig. 12. JV. hedfordi, terminal lobe bearing polyps, X 25. Fig. 13. Bendroncphthya diiciformis Kiikth. (aberrant specimen), spicules from the stem-cortex, X 72. Fig. 14. Stereonephthya lutea, sp. n., polyp and " Stiitzbiindel," X 72. Fig. 15. Wrightella robusta, sp. n., stem-spicules, X 72. * Figures marked thus are reproduced from photographs procured by Professor JJourue, and kindly given by him to me for the purpose. 528 MR, T. H. WITHERS ON 27. Some early Fossil CirripeJes of the Genus Scalpellum. By Thomas H. Withers, F.G.S.* [Received January 31, 1912 : Read March 19, 1912.] (Text-figs. 64 & 65.) Index, Page Structnre of curiua oi Scalpelht77t 528 Evolution of early fossil forms of Sea?peZZ«»j 630 Distribution (Geological) of Cretaceous Scalpelhim 537 Sea Ipellum arcuatum Darwin 532 S. trilineatum Darwin 531 ScaJjpellum s-p-p. — discussion of affinities 538 Among the Cirripede remains from the Albian (Ganlt) of Folke- stone in the collection of the British Museum (Natural History), three detached valves were noticed, which, for certain reasons to be explained later (see p. 532), appear to he the original valves upon which Darwin founded the species Scalpellum arcuaticm f. A portion of one of the valves, a carina +, happened to be broken away from its matrix, showing that the inti'aparietes, said by Darwin to be absent in this species, were really present. These parts, however, are developed in such a way that they can be seen only by cleaning the valve free from matrix. The finding of the intraparietes in Darwin's type of /S. arcuaium led to the examination of further examples of the carina, and the fact was established that intraparietes are developed in all carinas of this species, S. trilineatum Darwin (1851, p. 38, pi, i, fig. 5), from the Cenomanian (Grey Chalk) of Dover, was said by Darwin to come nearest to S. arcuatum, owing to the absence of intraparietes. The carina, which is the holotype of *S', trilineatum, is in the British Museum (Natural History), registered 38461, On freeing this specimen from the gum and matrix which obscured the inner portion of the valve, it was at once apparent that not only were the intraparietes present, but that they were developed almost exactly as in S. arcuatuin. Other Lower Cretaceous species of Scalpellum in which the intraparietes of the carina are said to be absent are S. simplex Darwin (1851, p, 39, pi, i. fig. 9) and S, accumidattcm Withers §, both of which come from the Aptian (Lower Greensand), The xniique carina of ^S*. simplex cannot at present be traced, but since the parietes do not reach to the basal margin of the valve (see text-fig. 64, 1) as in those carinas which have intraparietes, it does not appear probable that intraparietes could be present. In * Communicated by Dr. W. T. Calman, P.Z.S. t C. R. Darwin, 1851, Pal. Soc. Monogr. Foss. Lepadidse, p. 40, pi. i. fig. 7. J For the names of the various valves see text-fig. 65. § T. H. Withers, 1910, Gcol. Mag. dec. v. vol. vii. p. 152, figs. 1-1. FOSSIL CIRRIPEDES, 529 S. arcuatum (text-fig. 64, 3) and S. irilineatum (text-fig. 64, 4) the parietes reach to the basal margin, and in my opinion the intraparietes will be found to be developed in aS'. accumulatum (text-fig. 64, 2) as in those species, since *S'. accuvmlatum is similar in its general external form. The only carina known Text-fig. 64. Early types of tlie Carina of Scalpellum. 1 a-5 a, side view ; 3 S, 4 5, 5 5, inner view ; 1 c-5 c, transverse section near apex. t., tectum ; p., parietes ; i.p., intraparietes. of S. accumulatum is embedded in such a hard matrix that it is impossible to expose its inner siirface without danger to the speci- men, so that the matter must here remain until further specimens are forthcoming. So far, then, we have proved the existence of intraparietes in 530 MR. T. II. WITHERS ON the carinfe of the two species S. arciiatum and S. trilijieatum, and pointed out the possibiHty of their presence in /S'. accumulaluin. The intraparietes in these species, instead of forming a thin wall on each side of the carina as in the form represented by text- fig. 64, o, are bent inwards almost at right angles, and the upper regions of their inner margins meet to a greater or less extent (see text-fig. 64, 3 b, 4^); the upper part of the valve is solid, and must have projected freely to the extent indicated by the meeting of the intraparietes*. The peculiar development of the intra- parietes in the carinse of these species is therefore of importance as showing a development of the carina distinct from that in which the intraparietes form a thin wall on each side of the carina. Three types of carina, all having an apical umbo, were there- fore already developed amongst these early forms of Scalpellu.in, and the geologically oldest of these more closely resemble the carina of PoUicipes, from which /ScalpeUiom is considered to be derived. These are (1) represented by aS'. simplex (text-fig. 64, 1) from the Aptian (Lower Greensand), wdiich has no intraparietes, the tectum being flatly-arched transversely, the paiietes bent almost at right angles to the tectum and not extending to the basal margin ; this type is distinguished from the PoUicipes type of carina only in the parietes being separated from the tectum by a distinct angle ; (2) represented by S. accumulatum (text- fig. 64, 2) (Aptian, Lower Greensand), S. arc^uitum (text- fig. 64, 3) (Albian, Gault), and S. trilinecdum (text-fig. 64, 4) (Cenomailian, Grey Chalk), which have intraparietes, these parts being bent inwards almost at right angles and joining, the upper part of the valve being solid and px'ojecting freely to a consider- able extent ; in the solidity and free projection of the upper part of the valve this type is allied to PoUicipes ; (3) represented by S. hastatum (text- fig. 64, S) and other species from the Ceno- manian which have intraparietes also, but are characterized by these parts forming a thin wall on each side of the carina, the apex of which projects freely, slightly, or not at all ; this latter type is more typical of the genus ScalpeUttm, and, owing to the upward growth of the intraparietes in some forms, subsequently gave rise to the species with an angularly bent carina having the umbo in a subcentral position, a type which is not known below the tipper Senonian. The only species with an angularly bent carina known from the English Chalk is S. darioiniammn Bosquet t, but this has an early specialized form of carina in which the upward * A somewhat similar development of the intraparietes can be seen in the carinic of S. maximum var. cylindraceum Darwin (1851, p. 33, pi. ii. fig. 2) from the Upper Senonian {BeleniniteUa vmcronata-ior\e) of Norwich, Norfolk, and S. solidtdum Steenstrup (1839, Kr^yer, Natm-hist. Tidsskrift, Bd. ii. p. 412, pi. v. figs. 14, 14* ; Darwin, 1851, Pal. Soc. Monogr. Foss. Lepadidae, p. 42, pi. i. fig. 8) from the Upper Senonian of Kjuge, Scania. f J. Bosquet, 1854, Les Crust. Foss. du Terrain Cretace du Duche de Limbourg, p. 46, pi. iv. figs. 6-12 ; T. H. Withers, Jan. 1911, Geol. Mag. dec. v. vol. viii. p. 23, figs. 3-4. FOSSIL CIRRIPEDES. 531 extension of the valve is clue, not merely to the upward growth of the intraparietes, but to the ahnost equal upward and downward growth of the valve from the umbo, the whole external surface of the valve being ornamented. It is, therefore, probable that species with an angularly bent carina, due to the upward growth of the intraparietes, which is apparently the more primitive type, existed even earlier than in the Upper Senonian *. Extreme interest attaches to the remarkably complete example of S. arcuatum from the Albian (Gault) of Folkestone, Kent, here described and figured (p. 534, text-fig. 65, 6?). To find so many valves of the capitulum in position is a remarkable circumstance, very few specimens of Scalpellum having been discovered with several valves associated. Those known up to the present come from the Upper Senonian, and, with the exception of a single incomplete capitulum of S. ma.vimit7n, belong to the species S.fos- sula. This specimen, therefore, represents the only fossil species of Scalpelhmi obtained from below the Upper Senonian with any considerable number of valves in position. Previously described species of Scalpellum from rocks below the Upper Senonian have all been founded on detached valves, and in most cases on single valves. Neither the upper latus, rostral latus, or rostrum, are included in these descriptions, but are represented in the j^resent specimen. A further example of ^S'. arcuatum (described, p. 533) has furnished some scales of the peduncle, which so far are not known in any species found below the Upper Senonian. *S'. arcu- atum is the only representative of Scalpellimi known from the Albian (Gault) of England. Consequent on the discovery of the intraparietes in the carinEe of S. arcuatum and S^ trilineatum, and since the carina is the typical valve of the genus, it is here proposed to give fresh diagnoses of those species and to describe the new material of >S'. arcuatum^ From the two examples of S. arcuatum it is possible to construct a restoration of the capitulum, to which only two valves are diagrammatically added (see text-fig. 65, 7, p. 534). Scalpellum trilineatum Darwin% 1851v ScaljKllum trilineatum C. R. Darwin, Pah Soc. Monogr. Foss. Lepadidfe, p. 38, pi. i. fig. 5. 1854. /Scalpellum trilineaticm C. R. Darwin, Ray Soc. Monogr. Sub-class Cirripedia, Balanidse, Synopsis et Index Systematicus, p. 633. 1854» Scalpellum trilineatum J. Morris, Oat. British Fossils, 2nd ed., p. 97. 1877. Scalp)ellitm trilineatum H. Woodward, Brit. Mus. Cat. Brit. Foss. Crustacea, p. 143. * Since the above was written, two exceedinglj^ small, angularly bent carina) '(length respectively 1'7 mm. and 1"8 inm.) with the umbo subcentral, due to the Upward growth of the intraparietes, have been obtained by Mr. F. Mockler from the Cenomanian (Chalk Marl) near Cambridge. The specimens are incomplete and poorlj' characterized^ but of much importance as showing that species with an angularly bent catiua e.\isted even so far back in time as the Cenomanian. 532 MR. T. H. WITHERS ON This species was founded on a carina and tergum from the Cenomanian (Grey Chalk) of Dover. The tergum cannot now he found, but the carina, which Darwin considered to be the typical valve of the genus and can therefore be regarded as the holotype, is in the British Museum (Natural History) registered 38461. Diagnosis. — Carina with three prominent, rounded, longi- tudinal ridges on its tectum — one central, and one on each side separating the tectum from the parietes; intraparietes bent inwards almost at right angles, the inner margins meeting a short distance below the apex, upper part of valve solid and projecting freely. Description of Carina. — Carina narrow, widening very gradu- ally from the apex, considerably bowed inwards, basal margin obtusely angular. Tectum flatly-arched transversely, with a central, prominent, rounded ridge extending from the apex to the basal margin, and bounded on each side by a slightly coarser but flatter ridge on the angle separating the tectum from the parietes. Parietes narrow, less than half the width of the tectum, bent almost at right angles to the tectum, slightly concave. Intraparietes very narrow, bent inwards almost at right angles, the inner margins meeting about one-sixth the length of the valve from the apex ; the upper part of the valve is solid, and must have projected freely to the extent indicated by the meeting of the intraparietes. Lines of growth plainly marked. ScALPELLUM ARCUATUM Darwin. (Text-fig. 65.) 1851. Scalpellum arcuatum C. R. Darwin, Pal. Soc. Monogr. Foss. Lepadidfe, p. 40, pi. i. fig. 7. 1854. Scalpellum arcuatum C. R. Darwin, Ray Soc. Monogr. Sub-class Cirripedia, Balanidse, Synopsis et Index Systematicus, p. 633. 1854. Scalpellum arcuatum J. Morris, Cat. British Fossils, 2nd ed. p. 96. 1865. Scalpellum ao^cuatum J. W. Salter & H. Woodward Cat. & Chart Foss. Crustacea, p. 27, pi. i. fig. 14, 1877. Scalpellum arcuatum H. Woodward, Brit. Mus. Cat. Brit. Foss. Crustacea, p. 142. The species S. arcuatum was founded on three detached valves, namely, carina, scutum, and tergum, from the Gault of Folke- stone, Kent. These valves were considered by Darwin to belong to the same species, and the material now to be described proves such to be the case. Darwin further stated that these valves were in the Bowerbank Collection. This collection was acquired in 1865 by the British Museum (Natural History), and among the specimens are three valves, a carina (I. 13796), a scutum (I. 13797), and tergum (I. 13798), mounted together on Bower- bank's original tablet, and labelled in Darwin's handwriting " S. arcuatum." The carina has been much broken, presumably since Darwin described it, but the scutu.m and tei-gum are in good FOSSIL CIRRIPEDES. 533 condition. These specimens are also exactly half the size of Darwin's figures, which are enlarged to two diameters, and in these circumstances there seems little dovibt that they ai-e the original valves figured by Darwin. The carina, since Darwin considered it to be the typical valve of Scalpellum, is consequently regarded as the holotvpe. Two specimens in the British Museum (Natural History) add considerably to ova- knowledge of this species. One, registered (I. 13577), text-fig. 65,5, fii^st appeared to consist of a carina with fragments of the scuta and terga. Careful clearing away of the matrix and the abundance of gum which covered the specimen showed, however, that several of the valves of the capitulum were preserved. These consist of the carina, the paired scuta, the paired terga, an upper latus, a rostral latus, and the rostrum. The left side of the capitulum is uppermost, and the carina is so exposed that the intraparietes can readily be seen. Portions of the inner surfaces of the right scutum and tergum are also exposed owing to the fact that the upper valves are somewhat displaced and broken. The second specimen (I. 13580) has not so many valves pre- served and those present are somewhat broken. The i-ight side is uppermost, and the plates shown are the carina, the pair of terga, the upper portion of tlie left scutum showing its inner surface, and the left upper latus. This last valve showed only its inner surface, but on removal from the matrix it was found to be the left upper latus. This specimen is of interest since it furnished eighteen scales of the peduncle. A supposed shell- fragment was removed from between the two terga, and on being cleaned was found to be a peduncle scale. On removing and washing the remaining matrix from the same position, the number of scales was increased to eighteen. There is little doubt that these scales belong to *S'. arouatitm, since it is the only species of Scalpellum known to occur in the Gault. Moreover, they are ridged like the valves of the capitulum of that species, and the fact that they were found in such close association is prima facie evidence that they belonged to the same individual, and were washed into the position in which they were found. They are somewhat similar in shape to the scales of the peduncle of the Upper Senonian species S. viaxiiniom and aS'. fossula as figured by Th. Marsson *, but are easily distinguisha,ble from them'^by their longitudinal ridging. Diagnosis. — Capitulum composed probably of fourteen valves t, which are ornamented with numerous fine ridges radiating from their apices ; umbo of all valves apical. Carina with tectum flatly arched transversely, parietes rectangularly inflected * Th. Marsson, 1880, " Die Cirripedien vuid Ostracoden der weissen Sclireibkreide der Tiisel Riigeii," Mitth. natuvwiss. Vereine Neu-Vorpommern und Kiio-en xii pi. i. figs. 2 5, c, '7, 3rf, f. t It is possible tliat this species had a subcavina, in which case the number of valves would he fifteen. A higher number of valves is not likelJ^ 534 MR. T. H. WITHERS ON Text-%. 65. Fif Fig. Fig, Fig. Fig. Fig. Fig. Fig 6. Scalpellum arcuatmn Darwin, Nearly complete capitulum showing the left side uppermost, with the valves somewhat displaced and broken. X 2 diam, Albian, Gault: Folkestone, Kent. Brit. Mus. (Nat. Hist.), 1. 13577. c, carina, showing the inflected intraparietes ; t.\, oviter portion of left tergum ; t. 3, inner view of right tergum near apex, showing the ridges evidently connected with the firm attachijient of the corium j s. 1, outer portion of left scutum ; a. 2, inner view of right scutum showing the pit for the adductor scutorum j I., upper latus j r.l,, rostral latus with ijpper portion broken off; r-, rostrum. 6 a. Id. Outer view of incomplete rostral latus, "^ 6 6. Id, Inner view of same. | 6 c. Id. Outer view of rostrum. V X 4 diftin. 6 d. Id. Hypothetical transverse section of same. 6 e. Id, Side view of same. J 7. Restored capitulum of Scalpellum arcuatum Darwin. This figure is based on the neai-ly complete capitulum represented by flg. 6, with the addition of a cai'inal latus and an infrajmedian latus. These two valves have not yet been discovered in the Gault. X 2 diam, Albian, Gault : Folkestone, Kent. C.I., carinal latus ; i.l., infra-median latus, , 8, Scalpellum arcuatum Darwin. Two scales of the peduncle, a, outer view, 6, inner view ; c, outer view; c^, inner view. X 8 diam. Albian, Gnujt i Folkestone, Kent, Brit, Mus. (N?it. Hist,), I, X3580, FOSSIL CIRRIPEDES. 535 intraparietes bent inwards and meeting for about one-fourth the length of the valve from the apex, upper part of valve solid and freely projecting. Scutum with tergo-lateral angle almost in line with the middle of the valve, Tergum subrhomboidal, a delicate furrow extending from the apex to the basal angle. Upper latus i^ubtriangular, apex acutely angular, slightly bowed towards scuta, basal margin I'ounded. Kostral latus acvitely angular transversely, ftbout 2 2 times as long as wide. Rostrum subtriangular, with a strong median keel extending from the apex, widening towards the convex basal margin. Description of valves. — Carina narrow, widening gradually from the apex, considerably bowed inwards, basal margin obtueely V'^haped. Tectum flatly-arched transversely, obscurely carjnfite, and oniamented with numerous fine longitudinal ridges, Parietes niirrow, less than half the width of tectum, not longitudinally ridged, bent almost at right angles to the tectum, slightly concave, Intraparietes very narrow, bent inwards almost at right angles, the inner margins meeting about one-fourth the length of the valve from the apex, above which the valve is solid and must have projected freely. Scutum moderately convex, divided unequally by a prominent ridge running from the apex to the basi=lateral angle, the basi- lateral angle being slightly produced. Apex acuminate. Basal margin sinuous, about two-thirds the length of the valve ; occludent margin slightly convex and nearly parallel to the lateral margin; tergal and lateral margins of almost equal length and forming an angle of about 145°, either margin being about half the length of the valve. Surface of valve ornamented with fine, closely-set, longitudinal ridges ; a narrow slip along the tergal margin is somewhat bent downwards and is devoid of ridges, Tergum sub-rhomboidal in general outline, slightly convex, with a delicate furrow extending from the apex to the basal angle. Apex and basal angle acuminate, more so than is indicated in Darwin's original figure ; scutal angle somewhat protuberant, Carinal margin convex ; scutal margin sinuous, longer than the occludent margin, which is nearly straight. Surface of valve ornamented with numerous fine longitudinal ridges. Inner surface, in the region of the apex, is marked on its edges with oblique lines of growth, these indicating the extent to which the valve projected freely. A little below the apex, nearer to the occludent margin, are three or four small ridges ending abruptly about one-fourth the length of the valve from the apex. These ridges were evidently connected with the firm attachment of the corium, or membrane which lined the inner surfacesof the valves, and are homologous with the series of tubercles on the inner surfaces of the terga of *S'. darioinianum Bosquet. Upper latus subtriangular, slightly curved towards the scuta, almost flat transversely, convex longitudinally ; umbo slightly projecting, with a thick ledge formed beneath it, which thins out towards the lateral angles ; tergal margin slightly convex ; scutal 536 MR. T. H. WITHERS ON margin slightly concave and about the same length as the tergal margin, the two margins if represented by lines from the apex to the lateral angles would enclose an angle of about 55° ; basal margin rounded, and indistinctly marked off into three lines ; a portion of the valve on either side, parallel with the scutal and tergal margins, is somewhat raised, and the lines of growth on these parts are upturned sharpl}^ towards the umbo. Surface of valve between the raised portions ornamented with several longi- tudinal ridges. Rostral latus. — This valve is imperfect, the upper portion being broken off. Yalve very narrow, acutely angular transversely, about 2| times as long as wide, widening gradually from the inner acute extremity to the rostral margin, which is abruptly truncate ; umbo apical ; the inner extremity is marked by a strong rounded keel; outer (rostral) half of valve ornamented with fine longitudinal ridges, two fine ridges close together almost dividing the basal margin into two equal portions. At the point where these two ridges reach the basal margin the valve is, some- what convex, no doubt indicating the extent to which the valve was bounded by the inframedian latus. Rostrum sub-triangular, strongly convex transversely, bowed inwards ; lateral margins bounded by strong ridges ; basal margin convex ; a strong median rounded keel extends from the apex, widens considerably towards the basal margin, and is bounded on either side by indistinct longitu:dinal ridges. Peduncle scales varying in shape from semilunar to trapezoidal, the basal margin of the former being straight, while that of the latter is somewhat concave ; immediately below the apex the trapezoidal scales are slightly constricted, the truncated top appearing to overhang ; scales thickest at one-third from the base, below which, on the inner surface, they are somewhat excavated, this indicating the extent to which the scales were covered by the corium, the upper two-thirds no doubt overlapping the contiguous scales. Outer surface ornamented similarly to the valves of the capitulum with a number of longitudinal ridges. These ridges number about seven on the lai-ger scales, and four or five on the smaller scales. Measurements. — Owing to the fact that the valves in these two specimens of *S'. arcuatum are broken, it is impossible to give their accurate measurements. Since, however, it is desirable that we should have some idea of the relative sizes of the valves in an individual, approximate measurements are given where the connect measurements cannot be obtained. Specimen I. 13577, Carina. Length circa 21 mm. Scutum (right valve). Length (from apex to rostral angle) 13" 5 mm. ; bi'eadth 7*4 mm. Tergum. Length circa 17 mm. ; breadth 8*2 mm. FOSSIL CIRRIPEDES. 537 Upper latns. Length 6"8 mm. ; breadth circa 7'5 mm. Rostral latns. Length 6'3 mm. ; breadth 2'6 mm, Rostrum, Length 3'4 mm, ; breadth 2-2 mm. Specimen I. 13580. Cai'ina. Length circa 20 mm, Tergum. Length circa 17 mm. ; breadth 7"5 mm. Upper latus. Length 5-9 mm, ; breadth circa 4'2 mm. Scales of peduncle. Length from 0-5 mm, to 1*2 mm. ; breadth 1'3 mm, to 2 mm. Remarhs^ and comparison loith other Species. — The restoration of S. arcuatum (text-fig, 65, 7) is based on the nearly complete capitulum figured on the same page, and, to complete the capitu- lum, a carinal latus and an inframedian latus have been added. The carinal latus figured in the restoration was found amongst a nnmber of detached plates of S. arcuatum from the Cambridge Greensand, and is longitudinally ridged as in the valves of >S'. arcuatum. It possibly belongs to the same species. In any ease the only carinal latera known to the writer from the Lower Cretaceous rocks are of the type figured, although they evidently belong to several diflTerent species. Judging from the hiatus between the carinal latus and rostral latus, an inframedian latus was undoubtedly present, and was probably very like the homologous valve in S.fossula from the Upper Senonian, The specimen of *S'. arcuatum here figured (text-fig, 65, G') is, up to the present, the oldest fossil Scalpelhbm from Avhich one can gain any idea of the appearance of the complete capitulum. It comes from the Albian (Gault) of Folkestone, Kent, and the only undoubted valves of Scalpellum older than this occur in the Aptian (Lower Greensand) *", These Lower Greensand forms comprise only three species, two of which — S. simplex Darwin and S, accumiolatum Withers — are respectively i-epresented by a single carina ; the third, ^S. comptmn Withers, is represented by two detached terga. Our knowledge of these early forms of Scalpellum is therefore not very extensive, and the fact that they are represented by such a small number of valves, and those only of carinfe and terga, serves to emphasize the importance of this fine example of *S'. arcuatum (text-fig. 65, G). A comparison of the carina and tergum of ;S', arcitatum with the corresponding valves of S. SGliduhim Steenstrup, as figured by Darwin (1851, p, 42, pi. i. fig, 8), shows how closely these two species resemble each other. They are evidently related. The * This statement is made with full knowledge that certain valves from the Palaeozoic and Jurassic rocks have been referred bj' various authors to the o-enns Scalpellum. There is, however, not sufficient evidence to justify the reference of these valves to the genus Scalpellum. Notwithstanding this, it is possible that some of the Mesozoic Cirripedes referred to Pollicipes may eventually be shown to be ancestral forms of Scalpellum, but this cannot be done until moi'e is known of the various valves comprising the capitulum. Proc, Zool, Soc— 1912, No. XXX YI. 36 5:58 ox FOSSIL CIRRIPEDES. carinie are easily distinguished, for whilst in >S'. arcuatum the tectum is flatly-arched ti-ansversely, and the intraparietes are bent inwards almost at right angles, the tectum in S'. solidtdum is strongly convex, and the intraparietes join to form a prominent crest. Darwin considered the scutum of &'. solidulum to be like that of *S'. arcimtimi, with the exception of the longitudinal ridges being proportionally broader and further apart, closely resembling those in the carina of S. solidulum. Unfortunately the only known complete capitulum from the Cretaceous rocks with which that of *S'. arcuatum can be compared is that of the XJj)per Senonian species S. fossula. In the relative positions of the valves both species are alike, but in the structure of the carina and scutum there ai-e important diflferences. The intraparietes of the carina of >S'. arcuatum are sharply bent inwards, the upper part of the valve is solid and must have projected freely to some considerable extent. The carina of ,S'. fossula, on the contrary, projected freely but little, and the intraparietes form a thin wall on each side of the carina. In S. arcuatum the tergo-lateral angle of the scutum is situated much further from the apex than in ^S'. fossula, and in this respect is farther i-emoved from the more advanced scuta which have the tergo-lateral angle almost in line with the apex, above which the valve is added to, the umbo consequently being sub- central. The valves of *S'. arcuatum are longitudinally ridged, while those of S.fosstda are smooth. Affinities of the Species m.entiooied. >S. arcuatum is no doubt an ancestral form of a group of almost exclusively deep-sea species, which Dr. P. P. C. Hoek* has separated" as a sub-genus under the name Arcoscalpellum. S. trilineatum, &'. accumulatum, S. comptum, S. m.axi7num. var. cylindraceum, and *S'. solidulum., which appear to be related to ;S'. arcuatum, possibly belong to the same group, but we know too little of these species to say much about them. S. slmj^lex probably does not belong here. The species embraced by the sub-genus Arcoscalpellum have no sub-carina, and it is impossible at present to say whether S. arcuatum had a sub-carina or not. The Senonian species S. fossida also comes nearest to the subgenus Arcoscaljjellum,, but this species is said by Ed. Hebert t to have a sub-carina, and in view of this it is possible that the Albian *S'. arcuatum a,lso may have had one. Assuming this to be the case, we have two forms agreeing in all essential characters with the forms of Arcoscalpellum,, except that they (possibly) have a sub-ca,rina. Moreover, the two species differ in the form of the carina, -S'. arcua- tum having the intraparietes bent inwards at right angles and * P. P. 0. Hoek, Oct. 1907, Siboga-Expeditie, Cirripedia Pedunculata, p. 59. f Ed. Hubert, 1855, Mem. Soc. Geol. France, 2° ser. vol. v. p. 356, pi. xxviii. fig. 1 (S. gaJlicwn — S. fossula). P.Z.S,1912.P1.LXVI. 7 8 9 West, Newraan ad nat olrr FEATHERS FROM PHEASANTS, 9. P.Z.S.1912.P1.LXVIL 9 8 West.lSrswman ad nat. clir. ON EXPERIMEXTAIi PHEASANT-BREEDHSTG. 539 ioining, the upper part of the valve being solid and projecting freely, while the carina of S. fossida has the intrapai-ietes forming a thin wall on each side of the carina, which is a more advanced tj^pe. There is not as yet sufficient evidence to prove whether *S'. fossula was derived from 6'. arciudum or not. It is patent, however, that from the foregoing species subsequently arose the forms grouped as Arcosc(dpeUuvi and characterized by a reduction in the number of plates of the capitulum, by a sup- pression of the rostrum, and a tendency towards the reduction of the pair of infra-median latera. For help in connexion with this paper I wish to express my indebtedness to Dr. F. A. Bather, Dr. W. T. Caiman, Mr. 0. P. Chatwin, and Dr. P. P. 0. Hoek. Key to /Sjjecies ^mentioned. A. Carina without intvaparietes, parietes not reacliing to the basal margin and bent almost at right angles to the tectum S. simplex Darwin. B. Carina with intraparietes bent inwards and joining, the upper part of the valve solid and projecting f'reelj\ 1. Carina with basal margin rounded S. accunmlatum Withers. 2. Carina with tectum tlatly-arched transversely, and marked with numerous fine longi- tudinal ridges S. arcuatum Darwin. 3. Carina with tectum tlatly-arched transversely, and marked with three prominent longi- tudinal ridges, one central, and one on either side separating the tectum from the parietes_ S. trilineatum Darwin. 4. Carina with tectum strongly convex trans- versely and marked with several longitudinal ridges ^ S. solidulum Steenstrup sp. 5. Carina with tectum strongly convex trans- versely with smooth surface S. maximum var. cylindra- C. Carina with intraparietes forming a thin wall [ceum Darwin. on each side of the valve. 1. Carina with tectum and parietes smooth, dorsal surface and iimer margin much arcuated S. hastatum Darwin. 2. Carina with tectum bordered on each side by a large, protuberant, flat-topped ridge S. fossida Duxvi'm. 2di. Experimental Pheasant-hreedino-, By Rose Haig Thomas, F.Z.S. [Received December 4, 1911 ; Read February 6, 1912.J (Plates LXiy.-LXVII.*) The expei'iment with which I deal in this paper was under- taken to test the truth of the result of one previously made (P. Z. S. 1909, pp. 884-885), in which it was shown that a male Pheasant had transmitted to his female oifspring of the second * For explanation of the Pktps see p. 545, 3G* 540 MRS. B. HAIG THOMAS OX generation the female plnmage of liis species, under the following scheme of mating : — Silver ? X Swiuhoe S F. 1 ? X Swiiilioo (? (1) F. 2 ? S\v. (1) S Sw. (3) c? S\v. liyb. It is to be observed that the Swinhoe male transmitted the pure Swinhoe plumage only to one of his male offspring. As siich an isolated inst:ince, though interesting, had small value without the support of other evidence, the following experiment ■was ari'anged. I chose species which have matui-e phnnnge in five months from hatching, so as to get quicker nt the i-esults, instead of, as in the case of genna'ns, waiting eighteen montlis for the adult plumnge of F. 2. P/iaslanHf! f(y)'7)iosus, one of the species selected, is found only on tlie Island of Fonnosa ; somewhat resembling P. torquatus, it differs from it in one special feature ; in the cock, the collar, instead of being an even white I'ing forming a complete circle round the neck, is formed by a deep Vandyke on each side with a narrow connecting-line round the back of the neck but not connected round the throat, an incomplete circle. The other species, P. versicolor, the Japanese Phensnnt, is too Avell known to require description ; it has no white collar. Below is appended a, tiible of the differences between the males a,nd females of the two species which might be expected to prove Mendelian pairs. P. versicolor S • Moult: Slow, late. Leff : Colour : I1:ivlr, inUlisli grey. Dimension : Thin. Rill : Small. Crest : Dark. Breast : Metallic green. Tail : Short. Some laterals not banded. P.formosus 6 Moult : Rapid, early. Leff: Colour: Pale bluish grey Dimension: Thick. Bill: Large. Crest : Pale. Breast : Rich brown. Tail: All Long, rectrices banded. Neck: White collar. Bird; Large. P. formostis 5 . Leff : Colour : Pale bluish grej'. Dimension : Thick. Bill: Colour Crest : Breast : Tail: Bird: Egg: Large. Greyish gape. Pale ground, black spots. Pale bufl", no pattern. Banded pattern on ground. Large. Large. green, paler at oval pale Calm temperament : quiet, tame. Neck Bird : Lea- : Colour White collar absent. Small. P. versicolor 2 Reddish grey. Dimension: Tbin Small. Darker greyish green, same all over. Dark ground, irregular black jiatclu's. Dark buti; thickly patterned. Banded pattern on dark ground. Small. Small. Excitable temperament: wild, untame- able. Bill : Colour Crest ; Breast Tail: Bird : E.VO-: EXPERIMENTAL PHEASANT-BREEDING. 541 Scheme of mating in second experiment : — F.formosus ? X P. versicolor $ (3) F. \ $ Jb'.l ? " Y" F. 1 ? "Z" X P. versicolor $ (5) F. 2 ¥ Ve. (2) $ Ve. hyb. The second experiment confirmed the results obtained in the first. Four skins of the F. 2 5 's are exhibited (the fifth female has been kept to breed from), together with skins of pure Versicolor and pure Formosan females, also the skin of the F. 1 $ "Z" parent of F. 2. If these are all turned breast vippermost, it is seen at a glance that Versicolor pattern and coloration are present to a certain extent in F. 1 $ "Z," mother of the F. 2 generation, and that the F. 2 females appear to be identical with Versicolor in size, pattern, and coloration, except for small difi'erences which might be found existing between individuals of any species. Besides plumage, these five hens had the habit and temperament of the Versicolor, the leg-colour and eye-skin also. The records of this experiment constantly refer to the quick sharp movements, the wild scared appearance of the F, 2 females, so characteristic of the habits and ways of the untameable Versicolor. If the pen was entered for any purpose, even when exercising the greatest care, there was always a chance the birds might break their necks in their terrified flights and violent dashes against the wire netting, and in this manner one or two were scalped to the bone. A list of the Versicolor characters found in the F. 1 females is interesting, for there are certain character's which can only be classified in the living bird : Leg-colour, eye-skin, bill, moult, habit, temperament, voice. F. 1 $ " 7u" {F.formosus 5 X P. versicolor c?). Parent of Eye-skin. ^ F. 2 (Fo. X Crest. 1 Ve.x Ve.). Neck. ;^ Flank. 1 '~~ Breast. 1 o Interscapulars, >-- 00 Scapulars. - Secondaries, ^^ Back. Size of egg. J Tail Formosan Bill. 1 ^• Primaries. s Dimension. f>. Leg. ^ 542 MRS. R. HAIG THOMAS ON F. 1 2 " Y." Parent of Eye-skin. ^ inter se F. 2. Crest. Neck. ^ Brenst. o Flanks. Interscapulars, CO Scapulars. ~ Secondaries. ^ Back. Size of egg. Lesf Formofian o Bill. , Priniai'ies. '£ Dimension. Tail. In both F. 1 females the size of egg was transmitted by the Versicolor male. 5 F. 2 2 . Crest. Neck. Flank. Breast. Legs. Bill, size and colour. Interscapular >s . Scapulars. Secondaries. Primaries. Tail. Back. Dimension. Habit of moult. Temperament. J The fifth female, reserved to breed from, has a pale grey stripe down the back of the shank of one lee;. The presence of a mosaic of pale grey and dark grey seen in the legs of F. 1 in this experiment led to the inference that the two parents had severally pale grey legs and dark grey legs, which an immediate examination of the Formosan and Versicolor species confirmed. A curious independent double segregation of allelomorphs was observed in the autumn of 1911 in the crests of the F. 1 males. The centre feathers were dai-k — colour Versicolor ; fully developed, rapid early moult — habit Formosan : the feathers on the outer edge of the crest were pale — colour Formosan ; EXrEKIMIiXTAL PHEASAXT-BJIEKDJXG. 543 undeveloped, many still in the quill — habit Versicolor : two Mendelian pairs coupled in each parent yet repulsing one another in the same area in F. 1. An examination made at the same time of the Formosan male showed the crest fully developed, and an inspection of the Versicolor male showed the crest undeveloped, mostly still in the quill. F. 1 generation is to me always the most interesting in these artificial Pheasant crosses. For Mendelian segregation already shows, and it is sometimes possible to select the strain of parentage desired to reappear more strongly in the F. 2 generation. Also F. 1 occasionally produces remarkable mosaics of sex — a sort of sex-hybridism accompanied by sterility and extraordinary develop- ments of plumage in the female, phenomena I will not touch upon now. To illustrate more clearly the points I wish to bring to your notice, here is a selection of secondaries (3rds fi'om the last primary) extracted from the wings of the parent species and of all the birds connected with the experiment (PI. LXIV.) : the Formosan male secondai-ies have a peculiax- Vandyke pattern like a feather laid on a feather, and the female secondaries of this species are banded, both are extremely light in colour. The Versicolor male secondaries have a mottled grey oblique banding on a very dark grey ground ; the female secondaries of this species are rich brown, with wide bands of darker brown ; also placed in the same frame are the secondaries of F. 1 and F. 2 males and females. After examining the parents' secondaries, we perceive that the Versicolor male has transmitted to his female offspring of the F. 1 generation the female secondaries of his species, and that conversely the Formosan female has transmitted to her male offspring of the F. 1 generation a pattern resembling the male secondaries of her species on one vane, though not on the other vane of the feather. In the F. 2 generation (" Fo xVe. xVe." PI. LXV.) the influence of the Versicolor male on his female offspring continues in pattern and size and more or less in colour, for in F. 2 females the secondaries seem to be Versicolor, with slight differences of colour not unlikely to be found between individuals of the same species ; whilst the F. 2 male secondaries, though most resembling Versicolor, are somewhat hybrid in size and pattern, and still show slight traces of Formosan influence. A selection of interscapulars extracted from the parents and from the two generations of the Formosan Versicolor cross is shown ; the females are in one frame, the males in another, and in these the same phenomena appear. Those of the F. 1 fema.les seem to be Versicolor in pattern and coloui-, as also do the five F. 2 females, with certain modifications that might be readily found amongst individuals of a pure race (PI. LXVL). The frame containing the male interscapulars shows amongst the F. 1 males a Formosan and also a hybrid pattern, whilst the F. 2 males are also hybrid (PI. LXVII.). The phenomenon of pattern-transference has occuri'cd in all my Pheasant crosses ; sometimes fiom the male to the female, or 544 MRS. R. HAIG THOMAS ON conversely from the female to the male, or a pattern may be transferred from one area in the parent species to another area in the F. 2 offspring. I have noticed that these pattern-trans- ferences are inclined to remain fixed and constant. Colour-transference also takes place, and sometimes where it has occvirred seemed to inhibit the appearance of pattern. One instance was noticed where the breasts of the males of the two races crossed, Th. amhersti and Th. picta, differing widely in colour, both colours were found on the breast of F. 2 Th. ohscura, the red o1 j^icta overlying the metallic green of amhersti. These appear to be the results of the second experiment : — ■ (1) The male parent transmitted to his F. 1 female offspring much of the female plumage of his species and the dimension of the Qgg. (2) The female parent transmitted to her F. 1 male ofi'spring much of the male plumage of her species. (3) In the F. 2 generation, the offspi'ing of F. 1 female x Versicolor male, the Versicolor male seems to have trans- mitted every character — bill, leg-colour, plumage, habit, and temperament — of the female of his species to his F. 2 female offspring, whilst he has not transmitted every character of the male of his species to his F. 2 male offspi'ing; repeating exactly the results of the original expeiiment with geiinceus. Are we, then, to suppose that some of the gametes of this Versicolor male contained all the factors lepresentiug the tem- perament and liabit, the colou)-, pattern, and dimension of plumage, leg, and bill, and the bulk of the female of his species and even the factor for size of ^gg, with the one exception of the factor for the sex to which these belonged ? To the practical experimentalist, to the non-mathematical simple observer, the hypothesis is difficidt to conceive. These phenomena seem to be of the nature of a sex-limitation opposed to expectation. I have read with much interest Mr. Doncaster's account of game- togenesis iir the Gall-fly, also his researches on sex-liraitations published in ' Genetics,' and am interested to know how he w^ould consider the above facts in relation to his theory of sex : Male gametes c? female gametes J 5 with selective fertilisation between the male gamete O (a non-determinant of sex) and the female gamete S . I hope the material collected in these two experiments may be thought of sufficient importance for the higher students of Genetics to give it some attention, when probably the apparently complicated problem will receive a simple explanation. EXPERIMENTAL PHEASANT-BEBEDIXG. 545 EXPLANATION OF THE PLATES. Plate LXIV. > of (J versicolor. Coarser lines but same pattern as found in souie posterior interscapulars on ,:? ^^ ' > r< 1 .y 1' /7 ; 1 M '/ , L ^' k.i'si ♦. -« Text-fig. 75. — A portion of the medullary parenchyma of Inermicapsifer capensis, showhig centrally a condensation of tissue which is connected with the rudimentary uterus. Text-iig. 76. — A portion of the ovary of the same species with eggs (o.) detached from the ovary and lying in the parenchyma, and a portion of the network of tissue {x.) which ma3' possibly represent the uterus. already described, is seen to be prolonged into an extension not distinguishable from it, which runs for a little way into the meduUaiy parenchyma towards the median point of the proglottid. It is quite short, and ends more or less abruptly in a sti'and of condensed parenchymal tissue, which is apt to be branched, sending out shorter strands of a similar appearance in an oblique direction. Some of these ceased after a short course. In neigh- bouring regions of the proglottid (text-figs. 75 and 76) there were patches or rather stx-and-like parts of the medullary parenchyma NEW TAPEWORMS PROM THE HYRAX. 587 exactly similar to those which have been described as continuous with the prolonged oviduct, but not always continuous with it. They would seem to be produced indepeiidently of the genexvative ducts. A network arrangement often characterised these regions of the medullary parenchyma. This condition of the uterus characterised proglottids in the same stage of evolution as those which in Thysanotcenia gambiana possessed an uterus with a distinct lumen. But if we are to compare them with earlier proglottids in Thysanotcenia, then it must be remarked that in the present species of Inermicapsifer there is no such regular band of nuclei marking out the future uterus, as exists in the former worm! The appearance is, in fact, totally different. In Inermicapsifer the suggestion is merely of a crowding together in a quite irregular fashion and a local multiplication of the nuclei of the cells of the medullary parenchyma. In Thysanotcenia the nuclei are in orderly arrangement and regular sequence, obviously belonging to an organ which, as already pointed out, swells at its medianly situate end into an oval region of greater diameter. In Thysanotmnia we have obviously a regular outgrowth of the generative system ; in my species of Inermicapsifer what would appear to be a condensation of parenchymal tissue in contact with and continuing on a very short, and in this species hollow, outgrowth of the generative duct. I am not clear whether the species described by Janicki agrees most nearly with the species described in the present paper or with Thysanotcenia gamUana. I am disposed to think that the branched strand in which the slightly prolonged oviduct ends in this Inermicapsifer is not the homologue of the solid mass of cells in Thysanotoinia gambiana, which aftei^wards becomes hollowed out to form the uterus of that worm ; in this case Inermicapsifer hyracis agrees more closely with Thysanotcenia gambiana than with the species described here. It might possibly be held that this network, which permeates the segment when seen in horizontal section, is merely a stage subsequent to that figured by Janicki in a proglottid of some- where near the same stage of maturity as that which I am now considering in the species of Inermicapsifer studied by myself— that the lumen had, in fact, existed and had disappeared. That this is not the case is clear from the fact that in this stao-e there were not any ova contained in the meshwork, and in*' fact no ripe ova anywhere outside of the ovary. The meshwork formed by condensation of the parenchyma, in fact, precedes the extrusion of ova from the ovary. Still, it may represent the imperfect remains of a retiform uterus such as is characteristic of the o-enus Dipylidium or, better perhaps, in relation to the present g^enus, the Anoplocephalid genus Andrya. I have also been able to approach the question of the identity or non-identity of the cavity in which the eggs lie with the cavity of the uterus from another side. Text-fig. 77 represents portions of horizontal sections through a proglottid in which the 39* 588 DR. F, E. BEDDARD ON eggs are leaving the ovary to be scattered through the medullary parenchyma, though in this particular stage the whole paren- chyma is not filled with eggs as it is later and at the end when the whole medullary region is crammed with the paruterine organs. As will be seen by a reference to the figures cited, the eggs lie partly in round or oval cavities which might well be Text-fig. 77. •V A B. o c. S- \ C Three sections through the medullary tissue in the neighbourhood of the ovary of Inermicapsifer capensis. In A a single ovum is shown to the left. In the centre is an oval area with strongly marked walls, but filled with the ordinary medullary tissue. B shows eggs (o.) lying each in a single vesicle of the medullary tissue and three eggs lying in a cavity (o.c). C. Three of the cavities (o.c.) containing eggs and also showing remains of the medullary tissue which is shown unaltered in A. thought to be uterine cavities were no further information forthcoming. I believe, however, that these cavities are not remnants of a uterus, as is held by Janicki to be the case with apparently similar cavities in Inermicapsifer hyracis, A close NEW TAPEWORMS FROM THE HYRAX. 589 examination of the cavities seems to show that they are lined by a flattened epithelium, of which the nuclei are apparent and are represented in the figure, and there is a slight accentuation of the wall of the cavity, which is merely, as I think, the portion of the parenchymatous network which abuts upon the cavity and is not a special layer distinct from that ; in this case the ap- parent lining epithelium should be regarded merely as the nuclei belonging to this part of the parenchymatous network. I believe this to be the leal explanation ; for when we look carefully into the cavity itself, in which eggs lie in vai'ious numbers, there is nearly always, if not quite always, some granular detritus to be seen and which is shown in the figures referred to. This formless detritus (as it often, but not always, is) might, of course, be inter- preted as simply coagulated fluid which it would not be surprising to meet with in the interior ^of a uterus were this system of cavities the remains of a uterus. But, as will be seen, this detritus is susceptible of another explanation ; it is, as I think, the remains of the delicate parenchymal network originally present, as is shown in text-fig. 77 A, in these regions of the parenchyma set apart for the development of the eggs. The figure referred to shows plainly an area oval in section and with a slightly accentuated wall, marking it off from the surrounding tissue, which is filled with parenchyma network, but does not as yet contain any eggs. If my contention is right, then it must follow that the space containing developing eggs is at least not always to be referred to a uterus. Nor is it in the least against this view that it is possible to meet with these circular spaces in which nothing is apparent but eggs — that is to say, no remains of the originally present parenchymal network, for this may have completely disappeared. I would furthermore point out that the position of the eggs in the nearly empty cavity shown in text-fig. 77 B suggests that they have only just forced themselves into the cavity. I have yet another argument to show reason against regarding the cavities of the paruterine organs in which the embryos are finally lodged as detached fragments of the pre-existing uterus ; or, at any rate, to show that they cannot always be so regarded. In text-fig. 77 b and text-fig. 76 is represented an ovum lodged in one of the meshes of the parenchymatous network of the medullary region without any special relation to the larger egg- containing cavities which have been looked upon as detached fragments of a uterus. There are plenty of such examples to be seen in sections of this age, and it is plain to me that eggs are constantly lodged singly in the parenchymal network. I argue this from the fact that in these cases the cavities lodging the ' eggs are in every way indistinguishable in size and appear- ance from the cavities of the parenchyma in which an egg is not lodged. It may be, of course, that these are eggs which got extruded from a uterine cavity and forced into the sur- rounding parenchymatous network. But it is equally reasonable 590 DR. F. E. BEDtoARD ON to assume that they have got directly to their situation from the ovary (see text-fig. 76). The lax parenchymal tissue, the mesh- work of which is filled with a substance plainly visible as granular matter after staining with Ehrlich's hfematoxylin, but not to be seen after staining with borax carmine, can offer little obstacle to the immigration of eggs ; so that in any case some of the paruterine organs aie without vestiges of an uterine cavity. I believe, as a matter of fact, that all are so, and that there is no persistent uterus in this worm. The pamterine organs of this species resemble those of the species which I described as Thysanotcenia *. At the time when that description was written, I believe that Janicki's careful account of the genus Inermicapsifer with similar egg-capsules had not actually appeared. I had not realised from the descrip- tions of Zschokkeella that the organs containing the ripe eggs were doubtless of the same structure. I had considered that those oi'gans probably resembled the figure given by Ransom t, not entirely grasping the fact that that figure was intended rather as a diagram to distinguish between those species of Davainea which had several eggs enclosed in one capsule and those species in which each capsule had within it but one egg. I was thus misled, though not through Mr. Ransom's fault. There was, of course, no other genus with which I could directly compare Thysanotcenia ga7nbiana. In defining the genus Zschokkeella, Ransom speaks of the fate of the uterus in the following words : — " Uterus early breaks down into egg-capsules." Earlier in his resum6, Ransom defines the subfamily Linstowinse in the same way ; he remarks that the " uterus breaks down into egg-capsules." As the subfamily Thysanosominse is defined by the presence {intei^ alia) of numerous paruterine organs, I thought myself justified in placing my genus, as I regarded it, in the latter subfamily and marked its affinities by the use of the geneiic name Thysanotmnia. I was indeed of opinion that the uterus in Zschokkeella really persisted in separate pieces, each containing so many eggs. It appeared to me, in fact, after studying a tapeworm which I have lately described in the ' Proceedings ' of this Society as Otiditmnia %, that the fate of the viterus in Zschokkeella might be like that of Otiditce^iia. No figures are given by Fahrmann in his account of Zschokkeella linstovn§ which illustrate this particular point, and the only reference to the matter is the assertion that the eggs are surrounded by a " Parenchymhlille." Janicki ||, however, is apparently of my earlier opinion ; for, in distinguishing between the genus Zschokkeella (written, as originally — Zschokkea) of Fuhrmann and his own genus * P.Z.S. 1911, p. 1001. t Bull. U.S. Nat. Mus. No. 69, 1909, p. 17, fig. 8. t P. Z. S. 1912, p. 194. § Centralbl. f. Bakt. u. Paras. Bd. xxxii. (1902). II Loc. cit. p. 393. NEW TAPEWORMS FROM THE HYRAX. 