Irt* '^. Y^ i*- *\fi/s. ^V*^ ■!^f •J.>'-'?? ^.;«V:. ■■ / '<5*W-f tf'i > ;>-:/?* ^2? .^fT -? I o^- ^ . • HARVARD UNIVERSITY. LIBRARY MUSEUM OF COMPARATIVE ZOOLOGY. io^ll. D~ ^^-vnl^T. I J%'^) jhd.^ijjJ2 I L 0, /^JD THE TEANSACTIONS OF THE LINNEAN SOCIETY OF LONDON. SECOND SERIES.— VOLUME VII. ZOOLOGY. LONDON: PRINTED BY TATLOK AND FRANCIS. BED LION COUET, FLEET STREET. SOLD AT THE SOCIETY'S APARTMENTS, BTJRLINGTON HOUSE; AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW. 1896-1900. .■A AUG 10 1900 C O N T E N T S. PART I.— December, 1896. I. Heport on a Collection of Cephalopoda from the Calcutta 31tiseum. By Edwin S. Goodrich, B.A., F.L.S., Assistant to the Linacre Professor of Comparative Anatomy, Oxford. (Plates 1-5.) pages 1-24 PART II.— May, 1897. II. Amphipoda from the Copenhagen Museum and other Sources. — Part I. By the Bev. Thomas R. R. Steering, 31. A., F.B.S., F.L.S. (Plates 6^14.) . . 25-45 PART III.— June, 1897. III. The Origin of the Corpus Callosum : a Comparative Study of the Hippocampal Begion of the Cerebrum of Marsupialia and certain Cheiroptera. By G. Elliot Smith, Jf.Z)., Ch.M. {Sydney), "James King" Besearch Scholar of the University of Sydney ; St. John's College, Cambridge. (Comtnunicated by Prof. G. B. Howes, Sec. Linn. Soc.) (Plates '15 &'16.) 47-69 PART IV.— Jlne, 1898. IV. On the Muscular Attachment of the Animal to its Shell in .some Fossil Cephalopoda {Ammonoidea) . By George Charles Crick, F.G.S., F.Z.S., of the British Museum {Natural History). {Communicated by the President, Dr. A. GOnther, M.A., F.B.S.) (Plates'^lT-SO.) 71-113 PART v.— October, 1898. V. Observations on the Structure and Morphology of the Cranial Nei'ves and Lateral Sense Organs of Fishes ; toith special reference to the Genus Gadus. By Erank J. Cole, Demonstrator of Zoology , University College, Liverpool. [Communicated by Prof W. A. Herdman, D.Sc, F.B.S.) (Plates 21-23.) .... ' 115-221 [ iv ] PAUT VI.— November, 1898. VI. A Contribution towards our Knowledge of the Morphology of the Oiols. By ^N.V.VYCY^\Y■v, A.L.S.,M.B.O.U. (Plates '24-29.) pages 223-275 PAET VII.— January, 1899. VII. The Brain in the Edentata. By G. Elliot Smith, M.B. (Sydney), St. John's College, Cambridge. {Communicated by Prof G. B. Howes, F.E.S., Sec. Linn. Soc.) (With 36 Illustrations in the Text.) 277-394 PART VIII.— May, 1899. VIII. Amphipoda from the Copenhagen Museum and other Sources. — Part II. By the Bev. Thomas R. E. Steering, 3LA., F.B.S., F.L.S. (Plates ""SO^.) 395-432 PART IX.— November, 1899. IX. On Fossil and Recent Lagomorpha. By C. I. Porsyth Major, M.B. {Com- municated by Prof G. B. Howes, Sec. Linn. Soc.) (Plates ''36-39.) . 133-520 PART X.— March, 1900. X. On the Genus Lemnalia, Gray; with an Account of the Branching -systems of the Order Alcyonacca. By Gilbert C. Bourne, M.A , F.L.S. , Fellow and Tutor of New College, Oxford; TJnicersity Lecturer in Comparative Anatomy. (Plates'40-42.) 521-538 PART XL— March, 1900. Titlepage, Contents, and Index 539-550 ERRATA. Page 180, line 'I'l from toj), tor Fierasfer dentulus read Ficrasfir dcntutas. Page 228, line 9 from bottom, for Pterohjsis read Pterylosis. Page 242, line 10 from bottom, for cops letjcotis, Temm., read Scops LEncoiis, Temm. Page 285, line 3 from bottom, for Ohlamdyoj^Jiorus read Chlamydophorus. Page 516, line 17 from toji, for J'lvhiyiis iniiiKjctisis, Kon., read Prolagus ceniiigensis, Kon. '/ (/ "T^ 2nd Ser. ZOOLOGY.] [VOL. VII. PART 1. THE ^^/l. ^ - 12^ TRANSACTIONS OF THE LINNEAN SOCIETY OF LONDON. REPORT ON A COLLECTION OF CEPHALOPODA FROM THE CALCUTTA MUSEUM. BY EDWIN S. GOODRICH, B.A., F.L.S., ASSISTANT TO THE LINACHK PROFESSOR OF COMPAKATIVK ANATOMY, OXFORD. LONDON: PRINTED FOR THE LINNEAN SOCIETY BY TAYLOR AND FRANCIS, RKD LION COURT, FLEET STREET. SOLD AT THE SOCIETY'S APARTMENTS, BURLINGTON-HOUSE, PICCADILLY, W., AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW. December 1896. TRANSACTIONS OF THE L I N N E A N SOCIETY. I. Report on a Collection of Cephalopoda from the Calcutta Iluseum. By Edwin S. Goodrich, B.A., F.L.S., Assistant to the Linacre Professor of Comp. Anatomy, Oxford. (Plates 1-5.) Eead 19th December, 1895. TiAST year Professor E. Ray Lankester received for examination from the Calcutta Museum a large collection of Cephalopods, which he kindly handed over to me to be identified. The work was carried on at the Oxford Museum and at the British Museum, Cromwell E-oad, and I must express my thanks to Dr. A. Glinther and Mr. E. A. Smith for placing the collections in the latter Institution at my disposal. The present collection is composed almost entirely of specimens captured during the voyages of H.M.S. 'Investigator.' It contains 162 specimens, belonging to 28 genera. Eleven new species are described, belonging to the genera Inioteuthis, Sepia, Loliolus, Sepioteuthis, Abralia, Cheiroteuthis, Histiopsis, Taonius, and Octopus. No new genus has been foimded ; but four genera included are new to the Indian region. Of these the genus Calliteuthis has been recorded from the Atlantic and Pacific regions ; the ^genus Histiopsis from the Atlantic ; and the genera Loliolus and Inioteuthis from the Pacific and Japanese regions. SECOND SERIES. — ZOOLOGY, VOL. VII. 1 2 MR. E. S. GOODEICH ON A COLLECTION OF DECAPODA MYOPSIDA. Family SEPIOLINI. Iniotetjthis maculosa, n. sp. (PI. 1. figs. 1, 2, & 3.) One specimen from the Anclamans, and another from the Persian GuK ; both females. The principal measurements of the first are * : — Length of mautle (lower surface) TS centim. Breadth „ '95 „ „ . attachmeut to neck '4 „ ,, between fins '8 ,, J, across eyes '7 „ „ of fin '4 „ Length of fin-attachment '4 „ „ from extremity of mantle to tip of dorsal arm . 2'4 ,, Length of dorsal arm '8 „ „ dorso-lateral arm "10 „ „ ventro-lateral „ '11 >, „ ventral „ '7 » „ tentacular „ 2'1 „ „ „ club '5 „ The mantle is rounded at its apex. The round fins are attached about halfway up the mantle. The band uniting the mantle to the neck is narrower than in In. Morsei. The funnel has a small opening, and a wide base on which are l-shaped sockets. The two muscular bands which unite the base of the funnel above the sockets to the head are less prominent than in Sejnola oi' In. Morsei. There are glandular pads and a small valve inside the funnel. The edge of the bviccal membrane is notched, but the lobes are not distinctly marked. The first two pairs of arms are rounded ; the ventro-lateral arms are slightly keeled ; the ventral arms are provided with a well-developed keel on the upper edge. They all bear two rows of round suckers, obliquely set on slender stalks rising from a swollen base. The opening of the sucker is wide ; the horny ring has an ornamented surface and a smooth edge. The tentacular arms are flattened, and a groove runs down tbe inner surface. The club is of great Iciigth, slightly enlarged, provided with a lateral membrane on both sides, and a large niimber of minute suckers in eight rows. Each sucker is nearly hemi- spherical, set obliquely on a long slender stalk (fig. 3). The papillary area of the horny ring is wide, and the edge armed with alioiit 15 teeth (fig. 2). The ground-colour of the first individual is pale brownish yellow, inclining to orange * In the following descriptions the terms " upper " and " lower " are used to denote the surfaces generally called " dorsal " and " ventral ; " since, strictly speaking, they do not correspond to the morphological dorsal and ventral surfaces. CEPHALOPODA PEOM THE CALCUTTA MUSEUM. 3 on the arms and upper surface of the mantle. Scattered over the mantle, more especially on the lower surface, are large hrown chromatophores (fig. 1) ; similar clu'omatophores are seen on the upper surface of the fins, and on the head and arms. There is a con- spicuous row of five between the eyes on the upper surface of the head. The second specimen is of a duller tint and less well preserved. This species appears to he more closely related to Iniotenthls japoiiica (Tilesius, MS.), Verrill, which also has two rows of suckers on the arms, than to In. Morsel, Verrill, which has foiu*. Inioteuthis Moksei, Verrill. Seven specimens from the Andamans have been placed in this species. FamUy SEPIAEII, Stp. Subfamily Sepiadarii, Stp. Sepiadarium Kochii, Stp. FoLU' specimens, all female. One from off the south coast of Ceylon, lat. 6° 6' 30" N., long. 81° 23' E., from a depth of 32 fathoms. The other three from the Andamans. Sepia singaporensis, Pfeff'er. One female specimen from Singapore. Sepia aculeata, von Hasselt. Ten specimens apj)ear to belong to this species. One from the Irawaddy delta, at a depth of 20 fathoms ; four from Port Blair, and five from the Andaman Sea. Sepia singalensis, n. sp. (PI. 1. figs. 4, 5, 6, 7, & 8.) Two male specimens ; the first from Colombo, the second from off Point Galle (?). The principal measurements of the former are : — Length of mantle (above) 16 ceutim. (below) 14-1 „ from „ apex to mouth . . . 17'4. „ Breadth of „ at origin of fins . . . 4'9 „ „ » halfway 6-2 „ fin 1-2 )> Length of dorsal arm 7 „ „ dorso-lateral arm 6"5 „ „ ventro-lateral „ 6"3 „ „ ventral „ Q-7 „ „ tentacular „ 12'7 „ „ „ club 3-1 „ pen 15'5 „ Breadth of „ 4-8 „ 1* 4, ME. E. S. GOODEICH ON A COLLECTION OF The mantle is of a narrow oval shape ; the strong fins arise a little way below the mantle-margin, and do not join at the ajiex (fig. 4). The siphon is thick-walled, with a somewhat crescentic opening; it does not reach to the junction of the ventral arms. The arms have broad compressed bases (especially the ventral arms, which are 23 mm. broad), without well-developed keels. The inner surfaces, bounded on either side by narrow lateral membranes, bear four rows of moderate-sized suckers, largest on the lateral arms, where they attain a diameter of 2 mm. On the distal half of the arms the suckers become extremely small. The horny ring of the arm-suckers has a narrow paj)illary area, and a margin smooth on the proximal two-thirds of its circumference, and armed with very irregular teeth on the distal third (fig. 8). The stem of the tentacle is compressed, forming a sharp edge on the outer surface. The club is furnished as usual with a wide keel springing from the upper surface. Svickers of very unequal size are jjlaced in four rows on the proximal region ; the largest attain a width of 4 mm. in the two central rows, and are provided with smooth horny rings. The suckers diminish rapidly in size at either end ; these and the suckers of the outer rows have horny rings armed all round with numerous sharp teeth (fig. 7). In the distal region the small suckers are closely packed in six or seven rows, and near the extreme apex two suckers stand out from the rest on the upper margin as if utilized for some special purpose (an arrangement I have noticed in Sepia officinalis). The buccal membrane has seven lobes provided with a few suckers, the horny rings of which closely resemble those of the arm-suckers. A small Aveb is developed at the bases of the arms except of the ventral jiair. The mandibles are shown in fig. 6. The long oval pen of this Sepia is very remarkable (fig. 4). Above, the rough calcareous surface raised in three slight ridges narrows anteriorly, and is bounded on either side by a broad expanse of the chitinous margin {ch.m.) (in places about 1 cm. broad). Below, the striated area formed by the loculi reaches far up (loc), and is marked in the centre by a longitudinal groove. The margin of the inner cone is produced forwards and reflected over the posterior loculi, to which it is closely applied (i.e.). The apical spine has unfortunately been broken off ; it appears to have been small. It is not without hesitation that I have placed these specimens in a new species, as they closely resemble Sepia Bouxii, d'Orb., described in d'Orbigny and F^russac's monograph from the Indian seas (5). There are, however, several characters in which the two species differ. "Whereas 5. Mouxii is of a " forme generale racourcie," has very wide fins, and toothed rings to all the suckers of the tentacular club, S. singalensis is of narrow shape, with moderate fins, and smooth rings on the largest suckers. On the other hand, their pens are remarkably similar. The ground-colour of the best specimen is pale brown, speckled vnih. slate-coloured chromatophores on the lower surface, and conspicuously striped with broad dark bands on the ixpper surface of the mantle, head, and arms. The hectocotylized region is short and sitviated about halfway up the left ventral arm. CEPHALOPODA FBOM THE CALCUTTA MUSEUM. 5 At this point the suckers are small and resting on elevated transverse ridges. Scattered about the siphon of one specimen are spermatophores, conspicuous for the covering of black adhesive substance at the base (fig. 5). Sepiella inermis (van Hasselt, MS.). Among the 25 specimens referred to this species there is great variation as to general appearance, shape, width of the fins, and details of the structure of the pen. Neverthe- less I have not been able to distinguish any modifications of specific importance. The differences in the soft parts seem to be due to varied preservation. Six specimens were captured near Madras, three near Bombay, five at Sandlieads, one in the Chilka Bight, one at Mergui, two at Singapore, two at Penang, and five off the Ganjam coast at a depth of 10 fathoms. PamHy LOLIGINEI. Sepioteuthis indica, n. sp. fPl. 1. figs. 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, & 19.) Although the nine specimens in the collection belonging to the genus Sepioteuthis differ considerably in size and colour, I refer them all to one species. The best- preserved individuals are of a pale brownish-yellow ground-colour, sprinkled with slate- coloured chromatophores, very thickly on the upper surface of the mantle, head, and arms, more sparsely on the lower surface. Seven specimens come from the Andamans, and two from Singapore. Below are the principal measurements of a large male and of a large female : — Large Male. Large Female, centim. centim. Length of mantle (lower surface) .... 15'5 18"3 Breadth „ near margin 5"1 5'9 „ „ halfway 5-1 6*1 „ fin 7 cm. from mantle-edge . . . 2'9 2'9 ir, £ ii 1 .1 12-5 cm. from") n c- „ „ 10 cm. trom mantle-edge ... 3 \ o'b mantle-edge J Length of dorsal arm 5"2 5"5 „ dorso-lateral arm 6"3 6*5 „ ventro-lateral „ 8 9 „ ventral „ 7-4 8"6 „ tentacular „ 13'5 15 club 5-2 Q-7 Length of pen . . . ^ \2> cm. $ 20"5 cm. 12-7 cm. Breadth of pen . . . 2"4 4 3 Prom the outline figure given (fig. 9) it Tvdll seen that the fins are perfectly rounded, while the mantle from about halfway tapers to a blunt point ; the whole forming an almost perfect oval. In this respect this species resembles 8. BlaiiwilUana, Per., and S. mauritiana, Q. & G. The fins begin "35 cm. from the margin of the mantle and join 6 MK. E, S. GOODRICH ON A COLLECTION OF behind at the aj)ex. The siplion presents no peculiarity. The olfactory fold below the eye is prominent, undulating, and with its two extremities turned forwards (fig. 13). The buccal membrane is large (in one specimen extending 2 cm, beyond the mandibles), and provided with seven lobes bearing suckers. The horny ring of the latter is armed on its distal border with strong teeth (fig. 15). The jaws are shown in fig. 14. Adhering to the buccal membrane of tliis specimen (a large female) are spermatophores very similar to those oi LolUjo Indica (PI. 2. fig. 28), but slightly thicker. The powerful arms are provided on both edges of the inner surface with a lateral membrane, strengthened by transverse muscular ridges springing fi'om the base of the suckers. This membrane is especially well-developed on the lower edge oE the second or dorso-lateral arm. A median outer keel is developed along the dorsal and ventro-lateral arms ; a lateral keel on the outer lower edge of the dorso-lateral arm ; and two lateral keels on the ventral arms, the upper one being large and membranous. The suckers, closely set in two rows, are of moderate size, largest on the dorso-lateral arms, where they attain a diameter of -35 cm. The horny ring is armed with from 20 to 28 strong teeth (generally 21-22), rather larger on the distal border (fig. 17). The stem of the tentacular arm is thick, and flattened near the base (13 mm. broad aboixt 5 cm. from the base in the large female already mentioned), from the inner edge of which springs a membrane reaching down below the buccal membrane. The club is wide, provided with a keel on its outer edge and with a membrane on either side, and bears four rows of suckers of tmequal size. The largest suckers, situated towards the centre of the club, reach a diameter of 5 mm. ; their horny rings are armed usually with 15 teeth (figs. 18, 19) (sometimes with 16-18 teeth). The smaller distal and proximal suckers and the outer rows of suckers usually have 17 large sharp teeth on the distal border. At the tip of the club is a slightly spoon-shaped expansion, on which the suckers are modified, and arranged in two rows on either side enclosing a space (fig. 12). These suckers are highly specialized, being small, flat, and very short-stalked ; the papillary area of the horny ring is very wide and beset with rows of conical teeth reaching the margin (figs. 11 & 12). Such a special group of suckers constitute what I shall call the apical set, an arrangement which has not, so far as I am aware, been noticed by previous observers in the Myopsida. It is found in both sexes, and reminds us strongly of a somewhat similar group in Onychotenthls; the apical set of suckers, perhaps, forms a sort of " connective appai^atus." It occurs also in Loligo JPealei and L. indica ; on the other hand, it appears to be absent in L. marmora and in the genus Loliolus. ^ In the male the hectocotylized left ventral arm is modified from about the 23rd pair of suckers to the tip. In this region (fig. 16) the suckers are reduced to conical papillge. The right arm, as noticed in other species by Steenstrup (7), is also slightly modified, the suckers near the tip being unusually small. The pen is very like that of SejjioteutJns Blainvilliana, Fer., thin, lanceolate, convex above, with a strong, raised, convex ridge down the middle (fig. 10). It is often very difiicult to distinguish between the various species of this genus described by authors. Sepioteuthis Blainvilliana, Fer., and S. nlauritiana, Q. & G. (5), CEPHALOPODA FEOM THE CALCUTTA MUSEUM. '7 •seem to be tlie nearest allies of this new species. Prom the former it is distinguished by the lesser number of teeth on the homy rings of the arm and tentacular suckers, by the presence of suckers on the buccal membrane, and by the undulation of the olfactory ■.crest (?). From the latter, S. maurltiana, it is distinguished by the lesser number of teeth on the horny ring of the slickers and by the sliape of the pen. LOLIGO INDICA, Pfeffer. (PI. 2. figs. 20, 21, 22, 23, 24, 25, 26, 27, & 28.) To this species arc referred eight specimens from Mergui, one from Camorta Harbour, one from Daley Sandheads (1887), and two from the mouth of the Mutlah River; also, with some hesitation, a large female from the Ohilka Bight. The first 12 specimens differ little in size ; below are the measurements of a male and of the large female mentioned above : — 6- ?• ecntim. ceutim. Length of mantle (npper surface) 9"1 14 Breadth „ at beginning of tins .... \Q 3 Length of fin-attachment 5 7"6 Breadth of fins combined -i'S 7"65 Length of pen of a smaller ^ , 6'6 centira., breadth ] -2 centim. „ „ the large 2 , li centim., breadth 2-7 ceutim. I feel obliged to go into some detail concerning this species, since the figures and descriptions published by Mr. Hoyle of the ' Challenger ' specimens (i) differ widely from Dr. Pfeffer's original description (6). Concerning the horny ring of the suckers of the arms, Pfeffer says, " Armringe mit 6 ziemlich grad viereckigen Zabnen ; darauf folgt die erhabene Leistc, die vorn nocli jederseits cine Einkerbung als Paxdiment eines Zahnes zeigt." This hardly agrees with Hoyle's figures, but corresponds exactly to my figure of the horny ring of the female (fig. 26). With regard to the horny ring of the suckers of the arms, I find that there is a well- marked sexual difference, for in the males these rings are armed with from 9 to 11 rounded teeth (fig. 20). Again, concerning the horny ring of the tentacular suckers, Pfeffer says, " Grosse Tentakelringe mit etwa 21 spitz dreieckigen Zahnen, deren Basis schmaler ist, als die Zwischenraume." While this description differs much from Hoyle's figures, it agrees entirely with what I find (fig. 27). iVlthough 21 is the usual number of teeth in these specimens, there are sometimes as few as 15 in the large female. Pig. 24 shows the horny ring armed with G strong teeth from one of the very small distal suckers of the ckib ; while fig. 25 represents a portion of the horny ring from one of the suckers of the apical set at the tip of the club (cfr. Sepiotetdhis, j). 0). The lobes of the buccal membrane are provided with small toothed suckers, the horny ring of which is shown in fig. 21. I figure also a club-shaped spermatophore removed from the buccal membrane (fig. 28). '8 ME. E. S. GOODEICH ON A COLLECTION OF The hectocotylized arm of the male is strongly modified. On its proximal half are two rows of small suckers (fig. 22) ; on the outside of these the upper edge of the arm is drawn out into a wide thick flap (fig. 22, Ji.). On the distal half of the arm the suckers are reduced to rounded projections {m.s. fig. 23) which bend over a groove. The outer and upper edge of the arm is developed into a lateral membrane of considerable width {l.m. fig. 23). LoLiOLrs iNVESTiGATOKis, n. sp. (PI. 2. figs. 29, 30, 31, 32, 33, 34, 35, 36, & 37.) Nine specimens, some of which are much mutilated. The specimen figured and one other come from the Mutlah River, five from the Ye River, one from Mergui, and one from Sangor. They vary considerably in size ; the following are the principal measurements taken on the specimen figiu'ed, a male of average size : — Lengtli of mantle (upper surface) .... „ from apex of mantle to base of arms Breadth of mantle near edge „ across eyes .... „ „ combined fins Length of attachment of fins „ dorsal arm dorso-latcral arm ventro-lateral „ ventral „ tentacular „ 3'5 centim 4 1-3 1-35 2-9 2-1 •8 1-6 2-] 1-8 3-3 , The mantle is of conical shape, with a roimded apex ; its margin projects on the upper surface into a small point above the neck, and below on either side of the funnel (figs. 29 & 30). The fins have rounded anterior and lateral edges ; the two together form a heart-like figure. They join and are continued beyond the mantle apex. The head is large and provided with a transverse olfactory crest below the eye. The siphon possesses a valve, and muscular bridles hidden below the skin. The three dorsal pairs of arms bear on their outer sui'face a median keel, especially well-developed on the 3rd pair. The ventral arms are somewhat square in section, and both the outer edges are drawn out into lateral keels. Small lateral membranes spring from the inner surface on the ujiperside of the 1st and 2nd pairs of arms, and on both sides of the 3rd and 4th pairs of arms. The suckers, in two rows, are obliquely set on their stalks (fig. 36) ; the horny ring is provided with three large blunt teeth (fig. 37). The tentacular arms enlarge distally into small c]u.bs, bearing an outer keel and four rows of suckers. The horny ring of these is armed with sharp teeth on its distal border (fig. 35). A few small suckers are situated on the seven lobes of the buccal membrane ; their horny ring strongly resembles that of the tentacular suckers, but there are only from four to six large teeth. The mandibles are rather weak (fig. 34). The lower mandible has a sharp, hard, tooth-like point and a small basal tooth (b.t. fig. 34 b). CEPHALOPODA FROM THE CALCUTTA MUSEUM. 9 The pen is lanceolate; narrow anteriorly, it broadens out to a thin blade behind (fig. 31). In the male the left ventral hectocotylized arm is somewhat sickle-shaped (figs. 29 & 32). The upper outer edge is developed into a wide lateral membrane [l.m. fig. 33), and the suckers are more modified than in Steenstrup's species (7), the upper row being reduced to mere papillae, and the lower row having almost entirely disappeared (fig. 33). OIGOPSIDA. Family ONYCHII. Abralia andamanica, n. sp. (PL 2. figs. 38, 39, 40, 41, 42, 43, 44, & 45.) Three specimens, captured at a depth of from 188 to 320 fathoms in the Andaman Sea, belong to this species. The following are the measurements of the largest specimen, a female (fig. 38) : — Length of the mantle (lower surface) 3"8 centim. „ from mantle-edge to base of arms . . . 