591 Inermicapsifer, lie uses the differences between the egg-capsules in the two genera. He considers that in Zschokkeella the eggs lie " einzeln in einfache Bindegewebskapseln," a difference which also appeared to me to hold good. I do not, however, feel confident about this point, and in view of other points of likeness between the genera am disposed to compare more nearly the paruterine organs in the two. In my paper upon Otiditcenia just referred to I have dealt to some extent with the "egg-capsules" of Davainea, of which the various figures published are not quite in unison. I find a justification for this in the paruterine organs of Inermicapsifer capensis, where the appearances vary slightly among examples which I cannot refer to different species. To begin with, I prefer the name of paruterine organs for the structures in question, because they seem to me to be exactly comparable to structures so named in other tapeworms. They are, in fact, sacs formed out of the parenchymal tissue, whether they have or have not ultimately a connection with the uterus. I should prefer to term the egg-containing spaces in Otiditcenia " egg-sacs " which are formed in a different way, i. e. by a breaking up of the uterus. When specimens of this Inermi- capsifer were examined alive in salt-solution, the individual paruterine organs could -easily be squeezed out of the body by crushing it. They were then spherical or egg-shaped and appeared to be surrounded by a thick, colourless, and almost structureless membrane. This membrane exhibited only faint striae in a longitudinal direction, being thus concentric in reference to the whole body. The interior was filled with quite transparent spherical embryos, between which were abundant cells with granular contents. In stained preparations (text-fig. 78) the outer layer was also perfectly distinguishable. It was stained more lightly than the inner by carmine and more deeply by logwood. A figure of the mature organ in Inermicapsifer hyracis is given by Janicki *, with which may be compared my own figure of the same organ in the species described here. The appearances shown by the organ in the living condition are not borne out by preserved and stained material. The outer layer is not fibrous but cellular, as shown by Janicki and others. This layer is, however, as a rule, quite distinct from the inner mass of cells immediately surrounding the embryos. The distinction between the two partly depends, as already mentioned, upon stains, and is not always obvious, I do not see in my specimens of Inermicapsifer capensis so great a distinction between the outer and inner cells in point of size and shape as does Dr. Janicki, which is perhaps rather remarkable in consideration of the very different appearances they present when living. The cells of both layers are, in fact, rouxided and nucleated, and not greatly different in size and shape. Those of the inner layer are filled with larger spherules. They are * Loc. cit. pi. xiv. fig. 28. 592 DK. P. E. BEDDARD ON separated by a more deeply staining reticulum, which appears to be similar to that of the general parenchyma. This forms a some- what thicker layer enclosing the whole organ. The reticulum of the inner mass of cells is thicker than that of the outer layer, and the nuclei lying at intervals along its strands are very obvious. The reticulum also forms a distinct layer separating the outer from the inner mass of cells. In fact, the whole organ is of an appear- ance more like that represented by Fuhrmann * for Davainea than that by Janicki. This is so, at any i-ate, as regards the outer layer. The interstitial substance of the inner mass of cells forming the reticulum is represented by Janicki, but as of much greater relative extent than I have found it. Text-fig. 78. Pai'uterine organ of Inermicapsifer capensis, showing differentiation between outer and inner coat. The species that has just been described cannot, as 1 think, be referred to any of those enumerated by Janicki. It differs from /. hyracis in the form of the scolex, in the fact that the testes are arranged in two separate masses in each proglottid instead of forming a continuous row. From /. interpositus my species difi'ers also in the arrangement of the testes, and in the fact that the genital pores of /. interpositus are anterior in position and open into a well-marked cloaca, and also in the fact that the sexual products are ripe earlier in the body in /. interpositus. Nor can /. settii be confused with my species. For in the former the body is very short and consists of not more than 70 proglottids. Moreover, the excretory vessels do not approach * Rev. Zool. Suisse, iv. (1896). NEW TAPEWORMS FROM THE HYRAX. 593 SO nearly to the apex of the scolex, nor are they so coiled in that region as in my species. On the other hand, the two species agree in the posterior position of the genital pores and the separation of the testes into two masses. There may be a difference also in the lack of an excretory network in /. settii, of the existence of which Janicki is not certain. There remain certain less-known species referred by Janicki to the genus Inermicajjsifer with more or less certainty and of whose characters he gives some account. Of these, " Taenia paronai " cannot be identical with my species, since it possesses hooks; " Tcenia sjjatulcc" of von Linstow is too imperfectly described to admit of its definite inclusion in the genus Inermicapsifer. It cannot, however, be identical with my species, since the cirrus sac is apparently much larger (" Dei- Cirrusbeutel nimmt 1/7 des Querdurchmessers ein ") and the testes are scattered through the greater part of the segment. Tcenia ghondhorensis of Klaptocz is very imperfectly known, but a pit upon the scolex shows that it is not identical with my species. From /, criticics of Pagenstecher (which is perhaps identical with another species described below), the present species can be distinguished by the grouping of the testes into two masses. /. j9«^e9isfecAe?'i of Setti differs from my species by its few proglottids (not more than 80) ; otherwise it seems to pi'esent more resemblances to my species than any other form except 7. settii by virtue of the posterior position of the genital pores. To a species termed by Nassonov Anoplocephala hyracis Rud. var, hepatica, and by Janicki " Inermicapsifer spec. ? " I shall recur in considering some worms from the gall-bladder of Procavia capensis. In the meantime my own species may be thus defined and named : — Inermicapsifer capensis, sp, n. Length about 95 mm., hreadth 2 mjin.; number of proglottids 200. Scolex wider than the neck. Proglottids at end of body nearly as long as vnde. Genital pores unilateral near to posterior end of proglottids, not borne xipon a conspicuoics j^i^ojection. Testes in two separate groups, one on pore side and one on opposite side of proglottid. Vas deferens forms a netvjorh ; a large vesictda seminalis present ; cirrtis sac small, filled with slightly coiled and dilated sperin-duct. Uterus short and not persistent. Many paruterine organs, each containing several embryos. Hab. Procavia capensis. On Species of the Genus Hyracotsenia, gen. nov. Along with the numerous examples of the tapeworm which I have described here as Inermicapsifer capensis, I found in the gut of the Hyrax two complete or nearly complete examples of a tape- worm which has quite a different external appearance, and whose 594 DR. F. E. BEDDARD ON internal structure is different from that of the former species. Before examining it anatomically it was, of course, impossible to say whether or no these worms belonged to the genus referred to, and even now it is possible that they may be of the same species as one or more of those which Janicki has — provisionally, at any rate, — assigned to the genus Inermicapsifer. They are, however, Text-fig. 79. Syracotania hyracts. X 2. clearly not members of that genus, as the following account of their structure will prove. Nor can I, with any confidence, refer them to any other known genus. Hence I propose the above name, which is indicative of their habitat. I shall enquire later as to NEW TAPEWORMS PROM THE HYRAX. 595 the possibility of their identity with any other species known from the Hyrax. The two worms belong, as I think, to two different species, but are referable to the same genus without any doubt. I shall consider the anatomy of both of them together. The larger specimen (text-fig. 79) measured about 90 mm. in length by a greatest diameter of 5-6 mm. The proglottids are very short in an antero-posterior direction, but, at any rate some way back in the body, rather thick. If there is a neck present at all it is very short. The scolex is unarmed and distinctly marked off from the strobila, but not much wider than the ensuing body. The latter increases gradually in width up to the widest point, and towards the end of the body again decreases. Of the smaller specimen I cannot give such precise details, since, believing at first that both specimens were of the same species, I investigated this individual anatomically without making full notes on its external characters. It was, however, rather shorter and of less breadth, while the anterior narrower region of the body widened out to the full dimensions rather sooner than in the larger specimen. I have investigated the scolex by means of longitudinal sections only in the smaller specimen last refei-red to. The four suckers look directly upwards, their orifice being terminal in such sections. There is nothing remarkable that I could detect about their structure. They do not bulge to any extent from the sides of the scolex, and these sections show that the scolex is hardly, if indeed at all, wider than the immediately following strobila. The rudimentary rostellum is merely a hemispherical elevation lying between the suckers. There is no terminal pit of any kind and no hooks discoverable. The water vascular tubes extend into the rostellum. Of the larger specimen I only examined the scolex without destroying it. It is clear that the sti'ucture is the same in all the points mentioned above, but I am not able to report upon the water vascular tubes in this region. The structure of the body-wall (text-fig. 80) also differentiates these two species from the Iner'inicapsifer whose anatomy has been described above. The principal difference lies in the much more marked layer of transversely running fibres which bound the cortical layer internally and the medullary parenchyma externally. This layer is very much the same — I think exactly the same — in both of the two individuials of this genus which I refer later to difierent species. It is composed of very delicate fibres ; but the layer, as a whole, is rendered more conspicuous by the fact that lai-ge fibres belonging to the longitudinal layer occur between the trans- versely running fibres. The cortical parenchyma is nearly as thick as the medullary. The stout longitudinal fibres which run in the former are to be found in the greatest numbers at about the middle of the cortical layer, but they occur elsewhere. They are not massed into large bundles, but two or three are here and there closely associated. This massing of the longitudinal fibres is not obvious in the larger specimen. 596 DR. p. E. BEDBARD ON The excretory vessels consist of the usual dorsal and ventral tubes running continuously through the strobila. Anterioi'ly there is less difference in the calibre of these tubes than posteriorly. The position in relation to each other also becomes altered . In immature segmetits (where I have studied the excretory tubes) the rather smaller dorsal vessel lies obliquely above the ventral vessel to the Text-fig. 80. -(/. V .^ I- •• i \ "^t: Syracotcenia lujracis. Transverse section through a proglottid. d.v. Dorsal excretory' tube. T. Row of testes. V. Vitelline gland. v.v. Ventral excretory tube. pore side of the same. In maturer proglottids the two vessels are practically superposed, and the dorsal vessel is at times so minute as to escape observation. Both of these vessels are connected with a network of larger and smaller wEtter vascular capillaries which traverse the medullary region of the body. I have seen branches of these ending in a testis, and it becomes a matter of NEW TAPEWORMS PROM THE HTRAX. 597 great interest to enquire if this network is directly connected with the network of vasa efferentia. But I have at present no further facts to offer. The ovary Of this tapeworm is unquestionably single. It lies on the pore side of each proglottid in the neighbourhood of the two longitudinal water vessels of that side. The ovary is apparently sometimes quite anterior to the testes, and sometimes is sur- rounded laterally by them. The vitelline gland is posterior to the ovary, which is thus the most anterior of the generative organs. Text-fig. 81. d.v. V.J/ U 1- - ■ ■ ; \ X sp.d. r-s. v.g. Text-fig. 81. — Part of a transver-se section through a proglottid of Syracotania Tiyracis, to illustrate the .immature ovary. d.v. Dorsal excretory tube. o. Ovary, v.v. Ventral excretory tube. Text-fig. 82. — Transverse section of one-half of a proglottid of the same species. d.v. Dorsal excretory tube. o. Part of ovary, r.s. Receptaculum seminis. sp.d. Sperm-duct. t. Testes, v.g. Vitelline gland, v.v. Ventral excretory tube. Its exact position with reference to the two water vascular vessels is not always identical. It is more dorsal or more ventral, as the case may be, and is sometimes entirely to the pore side of the ventral tube of the excretor}'^ system, and sometimes entirely to the median side of that tube ; sometimes it extends to both sides. In any ease it lies more or less between the dorsal and ventral excretory tubes. The ovary is not solid and compact, but arranged in a series of finger-like outgrowths radiating outwards and 598 DR. F. E. BEDDARD ON dorsally. In young ovaries, such as the one figured in text- fig. 81, the riper eggs lie in straight lines connected by delicate threads with the central mass, and suggestive almost of the pseudopodia of a Rhizopod with the eggs carried along them. This radiating appearance of the ovary is retained until the organ is quite mature, when it consists of a group of sacs with, for the most part, definite walls, enclosed in which lie the ripe ova. I refer to this appearance of the ovary in considering the uterus on a later page. It seems possible that this condition of the ovary is to be compared to the testes, and that there are really several separate ovaries, which, however, are more closely ad pressed than are the testes. In any case we have instances, like Stilesia, where the single ovary consists simply of a single mass of egg- cells, a condition which is to be compared with one of the sub- divisions of the ovary in the present species. The vitelline gland in the immature segments lies exactly opposite to the ovary, the vagina dividing, as described later, into two ducts, which end respectively in the ovary and vitelline gland. It is large and conspicuous in the mature proglottid. The shell- yland is also quite conspicuous in this tapeworm. The vagina is wider and with thicker walls for a short space after its orifice on to the exterior. It then narrows and runs a veiy straight course towards the interior of the proglottid. It then becomes again wider, and opens gradually or abruptly into a dilated receptaculum seminis, which lies beside the vesicula seminalis. This region of the female duct is thin-walled. It is on a level with the ventral water-vessel. From the median end of the swollen receptaculum arise two tubes, one of them being the vitelline duct and the other the duct leading to the shell-gland and to the ovary. These two ducts are very much nariower than the receptaculum, into which they suddenly open. In immature proglottids the i-eceptaculum is rather wider than the vagina, and gradually widens towards the internally situated end, there diverging into two horns which are respectively the vitelline duct and the ovarian duct. These ducts are in these immature segments of hardly less calibre than the end of the receptaculum into which they open. These and the sperm-duct pass towards the exterioi- between the dorsal and ventral water vascular tubes. The terminal region of the vagina, i. e. that part nearest to the external orifice of the tube, has a lining which is very deeply stained by logwood, and so has the narrow region which imme- diately ensues ; the rest of the vagina is not deeply stained in this way. I noticed in the larger of the two specimens which I report upon in the present communication that the narrow region of the vagina lying between the terminal part and the portion which may be termed receptaculum was much shorter than in the other example. I am not certain, however, whether there may not be some variation in this matter from segment to segment, an ex- pansion of the lumen accounting for the different appearance. It NEW TAPEWORMS PROM THE HYRAX. 599 is certain, moreover, that the wide internal region of the vagina, before it divides into vitelline duct and ovarian duct, is susceptible to variation ; for in some cases it was an abruptly formed spherical sac, at other times merely a wider tube than the immediately pre- ceding region of the vagina. I mention later that the end of the receptaculum contains ripe ova in the mature proglottids. I have never seen spermatozoa therein. Uterus. — In both of the two examples of this species the last few segments became somewhat shorter from side to side than those in front, and were also longer in theantero-posteriordii^ection. One would naturally associate these changes in the facies of the pro- glottids with complete maturity and the existence in those segments of embryos. As I find in these segments completely ripe eggs quite detached from the ovary and associated together in small masses I shall assume that the anatomical structure of these proglottids is that of complete maturity. In this case the present species differs from those which we have been considering by the entire absence of numerous paruterine organs like those of Inermi- capsifer, etc. Even were these terminal segments not so fully mature as I presume them to be, there would be, I should imagine, at least some preparation for the formation of the paruterine organs. But there is none. The ripe eggs were massed into more or less spherical groups surrounded by a membrane. These were not to be distinguished from the lobes of the ripe ovary, and I imagine that the}^ were merely the persistent ovary. In addition to these masses of eggs, the end of the vagina, i. e. the dilated portion which I have termed receptaculum seminis, was found in many cases to be full or nearly full of ripe eggs unaccompanied by any interstitial cells. This was not only the case with a receptaculum which was swollen into a spherical contour at its base, but also in cases where the re- ceptaculum ended merely as a slightly wider tubular sac. I have no reason whatever to doubt these facts, as the receptaculum is quite easy of identification. I am therefore disposed to think that there is no uterus as a distinct and separate structure, but that the eggs are partly voided into the receptaculum and partly remain in situ awaiting the loosening and perhaps disintegration of the proglottids. It may be that one of the rounded sacs which I regard as an ovary is really a uterus. Of this I cannot be positive, especially in view of the very few completely mature proglottids in both specimens. That both specimens were identical looks as if the conditions above described are to be regarded as normal. The testes extend through a good part of each proglottid and as seen in sagittal sections, thei'e are five or six rows of them laterally, though not so many in the median region. In transverse sections they are seen to extend from edge to edge of each segment being nowhere interrupted save where they meet with the female reproductive organ. This row is mostly one deep, but in places two or even three deep. I counted from 40 to 50 or so of separate testes 600 DR, F. E. BEDDARD ON in a single complete transverse series. There are thus altogether two or three hundred of these gonads to each mature proglottid. In more immature segments the testes did not appear to extend to the pore side of the longitudinal water-vessels but to stop short before quite reaching the median side of those vessels. In a mature segment I found that 44 out of the 50 sections which displayed it in its entirety were occupied by the testes, which thus fill up most of the segment, though the proportions were not always exactly as stated in the above instance. In proglottids from the other example of this species 3 sections without testes were fol- lowed by 1 5 sections showing testes, and these again by 5 without testes, and thereafter 14 with testes. It is therefore obviously the case that the testes occupy a great deal of the segments. It will be observed that there is no grouping of the testes into two masses such as I have described in Inermicapsifer capensis. They lie mainly behind the ovary and vitelline gland, and in some proglottids the ovary lay rather more distinctly in front of the male gonads. The testes are more or less spherical or egg-shaped, and when ripe are seen to be surrounded by a layer of spermatozoa, which lie therefore, as I take it, in a cavity surrounding the testis, a coelomic cavity. I never found the testes of this tape- worm to be pear-shaped, like those of Inermicapsifer. Further- more, the testes, all of them, lie dispersed in quite unaltered parenchyma. As is veiy generally the case among the members of this group, the testes were mature much more anteriorly in the body to the ovary. It is, indeed, a striking feature of the pi^esent species, and one in which it contrasts, for example, with the species of Inermicapsifer that has just been desci'ibed, that the mature testes occupy so many segments of the body, while the mature ovaries are so exceedingly limited in the number of seg- ments in which they are found. The efferent tubules which collect the sperm form a very definite network (see text-fig. 83), which is copious and formed often of unequally sized vessels. A similar network has been described in othei- tapeworms, for example in Chapm,annia* . The vas deferens of this species (text-fig. 82, p. 597) is quite dif- ferent from that of the species which we have already considered. The reticulate efiei-ent ducts finally find their way into a large sac, which in the mature segments is stufied with sperm, and which lies in the female generative mass alongside of the receptaculum ovorum. This large vesicula seminalis is flask-shaped, and there- fore gradually narrows and emerges from the female generative mass as it passes towards the genital orifice. It is impossible to draw a hard-and-fast line between the vesicula seminalis and the sperm-duct proper with which it is continiious, for the gradual diminution in calibre of the entire tube forbids such a delimitation The tube pursues a winding course, narrowing gradually and but slightly ; it never forms an actual coil like the sperm-duct of so * See Puhrmanii, Swedish Zool. Bxped. Egypt, 1909, pt. iii. NEW TAPEAVORMS FROM THE HYRAX. 601 many other—indeed the majority of — Tetracotylea, but is at most once or twice bent upon itself. It becomes very narrow for a Text-fig. 83. - • . 1 .§ Portion of medullaiy region of a proglottid of SyracotcBnia liyracis, to illustrate network formed by vasa eft'erentia {v.d.). T. Testis, tr. Transverse muscular layer, v.v. Ventral excretor3' tube. Pboc. Zool. See— 1912, No. XL. 40 602 DR. F. E. BEDDARD ON short distance in front of its opening into the cirrus sac. The course of the sperm-duct is roughly parallel to that of the vagina, with which it might be sometimes confused in those cases where the vagina has not so abrupt a transition into the receptaculum. Along the course of the sperm-duct, which in ripe segments is gorged with sperm almost throughout, there lie masses of what appear to be prostatic cells, similar in the fact of their existence to those of Inertnicapsifer, but different in appearance. In the present species these cells are of a clear, almost hyaline, ap- pearance, which is possibly due to the state of their activity at the time when the worm was killed. In Inermicapsifer capensis and in the species which I originally named Thysanotoinia gain- hiana, the prostatic cells were darkly staining and granular. Nevertheless, they appear to be equivalent structures in the two tapeworms. In sections Avhere the sperm-duct appears in trans- verse section, these cells present the appearance of a winding duct cut transversely. This appearance is due to the clear cells clustered round the actual sperm-duct, which, as already said, is narrow of calibre close to where it opens into the cirrus sac, and thus not obvious in such sections. There can be no mistake, however, in transverse sections of proglottids, where the course of the sperm-duct is easily to be followed owing to its being filled with sperm. In the second and larger individual there are certain definite difierences in the form of the sperm-duct. The tube has no such great dilatation into a vesicula seminalis, and it is very much more coiled as it approaches the cirrus sac. It has, in fact, the large and close coil which is so typical of tapeworms. There is certainly nothing of the kind in the other individual. The clear cells already spoken of form a complete layer one cell thick round the mass of sperm in the sperm-duct, and are therefore, I take it, simply the epithelial wall of the sperm-duct. As the sperm-duct was in parts full of sperm, this difference cannot be owing, I believe, to the different stages of the maturity of the proglottids in this tapeworm as compared with those already described. It must, I think, be a specific difference, with which also, it will be observed, go differences in the position of the ovary and vitelline gland. The cirrus sac of this worm is not at all large as in the allied forms comprised in the genera Inermicapsifer and ZschokJceella. It can be seen in sagittal sections to lie straight in front of the vagina close to the external aperture, and I have not noticed any genital cloaca. There is certainly nothing of any size, and in one section the penis was seen to protrude on to the exterior directly from the cirrus sac without any intermediate and common chamber. The cirrus sac is oval in form and surrounded, as is usual, by a strong layer of muscles. I could not see any indications of a flask shape such as is so common in tapeworms. In the in- terior of the sac are the usual nuclei belonging, it is to be presumed, NEW TAPEWORMS FROM THE HYRAX. 603 to delicate muscles which retract the cirrus. The latter was relatively wide and short and the sperm-duct within the cirrus sac not coiled. The protruded male copulatory organ reminded one rather of the penis of Anoplotcenia * than of a cirrus, for it was wider at the free end than just within the cirrus sac. It is evident that this genus presents many resemblances to the genera ZschoJckeella and Ineriniccq^sifer f. It agrees with those genera in the following assemblage of characters : — The head is unarmed and the excretory tubules form a plexus, or at least a coil, at the very extremity of the rostellum, as in the species which has just been described ; the segments are narrow and the genital pores are unilateral. The excretory tubules furthermore form a plexus within the medullary parenchyma throughout the body. The ovary lies to the pore side of the segments and is distinctly not a double organ ; the vagina dilates into a wide receptaculum seminis. The cirrus sac, moreover, is small as contrasted with that of many other tapeworms. On the other hand, there are certain characters which argue against this placing of the worm whose anatomy has just been described. These are as follows : — The sperm-duct in our species is short and almost immediately dilates into a large and long seminal vesicle, a state of affairs which is not met with in the species of Inermicapsifer known at present. Finally, the network formed by the vasa efferentia is a feature hitherto unknown in the genus ZschoJckeella, though it occurs in Inermicapsifer. Inasmuch as a reticular disposition of the vasa efferentia is not necessarily diagnostic of a given genus as far as we know, for it occurs in Chapmannia lapica and Hymenolepis reticulata and not in other (at any late in some other) species of that genus, this fact alone would not perhaps necessitate the removal of the present species from the genus Zschohkeella to which other important characters would appear to assign it. But there is a negative character which may be of very great importance. In neither of the two individuals which I have studied was there the least trace of the formation of the characteristic " egg-capsules," which I prefer for reasons already given to call paruterine organs. In all the species of Zschokkeella and Inermicapsifer examined from this point of view, the formation of these capsules began perhaps rather far back in the body, but still a long way before the actual termination. Now both examples of the present genus in my possession ended posteriorly in a few segments which were rather longer than those preceding them and at the same time rather narrower, suggesting, in fact, the end of the body. They were, moreover, thicker than unripe segments. If this be not the completely mature end of the body, the worm would be very exceptional in the deferring of the egg-reservoirs to a point so very far behind the scolex. Besides, two specimens selected at * See Beddard, P. Z. S. 1911, p. 1015, text-fig. 215. t For the generic distinctions see below, p. 607. 40* 604 DR. F. E. BEDDARD ON random from the same host would hardly be likely to prove both abnormal in any way. Finally, we have in these segments rounded sacs with ripe eggs, though it must be admitted that these were eggs and not embryos. Some riper proglottids may, however, be missing ; but even then the commencement of the paruterine organs would surely be visible. It is therefore, as I think, impossible to include these worms in either of the genera with which I have just compared them. Of the remaining Anoplocephalidse (to which family I think that these worms must be referred) there are only the genera belonging to the subfamily Anoplocephalince. Of these Oittotcenia and Moniezia need not be considered, since their generative apparatus is double in each segment. Of the remaining genera none agree with the two worms under consideration in all of the following points, viz., uterus at most inconspicuous, cirrus sac small, genital pores unilateral, ovary to pore side of proglottids, genital ducts pass between excretory vessels, testes posterior. I believe, therefore, that they must be referred to a new genus. This new genus may be thus defined : — Hyracotaenia. Scolex unarmed, with four unarmed suckers. Proglottids wide and very short, a little longer at extreme end of body, hut always much wider than long. Genital pores itnilaleral, not borne upon papillce. Cortical parenchyma thick, separated from medullary by a thin layer of circular fibres. Water vascular tiobes four, dorsal and ventral, the latter larger, connected by a network of capil- laines. Testes nuvnerous, dorsal in position, lying behind and to sides of ovary ; vasa efferentia forming a network ; sperim-duct wide and. sinuous or coiled ; cirrus sac small ; a short blunt wide penis protrusible. Ovary near ivater-tubes of pore side, single, in front of vitelline gland; dilated receptaculum seminis and very narrow vagioia. Uterus small and sac-like ; paruterine organs absent. Hab. Procavia capensis. It is not possible for me to distinguish definitely at present between generic and specific characters. The above embody characters usually considered in generic definitions. The two species may be, for the present at least, defined as follows : — (1) Hyracot£enia procavise, sp, n. Length about 90 mm. ; greatest diameter 5-6 mm. Body attains its greatest width about 25 mm. from anterior end. Testes very numerous. Sperm-duct rather dilated posteriorly, much coiled anteriorly. Ovary ventral, on outer side only or both sides of ventral vessel ; vitelline glands dorsal. Vagi^ia not greatly dilated posteriorly. NEW TAPEWORMS FKOM THE HTRAX. 605 (2) Hyracotaenia hyracis, sp. n. Length about 70 ifrim. ; greatest diameter 4'5 m-m. Body attains its greatest loidth about 6 mm. from anterior end. Testes less numerous. Sperm-duct much dilated posteriorly, sinuous but not coiled anteriorly. Ovary more dorsal, to median side of ivater- vessels ; vitelline glands ventral. Vagina usually mu,ch dilated posteriorly. We now come to the consideration of the question of the possible identity of either or both of the above species with any of those enumerated from the Hyrax by Janicki. The only species of that series that can be considered (if, that is to say, there are really no paruterine organs in the forms described by myself) are Tcenia {Anoplocephala ?) gondokorensis of Klaptocz *, Tcenia (Anoplocephala) spatula of v.Linstowf, and Anoplocephala hyracis, YSbV.hepaticaoi ISTassonow J, termed Inermicapsifer spec. ? by Janicki. Of these Klaptocz's species has a small scolex like the species described b}"- me, but also an apical depression (? a rudimentary rostellum) which my species have not. Furthermore, the pro- glottids appear to be much smaller. In the species of v. Linstow we find too great a breadth, and the cirrus sac is too large for comparison with my species. The shape and proportions of Nassonow's species are like mine, but the scolex has a conical process. In all these species the details are insufficient. There are thus in African animals — chiefly in Mammals (mostly in the Hyrax (Frocavia), but also in E-odents and Lemurs), but extending to JBirds {Numida ptilorhyncha) — a group of worms which show at least specific difierences, but all of which have the following characters in common, viz., head unarmed and no neck, proglottids wider than long and as a rule very much so, genital pores unilateral, cirrus sac not very large, testes numerous, ovary not double. To these characters may possibly be added, if we exclude the species described above as Hyracotcenia spp., or are led by fui-ther material to interpret their anatomy difierently, the formation of numerous paruterine organs — or egg-capsules as they have been termed by others. These characters (excluding the fate of the uterus) seem to me to necessitate the inclusion of this group of worms in the family Anoplocephalidse. The various species which agree in the foregoing characters cannot, however, on these alone be massed into one and the same genus without further consideration. If we subtract from the assemblage the species which I have described as Hyracotcenia procavice, and which, as I think, must in any case be withdrawn from the group, the reasons for uniting the rest under a single generic name become more striking. For in this case all of the species possess paruterine organs of the same kind, unless, indeed, * S.B. Wien. Ak. 1906. t Jen. Zeitschr. Naturw. xxxv. 1901. X Arb. Zool. Lab. Univ. Warschau, 1897. 606 ON NEW TAPEWORMS PROM THE HYRAX. Janicki be right in inferring that Zschokkeella really differs, a point which I have already gone into above (see p. 590). The description of a second species of Zschokkeella* from a Cerco- pithecus does not throw any further light upon this particular matter. If it were not for the fact that Davainea seems in some of its species to possess paruterine organs of the same type, the African worms referred to might well be regarded as all congeneric, in which case, of course, Zschokkeella would have to be the name. Janicki appears to me to be rather hard put to it to separate his hiermicapsifer from Zschokkeella. The differences are certainly small. As already stated, he relies upon supposed differences in the paruterine organs of which I am disposed to doubt the existence. He also mentions the thickness of the muscular walls in Zschokkeella as compared with Inermicapsifer, and a few other points which seem to me to be of minor importance and not even collectively as of generic rank. Janicki's comparisons are based chiefly upon his own account of Inermicapsifer hyracis, which was the only species investigated by him in a detailed fashion. I do not think that a further examination of other species referred to by Janicki will necessarily prove the identity of Inermicapsifer and Zschokkeella throughout. I would point out that my own account in the present paper of /. capensis shows some differences between that species and /. hyracis. These differences are mainly the posterior position of the genital pore, the existence of a vesicula seminalis, the complete separation of two groups of testes, and the presence of a rete mirabile along the course of the sperm-duct. Finally, the uterus is much more rudimentary in /. capensis than in /. hyracis. In some of these characters it would appear that /. settii agrees with my species and differs from /. hyracis. A further examination of these species may show that they agree in other characters not referred to by Janicki in his r6sum6 of these forms. I would reserve the generic name Inermicapsifer for these forms and refer " Inermicapsifer " hyracis to Zschokkeella. There now remains my genus Thysanotcenia. Of that genus I have described two species which show many differences of structure. Thysano- tcenia ga'mbiana is, as I now think, undoubtedly to be referred to Zschokkeella, with which it agrees in all points, if we may assume that the paruterine bodies are identical in the two. On the other hand, it will be, as I think, advisable to retain the name Thysano- tcenia for the second species of the genus [T. lemtoris), which differs mainly in the following points : — There is no plexus of excretory tubes and the ventral vessel is very large, the dorsal being appa- rently absent in mature segments ; the ventral vessels are connected in each segment by the usual transverse trunks; the receptaculum seminis is quite different from that of the other forms ; the uterus * Z. remota, see v. Linstow, Zeitschr. wiss. Zool. Ixxxii. 1905. ON THE AUSTRALIAN LUNG-PISH. 607 is more rudimentary than in Zschokkeella. I should distinguish the various genera thus : — I. Excretory system forms a network in each segment. Recep- taculum seminislong and forming end of vagina. Cirrus sac small. A. Genital pores median on edge of segment. Testes forming a continuous row. No vesicula seminalis. Uterus well developed at first Zschokkeella. B. Genital pores posterior on edge of segment. Testes in two groups. Large vesicula seminalis present. Seminal ducts form a network. Uterus never well developed. Inermicapsifer. II. No excretory network. Receptaculum short and globular along the course of vagina. Cirrus sac rather large. A, Genital pores on conspicuous papilla. Testes forming continuous row. Uterus never well developed. ThysanotcBnia. This arrangement is naturally only tentative, since we are at present in need of more information concerning the majority of the species already known from the Hyrax and enumerated by Janicki in the paper which has been so often referred to. There are also points in the structure of the species referred to the genus Zschokkeella which demand further investigation. 32. Additional Notes on the Living Specimens o£ the Australian Lung-fish (^Ceratodus for steri) in the Collection o£ the Zoological Society of London. By Bashford Dean.* [Received February 9, 1912 : Read April 2, 1912.] (Text-figures 84 & 85.) Index. Page Development and Coloration 608 Regeneration of injured portion 611 The two specimens of the Australian Lung-fish in the Zoolo- gical Society's collection have been living under unchanged conditions since 1898, i. e. about fourteen years. In this time they have been observed repeatedly by zoologists, whose interest in these important and rare batrachian-like fishes has led them, in several instances, to publish their notes in detail. There is still, however, much to learn about the habits of these fishes, and it is to be hoped that the opportunity will be seized generally to observe the present specimens, especially * Communicated by the Seceetaet. 608 bK. BASHFORD DEAN ON since it is fair to assume that they are living under fairly normal conditions. In this connection it may be mentioned that the fishes have been subjected to no changes in their aquarial habitat ; indeed, they have remained practically undis- turbed for over a decade. Among the notes dealing with these specimens are those of the present writer, published six years ago in the ' Proceedings ' of the Zoological Society of London (1906, vol. i. pp. 168-178), a con- tribution which gave also data about Geratodus summarized from earlier literature. In this paper details were recoi'ded regarding the movements of the fishes, their mode of breathing, both with gills and lungs, their manner of feeding, their nocturnal activity, and in general their salamander-like habits. The writer had again the opportunity of observing these specimens during June 1911 ; his supplemental notes are as follows : — Colour. — One of the fishes, the larger one, remains notably darker than the other. This distinction in colour, therefore, is neither abnormal nor seasonal. Nor can there be vast adaptive colour changes in Geratodus, for the I'eaction to similar sur- roundings would then be the same in the two fishes. Is the darker specimen a male ? So far as could be learned, the fishes have shown no evidence of sexual activity. The colours have lemained constant, and there have been no signs of the brilliant tones noted by Schmeltz (1876, J. Mus. Godeffr., vol. viii. p. 138). According to his account, the ventral side of Geratodus is of a deep orange-ied, and several scales on the sides are margined with red ; nothing is aaid, however, of the relation of these colours to the season. From the characters of the present specimens, and by analogy with Amia, we can safely conclude that the tones of orange and red appear only at the time of spawning. In Amia, aquarium-kept fishes show no bright colours, but under native conditions the male fish develops wonderful brilliancy ; the spot at the base of the tail is conspicuous, red scale-margins appear, and the hinder abdomen glows with tones of orange. It was noted (1911) that the paired fins were margined with a narrow white band. Could this have been an indication of a breeding colour ? No coloration of this kind was seen on the edges of the unpaired fins. Size. — At present the darker specimen measures 33^ inches, the lighter 29^ inches, having grown but a very few inches during the past seven years. They have i-eached, accordingly, nearly their greatest length, Macleay leading us to infer that 3 feet is about their maximum (Cat. Austr. Fishes, p. 284). Exception- ally, however, a specimen may measure 45 inches, such a case having been cited by O'Connor (1897). Age. — -The present specimens give us an idea of the age to which Geratodus may attain. We have in the first place data that they have grown, broadly sjieaking, at the rate of three quai'teis of an THE AUSTRALIAN LUNG- FISH. 609 inch annually for the last thirteen years. At this rate the present fishes are over forty years old, and a fish of the record size (45 inches) would be over sixty. But this assumes that the rate of growth in length is approximately uniform in fishes of difi"erent ages. In point of fact it is known that fishes in all groups grow quickly when young, and slowly, if at all, when old : it is also known that under favourable conditions a fish may grow with far greater rapidity. In the case of Ceratodus the rate of growth of the young is as follows : — Length of specimen. -^ff^- inches 8 Af Viofr-Viino- -1 Notes from stages t\ At hatching. | reared in balanced 12 1 month. > aquaria in Gaj'ndah, 16 -,% ^, Queensland, by Mr. y9^ 1| month. J Thomas Illidge.) ii 2 months. I 21- months. II 6 months. 11 7 months. 2y\ 71 months. 21 8 months. In stages lately hatched the growth is seen to be rapid, its rate suggesting that a specimen 12 inches in length might be not older than a year. The rate, however, changes notably when the young fish no longer subsists on its yolk. In fact, for a period of about two months it actually decreases in size, a state of aflairs, however, probably abnormal and due to the lack of proper food in the aquarium. Young an inch long are nine weeks old; young two inches long are over seven months old. At the end of the first year the young Geratodihs measures, we infer, about five inches, a rate of growth which would be not unlike that in young Amia, gar-pikes, or in a number of teleostean fishes. If the analogy with known ganoids continues, a two- years' Ceratodihs would measure 8-9 inches; and a specimen 25 inches in length, approximately the size of one of the present fishes when it appeared in London, is estimated to be from eight to ten years old; this added to the thirteen years of their living in the Society's aquarium, makes the total age of the present specimens between, roundly, twenty and twenty-five years. Older and larger specimens it is fair to credit with great age, probably fifty years. Breathing. — The aquatic respiration of Ceratodus varies con- siderably according to the temperature of the water, but its range has not been observed. In September (1904), on a cool day, with water temperature not far from 65° Fahrenheit, the respi- ratory movements were " slow and regular ; the opercular cavity filled and emptied about twelve times a minute." In late June, when the water temperature was nearly 75°, the successive 610 DR. BASHPORD DEAN ON movements of the gills varied from twenty-two to thirty-one a minute. Both fish had long been quiet ; if their movements had been active there is no doubt that this rate would have been notably exceeded. Rhythmic movements in breathing are well shown in the opercular membrane. In the early stage of breathing the cheek in the subopercular region is seen to dilate slowly and strongly ; this dilated region is then passed (rather slowly) backward, and Text-fig. 84. Operculai' movement in breathing of Ceratodiisforsteri. The crest of the xindulatory wave is indicated by the asterisk. its enclosed water is discharged (text-fig. 84). During this process the dilated part becomes more and more conspicuous until the final discharge, and at that moment the free rim of the gill-opening is thrown outward and drawn forward, exposing the lighter coloured hinder border of the gill-slit. The free border up to that time has been closely apposed to the head. It was earlier noted that Ceratodus is a " nostril breather." THE AUSTRALIAN LUNG-PISH. 611 The mouth itself showed no noticeable movement of opening or closing ; it was indeed hardly open, the gape scarcely more than 1 inch. During the later observations, the mouth was seen slightly to open and close; its maximum gape noted (measured close to the glass) was | inch, its minimum | inch. The opening of the mouth was here doubtless correlated with the higher temperature of the water and the more rapid respiratory movements. In general, however, the mouth margin was almost motionless, the fish breathing through the nostrils. In the matter of breathing air at the surface, Ceratodus shows greater variability than earlier noted. On one occasion over seventy minutes elapsed without either fish coming to the surface. Feeding. — Little was added to the former notes. The only detail suggests that minute food, in the form of algse together with vegetable debris, forms a part of the normal diet. It was observed that the fishes would " nose" about in corners and suck in this finer material. In the process little pebbles would some- times be taken in and retained for a few moments, then rejected — the process several times repeated, in a fashion which suggested that the stones thus mouthed yielded food sought for by the fish. It was noted in this connection that the stones in the aquarium were in many cases well coloured with algse. Text-fig. 85. Pectoral fin of Ceratodus forsteri, showing regenerating margin. Movements,— The writer's preceding paper gave a number of figures showing characteristic movements of Ceratodus.^ The only addition to this series would be a variant of the fig. H there given; the right hand pectoral remained in its resting position (as in fig. 14) ; the left, however, was suddenly twitched up over the back several times (as in fig. 11), but in this case brushed close to the body, giving the observer the amusing impression that the fish was thoughtfully rubbing its head. The habit was curiously unlike that of a fish ; it suggested rather the movement of a tetrapod ; and a very similar movement is known in urodeles. Regeneration.— Thel&vg&v (dark coloured) specimen had suflered an injury to the left ventral fin; a portion of it near the tip 612 MR. c. H. o'donoghue on the had been lost and the margin was regenerating. Text-fig. 85 illus- trates how far this process had extended. It will be seen that the restitutive prohferations were most active along the free distal margin of the fin. Here several (four) eminences were present, each suggesting the pointed tip of the fin ; there can, however, be no doubt as to which of these is the terminal one, since the skeleton of the fin can be followed into the lowest of these lappets. The case is evidently akin to one known to teratologists, for when certain areas in injured limbs of batrachians are stimulated, there appears Polydactyly or polypody. It may therefore be worthy of record that a similar condition occurs in the lung-fish Ceratodus. 33, The Circulatory System o£ the Common Grass-Snake (Tropido7iotus natrix). By Chas. H. O'Donoghue, B.Sc, F.Z.S., Assistant to the Jodrell Professor o£ Zoology, University College, London. [Received April 1, 1912 : Read April 23, 1912.] (Plates LXX.-LXXII. and Text-figures 86-91.) Index. Page I. Introduction 612 II. The Heart. A. Development 614 B. Adult Form 617 III. The Arterial System. A. Development 619 B. Adult Form 621 (a) Anterior Vessels 621 (6) Posterior Vessels 623 IV. The Venous System. A. Development 626 B. Adult Form 630 {a) Anterior Vessels 630 (6) Posterior Vessels 631 V. The Vessels of the Head. A. Arteries 635 B. Veins 640 VI. List of References 643 VII. Explanation of Plates 645 I. Introduction. Our knowledge of the circulatory system in snakes is far from exhaustive ; indeed, we have only a complete account of the vessels in the Python by Hopkinson and Pancoat (25), and a later and a more full one by Jaquart (26), and in Pelophilus madagascariensis hy Gadoyv * . Although Tropidonotus natrix is * This is incorporated in the account given by Hoffmann (23). 00 P Z S. 1912. PI LXX fiJ- RA. BlA 4 ^' BV // s ^^f // pr.v // ^ 1 1 . HPV If A ' i> . n' > " PV__— ^l" > CIRCULATORY SYSTEM OF THE GRASS SNAKE. p. Z. S. 1912, PI. LXXI. .15.A. i.C.M. Xcr.A. Chas.H. O'Donoglme.del. Bale Is. Danielsson, L''.'^ imp CIRCULATORY SYSTEM OF THE GRASS SNAKE. P, Z. S. 1912, PI. LXXII. /on.Sn. CJL. Chas.H O'Donoglrue, del. CIRCULATORY SYSTEM OF THE GRASS SNAI-IE CJR. Bale i DaTAelsson, L*''* imp. CIRCLTLATORY SYSTEM OF THE GRASS-SNAKE. 613 the snake most commonly dissected in the laboratories of this country, no description of its vascular system has appeared since the anatomical account of the blood-vessels in snakes written by Schlemm (35) in 1827, which was based largely on Coluber (i. e. Tropidonottos) natrix and Trigonocephalus mutus. This account by Schlemm^ although excellent in many respects, is by no means complete, and, owing to the overlooking of the cerebi-al carotid artery, the remaining arteries of the head are misintei- preted. A great deal of work has been done, however, on different parts of the circulatory system of this animal by various authors. We are indebted to Rathke for a valuable account of its development (30) and also of the arteries of the head and neck (31) ; the last is the best general account of these vessels in snakes that has been written as yet. Hochstetter has dealt with the development of the posterior veins (20) and of the blood- vessels in general (22), Grosser and Brezina (19) with the development of the veins in the head and neck, and Bruner (12) with the veins and sinuses in the head of the adult. In addition to these works bearing directly on T. natrix, Beddard (1-6) has added considerably to our knowledge of the blood-vessels of other snakes, and the intracranial circulation has been dealt with in the vertebrate series in general by De Yriese (14) and Hofmann (24) and in reptiles by Dendy (13). The blood-vessels of the Grass-Snake were investigated by means of the dissection of a number of injected specimens. The injection fluid used for the main vessels was that recommended by Kingsley (28), i. e., Corn starch and 2 per cent, chloral hydrate (each) 400 vols. 95 per cent, alcohol 100 vols, and Colour and glycerine (equal parts) 100 vols. For the finer vessels a gelatine mixture advised by Tandler (36) was used, i. e., 5 gms. of gelatine in 100 c.c. of distilled water coloured with Berlin blue or carmine. 5-6 gms. of potassium iodide added slowly while warming gently. These are two very good mixtures, as the first will keep almost indefinitely, and the second, with the addition of a few crystals of thymol, will keep in a stoppered bottle for months, and, in addition to being useable when almost cold, will withstand acids. By mixing a quantity of the gelatine mass with about one-third or less of its volume of the solid residue that settles to the bottom of the starch mixture, an extremely useful general in- jection mass is obtained. The latter mass, which flows very readily if only slightly warmed, and sets firmly and fairly quickly in 70 per cent, alcohol or in 4-5 per cent, formalin, was the one most frequently employed in making the preparations for this investigation. 