1"2 „ ,, of attachment of fins 2'3 „ „ from mantle-ajDcx to angle of fin ... 2*7 ,, Bi-eadtli of combined fins 3'4 „ Length of dorsal arm 1-7 „ „ dorso-lateral arm 2 „ „ ventro-lateral „ 1*8 „ „ ventral „ 2 „ „ tentacular „ 4 „ club -8 „ Length of pen of a male, 2'4 centim., breadth 35 centim. The mantle is nearly conical, tapering gradually to the apex. At the free edge there is no pronounced dorsal point, but there is a small projection on each side of the funnel (figs. 38 & 39). The fins, rhomboidal in shape, reach about halfway np the mantle ; their posterior edge is slightly concave, their anterior convex and more rounded. The head is large, and fits closely on to the edge of the mantle. At the edge of the eyelid is a rudimentary sinus. On each side of the neck are two olfactory crests, the smallest being nearest the siphon. There is a well-developed valve in the funnel ; and the socket at the base of the latter is l-shaped. The buccal membrane has 8 lobes ; its nner surface is covered with pajiilte of considerable size. The first and second pairs of arms have a median external keel, and a small lateral membrane on the upper edge. The third, or ventro-lateral, pair has a large membranous keel {k, fig. 38), and a membrane along the lower edge. The ventral arms have the outer and upper edge drawn out into a lateral keel. All these arms bear two row^s of hooks for about three fourths of their length, and two rows of suckers on the distal quarter. On the ventral arm, for instance, 14 hooks and SECOND SERIES. — ZOOLOGY, VOL. VII, 2 10 ME. E. S. GOODEICH ON A COLLECTION OF 12 suckers can be counted. These suckers are very obliquely set on short stalks (figs. 42 & 43), and of a compressed conical shape. The horny ring is studded with tectli, and the margin is armed with small teeth on the proximal side and large blunt teeth on the distal side. The tentacular arms are scarcely at all enlarged to form the clubs (figs. 38 & 40). Each club bears a small keel externally, three hooks along the lower edge of the internal surface {Ji), and above these two rows of suckers. Beyond the hooks on the distal region small suckers are arranged in four rows. The horny ring of the suckers is furnished with small teeth round its margin (figs. 44, 45). Near the base of the club a set of three small short-stalked suckers alternating with three tubercles forms a connective apparatus {c. app. fig. 40). ^i§. 41 (a & b) shows the pen of a small male, the mantle of which is 3 cm. long ; it is lanceolate, with a strong median ridge. This specimen has a large bundle of sperma- toj^hores in the mantle-cavity. In colour the female is of j)ale yellowish-brown tinge, inclining to orange on the upper surface, covered with dark purjile-brown chromatophores, most numerous on the upper surface of the mantle, head, and arms, and on the buccal membrane. The male is paler, the chromatophores being less numerous. Both sexes are provided with a very large number of small dark tubercles, generally showing a white opaque lens in the centre, distributed over the lower surface of the mantle, head, funnel, and ventral arms (tig. 38). These tubercles, no doubt luminous or phosphorescent organs, are evenly scattered over the lower surface of the mantle ; on the siphon they are situated in six irregular longitudinal series ; on the head there is a row round the lower half of each eye and nine rows between these. Three rows extend on to the base of the ventral arms, but only two are continued to their extremities. Abralia lineata, u. sp. (PI. 3. figs. 46, 47, 48, 49, & 50.) Two specimens ; one, a male, taken at a depth of 265 fathoms in the Andaman Sea, the other, a female, at a depth of from 90 to 100 fathoms off the Ganjam coast. The principal measurements of the first specimen, which is the one figured, are given below : — Length of the mantle (upper surface) 1"5 centim. ,, „ (lower surface) 1"3 „ „ from mantle-apex to base of arms ... 2 „ ,, of attachment of fin "8 „ „ of jjosterior border of fin '9 „ Breadth of mantle "6 „ ,, combined fins 1'5 ,, Length of dorsal arm ■" ,, „ dorso-lateral arm I'O , „ ventro-lateral „ "9 „ ., ventral ,, I'l „ ,, tentacular „ 1"6 ,, CEPHALOPODA PBOM THE CALCUTTA MUSEUJM, 11' The sliort mantle tapers to a sharp point. The fins are rhomhoidal, with anterior rounded horder and nearly straight posterior border ; they extend forwards more than halfway up the mantle beyond the point of attachment. The head, of about the same width as the mantle, is provided with large eyes, and fits less closely on to the mantle than in the preceding species (this may be due to differences in preservation). A small olfactory crest is situated behind the eye. The funnel has a valve, two well-marked bridles, and l-shaped sockets. The buccal membrane has 8 lobes. The three dorsal pairs of arms are rounded proximally and slightly keeled distally. The ventral arms have a lateral membrane on the upper side. Two rows of hooks are borne by all the arms along almost their entire length. The ventral arms appear to bear no suckers at all distally, but their tips are swollen, and pi'ovidedin both sexes with three large convex pigmented organs forming a sort of club. These swellings are probably phosphorescent organs ; they correspond exactly to the modified arms described by Pfeffer in Enoploteuthis Iloylel, Pfelfer, and considered by him to be due to hectocotyli- zation. The other arms bear, near their extremity, a small number of minute suckers, the horny ring of which is armed on the distal margin with large teeth (fig. 49). The tentacular arm is slightly expanded to form a club, bearing in the middle region four hooks on the lower side, and three hooks alternating with suckers on the upper side (fig. 48). Beyond these are four rows of small suckers, with a wide ornamented papillary area and a smooth margin (fig. 50). At the base of the club is situated the connective apparatus of four suckers and tubercles (, Length of pen 40*4 „ Breadth „ "75 „ The colour of this specimen in spirit is of a dull purplish-brown, lighter and yellower on the lower surface, darker and inclining to heliotroije and blue on the uppper surface of the mantle, head, and on the outer surface of the arms. The buccal membrane, inner surface of the arms and suckers are paler. The mantle, as seen in the figure from the ventral surface, tapers gradually until the middle of the fin, whei-e it is very narrow ; it then enlarges slightly, and is produced beyond the fin at its apex {ap.2)rol.) very much as in Cheiroteuthis Ficteti, a species CEPHALOPODA FEOM THE CALCUTTA MUSEUJM. 13 admirably described by M. Joubin (2). This apical jirolongation of the mantle is shorter than in the latter species, and moreover is provided with a lateral fin on each side {ap-f.), a character which distinguishes the present species at once from any hitherto described. The mantle is laterally compressed, forming slight median longitudinal upper and lower ridges. Dorsally the mantle-edge is produced in a point above the neck ; below it is raised opposite the funnel. The fins are wide, the outlines of the right and left together forming an almost perfect circle {f, fig. 51). The funnel, provided with a large internal valve, is fixed to the head without distinct bridles, and at its posterior edge carries two sockets, each possessing two prominent lobes fitting into corresponding depressions in the mantle (fig. 56, a Sc h). The head is small ; the eyes moderately large and without sinus. About halfway between each eye and the attachment of the si2)hon is an olfactory jjapilla {olf.p.), about 2 mm. in length. The buccal membrane is well developed, with seven lobes and ridges. The mandibles are very sharp (fig. 54, a & h) ; the ujiper mandible has a basal tooth on either side {bt., fig. 54^), corresponding to which are small teeth on the lower mandible. The radula consists of seven rows of teeth (fig. 57), closely resembling those of C. Picteti, Joubin (2). All the arms are somewhat square in section, and are provided along the upper and outer edge with a membrane very narrow in the first two pairs, absent in the third, and very broad in the fourth (;«). The third and fourth pairs have narrow membranes along the lower and outer edge *. Two rows of small suckers are placed rather far apart along each arm (rather more closely set on the dorsal than on the ventral arms). Each sucker is obliquely set on a slender stalk, and has a narrow opening, the horny ring of which is armed on the distal side with square teeth (fig. 55). The soft rim of the sucker is more or less raised uj) into a peak above the toothed horny margin (fig. 55). On the ventral arms, at the base of the outer row of suckers are convex pigmented patches, no doubt similar to the " vesicules refringentes " described by Joubin in Ch. Veranyi (3). The tentacular arms, which have unfortunately lost their clubs, are long and slender, bearing at intervals the modified flattened suckers characteristic of the genus. The pen was somewhat injured owing to the contraction of the mantle. It is very long and narrow (fig. 52, a & h) ; the hollow cone reaches halfway up and opens by an oblique aperture. Diagrams of transverse sections of the cone and of the anterior region are given in figure 53, a & h. The pen thus closely resembles that of Ch. lacertosa, described by Verrill (9), and of Ch. Picteti ; yet the portion in front of the cone is not tubular, as figured by Joubin for the latter species Ch. Vei'anyi, on the other hand, has hitherto been described, by d'Orbigny and Ferussac (5), and by Verany (8), as possessing a pen flattened and expanded at both ends. On examining a siDecimen from the Mediterranean, I find, however, that its pen resembles in every essential that of Ch. macrosoma here figured ; in the case of the works referred to, the artist probably flattened out the slender cone before drawing it. * These membranes correspond rather to the keels in other forms, as they do not spring directly from the sucker-bearing surface of the arm. 14 ME. E. S. GOODRICH ON A COLLECTION OF Of the four species * of Cheiroteuthis hitherto known, two come from the Atlantic, one from the Mediterranean, and one from Aniboyna. In general shajoe and size the present species approaches most nearly the latter, Ch. Picteti ; yet it differs from it in the possession of a fin along the apical region of the mantle, in the beak, horny ring of the suckers, socket at the base of the siphon, and pen. f Cheiroteuthis pellucida, n. sp. (PI. 4. figs. 58, 59, 60, & 61.) This small specimen, cajitured at a depth of 922 fathoms off the Vizagapatam coast, lat. 16" 11' 15" N., long. 82° 30' 30" E., was brought up alive on deck, when it appeared transparent, with dark purple chromatophores. In spirit it is still remarkably trans- parent, and the chromatophores (not represented in fig. 58) are very large and few in nixmber, scattered at wide intervals over the body, fins, and arms. Tlie principal measurements are as follows : — Length of the mantle (lower surface) 4*8 eentim. Breadth „ '8 „ Length „ to sudden constriction . . 2'G „ „ „ beyond fin 'G ,, ,, of fin-attachment 2'1 ., Breadth of fin 1-8 Length from mantle-edge to base of arms . . . 2.2 „ Breadth across eyes -7 „ Length of dorsal arm 1-4 „ „ dorso-lateral arm 1'8 „ „ ventro-lateral „ 2"3 „ „ ventral „ 4'5 „ ,, tentacular ,, 8 „ >, ,> club 3 In general appearance this interesting little Cheiroteuthis reminds us strongly of Doratopsis vermiciilaris (Riippell) by its long and almost cylindrical mantle with a narrow apex, its long thick neck, and the great disparity in size between the ventral and remaining three pairs of arms. The mantle, as already mentioned, is nearly cylindrical until it suddenly narrows down in the region of the fins to a narrow straight prolongation, which extends beyond the fins, forming a rounded spine (fig. 58). The fins themselves are broad and rounded (/), but the outline is less circular than in the preceding species. Dorsally the mantle projects over the neck in a small point. The fiinnel is sharply recurved at its apex, and provided internally with a broad valve produced and thickened at the sides. The socket of attachment is shown in fig. 59. The head is hardly wider than the neck ; the eyes are somewhat prominent. Between these and the siphon are two olfactory papillae (olf.p.). The buccal membrane is furnished with seven well-marked ridges and lobes. Beyond the eyes the head narrows considerably, and enlarges again to support the * M. Joubin has lately described a new species from the Atlantic (4). CEPHALOPODA EfiOM THE CALCUTTA MUSEUM. 15 arms. The three small and dorsal pairs of arms bear two rows of suckers, aud have no lateral membrane. The large ventral arms have a membrane on the upper side ; toward, the base they bear two rows of suckers, but about halfway up the suckers come to alternate in so nearly the same sti*aight line that there appeal's here to be only one row. The horny ring of the suckers (fig. 60) is armed mth rounded teeth on the proximal, and powerful sharp teeth on its distal side. The tentacular arms are comparatively short, and provided at intervals with numerous flattened suckers. The long clubs bear four rows of small suckers, the outer row of which has the longest stalks. Round the margin of the club is a ribbed lateral membrane ; the apex ends in thickened, deeply pigmented, aud somewhat spoon-shaped knob. The horny ring of the suckers has a wide papillary area bearing sharp teeth, and characteristically notched on its distal border ; the proximal margin of the ring is smooth, the distal margin armed with powerful sharp teeth (fig. 61). I have not dissected out the pen ; it appears to closely resemble that of the foregoing species. Although, as the description shows, this specimen closely resembles JDoratopsis, yet the structui'e of the tentacular arms shows that it undoubtedly belongs to the genus Cheiroteutliis. HiSTiOPSis HoTLEi, n. sp. (PI. 4. figs. 62, 63, 64, 65, 66, 67, 68, 69, 70, & 71.) One specimen captured near the Andamans, at a depth of 190 fathoms. Length of mantle to notch between fins (upper surface) . 2"2 ceutim. „ „ to apex (lower surface) 1-9 „ Breadth ,, 1'2 ,, Length of fin-attachment to posterior notch "6 „ ,, fin beyond mantle-apex -'i ,, „ combined fins 1"3 „ „ dorsal arm (riglit) 3"5 centim. (left) 3"3 centim. „ dorso-lateral arm „ 3*4- „ „ 3*7 „ „ ventro-lateral ,, „ 3-1 „ „ 3'-i „ „ ventral „ „ 3 „ „ 3-2 „ „ tentacular „ „ 7 „ „ pen 2'1 centim. Breadth of ,, -6 „ The mantle swells out slightly midway, then tapers to a blunt point (figs. 62 & 66), Dorsally it projects above the neck ; the lower m.argin is nearly straight. The fins are rounded, joined distally above the mantle-apex, beyond which they project. The funnel is strongly recurved {si., fig. 68); the two lips which bound its aperture are so applied to each other as to leave a narrow U -shaped opening. The sockets at the base of the funnel are I -shaped {so.) ; two small bridles bind it to the head. Inside we find a small valve prolonged into the upper portion of the ' funnel-organ,' which covers the upper or attached Avall of the siphon, and into which fit two rounded flaps fixed on either side to the lower or free wall of the siphon. The whole arrangement closely 16 ME. E. S. GOODEICH ON A COLLECTION OF resembles that described and figvired below in Taonius abysslcola. Mr. Hoyle makes no mention of the ' funnel-organ' in liis species (i). The head is wide and sharply marked off from the narrow neck. Below each eye is an olfactory papilla {plf.j}., fig. 64). The eyes are large, with widely open eyelids (figs. 62 & 64) ; the left eye is injured and protrudes from the eyelid, it has been restored in the figure. The arms, which differ somewhat in length on the right and left sides, appear to be in the following order of length : 2, 1, 3, 4. Unlike H. atlantlca of Hoyle, the web at the base of the arm is quite rudimentary (fig. 63). The small suckers, placed in two rows on all the arms, are of a peculiar conical shape (figs. 69 & 70), with a wide opening furnished with a narrow horny ring armed round its margin with small sharp teeth. The left and only complete tentacle enlarges at its extremity into a small club (figs. 62 & 65), bearing four regular rows of small suckers on its distal half, and as many as six irregular rows on its proximal half. At the widest region of the club the central suckers attain a large size (fig. 65). They have round shallow cups, set almost straight on short thick stalks (fig. 67) ; the margin of the horny ring is armed with numerous long straight pointed teeth. Near the base there is a connective apparatus consisting of a row of three alternating suckers and tubercles {c.app., fig. 65) ; two or three small suckers arc scattered down the stem. The buccal membrane has seven lobes, with corresponding ridges extending on to the arms, as shown in figure 63. The delicate little pen is lanceolate (fig. 71). The colour of this specimen in spirit is pale yellowish-brown, covered with dark purplish-brown chromatophores, on the mantle, iipper surface of the fins, head, buccal membrane, and arms (especially on their iianer surface). As seen in the figiu'e, a large niimber of U-shaped pigmented organs, most probably phosphorescent (see Joubin, 3 & 4 a), are scattered in irregular transverse rows on the lower sm-face of the mantle, head, and arms ; a few extend over to the upper surface, and one row surrounds each eye (fig. 64). The first three pairs of arms have one row each extending to their tips, whilst the larger ventral arms have three rows proximally, reduced to two distally. The exact systematic position of this little Cephalopod is not very easy to determine. I have placed it provisionally in the genus Histiopsis of Hoyle, from whose specimen it differs in the absence of a distinct web at the base of the arms, in the horny rings of the suckers, in the buccal membrane, in the contimiation of the fins beyond the mantle- apex, in the possession of one row only of pigmented organs on the first three pairs arms, and perhaps in the presence of the ' funnel-organ ' (Verrill's organ). The horny rings of the tentaciilar suckers arc very like those oillistioteuthis, to which genus it is no doubt closely related. Calliteuthis eeveesa, Verrill. One specimen of this widely distribvited species was caught in the Andaman Sea at a depth of 265 fathoms. CEPHALOPODA FROM THE CALCUTTA MUSEUM. 17 Taonius abtssicola, n. sp. (PI. 5. figs. 72, 73, 71^, 75, 76, 77, 78, 79, & 80.) Two sj)ecimens from the Laccadive Sea belong to this genus. The first, the large specimen shown in figure 72, comes from a depth of 902 fathoms. It is somewhat injiu*ed ; the skin has been rubbed off the mantle, fins, and head, and the eyes have burst out of their lids. The second specimen, from a depth of 1370 fathoms, is very much smaller and in a less satisfactory state of preservation. The principal measurements of the large specimen are as follows : — Length of mantle (lowei- surface) 7*6 centim. Breadth „ near edge 3-8 „ Length of fin-attachment 2*4 „ Breadth across combined fins l-~ ^^ „ of head between eyes '7 „ „ across eyes 2'8 „ Length of dorsal arm 2'5 ,, „ dorso-lateral arm 2'8 ,, „ ventro-lateral „ 3"3 „ „ ventral „ 2-8 „ „ tentacular „ 4'2 ,, „ >, club 1-1 ,, The mantle is loose, enclosing an ample cavity ; it diminishes gradually to the origin of the fin, and then suddenly to the narrow apex. The outline of the fins is egg-shaped. The mantle is attached by a lozenge-shaped cartilaginous plate to the back of the neck, and on both sides to the base of the siphon. The aperture of the funnel is covered by the sharply recurved upper lip (fig. 72), All previous observers have described the funnel of this species as destitute of valve.* On slitting up the funnel of this specimen and turning aside the two sides as shown in figui-e 7i, I found that the inner and upper wall is sharply bent inwards near the opening so as to form a deep pit w^here the sij)hon is fused to the head. On either side of this pit are the two cushions [c) noticed by Verrill (9), which in their normal position almost close u.p the lumen of the funnel. When these two cushions are pushed aside, a well-developed valve is disclosed rising from the bottom of the pit (fig. 74 v.). Behind the cushions, further down the funnel, are two triangular flaps, flattened and fastened by their base to the sides of the funnel {t.Jl., fig. 7i) ; they appear to be of a glandular nature, and probably correspond to the lateral pads described in other forms by AVeiss (10). On the inner and upper wall of the siphon is a wide A-shaped plate — the funnel-organ — very similar in shape to that figured by Weiss in Verania sicula. In the present instance it was so loosely attaclied that it readily came off; it is represented in flgtire 75, and its position is indicated by a dotted line in figure 74i. On the inner surface this plate is produced on either side into a triangular cap {i.e.), which fits closely as a glove * Since this was written M. Joubin (4) has described a small valve in the siphon of a new species of Taonius from the Atlantic. SECOND SERIES. — ZOOLOGY, VOL. VII. 3 IS -MR. E. S. GOODEICH ON A COLLECTION OF on to the flaps already described lying opposite them.* I have confirmed these observations on the second specimen. Contrary to Verrill's observations in two other species (9), I find on either side between the eye and the funnel a truncated olfactory papilla (fig. 76). It is slightly expanded and flattened distally, the flat oblique surface being concave. The head appears to have been narrow. The two eyes are very large and protrude from their sockets. The arms are thick, and of considerable length for this genus. The three dorsal pairs are provided on both sides with a well-developed lateral membrane (largest in the third pair), springing from the inner edge and supported by muscular thickenings arising from the base of each sucker (fig. 73). The ventral arms are also provided with such a membrane on the lower edge ; but it is rudimentary on the upper side, where on the contrary the outer edge of the arm is produced into a lateral keel. On all the arms we find two rows of suckers, largest on the third pair. The suckers (figs. 77, 78), very obliquely set on a short stalk, have a wide opening provided with a horny ring, armed on the distal margin only with about 14 squarish teeth. The papillary area is narrow. The tentacular arms are thick- stemmed, the clubs only slightly enlarged and bearing a lateral membrane on either side similar to that of the arms. The suckers of the club are compressed and considerably injured in this specimen ; they appear to be set in four rows, and of a peculiar conical shape (fig. 80). The stalk of the sucker sw^ells gradually to the base of the cup, where there is a thickened ridge ; it then narrows and expands again into a bell-shaped cwp, with an oblique opening. The margin of the horny ring is armed with eight or nine strong curved teeth, extending round the distal two-thirds of the circumference. The striated outer margin is frayed out into a fringe, but this is probably artificial. From the club extending down the stem are small short- stalked round suckers, the horny margin of which is armed all rovmd with blunt teeth (fig. 79). Distally at the base of the club these suckers are placed in four rows, which dwindle gradually to two rows proximally, where the suckers are very small and set in pairs on either side of a median groove. The buccal membrane has seven ridges ; the lobes are indistinctly marked. Purplish-brown chromatophores colour the fragments of skin adhering to the head and arms, the lateral membranes between the ridges, and the buccal membrane. The pen can be seen, wdthout dissection, as a narrow ridge starting from the nuchal ]ilate and exjianding posteriorly into a thin plate, narrowing again to form a slender cone, as figured by Pfefi'er for Taonius {Ilegalocranchia) maximns (6). The second specimen is very much smaller, and is possibly a young individual of the same species ; it is, however, not in a sufficiently good state of preservation to allow one to make certain of its specific identity. * In their normal position the jioints of the flaps are directed away from the middle line ; in figure 74 they iirc directed inwards owing to the stretching open of the funnel. CEPHALOPODA FEOM THE CALCUTTA MUSEU.AI. 19 OCTOPODA. Family PTEEOTI. ClKRHOTEUTHIS PACIFICA, Hoyle ? A small and mutilated specimen, captured at a depth of 265 fathoms iu the Audamau Sea, has been referred to this species. Family OCTOPODIDiE.