614 MR, c. H. o'donoghue on the For the sake of clearness in description, the account of the blood-vessels of the head is not included in the general description of the vascular system of the whole animal, but is dealt with separately later. A brief account of the development of the heart and of the arterial and venous systems has been introduced in order to throw some light on the condition that obtains in the adult. Two or three features of general interest in connection with the elongation of the body and the loss of limbs in the Ophidia are clearly brought out in dealing with the vascular system of the Grass-Snake. The first is the marked asymmetry of the viscera and their blood-supply ; not only are the organs of the right side anterior to those of the left, but they are also considerably larger. Thus the right ovary, supra-renal body, and the kidney are in front of and larger than the corresponding organs on the left, and, as is well known, in the case of the lungs the left one is entirely suppressed. Secondly, the tendency to form longitudinal systems of vessels, common to all Ophidia, as Beddard (1) pointed out, is well marked. The various arteries supplying the intestine and the fat-bodies are in each instance indirectly connected into one long system. The ovarian artery forms a longitudinal trunk along the corresponding supra-renal body. Among the veins also we find that the hepatic portal vein runs from one end of the intestine to the other, and that each oviduct possesses a sinus running beside it for the greater part of its length. This oviducal sinus is very conspicuous in T. natrix, although it does not appear to have been described previously in other snakes. The liver, too, is greatly elongated, and the post-caval vein and the hepatic portal vein pass along its opposite faces from one end to the other. Lastly, the blood-vessels of the adult, with the exception of a small pair of veins in the cloacal region, which may represent the pelvic veins of Lacertilia, give no indication of their derivation from those of a limb-beai-ing ancestor. In conclusion I should like to express my sincere thanks to Professor J. P. Hill of this College for the kindly assistance and advice he has given me throughoiit the work. II. The Heart. (PI. LXX.) (A) Development. The development of the heart has been very fully dealt with by Rathke (30), and as this account difiers but slightly from that of Lacerta given by Greil (18) and Hochstetter (20), it does not appear necessary to give more than a brief outline here. The primitive heart is in the form of a simple tube stretching in an antero-posterior direction in the region of the gill-slits. Its posterior end is formed by the union of the two omphalo- CIRCULATORY SYSTEM OF THE GRASS-SNAKE, 615 mesenteric veins, and the anterioi^ is continued as the short common stem (truncus arteriosus) of the first pair of branchial arches. It soon bends towards the right, and as the two ends remain in approximately the same position, while the tube itself grows longer, it is forced to take on a curved form. In this twisted condition three portions can be distinguished : first, a posterior part running from the union of the omphalo-mesenteric veins, close to which now open the paired ductus cuvieri, ventrally and towards the left ; secondly, a median part situated ventrally and running obliquely from left to right ; and thirdly, an anterior part running from the right to the median line, where it bends sharply dorsally before passing over into the truncus arteriosus, from which, by this time, two pairs of branchial arches are given off. Grooves appear on the posterior part of the tube which indicate the divisions between sinus venosus and atrium and between atrium and ventricle. Into the sinus venosus now open the paired umbilical veins. The middle part, afterwards to become the ventricle, becomes dilated ventrally, and as the anterior part also dilates, the two pai-ts are separated by a deep furrow. According to Hochstetter (20) and Langer (29) this anterior part is homologous with the bulbus cordis of the Batrachia. The ventricle expands still further and moves caudally, so that the atrium, which has also become dilated, comes to lie close to the bulbus cordis. The further dilatation of the atrium takes place cranially and towards the left, causing it to take up a position on the left side of the bulbus. The constriction between the atrium and ventricle, corresponding to the auricular canal, becomes more marked as these two structures swell out. At this stage, too, the sinus venosus is sharply constricted off from the atrium and the truncus arteriosus gives ofi" the six pairs of branchial arches. Rathke erroneously described only five branchial arches, but this will be referred to again later. After this the ventricle gradually assumes its adult shape The base of the bulbus cordis, originally joining the ventricle on the left, moves into an almost mid -ventral position. Spirally twisted grooves appear between the branchial arches, now reduced to three in number, and extend downwards over the bulbus. The atrium now gives off another dilatation, but this time to the right, and consequently the truncus arteriosus and bulbus cordis lie in a deep groove between the outgrowths of the atrium. These two dilatations are the definitive auricles, and already the sinus venosus, which lies in the atrio-ventricular sulcus on the dorsal side of the heart, opens into the one on the right. The groove between the bulbus and the ventricle gradually disappears, and ultimately the proximal part of the bulbus becomes incorporated with the latter, while its distal portion becomes assimilated to the truncus arteriosus. The spiral grooves finally extend over the whole of the so-formed truncus, and they indicate its internal division into three parts by the 616 MB. C. H. O DONOGHUE ON THE backward growth of two septa. One, the septum aortico- pulnionale, arises from the edge of the pulmonary artery, and as it grows it divides the truneus cavity into two tubes, an aortic and a pulmonary. The other, the septum aorticum, arises between the two aortic arches, and so subdivides the aortic cavity of the truneus into two, a right and a left. Of the three tubes formed in this way, one lies to the left, ventrally, and leads to the pulmonary arch ; another lies to the right, ventrally, and leads to the left aortic arch ; and the third lies dorsally and leads to the right aortic arch. Text-fig. 86. Ventral view of the heart and adjoining vessels. Az. Azygos vein. C.A. Left common carotid artery. L.A. Left auricle. L.C.A. Left coronary artery. L.J. Left common jugular vein. L.S. Left systemic arch. P.A. Pulmonary artery. P.C. Primary carotid. P.V. Pulmonary vein. P.V.C. Post-caval vein. R.A. Right auricle. R.C.A. Right coronary artery. R.J. Right common jugular vein. R.S. Right systemic arch. Th.A. Thyroid artery. V. Ventricle. V.A. Vertebral artery. Lastly, the sinus venosus also assumes its definitive form. After the disappearance of the umbilical and omphalo-mesenteric veins, it has opening into it, on the right, the right ductus Ouvieri and the post-caval, while on the left it has the left ductus Ouvieri. Thus we have practically the adult condition of the CIRCULATORY SYSTEM OF THE GRASS-SNAKE. 617 heart, which, however, in the course of the further development moves caudally, and ultimately comes to lie a long way from its primitive position. Text-fig. 87. PV.C. Dorsal view of the heart and adjoining vessels. C.V. Coronarj' vein. L.P.C. Left pre-caval vein (left common jugular). R.P.C. Right pre-caval vein (right common jugular). S.V. Major part of sinus venosus. S.V.I. Minor part of sinus venosus. Other letters as in text-fig. 86. (B) Aclidt Form. Tropidonotus in common with all the reptiles, except the Crocodilia, possesses a three-chambered heart. This is situated a considerable distance behind the head and slightly towards the right. It is eiiclosed in a pericardium in which it lies freely, not being attached to it by a gubernaculum cordis as is the case in the heart of the Lacertilia and Crocodilia, Beddard (2) has pointed out that althovigh a gubernaculum cordis is generally absent in snakes, it is not completely so, as a homologous structure occvirs in some species*. The pericardium on the right side lies * E. g. Coronella getida, Coelopeltis monspesstdana, and Ophiopliagns lungarus, Beddard {loc. cit.). Proc. Zool. See— 1912, No. XLI. 41 618 MR. c. H. o'donoghue on the against the body-wall, while on the left side it is separated from it by the intervention of the CBSophagus. In conjunction with the elongated form of the body, we find that the heart also is long and narrow. The Sinus Venosus is situated on the dorsal surface of the heart, and appears as a saccular structure divided into two parts and formed by the swollen extremities of the common jugular veins and the post-caval veins (text-fig. 87). It is hardly dis- tinguishable externally from the right auricle, although internally the two cavities are separated by the bicuspid sinu-auricular valve. The right common jugular vein from the anterior end of the body and the post-caval from the posterior end join together to form the major part of the sinus venosus, which lies on the right of the dorsal surface of the heart. The left common jugular vein runs down along the outer edge of the left auricle and then across in the groove between the left aui-icle and ventricle (text-fig. 87). Its mouth opens into the smaller part of the sinus venosus *, which is partially separated from the major part by a valvular septum. The efficacy of this septum is seen when injecting, for while the right common jugular and post-caval veins may be easily injected from the major part of the sinus, it is almost impossible to inject the left common jugular from it. The thin- walled Auricles (text-fig. 86) are unequal in size, the right, of an elongated oval form, being much larger than the left, which is shorter and inore rectangular. Into the right auricle opens the sinus venosus and into the left the single pulmonary vein (text-fig. 87). The opening of the pulmonary vein is not guarded by a valve as Fritsch (16) pointed out, but it seems highly probable, as Sabatier (33) suggested, that during systole a fold of the auricle in this region functions as a valve and so prevents regurgitation. The auricles are completely separated by an imperforate inter-auricular septum which is continued caudally so as to divide the auriculo- ventricular aperture into two. The internal surfaces of the auricles possess a network of raised muscular ridges, the musculi pectinati. The Ventricle is somewhat oval in shape, but very asymmetrical. The posterior end forms a bluntly conical apex, and the base, although more or less transverse on the right side, is produced anteriorly on the left side into a conical process, so that the left side of the ventricle is nearly as long again as the right. It is extremely thick-walled, and its cavity contains a large number of muscular trabeculse, some of which interlace in such a way as to form an oblique, incomplete ventricular septum. This partial septum keeps the aerated blood brought in by the left auricle more or less completely separated from the non-aerated blood from the right auricle. Two valves, a right and a left, similar in arrangement to those in Lacertilia, guard the auriculo-ventricular apertures. * They do not open separate!}' into the auricle as stated in Rolleston (32). CIRCULATORY SYSTEM OF THE GRASS-SNAKE. 619 The Bidbus cordis, as lias been pointed out above, is not to be found as a separate structure in the adult, and so the three aortic arches arise directly from the ventricle. The base of each of these is guarded by two semilunar valves, which Langer (29) has shown to be homologous with the distal row of valves in the amphibian heart. III. The Arterial System. (PI. LXX.) (A) Beveloptnent. The development of the aortic arches in Tropidonotios is very similar to that of other Reptilia*. It was first described by Rathke (30), whose general account has been confirmed since by Van Bemmelen (7 & 8) except in one particular. Rathke Text-fiff. 88. Diagram of embrj^onic arterial arches. It represents the condition after the disappearance of arches 1, 2, and 5, and shows also the division of the truncus arteriosus into three arterial roots. C.A. Common carotid. D.A. Dorsal aorta. E.G. External carotid. I.C. Internal carotid. L.S.A. Left systemic arch. P.A. Pulmonary artery. P.Ar. Pulmonary arch. li.S.A. Right systemic arch. Adapted from Hochstetter (22). describes the development of only five visceral arches on each side, which he numbered from the anterior end 1-5, Yan Bem- melen, however, showed that there was another arch, which has however, a somewhat transient existence, between arches 4 and 5 of Rathke, so that the latter's fifth arch is in reality the sixth of the series, and thus the snake is brought into line with other Amniota. * For a general account of this see Hochstetter (22). 41* 620 MR. C. H, O DONOGHUE ON THE These arches soon become reduced to three on each side, viz. 3, 4, and 6, bj the disappearance of arches 1, 2, and 5. Of the remaining arches, 3 is the carotid, 4 the systemic, and 6 the puhnonary. By the separation of the trnncus arteriosus into three tvibes the two carotids and the right systemic have a common opening into the ventricle ; the left systemic opens separately, and the two pulmonaries open by a common vessel (text-fig. 88). The most remarkable change in the development is the enormous lengthening of the carotids, brought about partly by the elonga- tion of the neck but largely by the caudal shifting of the heart. Thus it happens that in the adult condition the 3rd arch is far removed from the 4th and 6th arches. Text-fiff. 89. Diagram to show changes in embryonic arterial arches. It shows the change from the condition in text-fig. 88 to the definitive state. The filled-in portions represent the vessels left in the adult, and those indicated by dotted lines the vessels that disappear in the course of development. A.B. Basilar artery. C.C. Left common carotid. Th.A. Thyroid. Other letters as in text-fig. 88. Adapted from Hochstetter (22). P.C. Primary carotid. The Carotid Arch goes through considerable changes in the course of its development. The two common carotids arise from the systemic by one root, the primary carotid (carotis primaria CIRCULATORY SYSTEM OF THE GRASS-SNAKE. 621 of Eatlike), which remains short as in most snakes *. After the common carotids have split into intei-nal and external carotids, each internal vessel gives off a branch which enters the neural canal with the first spinal nerve and joins the basilar artery on the ventral side of the nerve-cord. These branches quickly widen out and, forming as they do an anastomosis between the two internal carotids, make it possible for the right common carotid to degenerate. This it does completely f from the point where it divides into internal and external branches down to close to its union with the left common carotid, but the last part of it remains and is to be found in the adult as a small artery supplying the Thyroid gland (text-fig. 89). The Systemic Arch undei-goes very little change during development. The Pulmonary Arch degenerates almost completely on the left side. In conjunction with the suppression of the left lung in Tropidonotus, we find that the pulmonary branch of the 6th arch is only developed on the right side. In the adult only one pulmonary artery is to be found. (B) Adult Form, (a) Anterior Vessels. The Left Aorta bends dorsally around the cesophagus and trachea and then posterio)7ly to unite with the right aorta in the mid-dorsal line. During this course it gives off two very small branches to the oesophagus, but none whatever to the parietes. The Right Aorta takes a corresponding course on the other side, during which it gives off the following branches : — I. The Left and Eight Coronary Arteries arise behind the two semi-lunar valves which guard the base of the aorta. The right coronary artery runs in the groove between the auricle and ventricle, and is the chief supply of the dorsal surface of the heart. The left passes around the base of the pulmonary artery and spreads out over the ventral side of the heart. * In some snakes it is absent altogether, so that the common carotids come off separately from the systemic arch, e. g. Boa. t It is interesting to note, however, that in some variations of T. natrix this does not occur, and so the two common carotids persist in the adult. The first specimen of T. natrix that I examined was in this condition, although in all other respects it appeared a perfectly normal adult male. The two cominon carotids, left and right, sprang from a common stem, the primary carotid, and were about equal in calibre. On the left side the carotid pursued a normal course. The abnormal right carotid passed ventral to the oesophagus, just behind the thyroid gland, to which it sent a small branch, over to the right side of the neck. From here up to the posterior end of the skull it followed a similar course to its fellow on the left. Unfortunately the vessels of this specimen were not injected, so that the relation of the persistent right carotid to the basilar arterj' could not be ascertained. However, this apparent anomaly in the arterial system is quite readily understood in the light of the developmental history of these vessels. Only one other example of this peculiar abnormalitj' seems to have been described before, and tliat by Van Bemmelen (9), but in this case the right carotid was only a fine tube. 622 MR. c. H. o'donoghue on the II. The Primary Carotid (Carotis pi'imaria, Ratlike) is a short trunk which qiiickly divides into two unequal branches : — i. The Thyroid Artery is the smaller of the two, and, in addition to supplying the thyroid gland, it sends a twig to the right thymus glands. This is the sole remnant of the right common carotid. ii. The Left Common Carotid (Arteria carotis communis, Rathke, Arteria cephalica, Schlemm) i-uns along the left side of the oesophagus and trachea, to which it sends three or four slender bi-anches, until it reaches the posterior region of the head. Here it divides into internal and external branches, and supplies the whole of both sides of the head. (The distribution of this vessel in the head will be dealt with later.) III. The Vertebral Artery (Ai'teria vertebralis, Rathke, Arteria coUaris, Schlemm) arises from the anterior dorsal part of the right carotid arch and runs forward, a little to the right of the vertebral column about half-way to the head. It gives from three to seven branches to the parietes and one or more to the oesophagus before disappearing into the vertebral musculature in the mid- dorsal line, lY. Five Parietal Arteries are then given off. The first three are very slender and close together, while the remaining two are of the same size as the regular parietal arteries. V. One or two small (Eso2)hageal Arteries run to the oesophagus. After this the two systemic arches unite to form a single vessel, tlie dorsal aorta. The Left Ductus BotalU is not completely closed up in the course of development, and its proximal portion is to be found in the adult animal as a cul-de-sac rvmning cranially from a point low down on the root of the right carotid arch *. This remnant varies somewhat in size in different individuals, and although it is always more or less short, is of nearly the same calibre as the right carotid. It is completely hidden by the left auricle, but is readily seen if that body be carefully removed. From its somewhat bluntly conical end comes off a thin strand of tissue which runs forward into the left sj^stemic arch at the point where it bends over to run backwards. This represents the closed part of the left pulmonary arch, and is therefore the left ligamentum BotalH, such as has been described by Brenner (11) and Hoch- stetter (21). I find myself in agreement with the former author also when he states that he was unable to find a right ligamentum Botalli in Tropidonotus natrix. According to Brandt (10), quoted also by Hoffman (23), there * Such a saccular appendage appears to be present generallj" in those snakes with one lung suppressed, and has been recorded bj^ Hochstetter (21) in Tropidonotus natria^, T. tesselatus, Coluber (VscuJapii, CoroneUa lavis, Vipera berus, and Cerastes vipera. CIRCULATORY SYSTEM OF THE GRASS-SNAKE. 623 is present in the Grass-Snake a solid strand of tissue joining the primary carotid to the transverse part of the left aortic arch and called by him the ligamentum caroticum. This, he states, in exceptional cases may remain open and may then be described as a ductus caroticum, and is a vestigeal structure somewhat similar to the ductus Botalli. In the hearts I have examined no trace of this vessel or cord could be found, and, indeed, no such connection exists in the course of embryonic development, as a glance at text- fio's. 88 and 89 will show. A connection between the carotid and left aortic arches is present in the embryo, but in the adult it would run from the dorsal part of the left systemic arch along the whole length of the neck up to the point of origin of the internal carotid (text-lig. 89). This, however, does not fit in with Brandt's description of the ligamentum caroticum. (b) Posterior Vessels. The Eight Pulmonar'i/ Artery (Arteria pulmonalis, Schlemm) arises separately from the ventricle and leaves the heart the most dorsal of the three arterial roots. It runs backwards alongside the oesophagus almost parallel with the right systemic over which, however, it passes ventrally, and then runs dorsal of the post- caval vein to the anterior end of the lung. As it passes along the light border of that organ it gradually gets smaller and smaller until it disappears as a distinct vessel at the level of the posterior end of the liver, although the lung is continued on for some distance. In correlation with the suppression of the left lung no left pulmonary artery is found at any time. The Right and Left Aortic Arches unite posterior to the heart to form the dorsal aoi'ta. The Dorsal Aorta runs in the body-cavity just ventral to the vertebral column, back to the level of the cloaca. Just posterior to this it leaves the body-cavity and enters the hsemal canal, and in this is continued along the tail as the Caudal Artery. During its course through the ccelom it gives ofi' a number of branches. The Parietal Ai'teries form a numerous and more or less regular series of branches going to the body-wall, of which there are about tv/elve up to the point of origin of the superior mesenteric artery. These arteries enter the body-wall in the mid-dorsal line, a characteristic of most colubrine snakes, and do not split into two before so doing, as in the pythonine snakes (c/. Beddard, 4 and 1). The (Esophageal and Hepatic Arteries. — In front of the liver the dorsal aorta gives off two or three slender branches to the oesophagus. After these come a series of common trunks, about fifteen in number, which divide into two branches, one going to the liver and one to the oesophagus or posteriorly to the stomach. The last of this series is considerably larger than the others and has more branches, some of which go to the anterior end of the stomach. 624 MR. c. H. o'donoghue on the The following vessels then come off from the dorsal aorta in order : — 1. The Lieyio-gastric Artery is the first of these. Its gastric branch is the main artery supplying the stomach, and it also sends a branch to the spleen, and yet a third, the cystic artery, to the gall-bladder. No branch of it goes to the pancreas, nor does the superior mesenteric artery send twigs to the spleen and gall- bladder as Beddard (1) has described in Tropidonotios fasciatus. 2. The Superior Mesenteric Artery is the largest vessel arising from the dorsal aorta. Soon after its origin at about the level of the pancreas it divides into two branches ; a smaller one, the duodenal, running anteriorly supplies the part of the intestine immediately after the pylorus and also the pancreas; a much larger one running posteriorly supplies the manj^ coils of the intestine as far back as the posterior end of the right ovary. Small branches from it also supply the anterior part of the fat- body. I have been unable to find any branch of this artery running to the right ovary such as Beddard (1) recorded in Trojndonotus fasciatus. 3. The Right Ovarian Artery^ a moderate-sized vessel, runs to the right supra-renal body and, dividing into anterior and pos- terior branches, forms a longitudinal vessel along it. From this longitudinal trunk are given off : — (a) Six small ovarian arteries of equal size. {b) Three somewhat larger oviducal arteries. One of these arises from the anterior end of the longitudinal vessel and supplies the fimbriated opening of the oviduct, and the other two arise from the posterior end. (c) Three fat-body arteries — an anterior, a small median, and a posterior. (fZ) A number of fine twigs to the supra-renal body. The posterior of the three fat-body arteries in some cases has an independent origin from the aorta. 4. The First Inferior Mesenteric Artery supplies the coils of the intestine just posterior to the right ovary. Its point of origin varies, however, in different individuals and in the two sexes, as will be pointed out below. 5. The Anterior Right Renal Artery conveys blood to the anterior half of the right kidney, and also gives off a branch to the right oviduct. 6. The Left Ovarian Artery is distribiited in a very similar way to the right, and from the longitudinal trunk it forms along the supra-renal body come off : — (a) Five equisized ovarian arteries. (h) Three oviducal arteries. The anterior again supplies the oviducal funnel. (c) Two fat- body arteries. [d) A number of fine twigs to the supra-renal body. As on the right side the posterior of the fat-body arteries may arise separately from the aorta. CIRCULATORY SYSTEM OF THE GRASS-SNAKE. 625 7. The Anterio7' Left Renal Artery divides soon after its origin into two branches, one going to the fat-body and the other to the kidney. This latter branch supplies the anterior half of the kidney and sends a twig to the oviduct. 8. The Second Inferior Mesenteric Artery arises about the level of the anterior end of the left kidney, and supplies the intestine in the region of the median part of the right kidney. 9. The Median Right Renal Artery, in addition to taking blood to the posterior median part of the kidney, sends a branch to the right oviduct. 10. The Postet^ior Right Renal Artery feeds the posterior end of the kidney and gives off two branches to the oviduct. 11. The Third Inferior Mesenteric Artery, arising near the level of the posterior end of the right kidney, supplies the intestine in the region of the posterior part of the left kidney. 12. The Median Left Renal Artery is distributed very similarly to the corresponding vessel on the right. 13. The Fourth Inferior Mesenteric Artery, arising at the level of the posterior end of the left kidney, supplies the last part of the intestine. 14. The Posterior Left Renal Artery resembles its fellow on the right, save that it does not send a branch to the oviduct. 15. The Posterior Oviducal Arteries are a pair of arteries running to the posterior ends of the oviducts. In addition to this there may be one or two small twigs going to the rectum. 16. The Rectal Artery is a small vessel supplying the last part of the rectum. An Epigastric Artery is present, and it appears to be similar to that described by Beddard (1) in Ofhiopjiagus bungaris, that is, it runs along the body-wall in the mid-ventral line in close con- nection with the epigastric vein, and is fed by branches from the carotid anteriorly and the fat-body posteriorly. Its precise relations are somewhat difficult to make out, for it is a vessel too small for individual injection, and in order to get a satisfactory injection of the arterial system it is necessary to open the snake from end to end, dissect away the skin, and free the gut to some extent. This, however, necessitates cutting either the epigastric artery itself or some of its small tributaries. The arrangement of the vessels in the male snake is very similar to that just described for the female. All the arteries anterior to and just posterior to the heart are precisely the same, and it is not until the region of the urino-genital organs is reached that we find any difference. Each testis receives one spermatic artery as is general in snakes, which first runs to the supra-renal body, whence it sends branches to the testis and also to the vas deferens. These two spermatic arteries are each followed by another vessel that runs to the pos- terior part of the supra-renal body and also supplies the vas deferens for a considerable distance, that on the right side also sends a bi-anch to the fat-body. 626 Mil. c. H. o'donoghue on the The kidneys eacli possess an anteiior and a median renal artery as before, but instead of one posterior trunk there are at least two, but most often three *. These vessels also send small twigs to their respective vasa defer-entia, and the left anterior renal supplies the fat-body with a large artery. The supply to the alimentary canal is also slightly different. The first of the inferior mesenteric arteries arises posterior to the anterior left renal artery, whereas in the female it is anterior to the anterior right renal t. This is followed by a series of about four smaller inferior mesenteric arteries. The various mesenteric arteries are connected by their small branches and so form more or less of a longitudinal system along the gut. The origin of the arteries supplying the fat-bodies in both sexes is fairly similar and is somewhat interesting. The anterior part is supplied by branches from the superior mesenteric artery, the next portion by vessels from the right genital artery, and the posterior end is fed by branches from the left genital and left renal supply. In addition to which the posterior branch from the genital artery on each side may in some cases arise independently from the dorsal aorta. All these bi-anches are joined one to the other by small twigs into a longitudinal system running the whole length of the fat-body, but there does not appear to be one unbroken artery, an arteria epiploica, traversing the whole length of the fat-body. lY. The Venous System. (PI. LXX.) (A) Development. The general coux-se of the development of the venous system in Tropidonotus is similar to that of other reptiles [vide Hoch- stetter, 22), but it has some points peculiar to itself [vide Rathke, 30, and Hochstetter, 23). The first veins to appear are the two omphalo-mesenterics, of wliich the right is somewhat stouter than the left (the reverse is the case in Lacerta), and they open into the sinus venosus. Soon after their appearance the anterior and posterior cardinal veins arise on each side, and their common stem unites with the umbilical vein on each side to form the ductus Cuvieri, which becomes associated with the omphalo-mesenteric veins at the * The number of renal arteries varies in different species, being only one in 'Python spilotes and eight in Coronella catenifer, Beddard (1). t This differs from the account of T.fasciatus in Heddard (1), where " it springs from the aorta in both sexes close to the second {i. e. left) gonad artery, in front of it ill the male, behind it in the female." Some variation is to be found, however, in the position of this artery in different individuals, for in the females of T. natrix that I liave examined, although it has generally been in front of the right anterior renal and left ovarian arteries, it may be behind these vessels. In the male it is usually behind the left anterior renal arterj-, but it may be in front of it. It has not occurred antorior to the right anterior renal artery' in any male snake that I have examined. Again, we find considerable variation in the number of tliese gut-arteries in various snakes. LacJiesis gramineus has only one, while in the genus Coluber there may be ten or eleven (Beddard, loc. tit.). CIRCULATORY SYSTEM OF THE GRASS-SNAKE. 627 point where they open into the sinus venosns. An anastomosis between the two omphalo-mesenteinc veins forms on the dorsal side of the gut just posterior to the pancreas rudiment, and the portion of the left vein between this point and the siniis venosus disappears. A similar anastomosis between the two veins now forms on the ventral side of the gut, and thus a complete ring is formed. In a short time, however, the right lialf of this ring disappears, leaving a single vein which runs in a spiral manner round the gut. While these latter changes are taking place, the middle part of the right omphalo-mesenteric vein between the sinus venosus and the dorsal anastomosis spreads out and forms a venous network in the liver. The portion of the right omphalo- mesenteric vein in front of the hepatic network persists as the hepatic vein, and the part behind it always remains as the anterior end of the hepatic portal vein (text-f3g. 90). Text-fii;. 90. Diagram of posterior veins in the embryo. It shows tlie change from an earlj' to a late embryonic condition. The shaded vessels being the first to disappear. A.C. Alimentary canal. D.C. Ductus Cuvieri. H.N". Hepatic network. L.O. Left omphalo-raesenteric (soon disappears). L.U. Left umbilical. 0. Omphalo- mesenteric. P.O. Post-caval. R.O. Right omphalo-mesenteric (disappears later than L.O.) . K.U. Eight umbilical. S.V. Sinus venosus. Adapted from Hochstetter (20). The posterior cardinal vein originates at the caudal end of the mesonephros, and runs forward along the dorso-lateral edge of that body. After leaving the kidney, however, it breaks into several branches, which soon reunite and then it runs into the ductus Cuvieri. When the caudal vein develops it divides into two branches at its anterior end, each of which runs to the exti-emity of the corresponding cardinal vein (not along the mesial wall of the kidney as in Lacerta). The post-caval vein springs from the union of the right umbilical and omphalo-mesenteric 628 MR. c. n. o'donoghue on the veins and runs backwards in the mesentejy to the kidneys. Between them it breaks up into two branches, which run pos- teriorly along their mesial borders. These branches meet the kidneys at a point some distance from their front end, and there each receives a branch from the anterior part of the kidneys. Subsequently the parts of the posterior cardinal veins in front of each mesonephros disappear, and so all the blood from the caudal veins has to pass through those organs. Still later the adult kidney ai'ises caudally to the mesonephros, and then the end part of each posterior cardinal vein comes to lie on its ventral and external border, while a continuation of one of the posterior mesonephric branches of the post-caval vein extends along the inner border of the permanent kidney. The left umbilical vein loses its opening into the ductus Cuvieri, and all its blood is taken to the liver network. The right umbilical vein, on the other hand, retains its original opening into the ductus for some time (text-fig. 90). In the subsequent changes the post-caval vein increases in size, and the part of the omphalo-mesenteric vein joining it breaks up into a venous net- work with the caudal extension of the liver. As the two umbilical veins are joined by an anastomosis at the navel, it is possible for the right to disappear, which it does, leaving the left, which, however, disappears soon after birth *. Ultimately the omphalo- mesenteric vein, lying on the dorsal side of the livei-, can only communicate with the sinus venosus via the hepatic network and the post-caval vein, which is on the ventral side of the liver. The anterior cardinal veins originate in a similar way to those of Selachians and Amphibians, but the pai'ts of these veins in the head are completely replaced in an interesting way (vide Grosser and Brezina, 19). The original cardinal vein r-uns backwards from the infraorbital and the anterior cerebral veins ventrally to the cranial nerves into the ductus Cuvieri. Three venoiis rings are now formed in succession ; the first around the root of the facial nerve and the auditory vesicle, the second around the root of the glossopharyngeal nerve, and the third around the vagus root. Their median portions afterwards disappear, and the external ones unite to form one latei-al trunk. In the meantime two new vessels come to open into the anterior cardinal : one, the median cerebral vein, coming from the cerebellum, opens just posterior to the trigeminal nerve ; and the other, the posterior cerebral vein, coming from the medulla, opens posterior to the vagus, leaving the skull by the foramen magnum. Still later the lateral trunk opens anteriorly into the infra- orbital vein by means of an extension by the side of the second and third branches of the trigeminal nerve, and posteriori}'- past the hypoglossus into the anterior cardinal opposite to the pos- terior cerebral vein. In this way is formed a new, complete * Generally the umbilical vein disappears in the adult snake, but remains of it persist in some species, e. g. Soa constrictor, S, diviniloqtia, Fython regiits, JEunectes murmus, C'oralhts cookii (Boddard, 4 & 5). CIRCULATORY SYSTEM OF THE GRASS-SNAKE. 629 lateral trunk, the lateral cephalic vein (V. capitis lateralis, Grosser and Brezina). At the same time the three cerebral veins become connected by a median longitudinal vessel (text-fig. 91). Further, two new anastomoses arise from the median cerebral vein, one goes to the anterior cerebral vein and the other, the secondary median cerebral vein, leaving the skull with the trigeminus, goes to the lateral cephalic vein. This becomes the main vein leading from the anterior part of the brain. Text-fi^. 91. RC.V Diagram of veins in the region of tlie head in the embryo. It shows the original veins indicated by the shaded portions and the definitive vessels indicated in black. A.C.V. Anterior cerebral. A.V. Auditory vesicle. A.V.C. Anterior cardinal. I.O.V. Infra-orbital. J.V. Kight common jugular. L.C.V. Lateral cephalic. M.C.V. Median cerebral. P.C.V. Posterior cerebral. P.V. Prosencephalic. S.M.C. V. Secondary median cerebral. II, V, VII, IX, X, XII, Primordia of cranial nerves. Adapted from Grosser and Brezina (19). The adult condition is reached by the regression of the dorsal part of the anterior cerebral vein, the complete disappearance of the portion of the anterior cardinal vein in the head, leaving its cervical part, however, as the common jugular vein, and the formation of a vein bringing back blood from the upper and lower jaws. 630 MR. c, n. o'donoghue on the (B) Adult Form. («) Anterior "Vessels. The whole of the blood from the head is brought back to the heart by the two common jugular veins. The distribution of the veins in the head and their union to form the common jugulars will be dealt with later. The Left Common Jugular Vein * (V. jugularis sinister, Schlemm) runs fi^om the anterior end of the neck parallel with and close to the left common carotid artery to the heai't. It passes ventral to the left systemic arch to the dorsal side of the left auricle, and it runs along the dorso-lateral edge of this structure to the auriculo- ventricular sulcus. Here it bends sharply to the right and runs into a part of the sinus venosus somewhat sharply marked off from the remainder. It receives : — i. In its anterior part a few small tributaries from the anterior end of the oesophagus. ii. The Coronary Veins (text-fig. 87), a series of small vessels which join it in its course along the auriculo- ventricular sulcus. . The Right Common Jugular Vein * (V. jugularis dextra, Schlemm) is similar in disposition to the left, but at the level of the anterior end of the right auricle it receives a large tributary, the azygos vein. After this it runs straight on and unites with the post-caval vein to form the major part of the sinus venosus. I. The Azygos Vein (V. azygos, Schlemm) is a short trunk leading into the i-ight common jugular vein from the union of the anterior and posterior azygos veins. A. 1\\B Anterior Azygos Vein (Y. azygos anterior, Schlemm) lies in fi'ont of the heart between the oesophngus and the backbone, and extends from the beginning of the neck back to tlae heart. It collects blood from the dorsal body- wall along this region by means of a fairly regular series of intercostal veins, all of which come off to the right of the vertebral column. Just before it unites with the posterior azygos vein it receives a branch from the oesophagus. («) The (Esophageal Vein is formed at the level of the front end of the thyroid gland by the union of two branches. One of these branches comes from the anterior dorsal part of the oesophagus, and the other from the posterior dorsal part. Their common stem runs almost straight to the right, ventral to the vertebral artery and dorsal to the right common jugular vein, and opens into the anterior azygos vein. * It is to be noted that the proximal parts of these two veins are homologous with the precaval veins of Lacertilia. As, however, there is no representative of the subclavian veins to mark the beginning of the precaval portion, it is convenient to apply the one name to the whole vessel. CIRCULATORY SYSTEM OF THE GRASS-SNAKE. 631 B. The Posterior Azygos Vein (V. Azygos posterior, Schlemm) is a much shorter vessel than the anterior, and also than the homologous vein in Lacerta. It originates at about the level of the middle of the ventricle, and runs forward to meet the anterior azygos just anterioi- to the right auricle. In its course it receives three or four intercostal veins, all of which come from the body- wall to the right of the vertebral column. (&) Posterior Vessels. The Right Fulmonary Vein (V. pulmonalis, Schlemm) first becomes noticeable as a definite vessel on the left side of the lung at the level of the posterior end of the liver. It runs forward along that organ to its anterior end, receiving numerous branches and, leaving the lung, it runs parallel to the post-caval vein and vential to the right systemic arch to open into the right auricle. A left pulmonary vein never exists at any time. The Caudal Vein (V. caudalis, Schlemm) arises far back in the tail and runs forward in the hsemal canal, together with the caudal artery. It leaves this canal and divides into two branches, the renal portal veins, a short distance before reaching the cloaca. The Renal Portal Vein (V. venalis advehens, Schlemm) on each side runs forward from the bifurcation of the caudal vein over the cloaca and along the outer side of each kidney. It receives soon after its origin a lateral tribxitary, which from its position and distribution appears to correspond to a pelvic vein *. If this be the case, it is interesting as being the only indication in the circulatory system of the derivation of the snake from a limb- beai'ing ancestry. Each renal portal vein receives several small tributaries from the dorsal body- wall, the cloaca, and the lateral part of the tail. It then passes over the cloaca and lies between the oviduct which is on its outer side and the meter which is on its inner side. Shortly after passing the cloaca each vein gives ofi" a well-marked branch, the right being anterior to the left, which passes dorsally around the ureter and then runs ventrally to unite with its fellow on the dorsal wall of the gut. The vein so formed, the "Yeine mesenterique posterieure" of Jourdain (27), runs forward along the gut and is continiied as the hepatic portal vein. In its course from the cloaca to the kidney each renal portal vein receives a number of small tributaiies from both ureter and oviduct, one of which, the Posterior Oviducal Vein, is well marked, and also one or more from the dorsal parietes. On reaching the kidney it runs along its lateral border closely accom- panying the ureter, and, gradually getting smaller, disappears at the anterior end of that organ. It is not continued anterior to the kidney as in some snakes, e. g. the Boidse and also in Zamenis gemonensis (Beddard, 4 & 2), a feature recalling the condition in the Lacertilia. During its course along the posterior part of the kidney this vein receives a well-marked tributary from the oviduct, * A similar pair of veins is to be found in JEunectes murinus (Beddard, 2). 632 MR. c. H. o'donoghue on the Tlie Right Efferent Renal Vein (Y. venalis rovehens, Sclilemm) originates along the inner margin of the kidney at its posterior end, runs to the anterior end, and after a short independent course in fi-ont of the kidney unites with its fellow of the opposite side to give rise to the postcaval vein. The Left Efferent Renal Yehi also has its origin along the inner margin of the kidney. But, unlike the right, it receives branches from tlie supra-renal body, the ovary, and the left oviduct in its course from the kidney to the point where it unites with its fellow. Thus : — i. The Left Ovarian Veins are numerous small tributaries flowing into the left eflerent renal veins. ii. The Left Supra-Renal Veins ai^e a large number of small branches running from the supra-renal body into the efferent renal vein, to which it is closely attached. They return the blood brought to that body by the supra-renal portal veins. iii. The Left Oviducal Vein is a large vein leaving the oviduct at the level of the fimbriated funnel and flowing into the efterent renal vein just in front of the supra-renal body. It returns blood fi'om the oviducal sinus. A. The Left Oviducal Shins is a wide thin-walled vessel which runs from the extreme anterior end of the ovi- ducal funnel backwards to about the level of the middle of the corresponding kidney. It is very conspicuous in injected specimens and was present in all the female examples of 7\ natrlx that I have examined, although it does not appear to have been recorded in any other Ophidian. The Post-Caval Vein (V. cava posterior, Schlemm) is formed, as has been stated above, by the union of the two etterent renal veins. It passes forward, dorsal to the gut, freely in the mesen- tery to the posterior extremity of the liver, and then along in a groove in the ventral surface of that organ to its anterior end. After leaving the liver it runs almost straight forward, ventral to the right systemic ai-ch and unites with the right pi'e-caval to form the larger division of the sinus venosus. Soon after its origin it receives veins from the right ovaiy, the supra-renal body, and the oviduct. Thus : — i. The Right Ovarian Veins are similar to those of the opposite side, ii. The Right Supra-Renal Veins also resemble those on the left and return the blood gathered by the supra-renal portal veins. iii. The Right Oirlducal Vein is like its fellow on the left and comes from an oviducal sinus. A. The 2iight Oviducal Sinus extends along the oviduct from the .antei'ior end of the fmmel to about the level of the middle of the right kidney. CinCULATORY SYSTEM OF THE GRASS-SNAKE. 633 The Sajyra-renal Portal System. — Each supm-renal l)Ofly lias a portal supply, au aiTangement of vessels which is universally present in snakes according to Beddanl (1), who, however, attributes tlie discovery of this system to Gratiolet in 1853 (17), whereas it had been described seven years previously Ijy Ecker (15). The supply consists of two, but in more rare cases of three, intercostal veins which arise from the cori-esponding side of the dorsal bodj^- wall and also of a vessel from the oviducal sinus in the female, all of these pour their blood into the supra-renal network. The last of these vessels is rather small and runs at the level of the middle of the ovary on each side. The Ilejoatic I'ortcd Vein (Vena portre, Schlemm) arises by two roots, from tlie renal portal veins, which unite to form a single ves.sel on the dors^il wall of the rectum in a way that has already been described *. It passes forward through tlie gut mesentery, receiving on its way numerous bra-nches from the many coils of the small intestine and also from the fat-body. Towards its anterior end a tributary joins it bringing blood from the front part of the intestine, the pancreas, and the spleen. This is shortly followed by another vein coming from the gall-bladder. In the same region it is joined by the anterior abdominal vein which runs down from the fat-body. From this point instead of being on the ventral side of the post-caval vein it pfisses dorsally to the left of this vessel and enters a furrow on the dorsal side of the liver. Between the entrance of the vein fj-om the pancreas and the spleen and the posterior end of the liver, i. e. just anterior to the superior mesenteric artery, the hepatic portal vein receives three large intercostal veins from the parietes of the right side. The vessel runs in the groove of the liver right to its anterior end, gradually diminishing in calibre, and it receives in this part of its course a more or less regular series of intercostal veins arising to the left of the vertebral column and a greater number of snudl veins from the stomach and oisophagus. A very similar condition obtains in T.fasciaius (Jieddard, 1). I. The Anterior Abdominal Veinf, corresponding to the simi- larly-named vein in Lacertilia, is a single small vein arising at * According to Schlemm (35) the hepatic portal yein has only one root, and that arises from the right roiial portal vein. This statement is a)so made in Hoffmann (23), but it should Ijc noted tliat this author quotes nearly the whole of Schlenmi's account of the venous system almost verbatim without iiiilicatiiig jn any way that he is so doing. My own investigations confirm those of Jourdain (27) iiml Jrloch- stetter (20), vvho descrilie a disciroyus. According to .Jatjuart (26) there arc a number of anastomoses between the hepatic portal vein and the right renal portal vein in Vijthon. In Gadow's account of Pelnphihis madac/ascarieusis, qm>ti:d by Holtinann (23), it is stated that there is no connection between the hepatic portal and renal portal veins. t This vein is especially interesting, as it is subject to considerable variation among the Ophidia. In jjizards, as is well known, it aiises by two roots from the renal portal veins, and a similar condition is to be found in some snakes, \\7,.,Ei\i/.v jaculus, E.Johni. Fi/thon sebce, and lion diviniloqua, Beddard (2, 3, & 4). In other snakes Proc. Zool. Soc— 1912, No. XLIf. 42 634 MR. C. H. o'cONOGnUE ON THE the posterior end of the fat-body and running forward in it to the level of the spleen, Avhere it passes dorsally and opens into the hepatic portal vein. It is only connected in an indirect way by small anastomosing branches with the i-enal portal veins. Along its course it receives little twigs from the epigastric vein. The Epiyastric Vein lies beside the epigastric artery in the mid- ventral line of the abdominal wall. In the region of the liver it gives off five or six small branches, all of which enter directly into the left side of that body and are not connected with the hepatic portal vein. Behind the liver the epigastric vein is connected by a number of small venules with the anterior abdominal vein. As Beddard (4) has pointed out, this is one of the most constant veins in Snakes, and is single save in Lioheterodon madagascariensis, where it is alternately single and double. The veins in the male are, like the arteries, on the whole very similar to those in the female. Those in front of the heart are precisely similar in both sexes. The caudal vein bifurcates to form the renal portal veins which, at the level of the cloaca, receive the paired pelvic veins, and in addition, in the male a vein from each corpus cavernosiim. The renal portals give rise to the two branches which unite above the gut to form the beginning of the hepatic portal vein and then pass forwards to the kidneys between the vasa deferentia and the ureters. On the kidney they receive no specially marked tribu- tary from the vas deferens to correspond with the one from the oviduct in the female. Each testis gives off one spermatic vein, whereas in the female there are a number of small ovarian veins, just in front of the corresponding supra-renal body, that of the left side opening into the left efferent renalvein, and that of the right into the post- caval vein. There is no vessel in the male to correspond with the oviducal sinus, and consequently no branch from it to the supra-renal body. The supra-renal portal supply consists of two intercostal veins, one at each end of that body, which arise from the corre- sponding side of the vertebral column. Some variation is to be met with in the position of the union of the two efferent renal veins with regard to the kidneys. The junction may be as much as an inch in front of the right kidney, or, on the other hand, this kidney may overlap the point of union, in which case several small veins bring back the blood from the it has only a siiia;le origin from the left renal portal vein, viz., JSrijx cnnicus, Emiectes mnrinus, and E. noftens, Beddard (3). Lastly, it may have no direct connection with the venal portal veins, bnt only indirect ones by means of anastomosing twigs, viz., Zamenis gemonensis, Beddard (2), Coluher ascula, and Tropidonotns natrix, Hochstetter (20): This last observation I have been able to confirm. Further, the anterior abdominal vein may be partly double thronghout, as in Boa constrictor, a. divinlloqua, JS)\yx jaculus, Beddard (4), and Pt/tlion sehce, Jaquart (26), or single as in Zamenis gemonensis, Catisus rhombeatus, JEri/x jnhni, Beddard (2), and Tfopidonotus uatrix. CIRCULATORY SYSTEM OP TUB GRASS-3XAKE. 635 anterior end of the kidney into the right efferent renal vein. On the whole this distance appears to be greater in the female than in the male. The remaining vessels of the male, the hepatic portal factors, the anterior abdominal vein, and the epigastric vein correspond in all respects to those of the female. Y. The Vessels of the Head. (Pis. LXXI. & LXXII.) (A) Arteries. It has been pointed out above that the whole of the blood is brought to the head by the left common carotid, the right common carotid having disappeared early in the course of development. To compensate for this absence of an artery on the right ride we find developed three arterial anastomoses between the two sides of the head. The first lies beneath the medulla oblongata and joins the two internal carotids ; the second is situated beneath the fore-brain just in front of the optic chiasma and unites the anterior cerebral and facial carotids ; and the third is behind the symphysis of the lower jaw and joins the two external carotids. The Left External Carotid (Carotis externa, Rathke; Arteria inframaxillaris, Schlemm) arises from the common carotid internal to the articulation of the lower jaw and the quadrate bone. It runs forward between the floor of the pharynx and the broad mylohyoideus * muscle, first inwards towards the tongue sheath and then outwards to the inner side of the mandible, being accompanied throughout the greater part of its course' by its corresponding vein, the glossopharyngeal nerve, and the cutaneous branch of the hypoglossal nerve. At the anterior end of the lower jaw, just a short distance behind the symphysis, the left external carotid anastomoses with its fellow by a well-marked vessel. The Right External Carotid is similar to the left, save that the common carotid from which it originated has disappeared and is represented only by a small branch vessel It receives its blood- supply partly from the anastomosis just mentioned, and partly from the anastomosis between the internal carotids. The distribution of the arteries in the dorsal part of the head is the same on both sides, so that the one description will apply equally well to either side. The Internal Carotid (Carotis interna, Rathke ; Art. cephalica and Art. carotis communis, Schlemm) starts from the origin of the external carotid and bends in a sharp curve dorsally round the angle of the lower jaw on the inner side of the vagus and hypoglossal nerves. It then passes forward under the columella and along the inner side of the quadrate to a point behind the orbit and above the posterior pterygo-sphenoidalis muscle, where * The nomenclature of the muscles is that adopted hy Hoffiuann (23) 42* (536 MR. c. u. o'doxogiiue on the it divides into two branches, the cerebi-al carotid and facial carotid arteries. During this course it gives off the following branches : — I. One well marked and sometimes also a smaller artery that supply the oesophagus. II. A branch that supplies a part of the cervico-mandibularis muscle, the lateral and ventral walls of the pharynx and skin in this region. III. A branch (Ramus pterygoideus, Schlemm) which supplies the transverso-maxillo-pterygo-mandibularis, and the cervico- mandibularis muscles and the skin near them. IV. The First Spincd Artery (Ramus spinalis, Rathke ; Art. nervi spinalis I, Hofmann) which arises at the same level or slightly in front of the preceding. The vessel passes through the atlanto-occipital membrane between the axis and the basi-occipital bone and, running into the loop of the basilar artery beneath the medulla oblongata, forms an anastomosis with its fellow of the opposite side, a relationship first described by Schlemm (35), who called this vessel the Truncus anonymus. By means of this anastomosis the blood can pass from the left to the right internal carotid. Shortly before piercing the atlanto-occipital membrane this artery gives oflf two branches : — A. A small branch (Ramus pterygoideus, Schlemm) to the Pterygo-sphenoidalis m uscles. B. A fairly large Occipital artery (Art. occipitalis, Schlemm), which however soon divides into two vessels, a small and a large. The smaller is the occipital branch (Ramus occipitalis, Schlemm), supplying the muscles and skin of the occipital region. The larger is the Cervical Artery (Art. cervicalis, Schlemm), a long and fairly well-marked vessel. It I'uns backwards near the hypapophyses of the anterior vertebrae, covered by the i-ectus capitus anticus muscles to which it sends small branches, to the fourteenth vei"tebra. Here it joins Avith a small branch from the bifurcated extremity of the vertebral artery. During its course it gives off a series of branches that lie beside the spinal nerves, and each of these sends branches not only to the skin and muscles, but also a small one, a spinal artery, that enters the neural canal through the intervertebral foramen. V. The Maxillary Artery (Art. maxillaris, or dentalis inferior, Rathke ; Art. alveolaris inferior, Schlemm) arises about midway between the spinal artery and the point just behind the orbit where the internal carotid divides. It runs outwax^ls, accompany- ing the inferior maxillary braxich of the trigeminal nerve between the second and third parts of the parietali-quadrato-mandibularis muscle, to which it gives branches. After giving ofi' several twigs also to the transverso-maxillo-pterj^go-mandibularis muscle, it passes on through the posterior maxillary foramen into the lower CIRCULATORY SYSTEM OF THE OrRASS-SNAKE. 