* Octopus vuiiGABis, Lam. One large female from the Andamans, and one smaller male from Point Galle, Ceylon. These specimens resemble very closely our common Octopod, and I could find no important distinction between the European and Indian forms. Octopus granulatus, Lam. Eleven specimens have been placed in this species. Three come from Port Blair, three from the Andamans, one from Great Cocos Island, one from Bombay, two from the Southern portion of the Malacca Straits, and one from Maskat in Arabia. Octopus globosus, AppeUof. (PI. 5. fig. 81.) One specimen from the Nicobars, one from the Kabusa Islands (Mergui Archip.), five from the Southern jiortion of the Malacca Straits, three from Bombay, and one from Point Galle, Ceylon — in all eleven specimens have been referred to this species. The extremity of the hectocotylized arm of a male is shown in figure 81. Octopus pictus (Brock), ^SbV.fasciata, Hoyle. (PI. 5. fig. 82.) One male specimen from Port Jackson, Avhich agrees very closely with the specimen described by Hoyle in the 'Challenger' BejDort. The extremity of the hectocotylized arm, however, shown in fig. 82, difiers considerably from that described by Hoyle in Oct. maculosus (Proc. U. Phys. Soc. Edinb. vol. viii. 1883), w^hich he considers to be the same species. It seems probable that the var. fasciata may have to be separated as a distinct species. Octopus Januarii, Stp. Three specimens appear to belong to this widely distributed and abyssal species. One, a female, comes from a depth of 193 fathoms in the Bay of Bengal (lat. 20° 17' 80" N., long. 88° 30' E.) ; the othei- two are males captured at a depth of 271 fathoms in the Andaman Sea. The ridges on the hectocotylized tip of the arm in the male are more marked than in the figure given by Hoyle in the 'Challenger' Report (i), but in aU essential respects these specimens agree with his description. * I must express my thanks to Mr. W. E. Hoyle for kiudly cxaminiug some of these Octopods. 3* 20 ME. E. S. GOODEICH ON A COLLECTION OP Octopus macropus, Risso. With considerable doubt eight specimens have been referred to this species. Seven come from the Andaman Islands (three from Port Blair) and one from the Malacca Straits. Octopus aculeatus, d'Orb. Ten specimens from the Andaman Islands (six from Port Blair), one from Colombo, and one from Little Cocos Island. It is with some doubt that these have been placed in this species, wliich is not very thoroughly defined. Octopus levis, Hoyle ? One female specimen from Gopalpur, at a depth of 7 fathoms, has been provisionally placed in this species. It agrees closely, although not perfectly, with Hoyle's description. Octopus microphthalmus, n. sp. (PI. 5. figs. 83, 84.) One female specimen from Port Blair, Andaman Islands (fig. 83). This species is remarkable from the smooth cylindrical shape of the head and mantle, the eyes forming no prominence wliatever on the surface of the head. Tlie mantle ends obtiisely, and fits closely on to tlie head ; the mantle-opening, which is very narrow, reaches rather more than halfway from the siphon to the eye. The aperture of the eyelid is small. The funnel reaches to within 4 mm. of the edge of the web between the ventral arms ; it is attached to the lower su.rface of the head along almost its entire length. Tlie arms are rounded, and l)ear small somewhat conical suckers, which are arranged some distance apart. Although they are placed in two rows, the suckers are situated in zigzag fashion so as to approximate to the condition found in Eledone, for instance, where they are in one row. The mandibles are dark and strong (fig. 84, a Scb). The following are the princij)al measurements : — Length of mantle and head to lower edge of web between ventral arms . . 4'5 centim. ,, „ to mantle-edge 2"7 Breadth of mantle 2 „ head across eyes 1*5 Length of dorsal arm 8'6 , dorso-lateral arm 8'1 „ ventro-lateral ,, 6'9 „ ventral „ G'9 There is a small web at the base of the arms, reaching 1-6 cm. from the mouth between the dorsal arms, where it is longest (the inner surface of this web is the same between the bases of all the arms). The colour is yellowish -brown, inclining to dark brown on the upper surface of the mantle, head, and outer surface of the arms. CEPHALOPODA FEOM THE CALCUTTA MUSEUM. .21 TAst of Works referred to in the Text. 1. HoYLE, W. E. — Challenger Reports, vol. xvi., pt. 4i, "Report on the Cephalopoda/' 188G. 2. JouBiN, L. — " Cephalopodes d'Amboine." Revue de Zool. Suisse, vol. ii., 1894. 3. JouBiN, L. — " Quclques organes colores, etc." Mem. Soc. Zool. de France, vol. vi., 1893. 4. JouBiN, L. — "Resultats des Campagnes Scieutifiques de Prince de Monaco": fasc. ix., 1895. 4fi!. JouBiN, L. — "Note sur I'appareil photogene d'uu Cephalopode." C. R. Soc. Biol. vol. v., 1893. 5. D'Orbigny, a., et Ferussac. — Hist. Nat. des Cephalopodes acetabuliferes. Paris, 1835-48. 6. PfEFFER, G. — " Die Cephalopoden des Hamburger Nat. Mus." Abhandl. d. Naturw. Ver. Hamburg, Bd. 8, 1884. 7. Steenstrup, J. — " Hectocotyldannelsen, itc." K. Danske Vid. Selsk. Skriv., Rk. 5, Bd. 4, 1856. Transl., Ann. Mag. Nat. Hist. ser. 2, vol. xx., 1857. 8. Vkrany, J. B. — Cephalopodes de la Mediterranee. Giines, 1851. 9. Verrill, A. E.— "Ccphalopods of the N.E. Coast of America." Trans. Conn. Acad. vol. v., 1880* and vol. vi., 1884. lo. Weiss, E. — " On some Oigopsid Cuttle-fishes." Quart. Journ. Micr. Sci. vol. xxix., 1888. REFERENCE LETTERS. a. Arm. loc. Striated region of loculi. ap.f. Apical fin. lolf.f. Left olfactory fold. ap.prol. Apical prolongation of the mantle. m. jNIembrane. ap.s. Apical set of suckers. mo. Mouth. b.m. Buccal membrane. m.s. Modified sucker. b.t. Basal tooth. nit. Mantle. c.app. Connective apparatus. olf.p. Olfactory papilla. ch.m. Chitiuous margin. ph.org. Phosi)horescent organ. d.a. Dorsal arm. pp. Pigmented patch. e. Eye. r.s. Rudimentary sucker. e.l Eyelid. s. Sucker. /• Fin. si. Siphon. /•/• Lateral membrane developed into a fiap. so. Socket. h. Hook. t. Tentacular arm. hd. Head. t.c. Triangular cap. i.e. Inner cone. t.jl. Triangular flap. i. 1.171. Inner lateral membrane. V. Valves. k. Keel. v.a. Ventral (4th) arm. I.m. Lateral membrane. v.h.a. Ventral hectocotvlized arm 22 ME. E. S. GOODEICH ON A COLLECTION OF EXPLANATION OF THE PLATES. Plate 1. Inioteuthis maculosa, n. sp. Fig. 1. View of lower surface, nat. size. 2. Enlarged view of the horny ring of a tentacular sucker. 3. Enlarged side view of a tentacular sucker. Sepia singalensis, n. sp. Fig. 4. Outline of the mantle and fins, and figure of the pen (inner surface), nat. size. 5. Enlarged view of a spermatophore. 6. Upper {b) and lower (a) mandibles, nat. size. 7. Enlarged view of a portion of the horny ring of a small tentacular sucker. 8. Horny ring of a small sucker of an arm. Sepioteuthis indica, n. sp. Fig. 9. Outline of the mantle and fins, reduced. 10. Figure of the pen of another specimen, reduced. IL Enlarged view of a portion of the horny ring of an apical sucker of the tentacular club. 12. Enlarged view of the apex of the tentacular club. 13. Side view of the head, showing the olfactory fold, nat. size. 14. Upper (6) and lower (a) mandibles, nat. size. 15. Horny ring of a sucker of the buccal membrane, enlarged. 16. Portion of the left 4th arm of a male, enlarged. 17. Edge of the horny ring of a sucker of the 3rd arm, enlarged. 18. Large tentacular sucker, x 4 diam. 19. Portion of the edge of the horny ring of tentacular sucker, enlarged. Plate 2. Loligo indica, Pfeffer. Fig. 20. Edge of the horny ring of an arm-sucker of a male, enlarged. 21. Horny ring of a sucker of the buccal membrane, enlarged. 22. Lower view of the left 4th arm of a male, showing the hectocotylization, nat. size. 23. Portion of the hectocotylized region of the same, inner view enlarged. 24. Horny ring of a small tentacular sucker, enlarged. 25. Enlarged view of a portion of the horny ring of an apical sucker of the tentacular club. 26. Horny edge of an arm-sucker of a female, enlarged. 27. Enlarged view of a portion of a large tentacular sucker of the same. 28. Enlarged view of a spermatophore. CEPHALOPODA FROM THE CALCUTTA MUSEUM, 23 Loliolus Investigatoris, u. sp. Fig. 29. Upper view of a male, uat. size. 30. Outline lower view of the mantle and fins, nat. size. 31. Upper view of the pen, nat. size. 32. Lower view of the hectocotylized 4th left arm, uat. size. 33. Enlarged inner view of a portion of the same. 34. Enlarged views of the upper [b) and lower [a) mandibles. 35. Horny ring of a tentacular sucker, enlarged. 36. Side-view of an arm-sucker, enlarged. 37. Edge of the horny ring of the same, enlarged. Abralia andainanica, n. sp. Fig. 38. Lower view, nat. size. 39. Outline upper view of mantle and fins, nat. size. 40. Enlarged view of the tentacular club. 41. Upper («) and side (6) views of the pen of a smaller specimeUj nat. size. 42 & 43. Front and side views of an arm-sucker, enlarged. 44&45. Front and side views of a tentacular suckei', enlarged. Plate 3. Abralia lineata, n. sp. Fig. 46. Upper view, nat. size. 47. Lower view, enlarged. 48. Enlarged view of the tentacular club. 49. Horny ring of an arm-sucker, enlarged. 50. Horny ring of a tentacular sucker, enlarged. Cheiroteuthis macrosoma, n. sp. Fig. 51. Lower view, nat. size. 52. LTpper («) and side (6) views of pen, nat. size. 53. Diagram of a section of pen across anterior region (6) and cone («), enlarged. 54. Upper (6) and lower [a] mandibles, nat. size. 55. Sucker of the ventral arm, enlarged. 56. Socket («) at the base of the siphon, and cushion on the mantle {b), uat. size. 57. Teeth of the radula, enlarged. Plate 4. Cheiroteuthis pellucida , n. sp. Fig. 58. Lower view, nat. size. 59. Socket at the base of the siphon, enlarged. 60. Sucker of the arm, enlarged. 61. Tentacular sucker, enlarged. 24 ■ ON CEPHALOPODA FEOM THE CALCUTTA MUSEUM. Histiopsis Hoylei, n. sp. Fig. 62. Lower view, nat. size. 63. Enlarged view o£ the mouth, buccal membrane, and the base of the arms. 64. Enlarged view of the right side o£ the head. 65. Enlarged view of the tentacular club. 66. Upper view in outline of the mantle and fins, nat. size. 67. Enlarged side view of a large tentacular sucker. 68. Enlarged left-side view of the funnel and mantle-edge. 69 & 70. Enlarged side and front views of an arm-sucker. 71. Upper view of the pen, nat. size. Plate 5. Taonius abyssicola, n. sp. Fig. 72. Lower view, nat. size. 73. Inner view of a portion of the dorsal arm, enlarged. 74. Enlarged view of the base of the head and siphon and mantle slit up along the median line and turned aside. 75. Funnel-organ, lower view, removed from its place on the wall of the siphon, indicated by a dotted line in fig. 74, enlarged. 76. Enlarged view of the olfactory papilla. 77. Side view of an arm-sucker, enlarged. 78. Horny ring of the same, enlarged. 79. Enlarged view of a portion of the horny ring of a small sucker at the base of the tentacular club. 80. Front (a) and side (6) views of a tentacular sucker. Octopus globosus, Appellof. Fig. 81. Enlarged view of the extremity of the hectocotylized arm. Octopm pictus (Brock), vax.fasciatus, Hoyle. Fig. 82. Enlarged view of the extremity of the hectocotylized arm. Octopus microphthalmus, n. sp. Fig. 83. Lower view, nat. size. 84. Lower (a) and upper (6) mandibles, enlarged. I Goodricl'i Tr,ajnis Linn . Soc. Zool . Ser,2 Yol . VII Pl 1 .^ 1 -. - ■«a |l " ''■■*.,^-.--' ^■| />'''^. 5 ,^-^'-.-^'?3??V 0.9 30 U i 28 36 Jr»,? 66 . 'C7 E.SGoodricK del . ATl.Hamraonol lith . "^ af-p, rol. ph.Oiu. 47 CEPHALOPODA FROM CALCUTTA MUSEUM Hanhart iinp Gflodru'l! Iran,; : -'ht ;..i''oiw/! ' ■ '" -■ - ;> f. \. i#^^^ ^. -^^:-^ 1?^ ' ' ^ K 1 ^r \ X y -J- / '•■-■e'/ tes^, / I fi ,§ ,# f^ 4p^' /'% %1 r GEPHALOPOSA FROIJ ;,V,"l,0U7TA MUSEUlvr Tp^-^WS LlI'.'N 8oO.ZoOT,.o'KR 2\0I.^ll Pl 74- \ 't /i # ¥ •'^^ y ^v 4 ^ fe.: \^y (-1 rm S^i.,. 