677 jaw. It soon gives off a branch to the inferior labial gland *, and then runs on in the lower jaw to its anterior end. Here it comes out through the anterioi' maxillai'y foramen as the Mentalis Artery (Art. mentalis, Rathke) and supplies the anterior part of the inferior labial gland and the skin in the region of the chin. Just anterior to this branch the internal carotid splits into two bi-anches, the cerebral carotid and the facial carotid arteries. The Cerebral Carotid (Art. carotis cerebralis, Rathke and Hof- mann) is a vessel of moderate calibre running downwards just behind the infra-maxillary branch of the trigeminal nerve. It then turns inwards under the skull and enters the cranial ca.vity through a foramen in the basisphenoid. This vessel was overlooked by Schlemm. In its course it gives off a small artery just outside the skull going to the posterior pterygo-sphenoidalis muscle. Inside the skull near the hinder end of the pituitary- fossa it splits into three fairly equal vessels, the posterior, median, and anterior cerebral carotids. I. The posterior branch (Ramus caudalis, Hofmann) runs back- wards round the outside of the pituitary fossa, giving off branches to the cerebellum and, about half-way to the foramen magnum, unites with the similar vessel of the other side to form the median, basilar artery. The Basilar Artery (Art. basilaris, Rathke and Hofmann) passes backwards until just before the foramen magnum, where, it splits up into a loop in the form of an isosceles triangle. Into the corners of the base of this loop open the right and left first spinal arteries, and thus it forms the anastomosis between these two vessels. Before it divides to form the loop, the basilar artery gives off on either side a well-marked internal auditory artery (Art. auditiva interna, Rathke and Hofmann) that enters the ear with the auditory nerve, and also a series of smaller branches, some of which supply the medulla, and some run on to the small choroid plexus of the fourth ventricle. II. The median branch f runs outwards behind the cerebral hemisphere, to vvhich it gives some twigs, and a short distance from its origin divides into two. A. The Anterior Choroid Artei-y (Art. choroides anterior, Hofmann) is the anterior branch. It passes around the hemispheres, supplying them with small twigs, to the dorsal side of the brain near the pineal stalk, where it breaks up in the choroid plexus of the third and lateial ventricles and anastomoses with a branch from the olfactoiy artery. B. The posterior division passes behind the optic lobes and spreads itself out over their dorsal surface. * The uomenclatvire of these glands in the head of the snake is that given by West (37&38). f In some examples this median brniicli does not arise at the jjoint where the cerebral carotid splits into anterior and posterior branches, but a little way down the latter ; consequently it appears as a branch of the posterior cerebral carolid alid is described as such by Hofmann (24). 638 MR. c. H, o'donoghue on tue III. The anterior branch (Ramus ci-anialis, Hofmann) runs forward along the side of the pituitary fossa and vmites with the similar vessel from the other side under the bases of the optic nerves immediately in front of the chiasma. By the union of the two anterior and the two posterior branches of the cerebral carotids a complete arterial ring, the circle of Willis (circulus arteriosus cerebralis, Hofmann) is formed around the pituitary fossa. Duiing its course this anterior branch gives off : — A. A Median Cerebral Artery (Art. cerebri media, Hofmann), which is a large vessel running round to the doi'sal side of the brain and supplying the anterior end of the hemi- spheres and the olfactory lobes. B. An Ophthalmic Artery (Art. ophthalmica, Hofmann), which is given off immediately before the two anterior branches of the cei'ebral carotids unite to form the circle of Willis. This passes out of the skull with the optic nerve, and immediately on reaching the orbit anastomoses with a branch of the facial carotid. The Olfactory Artery * (Art. olfactoria, Schlemm) arises from the mid-point of the anastomosis of the two anteiior branches of the cerebral carotids and runs forward in the skull in the furi-ow between the two olfactory lobes, to which it sends branches. At the anterior end of the lobes it gives off two symmetrical branches. Each of these again divides into two (Aa. ethmoidales, Ratlike ; Aa. bulbi olfactorii mediales, Hofmann), which pass out of the skull Avith the olfactory nerve to ramify over the olfactory membrane. The main trunk then recurves dorsally and runs back in the fissure between the two hemispheres, to which it sends numerous small branches, and near the pineal body anastomoses on each side with a branch of the anterior division of the median branch of the ceiebral artery. T. e Inferior Sjjinal Artery (Art. spinalis inferior, Rathke ; Tractus spinalis ventralis, Hofmann) is situated just below the ventral fissure of the spinal cord, and runs in a fairly straight line caudally from the anastomosis between the right and left fii-st spinal arteries at the posteiior end of the basilar artery. On its course it gives off branches to the spinal cord, some of which pass around to the dorsal side and others enter the ventral fissure, and it also receives the paired spinal arteries which come in through the vertebral column at the points of exit of the sjiinal nerves. In some places where a pair of such arteries enter it, the inferior spinal artery splits into a diamond-shaped loop. The Facial Carotid (Carotis facialis, Rathke) takes a faii-ly S'.raight course forward through the temporal fossa close to the infra-'.uaxillary branch of the trigeminal nerve to the orbit. Here it passes under the post- frontal bone into the orbit and * Tliis artery is douLle for tlic greater part of its length in Fi/tlion molvrus, Beddard (6). CIRCULATORY SYSTEM OF inE GRASS SNAKE. 639 immediately divides into two fairly ec^uisizsd terminal brandies. On its way it gives off : — I. A branch (Ramus glandulfe maxilL-e superioris posterior, Schlemui) which arises a short distance behind the orbit and runs outwards to the posterior part of the parotid glaml, i. e., the gland corresponding to the poison-gland of other snakes. It ramifies in the gland, and one of its branches anastomoses with a branch of the artery supplying the anterior part of the gland. During its course it gives ofi' a palatine branch and branches to several muscles. II. Several small branches and one fairly well-marked one that suf)ply the Harderian gland. III. Two or three slender branches to the skin overlying the skull and one or two to the under side of the skull. One terminal branch (Carotis facialis, Schlemm) of the facial carotid passes along the back of the orbit downwards to its tioor, where it runs forward (Ramus palatinus anterior, Schlemm), accompanjang the infra-orbital branch of the trigeminal nerve. It goes on forward out of the orbit, close to the outer wall of the internal nares, and finally spreads out over the skin and muscles at the front end of the snout. During its course it gives off: — I. A well-marked vessel (Ramus glandulae maxillaris superioris anterior, Schlemm) that accompanies the supra-maxillary branch of the trigeminal nerve outwards in the posterior wall of the orbit and along the upper jaw. It supplies the teeth in the pos- terior part of the upper jaw, the superior labial gland, the anterior part of the parotid gland, within which one of its small twigs anastomoses with a similar twig from the artery supplying the posterior part of this gland, and finally it gives off small vessels to the skin in this region. II. A posterior palatine branch (Ramus palatinus recurrens, Schlemm) that goes to the posterior palatine teeth, III. Small branches to the anterior palatine teeth. IV. Branches to the teeth in the anterior end of the jaw. V. Branches to the nasal gland. The other slightly stouter terminal branch of the facial carotid artery (Art. carotis cerebralis, Schlemm) runs in an irregular arch along the upper, inner, and hinder sides of the orbit to the optic foramen, where it anastomoses with the ophthalmic artery. In this way the blood from the left facial carotid can pass via the ophthalmic artery into the circle of Willis, and thence to the encephalic arteries and also over into the right facial carotid. In its course this artery sends out the following branches : — I. Several well-marked twigs to the part of the Harderian gland within the orbit. II. Branches to the muscles of the eye (Aa. musculares oculi, Rathke). III. Two short ciliary branches (Aa. ciliares postica^ breves, Rathke). 640 MR. c. H. o'donoghue on the IV. Two longer ciliary branches (Aa. ciliares posticfe longfe, Kathke), one of which runs forward on the inner side of the eye and the other outward on its hinder side. V. A Retinal Artery (Art. centraHs retinae, Rathke) which enters the eyeball with the optic nerve and spreads out over the retina. (B) Veins. The veins of the head of Tropidonotus natrix have recently been described in great detail by Bruner (13), who also describes a muscular mechanism in the head whereby the blood-pressvire in it i veins and sinuses may be considerably increased. It is not possible by ordinaiy dissection to make out all the small vessels given by that author, and, as my own results agree closely with' his, it will only be necessarvj for the sake of completeness, to give a brief description of the cephalic veins. The Mandibular Vein (V. inf ram axillaris, Schlemm ; Y. mandi- bularis, Bi'uner) arises from a small sinus at the anterior end of the lower jaw and runs backwards close to and on the outer side of the external carotid artery. On its way it receives veins from the trachea, tongue-sheath, muscles of the floor of the mouth, and . the pharynx. It runs into the maxillary vein immediately before the latter joins with the lateral cephalic vein to form the common jugular vein. The bases of the mandibular, maxillary, and lateral cephalic veins and the anterior end of the coinmon jugular vein are surrounded by constrictor muscles whose morphology and function are described by Bruner (loo. cit.). The Maxillary Vein (V. palatina, Schlemm ; V. maxillaris, Bruner) also commences in a small sinus which is situated just behind the premaxilla. This sinus has a double anastomosis with the similar one on the other side. From this point it runs back- wards beneath the naSal cavity along the floor of the orbit and then above the palate to join the mandibular vein at the angle of the lower jaw. During its course it receives : — I. The Rostral Vein (Y. rostralis, Bruner), which enters at the level of the anterior anastomosis and brings the blood from a venous network at the front end of the snout. It also receives nasal veins (Y. nasales esternse, dorsalis, and ventralis, Bruner) from the nasal gland. II. The Suhnasal Veiii (Sinus subnasalis, Bruner), which enters at the level of the posterior anastomosis. It drains a somewhat complex system of subnasal sinuses, which a,nastomose with one another at the posterior end of the nasal cavity and which receive also the palato-pterygoid vein. The Palato-pterygoid Vein (Sinus palato-pterygoid eus, Bruner) runs from near the anterior to near the posterior end of the skull on the inner side of the palatine and pterygoid bones, and flows into the anastomosis between the sub- nasal sinuses. CIRCULATORY SYSTEM OF THE GRASS-SNAKE. 641 III. A vein that joins it at the anterior end of the orbit and forms an anastomosis between it and the orbital sinus. IV. An Inferior Fcdjyebrcd Vein (V. palpebralis inferior, Bruner) that also enters at the front end of the orbit and runs backwards in the lower eyelid to its j^osterior end, where it enters the orbital sinus together with the superior palpebral vein. V. Several small veins from the orbital siniis just posterior to the orbit. YI. An Oblique Palatine Vein (Y. palatina obliqua, Bruner) that runs obliquely forward beneath the skull at the level of the hypophysis to join the similar vessel of the opposite side and form a median palatine sinus which runs forward for a short distance. Before reaching the middle line it gives off a palato-cerebral vein. The Palato-cerehral Vein (Y. palato-cerebrales, Bruner) runs dorsally around the skull and enters the median cerebral vein soon after this leaves the cranium. The Orbited Sinus (hSiuus orbitalis, Bruner) is a fairly large well- defined sinus occupying the inner and hinder parts of the orbit. At the outer end of the hinder part it is prolonged outside the orbit beneath the Harderian gland. It receives the following branches :■=- I. A small vein at its anterior end which comes from the nasal gland. II. The vein joining it to the maxillary vein which also enters at the anterior end. III. The Siqyerior Falj^ebral Vein (Y. palpebralis superior, Bruner), which arises at the anterior end of the sinus and runs backwards in the upper eyelid, re-entering the sinus at its posterior end and receiving just before it does so the inferior palpebral vein. lY. The Secondary Anterior Cerebral Fern (Sekundare Yerbind- ung der v. cerebralis media mit der v. cerebralis anterior, Grosser und Brezina) which runs from the posterior internal corner of the orbital sinus backwards inside the skull into the median cerebral vein just as the latter is leaving the cranial cavity. The anterior segment of this vessel is formed by a part of the original anterior cerebral vein. The Lateral Cephalic Vein (Y. capitis lateralis. Grosser und Brezina ; Y. jugularis interna, Bruner) arises from the posterior prolongation of the orbital sinus and runs inwards and backwards to the side of the internal carotid artery. It passes backward closely accompanying this artery to the posterior end of the head, where it bends round to the ventral side and unites with the maxillary and mandibular veins to form the common jugular vein. On its course it receives : — I. A vein from the Harderian gland. II. The Median Cerebral Vein (Y. cerebralis media, Bruner). — This runs from the longitudinal cerebral vein on the dorsal side 642 Mu. c. n. o'donoghue on tue of the brain, outwards, and around the posterior face of the optic lobes to the ventral side of the brain. Here it goes forward and leaves the skull by the foramen for the trigeminal nerve. Out- side the skull it bends sharply backwards and joins the lateral cephalic vein as the latter reaches the internal carotid ai'teiy. The last part of this vessel outside the skull is a secondary con- nection (V. cerebralis media secundaria, Grosser and Brezina) developed between the median cerebral vein, which originally opened into the internal jugular, and the lateral cephalic vein. During its course it receives : — A. The Dorsal Cephalic Vein (V. capitis dorsalis, Bruner), which arises from its dorsal side within the skull and passes outward through a special foramen. It runs backwards between the pro-otic and squamosal bones, receiving one or two cutaneous veins, and then bends laterally and enters the lateral cephalic vein. B. The Secondary Anterior Cerebral Vein, which runs on the floor of the cranium and joins it to the orbital sinus {vide sup)ra). C. The Palato-cerebral Vein, which connects it with the oblique palatine vein {vide siopra). According to Bruner there is a,lso an external secondaiy anastomosis with the anterior cerebral vein, as well as the internal one described above. I have been unable to find this vein by dissection. III. A large vein from the parotid gland and the muscles of the head, which closely accompanies the maxillary artery and enters the lateral cephalic vein close to the place where the maxillary artery leaves the internal carotid. lY. A Dorsal Cejjhalic Vein {vide sup7-a), which joins it to the median cerebral vein. V. The Posterior Cerebral Vein (Y. cerebralis posterior, Bruner), which runs from the end of the longitudinal cei-ebral vein a little behind the posterior end of the optic lobes obliquely outwards over the medulla oblongata and leaves the skull by the foramen magnum. Just before leaving the cranium it gives off a S2nnal vein Avhich runs caudally on the ventral side of the spinal cord, where it unites with the similar vessel from the other side. YI. One well-marked and several smaller veins from the muscles of the posterior end of the skull. YII. A Cervical vein which returns blood from the muscles of the neck. The Longitudinal Cerebral Vein ( Y. longitudinis cerebri, Bruner) is a vessel running backwards along the mid-dorsal aspect of the brain from between the olfactory lobes. At the postei'ioi' end of the optic lobes it gives off the median cerebral veins, and a short distance further back divides to foi-m the posterior cei'ebral veins. CIRCULATOEY SYSTEM OF TUE GRASS-SNAKE. 643 VI. List of References. (1) Beddaed, F. E. — Contributions to our Knowledge of the Circulatory System in the Ophidia. Proc. Zool. Soc. 1904, vol. i. p. 381. (2) Beddaed, F. E. — Notes upon the Anatomy of Certain Snakes of the Family Boidfe. Proc. Zool. Soc. 1904, vol. ii. p. 107. (3) Beddaed, F. E. — Contributions to the Anatomy of the Ophidia. Proc. Zool. Soc. 1906, p. 12. (4) Beddaed, F. E. — Contributions to the Knowledge of the Vascular and Respiratory Systems in the Ophidia, and to the Anatomy of the Genera Boa and Corallus. Proc. Zool. Soc. 1906, p, 499. (5) Beddaed, F. E. — A Comparison of the Neotropical Species of Corallus, C. cookii, with C. madagascariensis ; and on some Points in the Anatomy of Corallus caninus. Proc. Zool. Soc. 1908, p. 135. (6) Beddaed, F. E. — A Contribution to the Knowledge of the Encephalic Arterial System in Sauropsida. Proc. Zool. Soc. 1905, vol. ii. p. 59. (7) Bemmelen, J. F. van.— Entwikkelung en metamorphose der kieuw of vesei'alspalten en deraortenbogen bij Embryonen van Trojndojiotus natrix en Lacerta, muralis. Kon. Akad. V. Wet. Amsterdam, Afd. Natuurk., 1885. (8) Bemmelen, J. F. van. — Die Visceraltaschen und Aortenbogen bei Reptilien und Vogeln. Zool. Anz., 1886. (9) Bemmelen, J. F. van. — Beitrage zur Kenntnis der Halsgegend bei Reptilien. 1. Anatomischer Teil. Mededeelingen tot de Dierkunde. Ti]dschrift van het Genootschap Natura artis magistra te Amsterdam, 1887. (10) Beandt, F. — Ueber einen eigenthumlichen, spater meist obliterirenden Ductus Caroticus der gemeinen Kreuzotter (Pelias berus). Melange biologique, Bd. v., 1865. (11) Beennee, a. — Tiber das Verhaltnisse des N. laryngeus inferior vagi zur einigen Aortenvarietfiten des Menschen und zu dem Aortensystem der durch Lungen athmenden Wirbelthiere Uberhaupt. Archiv fiir Anat. und Phj'siol., Anat.-Abt., 1883. (12) Beunee, H. L. — On the Cephalic Veins and Sinuses of Reptiles, with a description of a mechanism for raising the venous blood pressure in the head. American Jour, of Anat., 1907. (13) Dendy, a. — The Intracranial Vascular System of Spheiiodon. Phil. Trans., 1909. (14) De Veiese, B. — Sur la Signij&cation morphologique des Arteres cerebrales. Archives de Biologie, xxi., 1905. (15) EcKEE, A. — Der feinere Bau der Nebennieren beim Menschen und den vier Wirbelthiei'klassen. BraunschAveig, 1846. (16) Feixsch, G. — Zur vergleichenden Anatomie der Amjohibien herzen. Archiv fiir Anat. und Physiol., 1869. 644 MR. C. 11. O DONOGIIUE ON THE (17) Gratiolet, p. — Note sur le syst^me veineux des Reptiles. L'Institut xxi., 1853. (18) Greil, a. — Beiti-tige zui^ vergleichenden" Anatomie und Entwicklungsgescluchte des Herzen und des Truncns arteriosus der Wirbelthiere. Morioh. Jahr., Bd. 31, 1903. (19) Grosser, 0., und Brezina, E. — TJeber die Enfcwicklung der Venen des Kopfes unci Halzes bei Reptilien. Morph. Jahr., Bd. 23, 1895. (20) HocHSTETTER, F. — Beitrjige zui- Entwicklungsgescluchte der Amnioten. 2. Reptilien {Tropidonotics, Lacerta). Morpb. Jahr., Bd. 19, 1893. (21) HoCHSTETTER, F. — IJeber Yarietaten der Aortenbogen, Aorten wurzehi und der von ihnen entspringenden Arterien bei Reptihen. Mox^ph. Jahr., Bd. 29, 1902. (22) HocHSTETTER, F. — Die Entwicklung des Blutgefasssystem, In Hertwig's Handbuch der Entwicklungslehre der Wir- beltiere, 1901-06. (23) HoFFMANisr, C. K. — Schlangen und Entwicklungsgeschichte der Schlangen. Bronn's Thier-Reich, 1890. (24) HoFMANN, M. — Zur vergleichenden Anatomie der Gehirn- und Rlickenmarksarterien der Vertebraten. Zeitschrift fiir Morph., Bd. ii., 1900. (25) HoPKiNSON, J. P. , and Pancoat, J. — On the Visceral Anatomy of the Python (Cuvier) described by Daudin as the Boa reticulata. Trans. American Phil. Soc, 1837. (26) Jaquart, H. — -Memoire sur les Organes de la Circulation chez le serpent Python. Ann. des Sci. Nat., Tome iv., 1855. (27) JoURDAiN, S. — Recherches sur la vein Porte Renale. Ann. des Sci. Nat., 1859. (28) KiNGSLEY. — Guides for Vertebrate Dissection. New York, 1907. (29) Danger, A. — Tiber die Entwicklungsgeschichte des Bulbus cordis bei Amphibien und Reptilien. Morph. Jahr., Bd. 21, 1894. (30) Rathke, H. — Entwicklungsgeschichte der Natter. Konigs- berg, 1839. (31) Rathke, H. — Ueber die Oarotiden der Schlangen. Denkschr. Wien. Akad. Wiss., Math.-nat. Kl., Bd. xi., 1856. (32) Rolleston, G., and Jackson, W. H. — Forms of Animal Dife, 2nd Ed., p. 70. Oxford, 1888. (33) Sabatier, a. — Etudes sur le coeur et la circulation centrale dans la serie des Yertebres. Ann. des Sci. Nat., Tome 18, 1873. (34) Sabatier, A. — Observations sur les transformations du Systeme aortique dans la serie des Yertebres. Ann. des Sci. Nat., Tome 19, 1874. (35) ScnLEMM, F. — Anatomischer Beschreibung des Blutgefiiss- system der Schlangen. Treviranus's Zeitschiift fiir Physiologie, Bd. 3, 1827. CIRCULATORY SYSTEM OF THE GRASS-SNAKE, 645 (36) Tandler, J. — -Mikroskopische Injectionen mit kaltflussiger Gelatin. Zeitscli. f. wiss. Mikros. v. f. Mikros. Tech., 19U1. (37) West, G. S. — On the Buccal Glands and Teeth of certain Poisonous Snakes. Proc. Zool. Soc. 1895, p. 812. (38) West, G. S.— On two little-known Opisthoglyphous Snakes. Jour. Linn. Soc, Zool. 1896. VII. EXPLANATION OP PLATES LXX.-LXXII, Tiettering. A.A.L. Left Aortic Arch. A.A.K. Kight Aortic Arch. A.A.V. Anterior Abdominal Vein. A.C.A. Anterior Choroid Artery. A.V. Azygos Vein. A.V.A. Anterior Azygos Vein. A.V. P. Posterior Azygos Vein. B.A. Basilar Arterj'. B.A.L. Loop of Basilar Artery. C.A. Left Common Carotid Artery. Ca. A. Caudal Artery. CCA. Cerebral Cai-otid Artery. CCAut. Anterior branch of the Cerebral Carotid Artery. CCM. Median branch of the Cerebral Carotid Artery. CCP. Posterior branch of the Cerebral Carotid Artery. Cejj. V.D. Dorsal Cephalic Vein. Cep. V.L. Longitudinal Cephalic Vein. Cer. A. Cervical Artery. Cer. V. Cervical Vein. , Cer. Sem. Cerebral Hemispheres. ■ CW. Circle of Willis. CJ.L. Left C!ommon Jugular Vein. C.J.R. Right Common Jugular Vein. CV. Caudal Vein. C.V.L. Longitudinal Cerebral Vein. C.V.M. Median Cerebral Vein. CV.P. Posterior Cerebral Vein. C.V.S.A. Secondarj' Anterior Cere- bral Vein. D.A. Dorsal Aorta. E.C.A. Anastomosis of the two Ex- ternal Carotid Arteries. E.C.L. Leff External Carotid Artery. E.CR. Right External Carotid Artery. E.M.A. Eye Muscle Artery. E.R.L. Left Efferent Renal Vein. E.R.R. Right Efferent "Renal Vein. JSxt. Nar. External Nares. F.A. Pat-body Artery. P.CA. Pacial Carotid Artery. P.C.P. Palatine branch of the Pncual Carotid Artery. G.A, Gastric Arterj', Gall. JBl. Gall-bladder. H.A. Hepatic Artery. Sar. Gl. Harderian Gland. H.G.A. Harderian Gland Artery. H.P.V. Hepatic Portal Vein. H.P.V.' The posterior part of the Hepatic Portal Vein, the " Veine mesenterique posterieure" of Jourdain (27). H.P.V.O. Origin of the Hepatic Portal Vein from the Renal Portal Vein. I. A. A. Internal Auditory Artery. I. C.A. Internal Carotid Artery. l.C.M. Muscular branch of In- ternal Carotid Artery. I.C.O. ffisophageal branch of Ln ternal Carotid Artery. I.M.A. First Inferior Mesenteric Artery. I.M.A.2. Second Inferior Mesenteric Artery. I.M.A.3. Third Inferior Mesenteric Artery. I.M.A. 4. Fourth Inferior Mesenteric Artery. Inf. Infundibulum. I.P.V. Inferior Palpebral Vein. I.S.A. Inferior Spinal Artery. iLid. L. Left Kidney. Eiid. a. Right Kidnej\ L.A. Left Auricle. L.C.A- Ijong Ciliarjr Artery. L.G.A. Lieno-gastric Artery. L.R.A. Anterior Left Renal Artery. L.R.M, Median Left Renal Artery. L.R.P. Posterior Left Renal Artery. M.A. Maxillary Artery. M.C.A. Median Cerebral Artery. M.C.P. Posterior branch of the Median branch of Cere- bral Carotid Artery. lied. Medulla oblongata, M.V. Mandibular Vein. Mx. Maxillary Vein. Mx.A. The double anastomosis of the Maxillary Vein. O.A.L. Left Ovarian Artery. 646 ON THE CIRCULATORY SYSTEM OF THE GRASS-SNAKE. O.A.R Right Ovarian Artery. Oc.A. Occipital Artery. Od.A. Oviducal Artery. Od.A.P Posterior Oviduoal Artery Od.S.L Left Oviducal Sinus. Od.S.R Right Oviducal Sinus. Od. V.L. Lett Oviducal Vein. Od. V.L.F Left Posterior Oviducal Vein. Od.V.R. Right Oviducal Vein. Od. V.R.R. Right Posterior Oviducal Vein. Oe.A. Q<]sophageal Artery. Oe. V. (Esophageal Vein. OLA. Olfactory Artery. Olf. lo. Olfactory Lobe. Op. A. Ophthalmic Artery. O-pt. lo. Optic Lobe. Opt. Ner. Optic Nerve. O.P.V. Oblique Palatine Vein. Orb. Orbit. O.S. Orbital Sinus. O.S.A. Anastomosis between the Orbital Sinus and the Maxillary Vein. Ov.A. Smaller Ovarian Arterj'. Ovar. Zi. Left Ovarj'. Ovar. B. Right Ovary. P.A. Parietal Artery. Fan. Pancreas. Par. Parotid Gland. Par. A. Anastomosis between a branch of the Anterior and a branch of the Posterior Parotid Gland Arteries. Par. A.A. Anterior Parotid Gland Artery. Par. P.A. Posterior Parotid Gland and Superior Maxillary Artery. Par. V. Vein from Parotid Gland and also from neigh- bouring muscles. P.O. Primary Carotid Artery. P.C.V. Post-Caval Vein. Pel.V. Vein probably corres- ponding to the Pelvic Vein of Lacertilia. Fit. Fos. Pituitary Fossa. Pl.C.V. Palato-cerebral Vein. Pl.P.V. Palato-pterygoid Vein. Pul. A. Pulmonary Artery. Pul. V. Pulmonarjf Vein. P.V. Parietal Vein. R.A. Right Auricle. Ret. A. Retinal Artery. R.P.L. Left Renal Portal Vein. R.P.R. Right Renal Portal Vein, R.R.A. Anterior Right Renal Artery. R.R.M. Median Right Renal Artery. R.R.P. Posterior Right Renal Artery. Rt. A. Rectal Artery, R.V. Rostral Vein. S.C.A. Short Ciliary Artery. S.M.A. Superior Mesenteric Artery. S.N. A. Anastomosis of the Sub- nasal Sinuses. S.N.V. Sub-nasal Vein. Sp. A. First Spinal Artery, Sp. V. Spinal Vein. Spin. Cor. Spinal Cord. Spl. Spleen. S.P.L. Lett Supra-renal Portal Vein. S.P.R. Right Supra-renal Portal Vein. S.R.L. Left Supra-renal Vein. S.R.R. Right Supra-renal Vein, Sup. Lab. Ql. Superior Labial Gland. Sup. Ren. L. Left Supra-renal body. Sup. Ren. R. Right Supra-renal body. Th. A. Thyroid Artery. TJu/. ai. Thyroid Gland. Ton. Mns. Tongue Muscle. Ton. Sh. Tongue Sheath. V. Ventricle. V.A. Vertebral Artery, Pip Pis PiAXE LXX. 1. Diagram of the Heart and Blood-vessels in the anterior part of Tropidonotus natrix. The two main arteries and veins on the ventral side of the head are indicated, but their precise distribution is dealt with on Plate LXXII. Between the lines X.X. and Y.Y. a portion of the circulatory system is omitted, as the relations of the Hepatic and Pulmonary Vessels and the Dorsal Aorta are similar throughout the whole length «f the liver. The part omitted is about the same length as the distance from the thyroid gland to the line X.X. For the purposes of diagram the liver is repre- sented as pulled out to the right of the animal. The positions of the Liver and Lungs are indicated by simple outline and those of the Gall-bladder, Spleen, and Pancreas by shaded areas. 2. Diagram of the posterior Blood-vessels in a female Tropidonotus natrix. The positions of the kidneys and ovaries are indicated by simple outline and those of the supra-renal bodies by shaded areas. For the purposes of diagram the intestine is represented as pulled out to the right of the animal, the fat-bodies as external to the oviducts, and the oviducts as external to the ureters. Owing to the fact that the anterior end of the Fat-body is supplied from the right ovarian arter^^, while a little further ON THE COURTSHIP OF THE REDSHANK. 647 back it is fed from the left ovarian arterj', it is necessary to divide the Fat-body between the two gonads. Tlius it comes tliat the anterior part of the Fat-body is represented on the right side of the animal and the posterior part on the left. As the Anterior Abdominal Vein runs in the Fat-body it is also divided and is so represented in the diagram, in which its two ends are joined by a dotted line. Plate LXXI. Diagrams of the Blood-vessels in the head of Tropidonotus natrix. Fig. 3. Diagram of the Ci-anial Arteries seen on the ventral surface of the brain. The brain is represented as removed from the skull leaving behind, however, the Pituitary Kod3^ Pig. 4. Diagram of the arteries of the head seen from the dorsal side. The brain is removed, but some of the cranial arteries are left behind. The positions of the glands of the head are indicated by shaded areas. PiATE LXXII. Fig. 5. Diagram of the vessels on the ventral side of the lower jaw after the removal of the superficial muscles. Fig. 6. Diagram of the Veins and Sinuses in the head seen from the dorsal side. The deeper vessels are represented in lighter shading and with a dotted outline. The position of the brain is indicated in simple outline, while the positions of the glands are indicated by shaded areas. For the purposes of diagram the Superior Palpebral Vein is omitted. It runs from the uppei' anterior part of the Orbital Sinus above the Maxillary Vein and joins the Inferior Palpebral Vein as the latter enters the Orbital Sinus. In dis- secting out the vessels of the orbit it is almost alwaj's removed with the upper eyelid. 34. A First Account of tlie .Courtship of the Redshank {Totamis calidris L.). Bj Julian S. Huxley, Lecturer of Balliol College, Oxford. . [Received February 2, 1912 : Read April 23, 1912.] Index. Page 1. Introduction 647 2. Locality 648 3. The Courtship proper 649 4. Other habits of the Pairing-Season : (a) The Love-flight 651 (i) The Combats of the Males 653 (c) Calling from a conspicuous perch 652 5. Discussion ..; 652 1. Introduction. While staying last spring in a lonely corner of North Wales it was my good fortune to come across a number of rare and interesting birds. But great as was the pleasure of seeing, for the first time, such comparatively uncommon species as the Grey Plover and Black-tailed God wit, it was far surpassed by that of being able to study, under the most favourable conditions, the natural behaviour and home life of some of the commoner shore- birds. Of these I was particularly foi^tunate with the Kedshank, 648 MR. J. s. nuxLF.y on the find was able to see the whole of the courtship ami pairing. When I say the whole, I do not mean that I saw every detail, nor that every detail I saw is clear to me. But I mean that I know wliat is the general course of events, and can interpret the bii'ds' behaviour more or less consistently. On returning to civilization and libraries, to my surprise 1 could (ind ver}^ little on the subject : the observations recorded wore either fragmentary or inaccura.te. It was not until most of this paper was written that I discovered a fairly coiuplete account by Selous *. This had remained undiscovered owing to the absence of any I'eferenco to Kedshanks in the title of the paper or in the index of the volume. I luive thought it worth while to publish my observations, however, since they diller in several points from those of Belous. Meanwhile, 1 fully realize their incompleteness, and recogni/.o that they cannot as yet be properly used in any general discussion of the theory of sexual selection. I hope to continue my own observations when opportunity offers, but ventnre to publish this general outline at once as a stimulus to other bird-watchers and naturalists. 2. LocALrrv, etc. Befor-e passing on to the birds' actions I nmst tlrst just n\ention the theatre where I saw them played. This was p;irt of a svnall estuary in the northern half of Cardigan Bay : an ai-in runs out on one side at right angles to the I'iver, thus giving during high spring tides a land-locked sheet of water nearly a mile long and half a mile wide; dvu-ing neaps, even high tide failed to cover it. Numbers of Redshanks and other birds fre- quented this expanse, and especially its head or most landward end, where they were close to a thick bed of reeils and tussocks ; the nuid here was scarcely ever covered by the tide, though kept always moist by a little stream. At one side, of the head was a low ridge of grass-covei'ed dunes, about five feet high and thirty or forty feet long, Avith level gi'ound behind them. Tluis, by ci'awling over the flat, I could get up to the d.vnies into an excellent position for viewing the whole top of the liay ; every bird was easily seen against the wet mud. So I kept watch with the uaked eye until some disturb- ance or unusual behaviour attracted attention ; then, being armed with a telescope magnifying 30 diameters (for the loan of which I have to thank my brother, Mr. N. T. Huxley), I focussed this on the spot, and could see the minutest details of attitude and behaviour in the nearer birds, and even in those on the far side of the bay could quite well interpret wliat I saw. I made a number of notes on the spot, and usually within twenty-foiu- hours end)odied what I had seen the day before in a letter to an ornithological friend. * E. Sdous, "Olisorvatioiis toudino- to tlirow lidit on tlio Question of SL'xnal Solectjoii in ninls," etc., I'avt I : Zoologist (1) x., I'JOii, jin. i:(tl--2U1. COUKTSUIP OF THE REDSUANK 649 3. The Courtship proper. I will begin with an account of the typical course of the courtship and pairing, such as I have seen repeated, with but slight variations, a considerable number of times. Among the forty or fifty birds that usually would be quietly feeding on the flats, walking or running in short starts from mouthful to mouthful, a disturbance would every now and then be visible — two birds running, one pursuing the other. These two are cock and hen. A cock takes a fancy to one of the hens, leaves his feeding, and starts running towards her. She at once runs away from him, and there ensues a regular game of follow-my-leader. The hen never goes far in a straight line ; she usually runs in a series of curves, often doubling sharply back, and sometimes describing a complete circle or even a figvire of eight. Where she goes the cock goes after her, following exactly, but some yards behind. The couple would be ridieulous enough in their devious course, with heads somewhat down and V^ # " i. f \W L.m. t.m.. Section through the muscnlai- laj'er of the cortex, taken at a point further back than that represeuted in text-fig. 97. l.m. Bundles of longitudinal fibres, t.m. Transverse fibres. The water vascular system of this tapeworm is in more than one respect remarkable. In transverse sections through the ripe proglottids only two vessels are as a rule visible, one on each side of the body. The most careful examination often failed to reveal the presence of another, even minute, tube. Nor did longitudinal sections show any trace of this second vessel. It is not, however, really absent in this worm, as I found it to be in Thysanotcenia lemuris, for it is present as a very minute tube in some segments. The vessels which are obviously present NEW CESTODES FKOM THE TASMANIAN DEVIL. 687 are of very large size, and it appeared to me that that on the pore side of the segment was a little the larger of the two. This, then, is the first important point about these vessels, i. e. that there is a very lai-ge ventral tube on each side of the body and a very minute dorsal tube. In following out a series of transverse sections, it is seen that the lumen of the large ventral water- vessels is occasionally occluded by a delicate diaphragm-like sheet of membrane which is abundantly nucleated. There is no question of a narrowing of the calibre of the tube, but of an actual membrane which extends partly across it here and there. In sagittal sections the existence of these membranes stretch- ing partly across the lumen of the water vascular tube is quite obvious. They occur, moreover, on both sides of the body, that is to say in the case of both ventral tubes. The reason for emphasizing this fact will be apparent later. In the longitudinal sagittal sections referred to it will be seen that there are several of these membranes which stretch a good way across the water-vessel, and though two membranes arising from different sides of the vessel do not actually meet, the edge of each stretches beyond the edge of the other, so that the tube would appear, when viewed in the direction of its length, to be entirely occluded. _ It is note- worthy that these diaphragms, so to speak, arise indifferently from both sides. The exact arrangement will be plain from the annexed drawing (text-fig. 99, p. 688). I have noticed that in some of the posterior proglottids the lumen is actually occluded once in each proglottid. "The two ends of two oppositely projecting membranes are connected by a continuous though very thin membrane which connects the thicker extremities of the lateral projections. This median part appears to be nowhere deficient, and the water vascular system is thus divided up in these regions of the body into a series of chambers. I presume that these numerous membranes stretched across the large vessels correspond to what has been figured in other tape- worms as valves. I have emphasized the fact that they occur on both sides of the body, because they carry on the pore side the o-enital ducts which actually perforate theii- substance and lie in their thickness. This perforation is limited to the larger of these valve-like structures which arise from the outer side, and it has been produced I imagine by the extension round the ducts of the water-vessel, which is of much greater diameter in the posterior than in the most anterior segments. In the case of these latter, as already mentioned, the generative ducts pass to one side of the water-tube. Another important feature in the water vascular system of this tapeworm is the total absence of transverse vessels uniting the longitudinal trunks in such segments. This state of affairs is not unknown among other Cestoidea — it occurs, for example, in Hymenolepis acuta* — but it is not common. Nor is this lack of * V. Jauicki, Zool. Aiiz. Bd. xxvii. 1904, p. 776. 688 DR. F. E. BEDDARD ON transvei'se vessels compensated by any network of excretory tubes pervading the medulla, such as is met with in the genera Inermi- ccqysifer and Zscliokkeella *. Text-fig. 99. Sagittal section to illustrate ventral water vascular tube. V. Projecting valves, v.d. Vas deferens, v.a. Vagina. T. Testes. * Beddard, P. Z. S. 1912, p. 596, and literature quoted there. NEW CESTODES FROM THE TASMANIAN DEVIL. G89 The Genercdim organs of this tapeworm begin to be recog- nisable very early in the body, only a segment or two behind the head. But it is a long way back before the ovaries are ripe. As in the vast majority of Cestoidea, the testes ripen earlier than the ovaries. This being the case, the testes are recognisable earlier as distinct bodies, and only cease to be so clear in the more pos- terior segments, where the uterus is gorged with eggs. A remark- able point of interest in the generative organs is the fact that the duct leading to the exterior, or to be more exact the formative mass of cells which will be both vagina and cirrus sac, is seen to alternate in position in relation to the single water vascular tube and the nerve-cord. In all segments the actual opening is on the same side of the body, but the generative duct passes towards it either between the water-tube and the nerve-cord or outside of both ; in the latter case only to one side — there is no alternation between a dorsal and a ventral position. We shall see, when the cirrus sac and vagina come to be described later, that there is also variation in the exact relationship in position between these two. The ovaries are lai'ge and consist of two wings, which are symmetrical or very nearly so, the middle jDoint between them being the middle line of the body, where are situated the shell- gland etc. The ovaries are posterior in the segment, and behind them lie the vitelline glands. These latter are of much the same shape as the ovaries, and in rather immature segments differ only from them by their rather darker staining with htematoxylin. They also form two wings symmetrical with their middle point, and are in contact with the ovaries in front. The two glands are composed of many lobes, which reach as far as the testes at the sides ; altogether an ovary occupies fully half of the segment, and rather more when it is fully mature. Immediately in front of it are the sperm-duct and vagina. At the sides are to be found the testes, v/hich are also dorsal to it. The vagina offers no very remarkable character. It has at first a contracted lumen, which widens out for a consideiable space, and then contracts again before it suddenly opens into the rather pear-shaped receptaculum seminis. The course of the vagina is quite straight between its two ends and oblique in direction. The narrow part of the vagina which opens into the receptaculum is of some length, but shorter than the wider part. The vagina shows quite the same characters in the most mature proglottids. The receptaculum seminis is full of spermatozoa, and very frequently contained ova at its wider end, close to where the oviduct opens into it. The receptaculum and the vagina lie anteriorly in the segxnent in fi'ont of the ovary, but behind the uterus. The uterus of this worm is persistent and found as a large cavity extending right across the segment in the most mature proglottids that I have examined. It begins as a small rounded cavity lying in the front part of each segment. 690 DR. F. E. BEDDARD ON The testes of this worm are chiefly massed at the two sides of the proglottid, but these two masses are connected by a string of testes which pass dorsally along the proglottid. They therefore nearly surround the ovary and the female organs generally, so far partly resembling Cyclorchis. In transverse sections the testes appear circular in section ; but sagittal sections such as that represented in text-fig. 99 (p. 688) show that the form of the testes is that of a flat plate, for in those sections they appear more linear in shape. These sections also show quite plainly that each testis Text-fig. 100. v/;; A portion of the coil of the sperm-duct {sj>.) gorged with sperm. c. Interstitial prostatic cells. lies in a spa ce, unless, indeed, the appearances produced are due to shrinkage through reagents. In any case, however, the testes of these worms frequently present the appearance of being sur- rounded by the spermatozoa which they produce, thus showing that a chnik exists or can be formed for their reception when pressed out of the testis. The testes are very numerous and quite crowded together, — so much so that the delimitations between successive sea;ments so far as concerns these organs NEW CESTODES FROM THE TASMANIAN DEVIL. 691 are not at all visible in sagittal sections ; they appear as a continuous mass. The sjKrm-cluGt I'uns straight for a short way after it has left the cirrus sac. It forms a copious coil (text-fig. 100) occupying the middle of the body and lying partly dorsally to the receptacuhmi seminis and rather nearei- to the pore side of each proglottid. The rest of the coil — that part which is nearest to the cirrus sac — is Text-fig-. 101. Sagittal section showing in consecutive segments the varying relations of cirrus sac (the larger tube) and vagina. to the poi-e side of the receptaculum, and therefore ventral in position. In fact, regarded as one coil, this region appears oblique in direction in transverse sections of the progiottids. Between the individual loops of the sperm-duct are cells which quite fill up the interstices, and are thus numerous in proportion to the width 692 DR. r. E. BBDDARD ON of tliose interstices. Tlie cells are rather clear, with well- stained nuclei. They correspond exactl}^, as it seerns to me, with the prostatic cells of Inermicapsifer and Zschokkeella, dealt with in these genera by v. Janicki* and myself t, and which appear to occur also elsewhere. In immature segments the cells bulk more largely than the coils of the sperm-duct. But the reverse is the case in the mature proglottid. The genital ducts open into a common cloaca genitalis, which in its turn opens on to the exterior. The cloaca genitalis is of some depth, but it is not borne upon any process of the body. Into it open, close together, the vagina and the cirrus sac, whose mutual relations I have investigated by means of sagittal sections (text- fig. 101) through portions of the strobila. There is a con- siderable variation in these relations. Although the vaginal pore is apparently never in front of the male pore, it is not always directly behind the male pore. The commencing vagina lies obliquely behind the commencing cirrus sac — the direction of the obliquity being now dorsal now ventral in this segment and in that. There is, in fact, an irregular alternation from segment to seg- ment ; sometimes the two tubes are not merely oblique, but actual ly dorsal or ventral to each other as the case may be, lying there- fore side by side in sagittal sections. This recalls to mind the alternation that occurs in certain (but not all) species of the genus Moniezia, where the vagina may be dorsal or ventral to the cirrus sac. But in this latter genus the alternation is of the right and left set of generative organs of a single segment. The cirrus sac of this tapeworm is large and has the very common flask-shape. The neck-region has very thick circular muscvilar walls, forming a shea,th which thins out over the more distended region of the sac. In less mature segments (in which, however, the testes are fully developed, though with no mature sperma- tozoa) the cirrus sac is elongated, gradually diminishing in breadth towards the external pore ; there is no marked division into neck and flask. It is very long and extends towards the middle line of the body, a little beyond the water vascular tube of its side of the body, or at least reaches the internal side of that tube. The cirrus runs straight from end to end of the cirrus sac and anteriorly presents a monilifoi'm appearance, which is due to suc- cessive dilatations of the lumen of the cirrus. This region of the cirrus is both preceded and succeeded by a perfectly straight section of that tube with very narrow lumen and thick walls. In later segments the cirrus sac acquires the flask -shape ah'eady referred to. Coincidently with this is an actual shortening of the length of the entire sac and a coiling of the cirrus within it. The cirrus sac in these and in subsequent segments hardly reaches beyond the outer edge of the water vascular tube. It seems clear therefore that the shortening is due to an actual contraction of * Dciikschr. Ges. Jena, xvi. 1910. t P. Z. S. 1912, p. G02. NEW CEiSTODES FROM THE TASMANIAN DEVIL. 693 the length caused by a bulging of the walls of the cirrus sac due in its turn to the rapid growth and consequent coiling of the cirrus. When the cirrus sac is in this fully formed condition, the cirrus itself is differentiated more thoroughly into those regions less markedly indicated in earlier stages. The sperm -duct enters the cirrus sac at the apex and its lumen contracts to a fine line for a short space near to its entry. This is particularly obvious in the last few segments of the body, where the sperm-duct has become much dilated before its entry into the cirrus sac, and thus offers a greater contrast to this exceedingly nai^row region. In these more mature cirrus sacs the flask-shape has been acquired, as already mentioned. But the neck of the flask is much longer than the body part. The latter is so fully occupied by the coils of the cirrus itself that there is but little of the inter- stitial packing tissue to be seen. At its entry into the cirrus sac and for a considerable time thereafter the duct is thick-walled, with a very narrow lumen, and much coiled. This region of the cirrus is succeeded by a not very long but coiled tract, which is much wider and has thinner walls. The lining membrane bears numerous spinelets. Finally, the distal region of the cirrus is again thick- walled and with a narrow lumen : it opens into the genital cloaca without any alteration of character. In the most posterior segments of the body the greater part of the cirrus sac is filled with sperm, the posterior region alone showing a group of coils of the cirrus. Whether the anterior part of the cirrvis has become ruptured, as it appeared, or has been simply enoi-mously expanded and its walls reduced to extreme tenuity by the enclosed sperm, I am unable to say. The systematic position of this tapeworm is difl&cult to fix with any confidence. The generative system, and, indeed, the intei-nal anatomy generally, presents no differences of importance from many Tetracotylea ; and there are, indeed, no reasons so far why the worm should not be placed in the Anoplocephalidse, which family, as has been pointed out, contains nearly all the Marsupial tapeworms. On the other hand, the very much developed layers of longitudinal muscles in the body-wall suggest the family Acoleidae. The difficulty, however, of accurately placing the worm lies in the peculiarities of the scolex. There is no doubt that it contrasts very considerably with the general form of the scolex among the Tetracotylea in a number of points, of wdiich the principal ones are: — (1) its large size, both relatively to the body and actually ; (2) the presence of numerous grooves which cannot be, at any rate, entirely artefact, as they are converted here and there into tubes running witliin the thickness of the head ; (3) the relatively minute size of the four suckers and the fact that two of them and two onlj^ are furnished with hooks*. * Furthermore, these hooks are distinctly hoHow at their broader end, "like a Ruminant's horn," as Shipley (Willey's Zool. Res., Entozoa, 1900) notes of Callio- bothrium, one of the Tetraphyllidea. 694 DR. F. E. BEDDARD ON These characteristics are collectively different from anything met with among the Tetracotylea that is known to me. They are not, however, inconsistent with the conditions known to occur among the Tetraphyllidea, if we may admit the grooves upon the scolex to I'cpresent the bothria of such tapeworms. The suckers would then correspond with the small accessory suckers so frequently possessed by these latter worms, and their small size relatively to the scolex would be thus intelligible. The apparently numerous bothria not reducible (by me) to symmetiy is suggestive of a type like Phyllohothrmm * slightly modified, or perhajDS Peltidocotyle t. We cannot, however, place Dasyurotcenia among these Tetraphyllidea on account of the Tetracotylean character of its yolk-gland. But with reference to this gland it may be borne in mind that it is in structure much more diffuse than is visual with the generally solid vitelline gland of the Tetracotylea, The genus may be thus defined : — DasyurotaBuia, gen. nov. Stoutly huilt worms with large scolex bearing four small suckers, of which the inner two hear hooks. No rostellum, hut anterior end of scolex, including hooked suckers, retractile. Segments very short. Inner layer of (longittidinal) mtiscles very thick, consistivg of fo%ir to six roivs of hundles of fibres. Ventral excretory tabes large, with numerous valves not communicating with each other in the strohila. Dorsal vessels minute, not always visible. Genital pores uidlateral. Testes numerous, chiefly lateral, anterior, and dorsal. Vas deferens tvith a large coil in middle of segment sur- rounded' by prostatic cells ; cirrus sac large, cirrus with spinelets. Ovaries loith two icings, mediant and posterior and ventral in position, in front of vitelline gland, which is also symmetrical. Shell-gland median, dorsal. Beceptacicluin sem,inis present, nearly median, ventral. Uterus sac-like, persistent, fills nearly whole of ripe 2)7-oglottid. Eggs ihm-shelled. Hab, Marsupials. The species I term " rohwsta " on account of its very stout build. It is, however, quite impossible for me to venture upon an enumeration of the peculiar specific charactei'istics for the present. This genus and species cannot, as I think, be identified with any other form that has been desci-ibed from an Australian Marsupial, From the present genus we only know a species described by myself § a little time since as Anoplot(Bnia dasyuri. Nor can I identify it with any of the genei'a enumerated at the beginning of this paper in other Marsupials. In fact, Bertiella is * Bfonu's ' Tbierreicli,' Bd. iv. Cestoiden, pi. xli. fig. 10. f Ihid. pi. xliii. fig. 1. X Middle line of female app.iratus only slightlj' displaced towards pore side. § P. Z. S. 1911, p. 1003. NEW CESTODES FROM THE TASMANIAN DEVIL. 695 the only one of these genera to which it bears any Kkeness in the reproductive system, and from this genus the characters of the scolex at once distinguish Dasijurotcenia. Indeed, its inckision among the Tetracotylea ( = Tpenioidea) is not, to my mind, an obvious certainty. In any case the hooked suckers exclude it from the family Anoplocephalidse, to which nearly all the Marsupial tapeworms belong up to the present. The most salient points of anatomical interest in this worm appear to me to be the following : — (1) The immense size, relatively speaking, of the scolex and the small size in comparison with it of the suckers. The fact that two suckers are armed with hooks while the other two suckers are not so armed. (2) The great thickness of the longitudinal muscles, which consist of at least four layers of bundles each containing very many individual fibres of considerable stoutness. (3) The existence for the most pai-t of only a single water vascular tube on each side of the body, which is, moreover, in the posterior segments completely divided up into a series of com- partments, one to each segment, and whose lumen is also here and more anteriorly divided by delicate septa jutting into its cavity. Furthermore, by the fact that these tubes are not connected in successive segments by transverse vessels, as is so nearly universally the case. (4) As a remarkable structural feature, which is at present mysterious in nature, may be mentioned the isolated cavities in the medullary region of the head which have no connection with the water vascular tubes. (5) An anatomical feature of some importance is the very variable relation to each other in position of the extremities of the male and female ducts, which is correlated with an orifice upon one side of the body only. An alternation in the position of the external pore may, we know, be accompanied with difiference in the relative position of the ducts as, for example, in the double series of genital tubes of Moniezia, (6) In view of the very considerable peculiarities of structure briefly indicated in the foregoing resume, it may be worth men- tioning, as a remarkable fact, that the generative organs do not show any marked features of interest as compared with those of other tapeworms. 