'Yl / A.H.Hanin\onrl_ liti-i , iianhd-pi liii}. CE PH AL 0 P On/^. F ROM CA POUT T A T,4U ST. UTs I LINNEAN SOCIETY OF LONDON. MEMORANDA CONCEKNING THANSACTIONS. The First Series of the Transactions, containing both Botanical and Zoological contributions, has been completed in 30 Vols., and a few entire sets are still for sale. Only certain single volumes, or parts to complete sets, may be obtained at the original prices. The price of the Index to Vols. l-2.'5 is 8s. to the public, and 6s. to Fellows; to Vols. 26-30, 4s. to the public, and '3s. to Fellows. The Second Series of the Transactions is divided into Zoological and Botanical sections. The prices of the Zoological parts of these which have been published are as undermentioned (for tlie Botanical parts see Botanical wrapper) : — When Price to the Voli.n.e. Published. Public. £ s. d. Second Series. — Zoology. I. Part I. 1875 1 4 0 . Part II. 1875 0 6 0 . Part III. 1876 1 8 0 . Part IV. 1877 0 16 0 . Part V. 1877 0 18 0 . Part VI. 1877 1 2 0 . Part VII. 1878 110 0 . Part VIII. lS7t) 1 0 0 . II. Part I. 1870 1 4 0 . Part II. 1881 0 15 0 . Part III. 1882 1 8 0 . Part IV. 1882 0 7 6 . Part V. 1882 0 3 0 . Part VI. 18S3 1 0 0 . Part VII. 1883 0 5 0 . Part VIII. 1883 0 3 0 . Part IX. 1883 0 3 0 . Part X. 1884 0 4 G . Part XI. 1«84 0 10 0 . Part XII. 1885 0 6 0 . Part XIII. 1884 0 6 0 . Part XIV. 1885 0 6 0 . Part XV. 1885 0 4 6 . Part XVI. 1885 0 5 0 . Part XVII. 1886 0 3 0 . Part XVIII. 1888 0 2 6 . III. Part I. 1884 1 14 0 . Price to Fellowa. £ s. d. 0 18 0 0 4 6 ] 1 0 0 12 0 0 13 6 0 16 6 1 7 0 0 15 0 0 18 0 0 11 6 1 1 0 0 5 6 0 2 3 0 15 0 0 3 9 0 2 3 0 2 3 0 3 6 0 7 6 0 4 6 0 4 6 0 4 6 0 3 6 0 3 9 0 2 3 0 2 0 16 6 When Price to the volnnie. t> 1 1- i i t» ui- Publislied. Public. £ s. d. Second Series. — Zoologt {(onllnucd). III. Part II. 1884 112 0 . . . . Part III. 1885 1 10 0 Part IV. 1885 0 8 0 . . . . Part V. 1887 0 8 0 Part YI. 1888 0 6 0 IV. Part I. 18SC 1 4 0 . . . . Part II. 1887 I 8 0 Part III. 1888 0 16 0 . . . . V. Part I. 1888 0 12 0 . . . . Part II. 1888 0 5 0 Part III. 1889 1 7 0 . . . . Part IV. 1890 0 12 0 .... Part V. 1890 0 6 0 Part VI. 1891 0 12 0 .. .. Part VII. 1891 0 6 0 .... Part VIII. 1892 0 8 0 . . . . Part IX. 1892 0 12 0 . . . . Part X. 1893 1 8 0 . . . . Part XI. 1894 0 2 6 . , . . VI. Part I. 1894 2 0 0 .... Part II. 1894 1 11 0 .. .. Part III. 1894 0 10 0 Part IV. 1896 1 4 0 .... Part V. 1896 0 10 0 . . . . Part VI. 1896 0 8 0 . . . . Part VII. 1896. . . . 0 12 0 . . . . VII. Part I. 1896 (I 10 (.1 Price to Fellows. 1 4 0 1 2 6 0 6 0 0 6 0 0 4 6 0 18 0 1 1 0 0 12 0 0 9 0 0 3 9 1 0 0 0 9 0 0 4 6 0 9 0 0 4 6 0 6 0 0 9 0 1 1 0 0 2 0 1 10 0 1 3 3 0 7 6 0 18 0 0 7 6 0 6 0 0 9 0 0 / 6 2nd Ser. ZOOLOGY.] [VOL. VII. PART 2. THE (c^f/l. ri^ IRANSACTIONS 16i/7 OF THE LINNEAN SOCIETY OF LONDON, AMPHIPODA FE(.)M THE COPENHAGEN MUSEUM AND OTHER SOURCES BY The Rev. THOMAS R. R. STEERING, M.A., F.R.S., F.L.S. LONDON PRINTED FOR THE LINNEAN SOCIETY BY TAYLOR AND FRANCIS, RED LION COURT. FLEET STREET. SOLD AT THE SOCIETY'S APARTMENTS, BURLINGTON-HOUSE, PICCADILLY, W., AND BY LONGMANS. GREEN, AND CO., PATERNOSTER-ROW. May 1897. [ 25 II. Amphipoda from the Copenhagen Mmeiuii and other Sources. By the Rev. Thomas R. R. Stebbing, M.A., F.H.S., F.L.S (Plates 6-14.) Head 19th November, 1896. Introductom Remarks. The Zoological Museum at Copenhagen is rich in Amphipoda. It is rich also in living authorities on this group of Crustaceans, since Inspektor Dr. Meinert and Professor Liitken are two of its Directors, and Dr. H. J. Hansen is on the staff. This might well seem a happy concurrence of a fine collection in the hands of those well qualified to make its value known to the world. But the masters in science find their work continually expanding, while time remains remorselessly inelastic. Hence it is that these gentlemen, being themselves beset by other duties, have passed over to me the task of reporting on the Amphipoda of the Danish University. In this first memoir on the subject some of the more striking rarities are described, together with one or two of a less uncommon type. As the collection is not local but cosmopolitan, the opportunity has been taken of bringing mto notice certain other new or insufficiently known forms, in addition to those received from Denmark. For some of these I am indebted to Professor W. A. Ilaswell, D.Sc, of Sydney, New South Wales, and G. M. Thomson, Esq., E.L.S., of Dunedin, New Zealand. A specimen, which in the Report on the ' Challenger ' Amphipoda was unavoidably presented without adequate ceremony, and which in conseqiience subsequently became the subject of misunderstanding, is now set forth with due illustration, and a specimen from the Clyde, some years ago described without figures, now in like manner makes a more formal appeal for acceptance as the representative of a valid species. The range of the various specimens described extends from the shore to the deep sea, from Cuba to Ceylon, from the North Atlantic to the South Pacific, from the western coast of Scotland to the eastern coasts of Australia and New Zealand. Nine genera and ten species are discussed. Six of each are new. The species afford an illustration of two difficulties which not unfrequently arise in systematic natural history. Some of them are so like their jireviously known neighbours that a short-sighted jierson might think them not worth distinguishing. Others stand oddly apart, with so queer a combi- nation of characters that more than one existing family must look at them askance, imwilling to embrace, reluctant to repulse, in truth not very well able either " to do with them or without them." Opinions will differ on the policy of promptly establishing new families for eccentric forms, or of postponing that responsibility to as late a date as SECOND SERIES. — ZOOLOGY, VOL. VII. i 26 EEV. T. E. E. STEBBING ON AMPHIPODA FEOM possible. In the amiable endeavour to oblige the partisans of either view, I offer tentatively a new family for one of these perplexing species, boldly assign one to an old family, and leave one for the present homeless. Fam. ORCHESTIID^. Parhtale, n. g. First cmtennce longer than the peduncle of the second. First maxillm with the palp one-jointed, not extending beyond the distal margin of the outer plate. Max'illipeds with the palp four-jointed. Both pairs of gnathopods subchelate, differing in the two sexes. The third uropods carrying a minute inner ramus. Telson bipartite. The character of the Orchestiidse has to be slightly modified for the reception of this genus, since in it the third uropods are not absolutely one-branched, but show a sradation towards the more normal biramous condition. Parhyale fascigek, n. sp. (Plate 6.) Specimens in spirit are slightly iridescent. The skin has some minute setules scattered over it. The second and third pleon-segments arc squared at the postero- lateral angles. The fourth segment almost reaches to the telson, overlapping the very short fifth segment, whUe the sixth is dorsally obsolete. The telson is divided to the base, its two oblong or sometimes somewhat triangular leaves standing nearly upright. Eyes oval, dark, obliquely set near the top and fi'ont of the liead ; ocelli numerous. First antennce. The second and third joints together approximately equal in length to the much stouter first joint ; the flagellum of ten or eleven joints is longer than the peduncle. Second antennce. The last two joints of tlie peduncle nearly equal in length ; the flagellum not once and a half as long as the peduncle, with about twenty joints in the male and fourteen to sixteen in the feniale. Upper lip. The apical margin is rounded and furred in the usual manner. Mandibles. The cvitting-edge has six teeth. The secondary plate on the left mandible is strong and quinquedentate, on the right it is slighter, with coml>like, slightly concave edge ; the spine-row on the left attains the number of six spines, which have the appearance of plumose sette ; the molar tubercle is strong, prominent, finely denti- culate, with a hairy tuft at one corner, and a long feathered seta at the other. Above it and a little in advance is the articulating process. Lower Up. The lateral processes are short. First maxillce. The inner plate is slender, surmounted by two feathered setse, of Avhich the inner is the shorter. The outer plate is rather liroad, surmounted by nine spines THE COPENHAGEN MUSEUM AND OTHEE SOUECES. 27 variously denticulate, in two rows, which are set so closely together that it is difficult to count and discriminate the spines. The palp is slender, tipped with a minute hair and small seta. There is often a slight constriction of the margins near the apex, as though a second joint were being thought of. Second maxillcc. The setae round the apex of the outer plate are longer than those of the inner. The fringe on the inner plate is bounded by a seta conspicuously longer than the rest. MaxilUpeds. The inner plates have the usual setse and three apical teeth. The outer plates reach but little beyond the first joint of the palp, and have their spinules set a little Avithin the margin. Both the second and third joints of the palp are broadly lobed at the inner distal extremity. The third joint has a little rounded and shining process for the aj^ex of its outer margin. On the inner surface there is a graduated row of spines, and in the male a dense tuft of setae on the outer surface, from among which the sharjD curved finger projects. The inner margin of the finger carries some spinules or small hairs. First (jnuthopods. The side-plates are widened below, and have the front margin straighter than the hinder one. The fifth joint is distally widened, and has the projecting hind margin fringed with setae. The hand or sixth joint is as long as the preceding, and in the male fully as wide, but narrower in the female. The hind margin has a central fringe of setix', bounded by a slight contraction in the width of the hand. The palm is rather oblique, fringed with little hairs and setules, and carrying two unequal but stoutish spines at the extremity, against which the rather broad finger impinges. There are spines on the inner surface of botli hand and wrist, and minute hairs on the inner margin of the finger. Second gnathopods. The side-plates are oblong, with rounded corners below, but with the hind margin somewhat excavate above. The branchial vesicles of these limbs and the four following pairs have at the base a small vesicle accompanying the ordinary large one. The marsupial plates of the female are long, distally acute, and, at least in the distal part, closely fringed. As in the first pair, the second joint widens distally, the third is short, the fourth is of nearly uniform breadtli, Imt having the hind margin longer than the front. The wrist in the male is almost triangular, with convex front border, and behind masked by the fourth joint, whereas in the female it protrudes a spine- fringed lobe between that joint and the hand. In the male the band is massive, with very oblique palm, uniformly convex till near the distal end, bordered Avith nuniei'ous little spines, over Avliich the long and broad finger closes tightly, sending its point into a little pocket on the inner surface of the hand, from which arises a small group of stoutish spines. The hind margin is almost smooth, but has a small indent similar to that in the first pair. In the female the hand is less massive, but still much larger than the Avrist, Avith a very oblique palm, leaving a shorter hind margin, Avhich is fringed with spines. In both sexes the finger has little hairs set at right angles to the inner margin. First peraopods. The side-plates are a little broader than the preceding pair, but similarly shaped. The fourth joint is longer and much broader than the fifth or sixth. 4* 28 EEV. T. R. E. STEBBING ON AMPHIPODA FEOM The finger is short, curved, with a spinule from its inner margin cnrviug towards the small but decided nail. The fingers are similarly shaped in all the perteopods. Second perc^opods. The side-plates are broader than the preceding pairs, with the hind margin doubly excavate above. The limbs closely resemble the preceding pair. Third perceopods. The side-plates are broad, bilobed, not very deep. The second joint has the breadth and depth about equal, with the margins very slightly indented. The fourth joint is much broader, but not longer than the sixth, which is narrower but a little longer than the fifth. The fourth joint has spines on both margins, the apical groups l)eing large. The same applies to the fifth joint, except that its hind margin is only armed at the apex. The sixth joint has spines along the front. Fourth perceopods. These are similar to the preceding, but with the joints from the second to the sixth decidedly larger. Fifth 2JercBopods. Tlie side-plates are not bilobed. The limbs are like the jireceding pair, but larger. Fleopods. There are two, or in some cases three, small coupling-spines. The spines of the inner margin of the first joint of the inner ramus have a distal thickening, but no cleft could be distinguished. The rami appear to be as nearly as possible eqvial in length, with about thirteen joints apiece, each carrying the usual pair of plumose setae. TJropods. The first pair are the longest, with peduncle longer than the rami, having spines on its iipper margin and a large spine at the apex. The rami are of equal length, with one large and other small spines at the apex, the inner having two in addition on the upper margin. The second pair are much shorter, but similarly armed. The third pair are short, the peduncle not being so long as the telson. It carries some apical spines. So does the outer ramus, Avhich is a little shorter. The inner ramus is conical, almost rudimentary, tipped with a minute seta. Lenyth. In the partially-folded position which seems natural to animals of this genus, the length of the specimens averages a quarter of an inch. ILah. Harbour, Antigua, West Indies (specimens received from W. E. Forrest, Esq.). Specimens from the Copenhagen Museum were labelled " St. Thomas, Havnen, 1877. Eggersr The specific name refers to the appearance presented by the extremity of the maxil- lipeds, whicli in the male is very striking. Fam. LYSIANASSID^. Lysianax cinghalensis, n. sp. (Plate 7 A.) First three segments of plcon large, postero-lateral angles of third segment rounded. The etjes large and black, occupying most of the svu-face of the head, on the top of which they meet. First antennce. First joint tumid, scarcely longer than deep, nearly twice as long as the second and third combined ; flagellum six-jointed, shorter than first joint of peduncle ; accessory flagellum three-jointed, less than half the length of the primary. THE COPES^HAGEN MUSEUM AND OTHEK SOUECES. 29 Second antennm. In the male about two-thirds the length of the animal, the ultimate joint of the peduncle considerably longer than the penultimate, flagellum of about thirty- five slender joints. Mouth-organs closely agreeing with the generic account given by Sars in ' The Crustacea of Norway.' The slender mandibular palp has the first joint characteristically elongate, and there are but few spines on the paljj of the first maxillae. First gnathopods. The side-plates widened below, so that they are as wide as deep ; the second joint as long as the four succeeding joints combined, the third larger than the fourth, the fifth much wider but rather shorter than tlie narrow, taj)ering sixth ; the finger curved, small. Second gnathopods. The second and third joints slender and elongate, the fifth as long as the third, the sixth more than half the length of the fifth, gradually widening distally, with short transverse palm, on which the finger closes tightly, without overlapping it. First and second permopods. The side-plates of the second pair are deeply and rather widely excavate behind. The second joint of the limbs comparatively short ; the fourth, fifth, and sixth joints rather long, with some plumose setae. Third percBopods. The side-plates wider than deep, pretty evenly bilobed ; the second joint as broad as long, smaller than the side-plates ; the fourth joint rather broad, produced to an acute point. Fourth and fifth perceopods. The second joint broader and much longer than in the preceding paii", roundly produced behind. Branchial vesicles strongly pleated. TJropjods slender. First pair with the rami subequal, more than three-fourths as long as the peduncle, each with three marginal spines. Second pair, with subequal rami, almost as long as the pedimcle ; the outer ramus with four marginal spines, the inner constricted near the end, with a spine in the notch. Third pair with the peduncle longer than the rami, produced to a point at the outer angle ; the inner ramus a little shorter than the outer, both with long setae on the inner margin. Telson oval, not reaching far along the peduncle of the third uropods, having a feathered setule on each lateral margin, and the apex smootlily rounded. Length three-tenths of an inch. Labelled " Trincomali (Ceylon), 3/89, surface. K. Fristedt," Copenhagen Museum. Lysianax cubensis, u. sp. (Plate 7 B.) Lateral angles of the head considerably produced, broadly rounded. Postero-lateral angles of third pleon-segment quadrately rounded. FJi/es reniform, moderately large. First antenncB. First joint large, longer than deep, not very much longer than second and third combined, neither of tliese being extremely short; flagellum six-jointed, shorter than the peduncle ; accessory flagellum three-jointed, less than half the length of the primary. Second avtennce. Last two joints of the peduncle (in the female) subequal ; flagellum broken. 30 EEV. T. R. R. STEBBING ON AMPHIPODA FROM Mouth-organs of the character usual in the genus, but palp of first maxilla with the apex rather Avide, carrying one little spine and six very distinct teeth. First QRcdhopods. Side-plates widened below, deeper than Avide. The second joint shorter than the four succeeding joints combined, the fifth rather stouter than the tapering sixth and quite as long ; finger small. Second gnathopods. The second and third joints rather long and slender, the fifth as long as the third, the sixth much more than half as long as the fifth, at first gradually but distally rather abruptly widened ; palm transverse, finger very small. First and second prrtropods. Side-plates of the second pair broad, the excavation being moderate in breadth and deptli. The joints of the limbs of moderate dimensions, fringed with smooth setse or spines. Third perccopods. Side-j^lates much wider than deep, the front lobe deeper than the hinder. Second joint much wider than deep, almost as large as the side-plate, the fourth joint little widened. Fourth per mopods. Second joint longer than broad, front margin strongly bowed, hind margin sinuous, fourtli joint scarcely widened. Fifth perceopods. Second joint much larger than in the preceding pairs, very broad, but rather longer than broad ; fourth joint scarcely widened. Marsupial plates slender, branchial vesicles rather strongly pleated. TJropods. First pair Avith the rami subequal, slender, carrying a few margmal spines, somewhat shorter than the peduncle. Second pair shoi-ter than the first ; peduncle scarcely as long as outer ramus, which lias four marginal spines ; inner ramiis slightly shorter, stout, with two spines, then strongly constricted with a spine in the notch. Third pair much shorter than second ; peduncle rather longer than the rami, with one lateral margin running out into a point, the other notched near the middle, the rami almost smooth, eqiial. Tel son reaching little beyond the sides of the sixth pleon segment, little longer than broad, boat-shaped, apex truncate. Colour of specimen, in spirit, brownish. Length about three-tenths of an inch. Specimen labelled " Zi/sianasscc aff". Paa en Reise til Cuba." Copenhagen Museum, Studer Collection. Fam. STEGOCEPHALIDiE. AXDANIOTES, n. g. Head wdth small rostrum. First segment of peragon the longest. Sixth segment of pleon longer than the fourth or fifth. Each of the first three side-plates socketed in a groove on the underside of that next succeeding ; the fourth very large, subequal in breadth and depth, with a wide shallow excavation behind, overlapping the fifth and in part the sixth. Eyes wanting. First antennse with very thick first joint ; flagellum tfipering, the first joint A'cry long, longer than the remaining three combined ; the accessory flagellum small, one-jointed, tipped Avith a long spine. Second antennae with the last joint of the peduncle shorter than the penultimate, flagellum short. Epistoma carinate. U2:)per THE COPENHAGEN MUSEUM AND OTHER SOUECES. 31 lip broader tlian deep, slightly bilobed. Lower lij) with the front lobes wide apart, not larger than the hind lobes, and snrmonnted with a single spinule at the outer corner. Mandililes a slightly sinnous oblong in shape, with straight smooth cutting-edge ; a tooth-like accessory plate on tbe left * mandible. First maxilla? with inner plate rather large, bordered with 7-11 long plumose setae ; outer plate moderately broad, carrying nine slender spines and a tuft of setules ; the palp one-jointed, nearly reaching the end of the outer plate, tipped with seven setae or slender spines. Second maxillae with inner plate very broad, with a fringe of 18 or 19 setae, most of them plumose ; the outer jilate shorter and greatly narrower, tipped with nine setae. Maxillipeds with inner plates very broad, the apical margin sloping outward, with three little spine-teeth; the outer plates not nearly reaching the end of the palp's second joint, fringed with 13 spine-teeth on the inner margin ; the joints of the palp successively diminishing in size, the third and fourth slender and small. Appendages of the perteon nearly as in Andania. Uropods with the peduncles robust, much longer than the rami; both peduncles and rami successively diminishing. In the male, outer ranuis of first pair thick, curved, smooth, inner rather shorter, straight, slender, each with two marginal spinules ; rami of second pair slender, smooth, subequal ; rami of third pair minute, the outer nearly twice as long as the inner, with a tiny second joint. In the female all the rami slender ; those of the third pair- not minute, subequal, nearly as long as the peduncles. Telson broadly oval, but ending almost acutely, cleft nearly to the middle, but without any dehiscence, shorter than the peduncles of the tliird uropods. Andaniotes corptjlentus (Thomson). (Plate 8.) 1882. Anonyx corpulentus G. M. Thomson, Trans. New Zealand Institute, vol. xiv. p. 231, pi. 17. figs. 1 a-f. 1888. Andania abijssorum Stabbing ' Challenger ' Amphipoda, p. 739, pi. 37. 1893. Stegocephalus abyssorum Delia Valle, Gammaiini, p. 629, pi. 59. fig. 38. Head vdth lateral angles rounded and below them produced downward in long straight triangular lobes hidden under the first side-plates, below which the mouth-organs of the down-bent head are visible. Third pleon-segment with the postero-lateral angles narrowly rounded, the dorsal line distally bent abruptly downward, having (in the male) two little eminences below the bend, the distal margin somewhat squared ; the fourth segment dorsally convex ; the fifth depressed ; the sixth long, dorsally convex, forming two longitudinal eminences, between which the telson is concealed in a lateral view. First antenncp. The third joint of the pedimcle is transversely oval ; the first joint of the flagellum carries transparent filaments ; the setiform spine at the apex of the accessory flagellum is longer than the joint. Second antenme. The flagellum is eight-jointed. First (inathopods. The side-plates are roughly triangular. The second joint is long and broad, stronger in the male than in the female ; the third joint longer than broad ; the fifth joint in the male longer and much wider, in the female wider but only little * Not on the right, as stated in tlie ' Challenger " Arnphipoda, p. 739. 