696 MR. K. E. TURNER ON 39. Studies in the Fossorial Wasps o£ the Family Scoliidse, Subfamilies Elidinse and Anthobosciuse. By Rowland E. Turner, F.Z.S., F.E.S. [Received March 29, 1912 : Read May 7, 1912.] (Plates LXXXI.-LXXXIII.*) Index. Geogvapliical Zoologj^ ; Page Antliohosca, Distribution of 726 Sj'steniatic : Antliohosca occipitalis sp. ii 734 Braunsomeria, gen. n 697 „ quadraticeps, ^]).\^ 698 „ atriceps, s^. n 699 Wis major, sp. n 723 „ sellotui, S'^. w 722 „ {Mesa) alicicB, s^. n 704 „ „ apicipennis, s^. n ., 707 „ „ auriflua, S'p. n 705 „ „ ineerta, sp. n 710 „ „ lonffiventris, sp. n 712 „ „ spoUata, sp. n 711 Myzine hraunsi, sp. n 700 „ combusta Sva. to Elis 724 „ constrictiventris, sp. n 701 „ satiakinensis Gvih. to AntJiobosca 740 „ stiffma, sp. u 699 „ umhr atica, sp. n 702 Odontotliynnus Cam . $ to Antlioh osca 724 Plesia contimia Cam. $ of Mysina ahdominalis Gner. 703 „ erytlironota Qaxa.. to Antliohosca 738 „ leucospila Cam. „ „ 739 „ m el anari a Cam. „ „ 736 The following notes on the Elidinje and AnthoboscinjB will, it is hoped, facilitate the study of these neglected groups, the latter of which especially has been very little understood by many aiithoi's, who have touched on it merely as describers of new species, Saussure alone having seriously studied the group. The best work on the Elidinaj has been done by the same author, but the material at his disposal was veiy limited. I am indebted to Dr.. Brauns, of Willowmore, S. Africa, for valuable assistance with many carefully collected specimens. The material available is still insufficient for a revision of the sj)ecies of Myzine, as to which much confusion still exists. The species which I have not seen are marked with an asterisk. * For explanation of tlie Plates see pp, 753-754. P.Z.S.1912.P1.LXXXI. Horace KnigM del.etlith.. West, Newman chr. FOSSORIAL WASPS. P.Z.S.1912. Pl.LXXXII. Catlierme A.M.Pearce del. West,NewitiaTLlith.. WING NEURATION OF FOSSORIAL WASPS. P.Z.S.1912. Pl.-LXXXIIL. 12 16 13 ^ * 15 Catherme A.M.Pearce del. West.Ne-wman lith. EXOSKELETAL STRUCTURES OF FOSSORIAL WASPS. STUDIES IN THE FOSSORIAL WASPS. 697 Family Scoliid^e. Subfamily Elidin^e. Beaunsomeria, gen. nov. 5 . Apterous ; mandibles acute at the apex, with a rather in- distinct tooth on the inner margin near the apex ; antenna3 twelve- jointed, the first joint of the flagellum very small and almost con- cealed by the apex of the scape. Head almost rectangular, the posterior angles slightly rounded ; eyes oval, touching the base of the mandibles, rather small, separated by a distance at least as great as their own length from the posterior angles of the head ; ocelli absent, their position indicated by large punctures. Thorax much narrower than the head ; pronotum rather longer than its greatest breadth ; mesonotum very short, almost covered by the pronotum, the tegute more or less developed ; scutellum narroAver than the pronotum, broader than long ; median segment nearly as long as the pronotum, flattened on the dorsal surface, broadened from the base to the apex. 3ides of the head and thorax and base of the abdomen thinly covered with long hairs. Abdomen longer than the head and thorax combined, shining, the apical segment long and more or less acute at the apex, stricture between the first and second ventral segments well developed. Intermediate coxre rather widely separated, posterior coxse contiguous, intermediate and posterior tibiee spinose, tarsal ungues simple. cJ . Winged ; stigma rather large, situated at about three-fifths from the base of the wing ; radial cell shorter than the stigma ; three cubital cells, the second and third small, not reaching the apex of the radial cell, each receiving a recurrent nervure ; cubital and discoidal nervures not continued beyond the cells. Medial cell of the hind wing not emitting veins from the apex. Antennre in the typical species long and slender, thirteen-jointed, the first joint of the flagellum almost concealed in the apex of the scape, the antennte much longer than thp abdomen ; antennal tubercles well developed. Head strongly convex ; ocelli present. First abdominal segment with a short petiole, the segment, including the petiole, a little longer than the se.cond segment, suddenly widened at tlie apex of the petiole. Apical segment with a re- curved spine, the apical emargination of the dorsal segment shallow. Eyes entire, not emarginate. The characters givep here for the male will doubtless be found not to apply ]bo all species of the genus ; but the important characters in the neui'ajtion separating the males from Ilyzine are the larger stigma, the blunter apex of th.e radial cell, the fact that the cubital and discoidal nervures are not continued beyond the cells as in j^Iyzine, and that no v.eins are emitted from the median cell of the hind wing, there being two veins in Myzine. Type of the geiius, Braunsomeria quadfatice2)s. Proc. Zool. See— 1912, No. XLVI. 46 698 MR. R. E. TURNER ON Braunsomeria quadraticeps, sp. 11. (PI. LXXXI. figs. 9, 10; PI. LXXXII. fig. 7.) 5 . liafo-ferruginea ; mandihulis apice, vet^tice, capite laterihus ; segmentis abdominalihus trihus basalihus nigris ; feinoribus tibiisque Jvscis, Gcdcariis albidis. Long. 8 mm. § . Head broader than long, fully half as broad again as the pi'onotum ; mandibles with a blunt tooth on the inner margin near the apex, another near the middle of the inner margin, acute, and another smaller nearer the base. The whole insect shining, with a few scattered punctures. Pronotum longer than broad, a little longer than the median segment. Dorsal abdo- minal segments bi'oadly depressed at the apex, the basal portion of the segments produced into a slightly raised lounded mark on each side ; apical segment very narrowly rounded at the extremity. (S . jS iger ; inandihidis, clypeo Qnacida mediana nigra, scapo subtus apice, tubercidis aniennalibus, macida parva fronUdi, mar- gine interiore ocidorum, linea' undidata verticcdi, pronoto macida uirinque antice et fascia lata postice, mesonoto macida quadrata, scutello fascia lata, j)ostscutello, mssopleuris fascia, segmento dor- sali primo fascia apicali, cceteris fascia apicali macula nigra utrinque, segmentisque ventralibus 2-6 fascia bisinuata ap)icali pallide ftavis ; alis hyalinis, venis testaceis. Long. 6 onm. d" . Antennfe slender, longer than the abdomen, the inter- antennal tubercles prominent. Clypeus very short and broad, very shallowly emarginate at the apex. Head very strongly convex, cheeks as broad as the eyes. Head and thorax coarsely but not very closely punctured, median segment finely and closely j.unctured-rugulose; abdomen shining, very sparsely and shallowly punctured. Pronotum shorter than the mesonotum, narrowed anteriorly, the anterior margin straight, posterior margin very feebly arched. Median segment steeply sloped posteriorly, nob truncate. Petiole of the basal abdominal segment occupying less than half the length of the segment, the remainder of the segment slightly inflated ; the first segment, including the petiole, only a little longer than the second. The segments not constricted ; hypopygium forming a long recurved spine ; apical dorsal segment convex, shallowly emarginate at the apex. Stigma large, twice as long on the costa as broad, nearly twice as long as the radial cell, which is broadly rounded at the apex. Three cubital cells on the right side, two on the left ; on the right the second abscissa of the radius is very short, the third about e(|ual to the first and second combined, but shorter than the second transvei'se cubital nervure, second recurrent nervure received close to the apex of the third cubital cell. Hab. Willowmore, Cape Colony ; January {L>r. Brauns). The female is the type. STUDIES IN THE FOSSORIAt WASPS. G99 1 have little doubt that links will be discovered connecting both sexes of this genus with Myzlne through the short- winged Pseudo- meria section in the females, and through species with somewhat more extended neuration in the males. But the apterous con- dition of the female and the difierences of neui-ation pointed out in the description of the male seem to me to be sufficient reason for founding a new genus. The female shows a strong resem- blance to female Thynnidse of the genus Eirone, also to the Bethylid genus Aj)enesia. I have not been able to examine the mouth-parts, but it is likely that they wovild show atrophy of some parts. The entire eyes of the male are also noticeable as contrasted with the shallowly emarginate eyes of Myzine. Braunsomebia atriceps, sp. n. $ . Nigra ; "mmulihtdls hasi, clypeo, antennis, thorace, segmento 7nediano, 2'>6dihus pyyidiocji^ue apice ferrihgineis. Long. 5 mm. 2 . Mandibles acute, with a very small ill-defined tooth on the inner margin near the apex, Head rectangular, a little broader than long, very slightly convex, shining, with a few scatteied punctures. Thorax shining, with a few scattered punctures on the pronotum, the median segment more closely punctured. Pronotum longer than broad, slightly naiTowed anteriorly, narrower than the head by about one-third ; mesonotum very short, the tegul?e rather better defined than in (luadraiicepjs ; scutellum rounded posteriorly, broader than long. Median seg- ment a little shorter than the pronotvnn, slightly broadened from the base, a little broader than long and obliquely sloped poster- iorly ; sides of the thorax and median segment sparsely clothed with long yellowish hairs. Abdomen shining, finely aciculate, the basal segment truncate anteriorly, with a short petiole not more than half as long as the posterior coxse ; the third segment the broadest ; a semicircular small raised mark on each bide of dorsal segments 2-5 ; sixth segment smooth, pointed at the apex. The constriction between the two basal ventral segments is well marked. The eyes are smaller than in quadraticeps and are separated from the posterior angle of the head by a distance equal to about three times their own length. Hah. Algoa Bay, Gape Colony; November {Dr, Braims). Myzine (?) stigma, sp, n. (PL LXXXI. fig. 11 ; PI. LXXXII. fig. 13.) S . Niger ; mxindihulis hasi., pronoto anguste postice, tegidis, tibiis suhtus tarsisque basi pallide Jlavis ; alis hyalinis, vends perlucidisy stigmate maximo, pallide jlavescenti ; ocidis haud emarghiatis, antennis ahdoviine hrevioribus ; cellula radiali ohlite- rata, cellula cubitali secundo pce7ie obliterata. Long. 7 m,m. (S . Clypeus very short, transverse. Antenna? about as long as 46* 7C0 MR. R. E. TUUXER 0J7 the thorax and median segment combined, not very slendei', of about even thickness throughout, inserted a little nearer to the eyes than to each other. Eyes not emarginate, their inner margins parallel ; posterior ocelli a little nearer to each other than to the eyes. Head convex, closely punctured, with a frontal sulcus reaching to the ocellus ; antennal tubercles not developed. Thorax rather sparsely punctured. Pronotum rather short, as broad as the head, the anterior margin straight, the postei'ior margin widely nnd feebly arched. Scutellum large, a little shorter than the mesonotum. Median segment short, almost smooth, with a median sulcus, truncate posteriorly. Abdomen subsessile, the basal segment broad, not constricted at the apex on the dorsal surface, deeply divided from the second on the ventral surface, all the segments sparsely punctured and shining ; the apical segment rather deeply triangulaily incised for the reception of the long aculeus of the hypopygium ; the whole abdomen about equal in length to the head, thorax, and median segment combined. Stigma very large, about twice as long as the greatest breadth ; only one cubital cell and one recvn-rent nervure, which is received on the cvibitus just beyond the angle of the cubital cell, the cubitus continued just beyond the point of I'eception of the recurrent nervure, the radial cell and all neura- tion beyond the stigma obliterated. Median and submedian cells of the hixid wing present, but no neuration beyond them, the median cell not extending veiy far beyond the submedian. Ilah. '^Yillowmore, Cape Colony (Z>?'. Brauns). This very distinct species will probably prove to be generically distinct from Myzine. It approaches most nearly to M. swalel Turn, and JL brcmnsi Turn., but differs in the reduced neui\ation, the entire eyes, the more robust and subsessile abdomen, and the much broader stigma. But until the female is known I prefer to leave it provisionally in Myzine. From JBraunsomeria, to which the neuration approaches more nearly than to Myzine, it may bo distinguished by the much more robust build, the much shorter and stouter antennre, and the deeper emargination of the apical segment. Myzike braunsi, sp. n. (PI. LXXXI. fig. 14 ; PI. LXXXT.I. fig. u.) S . Niger ; mcmdibulis basi, pronoto fascia angusta postice, seqmentis dorsalihus 2-6 viacula transversa mediali apice, macu- laque curvata laterali utrinque, tegulis, tarsis articulo apicali excepto, tibiis anterioribus omnino, intermediis posterior ihvsqne hasi pallid e fl avis ; alis hyalinis, vends perlucidis, stigmate 2^ullide testaceo. Variai segmentis dorscdibus macula mediali obliterata. Long. 5-8 mm. S . Clypeus much broader than long, closely punctured and shallowly emarginate at the apex. Antennte gradually thickened STUDIES IN TUE FOSSORIAL WASPS. 701 towards the apex, the terminal joint twice fis thick as tlie fourth ; the anteunse about as long as the head, thorax, and median segment combined. Eyes slightly emarginate on the inner margin ; posterior ocelli nearly as far from each other as from the e3'es. Head and thorax closely and not very finely punctured ; pronotum narrower than the head, shorter than the mesonotum, narrowed anteriorly, the anterior margin straight, the posterior margin widely but not strongly arched. Median segment ti'ansveisely rugose, with a longitudinal depression in the middle, truncate posteriorly, the surface of the truncation coarsely transversely striated. Abdomen narrower than the thoiax and nearly half as long again as the head, thorax, and median segment combined, finely and sparsely punctured, the segments scarcely constricted at the base ; petiole of the basal segment very short, the segment abruptly broadened and slightly swollen, nearly as long as the second segment without including the petiole. Apical segment convex, the incision at the apex subtriangular, not quite as deep as its apical breadth. Tarsal ungues simple. Cubital and discoidal nervures not continued beyond the cells, stigma not rounded on the inner mai'gin, three times as long as the greatest breadth; radial cell acute at the apex, produced far beyond the third cubital cell, second and third abscissje of the radius neai-ly equal in length, second recurrent nervure received just before the middle of the thiixl cubital cell. Median cell of the hind wing not emitting any veins from the apex. Hah. Willowmore, Cape Colony; January to March [Dr. Brauns). This is allied to M, svKtlei Turn., but differs in the shorter and stouter basal abdominal segment, in the sculpture of the median segment, and in the translucent nervures of the wings. Both differ from typical Myzine by not having the cubital and discoidal nervures continued beyond the cells and in the absence of the two veins emitted from the apex of the median cell of the hind wings. Myzine coNSXKicTivENTms, sp. n, (PI, LXXXII, fig. 15; PI. LXXXIIl, fig, 12.) S . Niger, dense albo-jnlosics ; mandihulis basi, prouoio margine posteriors, iegidis basi, segmentis dorsalibus 2-6 macula mediaiia apicali et macula transversa utrinque, t'lbiis basi iarsisque pallide Jlavis ; alis hyalhtis, vents bi'unneis. Long. 10-12 mm. (S . Clypeus broad and short, shallowly emarginate at the apex. Eyes distinctly convergent towards the clypeus, the inner margin not emarginate, almost straight. Antennje insei-ted nearer to each other than to the eyes, almost as long as the abdomen, moderately stout and of almost even thickness throughout. Posterior ocelli a little nearer to the eyes than to each other. Head and thorax closely and strongly punctured ; pionotum short, narrower than the head, the anterior margin straight, the posterior margin very feebly arched. Median segment coarsely 702 ME. R. E. TURNER ON rugose, almost vertically truncate posteriorly. Abdomen much longer than the head and thorax combined, not very slender, sparsely punctured, the segments rather strongly constricted at the base, subsessile, the first segment no longer than the second. Sixth dorsal segment more coarsely punctured, not convex, the lateral margins raised, the apical emargination shallow, much broader at the apex than deep. Radial cell broad, nearly twice as long on the costa as the greatest breadth ; second abscissa of the radius distinctly longer than the thii'd, second recurrent nervure received close to the middle of the third cubital cell. Hah. Willowmore, Cape Colony ; October to January {Dr. Jiraxms). This species is easily distinguished by the absence of any emargination of the eyes, the shallow emargination of the apical dorsal segment, the strongly constricted abdominal segments, and the broad radial cell. The cubital and discoidal nervures are continued very little beyond the cells, and the two nervures emitted from the apex of the median cell of the hind wing are very short. Myzine umbratica, sp. n. 5 . Nigra, mandihulis pygidioqiie fusco-ferrngineis ; segmentis dorsalihua 2-3 macida laterali totrinque alba; alls fusco-violaceis. Long. 10 mm. $ . Shining and almost smooth, coarsely but not very closely punctured round the base of the antennae ; a few scattered punctures on the vertex, pronotum, mesonotum, and scutellum, pro- and mesopleui'8e strongly but not very closely punctured ; median segment closely and rather finely punctured at the base, almost smooth in the middle and at the apex, a few obscure striae at the posterior angles ; abdomen almost smooth, with a few small punctures on the apical portion of the segments. Long black pubescence on the sides of the thorax ; calcaria whitish. Eyes rather narrowly ovate, cheeks as broad as the eyes ; ocelli small, the posterior pair about as far from each other as from the eyes; Antennae smooth and shining, the scape beneath punctured and clothed with long hairs. Head subrectangular, half as broad again as long, much broader than the thorax. Pronotum about twice as broad as long, the posterior margin almost straight. Mesonotum only half as long as the pronotum and a little shorter than the scutellum, the parapsidal furrows very distinct. Apical segment of the abdomen convex, long, and pointed. Wings of moderate length, reaching to the fifth abdominal segment, the stigma situated about halfway between the base and apex ; iieuration similar to that of rufifrons Fabr. IIah> Fourteen Streams, Cape Colony ; January (Dr. Brauns). There is only a very obscure median sulcus on the median segment; The slit in the fore wing extends from the termination of the cubitus just beyond the third cubital cell to the margin of the wingi STUDIES IN THE FOSSOKIAL WASPS. ' 703 Myzine abdominalis Guer. (PI. LXXXII. ficj. 16; PI. LXXXIII. fig. 5.) Meria abdominalis Guer. Rev. de Zool. iii. p. 365 (1839), $ . Plesia continua Cam. Rec. Albany Mus. i. p. 299 (1905), S • Hah. Willowmore ; Burghersdorp, Cape Colony. Taken in cojndd by Di\ Brauns. The colour in the female is variable, the head being sometimes ferruginous. Key to the Ethiopian Species of Elis (Mesa). Females. 1. Basal joint of posterior tarsi with a row of spines beneath 2. Basal joint of posterior tarsi unarmed or with a scopa only ])eneath 7. 2. Radial cell distinctly separated from the costa for more than half the length. Black, two apical abdominal segments red IS. alicice Turn. Radial cell separated from the costa at the apex only 3. 3. Abdomen ferruginous, the basal segment only black. *_E. abdominalis Guer. Abdomen 1)lack 4. 4. Thorax and abdomen entirely black 6. Thorax more or less red, the pronotum shorter than the scutellum 6, 5. Sixth dorsal segment punctured ; calcaria of the hind tibia; black *JS. peringueyi Sauss. Si xtli dorsal segment striate ; calcaria whitish *JS. /ioi^ewfoia Sauss. 6. Vertex red, much more sparsely punctured than the front, mesonotum and scutellum black. Spur of the posterior tibiae not strongly bent near the base ,.. ,.., IS. adeJogamia Turn. Vertex black, as closelj^ punctured as the front, mesonotum and scutellum red. Spur of the posterior tibiaj strongl^y bent near the base JS. anrijiua Turn. 7. Abdomen wholly bright ferruginous E. torrida Hm. Abdomen black, the two apical segments sometimes ferruginous red 8. 8. Lateral margins of the median segment acute ; head, thorax, legs, and abdomen black 9. Lateral margins of the median segment not acute ... 10. 9. Median segment twice as broad as the length in the middle, the lateral carina; sharply defined. An- tennae fusco-ferruginous at the base, black at apex JE. saussurei Tuvn. Median segment about three times as broad as the length in the middle, the lateral carina; not sharply defined. Antenna; wholly orange JE. xanthocera Gerst. 10. Head, thorax, and abdomen black 11. Head, thorax, and abdomen more or less red 12. 11. Legs black. Punctures of front and pronotum coarse and confluent longitudinally JE. innotata Turn. Legs ferruginous. Punctures of front and pronotum not coarse and well separated JE. erytliropuda Turn . 12. Apical or two apical segments of abdomen ferru- ginous red, head and thorax black 13. Abdomen wholly black, head and sometimes thorax more or less red 14. 13. Two apical segments of the abdomen red ; fore wings fuscous except at the apex ; sixth dorsal segment punctured S. apicipennis Turn. Apical segment of abdomen only red ; wings hyaline ; sixth dorsal segment striate JE. pyxidata Turn. 704 MR. R. E. TURNER ON 14. Wings subhyaline ; head, pronotum, mesonotum, and scutellumred; vertex punctured E. Iieteroffamia Sauss, Wings fusco- violaceous 15. 15. Vertex punctured ; pronotum closely punctured; legs black jE. Jiova Tm-n. Vertex smooth; pronotum sparsely punctured ; legs red 16. 16. Head red, thorax black I], ruficeps ruficeps Sm. Tnorax more or less red E. riificeps atopogamia [Sauss. Males. 1. Apical dorsal segment not incised at the apex 2. Apicil dorsal segment more or less incised at the apex 4, 2. Third dorsal segment of the abdomen measured from the transverse basal furrow distinctly shorter than its basal width. Abdomen wholly without yellow markings JE. incertalnm. Tliird dorsal segment measured from the transverse basal furrow distinctly longer than its basal width. Abdomen with very small yellow markings 3. 3. Apical dorsal segment flattened, with raised margins and a median carina ; pi-onotum entirely black ... JS. asmarensis Turn. Apical dorsal segment convex, the margins not raised, witliout a distinct median carina ; pronotum with a yellow band on the posterior margin JS. anietalla Turn. 4. Abdomen entirely black, sometimes with blue sheen . 5. Abdomen banded with yellow 6. 5. Wings more or less shaded with fuscous or viola- ceous ; incision of the apical segment half as deep ^ as its apical width. Length 15-18 mm E. ■)~uficeps Sm. Wings clear hyaline ; incision of apical segment much less than half as deep as its apical width. Length 13 mm E. nodosa Guer. 6. Apical dorsal segment flattened with rai.sed margins ; head large, the cheeks more than half as broad as the eyes E. oapitata Sm. Apical dorsal segment convex, with a median carina ; head small, the cheeks much less than half as broad as the ej^es 7. 7. Not very slender ; the apical dorsal segment with an incision nearly as deep as broad at the apex E. spoliata Turn. Very slender ; incision of the apical dorsal segment not mOre than half as deep as its apical breadth ... E. longiventris Tm-n. Ems (Mesa) ALXciiB, sp. n. (PI. LXXXI. fig. 12 ; PI. LXXXII. fig. 8 ; PI. LXXXIII. fig. 9.) $ . Nigra ; segmentis abdominal ibus qtointo sextoque rufo- ferrugineis ; pedibus albo-pilosus, calcariis albidis ; alis nigro- coiruleis. Long. 18 m7)i. 2 . Head and thorax coarsely punctured-rugose, the punctures on the front finer than on the vertex ; clypeus very broadly rounded at the apex, with two or three indistinct teeth on the margin ; scape shining and sparsely punctured, clothed beneath Vith long whitish hairs ; the nine apical joints of the flagellum Opaque. Front thinly clothed with white haii«, the interantennal prominence well developed and feebly bilobed. Inner margin of the eyes slightly sinuate ; posterior ocelli about twice as far from STUDIES IN THE FOSSORIAL WASPS. 705 the eyes as from each othei\ Pronotum nearly as broad as the head, widely emarginate anteriorly ; scutellum with a large triangular rugose area from the base to the narrowly truncated apex, the sides smooth and opaque. Postscutelluni and median segment smooth and opaque, a few large punctures on the middle of the postscutelluni ; median segment raised towards the median line, with a narrow margined median groove, the posterior trunca- tion of the segment coarsely but shallowly punctured, the sides of the segment closely obliquely striated. Abdomen shining, finely and very sparsely punctured, the punctures larger on the ventral than on the dorsal surface, and rather closer at the apex of the segments than at the base ; the apical dorsal segment broadly rounded, and longitudinally punctured-striate. Second abscissa of the radius a little shorter than the third ; first recurrent nervure received beyond the middle of the second cubital cell, second at two-thirds from the base of the third cubital cell. Radial cell detached from the costa for about half its length. Hah. British liast Africa, Makindu, 3300 ft. {8. A.^N'eave); April 5-7, 1911 (.4. E: R. C). This fine speeies does not seeih to be very nearly allied to any other, being well distinguished from the rvficeps group by the very coarse sculptUl-e of the head and thorax. The basal joint of the posterior tarsi is furnished with a closely-set comb of small spines beneath. The description of Cosila donaldsoni Fox corresponds rather closely to this species, but the clypeus is tridentate on the apical margin, not rounded, with indistinct teeth as in the present species, and I think Fox was too careful a woi'ker to have con-- fused the genera. Elis (Mesa) auhiflua, sp. n-. 5 . Nigra ; mandibulis basi, 2}ronoio, mesonoto scuielloque rufo- ferrugineis ; alls inficscatis. Lovg. 12 TiiTn. 5 . Head and pronotum closely and rather strongly punctured ; mesonotum and scutellum much more sparsely punctured, smooth in the middle ; pleurae coarsely punctured ; median segment finely and closely punctured, the median groove harrow and shallow, margined by low carinse, the posterior truncation more shallowly punctured. Abdomen rather closely and not very finely punctured, the apical segment closely longitudinally striated-. Sides of the median segment obliquely striated. , Clypeus transverse at the apex. Eyes vel-y feebly and widely emarginate on the inner margin ; posterior ocelli f ul'-ther from the eyes than from each other. Head more than half as broad again as long, broader than the thorax. Pronotum short, not as long in the middle as the scutellum, the anterior margin stiaight, the posterior margin "widely arched. Pubescence sparse and whitish, calcaria whitish. Radial cell only separated from the costa at the apex, which is subtruncate ; first abscissa of the I'adius as Ion"- as the second. 706 MR. R. E. TURISTER ON but sborter than the third ; recurrent nervures received close to the middle of the second and third cubital cells. Hah. Johannesburg, Transvaal (Kobrow). Received from Dr. Brauns. Nearly allied to adelogamia Turn, and diapherogamia Sauss. From the former it may be distinguished by the entirely black head, and by the closer and more even punctiiration of the head and pronotum, the vertex being almost smooth in adelogamia; from dlapherogaviia it may be distinguished by the much shorter pronotum, the stronger and closer puncturation, and the colour of the head. E. hova has the pronotum longer than in the present species or adelogamia. The comb on the underside of the basal joint of the posterior tarsi is present, but the teeth are few. The upper spur of the hind tibiae is strongly bent near the base. Elis (Mesa) adelogamia Turn. Plesia {Mesa) adelogamia Turn. Ann. & Mag. Nat. Hist. (8) i. p. 503 (1908), 2 . Hah. Maseru, Basutoland ; Lichtenburg, Transvaal. This is nearly allied to E. auriflua, but differs as noticed in the key, Elis (Mesa) ruficeps Sm. (PI. LXXXII. figs. 9, 10; PI. LXXXIII. figs. 2, 4, 10, 15, 16.) Myzine ruficeps Sm. Cat. Hym. B. M. iii. p. 75 (1855), $ ; Turner, Ann. & Mag. Nat. Hist. (8) i. p. 503, S ; Turner, Ann. & Mag. Nat. (8) vii. p. 304 (1911), J $ . Elis (Mesa) ruficeps, subsp. atopogamia Sauss. Plesia (Mesa) atopogamia Sauss. in Grandidier, Hist. Madagascar", XX. p. 244 (1892), $ . Plesia (Mesa) diapherogamia Sauss. ; Distant, Naturalist in the Transvaal, p. 225 (1892), $ . Plesia (Mesa) disjuncta Turn. Ann. & Mag. Nat. Hist. (8) i. p. 502(1908), c?. Elis (Mesa) ruficeps, subsp. atopogamia and diapherogamia Turn. Ann. & Mag. Nat. Hist. (8) vii. p. 304 (1911), S $ • Hah. Zanzibar ; Nyasaland ; Transvaal. The colour-differences between atopogamia and diapiherogamia are not constant, though in the former the mesonotum is usually red and in the latter black. The wings of the male are darker in specimens from Nyasaland than in those from the Transvaal. A single male received from Harar, in S. Abyssinia, has the wings entirely hyaline, slightly iridescent. Elis (Mesa) heterogamia Sauss. Plesia (Mesa) heterogamia Sauss. in Grandidier, Hist. Mada- gascar, XX. p. 244 (1892), 5 . Hah. Delagoa Bay ; Manica ; South Nyasaland. STUDIES IN THE FOSSORIAL WASPS. 707 Elis (Mesa) hova Turn. Plesia {Mesa) hova Turn. Ann. & Ma,g. Nat. Hist. (8) i. p. 504 (1908), $. Hub. Madagascar. Probablj the female of nodosa Gner. *Elis (Mesa) abdominalis Guer. Plesia abdominalis Guer. Rev. Zool. i. p. 57 (1838), $ . Plesia [Mesa) abdominalis Sauss. in Grandidier, Hist. Madagas- car, XX. p. 244 (1892), $ . Ifab. South Africa, Elis (Mesa) apicipenxis, sp. n. 2 ■ jSi^igra ; segmentis abdominalibus quarto apice, quinto sextoque rufo-ferrugineis ; inandihidis basi antennisque ferrugineis, cal- cariis albidis ; alis fuscis, apice anguste hyalinis. Long. 10 mm. 5 . Clypeus finely punctured, broadly rounded at the apex. Head sparsely and rather finely punctured, shining, almost smooth round the anterior ocellus ; interantennal prominence bilobed, divided by a longitudinal sulcus which does not reach the anterior ocellus. Scape shining, very sparsely punctured ; flagel- lum opaque, the two basal joints shining. Posterior ocelli nearly as far from each other as from the eyes. Pronotum and pleurae closely but not coarsely punctured ; mesonotum and scutellum^ more sparsely punctured ; median segment opaque, finely punctured, the ]3unctures more or less confluent longitudinally, the median groove shallow and not distinctly margined. Abdo- men shining, finely and rather closely punctured, with a very short petiole, the first segment broadly truncate at the base ; the sixth dorsal segment punctured, rounded at the apex. The sides of the median segment are closely obliquely striated. Third abscissa of the radius at least half as long again as the second ; first recvirrent nervure received just beyond the middle of the second cubital cell, second at two-thirds from the base of the third cubital cell. Radial cell not distinctly separated from the costa, narrowlv rounded at the apex. Hab. British East Africa, Makindu, 3300 ft. (.S'. A. Neave); April 5-7, 1911 {A. E. R. C). ISTearly allied to E. pyxidata Turn, from IST.E. Rhodesia, but di&ers in the broader sliape of the third cubital cell, the position of the second recurrent nervure, the colour of the scape, of the Avings, and of the fourth and fifth abdominal segments, in the finer and sparser puncturation and in the sculpture of the pygidium. The wings are clear hyaline beyond the radial and cubital cells. Basal joint of the posterior tarsi with a scopa of white hairs beneath ; spur of the posterior tibiie bent near the base. 708 MR. B. E. TURNER ON Elis (Mesa) torrida Sm, Myzifie torrida Sm. Desc. new spec. Hymen, p. 178 (1879), $ . $ . Nigra ; onandihulis basi, scapo apice abdomineque toto ferrugeneis ; tegulis iestaceis ; tibiis tarsisque fusco-fen-ttiffineia; alis hyalinis, venis nigris. Long. 12 mm. 2 . Olypeus short, transverse at the apex, subcarinate in the middle, finely punctured at the base, smooth at the apex^ I^ront, pronotum, aiid mesopleurse closely and strongly punctured, vertex a little more sparsely punctured ; mesonotum and scutellum . coarsely but sparsely punctured ; pronotum as long as the scutellum. Median segment twice as broad as the length in the middle, not margined, almost smooth, but not shining ; the median groove narrow but well defined, with the margins of the groove raised. Abdomen shining, with a few scattered punctures, the apical dorsal segment finely longitudinally striated. Basal joint of the posterior tarsi unarmed beneath, with a scopa of very fine hairs. Radial cell detached from the costa at the apex, third abscissa of the radius longer than the second, but a» little shorter than the first, second recurrent nervure received just beyond two- thirds from the base of the third cubital cell, .Hab. Gambia (ex coll. Shuckajxl). *Elis (Mesa) peringueyi Sauss. Plesia {Mesa) j)eringuey% Sauss. in Gi-andidier, Hist. Madaga,s- car, XX. p. 245 (1892), $ , " Areola radialis apice minute truncata. Secunda v. recurr. in ipso medio margine tertife ar. cubitalis exserta. Metatarsus posticus subtus scopa spinarum brevium instructus. Majuscula, nigra, nitida, cinereo-hirta. Caput et thorax cribrosa. Caput validum, quam pronoto paulo latius. Metathorax Iseviusculus, superne tenuiter punctatus, subtiliter carinatus (carina a latere visa subbitubercnlata) utrinque pago polito ; ejus facies postica plana, subrugulata ; metapleura strigata. Epipygium elongato- trigonale, punctatum. Spinse tib. post, nigrse, acutcie. Metatarsus posticus subtus pectinatus, pilis vel spinis albidis intermixtis, spinisque nonnullis dilatatis. Ala3 bruneo-nebuloste, venis fuscis ; 2^^ ar. cubit, intus valde acuta ; 1^ vena recurrens transversalis, cum V. discoidali angulum fere rectum elficiens." *Elis (Mesa) hotxentota Sauss. Plesia {Mesa) hottentota Sauss. in Grandidier, Hist. Madagascar, XX. p. 245 (1892), $ . " Areola radialis apice minute truncata. Secunda v. recurr. in ipso medio margine tertise ar. cubitalis exserta. Metatarsus posticus subtus scopa spinarum brevium instructus. Minor, nigra, cinereo-hirta. Antennae imo apice flavo. Caput et thorax sat tenuiter cribrosa. Caput quam thorax vix latius. Meta- thorax keviusculus, subtilissime puuctato-rugulatus, superne STUDIES IN THE FOSSORIAL WASPS. 709 obsolete roundato-carinatus ; facie postica punctulata, supra distincte angulata, obtusaiigula, fere rectangula. Epipygium trigonale, striolatum, margine laevi. Spiiue tib. post, albescentes, sqiiamosfe. Alee subhyaliufe venis bruneis et fuscis ; parte apicali nebulosa, 2** ar. cubit, intus breviter acute producta ; 2^ vena recurrens arcuata, obliqua, In alis posticis v. discoidalis longius ultra venulam transverso-discoidalem furcata. Long 14 mill. ; al. 10 mill, (Africa meridionalis)." *Elis (Mesa) capensis Lep. Tiphia capensis Lep. in Hist, Nat. Insect., Hym. iii, p, 554 (1845), $ . _ " Caput nigrum, supra nigro, subtus rufo subvillosum. An- tennae nigrfe, articulo primo nigro hirto. Thorax niger rufo subhirtus. Abdomen nigrum, subnndnm. Pedes nigri, rufopallido subvillosi, femoribus duobus posticis angulatis compressisque, Alse rufo-fuscse, nervuris costaque rufo-fuscis ; sqiiama nigra, Cellula radialis clausa. Femina." " Long. 7 lignes. " Cap de Bonne Esperance. Musee de M. Serville," The figure shows that this species is an Elis. Elis (Mesa) innotata Tuni. Flesia {Mesa) innotata Turn. Ann. & Mag, Nat. Hist. (8) i, p. 506(1908), $. Hah. Loangwa River, N,E. Rhodesia ; S. ISTyasaland, Elis (Mesa) saussukei Turn. Plesia [Mesa) saussurei Turn. Trans. Ent. Soc. London, 1910, p. 394, 2 . Eah. Madagascar. This is near E. xanthocern, hwt the median segment is longer in proportion and much more distinctly margined ; the colour of the antennse is also different. Elis (Mesa) xanthocera Gerst. Myzine xanthocera Gerst. Arch. f. Naturg, xxxvii. p, 353 (1870), 5 ; V. d. Decken, Reise in Ost-Afrika, Gliedethiere, p. 339, pi. 14, fig. 5 (1873), $ . Plesia {Mesa) xanthocera Sauss, in Grandidier, Hist, Madagas- car, XX. p. 245 (1892), $ . Hah. Howick, Natal ; Zoutpansberg, Transvaal ; Mozambique ; Harar, Abyssinia. Elis (Mesa) erythropoda Turn, Plesia {Mesa) erythro2ioda Turn. Ann. & Mag, Nat. Hist. (8) i, p. 505 (1908), $. Hah. Lake Ngami. The scopa beneath the basal joint of the hind tarsus is rather 710 MR. R. E. TURNER ON coarse at the base, but I do not think that any spines are present. Elis (Mesa) incerta, sp. n. c? . Niger, caoio-pilosus ; cl)/peo onacula viediana -minuta, tibiis anticis su2)ra 7;»«^/iVZe Jiavis ; tegulis ■pedihusqiie fuscis ; alls hyalinis, venis nigris ; pygidio haud inclso. Long. 13 wim. cJ . Front coarsely reticulated, the vertex punctured ; thorax closely but not coarsely punctured, median segment punctured- rugose, abdomen finely and closely punctured. Pronotum shorter than the scutellum, the anterior margins straight, the angles subacute. First abdominal segment petiolate, the narrow petiole scarcely more than half as long as the dilated apical portion of the segment, which is constricted at the apex ; second segment twice as broad at the apex as at the base, all the segments slightly constricted at the base. Apical dorsal segment sparsely punctured, without an incision, somewhat convex, subcarinate in the middle, pointed at the apex, the lateral margins raised near the apex. Third abscissa of the radius a little longer than the second, which is fully twice as long as the fourth. First transverse cubital nervure oblique, sharply bent close to the cubitus, second recurrent nervure received just before one-third from the base of the third cubital cell, curved outwards below the middle, the ends slightly -curved inwards. Hah. Ho wick, Natal [J. P. Cregoe) ; Cape Colony. This is very near the description of Plesia carbonaria Cam., but in that species the seventh dorsal segment is said to be shortly incised at the apex, and the fourth abscissa of the radius is almost as long as the second or third. Elis (Mesa) capitata Sm. Myzine capitata Sm. Cat. Hym. B. M. iii. p. 74 (1855), S . S . Niger, alhido-pilosus ; clypeo, mandibuUs, linea ohliqua utrinque supra antennas, 2yronoto fascia angusta p)Ostice, tegulis hasi, segmento dorsali primo fascia apicali emarginata, segmentis dorsalibus et ventralibus 2-6 fascia hisinuata angusta favis; pedibus favis nigro-variegatis ; alis hyalinis, venis testaceis. Long. 14-17 vnm. cJ . Head large, much broader than the thorax, the cheeks more than half as broad as the eyes ; antennae stout, the apical joint truncate at the apex. Head rugose, thorax closely punctured, median segment punctured-rugose ; abdomen shining, glossed with blue, very finely and closely punctured. Pronotum shorter than the scutellum, slightly narrowed anteriorly, the anterior margin straight, the angles not prominent, posterior margin widely arched. First abdominal segment petiolate, the petiole only about half as long as the rather strongly swollen apical STUDIES IJT THE FOSSORIAL WASPS. 711 portion, the npex slightly constricted. Apical dorsal segment flattened, subcarinate in the middle, the lateral margins raised and nearly parallel, the apical incision subtriangular, not quite as deep as the apical width. Second and third abscissiie of the radius nearl}^ equal in length ; second recurrent nervure strongly curved outwards below the middle, joining the cubitus just before one- fourth from the base of the third cubital cell. Hah. Johannesburg, Transvaal ; Kroonstad. The type is much damaged and without wings ; the details of neuration are taken from a more recent specimen in which the cubitus of the hind wing on the right side is almost interstitial with the transverse median nervure, but is further removed towards the base on the left side. Elis (Mesa) spoliata, sp. n. (5 . Niger ; mandibulis, clypeo macula parva nigra utrinque, linea uirinque supra antennas^ i^ronoto angulis aniicis et fascia postice, tegulis, segmento dorsali pri77io fascid angusta apiccili, segmeniis dorsalibus et ventralihus 2-6 fascia blsinuata angusta, coxis apice, femorlhiis posticis supra, tlbiis aniicis et intermediis supra tars isque subtus pallide Jiavis ; jjedibits rufo-testaceis ; alls hyalinis, venis nigris, stigmate rvfo-testaceo. Long. 13 inm. S . Clypeus short and broad, rather narrowly produced in the middle and very feebly emarginate at the apex. Eyes widely emai'ginate ; cheeks very mn.ich narrower than the eyes ; posterior ocelli more than half as far again from the eyes as from each other. Antennal tubercles well developed ; antennae longer than the head, thorax, and median segment combined, of even thick- ness throughout. Head punctured-rugose, much wider than the thorax ; the whole thorax finely and closely punctured, with sparse white pubescence ; pronotum a little longer in the middle than the scatellum, nai-rowed anteriorly, the anterior margin straight, posterior margin strongly arched. Median segment rounded posteriorly, not truncate, very closely but not coarsely punctured. Abdomen slender, slightly shining, very finely and sparsely punctured, petiolate ; the petiole occupying the basal third of the first segment, the apical two-thirds elongate pyriform, the whole segment half as long again as the second, which is gradually broadened from the base, a little longer than the apical width ; third segment broader than long. Apical dorsal segment slightly convex, subcarinate longitudinally in the middle, shining, with a few large punctures, the apical emargination almost as deep as its breadth at the apex. Radial cell pointed ; second abscissa of the radius shorter than the third, but much longer than the first ; second recui-rent nervure received at one-third from the base of the third cubital cell. Hah. Algoa Bay, Cape Colony ; March {Dr. Brauns). Nearly allied to E. capitata Sm., but m that species the head 712 MR. K. E. TURX-EU ON is lai'i>'or, tlio chocks much broader, and the first abdominal KCgniont a little shorter and nioi'o swollen towards the apex. The apical ilorsal segment in capifaUt is Hat, not convex, and has the sides distinctly raised into marginal ca.rinjB, Elis (Mesa) longiventris, sp. n. (Tl. LXXXI. lio-. 13 ; pi.Lxxxni.fig.il.) c? . JMger ; mandibidis, di/pco, liiiea ohliqna itfriiK^ite supra aritonuh'?, pnmoto faticia ani/i(i>ia poslicc, seginexto dorsali primo /((scia ampisfa apicali, ^('(/Dicntis ilorsalibus et roitntlihus '2-G fascia angu^ta apicali bisiiutata, coxis si(hti(s, tibiis tarsisijae auicrioribas supra, iiiierDHkUisqiie subfKS alhiJojlavis : alis /ii/alinis, renis nigris. Long. 10-13 mm. c? . Yery slender. Head rugose; thorax strongly and closely pnnctuvetl, median segment jnnictnred-rngose. Antemuv rather slender, as long as the he:\d, thorax, and luedian segment com- bined, of even tliicknoss throughout. Pronotum longer than the scutellum, the sides almost parallel, the anterior margin straight, with acute angles, the posterior margin widely arched, fc^ides of the head and thorax rather thicky clothed with long white pubescence. ^ledian segment longer than broad, rounded })os- teriorly. Abdomen slender, shining, very finely punctured ; the basal segment as long as the second and third combined, petiolate, the narrow petiole nearly as long as the slightly swollen apical portion ; second segment broadened fti'om the base, about equal in length to the third, tlie segments scarcely constricted at the base ; the apical doi'snl segment convex, subcarinate in the middle, sparsely pnnetni'ed, the punctures lai-ge, the apical incision shallow, not as deep as its breadth at the apex. "Wings reaching to the apex of the fourth doi'sal segment ; second and third abscissa^ of the I'adius usually about equal in length, the fourth distinctly shorter; second recurrent nervure \-eceived just before one-third from the base of the third cubital cell, curved outwards in the middle. I fab. AVilhnvmore, Oape Ooloiiy (Dr. Jhriiois). This corresponds i-ather closely with the description of Plesia ■incisa Cam., but in that description the incision of the apical dorsal segment is s;\id to be twice longer than wide, and some of the details of neuration are not quite the same. But the latter character is of little impoi-tance in the genus, especially as to the length of the second htuI third abscissae of the radius. Many variations occur which are not of specific value. Elis (Mesa) asmaeensis Turn, ricsia asmareims Turn, Aim. vt Maa". Xat, Hist. (8) iii. p. 481 (1909), cT. J fab. Erythvea, STUDIES IN THE FOSSORIAL WAf3P,S. 713 Ells (Mesa) ametalla Turn. Elis {Mesa) ametalla Turn. Ann. & Mag. Nat. Hist. (8) vii. p. 305(1911), c^. Ilab. S. ISTyasaland. Almost certainly the male of iniiolata. Elis (Mesa) nodosa Guer, Myzine nodosa Guer. Diet, pitt. hist, nat. v. p. 584 (1837), r^ ; Sauss. in Grandidier Hist. Madagascar, xx. p. 240 (1892), c? . Hah. Madagascar. *Elis (Mesa) cIiAvata Sauss, Myzine clavata Sauss. in Grandidicr, Hist, Madagascar, xx. p. 242 (1892), J. Hah. Transvaal, *Elis (Mesa) cahbonaria Onm. Plesia carhonaria Cam. Rec, Albany M'us. i, p, 317 (1905), c? . Hah. Dunbrody, Oape Oolony. *Elis (Mesa) reticulata Cam, Plesia reticulata Cam. Rec, Albany Mus, i. p, 300 (1905), J • Hah. Bra.k Kloof, Cape Oolony, *Elis (Mesa) rufo-femorata Cam, Plesia rufo-femorata Cam, Rec, Albany Mus, i.p. 298(1905), 6 ■ Hah. O'okiep, Cape Colony, *Elis (Mesa) incisa Cam. Plesia incisa Cam. Rec, Albany Mus, i, p, 320 (1905), S , Hah. Dunbi'ody, Cape Colony, iLcy to the Oriental Species of Elis (Mesa), Females. 1. Sixth dorsal segment longitudinally striated 2. 8ixth dorsal segment punctured 8. 2. Abdomen more or less ferruginous .3, Abdomen wholly black 4. 3. Clj'peus with a median carina ; two apical abdominal segments and the apex of the fourth black JS. dimidiata Gudr. Clypeus without a distinct carina; apical abdominal segment only black JiJ. mandaJensix Magr. 4. Median segment distinctly margined posteriorly JS.fiiscijjennis 8m. Median segment not distinctly margined posteriorly... 5. 5. Wings fuscous 6. Wings subhyaline 7- 6. Pronotum longitudinally rugo.se; second abscissa of the radius as long as the third. Length 16 mm.... JE. mandibular is 8u\. Pronotum coarsely punctured ; second abscissa of the radius much shorter than the third. Length 10 n)\\\. E. jietiohita Sm. Proc. Zool. Soc— 1912, No. XLVII, 47 714 BIR. 11. E. TUKNER ON 7. Propleuiw sparsely punctured JS. claripennis Bingh. Propleur;e tinely striated JB, ustulata Turn. 8. Abdomen not marked with yellow 9. Abdominal segments with yellow apical bands *_B. picticoUis Mor. 9. Head black ; abdomen black or black and ferruginous. 10. Head red ; abdomen steel-blue JS. tricolor Sm. 10. Abdomen mostly ferruginous 11. Abdomen entirely black, or with the apical segment only ferruginous 12. 11. Four basal segments of the abdomen ferruginous JE. bengalen.iis Cum. Four apical segments of the abdomen ferruginous E. api macula Cam. 12. Apical segment of the abdomen ferruginous 13. Apical oegmeut of the abdomen black 14. 13. Mesouotum closely punctured *E.fedtschenkoi Sauss. Mesonotum shining, ahnost impunctate *JS. dubia Mor. 14-. Legs ferruginous; head coarsely and closelj' punctured. JE. rothneyi Cam. Legs black ; head sparsely punctured E. opacifrons Turn. Myzine anthracina recoi-deil by Bingliam as Indian is Atisti'alian and belongs to the genus Anthohosca ; and Myzine combibsta also i^ecorded as Indian is undoubtedly West. Indian, and a true Elis. 2Iales. 1. First dorsal segment elongate, petiolate, more or less nodose at the apex 2. First dorsal segment subsessile, not nodose at the apex, broader at the apex than long E. dimidiaticornis 2. Seventh dorsal segment distinctly incised at the apex ; [LJingh. abdomen wholly black 3. Seventh dorsal segment not incised at the apex 4. 3. Legs black; seventh dorsal segment smooth and shining at the apex E. htirmamca Bingh. Legs fusco - ferruginous ; seventh dorsal segment coarsclj' punctured E. Iceta Bingh. 4. Abdomen wholly black, without yellow markings, sometimes glossed with blue 5. Abdomen more or less marked with yellow 6. 5. Median segment without a distinct sulcus; pygidial area clearly detined, with marginal carinse reaching nearly to the base of the segment E. dimidiata Gu&. Median segment with a distinct median sulcus; pygidial area less clearly defined, the marginal carinas lower and only on the apical third of the segment E. mandibuJaris Sm. 6. First abdomnial segment very little, if at all, longer than the hind femur and trochanter combined 7. First abdominal segment much longer than the hind femur and trochanter combined 8. 7. Head punctured-rugose ; pronotuni without a yellow band E. nursei Turn. Head punctured ; pronotum with a yellow band on the posterior margin E. manda lensis Magr. 8. Wings hyaline ; pronotum with a yellow band on the posterior margin, closelj' punctured 9. Wings pale fusco-hyaline ; pronotum without a yellow band, rather sparsely punctured E. e.vtensa Turn . 9. Narrow petiole of first abdominal segment scarcely more than half as long as the swollen apical portion of the segment E. petioluta Sm. Narrow petiole of first abdominal segment nearly as long as the swollen apical portion of the segment ... E. claripennis Bingh. Myzine pallida Sm. and Myzine orientalis Sm,, placed by Bing- ham and Smith in Myzine, together with many species of Elis, STUDIES m THE FOSSOUIAL WASPS. 715 belong to the genus Iswara Westw., which also belongs to the Elidinse. Elis (Mesa) dimidiata Guer. Myzine dimidiata Guer. Diet. pitt. hist. nat. v. p. 584 (1837), Methoca orientalis Sm. Cat. Hyni. B. M. iii. p. 66 (1855), d" (nee Smith, 1875). Myzine madraspatana Sm. Cat. Hym, B. M. iii. p. 72(1855), $ . Myzine violaceijyennis Cam. Mem. Manchester Lit. & Phil. Soc. xlii. p. 21 (1898), ? , Hah. The whole of India, except the ^ofth-'West, $ . Head ai:!.d thqrax coarsely rugosely punctured, smooth round the anterior ocellus and at the base pf the mespnotum ; median segment opaque, wit;h a nari'pw median sulcus in which is a low carina, the sides of the sulcus yaised and forming low carinse, the segment not margined at the base of the posterior truncation. Eadial cell distinctly separated from the costa for more than half its length, third abscissa of the radius very distinctly longer than the second. Sixth dorsal segrnent finely longitudinally striated ; basal joint of hind tarsi with a scopa of short hairs beneath. Black ; second, third, and the base of the fourth abdominal segments ferruginous. Oalcaria whitish ; wings fuscor violaceous. S . Head and tliorax closely punctured ; median segment rugose ; basal third of first abdominal segment forming a narrow petiole, apical two-thirds swollen and slightly constricted at the apex, the whole segment more than half as long again as the second segment. Abdomen shining, very finely punctured ; pygidial area coarsely punctured, fully twice as long as broad, the sides raised and forming carinse, a well-definpd median carina, the apex narrowly truncate, not emarginate. Black ; the abdomen slightly glossed with blue ; base of the mandibles, fore tibiae in front, fore tarsi in front, base of the tegulae and the hind margin of the pronotum naiTowly and obscurely pale yellow. Wings hyaline at the hasp, fusco-hy^-line beyond the basal nervure^ Length 17 mm, The association of the sexes was suggested by me in 1908, and positive proof was published by Mr. Lefroy a year later, a pajr having been taken in coitu by Mr. Dutt at Pusfi, Elis (Mesa) mandibularis Sm, Methoca mandibularis Sm. Trans. Ent. Soo. London, p. 301 (1869), J. Plesia {Mesa) mandibidaris Turn. Ann. & Mag. Nat. Hist. (8) i. p. 509 (1908), c?. Plesia {Mesa) pur-pureipennis Turn. Ann. k. Mag. Nat. Hist. (8) i. p. 508 (1908), 2 . This will almost ceatainly prove to be the Chinese subspecies 47* 716 MR. R. E. TURNER ON of E. dimidiata Guei\, from wliicli the female differs in the entirely black abdomen and the greater length of the second abscissa of the i-adius, which is nearly as long as the third. The male differs from dimidiata in having the second abscissa of the radius very distinctly longer than the third, in the paler colour of the ajjical portion of the wings, in the presence of a distinct longitudinal sulcus on the median segment, and in the somewhat shorter and broader form of the first abdominal segment. The pygidial area of the male is also wider and less distinctly margined than in dimidiata. Hah. Shanghai. Elis (Mesa) bengalensis Cam. Myzine bengalensis Cam. Mem. Manchester Lit. & Phil. Soc. xlii. p. 21^(1898), $. The four basal abdominal segments are ferruginous ; the wings violaceous, the base of the hind wings hj^aline. Pronotum spaivsely but coarsely punctured. Length 15 mm. This is quite distinct from dimidiata, and seems, as Camei'on suggests, to be more nearly related to mandalensis. Elis (Mesa) mandalensis Magr. Plesia mandalensis Magr. Ann. Mus. Civ. Gen. (2) xii. p. 257 (1892),$. $ . The live basal abdominal segments are ferruginous. Calcaria and spines of the bind tibite whitish. Wings hyaline, dark fusco-hyaline beyond the basal nervure of the fore wing- to the apex, the apex of the hind wing fusco-hyaline. Pronotum closely and somewhat coarsely punctured ; sixth dorsal segment liuely longitudinally striated. J . Head and thorax closely and rather finely punctured ; median segment punctured-rugose ; abdomen shining, micro- scopically punctured. First abdominal segment scarcely longer than the hind femiu' and trochanter combined, the narrow petiole a little more than half as long as the moderately swollen apical portion of the segment ; second segment longer than the third by about one-quarter. Seventh dorsal segment pointed, not incised ; pygidial ai-ea well defined, very narrow, coarsely punctured and with a well-marked median carina. Third abscissa of the radius distinctly longer than the second. Black ; mandibles at the base, clj^peus, apex of the inter- antennal prominence, posterior margin of the pronotum, tegulfe, an apical band on dorsal segments 1-6, strongly bisinuate on segments 2-6 and on ventral segments 2-6, tarsi, fore and intermediate tibia; in front, and the base of the hind tibiaj yellow. Wings hyaline, nervures fuscous. Length, $ 10 mm., S 10-11 mm. Hah. Mandalay, Burma. Taken in copula by Colonel Bingham. STUDIES IN THE FOSSORIAL AVASPS. 717 Elis (Mesa) rotiineyi Cam. Myzi7ie rothneyi Cam. Ann. & Mag. Nat. Hist. (7) x. p. 88 (1902),?. $ . This fine species is easily distinguished, being black with ferruginous legs ; the wings f usco-hyaline, flushed with purple. The sixth dorsal segment is coarsely punctured, the punctures tending to become confluent longitudinally towards the a,pex. Head and thorax very coarsely punctured, rugose on the pronotum. Hah. Khasi Hills, Assam. Elis (Mesa) fuscipennis Sm. Myzine fuscipennis Sm. Cat. Hym. B. M. iii. p. 72 (1855), $ ; Bingh. Fauna Brit. India, Hymen, i. p. 67 (1897), $ . 2 . This species may be distinguished from the female of petiolata Sm., which it closely resembles, by the sharply margined median segment, the posterior slope of which is a,brupt and steep, not gradual as in petiolata. The colour is black ; the calcai-ia whitish, mandibles fusco-ferruginons, wings fuscous. The spines on the outer margin of the hind tibise are black in fuscipennis, whitish in petiolata. Length 12 mm. Ilab. India. The type is unique in the British Museum collection and was obtained from Shuckard in exchange, so that the locality is uncertain. Bingham's description is taken from the type, but he evidently confused the species with ^;e//i:'cZrtto, a specimen of which is labelled ftiscipennis by him in the British Museum collection. I have not seen the specimens he records from Burma, but consider it very doubtful if they belong to this species. Elis (Mesa) petiolata Sm. Myzine petiolata Sm. Cat. Hym. B. M. iii. p. 72 (1855), J. Myzine ceylonica Cam. Ann. & Mag. Nat. Hist. (7) v. p. 18 (1900), ?. riesia {Mesa) petiolata Turn. Ann. & Mag. Nat. Hist. (8) i. p. 512 (1908), d $ . 5 . Differs from fuscipeomis Sm. in the absence of a distinct margin separating the dorsal surface of the median segment from the surface of the posterior slope and in the white spines of the hind tibiae. From claripennis Bingh. it differs in the fuscous colour of the wings, the somewhat coarser punctumtion, and the lesser length of the second abscissa of the radius, which is only about half as long as the third in petiolata and almost or quite as long as the third in claripennis. In fi(,sci2Jennis the second abscissa of the radius is shorter than in petiolata, being distinctly less than half as long as the third. 718 MR. R. E. TURNER ON c? . This is very near the male of claripennis, but may be distinguished by the less elongate petiole, the narrow basal portion of which is scarcely more than half as long as the moderately swollen apical portion of the segment. Length, c? $ , 10 nim. Hah. Bengal, Bombay, and Ceylon. Elis (Mesa) claripennis Bingh. Myzhie claripennis Bingh. Fauna Brit. India, Hymen, i. p. 68 (1897), ?. Myzine hortata Nurse, Journ. Bombay Nat. Hist. Soc. xiv. p. 81 (1902), $. 5 . Differs from petiolata as noticed under that species. Bingham's description of the species is inaccurate as to the median segment, and this has misled Nurse. The segment is not smooth and shining, and the longitudintil impression, though not very long, cannot be said to be triangulari S . The male, which has not been previously described, closely resembles petiolata Sm. Head and thorax closely punctured, most coarsely on the front, median segment rugose; abdomen veiy slender, shining, minutely punctured. First abdominal segment more than half as long again as the hind femur, the narrow petiole very little shorter than the feebly swollen apical portion ; second segment very narrow at the base, nearly half as long again as the third segment. Seventh dorsal segment not incised at the apex, pointed, convex and without a distinct pygidial are^. Black ; mandibles, clypeus^ the apex of the interantennal prominence, posterior margin of the pronotum, an apical narrow- band on dorsal segments 1-6, strongly bisinuate on segments 2-6, base of the tegulse, fore tarsi, antei"ior and intermediate tibise in front, base of hind tibise and base of intermediate and hind tai'si pale yellow. Wings hyaline, nervures fuscous. Length, J 8 mm., § 10 mm. Hah. Burma, Ceylon, Bengal j and Defesa. The male has the two basal abdominal segments mOre slender than in petiolata, the first with the apical portion less swollen, the basal narrow petiole longer in proportion ; the second narrower at the base. Elis (Mesa) ustulata Turn. Plesia {Mesa) ustidata Turn. Ann. & Mag. Nat. Hist. (8) i. p. 510 (1908), $. 5 . This is nearest to claripennis Bingh., but is a larger and more robust species, somewhat more closely punctured, with the wings distinctly darker and the punctures oh the abdomen larger. Hah. Yuhzalin Valley, Tenasserim, STUDIES IN THE FOSSORIAL WASPS. 719 Elis (Mesa) orACiFROxs Turn. Plesia {Mesa) ojjacifrons Turn. Ann. & Mag. Nat. Hist. (8) i. p. 509 (1908), $. $ . This black species may be distinguished by the very sparse puiicturation of the head, pronotum, and sixth dorsal segment. The wings are pale f usco-hyaline. It is a larger and more robust species than petiolata. Ilab. Salwen Valley, Tenasserim. *Elis (Mesa) dubia Mor. Plesia dubia Mor. Hor. Soc. Ent. Eoss. xxiv. p. 627 (1890), $ . This species is black, with the apical abdominal segment, tarsi, hind tibite, mandibles, and antennfe beneath ferruginous, the mesonotum shining and almost impunctate, the sixth dorsal segment finely and closely punctured. Length 1 1| mm. Hah. Turkestan. Plesia (Mesa) apimacula Cam. Myzine apimacula Cam. Journ. Bombay Nat. Hist Soc xiv p. 272 (1902), $. $ . This is allied to duhia Mor., but differs in having the four apical abdominal segments ferruginous, the wings palei', and the puncturation somewhat different. Hah. Deesa, N.W. India. The male described as belonging to this species by Colonel Nurse is Pcecilotiphia alhomacidata Cam., but has three cubital cells instead of two as in the type of the species, which is evidently an aberration from the usual neuration. The male has the form of a Myzine, not of a Mesa, but it is by no means improbable that Nurse is correct in the association of the sexes though he informs me that he does not recollect his reasons for placing them together. The apical dorsal segment is deeply incised as in Myzine, Cameron's action in forming a new genus on a specimen with abnormal neuration is quite unjustified and due to ignorance of the vai'iable character of neuration in the Scoliida?, but I am at present doubtful if Pcecilotiphia should be treated as a synonym of Myzine ov of Elis {Mesa). *Elis (Mesa) fedtscjenkoi Sauss. Plesia fedischenkoi Saussure in Fedtschenko : Turkestan, Scoliidse, p. 29 (1880), $ {Plesia tariara on plate). Plesia tartara 8auss. I.e. pi. ii. fig. 12, $ . This seems to differ from duhia Mor. in the coarser and closer puncturation of the mesonotum. The name tartara is used on the plate through a fault in the editing, and has unfortunately been recorded in Dalla Tori-e's Catalogue as a distinct species. 720 MR. R. E. TURNER ON *Elis (Mesa) picticollis Mor. Flesia piciicoUis Mor. Hor. Soc. Ent. Ross. xxiv. p, 624 (1890), $ . This appears to be a very distinct species, strongly marked with yellow on the head, thorax, and abdomen. The colour wovild suggest an Anthohosca alHed to A. arabica Turn., but the structure of the apical abdominal segment and the indistinct striation of the metapleurse render it unlikely that it belongs to that genus. Elis (Mesa) tricolor >Sm. (Pl.LXXXI. fig. 16; PI. LXXXII. fig. 11.) Myzine tricolor Sm. Jovu^n. Linn. Soc, ZooL ii. p. 91 (1858), 5 . This fine species is easily distinguished by the large size (19 mm.), the bright red head, and steehblue abdomen. The wings are fusco-hyaline, flushed with purple, almost hyaline at the base. The head is large, quadrate^ sparsely punctured, smooth on the vertex ; pronotum coarsely punctvired, median segment subconcave on the posterior slope ; sixth dorsal segment strongly punctured. Second abscissa of the radius nearly as long as the first and third combined. Hah. Borneo (typical) ; Dibrughar, Assam ; W. India [T. R. Bdl). The only specimen I have seen from Assam has the head distinctly longer than broad, lohger behind the eyes than in the typical form, and the scape and three basal joints of the flagellum are red ; the median segment is not at all concave on the posterior slope. The differences will probably prove to be subspecific. Ml\ Bell iliformed me that he bred this species from the larva of a longicorn beetlei Elis (Mesa) ? diMidiaticornis Bingh. (PI. LXXXI. fig. 15.) Myzine dimidiaticnrnis Binghi Journ^ Linn. Soc, Zool. xxv. p. 423 (1896), c^ ; Turn. Ann. & Mag. Nat. Hist. (8) i. p. 501 (1908), 6 ^ •S . Antennpe stout^ a little longer than the thorax and median segment combined. Front rugose, vertex sparsely punctured. Plfonotum transversely rugulose, much longer than the meso- iiotum, the sides almost parallel. Head slightly narrowed and produced from behind the eyes. Thorax and median segment rugose, the median segment distinctly margined posteriorly and vertically truncate. First abdominal seglnent vertically truncate anteriorly, with a distinct transverse caHna above the base of the truncation, attached to the abdomeli By a very short petiole, nearly as broad as the second segment-, the sides parallel. Apical dorsal segment not iiicised at the apex. Without a pygidial area. Second abscissa of the radivis as long as tho first and third cDmbined. Black ; scape and four basal joints of the flagellum, clypeus, and apex of the interantennal prominence dull ferruginous ; abdomen glossed with steely blue. Wings hyaline ; fusco- STUDIES IN THE FOSSORIAL WASPS, 721 violaceous from the basertice ynacula utrinque sup^'a oculos fusco-sanguinea ; unguic-ulis hijidis ; alis suhhyalinis, venis fusco-ferru,gineis. Long. 14 mm TUDIES IN THE FOSSORIAL WASPS. 735 5 . Radial cell rounded at the apex ; third abscissa of the radius shorter than the second ; third cubital cell as long on the radius as on the costa ; first recurrent nervure received just beyond one- third from the base of the second cubital cell, second at one-fifth from the base of the third cubital cell ; second cubital cell not very strongly produced towards the base. Tarsal ungues bifid, without a basal lobe ; basal joint of hind tarsi with rather long haiis beneath ; lobe beneath the apical half of the hind femora broadly rounded. Closely punctured ; rather strongly on the head, finely on the thorax, coarsely on the scutellum, rather sparsely and shallowly on the abdomen ; sixth dorsal segment rather broadly truncate at the apex. Black, with grey pubescence, a fusco-sanguineous spot on the vertex on each side close to the summit of the eyes ; sixth dorsal segment fuscovis at the apex, calcaria whitish. Hah. S. Australia ; probably from Adelaide. This species is remarkable for the very long second abscissa of the radius, and may be easily distinguished from the majority of Australian species by the bifid tarsal ungues. 1.3. Anthobosca nubilipennis Turn. Anthohosca nuhilipennis Turn. Proc. Zool. Soc. London, p. 307 (1910), $. 5 . Radial cell detached from the costa for fully half its length, slightly angular at the apex ; second abscissa of the radius longer than the first and about equal in length to the third ; first recurrent nervure received beyond two-thirds from the base of the second cubital cell, second before one-quarter from the base of the third cubital cell, which is nearly twice as long on the cubitus as on the radius. Tarsal ungues bifid, without a basal lobe ; hind femora convex beneath, the lobe not apical but almost medial ; basal joint of hind tarsi with three or four short spines beneath. Deeply but not very closely punctured ; meso- notum and scutellum very sparsely punctured ; sixth dorsal abdominal segment coarsely striated at the base, rounded at the apex. Black, with white pubescence ; calcaria black. Wings fusco- hyaline, nervures fuscous ; the stigma small. Median segment abruptly truncate. Length 16 mm. Hab. Perth district, S.W. Australia. A very distinct species in the characters of the hind legs, neuration, and median segment. 1 4. Anthobosca clypeata Sm. (PI. LXXXI. fig. 2 ; PI. LXXXII. fig. 5 ; PI. LXXXIII. fig. 7.) Dimorphoptera clypeata Sm. Trans. Ent. Soc. London, p. 239 (1868), 2. $ . Radial cell rounded at the apex, detached from the costa 73^ MR. R. E. TURNER ON at the extreme apex only; second abscissa of the radius twice as long as the first and about equal in length to the third ; second cubital cell receiving both recurrent nervures, the first at one- third from the base, the second just before the apex ; third cubital cell as long on the radius as on the cubitus. Tarsal ungues bifid ; lobe beneath the apical half of the hind femora strongly rounded : basal joint of hind tarsi with a row of short hairs beneath, in which are a few longer spines. Closely punc- tured ; the thoi"ax very closely and finely punctured, with lai^ger and sparser punctures intermixed ; abdomen shining and sparsely pnnctured. Black, with whitish pubescence ; sixth dorsal segment densely clothed with long golden hairs ; clypeus except at the apex and the margins of the eyes broadly interrupted on the summit, yellow ; basal two-thirds of doi-sal segments 2-4 and of ventral segments 2-3 emarginate in the middle posteriorlv, rufo- testaceous. Wings subhyaline, nervures fusco -ferruginous. Length 22 mm. Hah. Swan River, W. Australia. An obscure scar runs from the base of the stigma to close to the base of the first transverse cubital nervure. I have little doubt that the male of this species will prove to be A. crassicornis Sm,, which corresponds with it in the position of the recurrent nervures, the bifid tarsal ungues, and also to some extent in the colour of the clypeus and abdomen, *15, Anthobosca fasciculata Sichel. Cosila (Colobosila) fa^ciciUata Sichel, Sauss, et Sichel, Cat. spec. gen. Scolia, p, 263 (1864), 5 , Hah, Australia. I fail to recognise this species with any certainty, but do not consider that the subgenus Colohosila can stand, the truncation of the apex of the radial cell being insutiicient' a,s a subgeneric character unless supported by others, *16, Anthobosca inornata Sauss. Cosila inornata Saussure in Grandidier, Hist, Madagascar, xx. p, 233 (1892), S . T do not recognise this species as a synonym of any known to me, though it seems to be related to A. anthracina, differing in the position of the first recurrent nervure. Considering how unreliable the details of neuration are in this genus, it is quite possible that it is only a variety of anthracina. *17, Anthobosca melanaria Cam, Flesia 7nelanari€(, Cam, Rec. Albany Mus, i. 5, p. 297 (1905), 2 - 5 . From the description I have no doubt that this is an Anthobosca, the characters " apex of radial cell sharply pointed," find " pygidiuni piceous red, its base fringed with long, bright pufons hair," and " metapleurse smooth," agreeing much better STUDIES IX THE FOSSORIAL WASPS. 737 with Anthobosca than Plesia. Although the two genera belong to different subfamilies of the Scoliidae, the females are likely to be confused by a beginner. Cameron places male Scoliidse of the subfamily Anthoboscinae in the Thynnidse, and females of the same group with the Scoliidfe of the subfamily Elidinse. *18. Anthobosca donaldsoni Fox. Cosila donaldsoni Fox, Proc. Acad. Philadelphia, p. 549 (1896), $ . 5 . Third transverse cubital nervure received close to the apex of the radial cell. Clypeus tridentate on the anterior margin. Head and thorax very coarsely punctured, pronotum and scu- tellum scabrous ; median segment finely striato-punctate, the sides obliquely striated ; abdomen strongly but sparsely punc- tured, sixth dorsal segment striato-punctate : tarsal ungues cleft. Black, with greyish pubescence ; two apical abdominal segments red ; calcaria whitish. Wings black, strongly violaceous. Length 18 nam. Hah. Somaliland. I have not seen this species, but the description corresponds almost exactly Avith Elis alicioi described in this jjaper, though the clypeus in that species is not distinctly tridentate. But I do not believe that Fox would have confused the genera. 19. Anthobosca insularis Sm. (PI. LXXXII. fig. 6 ; PL LXXXIII. fig. 8.) Myzine insidaris Sm. Descr. new spec. Hymen, p. 178 (1879), $. Cosila insidaris Saussure in Grandidier, Hist. Madagascar, xx. p. 231 (1892), $. 5 . Radial cell subacute at the apex, receiving the third trans- verse cubital nervure very near the apex, in some specimens almost at the apex ; second abscissa of the radius longer than the first, the two combined not quite as long as the third ; first recurrent nervure received at two-fifths from the base of the second cubital cell, second at one-quai-ter from the base of the third cubital cell, which is nearly or quite as long on the radius as on the cubitus. Tarsal ungues bifid ; hind femora very broadly rounded beneath on the apical two- thirds, basal joint of hind tarsi with three long spines beneath. Sliining and sparsely punctured, closely on the front and closely and very finely on the median segment. Black, with black pubescence ; calcaria brown. "Wings fiavo- hyaline, nervures ferruginous. Lengt.h 23-29 mm. Hah. Madagascar. The type has a rufous spot in the middle of the first dorsal segment, and obscure reddish shading on the head and thoi^ax. 738 MR. R. E. TURNER ON 20. Anthobosca natalica Turn. Anthohosca natalica Turn. Trans. Ent, Soc. London, p. 85 (1908), 2 . 5 . Radial cell narrowly rounded at the apex, subacute ; third abscissa of the radius twice as long as the second, the fourth shorter than the first ; first recurrent nervure received just beyond the middle of the second cubital cell, second before the middle of the third cubital cell, which is as long on the radius as on the cubitus. Tarsal ungues bifid ; basal joint of hind tarsi with a scopa beneath. Shining and sparsely punctured, more closely on the front and pronotum ; median segment very finely and closely punctured ; abdomen finely but more sparsely punctured. Black ; pubescence on the apical dorsal segment brown and long ; tegulfe and extreme apical margin of the abdominal segments testaceous brown ; calcaria whitish ; fore tarsi fusco-feiruginous. Wings fusco-hyaline, nervures fusco-ferruginous. Length 12-17 mm. Hah. Malvern, JSTatal. The neuration given is as in the type, but in another specimen the second abscissa of the radius is much more than half as long as the third, and both recvirrent nervures are received nearer the base of the cells. 21. Anthobosca erythronota Cam. (PI. LXXXI. figs. 5, 6 ; PI. LXXXII. fig. 4.) Plesia erythronota Cam. Rec. Albany Mus. i. 5, p. 320 (1905). $ . Nigra ; iwoihorace. viesothorace, tibiis tarsisque rufo-ferru- gineis ; segmentis abdominalibus 2-4 macula laterali alba utrinque ; alls fusco-hyalinis, venis nigris. (S . Niger, gracilis ; 'mandibtdis basi, clypeo, ocidis margine interiore, pronoto late postice, tegidis, mesonoto macida, mesopleuris inacida, sctttello macida tjiediana, postsctctello macula mediana transversa, segmentisque abdominalibtcs 2-4 macula magna trans- versa utrinque albido-Jlavis ; pedibus nigris albo-variegatis ; alis hyalinis, venis nigris. Variat $ prothorace mesothoraceque nigris. Long., 5 10-11 mm., S 8-11 mm. $ . Clypeus shining in the middle and somewhat flattened, narrowly truncate at the apex. Head convex, broader than long, not much broader than the thorax ; rather sparsely punctured, the space round the base of the antennje closely punctured and clothed with fulvous hairs. Eyes touching the base of the mandibles, the line of the interior margin slightly undulating ; posterior ocelli much nearer to each other than to the eyes. Thorax rather spai'sely punctured ; pronotum slightly emarginate anteriorly, the posterior margin widely but not strongly arched ; the projection of the mesosternum between the intermediate coxse very deeply bilobed. Median segment veiy finely and closely STUDIES IN THE FOSSORIAL WASPS. 739 punctured, the posterior slope very steep. Abdomen lather closely punctured ; the apical margins of the segments rather bi'oadly depressed and smooth, more broadly in the middle than at the sides, the apical segment testaceous and thickly covered with fulvous bristles. Radial cell bluntly rounded at the apex, not detached from the costa ; second abscissa of the radius a little longer than the first, but scarcely more than half as long as the third, the third cubital cell longer on the radius than on the cubitus, second recurrent nervure received at about one-quarter from the base of the third cubital cell. Ungues cleft. (S . Clypeus slightly convex, rather narrowly truncate at the apex and armed with a row of three very minute teeth. Antenna) no longer than the thorax and median segment combined, tapering slightly towards the apex. Eyes convergent towards the clypeus, the inner margin almost straight. Head and thorax finely and closely punctured ; pronotum strongly narrowed anteriorly, the anterior margin slightly emarginate, posterior margin strongly arched. Abdomen shagreened, very slender, slightly tapering to the extremities, the basal segment distinctly longer than the second. Hind tibife finely serrate, all the tarsal ungues cleft. Second abscissa of the radius half as long again as the first a.nd only a little more than half as long as the third, second recurrent nervure received at about one-quarter from the base of the third cubital cell. Ilab. Willowmore, Cape Colony ; November to January (i)r. Brauns). The male comes very near Cameron's genus Odontotliynnus, which I have elsewhere treated with some doubt as a synonym of Anthobosca ; but Cameron states that the posteiior tarsal ungues in his genus are simple. Even if he is correct as to this character it would not be sufficient to justify the formation of a genus on one sex only, and his remarks show that he has no knowledge of the genus Anthobosca. I suspect that the present species may prove to be identical with Plesia leucospila Cam., with the description of which it agrees fairly well, but the mesopleurse are rather strongly though not very closely punctured, not almost smooth as in Camei'on's description. In the broadly rounded apex of the radial cell this species diflfers from A. natcdica Turn., in which the cell is subacute, but agrees with A. arahica Turn. *22. Anthobosca leucospila Cam. Plesia leucospila Cam. Rec. Albany Mus. i. 5, p. 319 (1905), $. 5 . It is almost certain that Cameron has misplaced this species. It is possibly identical with A. eryihronota Cam., which has a similar colour variety, in which case the name leucospila should be used for the species. But Cameron states that the pleurae are almost smooth, whereas in eryihronota the mesopleurse 740 MR. R. E. TURNER OX are strongly though not closely punctured. But this is possibly an error in Cameron's description. *23. Anthobosca sauakinensis Magr. Myzine sauakinensis Magr. Ann. Mus. Civ. Genova, xxi. p. 560 (1884), $ . I think it probable that -4. a?'a&^ca Turn, is only a colour variety of this species, with the abdomen black instead of ferruginous. The difference in size between the two forms is considerable, but both size and colour vary much in this genus. But as I have not seen sauahinensis I think it better to keep the two separate at present. 24. Anthobosca arabica Turn. Anthobosca arabica Tvmi. Trans. Ent. Soc. London, p. 397 (1910), ?. 2 . Radial cell broadly rounded at the apex ; second abscissa of the radius longer than the first, the two combined equal in length to the third, first recurrent nervure received a little before the middle of the second cubital cell, second before one-third from the base of the third cubital cell, which is as long on the radius as on the cubitus. Tarsal ungues bifid ; the lobe beneath the hind femora occupying nearly the whole length of the joint and scarcely rounded ; basal joint of fore tarsi with a comb of nine rather short spines on the outer margin and a row of short fine spines on the inner margin, the outer angle strongly produced and almost reaching the apex of the second joint. Shining, finely and sparsely punctured, more closely on the front and pronotum. Black, with grey pubescence, a narrow band on the inner margin of the eyes, continued and ai-ched on the vertex, a spot on the front, hind margin, and anterior angles of the pronotum, a curved band on the scutellum, a median spot and the posterior angles on the median segment, and a transverse band on each side on dorsal segments 1-4 pale yellow; mandibles, tegulae, tarsi, and pygidium testaceous brown. Length 9 mm, Uab. Aden district. 25. Anthobosca minima Turn. Anthobosca viinima Turn. Trans. Ent. Soc. London, p. 398 (1910), $. 2 . Radial cell narrowly truncate at the apex ; second abscissa of the radius more than twice as long as the first, but much shorter than the third. Tarsal ungues bifid. Shining, sparsely and finely punctured ; apical dorsal segment strongly punctured and covered with stiff' fulvous hairs. Black ; mandibles, antennte, and legs testaceous brown ; STUDIES IN THE FOSSORIAL WASPS. 741 abdomen dark reddish brown ; yellow marks as in arahica, but the yellow band on each side of the first dorsal segment is absent. Length 5 mm. Hab. Mombasa. 26, Anthobosca chilensis Guer. Gosila chilensis Guer. Yoy. 'Coquille,'Zool. ii. p. 249 (1839), 5 ; Spinola, in Gay's Hist. Fis. Chile, Zool. vi. p. 312 (1851), J $ . ' Myzine Jlavicornis Sm. Descr. new spec. Hymen, p. 183 (1879), 2. 5 . Eadial cell rounded at the apex ; second abscissa of the radius twice as long as the first, but distinctly shorter than the third ; first recurrent nervure received close to the middle of the second cubital cell, second at about one-fifth from the base of the third cubital cell. Tarsal ungues bifid ; lobe beneath the apical third of the hind femora scarcely rounded, basal joint of hind tarsi with a thinly-set row of very short spines beneath. Finely and rather closely punctured, median segment finely rugulose ; abdomen shining, very finely and sparsely punctured . Basal joint of fore tarsi not strongly produced at the outer apical angle. Black, with long black pubescence ; calcaria black ; flagellum bright orange. Wings fusco-violaceous. S . Third abscissa of the radius twice as long as the second in some specimens, shorter than the second in other's ; first recurrent nervure received by the second cubital cell at a distance from the base slightly less than the length of the first transverse cubital nervure, second either interstitial with the second transvei-se cubital nervure or received a little before the apex of the second cubital cell. Apical joint of antennte no longer than the penultimate. Basal abdominal segment nearly twice as broad at the apex as long. Very finely and closely punctured, minutely on the abdomen. Colours as in the female. Length, 9 22 mm., c? 16 mm. Hah. Chile. 27. Anthobosca carbonaria Burm. Myzine carbonaria Burm. Stett. ent. Zeit. xxxvii p 168 (1876), 5 . _ Anthobosca carbonaria Turn. Trans. Ent. Soc, London p 83 (1908), $. Gosila carbonaria Brethes, Ann. Mus. Buen. Aires, xx. p 256 (1910), $. 5 . Radial cell narrowly rounded at the apex ; second abscissa of the radius longer than the first and third combined ; first recurrent nervure received at about two-fifths from the base of 742 MR, R. E. TURNER OX the second cubital cell, second before one-third from the base of the third cubital cell, which is nearly twice as long on the cubitus as on the i-adius. Basal joint of hind tarsi with a row of fine spines beneath ; basal joint of fore tarsi with seven spines above and a row of short spines beneath ; tarsal ungues cleft ; apical half of hind femora broadly rounded beneath ; tarsal ungues cleft. Shining, finely aiid sparsely punctured ; front and pro- notum more closely and coarsely punctured ; median segment subopaque, very finely and closely punctured, the posterior slope rather indistinctly transversely sti'iated . Pygidium densely clothed with fusco-ferruginous pubescence. Basal joint of fore tarsi not very strongly produced at outer angle. Black, with grey pubescence ; calcaria pale brownish. Wings rather light fuscous, fusco-hyaline at the apex. Length 20 mm. Hah. Nova Fribourg, S. Brazil. 28. Anthobosca bipunctata Perty. Tiphia bipunctata Perty, Delect, a.nim. artic. Brasil. p. 139 (1833), 2 . _ Myzine bipunctata Sm. Cat. Hym. B. M. iii. p. 76 (1855). Anthobosca bipunctata Turn. Trans. Ent. See. London, p. 83 (1908). $ . Radial cell narrowly rounded at the apex ; first and fourth abscisspe of the radius about equal in length, second and third also nearly equal to each other and twice as long as the first ; the first recurrent nervure received just beyond two-fifths from the base of the second cubital cell, second at the middle of the third cubital cell, which is shorter on the radius than on the cubitus. Tarsal ungues bifid ; lobe beneath the posterior femora extending over the apical two-thirds of their length and scarcely rounded. Shining and very sparsely punctured, the front smooth, apical dorsal seg- ment covered with long fuscous hair. Basal joint of fore tarsi strongly produced at the outer apical angle. Black ; a yellow spot on each side of the third dorsal segment. Wings fusco-hyaline, nervures fuscous ; calcaiia whitish. Length 16 mm. Hab. Minas Geraes, Brazil. Allied to carbonaria Burm., but has the lobe beneath the hind femora longer, the basal joint of the fore tarsi is more produced, and the puncturation is sparser, A. bipttstulata Turn, is probably the male of this species, 29. Anthobosca albomaculata Sm. (PI. LXXXI. figs. 7, 8.) Myzine albomaculata Sm, Descr, new spec. Hymen, p. 181 (1879), $ S. Myzine lecointei Ducke, Rev. entom. p. 146 (1907), 5 . $ . Radial cell narrowly rounded at the apex ; second abscissa STUDIES IN THE FOSSORIAL WASPS. 743 of the radius a little longer than the first, the third longer than the first and second combined ; first recurrent nervure received just beyond two-fifths from the base of the second cubital cell, second close to the middle of the third cubital cell. Tarsal ungues bifid ; lobe beneath the basal half of the hind femora not very prominent and very feebly rounded ; basal joint of hind tarsi with a row of short hairs beneath ; basal joint of fore tarsi strongly produced at the outer apical angle. Shining and sparsely punctured ; median segment opaqvie, closely and minutely punc- tured ; apical dorsal segment coarsely punctured and thickly clothed with long fuscous hairs. Black ; a spot on each side near the anterior angles of the pro- notum, a spot on the mesonotum, a spot at the base of the median segment and one at each of the posterior angles, and a spot on each side of the second and third dorsal segments yellowish white ; calcaria whitish. Wings pale fusco-hyaline, nervures fuscous. d . Second abscissa of radiiis twice as long as first, third about half as long again as second ; first recurrent nervure received a little beyond the middle of the second cubital cell, second at two- fifths from the base of the third cubital cell. Tarsal ungues bifid ; hind tibise serrate on the outer margin. Basal abdominal segment nearly half as long again as the breadth at the apex. Finely and rather closely punctured, abdomen finely sha greened. Black ; mandibles, clypeus, scape beneath, a spot on each side of the pronotum, a spot on the mesonotum, a spot on the scutellum, one on the postscutellum, another at the apex of the median segment, the base of the tibiae, and the tarsi pale yellow. Wings hyaline iridescent, nervures nearly black. Length, $ 12 mm., c? 9 mm. ITab. Amazon, from Para to Ega. 30. Anthobosca antennata Sm. Anthohosca antennata Sm. Descr. new spec. Hymen, p. 174 (1879), cT_. _ Cosila jheringi Saussure in Grandidier, Hist. Madagascar, xx. p. 234 (1892), 2 . Thynnus antennatus D. T. Cat. Hymen, viii. p. 101 (1897), c? . $ . Radial cell very narrowly rounded at the apex ; third abscissa of radius half as long again as the second, which is nearly twice as long as the first, the fourth shorter than the first ; first recurrent nervure received a little before the middle of the second cubital cell, second at one-fifth from the base of the third cubital cell. Tarsal ungues bifid ; lobe beneath the hind femora commencing near the base and extending to the apex, broadly rounded ; basal joint of hind tarsi with a row of very short fine spines beneath ; basal joint of fore tarsi strongly produced at the outer apical angle. Shining, very finely and rather sparsely punctured ; median segment subopaque, very closely and minutely 744 MR. R. E. TURNER ON puncturetl; apical dorsal segment densely clothed with long fulvous hairs. Black ; mandibles at the base, antennpe, and legs fei-ruginous : an interrupted band on the posterior margin of the prouotum, a spot on the mesonotum, one at the base of the median segment and another at each of the posterior angles, and an intei'rupted band on the four basal doisal segments, more bioadly interrupted on the second than on the other segments, yellow. Wings very pale flavo-hyaline, nervures ferruginous, cf . Third abscissa of the radius half as long again as the second, which is about equal to the fourth and twice as long as the first ; first recurrent nervure received at or a little beyond the middle of the second cubital cell, second at one-third from the base of the third cubital cell. Tarsal ungues bifid ; hind tibife serrate on the outer margin. Antennae short and stout, tapeiing to the apex ; pronotum rounded anteriorly ; first abdominal segment a little longer than the breadth at the apex, the apical dorsal margin of the segment broadly rounded. Finely and closely punctured ; abdomen finely shagreened. Black ; mandibles, clypeus, scape beneath, posterior margin of the pronotum broadly, a spot on the mesonotum, one on the scutellum, another on the postscutellum, and the greater part of the tibife and tarsi yellow. "Wings hyaline, nervures fuscous. Length, 5 13-16 mm., (5 12-14 mm. Hah. S. Brazil ; Rio Grande do Sul. As I have noticed before, the anteun?e in the t5^peof antennata are a little shorter and stouter than in the males sent with jheringi. This diflTerence may possibly prove to be specific, but I do not think it is. *31. Anthobosca erytiiropyga Burm. Myzhie erytliro'pyga Burm, Stett. ent. Zeit, xxxvii. p. 169 (1876), $ c?. Anthobosca erythrojjjjga Turn. Trans. Ent. Soc. London, p. 83 (1908). Cosila erythropyga Br^thes, Ann. Mus. Buen. Aires, xx. p. 256 (1910). I have not seen this species. Schrottky suggests that it is identical with A. jheringi Sauss,, but Burmeister's desci'iption gives the legs of the male as red, of the female black, and the five basal abdominal segments in both sexes with lateral yellow sjjots, whereas in jheringi the four basal segments are banded with yellow and the legs i-ed in the female, and the male has the abdomen without yellow marks and the legs yellow and black. I do not think there is sufficient ground for consideiing the two to be identical, especially as Schrottky had not seen specimens of jheringi. STUDIES IN THE FOSSORIAL WASPS. 7-45 *32. Antiiobosca apicalis Sicliel. Cosila apicalis Sichel, Sauss. et Sichel, Oat. spec. gen. Scolia, p. 262 (1864), 2 • I have not seen tliis species, which may possibly be identical with carbonaria Burm., but the dorsal segments of that species are not reddish laterally. Males, 1. Species from the Old World ;..,.;.... 2. Species from South America 18. 2. Australian spQcies 3. African species ; 14. 3. Tarsal ungues with a blunt lobe at the base ; second and third cubital cells each receiving a recurrent nervure 4. Tarsal ungues bitid, without a basal lobe ; second cubital cell receiving both recurrent nervures A. crassicornis Urn. 4. Eighth and three following joints of the liagellum sub- tuberculate beneath, apical joint longer than penulti- mate and slightly curved. Wings fusco- violaceous... A. australis Sichel None of the joints of the flagellum subtuberculate {==n7ffri/pen7iis Sin.)- beneath, apical joint no longer than penultimate. Wings hyaline or subhyaline 5. 5. First abdominal segment distinctly longer than the second ; fully half as long again as its breadtli at the . apex A, australnsice Giu'v. First abdominal segment not distinctly longer than the ■ second; scarcely, if at all, longer than its apical breadth 6, 6. Thorax and abdomen wholly black " 7. Pronotum at least marked with yellow , 9. 7. Legs ferruginous A. varipes Sn\. Legs black g. 8. Median segment transversely rugulose, very short and broad ; anterior tibiae black beneath A. ceth inps Sm. Median segment finely punctured, not very short or broad ; anterior tibiae ferruginous beneath A. nigra Sm. 9. Legs ferruginous 10. Legs black, more or less marked with yellow 11. 10. Postscutellum yellow, legs marked with yellow ; first transverse cubital nervure as long as the second or longer and strongly bent near the cubitus A. torresensis Turn. Postscutellum and legs without yellow markings ; first " transverse cubital nervure much shorter tlian the second, and not bent near the cubitus A. gilesi Turn. 11. Five basal abdominal segments with yellowish lateral spots; cubitus of hind wing originating just before the transverse median nervure, almost interstitial ... A. moderata Turn. Five basal abdominal segments immaculate ; cubitus of hind wing originating very distinctly before the transverse median nervure 12. 12. Seventh dorsal segment with yellowish lateral spots ; postscutellum yellowish ]3. Seventh dorsal segment and postscutellum immaculate . A.frencJii Turn. 13. Clypeus black ; mesonotum without a yellow spot ; hind tibiae feebly spinose A. lonff ipalpa Tavn. Clypeus yellow ; mesonotum with a yellow spot ; hind tibiae feebly serrate A. lagardei Twrn. • 14. Apical margin of clypeus not toothed 15. Apical margin of clypeus with two small teeth 16. Puoc. ZooL. Soc— 1912, No. XLIX, 49 746 MR. R. E, TURNER ON 15. Legs wholly ferruginous; mesouotum with a yellow spot ; tegulse black A. errans Sin. Legs ferruginous marked with j'ellow; mesonotum - immaculate; tegulee -yellow A.fiavopictaTwvn, 16. Wings hyaline 17. Apical half of wings with a distinct fusco-violaceous tinge A. bidentata Cam. 17. Abdomen immaculate A. lactei_pennis Cam. Dorsal abdominal segments 2-4 with large yellowish lateral spots A. erytlironota Cam. 18. Antennae orange ; wings fusco-violaceous A. chilensis G:ue.v . Antennae black ; wings hyaline 19. 19. Abdomen spotted with yellow *A. erythropyga Burm . Abdomen immaculate 20. 20. Legs ferruginous A hipustiilata Turn. Legs black, marked with yellow 21. 21. Length 13 mm. Yellow band on pronotum entire A. antennata Sm. Length 10 mm. Yellow band on pronotum interrupted. A. albomacuJata Sm. 33. Anthobosca AUSTRALASIA Guer. (PL LXXXIII. fig. 13.) Anthohosca australasice Guerin ; Duperry, Yoy. ' Ooquille/ Zool. ii. p. 237 (1839), 6 . Anthohosca crabroniformis Sm. Cat. Hym. B. M. vii. p. 59 (1859), c^. , Thynnus cathreinii D. T. Cat. Hym. viii. p. 103 (1897), . T. Cat. Hymen, viii. p. 114 (1897), d' . c5' • Clypeus truncate at the apex ; antennse nearly as long as the thorax and median segment combined. Pronotum short, scarcely more than half as long as the scutellum, very slightly narrowed anteriorly. First dorsal segment longer than the apical breadth ; seventh dorsal segment broadly rounded at the a,pex. Hind tibife rather feebly serrate ; hind femora without a lobe at the apex beneath. Second abscissa of the radius longer than the first, the two combined much shorter than the third; first ti-ansverse cubital nervure oblique, sharply bent close to the cubitus ; first reciu^rent nervure received at the middle of the second cubital cell, second at two-fifths from the base of the third cubital cell. The distance between the cubitus of the hind wing and the transverse median nervure is greater than half the length of that nervure. Black ; the fore tibiae ferruginous within ; calcaria whitish. Wings hyaline, nervures fuscous. , Length 9-12 mm. JIab. Victoria and South Australia,. , Differs from varipes in the absence of a carina on the scutellum, in the colour of the legs, and the more distinct serration of the hind tibise. 38. Anthobosca ^thiops Sm. Anthohosca cethiovs Sm. Descr. new spec. Hymen, p. 175 (1879), s. Thynnus stolzii D. T. Catal. Hymen, viii. p. 116 (1897), (5 • c? . Antennae stouter than in other Australian species ; pro- notum very slightly narrowed anteriorly, shorter than the scutellum. Head and thorax very closely and finely punctured, median segment transversely rugulose, short and not as strongly convex as in other species. First dorsal segment very little longei' than the apical breadth ; seventh dorsal segment narrowly STUDIES IN THK FOSSOKIAL WASPS. 749 truncate at the apex. Hind tibise spinose ; hind femora without a rounded lobe at the apex beneath. Second abscissa of the radius equal to the first, the two combined shorter than the third ; first transverse cubital nervure oblique, sharply bent close to the cubitus; first recurrent nervure received close to the middle of the second cubital cell, second just before one-third from the apex of the third cubital cell. Cubitus of the hind wing separated from the transverse median nervure by a distance equal to one-third of the length of that nervure. Black ; the inner margin of the eyes narrowly whitish ; calcaria whitish ; wings hyaline, tinged with fuscous, nervures fuscous. Length 14 mm. Hah. Champion Bay, W. Australia. 39. Anthobosca moderata Turn. (PI. LXXXI. fig. 4.) Anthohosca moderata Turn. Ann. & Mag. Nat. Hist. (8) iii. p. 482 (1909), S. S . A minute tubercle in the middle of the clypeus ; pronotum not much narrowed anteriorly, not more than half as long as the scutellum ; first dorsal segment a little narrower at the apex than long, a little longer than the second segment. Hind tibiae very distinctly spinose ; hind femora with a very distinct rounded lobe at the apex beneath. Second abscissa of the radius longer than the first, the two combined about equal to the third ; first recurrent nervure received at ,the middle of the second cubital cell, second beyond one-third from the base of the third cubital cell. Cubitus of the hind wing originating just before the transverse median nervure. Black; mandibles at the base, clypeus, posterior margin of pronotum, tegulse, a spot on the mesonotum, postscutellum, a spot on each side at the apex of the median segment, a small lateral spot on each side of dorsal segments 1-5, a large spot on each side of the seventh segment, the base of the tibise and the base of the tarsi whitish yellow; the apex of the seventh dorsal segment testaceous. Wings hyaline, tinged with fuscous ; nervures fuscous. Length 12 ixim. Hab. Townsville, Queensland. 40. Anthobosca frenchi Turn. Anthobosca frenchi Turn. Proc. Linn. Soc. N.S.W. xxxii. p. 518(1907), d. o' . Clypeus truncate at the apex ; pronotum much shorter than the scutellum, very slightly narrowed anteriorly. First dorsal segment no longer than the apical breadth ; seventh dorsal segrnent broadly rounded at the apex. Hind tibise very feebly serrate; hind femora without a lobe at the apex beneath. Second abscissa of the radius longer than the first, the two 750 MU. K. E. TURNER ON combined as long as the third ; first transverse cubital nervure oblique, strongly bent near the cubitus, longer than the second ; first recurrent nervure received at the middle of the second cubital cell, second at one-third from the apex of the third cubital cell ; the distance between the cubitus of the hind wing and the transverse median nervure equal to less than half the length of that nervnre. Black ; base of the mandibles, a small spot round the base of each antenna, the hind margin of the pronotum very narrowly, tegulse, anterior tibire in front, and the base of the hind tibite pale yellow ; apex of the mandibles and fore tarsi fuseo-ferruginous. "VVings hyaline, nervures dark fuscous. Length 7-8 mm. Hah. Victoria. 41, Anthobosca lagardei Turn. Anthohosca lagardei Turn. Trans. Ent. Soc. London (1908), p. 86, c?. c5 . Differs from A. frenchi in the yellow clypeus, yellow spot on the mesonotum, postscutellum, apex of median segment, and on each side of the seventh dorsal segment ; in the rounded lobe at the apex of the hind femora beneath, which is well developed in lagardei, and in the yellow basal joints of the tarsi. Length 8 mm. Hah. Sydney, New South Wales. 42. Anthobosca longipalpa Turn. Anthohosca longipalpa Turn. Proc. Linn. Soc. N.S.W. xxxii. p. 517(1907), 6. S. Clypeus very feebly emarginate at the apex; the apical joint of the maxillary palpi filiform, longer than the penultimate joint, the three apical joints much more slender than the basal ones. Pronotum short, very slightly narrowed anteriorly. First dorsal segment a little longer than the apical breadth ; seventh dorsal segment narrowly truncate at the apex. Hind tibise feebly spinose ; the lobe at the apex of the hind femora beneath not distinct. Second abscissa of the radius nearly twice as long as the first, the two combined longer than the third. First transverse cubital nervure curved, much longer than the second, and rather sharply bent near the cubitus ; recurrent nervures received a little before the middle of the second and third cubital cells. Cubitus of the hind wing originating before the transverse median nervure, the distance between it and that nervure equal to about one-third of the length of the transverse median nervure. Black ; posterior margin of the pronotum very narrowly, a transverse line on the postscutellum, a spot on each side of the seventh dorsal segment, and the base of the hind tibiee whitish STUDIES IN THE FOSSORIAL WASPS. 751 yellow ; tibi?e fusco- ferruginous ; calcaria white. Wings hyaline, nervures fuscous. Length 12 mm. Uah. Cairns, Queensland. 43. Anthobosca crassicornis Sm. (PI. LXXXI. fig. 1.) Tachypterus crassicornis Sm. Cat. Hym. B. M. vii. p. 64 (1859), J. Trachypterus crassicornis D. T. Cat. Hymen, viii. p. 120 (1897), c?. Anthohosca crassicornis Turn. Proc. Linn. Soc. N.S.W. xxxii. p. 519(1907), 6. 5 . Clypeus produced in the middle and rather narrowly trun- cate at the apex. Antennae slightly longer than the thorax and median segment combined. Head and thorax very closely and not very finely punctured. First abdominal segment scarcely longer than the second ; the apex of the dorsal segment rounded and a little, but not very distinctly, narrower than the first segment is long. Posterior tibise very distinctly serrate, tarsal ungues bifid. Second abscissa of the radius longer than the third ; both recurrent nervures received by the second cubital cell, the first at two-fifths from the base, the second at about nine-tenths or more from the base. Third cubital cell as long on the radius as on the cubitus. Black, with long grey pubescence ; mandibles, clypeus, lower portion of the inner mai^gin of the eyes, anterior coxae, trochanters and femora beneath, and a line beneath the intermediate femora yellow ; second and third abdominal segments and the base of the fourth, tibiae, tarsi, hind femora, and intermediate femora above ferruginous. Wings hyaline, nervures fusco-ferruginous, stigma fuscous. Length 13 mm. Hah. Australia. A variety has the ferruginous colour much obscured. As mentioned under that species, I believe this to be the male of A. clypeata Sm , owing to the similarity of the structure of the tar«al ungues, the position of the recurrent nervures, the length of the second abscissa of the radius, and some similarity of colour. 44. Anthobosca errans Sm. Anthohosca errans Sm. Descr. new spec. Hymen, p. 174 (1879), 6. 6 ■ Clypeus broadly truncate at the apex, not toothed ; an- tennae short and stout. Pronotum shorter than the scutellum, distinctly narrowed anteriorly ; scutellum and median segment obscurely longitudinally carinated in the middle. First dorsal segment longer than the apical breadth ; seventh dorsal segment 752 MR. R. E. TURNER OK" narrowly truncate at the apex. Hind tibise serrate ; hind femora with a rounded lobe at the apex beneath. Second abscissa of the radius nearly twice as long as the first, the two combined shorter than the third ; first transverse cubital nervure oblique, strongly bent close to the cubitus, longer than the second ; the distance between the cubitus of the hind wing and the transverse median nervure equal to about one-third of the length of that nervure. Tarsal ungues bifid. Black ; base of the mandibles, clypeus, a spot on each side of the pronotum, a spot on the mesouotum, one on the scutellum and a transverse line on the postscufcellum yellow ; legs ferru- ginous. Wings hyaline, nervures fusco-ferruginous. Length 13 mm. Hah. Zululand. 45. Anthobosca flavopicta Turn. Anthohosca flavojncta Turn. Trans. Ent. Soc. London, p. 399 (1910), 6 pi. L. fig. 4. S . Clypeus subtruncate at the apex, not toothed. Postei'ior tibiae serrate, tai'sal ungues bifid. Black ; clypeus with a black spot in the middle, inner margin of the eyes narrowly, a small spot behind the eyes near the summit, a broad band on the posterior margin of the pronotum, tegulse, a large spot on the scutellum, one on the postscutellum, the apex of the fore femora, and the anterior and intermediate tibiae and tarsi above pale yellow ; tibiae and tarsi ferruginous brown. Wings hyaline, nervures black. Length 11 mm. Hah. Zanzibar. 46. Anthobosca bidentata Cam. Odontothynnus hidentatus Cam. Rec. Albany Mus. i. p. 162 (1904), c^. (S . Apex of clypeus bidentate. Hind tibiae serrate. First transverse cubital nervure rounded. Wings hyaline, the apical half fusco -violaceous. Black ; lower half of inner orbits of the eyes, face, clypeus, base of mandibles, base and outer side of fore and intermediate tibiae, calcaria, tarsi, base of hind tibia?, and the posterior margin of the pronotum broadly yellow. Length 11-12 mm. Hah. Grahamstown, S. Africa. 47. Anthobosca lacteipennis Cam. Odontothynnus lacteipennis Cam. Rec. Albany Mus. i. p. 162 (1904), d- J . Apex of clypeus bidentate. Hind tibiae serrate. First STUDIES IN THE FOSSORIAL WASPS. 753 ti-ansverse cubital nervure obliqvie, "Wings hyaline, nei^vures white. Colours as in bidentata. Length 10 mm. Hab. Grahamstown, S. Africa. I have seen neither bidentata nor lacteipennis. Cameron's statement that the ungues of the hind tai'si are simple, not bifid, needs confirmation. 48. Anthobosca bipustulata Turn. Anthobosca bipustulata Turn. Ann. & Mag. Nat. Hist (8) iv. p. 343(1909), 6. S . Slender ; antennae stout, tapering towards the apex ; pronotiim scarcely more than half as long as the scutellum, moderately narrowed anteriorly. First dorsal segment nearly half as long again as the apical breadth ; seventh dorsal segment rather narrowly rounded at the apex. Tarsal ungues bifid ; hind tibia? serrate ; hind femora without a distinct lobe at the apex beneath. Apical joint of maxillary palpi filiform. Second abscissa of the radius twice as long as the first, the two combined nearly as long as the third; first transverse cubital nervure oblique, sharply bent near the cubitus, longer than the second ; the distance between the cubitus of the hind wing and the transverse median nervure equal to half the length of that nervure. Black : mandibles, clypeus, scape beneath, a spot on each side on the posterior margin of the pronotum, the base of the tegulfe. a small spot on the mesonotum, one on the scutellum, another on the postscutellum, the two latter spots sometimes absent, and the anterior coxae beneath yellow ; legs ferruginous. Wings hyaline, nervures fuscous. Length 8-12 mm. Hab. Barbacena, Minas Geraes. A variety has the legs black, marked with yellow. EXPLANATION OP THE PLATES. Plate LXXXL 1. AniJiol osca crassicornis Sm. $. 2. AntJiolosca olypeata Sm. $ . 3. Antliohosca fastuosa Sm. $ . 4. Anthobosca moderata Turn. (J . 5. Antliohosca erytlironota Cam. $ . 6. Antliohosca erytlironota Cam. ^ . 7. Anthobosca albomaculata Sxa. (J. 8. Antliohosca albomaculata Sm. $ . 9. Braunsomeria quadraticeps Turn. (?. Proc. Zool. Soc— 1912, No. L. Fig. 10. Braunsomeria qiiadraticeps Turn. 11. Myzine stigma Turn. $ . 12. Mis (Mesa) alicia Turn. $ . 13. JSlis (Mesa) Innc/lventris Turn. $. 14. Mi/zine braunsi Turn. $ . 15. JElis (Mesa) dimidiaticornis Bingh. $. 16. Jiilis (Mesa) tricolor Sm. $ . 50 754 MR. ABEL CUAPMAN ON PXATE LXXXII. Nenration of Wings. Fig, 1. AntJwhosca australis Sicliel. ? . 2. Antholosca australis Sichel. ^ • 3. Anthohosca anthracina Sm. ^ . 4. Anthohosca erythroiiota Cam. $ . 5. AntJwhosca cli/peata Sm. $ . 6. Anthohosca insularis Sm. $ . 7. Sraunsomeria qwadratieeps Turn. 8. JEZis (Mesa) alicice Turn. Y • rig. 9. JEIis (Mesa) rwficeps Sm. '■^ . 10. JBZis (Mesa) rujiceps Sm. (j . 11. JE??«s (Mesa) tricolor Sm. $ . 12. ^Ws comhusta Sm. 9 • 13. Myzine stigma Turn. $. 14. Myzine braunsi Tura. (? . 15. 3Iyzine constrictiventris Turn. 10. Mysitve ahdomiTicdis Guer. § . Plate LXXXIII. Exoskeletfil Stractnres. Suture between two basal abdominal Suture between two basal abdominal Fig. 1. Anthohosca australis Sichel. segments. 2. Slis (Mesa) rujiceps Sm. segments. Z. Anthohosca australis ^\c\\e[. $. Apical ventral segment. 4. Bilis (Mesa) ruficeps Sm. S ■ -Apical dorsal segment. 5. 3£ysine ahdominalis Guer. S • 6. Anthohosca australis Sichel. 7. Anthohosca clypeata Sm. $ . 8. Anthohosca insularis Sm. ? . 9. J7Zis (Mesa) aliciaiT\xm. ?. 10. JEZjs (Mesa) ruficeps Sm. $ . 11. JSlis (Mesa) longiventris Turn. $. 12. Myzine constrictiventris Turn. $ . 13. Anthohosca australasicB Guer. ^. „ „ 14. Anthohosca australis Sichel. $ . „ „ 15. JSlis (Mesa) ruficeps Sm. $ . „ „ 16. JSlis (Mesa ) ruficeps Sm. ^ . Intermediate and hind coxae. l . Tarsal unguis. Intermediate and hind cosse. Basal joint of hind tarsus. Two basal segments of abdomen. 40. Notes on the Spanish Ibex. By Abel Chapman, F.Z.S. [Eeceived April 13, 1912 : Eead May 7, 1912.] Inpes. Page Distribution 755 In Profe.s.sor Angel Cabrera's most informative paper on Ccqwa pyrenaica* two small points occur to me as worthy of brief comment. In discussing the present distribution of the Spanish Wild- Goat in some six isolated colonies, Dr. Cabrera takes exception to the remark in ' Unexplored Spain ' (by Walter Buck, C.M.Z.S., and myself) that they hacl been so isolated "during ages." Well, the term used in our former book ('Wild Spain') was " during centuries," and that is certainly more definitive and probably more accxirate. Dr. Cabrera, however, goes on to state that there exist "strong reasons for believing that in the past * P.Z.S. 1911..P. 963. THE SPANISH IBEX. 755 [inferentially as late as the middle of the seventeenth century — say 250 years ago] Ibex inhabited every suitable point of almost every mountain-ridge in Spain," The only reason actually adduced, however, is the prevalence of place-names based upon, or compounded with, the Spanish word Cobra = goat. Such names, it is true, are ubiquitous ; but it would never have occurred to me that those names necessarily refer to the loild goat. Spain is a land of goats, and many localities bearing names such as Sierra de las Cabras, Cabrales, Cebrero, and so on, are not at all adapted to the nature and requirements of the wild Ibex. I would suggest that, in many cases, the names merely indicate the existence of suitable local pasturage for domestic goats, which are herded everywhere. Again, Dr. Cabrera translates the Spanish name of the Ibex, Cobra monies^ as equivalent to " Mountain-Goat." Now it woidd be nothing less than presumptuous for me, a foreigner with but a limited colloquial knowledge of the Spanish tongue, to question his rendering. I do not do so. I accept that as the pure classic Castilian of Madrid. But I do venture to say that, in wilder Spain, the term monte, with its derivative adjective inontes (pronovmced niotitess), possesses quite a different signification, Monte may occasionally, and in combination, be used to indicate a hill or mountain ; but in its ordinary provincial sense, it signifies scrub or brushwood. Thus the wild-cat, which is equally common on lowland or sierra, is known as Gato mooites = Scrub-Cat : on the low-lying plains of Andalucia or Estremadura, the expression Reses mo7iteses includes all the scrub-haunting animals — such a,s deer, wild-boar, lynx, etc. Now, viewed in this light, it has always appeared to me uttei-ly inexplicable and incongruous to apply the name Cabra mantes, or Scrub-Goat, to the Ibex of the higher ranges, such as the Sierias de Gredos and Nevada, where the Ibex live exclusively amidst rock-regions far above the topmost levels of scrub. But such incongruity would disappear if Dr, Cabrera's assumption were correct, that the Ibex, up to a couple of centuiies ago, occupied the Avhole vast area shown in the map at p. 965, supra. Was such the distribution, that name would become appropriate enough, since an immense proportion of the dotted area consists, not of high mountains at all, but of low scrub-clad hills. Such country might appear, to preconceived ideas, totally unsuitable for Ibex : but we have the fact before us (as fully explained in our books on Spain) that in several of the lower Mediterranean sierras (some of which are bush-clad to the summits) the local Ibex do to-day take kindly to a bush-haunting habit. Indeed, in such situations, it is obvious, they have no other option. This latter point tends to support Dr. Cabrera's assumption, and equally, of course, undermines our own. In our two books we had pointed out that the Ibex of the two extreme ends of Spain [i. e., those of the Pyrenees and those of 756 ox THE SPAXISH IBEX. the Mediterranean sierras) most nearly assimilated to each other in their more flattened and laterally compressed horns*. It is gratifying to find that our rough field -observations are now corroborated by Dr. Cabrera's skilled investigation and careful cross-sections. But again, it appears anomalous — assuming that complete isolation only commenced some two centuries ago — that the central group (now honoured with subspecific i^ank as C. p. victorice) should have developed the greater difterence. The females and young males of the Spanish Ibex are devoid of the dark dorsal stripe, as is correctly shown in the plate in ' Unexplored Spain,' at p. 140. They are of a dun-brown, uni- colorous in coat as the Spanish Red-Deer ; but Dr. Cabrera is quite justified in criticizing the second plate (op. cit. p. 967), at p. 216. That slip should be debited, in the first instance, to the artist, Mr. E. Caldwell, but the fault is wholly mine, since I should have detected the mistake and had it corrected before passing the drawing for reproduction. If permissible to express an opinion on the three beautiful plates given by Dr. Cabrera, I would say that in life the Spanish Ibex is rounder and bulkier in the barrel than can ever be gathered from museum specimens, since skins shrink. In conclusion, may I express a fervent hope (since interest in her vanishing Ibex has been aroused in Spain) that further protection may be extended to the few surviving colonies ? Within my own time, Ibex have been exterminated in several of tlieir earlier haunts. To-day they are at their last gasp in the Pyrenees and in the Gerez (Portugal). Fortunately, in Gredos, Morena, and Bermeja, their future has been assured — though only at the eleventh hour. Can our Spanish friends not see to safeguarding the much-menaced remnant that yet survives on the main chain of Nevada ? * Unfortunatelj', in ' Unexplored Spain,' in an effort to be concise, and to avoid repetition, we omitted the word " laterally " ; but a reference to ' Wild Spain,' p. 129, makes our meaning clear. Papers {continued). I'agft 23. Experimental Pheasant-breeding. By Eose Haig Thomas, F.Z.S, (Pis. LXIV.-LXVII.) 53S* 129. A List of Moths of the Family Pyralidae collected by Felix B. Pratt and Charles B. Pratt in Dutcli New Guinea in 1909-10 ; with Descriptions of new Species. By Sir GEoiiGB H. Kb-vkick, F.Z.S, (PI. LXVIII.) 546 30. On a rare Stag (Cervus wallichii) from Nepal recently presented to the Zoological Society by His Majesty King George. By E. I. Pocock, F.E.S., F.L.S., F.Z.S., Superintendent of the Gardens. (Text-figs. 66-71.) .558 31. Contributions to the Anatomy and Systematic Arrangement of the Cestoidea. — IV. On a Species of Inermicapsifer from the Hyrax, and on the Genera ZschoJcJceella, Thy- sanotmnia, and Hyracotania. By Feank E. Bbddard, M.A,, D.Sc, F.R.S., F.Z.S. , Prosector to the Society, (Text-figs. 72-83) 576 32. Additional Notes on the Liring Specimens of the Australian Lung-fish {Ceratodus forsteri) in the Collection of the Zoological Society of London. By Bashford Dbajt, (Text-figs. 84 & 85.) GG7 33. The Circulatory System of the Common Grass-Snake {Tropidonotus natrix). 'Bj Chas. H. C'Donoghue, B.Sc, F.Z.S., Assistant to the Jodrell Professor of Zoology, University College, London. (Pis, LXX.-LXXIL and Text-figs. 86-91.) 012 34. A First Account of the Courtship of the Redshank {Totanus calidris). By Julian S. HtjXley, Lecturer of BaUiol College, Oxford 647 35. Some Brackish-water Araphipoda from the mouths of the Weser and the Elbe, and from the Baltic. By E. W. Sexton, Marine Biological Laboratory, Plymouth. (Pis, LXXIII. & LXXIV.) 656 36. Descriptions of new Fishes of the Family Loricariid» in the British Museum Collection. By C, Tate Eegan, M.A., F.Z.S, (Pis. LXXV.-LXXVII.) 666 37. On a Collection of Fishes made by Mr. A. Blayney Percival in British East Africa to the East of Lake Baringo. By G, A. Boulesgek, F.E.S., F.Z.S. (Pis. LXXVIII.- LXXX.) 672 38. Contributions to the Anatomy and Systematic Arrangement of the Cestoidea. — V. On a new Genus {Basyurotcsnia) from the Tasmanian Devil (Dasyurus ursinus), the type of a new Family. By Feank E. Beddard, M.A., D.Sc, F.E.S., F.Z.S., Prosector to the Society. (Text-figs. 92-101.) 677 39. Studies in the Fossorial Wasps of the Family Scoliidse, Subfamilies Eliding and Anthoboscinse, By Eowland E. Turner, F.Z.S., F.E.S. (Pis. LXXXI.-LXXXIII.) 696 40. Notes on the Spanish Ibex. By Abel Cuapman, F.Z.S 754 LIST OF PLATES. 1912, Part III. (pp. 505-576). Plate LXI. LXII. LXIII. LXIV. LXV. LXVI. LXVIl. LXVIII. LXIX. LXX. LXXI. LXXII. LXXIII. LXXIV. LXXV. LXXVI. LXXVII. LXXVIIT. LXXIX. LXXX. LXXXI. LXXXII. LXXXIII. Alcyonaria from Singapore 505 I y Feathers from Pheasants 539 I ; Pyralidse from Dutch New Guinea 546 Young Cariama cristata 557 j. Circulatory System of the Grass-Snake 612 f Amphipoda from Bremerhaven , 656 1. Chcetostomus lepturus. 2. C. patccispinis. Z, C.palmeri.\ 1. Xenocara muliispinis. 2. JT. heterorhynchus. 3. Pleco- i stomus hondcB \ 66d 1. Oxyloricaria tamance. 2. 0. leightoni. 3. Otocinclus I maculipinnis ) Laheo percivali ^ . 1. Barhus argyrotmnia. 2. B. mimus. 3. Tilapia I percivali f'oi^ Amphilius oxyrhinus i Fossorial Wasps Wing Neuration of Fossorial Wasps \ 696 Exoskeletal structures of Fossorial Wasps NOTICE. The ' Proceedings ' for the year are issued in foi.ir parts, paged consecutively, so that the complete reference is now P. Z. S. 1912, p. is as follows : — Part I. issued in March. „ II. „ June. „ III. „ September. I, IV. „ December. The Distribution Proceedings,' 1912, Part II. (pp. 241-604), M-ere published on June 22nd, 1912. L^^ m M. mJC' PROCEEDINGS OP THE aENERiL MEETINaS FOR SCIENTIFIC BUSINESS OF THE ZOOLOGICAL SOCIETY OF LONDON. 1912. PART IV. CONTAINING Pages 757 TO 913, WITH 14 Plates AND 35 Text-figures. y^ 871:41918 DECEMBER 1(913. FE PllINTED FOR THE SOCIETY, SOLD AT THBIE HOUSE IN REGENT'S PARK. LONDON : MESSRS. LONGMANS, GREEN, AND CO., PATERNOSTEa ROW. ./ y [Price Twelve Shillings.] LIST OF CONTENTS. 1912, Part IV. (pp. 757-913). EXHIBITIONS AND NOTICES. Page The Secretary. Eeport on the Additions to the Society's Menagerie during the month of April 1912 806 Mr. A. BiiAYNEY Percival, F.Z.S. Exhibition of photographs and lantern-slides of Game Animals from British East Africa 806 Mr. D. Seth-Smith, F.Z.S. Exhibition of two living specimens of the Lory, Calliptilua solitariics, from Fiji 806 Mr. G. A. BouLENGEE, F.E.S., F.Z.S. Notice of a paper entitled " Second Contribution to our Knowledge of the Varieties of the Wall-Lizard " 807 Mr. E. G. BouLENGER, F.Z.S. Exhibition of a cocoon of the African Lung-fish {Protopterus annectens) 807 Mr. E. G. B. Meade- Waldo and others. Discussion on the Preservation of the Native Fauna of Great Britain 807 The Secretary. Eeport on the Additions to the Society's Menagerie during the months of May to September 1912 • 908 Mr. E. Lydekker, F.E.S., F.Z.S. Note stating that " Gazdla hayi "= Gazella fuscifrms . 911 The Eev. T. E. E. Stebbing, M.A., F.E.S., F.Z.S. Notice of a memoir on the Crustacea Isopoda collected by the ' Porcupine ' Expedition in 1869-1870 912 Sir Edmund G. Loder, Bt., V.P.Z.S. Demonstration of the capacity of the electric lantern presented by him to the Society 913 Mrs. EosE Haig Thomas, F.Z.S. Exhibition of the Eggs of Phasianus formosanus, P. versicolor, and their F. 1 and F. 2 offspring. (Text-fig. 126.) 912 PAPERS. 41. The Local Eaces of Burchell's Zebra. By Major J. Stevenson-Hamilton, C.M.Z.S., Game Warden of the Transvaal. (Text-figs. 102-106.) 757 42. On Dipteropeltis, a New Genus of the Crustacean Order Branchiura. By W. T. Calman, D.Sc, F.Z.S. (PI. LXXXIV.) _. 763 43. On two new Larval Trematodes from the Striped Snaie (Tropidonotus ordinatus sirtalis). By William Nicoll, M.A., D.Sc. M.D., F.Z.S., Lister Institute of Preventive Medicine, London. (Text-fig. 107.) 767 Contents continued on page 3 of Wrapper. THE ZOOLOGICAL SOCIETY OF LONDON. This Society was founded in 1826 by Sir Stamford Raffles, Mr. J". Sabine, Mr. N. A. Vigors, and other eminent Naturalists, for the advancement of Zoology and Animal Physiology, and for the introduction of new and curious subjects of the Animal Kingdom, and was incorporated by Royal Charter in 1829. Matron. HIS MAJESTY THE KING. COUNCIL. HIS GRACE THE DUKE OF BEDFORD, KG., President . The Earl of Altamont, P.S.A. SiE John Rose Bradford, K.C.M.G., M.D., D.Su., F.R.S., Vice-President. Richard H. Burne, Esq., M.A. Lt.-Col. Sir R. Havelocic Charles, G.C.V.O., M.D. Alfred Heneage Cocks, Esq., M.A. F. G. Dawirey Deewitx, Esq., M.A., M.D. Charles Drummond, Esq., Treasurer. Sir Edward Durand, Bx., C.B. F. Dp Ca^b Godman, Esq.,D.C.L., F.R.S. Sir Edmund G. Loder, Bt., Vice- President. Edmund G. B. Meade-Waldo, Esq., Vice-President. Professor Edward A. Minchin, M.A., F.R.S., Vice-President. P. Chalmers Mitchell, Esq., M.A., D.Sc, LL.D., F.R.S., Secretary. W. R. Ogiltie-Grant, Esq. Albert Pam, Esq. Adrian D. W. Pollock, Esq. Oldfield Thomas, Esq., F.R.S. Anxhony H. "Wingfield, Esq. A.Smith WoODWARD,EsQ.,LL.D., F.R.S., Vice-President. I Henry Woodward, Esq., LL.D., F.R.S., Vice-Presidenti The Society consists of Fellows, and Honorary, Foreign, and Corresponding Members, elected according to the By-Laws. It carries out the objects of its foundation by means of the collection of living animals, by its Library, and by its Scientific Publications. The Office of the Society, Regent's Park, N.W., where all com- munications should be sent, addressed to " The Secretary," is open from Ten till Five, except on Saturdays, when it closes at Two p.m. The Library, under the superintendence of Mr. F. IS. Waterhouse, is open daily at the above hours, except in September. The Meetings of the Society for General Business are held in the Meeting Eoom at the Society's Office on the third "Wednesday in every month of the year, except in September and October, at half- past Four o'clock p.m. The Meetings for Scientific Business are held in the Meeting Hoom at the Society's Office fortnightly on Tuesdays, except in July, August, September, and December and January, at half-past Eight o'clock P.M. The Anniversary Meeting is held on the 29th. of April, or the nearest convenient day, at Four p.m. The Society's Gardens are open daily from Nine o'clock until Sunset. Mr. R. I. Pocock, F.E.S., F.L.S., is the resident Super- intendent and Curator of Mammals, Mr. D. Seth-Smith is Curator of Birds and Inspector of "Works, and Mr. E. G Boulenger is Curator of Eeptiles. The Prosectorium for Anatomical and Patho- logical work is under the charge of Mr. Frank E. Beddard, M.A., D.Sc, F.E.S., Prosector, assisted by Mr. H. G. Plimmer, F.E.S., M.E.C.S., Pathologist to the Society. TEEMS FOE THE ADMISSION OF FELLOWS. Fellows pay an Admission Fee of £5, and an Annual Contri- bution of ,£3, due on the 1st. of Janiiary, and payable in advance, or a Composition of £45 in lieu thereof; the whole payment, including the Admission Fee, being £50. No person can become a Fellow until the Admission Fee and first Annual Subscription have been paid, or the annual payments have been compounded for. Fellows elected in November and December are not liable for the Subscription for the year in which they are elected. PRIVILEGES OE EELLOWS. Fellows liave Personal Admission to the Gardens upon signing their names in the book at the entrance gate, and may introduce Two Companions daily. The "Wife or Husband of a Fellow can exercise these privileges in the absence of the Fellow. For the year 1913, and until further notice, Fellows will receive 40 undated Green Cards, available on any Sunday or week- day up to the end of February of the year following the year of issue, and 20 White Cards available on any week-day up to the same date. Twenty of the Green Cards may be exchanged for a book containing two Orders for each Sunday in the year, and the twenty White Cards may be exchanged for a similar book of Saturday Orders, Special children's tickets will no longer be issued, but the Green and White Cards will be perforated, and each half will be valid for a Child under twelve years of age. Fellows are not allowed to pass in friends on their written order or on presentation of their visiting cards. Fellows have the privilege of receiving the Society's ordinary Publications issued during the year upon payment of an additional Subscription of One Guinea. This Subscription is due upon the 1st. of January, and must be paid before the day of the Anniversary Meeting, after which the privilege lapses. Fellows are likewise entitled to purchase these Publications at 25 per cent, less than the price charged to the public. A further reduction of 25 per cent, is also made upon all purchases of Publications issued prior to 1881, if above the value of Five Pounds. Fellows also have the privilege of subscribing to the Annual Volume of ' The Zoological Record,' which gives a list of the Works and Publications relating to Zoology in each year, for the sum of One Pound Ten Shillings. Separate divisions of volumes 39 to 42 can also be supplied. Full particulars of these publications can be had on aj^plication to the Secretary. Fellows may obtain a Transferable Ivort Ticket admitting two persons, available throughout the whole period of Fellowship, on payment of Ten Pounds in one sum. A second similar ticket may be obtained on payment of a further sum of Twenty Pounds. Any Fellow wlio intends to be absent from tbe United Kingdom during the space of at least one year, may, upon giving to the Secretary notice in writing, have his or her name placed upon the " dormant list," and will then be called upon to pay an annual subscription of ^1 only during such absence, but after three years must make a further application to be retained on that list. Any Fellow, having paid all fees due to the Society, is at liberty to withdraw his or her name upon giving notice in ivriting to the Secretary. Ladies or Gentlemen wishing to become Fellows of the Society are requested to communicate with " The Secretary." Regent's Park, Lonclon, N.W., December, 1912. P. CHALMEllS MITCHELL, Secretary. MEETINGS OF THE ZOOLOGICAL SOCIETY OF LONDON POR SCIENTIFIC BUSINESS. 1913. TrESDAT, Febeuart 4 and 18 Makch 4 „ 18 April 8 „ 22 May 6 „ 20 June 3 October 28 November 11 „ 25 The Chair will he talen at half-jjast Eight o'clocJc in the Evening precisely. ZOOLOGICAL SOCIETY OF LONDON. LIST OF PUBLICATIONS. The scientific publications of the Zoological Society of London are of two kinds — " Proceedings,^^ published in an octavo form, and " Transactions/' in quarto. According to the present arrangements, the '' Proceedings" contain not only notices of all business transacted at the scien- tific meetings, but also all the papers read at such meetings and recommended to be published in the ''Proceedings" by the Committee of Publication. A large number of coloured plates and engravings are issued in the volumes of the " Proceedings," to illustrate the new or otherwise remark- able species of animals described therein. Amongst such illustrations, figures of the new or rare species acquired in a living state for the Societj^'s Gardens are often given. The " Proceedings " for each year are issued in four parts, paged consecutively, in the months of March, June, September, and December. From January 1901 they have been issued as two half-yearly volumes, indexed separately. " An " Abstract of the Proceedings "" is published by the Society on the Tuesday following the date of Meeting to which it refers. It is issued along with the " Proceedings," free of extra charge, to all Fellows who subscribe to the Publications, but it may be obtained on the day of publi- cation at the price of Sixpence, or, if desired, sent post free for the sum of Six Shillings per annum, payable in advance. The " Transactions " contain such of the communications made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are issued at irregular intervals. Fellows and Corresponding Members, upon payment of a Subscription of One Guinea before the day of the Anni- versary Meeting in each year, are entitled to receive the Society's Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent, less than the price charged for them to the Public. A further reduction of 25 per cent, is made upon purchases of Publications issued prior to 1881, if they exceed the value of five pounds. Fellows also have the privilege of subscribing to the Annual Volume of the Zoological Record for a sum of 305. (which includes cost of delivery), payable on the 1st. of July in each year ; but this privilege is forfeited unless the subscription be paid before the 1st. of December following. The following is a complete list of the publications of the Society already issued. TRANSACTIONS* OF THE ZOOLOGICAL SOCIETY OF LONDON. Yol. Vol. 4to. 19 vols. and I ndex. Price to Fellows. Price to the Public. I., contaiiiiiis- 50 Plates. . . . (1833-35) ... . £3 13 6 .. .. £4 18 or 11.; }> ''1 » (1835-41) ... . 4 .. .. 5 6 6t III., V 63 ,, (1842-49) ... . 3 8 3 .. .. 4 11 0+ IV., „ " » (1851-62) ... . 6 2 .. .. 8 2 6t v.. ,. ^'7 „ (1862-66) ... . 5 4 3 ., .. 6 19 VT., Q'2 (1866-69) ... . n 5 .. .. 15 vir., j> '^3 ,, (1869-72) ... . 10 4 .. . . 13 12 YIIT., 82 „ (1872-74) ... . 9 8 3 .. .. 12 11 IX., ^9 „ (1875-77) ... . 12 1 6 .. .. 16 2 X., 95 „ (1877-79) ... . 10 -0 3 .. ,. 13 7 X, Vols. I.-X (1833-79) ... . 7 6 .. .. 10 XL, containing 97 Plates. . (1880-85) ... . 9 12 .. .. 12 16 XII., 65 » • • (1886-90) ... . 5 8 .. .. 7 4 XIII., „ 62 )) (1891-95) ... . 6 8 3 ,. .. 8 11 XIV., » 47 » (1896-98) ... . 5 5 .. .. 7 XV., „ 52 >? • ■ (1898-1901) . . 5 15 6 .. . . 7 14 XVI., „ 38 j> (1901-1903) . . 5 8 .. .. 7 4 XVII., „ 41 (1903-1906) . . h 18 6 .. .. 7 18 XVIII „ 43 (1907-1911) . . 4 1 .. .. 5 8 XIX. „ 24 »> • • (1909-1910) . . 10 4 .. .. 13 12 XX.- -Part 1. rPls. I. -V.) . . (Feb. 1912) . . 18 .. .. 1 4 XX.- - „ . (Pls.VI.-XV. ) (April 1912) . . 2 5 .. .. 3 PROCEEDINGS OF THE COMMITTEE OF SCIENCE AND CORRESPONDENCE OF THE ZOOLOGICAL SOCIETY OF LONDON. 8vo. 2 vols. (Letterpress only). Price to Fellows. Price to the Public. Part I. „ II. 1830-31. 1832. 1 vol. Svo. 4s. 4s. Q(l. Qd. 6s. s.t PROCEEDINGS OF THE ZOOLOGICAL SOCIETY OF LONDON. Svo. 15 vols. (Letterpress only) and Index. (First Series.) Price to Fellows. Part I. 1833. 1 vol. 8vo. 4s. 6f/. II. 1834. III. 1835. IV. 1836. V. 1837. VI. 1838. VII. 1839. VIII. 1840. „ 4s. Qd. „ 4s. iSd. „ 4s. 6^/. ,, 4s. Qd. „ 4s. Gd. „ 4s. 6d. „ 4s. M. 8vo. 13 vols. Price to the Public. .. 6s.t .. 6s. .. 6s. .. 6s. .. 6s. . . 6s.t . . 6s.t and Index Part IX. 1841.: „ X. 1842. „ XI. 1843. „ XII. 1844. „ XIII. 1845. „ XIV. 1846. „ XV. 1847. Index 1830-1847. Price to Price to the Fellows. Public. 4s. Gd. .. 6s.t 4s. 6^. .. 6s. 4s. Gd. .. 6s.t 4s. Gd. .. 6s. 4s. Gd. .. 6s. 4s. Gd. .. 6s.t 4s. Gd. .. 6s.t 4s. Gd. .. 6s. (Second Series.) Letterpress only. Price to Price to the Fellows Part XVI. 1848. 1 vol. Svo. 4s. Gd. „ XVII. 1849. „ XVIII. 1850. „ „ XIX. 1851. „ „ XX. 1852. „ „ XXI. 1853. „ „ XXII. 1854. „ XXIII. 1855. „ „ XXIV. 1856. „ „ aX \ . Ibo/ . „ „ XXVI. 1858. „ „ XXVII. 1859. „ „ XXVIII. 1860. Index 1848-1860. Public. G.S. Gs. With Plates coloiirecl. Price to Price to the 4s. Gd. „ 4s. 6(;. . . 6s. „ 4s. 6(/. . . 6s. , „ 4s. Gd. . . 6s. „ 4s. 6a'. . . 6s. „ 4s. Gd. . . 6s. „ 4s, 6c/. . . 6s. „ 45. Gd. . . Gs. „ 4s. Gd. . . 6s. „ 4s. Gd. . . 6s. „ 4s. 6d. , . Gs. „ 4s. Gd. .. 6s. „ 4s. Gd. . . Gs. t Out of print. * In consequence of a re-arrangement of the stock of the ' Transactions,' the Society is now able to offer for sale, at the reduced price of .£30, sets of Vols, v.-xvi. inclusive, and separate papers, of which a list can be supplied, at about one-fourth their published price. Fellows. Public. £1 8 . £1 7 6t 1 8 . 1 7 6t 1 8 6 , 1 18 Of 15 9 . 1 1 Of 15 9 . 1 1 Of 18 . 1 4 Of 19 6 . 1 6 Of 1 8 6 . 1 18 Of 1 8 . 1 7 6t 1 8 „ 1 7 6t 1 11 6 . 2 2 Of 1 11 6 . 2 2 Ot 1 11 6 . 2 2 Of PROCEEDINGS OF THE SCIENTIFIC MEETINGS OF THE ZOOLOGICAL SOCIETY OF LONDON. 8vo. 40 vols, and 4 Indices. Letterpress only. With Plates uncoloured. With Plates coloured. Price to Price to the Price to Price to the Price to Price to the Fellows. Public. Fellows. Public. Fellows. Public. 1861 . . 4s. 6cZ 6s .. 9s. ... 12s ,. 33s. 9d. .. .. 45s.t 1862 . 4s. 6^/ 6s .. 9s. ... 12s .. 33s. 9d. ,. ,. 45s. 186.3 . 4s. 6(1 6s .. 9s. ... 12s .. 33s. 9d. .. ., 4.5S.+ 1864 . 4s. 6(1 6s.* .. 9s. ... 12s.t.... .. 33s. 9d. .. . . 45s.* 1865 . 4s. 6d 65.+ .... .. 9s. ... 12s .. 33s. 9d. .. . . 4.5s. 1866 . 4s. 6d 6s. 1 .. 9s. ... 12s.t.... .. 33s. 9d. .. . . 45s. 1867 .. 9s. ... 12s.*.... .. 3.3s. 9d. .. .. 45s. 1868 .. 9s. ... 12s .. 33s. 9d. .. .. 45s. 1869 .. 9s. ... 12s .. 33s. 9d. .. .. 45s. 1870 .. 9s. ... 12s . . 33s. 9d. .. .. 45s. Index, 1861-1870 .. 4s. 6d. . . . . 6s. 1871 .. 9s. ... 12s.*.... .. 33s. 9d. .. . . 45s. 1872 .. 9s. ... 12s.*.... ... 12s . . 33s. .. 335. 9d. .. 9d. .. . . 455.t .. 45s. 1873 .. 9s. 1874 .. 9s. ... 12s.t.... .. 36s. .. 48s.t 1875 .. 9s. ... 12s .. 36s. .. 48s. 1876 .. 9s. ... 12s .. 36s. .. 485.t 1877 .. 9s. ... 12s .. 36s. .. 48s. 1878 .. 9s. ... 12s .. 36s. .. 48s. 1879 .. 9s. ... 12s .. .365. .. 48s. 1880 .. 9s. ... 12s .. 36s. .. 48s. Index, 1871-1880 .. 4s. 6d. . . . . 6s. ... 12s .. 36s. 1881 .. 9s. ,, 48s. ]882 . . 9s. ... 12s ... 12s ,. 36s. .. 36s, 48s 1883 .. 9s. .. 48s, 1884 .. 9s. ... 12s .. 36s. ,. 48s, 1885 .. 9s. . . . -125 .. 365. .. 48s. 1886 .. 9s. ... 12s .. 36s. .. 48s. 1887 .. 9s. ... 12s .. 36s. .. 48s.t 1888 .. 9s. ... 12s ... 12s .. 36s. .. 36s. . 48s 1889 .. 9s. .. 48s. 1890 .. 9s. ... 12s .. 36s. .. 48s. Index, 1881-1890 .. 4s. 6d. . ... 6s. .. 36s. 1891 . . 48s. 1892 .. 36s. . . 48s. 1893 36s. 48s 1894 .. 36s. . . 48s. 1895 .. 36s, . . 48s. 1896 .. 36s. .. 48s. 1897 . . 36«. .. 48s 1898 . . 36s. . . 48s 1899 .. 36s. . . 48s. 1900 . . 36s. . . 48s, Index, .. 4s. 6d. . ... 6s. # t Out of p rint. PROCEEDINGS of the GENERAL MEETINGS for SCIENTIFIC BUSINESS OF THE ZOOLOGICAL SOCIETY OF LONDON. 8vo. 24 vols, and Index. 1901, vol. 1.. 18s. >) )) ^^ 1902, „ 1 . „ n 1003, „ I . „ n 1904, „ I ,, . II 1905, „ I „ „ II 190G, „ I „ ,, II 1907, „ I „ „ 11 1908, „ I „ „ II 1909, ., I „ '„ II 1910, „ I „ „ 11 ••••■ Index, 1901-1910 4s. 1911, „ I „ „ II 1912, „ 1 „ „ II Price to Price to the Fellows. Public. 18s. ... ... 24s. IBs. ... ... 24s. 18s. ... ... 24s. 18s. ... ... 24s. 18s, ... ... 24s. 185. ... ... 24s. 18s. ... . . . 24s. 18s. ... ... 24s; 18s. ... ... 24s. 18s. ..; ... 24s. 18s. ... ... 24s. 18s. ... ... 24s. 18s. ... ... 24s. 18s. ... ... 24s. ]8s. ... ... 24s. 18s. ... . .. 24s. 18s. ... ... 24s. 18s. ... ... 245. 18s. ... . .. 24s. 18s. ... ... 24s, 4s. 6d. . ... 6s. I8s. ... . .. 24s. 18s. ... . .. 24s. ]8s. ... . . . 24s. 18s. ... ... 24s. LISTS OF THE ANIMALS IN THE SOCIETY'S GARDENS. List of the Vertebrated Animals now or lately Living in the Gardens of the Zoological Society of London. (Eighth Edition.) 8vo. 1883. Cloth, 4s. 6cl. List of the Yertebrated Animals now or lately Living in the Gardens of the Zoological Society of London. (Ninth Edition.) 8vo. 1896. Cloth, 6s. Catalogue of the Library of the Zoological Society of London (Eifth Edition.) 8vo. 1902. Cloth, 6s. THE OFFICIAL ILLUSTRATED GARDEN GUIDE— lOthEdition — with (1) a Ttailway and Street Map, showing a direct ronte to the " Zoo " from all parts of London and the Suburbs ; (2) a Plan of the Grounds, showing at a glance the location of the animals ; (3) a short description of some of the principal animals in the Collection (now containing over 3000 spe- cimens), together with 50 Photographic Illustrations and Index. Price 6cl. in Stiff Paper Cover, postage l^d., or in Green Cloth Cover price Is. 2cL post free. P. CHALMERS MITCHELL, Secretary. Eegent's Pai-k, Lonrlon, N.W., December, 1912, * at Me. hooJcseller. These puhlications may be obtained at the Society s UpricE Uessrs. Longmans' {Paternoster Row, E.C.), or iltronr/h any ■eller. ZOOLOGICAL SOCIETY OF LONDON THE ZOOLOGICAL RECORD. THE object of the Zoological Eecoed is to give, by means of an -^ annual Volume, complete lists of the Works and Publications relating to Zoology in all its branches that have appeared during the year preceding the issue of the Yolume, together with full information as to the points they deal with, arranged in such a manner as to serve as an Index to the literature of Zoology in all parts of the globe, and thus to form a repertory that will retain its value for the Student in future years. The ' Zoological Kecord ' having been amalgamated with the International Catalogue of Scientific Literature, Zoology, Volumes from 43 onwards can now be obtained only from Messrs. Harrison & Sons, escept when purchasing complete sets from the Zoological Society. Under the scheme of amalgamation. Fellows of the Society, and Institutions already on the subscription-list, have the privilege of subscribing at the old rate of 30s. per annum, which covers the cost of carriage of the volume. The subscription becomes due on July 1st. in each year, and lapses if not paid by the 1st. of December following. The Society is able to supply complete sets of the Eecord on the following terms : — Vols. 1 to 37, price £14 10s. net. Vols. 38 to 42 at 10s. each net. Vol. 43 and onwards at 40s. each. The prices for separate volumes are as follows : — Vols. 1 to 42 (except Vols. 4 and 6) 10s. each net. The price of the 'Zoological Eecord,' Vol. 43 and subsequent volumes, obtainable separately only from Messrs. Harrison and Co., is 40s. each. Index Zoologicits. An alphabetical list of names of genera and subgenera proposed for use in Zoology, as recorded in the ' Zoological Eecord,' 1880-1900 ; together with other names not included in the ' Nomenclator Zoologicus ' of S. H. Scudder. Com- piled (for the Zoological Society of London) by Chaeles Owen Watekhottse and edited by David Shaep, Editor of the ' Zoological Record.' London, 1902. Price to Fellows, 18*.; price to the public, 20*., or if sold with a set of the ' Zoological Eecord,' 10s. Index Zoologicus, Iso. II. An alphabetical list of names of genera and subgenera proposed for use in Zoology, as recorded in the ' Zoological Record,' Vols, 38-47 inclusive (1901-1910), and the Zoology volumes of the ' International Catalogue of Scientific Literature,' Annual Issues 1-10. Compiled (for the Zoological Society of London) by Charles Owen Waxebhouse, I.S.O. and edited by David Shaep, M.A., F.R.S., Editor of the 'Zoological Record.' London, 1912. Price to Fellows, 12s. M. net; price to the public, 15s. net. Divisions of Vols. 39 to 42 of the ' Zoological Record ' can be supplied by the Society, but those of Vol. 43 onwards can be had only of Messrs. Harrison & Sons, 46 St. Martin's Lane, W.C. [p. T. o. SEPARATE DIVISIONS OF THE ZOOLOGICAL RECORD. Divisions of the 'Zoological Record,' Vols. 39-42, containing the literature of the years 1902-1905, may be obtained separately as follows : — List of abbreviations of journals, etc. Special Records, viz. : — I. General Subjects 2 II. Mammalia III. Aves IV. Reptilia and Batrachia 2 V. Pisces VI. Tunicata YII. Mollusca VIII. Brachiopoda IX. Bryozoa X. Crustacea XI. Arachnida XII. Myriopoda XIII. Insecta XIV. Echinoderma XY. Vermes XVI. Coelenterata XVII. Spongise XVIII. Protozoa Index of new names of genera and subgenera , Divisions from Vol. 43 onwards are now supplied by Messrs. Harrison & Sons, 46 St. Martin's Lane, London, W.C. P. CHALMERS MITCHELL, Secretary. Eegbnt's Park, London, N.W. December, 1912. s. d. 2 net 2 6 ,, 2 6 „ 6 „ 2 6 „ 2 6 „ 1 „ 4 „ 1 „ 1 „ 2 6 „ 2 „ 1 6 „ L2 „ 3 6 „ 3 „ 1 6 „ 2 „ 2 „ 2 „ ON burchell's zebra. 757 41. The Local Races of Burchell's Zebra. By Major J. Stevenson-Hamilton, C.M.Z.S., Game Warden of the Transvaal. [Received April 1, 1912 : Read May 21, 1912.] (Text-figures 102-106.) Index. Page Variation 757 Burchell's Zebra (Equits hurchelli) south of the Zambesi is lasually divided by naturalists into several local races — tvahlbergi, to-ansvaalensis., chwptnanni, selousi. These derive their subspecific distinctions from certain assumed diSerences in their leg and body striping as observed in museum specimens. Thus : — " In tvahlbergi the body stripes meet the ventral stripe inferiorly, while the legs are more or less fully striped ; the shadow stripes on the hind quarters are strongly developed and not much narrower than the main stripes, wliich are narrower than the intervening spaces, while the fetlocks and pasterns are devoid of stripes and spots. A female zebra from the Transvaal differing from the typical wahlbergi by the extension of the shadow stripes to the neck, has been named £J. b. transvacdensis. In E. b. chapmanni the shadow stripes have become faint and narrow, the legs are marked to the hoofs, but the stripes on the lower portions tend to break up into spots, and the inferior j)art of the pasterns is not wholly black. This race inhabits the country between Damaraland and Matabeleland. The last repre- sentative of the species with distinct shadow stripes is the Mashonaland bonte quagga (ii'. b. selousi), which differs from the last in that the sti-iping of the legs is complete right down to the hoofs, the pasterns being striped on both sides, and their lower portion, owing to the fusion of several stripes, wholly black. The sides of the tail are also striped." (Lydekker.) Long observation of the herds of Burchell's Zebra, running in the Transvaal Game Reserves, has convinced me that all or nearly all the above distinctions are found among them, frequently in the same herd. I have often discussed the matter with Dr. Gunning, of the Transvaal Museum, and have with him examined skins and live specimens obtained from the "Western and ]Srorthern Transvaal. There is no doubt that any of these might, from their markings, have come from the Eastern part of the Province. Among the herds found in the latter occurs a very great variety of striping. Some animals show heavy, wide, and deeply tinted shadow stripes, while others display only the slightest indications of them. Some are strongly ringed down to the fetlocks, while others have no signs at all of any markings Prog. Zool. Soc— 1912, No. LI. 51 758 MAJOR J. STEVEXSO>C-HAMILTON ON below the knees and hocks. In some cases the body stripes are continued right round the barrel to the ventral stripe, in others Text-fia-. 102. Bui-cLell's Zebra, $ , six j'ears old, from south of the Sabi River, Eastern Transvaal. they stop far short of it. The striping is usually carried right round the buttocks, but in I should say 20 per cent, at least of the BUBCHELL S ZEBRA. 759 animals in each herd this is not the case. In fact there appears to be no special distinctive marking common to any locality or Text-fio-. 103. BurcheH's Zebra, $ , ten years okl, from south of tlie Sabi River, Eastern Transvaal. 51* 760 MAJOR J. STEVENSON-HAMILTON ON even herd. The shadow stripes on the neck, said to he dis- tinctive of E. b. iransiiaalensis, have been occasionally, but very rarely, noticed in individuals, and I cannot conceive their presence to be more than an individual eccentricity, occasionally perhaps ti'ansmitted to offspring, but by no means constant. Text-liii'. 10-1. Burchell's Zebra, $ , eight years old, from south of the Sabi Eivev, Eastern Transvaal. During the year 1911 I was asked to obtain a couple of skins for the Transvaal Museum, and selected two stallions from the same herd at a point a few miles south of the Sabi River (roughly 25 degrees S. lat. by 31 E. long.). I chose them merely as being typical of the darker and lighter forms of striping respectively, neither of them being in any way exceptional. Some animals present displayed deeper striping than the specimen shown in text-fig. 102, while others showed more white than that in text- fig. 103. These figures are from photographs of the skins. Text- BURCUELLS ZEBRA. 761 "fig. 102 is that of a stallion six years old, and text-fig. 103 is of one about ten years of age. A little later I obtained the skins of a mare and her foal killed at a spot less than ten miles from Avhere I had shot the two stallions. The photographs of these are reproduced as text-figs. 104 and 105. Text-fig. 105. "JJuvclioll's Zebra, S > foaJj train sovitli of the Sabi River, Eastern Transvaal. In the first stallion (text-fig. 102) both the ordinary and the shadow stripings are exceptionally strong, and the former is con- tinued down and nearly all round the legs to the pasterns, which are quite black. The body stripes meet the ventral line thi'ough- -out. The dorsal stripe is wide and very strongly marked. The neck sti-ipes are exceptionally deep, and set very close together. In the second stallion (text-fig. lO.'i) the shadow striping is not nearly so well defined and the black stripes also are n.-irrower The body stripes do not extend all round the barrel, and there is wnictically no connection between them and the not very distinct 762 ON BUKCHELL S ZEBRA. ventral stripe. The dorsal stripe is a very narrow black line, and" there is a distinct gridiron pattern on the back. The legs are very slightly striped below the knees and hocks, and there are only a few black hairs on the pasterns. In the case of the mare (text-fig. 104), while there is a well- marked ventral and a wide dorsal stripe, and complete union between the former and the barrel stripes, there is an entire- absence of any black markings from all the pasterns, and from the fore legs below the knees. Text-fio-. 106. Burdiell's Zebra, $ ; the same animal as that sliowii in text-fig. 103, pliotogvaphcd just after death. In the foal again (text-fig. 105), while the legs all bear faint indications only of stripes, the pasterns are all qitite black, and there is some indication of a giidiron pattern on the ba,ck. In none of the specimens is there the smallest indication of any shadow striping on the neck. It will be noticed also that in the younger stallion and the mare (text-figs. 102 & 104) the- frontal stripes on the face run straight from forehead to nose, while in the older stallion (text-fig. 103) (and to some extent in the foal (text-fig. 105), though it does not show up well in the figure) the frontal stripes tend to form concentric rings. p. Z. S. 1912. PI, LXXXIV. >Vl f G. M. Woodward, del. Bale & Daniels son, L*"^ imp. DIPTEROPELTIS HIRUNDO. O:^ A NEW CRUSTACEAN GENUS. 763 It would \)e interesting to know whether this absolute want of uniformity found among Burchell's Zebras in the Transvaal is peculiar to this country, or whether it is the case throughout Africa south of the Zambesi. Obviously it would be very dangerous to attempt any classification from isolated museum specimens obtained from the Transvaal at any rate. Dr. Gunning proposes, I understand, to deal more fully in the forthcoming ' Annals ' of the Transvaal Museum Avith a matter which I here only lightly touch upon. 42. On Dlpteropeltis *, a Now Genus of the Crustacean Order Branchiura. By W. T. Calman, D.Sc, F.Z.S.f [Received April 13, 1912 : Eead May 21, 1912.] (Plate LXXXIY.i) Index. Paa Structure or Morphology 763 Ethology (Parasitism) 763 Geographical Zoologj- 763 Systematic ; DipteropeUis li irimdo 763 The specimens described in this paper were presented to the British Museum twenty years ago by Spencer Moore, Esq., by whom they were collected in Southern Brazil. They were found among a number of specimens of Dolojjs longicauda Heller bearing the label " Parasite on the gills and body of the fish known all up the river as ' Dorado ' from its golden colour. From Corumba and neighbourhood, Matto Grosso." DiFTEROPELTIS HIRUNDO §. (PI. LXXXIY. figS. 1, 2.) Caiman, Abstract P. Z. S. 1912, p. 34 (May 28th). -Descrijytion offenude : — The most striking feature of the species is the form of the carapace, the lateral lobes of which are drawn out into narrow lanceolate wings directed backwards and ex- tending far beyond the tips of the long abdominal lobes. The wings are fleshy in texture, traversed by numerous branching blood-channels, but without spines or other armature on the under surface. The head is small, defined on each side from the carapace wings * The complete account of the new form described in this paper appears here, but since the name and a preliminary diagnosis were published in the 'Abstract,' it is distinguished by being underlined. — Editoe. t Published by permission of the Trustees of the British Museum. X For explanation of the Plate see p. 766. § For definition of the genus see p. 766. 764 DR. W. T. CALMAN ON A by a shallow antero -lateral sinus from wliicli two converging grooves run in on the dorsal surface. These grooves unite into one on each side, but the resulting groove dies out before reaching the middle line. A pair of inter-ocular chitinous rods are well- marked, closely approximated in front, diverging posteriorly, and united by an indistinct articulation with a nearly parallel posterioi- pair. The paired eyes are very small and close together near the front margin. The unpaired eye could not be detected. The free thoracic region is not distinctly segmented ; it is of equal width throughout and more than three times as long as wide. The abdominal lobes are very long, narrowly lanceolate, and cleft nearly to the base. No trace of furcal rami can be found. The antennules and antennte (fig. 3) are very minute and nearly hidden from Aaew under the in-turned front margin of the head. The antennules, in particular, are very easily overlooked ; they are not divided into segments but have a blunt lobe at the base carrying a short apical tooth. The antennfe are about the same length, and consist of a stout basal part and a subgiobular terminal part separated from it by a constriction. The mouth-parts form a prominent cone (fig. 4). The two large and well-separated lobes of the upper lip are unarmed. The lower lip bears a pair of slender conical papillas ("maxillae" of Claus) which project slightly from the opening of the mouth. The mandibles are crescentic in form, with the concave edge very finely serrated and the convex edge bearing a few teeth near the tip. In front of the oral cone and continuous with it at the base is a papilla., directed forwaid, having at its tip the opening of a duct. This, no doubt, represents the sheath of the preoral "sting" of Argidus, but no trace of the spine or sting itself could be detected in the dissection of two specimens. At the base of this papilla and of the mouth-cone are groups of very large cells, presumably glandular, with deeply-staining nuclei. The suckers (first maxillse, second maxillae, or first maxillipeds of various authors) are placed close together in front of the mouth-cone, generally concealing altogether the preoral papilla. In place of the usual radial supports, the whole of the membranous border of the sucker is covered with discoidal scales (fig. 5). The maxillipeds (fig. 6) (maxillse of some authors) are very short, very stout at the base, and tapering rapidly to the tip. They are each composed of five segments, of which the second and third have numerous pectinate or branched spines on the ventral and anterior surfaces. The terminal segment bears two minute claws and a process which lies alongside of them. The successive pairs of legs (figs. 7-10) are set wide apart from each other along the sides of the thoracic region. The rami are in all cases shorter than the protopodite and carry only a few short set£e ; there is no flagellum on any of the legs. In the first NEW CRUSTACEAN GENUS. 765 pair the endopod may be a good deal sliortei* than the exopod, or, as in the specimen figured, nearly equal to it ; both rami are un- segmented. In the second pair also the i-ami are unsegmented, and the endopod is from about one-half to two-thirds as long as the exopod. In the third and fourth pairs the rami are subequal and the endopod is divided into two subequal segments. The basal lobe of the protopodite in the fourth pair is tongue-shaped, with a slight protuberance at the base of its distal edge. Dimensions of Rolotype in millimetres. Length of body to tip of abdominal lobes 20'0 Total length to tip of cai'apace wings 26"0 Breadth of head 2-5 Length of thoracic region 8*0 Breadth of thoracic region 2*5 Breadth of wing at base 1*8 Greatest breadth of wing, at about 8 mm. from base 4'8 Length of abdominal lobes 6 'S Greatest breadth of abdominal lobes 1 '3 Length of antennules and antennre, about 0"13 Length of oral cone 0"5 Greatest diameter of sucker 1*1 Length of second leg 1 "8 Diameter of eyes, about 0' 1 Distance apart of eyes, about 0'45 Locality. — Corumba, on the Paraguay River, Matto Grosso, Brazil. Four female specimens taken (in company with Dolops longicaucla) on the fish known as " Dorado " (probably a species of Salminus). Rolotype. — Female, No. 92.10.24.2 in British Museum Register of Crustacea. Affinities. — In having the so-called maxillse or first maxillipeds modified into suckers this species diS"ers from the genus Dolops, and in possessing anteniiules and a preoral papilla it differs from Chonopeltis. From all the species hitherto referred to Argulus it diflfers in (1) the remarkable form of the lateral wings of the carapace ; (2) the length of the abdominal lobes and the absence of furcal rami, in which characters it resembles some species of Dolops ; (3) the entire absence of conspicuous spines or hooks on the under side of the body and appendages, in which it resembles Chonopeltis ; (4) the vestigial condition of antennules and antennge ; (5) the absence of a spine or sting on the preoral papilla; (6) the absence of the usual radial supports on the disc of the suckers. Some of these characters, especially Nos. 1, 3, and 6, are possibly not of great systematic importance, but together they seem to show that the new form is less closely related to any of the species included in the genus Argulus than 766 ON A NEW CRUSTACEAN GENUS. these are to one another. I liave proposed, therefore, to establish for it a new genus, which may be defined as follows : — DiPTEROPELTIS. Caiman, Abstract P. Z. S. 1912, p. 34 (May 28th). Argulidaj with the so-called maxillse modified as suckers ; with the preoral papilla present, but without a spine ; with antennules and antennte veiy minute, imperfectly segmented ; without large spines or hooks on under side of carapace, body, or appendages : without furcal rami on the abdominal lobes ; with the lateral wings of the carapace greatly elongated. Genotype. — D. hirundo Caiman. It is to be noted that the single species of Chonopeltis, C. inermis Thiele *, is known only from a single specimen (from Lake Nyasa). It is not at all impossible, therefore, that the antennules, if they are as small as in Diptero2:>eltis, may have escaped observation, and it is even possible, though less probable, that an unarmed preoral papilla may be present. Should this prove to be the case it is doubtful whether, in spite of the great difference in the form of the carapace, the separation of Diptero- peltis from Chonopeltis covild be maintained. EXPLANATION OP PLATE LXXXIV. (The magnifications given are only approximate.) Fig. 1. Di^teropeltis Mrundo. Female (holotype), from the dorsal side, x 3. 2. Anterior part of body of same specimen from \entral side. The suckers have been drawn apart and slightly backwards. In the preserved specimens they meet in the middle line and entirely conceal the preoral papilla. X 6. 3. Antennule and antenna, x 450. 4. Tip of oral cone from antero-ventral aspect, x 100. 5. Part of membranous margin of one of the suckers, x 160. 6. Maxilliped. x 20. 7. Leg of iirst pair, x 20. 8. Leg of second pair, x 20. 9. Leg of third pair, x 20. 10. Leg of fourth pair, x 20. Figures 3-10 are drawn from one of the paratypes. * Thiele, Zool. Anz. xxiii. p. 47 (1900) ; Mitth. Zool. Mus. Berlin, ii. p. 44. figs. 110-116 (1904). ON NEW LARVAL TREMATODES. '67 43, On two new Larval Trematodes from the Striped Snake ( Trojndonotus ordinatus sirtalis). By WiLLlAM NiCOLL, D.Sc, F.Z.S., Lister Listitute o£ Preventive Medicine, London. [Received April 9, 1912 : Read May 21, 1912.] (Text-figure 107.) Index. Page Ethology : Two larval Trematodes in the mesenteric fat of Stri-ped8\\ake{Tropidonotusordinatt(,s sirtalis). Adult form probably in a bird ■ 767 Geographical Zoology : North America ; Striped Snake infected with two larval Trematodes 767 Systematic: Gercaria ordinata, sp. n., and Diplostomiim sirtale, sp. n., from cysts in themesentery of the Striped Snake 768 The occurrence of encysted larval parasites in snakes is evidence, if such were wanting, that some snakes are eaten by other animals. What is more important, the character of the parasites may indicate what variety of animal is in the habit of eating the snake in question. Conversely, the presence of any particuLar species of adult parasite in an animal is ahnost always a sure proof that such animal eats the snake in which the larval stage is found. It is unfortunately in many cases a matter of difficulty to diagnose the systematic characters of a larval parasite. In a number of cases, however, it is possible to assign it to a definite genus, rarely to a particular species. The two cases to be dealt with here present a certain amount of difficulty. The first larva is evidently a Distomate Trematode, but beyond that it is impossible to go ; the second larva is just as obviously a Holostomid, and almost certainly belongs to the genus Hemistomum. Such a diagnosis does not lead very far, but it at least enables one to say that in all probability the Striped Snak-e (Tropidonotus ordinatus, var. sirtalis) is eaten by some bird, for adult Holostomata are known to occur only in birds. This will possibly be confirmed by direct observation. _ Both forms were met with together in each of three Striped Snakes from N"orth America, which died in the Society's Gardens on the 5th and 20th December, 1910, and on 10th March, 1911. They occurred in enormous numbers in the mesenteric fat along the whole length of the intestine. Each was enclosed in a small spherical or ovoid cyst w4th unusually thin and soft walls. Unlike what is generally found in a Trematode ^ cyst, the wall gave the impression of being a thin membrane instead of the more usual tough chitinous investment. On this account, not only could the larvte be extracted from the cysts without dif- ficulty, but when placed in water they escaped readily of their own accord. When a piece of the cyst-infested mesentery was. 768 DR. WILLIAM NICOLL ON suspended in water, a continuous shower of larvpe was observed to fall to the bottom of the vessel. The first form, which I name Cercaria ordinata, sp. n. (text- fig, 107 a), was much more numerous than the other. It is a typical tailless encysted cercaria, about "5 mm. (•4-'55 mm.) in length and •2-*25 mm. in greatest breadth. In shape it is ovoid and flattened dorso-ventrally. The entire surface of the body is covered with minute regularly-arranged spines. The oral sucker is almost terminal and has a diameter of "07 mm. (•06--08 mm.). Text-fie-. 107. \ ^ \ A 3 A. Cercaria ordinata, sp. ii. Ventral view, X 150. B. Di]plostomum sirtale, sp. ii. Ventral view, X 150. The length is usually slightly greater than the breadth, and the sucker has a somewhat characteristic funnel-shaped appearance. The globular ventral sucker is situated rather in front of the centre of the body, and has a diameter of "083 mm. (•075-'097mm.). Its distance from the anterior end of the body is on an average "22 mm. (•17-'25 mm.). On the dorsal lip of the oral sucker are the two symmetrical apertures of the cystogenous ducts. The ■cystogenous glands are conspicuous structures, and consist of four NEW LARVAL TREMATODES. 769 large cells situated in a transverse row immediately in front of the ventral sucker. There are two pairs, a right and a left. The ducts from each pair unite almost at once, and the united ducts then pass foi'wai-d in an irregular course ; but just before they reach the oi-al sucker each makes a characteristic twist, following which there is a gradual increase in calibre until near the termi- nation, when they contract slightly again. The cystogenous glands have an irregular rectangular outline, and measure •048 X "037 mm. In direct contact with the oral sucker is the small muscular pharynx measuring -024 mm., which is continued by a short, somewhat dilated oesophagus of the same length as the pharynx. The intestinal bifurcation is about midway between the two suckers. The diverticula are simple, somewhat wide tubes, which terminate not far behind the ventral sucker (i. e. a third of the distance from the sucker to the posterior end of the body). The excretory vesicle is Y-shaped. A common trunk is practically absent, and the limbs extend forward to near the terminations of the intestinal diverticula. The excretory tubules are very fine ; the main tube on each side extends forward to near the oral sucker, where it turns back. The excretory aj^erture appeai^s to be slightly dorsal. No trace of other organs could be made out, so that no accurate idea can be obtained of the systematic position of this larva. The peculiar configuration of the excretory vesicle may, in con- junction with the shape of the alimentary canal, eventually lead to the identification of its adult form, but at present, so far as I am aware, there is no known adult Distome to which this larva can be ascribed. The second form, which I name Diplostomum sirtale, sp. n. (text-fig. 107 b), differs markedly from the first. It is about the same size, and occurs in somewhat similar cysts, but its shape and colour are entirely different. The body of Cercaria ordinata is light and transparent, whereas that of Diplostomum sirtale is dark and almost opaque. The opacity is due to the presence of innumerable small granules distributed throughout the whole body. The shape is that of a typical Biplostoimcm larva, being scoop-like with a short handle. The shape is due to the rollino- over of the postero-lateral margins of the body. In life, however these margins are capable of more or less eversion, so that on occasion the body may appear almost flat. The dimensions of this larva are •48--55 x •28-'32 mm., the short stumpy tail beino- •06 mm. long. The oral sucker measures •OS 9 mm. in diameter ; the ventral sucker •042-^045 mm. The latter is situated a little in front of the middle of the body, •25--28 mm. from the anterior end. Midway between it and the posterior end of the body occurs the characteristic Holostomid fixing disc, which appears as a transparent disc about the same size as the ventral sucker. Of the internal organs only the alimentary canal and excretory vesicle were visible. The former comprises a pharynx contiguous 770 SIR CHARLES ELIOT ON A RARE XUDIBRAXCH. with tlie oral sucker and measuring -024 x "020 mm. This is followed by an oesophagus about twice as long as the pharynx. The intestinal bifurcation takes place rather nearer the ventral sucker than the oral sucker, and the simple diverticula extend a little beyond the posterior border of the fixing disc. The excretory vesicle consists of a wedge-shaped sac, which extends forward as far as the ends of the intestinal diverticula. This is in all probability the larval stage of some species of Hemistomum, parasitic in a bird. •14. A Note on tlie rare Britisli Nudibrancli Hancocha eudactylota Gosse. By Sir Charles Eliot, Iv.O.M.G., C.B., F.Z.S. [Received April 18, 1912 : Read May 21, 1912.J (Plate LXXXV.) See Gosse, On Hancoclia eudactylota, Ann. Mag. Xat. Hist. ser. 4, XX. 1877, pp. 316-318 ; Gamble, On two rare British Nudi- branchs, Lomanotus genei and Hancockia eudactylota, ib. ser. 6, ix. 1892, pp. 378-385; Trinchese, Ricerche anatom. sul genere Oovia ( = Hancockia), Mem. della E. Accad. delle Sci. dell' Istituto di Bologna, ser. 5, vii. pp. 183-191, 1886; Eliot, Sup- plement to Alder and Hancock's Monograph of the British Nudibranchiate Mollusca, Ray Society, 1910, pp. 17, 72, 118-120, 163. No coloured figure of this rare British ISTudibranch has yet been published so far as I am aware. I endeavoured to include one in my Supplement to Alder &, Hancock's Monograph, but no specimen of the animal could be obtained before the time fixed for publication. Shortly afterwards a single individual was captured at Plymouth and drawn by a local artist under tlie supervision of the natui'alists w4io were working in the Laboratory of the Marine Biological Association. I have not seen the animal alive, but these drawings agree with what is known of its struc- tui-e, and I have no doubt that they faithfully reproduce its appeai'ance. For an account of the genus and species see the references to my Supplement given above. EXPLANATION OF PLATE LXXXV. JSancocTcia etidactt/lota. Fig. 1. Dorsal view of animal. 2. Veutral view of animal. 3. Side view of liead. 4. A rliinopliore. 5.) 6. > Difterent views of lateral processes. 7.) P.Z.S. 1912. PI. LXXXV. ' ^^^ J"'^^^ 1;^ ^^ ^'-^ "^ o::^o London Stereoscopic Co. imp. HANCOCKIA EUDACTYLOTA. P.Z.S. 1912. PL LXXXVI. London Stereoscopic Co, imp. GIRAFFA CAAELOPARDALIS THORN ICROFTl, ox THE NORTH RHODESIAN GIRAFFE. 771 45. The North Rhodesiau Giraffe. By R. Lydekker, F.R.S., F.Z.S.* [Received April 25, 1912 : Read June 4, 1912.] (Plate LXXXYI. f) Index. Page Rliodesia, Giraffe of 771 Giraffa camelopardalis thornicrofti, Description of ... 771 On the 20th of October, 1910, the British Museum (Nat. Hist.) received from Mr. H. Thornicroft, Native Commissioner, Petauke, North-east Rhodesia, the skin, skull, and limb-hones of an adult male Giraffe shot by himself in tiiat district. Mr. Thornicroft had previously called on me in London, and. expressed his willing- ness to shoot and present to the Museum a Giraffe from the single herd in this part of Rhodesia, if the necessary permit could be obtained. This was in due course procured, and was followed, after an interval, by the arrival of the above-mentioned skin and bones. The skin was soon afterwads set up by Rowland Ward Ltd., and the mounted sjDecimen placed in the big case at the head of the staircase leading to the East Corridor of the Museum, along- side the male and female of the East African Girajfa camelo- jjcirdalls rotkschildi. As it is mounted with the neck bent, it is difficult to ascertain the exact height, but I estimate this at close on 18 feet, or possibly rather more. When the specimen was installed in its case, it became essential that it should receive a distinctive name ; and I accordingly communicated the foUowing preliminary note to ' Nature ' % :— " This Giraffe is characterised by the low and conical frontal horn, the grey colour and scattered spotting of the sides of the face, the chestnut-brown forehead, deepening into black on the tips of the horns, the absence of a distinctly stellate pattern in the neck and body spots, Avhich are light brown on a yellowish fawn ground, and the uniformly tawny colour of the lower portion of the limbs. This Gii-affe, which I propose to call Qiraffa caraelopardalis thornicrofti, appears to be related to the Kiliman- jaro G. c. tijjpelskirchi, but differs by the more compact frontal horn, the brown, in place of grey, forehead, and the uniformly fawn lower part of the legs, the latter being whitish in adult bulls [of tlp2}elskirchi\, but fawn and spotted in cows and youno- bulls." The last statement rests on the authority of Messrs. M. de * Published bj' permission of the Trustees of the British iluseuin. t For explanation of the Plate see p. 773. X Vol. lx.Kxvii. p. 484 (1911). 772 ox THE NORTH RHODESIAX GIRAFFE. Rothschild and li. Neuville*, who state that in the East African Giraffe which they describe as rothschildi, but which — despite tlie locality whence it is stated to come — is certainly tippelskirchi, these age and sex differences are observable. I have, however,^ doubts whether they hold good in all cases ; and it is still possible, in spite of Avhat I have previously written, that there may be one form (schiUingsi) in which the shanks of adult bulls are white and another [tippelslcirchi) in which they are fawn and spotted f, and further, that these two types may intergrade. Tliat the nearest relative of the Rhodesian Giraffe is G. c. ti])- pelskirchi, may be considei'ed certain. Of the latter I have had for comparison the mounted head and neck of an adult male, a mounted immature female, and the mounted liead and neck of a calf, as well as a coloured plate in Messrs. de Rothschild and Neuville's memoir J. Elaborating to a certain extent the foregoing brief diagnosis, attention may be directed to the fact that tijypelsldrchi and thornicrofti agree (and thereby differ mai/kedly from rothschildi) in having the triangular space between the eye and the nostril devoid of spots. In the adult male of tippelskirchi, however, the ground-colour of the whole head is dirty greyish white, whereas in thornicrofti the forehead is chestnut or umber-brown, deepening into black at the tips of the horns, which are grey in the Kilimanjaro race. In the Rhodesian Giraffe the spots on the region behind the eye and the side of the lower jaw ai'e very faintly mai'ked, and blackish grey in colour ; wliereas in the Kilimanjaro bull they are larger, more numei'ous, and chocolate-brown in colour, being deeper in tint than the neck-spots (this feature being also shown in the immature female and the calf). In thornicrofti the spots on the neck are burnt-umber in colovir and markedly elongated in form, with their terminal ends jagged. Thei'e ai-e about eight of them in the longitudinal row which starts immediately in advance of the point of the shoulder. In tippelskirchi they are more numerous (ten or eleven in what appears to be the corresponding row), less elongated, and much more irregular in shape. Compai'ed with the young cow tippelskirchi, the spots on the body of thornicrofti are less numerous, more especially on the hind-fpiarters, while many of them are more deeply incised on one side, although they are less jagged in general contour. The spotting on the inner side of the thighs and of the upper part of the foi'e-legs is also much less pronounced. In the original de.scription (which was drawn up when the specimen was in the basement of the Museum) it is stated that the shanks of the legs are uniformly fawn, but, as a matter of fact, they are * Anil. Sci. Nat., Zool. ser. 9, vol. xiii. pp. 124, 129 (1911). t See Proc. Zool. Soc. 1904, vol. i. p. 219. J Op. cit. 1)1. ii. tig. 1, lettered O. c. rothschildi. ON ANTLER-GROWTH IN THE CERVID^.. 773 rufous-fawn with very faint traces of spotting nearly clown to the fetlocks ; while from the latter to the hoofs they are dirty greyish white. The foregoing evidence clearly establishes the right of the North Rhodesian Giraffe to rank as a distinct local race ; and if it be true that the one herd is completely isolated, there is pi'obably no intergradation with the Kilimanjaro race. EXPLANATION OF PLATE LXXXVI. Adult bull of Giraffa camelo^ardalis tliorn'icrofti. 46. On Antler-Growth in the Cervid^e, with special reference to Elaphurus and Odocoileus (Dorcelaphxs). By R. L PococK, F.H.S., F.L.S., F.Z.S., Superintendent of the Gardens and Curator of Mammals. [Received and Read June 4, 1.912.] (Text-figures 108-112.) Index. Page Classifications of the Cervidie 773 Mode of. growth of Antlers in tj-pical Old World Deer 775 Interpretation of the Antlers of Elaphurus davidianus 777 Interpretation of the Antlers of Odocoileus sp. incert 780 Date of Antler-change in American and Old World Deer ... 783 (Note) Introduction. Most, if not all, the attempts that have hitherto been made to understand the antlers of Deer and arrive at correct conclusions regarding the homology of the tines have been based upon com- parisons between the fully formed antlers of diiferent species. This, in my opinion, is the reason why there has been failure in some cases to detect homologies which study of the growth of individual antlers reveals. The importance of this question depends upon the circumstance that twenty years ago Mr. Gordon Cameron * proposed a classi- fication of the Cervidfe, based upon the antlers, as a substitute for the classification, founded upon the skeletal structure of the fore feet, which Sir Victor Brooke had suggested f. To make clear the purpose of the present paper, it is necessary to summarise briefly the rival classifications put forward by these two authors. Sir Victor Brooke divided the Cei'vid^e into two * 'The Field,' 1892, pp. 265, 703, 741, 860. t P. Z. S. 1878, pp. 883-928. Proc. Zool. Sog. — 1012, No. LII. 52 774 ME. R. I. POCOCK ON sections. The first, which he called Telemetacarpi (Telemeta- carpalia) because the distal ends of the lateral metacarpals persist, comprises the Roe {Capreolus), the Chinese Water-Deer \Hydropotes), the Reindeer (Bangifer), the Elk (Alee), and all the exclusively American deer with the single exception of the typical Wapiti [Cervtis canadensis) ; the second, called Plesiometacarpi (Plesiometacarpalia) because the proximal ends of the lateral metacarpals are usually present, whereas their distal ends are suppressed, comprises all the deer of the Old World, except the four genera mentioned above, but none of those of the New World apart from the Wapiti. Amongst the Old World forms the most important species for the moment figuring amongst the Plesiometacarpalia is Pere David's Chinese Deer (Elaphurus davidianus) . Mr. Cameron's classification was widely different. Dismissing as vmimportant the character relied upon by Brooke, he divided the Cervidfe into three sections : one for the Reindeer with antlers in both sexes, the second for the Elk with laterally extended antlers, the third for the remaining species with antlers i-estricted to the male and erect or suberect. This third section, which alone concerns us now, was subdivided into two categories of species, one comprising those in which the antlers consist, as in the typical Old Woi^ld deer and the Wapiti, of a " brow -tine " and a " beam," to use Gordon Cameron's terminology, and the other those in which the antler has, as he thinks, no brow-tine but consists of a " forked beam," as in all typical American deer (except the Wapiti) and in the Roe and Pere David's Deer amongst the Old- World species. Now with regard to the affinities of the species comj)osing Cameron's third division, there is only one point in which there is complete divergence between him and Brooke. This concerns Pere David's Deer, a species classified by Brooke with the Red Deer, Sambai', and other Elaphine stags, and by Cameron with the American forms allied to the Virginian and Mule Deer, the correct name of which seems to be Odocoileus *. So far as I can see, the only a priori objection to be raised against Mr. Cameron's system, if we accejit his premises, is that it is based upon a secondary sexual character. But although it cannot be justifiably consigned to oblivion on that account, it may be doubted if it Avould ever have come into sufficient prominence for serious discussion had it not been for the unqualified acceptance accorded it by Mr. Lydekker. However that may be, it is clear that if Mr. Cameron's assumption that there is a fundamental difference in structure between the antlers of the groups of deer mentioned above is Avrong, his classification, based on that claim, goes by the board. In the following pages I shall endeavour to show that his classification is untenable, because a study of the seasonal growth * Dorcelaplius and Cariacus are better known but sxiperseded terms. ANTLER-GROWTH IN THE CERVID^. 775 of an antler of Pere David's deer and of an American deer, allied to the Virginian, proves that the homologue of the brow-tine of the Elaphine stage is present in both — a conclusion which is by no means evident from an examination of the fully-formed antlers. Antler-Growth in tyjnccd Old-World Deer. In the Zoological Gardens I have repeatedly watched, year after year, the growth of the antlers of deer belonging to the Elaphine, Sikine, and Rusine types without finding any variation Text-fia. 108. D Earlj' growth-stages of Antlers of some Old- World Deer. A & B. Successive stages observed m Cervus hancjlu. C. Cervus canadensis. D. Stisa aristotelis. a, the anterior branch or " brow-tine " ; p, the posterior branch or " beam " ; h, the rudiment of the bez-tine arising from the posterior branch. of moment in the method of their development. The antler starts as an undivided bud. This bud then shows signs of division into two buds, an anterior and a posterior. These*buds 52* 776 MR. R. I. POCOCK ON gradually, and with nearly equal rapidity, increase in length, the anterior growing forwards and the posterior backwards. In the Sambar [Rusa, text-fig. 108, D) and some other species they show a marked inclination upwards ; so that at one stage the antler may be likened somewhat to a short-stalked Y, and at this or even at a later stage in deer like the Sambar {Rusa) and others which have no " bez "-tine, the antler may be indift'erently described as an " unbranched beam with a brow-tine" or as a " forked beam " or as a biramous antler. The anterior and posterior branches some- times, as in Cervus eldi, grow at approximately the same speed until the anterior has almost attained its limit ; but usually the growth of the posterior tine is from the first more rapid. However that may be, the equivalence of the two branches in the early stages is plain enough ; but afterwards this becomes less and less evident as the posterior branch continues to lengthen and develops its accessory tines. Text-fig. 109. Five stages (A to JE) in the growth of an antler of Hucervus diivaucelli. a, anterior branch or " brow-tine " ; p, posterior branch or " beam." These facts are shown in the annexed figure (text-fig. 109), representing five stages in the growth of an antler of a specimen of the Swamp Deer or Barasingha (Racervus duvaucelli). These were ANXLER-GROWTH IN THE CERVID^. 777 sketched on May 13, 16, 22, June 6 and 12. Similar stages may be observed in other typical deer of the Old World *. In the Elaphine stags, however, which normally grow a " bez "-tine, the biramous stage is early complicated by the appearance of the bud of this tine. Now this tine has been regarded as a dupli- cation of the brow-tine ; and in Max Weber's t diagram showing suggested homologies of the tines in certain deer the brow- and bez-tines are tinted alike, suggesting his adoption of this view. Nevertheless I believe it to be quite incorrect, for in all cases where I have watched its origin, the bud of the "bez "-tine arises, not from the brow-tine at all, but from the " beam." It is, in fact, the basal or proximal tine of the posterior branch of the antler. This is illustrated in text-fig. 108, A-C, showing the early stages of the growing antler of the Hangul [CWvus hanglu) and of the Wajjiti {^Cervus canadensis). Antler-Growth in Pere David's Deer (Elaphurus davidianus). There is no stag whose systematic position has troubled zoologists so much as Elaphurus. On the one side are those, like Dr. Gray, Mr. Cameron, and Mr. Lydekker, who, relying upon the structure of the antlers of the adult, placed the genus with the American deer. On the other side are those, like Sir Victor Brooke, Flower, Max Weber, and others, who, adopting the skeleton of the foot as a basis, classified it with the typical Old- World species. The antlers of this stag have often been figured and described, and a good idea of their form in the adult may be gathered from text-fig. 110, C, and text-fig. Ill, /. They typically consist of a comparativ'ely long basal portion from which two branches arise : one long, slender, simple or divided, projects backwards parallel, or nearly so, with the animal's back ; the other stout, erect, or curved slightly forwards, terminates in a pair of strong tines. At first sight, these antlers appear to have no trace of a brow- tine. This was evidently Sir Victor Brooke's opinion, and it was adopted by Mr. Cameron and Mr. Lydekker, who, on the strength of this belief, boldly claimed that this stag belonged to the same group as the American deer, also held to have no brow-tine, despite the resemblances in other respects pointed out by Brooke between Elaphurus and the typical Cervidse of the Old World. Prof. Garrod was more cautious, and frankly gave vip the attempt to interpret the antlers of ElaphuTus when he remarked that they " are at present quite beyond my comprehension." This, then, was the state of the case when my researches on the specialised cutaneous glands of Ruminants % showed that the * Mr. J. G. Millais (' Mammals of Great Britain and Ireland,' iii. plate facing p. 140, 1906) lias published a series of figures of antler-growth in the Fallow Deer (TJama) illustrating precisely the same phenomenon. t Die Saug. p. 667, 1904. X P. Z. S. 1910, p. 840. 778 MR. R. I. POCOCK ON absence of interdigital glands on the feet and the smoothness of the integument between the hoofs in Elaphurus corroborated Sir Victor Brooke's views as to the relationship between this animal and such Old World deer as lliisa^ Rucervus, and Cervus, and weakened to a cori^esponding degree the claim for affinity between it and the Text-fis. 110. Diagram of the Antlers of four genera of CervidtB, to illustrate the homologies established in this paper. A. Cervus. B. Eusa. C. Elaplmms. D. Odocoiletis. a, anterior and^, posterior branch. In A and B the anterior branch is called the " brow-tine " and the posterior branch the " beam." JB is somewhat inter- mediate between A and C. They difier collectively from 1> in having the anterior branch well developed. In D it is small and concealed behind the highly developed posterior branch. Telemetacarpal species, in all the members of which examined by me, belonging to the genera Mazama, Odocoileus [Dorcelaphus), Capreolus, Rcmgifer, and Alee, the skin between the hoofs is ANTLER-GROWTH IN THE CERVIDiE. 779 thickly hairy and, in all but Alee, a large pouch-like interdigital ^land is present at least in the hind foot. Nine stages (A to J) in the growth of an Antler of JElaphurus davidianus, showing that the branches marked a and p correspond precisely in origin with the brow-tine and the beam of other genera of Old World Deer. Compare D and JB with fig. 108, B. (Prom sketches made at Woburn and kindly supplied by Lord Tavistock.) 780 MR. B. I. POCOCK ON Thinking, for these reasons, that there must be some flaw in the claim that the antlers of Elapliurus differ fundamentally from those of, say, Rusa, I suggested the following homologies : — In Rusa the antlers have a short base, a short undivided anterior branch or brow-tine, and a large divided posterior branch oi- beam ; in EJaphurus they have a longer base, a veiy large divided and more erect anterior branch or brow-tine, and a correspondingly reduced, comparatively slender, divided or undivided posterior branch or beam. This, however, was a mere guess, which I was unable to substantiate by any evidence of much value. Believing, how- ever, that the growth of the antlers in Elaphurus would finally settle the question one way or the other, I asked Lord Tavistock if he would kindly observe the process for me on one of the stags, at Woburn. This he was good enough to do, and sent me in addition the series of sketches reproduced in text-figure 111. These sketches show, in my opinion, that my guess was, as I expected, correct. In the first three stages the antler is little more than an excrescence dividing into an anterior and a posterior bud. In the fourth stage the base is beginning to lengthen, the anterior bud to grow upwards, and the posterior bud nearly straight backwards. This pi-ocess continues during the succeeding stages, the anterior branch gradually taking the lead in size and importance, and becoming divided distally into two tines. I can see no escape from the conclusion that the anterior and posterior buds of the very young antler in this stag are the homologues of the corresponding buds in the young antler of the Barasingha {Rucervus duvaucelli) shown in text-fig. 109. That being so, it is clear that the anterior branch of the antler of Elapliurus is homologous with the " brow-tine" and the posterior branch with the " beam " of the antler in the Red Deer, Sambar, Barasingha,, and other deer characteristic of the Old World. The difierences between them are mainly a matter of size and direction of growth ; that is to say, they are difierences of degree and not of kind *. Antler-Groioih in a Sjyecies q/" Odocoileus. Writing of the antlers of the typical American Deer, Mr. Lydekker said f: — " A large amount of misconception has arisen with regard to the structure of the antlers of this group. In 1872 the late Dr. Gray rightly termed the single upright prong- arising from the inner side of the lower part of the beam of the antlers of the Virginian Deer the ' subhasal snag ' ; but this snag * The siibdivision of the anterior branch of the antler in ISlapliurus is, of coiirse, no argument against it being the homologue of the "brow-tine," for the latter not infrequently, though abnormallj', produces an additional snag in Elaphine and allied groups of deer. In some species indeed, as in the Irish Elk and Cervus eJdi, it is commonly and normally provided with supplementary processes. t ' Deer of All Lands,' p. 246. ANTLER-GROWTH IN THE CERVID^. 781 Sir Victor Brooke incorrectly identified with the brow-tine of the typical Old AVorld deer. This error has been pointed out by Mr. A. Gordon Cameron, [who stated that] these characteristic tines have nothing in common with the true brows of Old World types, and rise vertically from the inner side of the beam between the coronet and the main furcation, usually converging at the apex. They are subject, in common with the antlers that produce them, to ail kinds of eccentricities ; are frequently forked or sub- palmate." Mr. Lydekker writes as if Mr. Cameron's dictum settled the question at issue ; but it does not appear to me that much weight, can be attached to the reasons adduced by the latter for his dogmatic denial of the truth of Sir Victor Brooke's interpretation of what Gray called the " subbasal snag " in the Virginian deer. Except that the tine in question is situated on the inner side of the antler, there is no great difference between it and the brow- tine of the Old World stags, which is highly variable in direction, as a comj)arison between the antlers of, e. g., Cervus affinis and Rusa aristotelis will shoT\^. Not less does it vary in size and structure even in nearly allied forms, as is testified by Dama dama, where it is large, by Dama mesopotami-ca, where it is sometimes almost suppressed, and by the Irish Elk, believed to be a Damine stag, where it may be palmated and branched. The question to be settled, then, is this :— Does the position of this tine on the inner side of the antler in the Virginian deer preclude its being the homologue of the brow-tine situated on the front of the antler in the Old World deer ? Study of the growth of the antler justifies, in my opinion, a negative answer to this question and shows that Sir Victor Brooke's opinion was correct. Early last year the Society received from the northern part of South America a male specimen of Odocoileits, which I cannot determine accurately. It is smaller and browner than a Venezuela specimen identified as 0. savannarvAn, but is otherwise very like it. Its antlers are short, with the beam curved forwards in the upper portion and ending in two tines, an anterior and a posterior ; while on the inner side, near the base, arises the so-called " sub- basal snag." The growth of these antlers was very instructive. They started as a simple excrescence, which soon began to divide into an anterior and a posterior bud, the only difference l^etween the antlers at this stage and those of a typical Old World deer being that the anterior bud was slightly internal and projected a little inwards as well as forwards. Nevertheless the two buds were perfectly visible in profile view. The appearance of the antler at this stage is shown in text-fig. 112,^, taken on May 12th. Four more stages of the growth are represented in the following figures, B-E, taken respectively on May 22nd, May 30th, June 6th, and June 17th, which show very markedly the gradual assumption of an apparently more internal position by the anterior branch, its 782 MR. R. I. POCOCK ON Text-%. 112. Five stages (-1 to E) in the growth of an Antler of an Anieriean Deer {Odocoilens sp. iucert.), showing that the " subbasal snag'"' («) and the "forked beam" (p) were respectively the homologues of the "brow-tine " and the "beam " of the typical Old World Deer. Compare B with lig. 108, D. ANTLER-GROWTH IN THE CERVID^. 783 point of attachment to the posterior branch being completely •concealed from the external aspect in the last three stages *. In view of these facts, I do not think it can be doubted that the anterior Inid which develops into the " subbasal snag " in Odocoileus is the homologue of the anterior bud which forms the brow-tine in Cervics. In that case the " subbasal snag " and the "brow-tine" are homologous structures passing under different names, and to state that Odocoileus has no brow-tine is merely playing with terminology. If this interpretation of the structure of the antlers 'wiElaphurus and in the species of Odocoileus above referred to be, as I believe, ■correct, it shows that these two genera are widely divergent in the very point upon which relationship between them has been claimed to exist, and that the likeness, such as it is, between the antlers of Elaphitrus and of the Mule Deer {0. hemionus), for instance, which has the so-called forked antlers without a brow-tine or with the mei^est vestige of it, is purely a question of parallelism in development ; that is to say, it has been brought about by growth and modification of fundamentally different parts of the antler. In the Mule Deer the anterior branch or brow-tine is to all intents and purposes suppressed, practically the whole antler being composed of the posterior branch or '• beam," which is highly developed and heavily tined. In Mcq^hicrus, on the contrary, the principal part of the antler is composed of the finterior branch or " brow-tine," which attains a large size and is divided into two prongs, while the posterior branch or beam remains comparatively small and slender and projects straight backwards as a long often undivided prong. * III connection with the date of antler-change in this Stag, attention may be acijtca, sp. n 788 Classitication of Sjyirorbis 792 Frofolteospira, subgen. ii 798 Spirorbis anihilateralis, sp. ii 796 S. raceniosus, sp. n 799 S. wedins, si^. \\ 800 General Characteristics of the Famibj Sex'pulidfe. 1. Tube calcareous, neai-ly always attached to rocks or other substratum for some part of its length. 2. Generally one or more branchife on dorsal side terminated by an operculum. 3. Thorax, generally provided with a thoracic membrane, representing the fused cirri and having 3-9 (iisually 7) segments. 4. Gland shields in thorax only. Genus Serpula Linne (13) 1767. Philippi (21) 1844. Generic characteristics f : — 1. Collar setiv of bnyonet-shape, with spines at base of blade. 2. Operculum funnel-shaped, with numerous radii ending in serr.'itions on maigin. 3. X^ncini with only a few lai'ge teeth. 1. Sektula columbiaxa Johnson (9), 1901. (PI. LXXXYII. fig. 1.) Serpula splendens Bush (3), 1905. Serpula colwmhiana INIoore (19), 1909. Specific characteristics : — 1. Anterior abdominal setfe with flaring fringed ends, short and deeply embedded, posteriorly replaced by small fascicles of very long stiff spines. * For explanation of the Plates see p. 805. t An attempt is here made to summarise bvietly the generic and specific characteristics in every case. Such a procedure has not previously been adopted, so far as I know, and it will, no doubt, in some cases be necessary at some future time to modify such characteristics, but in the present confused state of our systematic knowledge of the Serpulids, this seems to be a course likely to eliminate some of the ditficulties. p. Z.S. 1912.PI.LXXXVII, scopio Lo. :rnp. NORTH AMERICAN SERPULIDtE . p. Z.S. 1912.PI.LXXXVIII London. 3tepeoscopic Co. trap NORTH AMERICAN SERPULIDtE . p. Z.S. 1912.PI.LXXXIX. London Stepeoscopio Co, imp. NORTH AMERICAN SERPULIDiE . POLYCH.ETA FROM NORTH AMERICA. 785 2. Uncini generally 6 or 7 teeth, the anterior one being the largest (figs. 202 & 203 Johnson, 9). 3. Very large size (fig. 1) with numerous abdominal segments, 30 to 54 branchiae and from 80 to 160 serrations on edge of operculum. Numerous specimens from Departure Bay, Dodds Narrows, and two from Puget Sound (Prof. Kincaid's collection). Isolated tubes are to be found attached to the undersides of stones on rocky shores near extreme low-water mark, and small colonies may be found above this level in rock-pools. In other places attached to rocks near low-water level the masses of large white calcareous tubes are very striking — they are thick and often finely ridged, the lower parts being much intertwined, the free distal ends often overgrown with Polyzoa, iSpirorbis, etc. The brilliant red colour on the branchial crown may involve the whole of the gills and opei-culum, or these may be colourless except for the tips, or they may be barred and mottled in a large variety of ways. The remainder of the body is generally yellowish. As regards collar, thoracic membrane, and operculum with its tubercles, this species agrees very closely with &'. vermicularis, so fully described by St.-Joseph (24. pp. 328-335). Johnson (9) presumably has made it into a new species on account of its very much greater size with corresponding greater number of abdominal segments, branchife and serrations on operculum, together with small difierences in the setse and uncini. He cannot, I think, have examined many specimens, for he states that the functional operculum is on the right side (9. p. 432). The position of this in the genus Serpula may be right or left as shown by Zeleny (27. p. 34), but out of 50 specimens that I have examined 28 had it on the left, 21 on the right, and the remaining specimen had one on each side. Consequently, when he says that there are about 100 serrations on the edge of the operculum, I cannot think that Bush (3) is justified in recording this as a distinction between this species and S. splendens with 127 to 150 serrations. Moore (19) gives 140 for the one specimen of S. columbicma in which he counted the serrations. In my specimens they vary from about 80 to 160 — the number apparently increasing with Another distinction given by Bush is that there are in *S'. columhiana " but . 250 abdominal segments in a length of 55 mm.," whereas Johnson says " 250 or more" (9. p. 432), and she gives 313 as the number in a specimen of her so-called jS'. splendens, of which she does not state the size, and 190 in a specimen 35 mm. long. These figures speak for themselves I think as creating nothing but confusion. In 15 of my specimens the average number of abdominal segments was 236, including 79 in a specimen 10 mm. long, 142 in a specimen 41 mm. long, and 310 in one 81 mm, long. 786 MISS HELEN PIXELL ON The third point of difference given by Bush is that S. Colum- biana has more numerous branchiae — 54 in each lobe as given by Johnson — than her *S'. sjjlendens, for which she records 45 to 50 pairs. In the specimen of the former species counted by Moore (19) there were only 38. One small specimen of mine had only 18, but the general number was from 30 to 50 on each side, one or two large specimens having as many as 54 — the number given by Johnson. These facts show clearly I think that «S'. splendens is a quite unnecessary species and can be included in S. colimihiana^ which in turn, as pointed out by Johnson (9. p. 433), may be identical with S'. jukesii Baird, for which however no satisfactory desci"iption was given. Numerous large free Selenidia in the trophozoite stage wei-e found in the alimentary canal of nearly every specimen of 6'. Columbiana examined. Genus CiiuciGERA Benedict (1), 1886. Generic characteristics : — 1. Collar setfe and uncini similar to those of Se^yula. 2. Operculum with comparatively few radii forming a scalloped margin to the funnel and with conspicuous basal pro- cesses. 2. Crucigera zygophora Johnson. (PI. LXXXVII. fig. 2.) Serpula zygophora Johnson (9), 1901. Cntcigera zygophora Bush (3), 1905. Specific characteristics : — 1, Branchiae about 30 pairs with long filamentous ends to rachises. 2. Operctilum thick, shallow, with about 30 radii and 3 rounded processes at its base ; attached by a long pedicle. One specimen from Puget Sound was 45 mm. long. A smaller incomplete specimen came from Victoria (fig. 2). Another specimen was only 7 mm. long and had a much thinner operculum, but seemed otherwise similar. 3. Crucigera irregularis Bush (3), 1905. (PL LXXXVII. fig. 3.) Specific characteristics : — 1. Branchiae much coiled and with comparatively short fila- mentous ends to rachises. 2. Operculum irregular, apex of funnel displaced ventrally and the posterior and lateral walls deeper and rolled over to some extent. Not more than two basal processes which may, however, be bi-lobed and attached by long stout pedicle (fig. 3). About 12 specimens from the Channel outside Departure Bay POLTCH^TA FROM NORTH AMERICA. 787 and one from Dodcls JSTarrows. Depth 15 to 25 fathoms. Tubes generally solitary, attached to stones, shells, old wood, etc. Only one specimen is recorded by Bush from Juneau, Tubes much coiled, with flaring ends and one or two other conspicuous ridges at intervals indicating the flaring end of a younger tube. Young tubes seem to develop with a centimetre or so attached more or less straight along the substratum, then to coil indiflferently to right or left, and only at a much later stage, if at all, to ascend and form the flaring end. General colour pale orange, the branchiae and operculum variously mottled and barred with red. Pinnfe sometimes golden ; ova greenish. Length varies from about 14 mm. to 50 mm, (fig. 3). Largest diameter of opercidum 1*3 to 4"5 mm. ; the latter had 32 radii foiining a thick scalloped edge. The j)edicle was bifid and contx'acted at the top just before joining the basal processes of the operculum (fig. 3). Setfe as figured by Bush (3) ; in the posterior region of the abdomen the ordinary setae are replaced by small fascicles of long slender spines. Genus Apomatus Philippi (21), 1844. Generic characteristics : — 1. Operculum globular, terminating a gill retaining its pinnae. 2. Some thoracic setae bladed sickles (setae of Apomatus) (fig. 4 c). 3. Terminal dorsal gland present. 4. Apomatus timsii, sp, n, (PI, LXXXYII. figs. 4«-4/.) Specific characteristics : — 1. Collar setfe simple tapered blades (fig. 4 6). 2. Branchiae about 40 pairs with pinnae nearly to the ends of rachises. 3. Uncinigerous tori begin on third setigerous segment. 4. Uncini with numerous small teeth, the posterior one larger with terminal enlargement (fig. 4 /). 5. Abdominal setae more or less sickle-shaped with some long filiform ones in the last segments. Eleven specimens from the Channel outside Departure Bay in about 20 fathoms. This is the first time apparently that the genus has been recorded from the Pacific Coast of America. The tubes are solitary, adherent, and sinuous — one was attached along its whole length to a portion of the Hexactinellid sponge Aphroccdl istes lohiteavesicmus. When full-grown about 80 mm. with 150 abdominal segments (fig-4«). The branchial rachises are much coiled and almost colourless, with pairs of red spots up their outer surfaces and only short filiform extremities — the pinnae appear green due to the contained blood. The thorax is more or less orange or red, the thickened 788 MISS HELKX riXELL ON ridges bearing the tori being espi^oially deep in colour. Ab- domen pinkish ; there is a short anterior asetigerous region, and posteriorly nearly covering the last centimetre or so of dorsal surface is a chalky white strap-shaped raised glandular area. This is the terminal dorsiil gland mentioned by 8t. -Joseph (24) as possibly serving in the construction of the tube. It tapers anteriorly and ends abruptly posteriorly just above the vertical slit-like anus.* Along nearly the whole of this glandular region the ordinary abdominal set:e are replaced by fascicles of about li\e slender cajiilliary seta^ without blades and nearly equal in length to the width of the abdomen at this place. The collar is entire ventrally and has a deep incision on each side — the lateral lobes being continuous with the wide thoracic membrane. The functional operculum is large and transparent and carried by the second branchia fi-om the dorsal side. The pedicle is generally coiled twice like the other gills ; except in very young .specimens there is a small club-shaped one terminating the branchia on the opposite side. All the coUai" seta^ and most of the other thoracic set:« have ^simple narrow blades ; amongst them in the posterior thoracic fascicles ai-e to be found a few of the typical bladed sickles (fig. 4:c). The abdominal seta^ are more or less sickle-shaped, though they easily become folded (figs. 4 d, e) or straightened out. Several of the s})ecimens examined were much smaller and obviously young forms, with 70, 83. 101, 11 -i abdominal segments ;\nil fewer branchiae than in adult specimens. Genus Protula Risso (23), 1826. Cieneiic chai-acteristics :- — - 1 . No operculum. 2. Collar setiB simple tapered blades (fig. 5h). 3. Terminal dorsal gland present. 5. Protul.\ PACiFicA, sp. n. (PI. LXXXVII. figs. 5 a-5f.) Specific characteristics : — 1. Ventral lobe of collar notched. 2. Abdominal setje somewhat sickle-shaped in the anterior region, narroAV terminal bladed ones posteriorly, o. Uncinigerous tori extend from segment 3 to the end of the abdomen. 4. Uncini with numerous small teeth, the posterior one long with a bulbous extremity (fig. 5f). 5. Thoracic seta^ simple blades with sometimes a few seti>3 of Apomatiis posteriorly. * In section the gland is seen to consist of epitlielial cells crowded with spherical irranules which stain easily with iron luvmatoxylin but not with Delatield's. The iiland is apparently unchanged by the presence of acid in the preservatives and therefore does not appear to be calcareous. i I'OLYCIf/KTA FJIOM NORTJJ AMKUICA. 789 Three specimens from Fairway Channel outside j)eparture Bay- in about 30 fathoms ; one from Puget Sound (Prof. Kincaid's collection). No tube. The specimens were colourless except for the branclii-ie, which appear greenish due to the contained blood. Down the outer side of each branchial rachis was a line of opaque white spots. The branchial crown (10-15 mm. high) easily falls off, leaving a scar ; the remainder of the body is 38 to 60 mm. long, rather more than one-third of which is thorax. This is broad and flattened with nearly parallel sides (fig. 5 a) and the usual 7 setigerous segments. The thoi'acic membrane is very wide, with an entire margin which can extend beyond the setae but is generally considerably crumpled. The collar is notched in the median ventral line and has a deep fissure on each side. The inward coil of the stout base of the branchial ci'own is seen on the right side of the specimen in fig. 5 a. It takes Ig to 2 turns inwards and upwards, the short inner gills reaching about the same height as the outer long ones. An interbranchial membrane connects the lower third of the gills, of which there are about 60 pairs having small closely placed pinnae almost to their extremities. At the junction of the branchife with the stout basal membrane which carries them there is a slight ridge visible externally, and at this place on the inner side arises the oral membrane, which is continuous across the median line and up each of the spirals. On the dorsal side of the mouth is another shorter membranous lip. Uncini begin on the third setigerous segment. They have the characteristic shape (fig. 5/). The first two millimetres or so of the abdomen is achietous, nearly round, and of smaller diameter than the wide dorso-ventrally depressed part which follows. The ventral surface has deep segmental grooves showing 83-110 segments, and a wide faecal groove which turns to the right on reaching the thorax. Laterally the short tori are raised on distinct parapodial processes which extend to the posteiior end of the body. The last 30 or so segments on tlie dorsal surface are covered with the calcareous-looking gland which anteriorly tapers off to two points, suggesting a paired origin. In section the gland is seen to occupy nearly the whole thickness of the dorsal body-wall, the longitudinal muscles being pushed towards the sides. Nearly all the cells are crowded witli the spherical granules or globules which stain easily with iron hfematoxylin, and are apparently similar to those in the glandular cells so frequently found in the ej^idermis. This hind region of the abdomen has very long setse which extend 2 mm. or more on each side but are easily broken. At first sight they appear to be simple spines very slightly bent at the extremities, but, with high magnification, a narrow striated wing may be made out (fig. 5 e), shorter setae found with these are more distinctly winged (fig. 5 d). The other abdominal segments have ventral fascicles containing about 13 short, stout, Puoc. Zoou 8of.— 1912, No. TJII. r,:i 790 MrSS HELEN I'lXELL OS somewliat sickle-sliaped setfe (fig. 5 c). These setre were, iin- fortunately, not examined before being preserved, and as St.-Joseph (24. p. 338) points out, a lengthened immersion in alcohol tends to reduce the curve of the sickle. In general structure this species resembles F. capensis Mcintosh (13) fairly closely, but differs in the shnpe of the seta? and nncini : in the latter respect and in some other points it differs, too, from P. diomedece Benedict (1). Thei^e are mnny character- istics distinguishing it from F. superba Moore (19) and other Pacific species that have been described. P. aiypha Bush (3) might possibly be a young specimen of this same species. Genus Chitinopoma Levinsen (10), 1883. Generic characteristics : — 1. No thoracic membrane. 2. Collar seta? with fin -like expansion at base of blade. 3. Some of the other thoracic seta^ are siclde-shaped. 4. Abdominal seta^ geniculate. 5. Uncini with 9 or 10 fine teeth, the anterior one being larger and blunter than the others. 6. Operculum with horny plate. 6. Chitinopoma greenlandica. (PI. LXXXVIII. figs. 6 a-6 e.) Serpula triqueter Fabricius (7), 1780. Hydroides norvegica var. grdnlandica Morch (20), 1863. Hydroides (?) gronlandica Malmgren (14), 1867. Chitinopoma fahricii Levinsen (10), 1883. Specific characteristics : — 1. Bodies elongated, somewhat cylindrical. 2. About 6 pairs of branchi;>? with ends fi-ee from pinnse. 3. Operculum enclosing central stalked vesicle. Numerous specimens from Departure Bay and neighbourhood, one incomplete one from Victoria. In thick sinuous tubes adherent to shells, stones, etc., and having a very conspicuous dorsal keel generally ending in a spine oveilianging the apertiu-e. One tube was U-shaped with the two ends close together. The largest specimen was about 12 mm. long (fig. 6 «). The whole animal was practically colourless, the pedicle of the operculum and the branchiae sometimes having faint ti-ansverse bands, and the contents of the alimentary canal were in some specimens dark red. The branchiae varied very much in number (from 6 to 8 in each lobe) and some were frequently found in a rudimentary state : one specimen had 7 in the right, and only 3 fully-developed functional ones in the left. The operculum was in every case on the left dorsal side, and I have found no trace of a secondary one on the other side. POLYCH^TA FROM NORTH AMERICA. 791 The full-grown branchiae had about 17 pairs of ciliated pinnse, which stopped short some distance from the top of the rachis leaving a fairly long filamentary end. In sections the rachises are seen to be strengthened externally by a thick layer of chitin, which also forms a protecting layer round the pedicle (fig. 6 d), and is continuous over the operculum with the thick horny plate on its top. This plate was often covered with sand, and many specimens had infusorians, etc. attached to it. The central coelomic space in the pedicle is lined with peritoneum provided with very large conspicuous nuclei. There is a small vessel running along its whole length and enlarging in the opercular cup into a spherical vesicle (fig. 6 e). This is filled with a finely granular precipitate, and from its wall and general appearance seems to correspond with the branchial blood-vessels of the ordinary gill rachises, though I do not see that it can have any respiratory function. It is apparently suspended in a fine reticulate connective-tissue which easily shrinks away from the epithelial cells ; Levinsen,in his original description (10. p. 203), suggested that it might be a new operculum forming in the old one. The collar is very wide — the entire ventral lobe being generally reflexed — the latero-dorsal lobes are continued down the dorsal side to between the second and third thoracic setfe, where they end abruptly, giving the aj^pearance of a short thoracic membrane. This was a constant characteristic in both small and large specimens. I have not seen it referred to before, but do not think it necessary on that account to separate this as another species. With the large collar setae (fig. 6 b) are a few shorter curved forms with very narrow blades. The abdominal setae agree with those described by St.-Joseph (24) for the genus (fig. 6 c). The abdomen is long and slightly dorso-ventrally flattened, with 25-40 segments. The tori contain about 17 uncini, but there are only a few setfe to each segment. The doi'sal longi- tudinal muscles are greatly developed — there is a small fascicle of ventral ones on each side of the wide faecal groove. The large ventral nerve-cords are separated from one another. The epitheliuin consists of low columnar cells with numerous gland- cells containing the usual spherical masses which stain easily with iron haematoxylin. This species seems to agree very closely with that described by Bush (3) as Hycdojioinatopsis occidentalis. I cannot under- stand why this form with no thoracic membrane and " trumpet- shaped " abdominal setae is put in a genus, of which St.-Joseph writes (24. p. 264) : " II m'a fallu creer un genre nouveau Hyalopomatopsis pour le Hyalopomatus langerhansi Ehl. et le H. marenzelleri Lang., qui par la presence d'une membrane thoracique ... ne pouvaient rentrer, comme I'avait dxi reste prevu Langerhans, dans le genre Hyalopomatus tel que I'avait r>:{* 792 MISS HELEN PIXELL OX defini von Marenzeller." St.-Joseph also gives as another characteristic of his genus Hycdopomatopsis, the presence of capillary setfe in all the abdominal segments. Genus Spirorbis Daudin (6), 1800. (Pis. LXXXVIII., LXXXIX. figs. 7-16.) Generic characteristics and Schemes of classification : — 1. Calcareous tubes coiled in a dextral or sinistral spiral. The method of coiling and the markings on the tube have been used by Bush (3) in drawing up Table I., but the coiling of the tubes is variable {cf. figs. 8 a and b), being- determined to a great extent by the natm-e of the sub- stratum, and as Caullery & Mesnil (6) have already pointed out there is no constancy in either coiling or markings. 2. Branchiae are constant in difi'erent species, the operculum with terminal calcareous plate, alwaj's occurring as the second on the concave side {i. e. the right in dextral forms and left in sinistral ; since the animal lies with its back towards the substi-atum). 3. Thoracic segments generally 3 — the first having only dorsal setae — the two following have on each side an uncinigerous torus as well. In the subgenera given the prefix Fara- by Caullery & Mesnil an extra torus (and in S})- can- cellat'us a fascicle of dorsal setse also) is developed on the concave side of the animal ; this condition has been described in the following Table of Classification (p. 794) as 3| setigerous segments. In S'})- ambilateralis, sp. n., there are four complete setigerous segments, although the fourth on the convex side is very reduced. This specimen therefore approaches the hypothetical Prosjnrorhis, described by Caullery & Mesnil (5. p. 233), who point out that the genus Spirorbis has been evolved from other Serpulidoe, which have the characteristic greater number of thoiacic segments, by a gradual i-eduction. I have therefore placed this species in a new sub-genus Froiolceospira. 4. Abdominal segments 8-40. Between the thorax and the abdomen is a more or less long asetigerous region — often crowded with ova. The spermatozoa develop in the postei'ior setigerous segments. 5. The thoracic sette are distinctive and the differences ai'e of use as specific characteristics. Table II. given by Bush (3. p. 261) is drawn up with regard to these and the direction of the coiling alone. As a rule, the first thoracic segment hns some slender capillaiv (seta^ forming the inferior part of the fascicle (fig. 14 b) — the POLVCH.t;TA FKOM ]SORTH AMERICA. 793 superior ones, referred to as collar seke, are distinctive, tliey may have simple blades (fig. 13) or there may be a distinct fin-like expansion at the base of the blade (figs. 7, 10 a, 11 a). In the case of sinistral forms Caullery & Mesnil referred the former t,o their sub-genus liomancheUa, putting only those with a distinct tin in the sub-genus Lmosjnra. In the species S. verruca and S. evoliUus Bush and S. onedius, sp. n., however, the setfe are intermediate between these two types, having blades which are faintly notched (fig. 14 a), shoAving an indication of a superior blade and inferior fin. Consequently I propose to do away with Caullery & Mesnil's sub-genus Romanchella altogether — taking Lceospira to include all sinistral forms with three setigerous segments : it thus comes into line with the other three sub-genera proposed by Caullery & Mesnil {cf. Table of Classification). The second thoracic segment has only ordinary bladed setee, differing very little from one species to another. The third thoracic segment has some ordinary bladed setae, but generally also some bladed sickles as found so generally in the genus Apomatus (fig. 14 e). In many Pacific species there are present instead of the ordinary bladed sickles a peculiar shorter form which appears almost fringed at the extremity (fig. 10 c). 6. Abdominal ventral setse generally geniculate. 7. Uncini similar in thorax and abdomen — plates with the free edge provided with fairly numerous fine teeth, the anterior one being larger than the others. The following Table of Classification has been adapted from that given by Caullery & Mesnil in their excellent paper on ^'pirorbis (5) to include such of the Pacific forms as have so far been studied. These authors pointed out that such a modification would possibly be necessary. Unfortunately the majority of the new species described by Bush (2, 3, 4) from California, Alaska, and Japan cannot be included owing to absence of information as to the number of thoracic segments. Two or three of the species are established on details as to the tubes alone. Sub-genus Paradexiospira Caullery &, Mesnil (5), 1897 (modified). Characteristics : — 1. Tube dextral. 2. Thorax with 3| setigerous segments. 7. Spirorbis vitreus Fabr., 1780. (PI. LXXXVIII. fig. 7.) Serpida vitrea Fabricius (7), 1780. Spirorhis vitreus Morch (20), 1863; Malmgren (14), 1867; Levinsen (10), 1883; Caullery & Mesnil (5) 1897; Moore (17), 1902 j Bush (3), 1905. 794 MISS HELEN nXELL OX ft) 02 rfl ^s 8 S 3 Sog ^i ^ » ^ P'';^ s ■= o 5^ ^ g§. ■5 S a i •^ "^ ? s <0 1 i: » n 5 ? 1- 5» 1 1 a 1 S 1 2 s i.3 &J so 50 00 so CO 00 J « a w o ^ ;:^ < So H u 3 =Q is o of a o » ffi-g ffi.s o c o3 g ^H o O-^ o3 — ph::, ft. A CO u HlJ 796 MISS HELEN PIXELL ON Specific cliaractevistics : — • 1. Last thoracic segment on the right has no dorsal seta?. 2. Collar setae with fin-like expansions at base of deeply serrated blades. 3. Some setje of third segment bladed sickles. 4. Embryos incubated in the tube. This species was quite common on stones and i-ocks from the Departure Bay region and from Victoria. The tubes vaiy a good deal but are always ti\anslucent, Avith the whorls piled on one another. The whole tube measures about 2"5 mm. in diameter and the aperture about 1 mm. across. There are various markings on the exterior. Sometimes a ridge along the median line ends in a sharp pi'ojection above the aperture — in the grooves on either side of this there may or may not be scalariform markings. Young specimens have shells as figured for this species by Levinsen (10. fig. 11), older ones were ridged more like /Sp. cancellatns (10. fig. 18). The living- animals were bright pink in colour. Gills 7, each with six pairs of long opposite pinme, opercular plate a shallow funnel. About 20 abdominal segments. Sub-genus Dexiospira (Oaullery & Mesnil, 1897). Characteristics : — 1. Tubes dextral. 2. Thorax with 3 setigerous segments. 8. Sptrorbis spirillum Linne, 1767. (PL LXXXVIII. figs. 8rt-8c.) Serpida sinriUum Linne (12), 1767 ; Fabricius (7), 1780. tSpiroybis spirillum Malmgren (14), 1867; Levinsen (10), 1883; Caullery& Mesnil (5), 1897; Moore (17), 1902; Bush (3), 1905. Spirorbis Imldus Montagu (16), 1803; Morch (20), 1863; Malmgren (14), 1867." Circeis armoricaiia Saint- Joseph (24), 1894. Spirorbis borealis Fewkes (8), 1885. Specific characteristics : — 1. Collar setje geniculate (fig. 8 c). 2. Operculum without brood-pouch. 3. Concave plate of operculum has a slight projection (talon) on under side. Two vai-ieties of this species were fairly common in the Departure Bay region and at Victoria, growing on calcareous polyzoa, seaweeds, etc. The discoid form grows only on smooth surfaces, Larainarian thalli being often extensively covered with specimens. These were very fiat, regularly coiled tubes — the spiral with I5 to 3 coils measuring "5 to 2 mm. in diameter (tig. 8 a). The POLYCH.ETA FROM NORTH AAIERICA. 797 ascending variety in some localities is much more common than the discoid one, and it grows to a much larger size (fig. 8 b), often attaining a height of 5 mm. It was generally found with its lower coils overgrown with an orange Bryozoan growing on Ghcetopterus tubes. The living animal had a reddish colour, with a colourless ti-ans- parent operculum. A string of pink ova extended along the tube beyond the posterior end of the body in some specimens. There Avere 3 thoracic and from 12 to 20 abdominal segments. The collar setae attain a length of '27 mm., much larger than those figured by Caullery & Mesnil (5. fig. 4 5), they resemble more closely in shape those given for S. armor^icanus (5. fig. 5 h). The ditterences between these two species were pointed out by Caullery &, Mesnil (5. p. 199) to be quite unimportant. All that they could summarise were that 8p. armoricanus (Circeis armo7-icanus St.- Joseph (24. p. 350)) was slightly larger, had more abdominal segments (16-20) and a reduced talon to the operculum. They record having seen intermediate forms themselves, and the above observations as to variation in size, the number of abdom- inal segments, and the collar setaj conseqviently confirm their opinion that Sp. armoricanus should be considered as a variety only of Sjy. spirillum,. 9. Spirorbis pusilloides Bush (3). (PI. LXXXVIII. figs. 9«, 9 b \ S. pusillus Caullery & Mesnil (5), 1897. *^ Mera pusiUa St. -Joseph (24), 1894. Xon S.2ncsiUas Ilathke (22 a), 1836. Specific characteristics : — 1. Collar setfe of a more or less geniculate form (fig. 9 a). 2. Embryos incubated in operculum. 3. Seta? of 3rd thoracic segment sickle-shaped. 4. Hepatic pigment reddish brown. On stones from Taylor's Bay, Cabriole Island. This animal agrees in all important points with the full description by St.-Jos3ph (24. p. 351) for Mera jjusilla. The collar setee are a little more distinctly angulated at the base of the blade (fig. 9 a), but they have not the typical geniculate form described by Caullery & Mesnil (5. p. 202). They are much shorter than the setae of the second segment and are decidedly hooked. The extent to which the setae are hooked has been pointed out by St.-Joseph to be reduced by preservation (24. p. 338), so that this does not seem to be a very important point. The opercular brood-pouch is somewhat cylindrical and bounded proximally and distally by calcareous plates, the talon is red\xced to a small quadrangular projection. The abdomen has eight segments, with large, more or less sickle-shaped setae (fig. 9 b) and an asetigerovis anal segment. ^^98 MISS IIELK.N PIXKLL UN Sub-genus PROTOL.tiusi'iRA, now 1. Tube sinistral. 2. Four complete setigerous segments to thorax. 10. SriRORBIS AMBILATERALIS, Sp. 11. (PI. LXXXVIII. Hgs. 10«-10e.) Specific characteristics : — 1. Collar setie very large, conspicuously serrated blades with fin-like expansion at base (tig. 10 a). 2. Operculum without brood-pouch. 3. Talon of operculum with large hook-like process (fig. 10 6), 4. Some setae of 3rd segment have fringed ends (tig. 10 c). 5. Some sette of 4th segment ordinary bladed sickles. 6. Abdominal setaB brush-like (tig. 10 e). 7. Dorso-lateral brood-pouch. Several specimens on the inner sides of shells of Balanus nubilus from Dodds Narrows, 15-25 fathoms. Tubes forming translucent sinistral spirals measuring 3 to 4 mm. across. The surface is distinctly corrugated outside, highly polished iuvside ; the aperture measures 1 mm. in diameter. Branchiaa 12 — G on the right, and the operculum with 5 others on the left. The pinnie extend xipwanls and reach the same height as the rachises. The opercular plate (fig. 10 6) is very like that of