32 REV. T. R. E. STEBBING ON AMPHIPODA FROM longer thau the sixth, with many setse on both margins, those in front elongate ; the sixth joint abruptly narrows distally, so as to leave no palm, carrying long setse in front, more in the male than in the female, and a row of shorter ones at right angles to the hind margin ; the finger is small, curved. Second gnatlioimds. The second joint is slender, curved at the base, the third longer than the fourth, the fifth and sixth rather slender, subequal ; the curved finger impinging against the sinuous hind margin of the sixth joint, which is armed with plumose setae. Third jyerceopods. The second joint is scarcely dihxted ; the fourth is not very wide, acutely outdrawn below. Fourth i^erceopods. Second joint twice as broad as in the preceding pair, fourth joint also larger. Fifth perceopods. Second joint longer and much broader than in the preceding pair, subequal in length to the rest of the limb, broadly overlapping the third joint ; the fourth joint rather wide, but much shorter than in the preceding pairs. Pleopods. Coupling-spines rather long. Length nearly three-tenths of an inch. Hub. New Zealand. A male and a female specimen received from G. M. Thomson, Esq., F.L.S., the first describer of the species. A female specimen was also taken by the ' Challenger ' Expedition. Earn. PHOXOCEPHALID^. PONTHARPINIA, n. g. 1880. Urothoe Haswell, Proc. Linn. Soo. N. S. Wales, vol. iv. p. .^^o. 1891. Harpinia? Stebbing, Trans. Zool. Soo. London, vol. xiii. pt. 1, p. l. Back very broad, tapering to the rostrum and telson. Third pleon-segment setose on the lower hind margin. Eyes small, lateral, set rather far back. Eirst antennae with first joint tumid, second more slender, setose, third short ; accessory flagellum many- jointed, more than half as long as the primary. Second antennae with the basal joints of the peduncle short and stout; the penultimate long, broad, and setose; the ultimate much shorter and nai'rower, but also carrying long setae ; the flagellum shorter than the peduncle. Upper lip distally rounded, wide. Lower lip quadrilobate. Mandililes with cutting-] )late and left mandible's accessory plate small, divided into a few teeth ; spine-row of 8-10 spines, the first on the right mandible conspicuously the largest (perhaps representing the accessory plate) ; molar small, prominent, tipped with a few imeqtml spinules ; palp much longer than the trunk of the mandible, third joint rather shorter than second, Avith many long setae near the base and the apex. Eirst maxillae : inner plate small, carrying three short setae, outer with eleven spines, very slight in size and armature; second joint of palp with 4 spinules and 4 setules on truncate apex. Second maxillse : inner j)late shorter and narrower than the outer, its armature not occujiying more than half the inner margin. Maxillipeds with the plates narrow and not very long ; the outer scarcely reaching the middle of the long and broad second joint of the palp, fringed with half a score of graduated spines ; the fourth THE COPENHAGEN MUSEUM AND OTHER SOUECES. 33 joint of the palp slender and curved, considerably longer than the third. Gnathopods apparently varying in the two sexes. First and second j)erajopods with the fourth joint much longer than the fifth or sixth ; fifth much wider than the sixth, fringed with long spines at the hinder apex ; the finger small, spine-like, scarcely so large as a knife-like spine on the apex of the sixth joint. Third and fourth peneopods with the second, fourth, and fifth joints greatly widened ; fifth pair much smaller than the others, its second joint much expanded, Avidest distally, produced to the end of the fifth joint ; third joint wider than those folloAving, which are all narrow and small ; the finger straight, acute. First uropods with a long spine at apex of peduncle, rami suhequal, rather longer than peduncle, spinose ; second pair short, rami equal, one sjjinose ; third pair with lanceolate rami. Telson divided into two suboval plates. Marsupial plates of the female narrow. The generic name is intended to indicate the intermediate character of this genus between such forms as those in the genera Urothoe and Hailstorms of the Pontoporeiidte and Harpinia in the Phoxoeephalidre. By the character of the mandibles it is better placed in the latter family. PoNTHARPiNiA PiNGUis (Haswell). (Plate 9B.) 1880. UrotJwe jnnffuis Haswell, Proc. Linii. Soc. N. S. Wales, vol. iv. p. 3.25, pi. 19. fig. 2. 1882. „ „ „ Catalogue of the Australian Stalk- and Sessile-eyed Crustacea, p. 240. 1891. Harpinia t jnngiiis Stebbing, Trans. Zool. Soc. London, vol. xiii. pt. 1, p. 4. 1893. Urolfioe irrostruta ? Delia Valle, Gammarini, p. 667. Head long, tapering from a broad base. First side-plates tending to quadrate, but widened below ; fourth with rounded lower margin, widest just below the shallow exca- vation ; fifth broad, not deep, the binder lobe the deeper ; seventh very small. Third pleon- segment with a very small produced point at the lower hinder angles, above which the hind margin is finely serrulate. The fourth segment ajjpears dorsally to overlap the fifth and sixth. First antennce. Fhigellum consisting of 15 joints, accessory flagellum of 10 ; in a female specimen of 9 and 7 respectively. A peculiarity of these joints is that the apical margin being oblique, they overlap one another. This is shown in Haswell's figure of the primary flagellum, though he does not mention it in his text. Second antennce. Last joint of peduncle narrow at base, thence a little widening ; flagellum in female with 10 joints, shaped as in the first pair. Haswell's figure shows 21 joints. Tipper lip wider than deep. Lower lip with very small principal lobes. The mouth-organs in general are small, except the palps of the mandibles and maxillipeds. Tirst gnathopods. 2 . Second joint reaching much beyond the side-plate ; fifth nearly as long as second and rather wider, fusiform; sixth shorter and rather narrower, widening from a narrow base, then preserving an almost iiniform width to the transverse slightly convex palm, wliich ends in a short tooth, tlie convex front margin of the widened part SECOND SEIIIES. — ZOOLOGT, VOL. VII. 5 Si EEV. T. K. K. STEBBING OX AMPHIPODA EKOM li'ing-ed with many seta-like spiues ; the finger closely fits the palm and projects a very little beyond it. Second (jnathopods. S . Almost exactly like the first, but the sixth joint rather longer and more slender, -with a smaller palmar tooth, and tlie finger not extending beyond the palm. The marsupial plate is as long as the second joint but much narrower, while the branchial vesicle is considerably longer and much wider ; the side-plate is oblong, m ith the lower margin slightly convex, and like the side-plates in general partly fringed Avitli setoe. Gnathopods. 6 . The species now being described is assigned to Haswell's Urothoci pinguis on the supposition that the account gi\en by Haswell refers to the male, and that in that sex the gnathopods are strikingly different from those of the female. Haswell's description says : — " Gnathopoda large ; anterior pair Avitli the carpus three- fourths of tlie length of the propodos ; the propodos ovoid, swollen ; the palm not defined ; the dactylos half as long as the propodos ; posterior pair longer than the anterior ; carpus small, subtrian<];ular ; ]n'opodos ovate, palm defined by a prominent angle ; dactylos rather more than half as long as the propodos." Third pera'Ojjods. Second joint expanded, longer than broad, front margin sinuous, serrulate, carrying slender spines and setoe, hind margiu smooth ; fourth joint broader than long, with a long free upper margin armed with slender spines, hind margiu cut into 5 teeth, this and the lower margin spinose; fifth joint narrower than fourth, but still very broad, breadth and length equal, fringed with many spines and setai ; sixth joint not longer, and less than half as broad ; finger quite small. i^o^-;-;'/^ j>er^q2;o(/* similar to third, but larger ; second joint broader in proportion to length; fourth very large and setose, with 7 teetli on hind margin, of Avhich the lowest two are formed by deep incisions ; fifth joint rather longer than broad, about twice as broad as the sixth. Fiftliperceopods. The large expansion of the second joint has a smooth front margin, the hind one serrate, the lower sei-rulate and fringed with sette ; the fifth joint is apically finely pectinate; the finger is more than half the length of the sixth joint. Pleopods. The peduncles are short; the rami with about 21 ratlier broad joints. (Jropods. The rami of the first pair are slightly curved, each with a short stout nail ; the rami of the second pair are straight ; those of the third pair (at least in the male) lanceolate, long, plumose. Telson with length and breadth equal, the apices rather divergent, each with a spinule in a small emargination, the outer margins (in the male) setose. Length of male yj inch, of female -^^ inch. " Bondi, New South Wales, cast on the beach during a storm," is the source assigned for the specimen described by Professor Haswell ; the specimen here described was contained in a gathering from Jervis Bay, kindly forwarded to me by the same author. THE COPENHAGEN MUSEUM AND OTHEE SOURCES. 35 Fam. LEUC0TH0IDJ3. Leucothoe INCISA, Robertson. (Plato 10.) 1888. Leucotho'e farina Chcvreux (not Savigny), Bulletin do la Socirte cVetiulcs scientifiqucs de Paris, W annee, U"' sem., Extr. p. 9. 1892. Leucothoe incisa Robertson, Transactions of the Natural History Society of Glasgow, vol. iii. p. 217 (sep. copy, p. 23). The drawings of this species were made many years ago, hut have waited till now for a suitable opportunity for publication. The species of Leucotlioe axe se^iarated by rather fine distinctions. In 18G0 Boeck named a new species LeucofJioe LiUjehorgii, hut in his later works he was willing to let this lapse as a synonym of the long known Leiicolhoii spinicarpiis (Abildgaard). In 18S9 Norman gave the name imparicornis to a form from Shetland, which Sars in 1892 determines to be the same as 'Boeck's I/Hljcborffli,, the latter name having priority. To tliis species, as figured and described by Sars, Leucothoe incisa makes a tolerably close approximation, yet it seems difficult to unite the two. In hicisa tlie fourth side-plates have the front angle rounded, not acute; the first gnathopods have the tip of the process of the fifth joint strongly hooked, the inner margin of the hand not quite smooth but faintly crenulate, and the finger not very small; in the second gnathopods the ^^alm is convex and faintly but broadly crenulate, the finger is not abruptly bent at the base ; the tclson, instead of being little longer than broad, is fully twice as long as broad, with the apex almost acute. As in Lllljehorgil, the inner margin of the wi'ist in the first gnathopods is serrate, and the postero-lateral angles of the third pleon-segment are sharply upturned, forming a sinus with the bulging hind margin. There is a tendencj^ to this in the preceding segment. The mandibles have the cutting- edge divided into about eight teeth ; the secondary plate on the left mandil)le is large, witli ten teeth, that on the right is much slighter (Sai's denies its existence in this genus) ; the second joint of the palp is not much longer tlian the narro'^v^er third. In tli(> maxil- lipeds the first and second joints of the palp have the length and breadth subequal. Length not quite three-tenths of an inch. Taken off Cumbrae, in the Clyde, at low water and also in 20 fathoms, by Dr. David Robertson, LL.D., F.L.S. Anamixis, n. g. Head hood-like. Eirst side-plates small, three following pairs very large. Eyes well developed. First antennae attached below tlie apex of the head, witli elongate peduncle and no secondary flagellum. Second antennae remote from the first, shorter and thinner, Avith small flagellum. Mouth-organs (at least in the adult) degraded and abnormal. Maxillipeds Avith the full number of joints, the third simple. First gnathopods delicately chelate ; second massive, complexly subchelate ; the other limbs slight, normal. Branchial vesicles small, simple. Pleopods small. First and second uropods biramous, third at present unknoAvn. Telson simple. The shape of the head calls to mind the genus Bnlichia, the situation of the antennae Plat ijiscUn opus, the size of the side-plates Metopa, themassiveness of the second gnatho- pods Jlicrojjrotopus, the structure of both pairs of gnathopods and of the antenna) and 36 EEV. T. E. E. STEBBING ON AMPHIPODA FEOM maxillijieds Lencotho'e. In reference to this combination of characters tlie name of the genus has been chosen, from the Greek word ai'a^tgi?, mingling. Among existing families it stands nearest to the Leucothoidse, l)ut the extraordinary nature of the mouth-organs — whether due to a parasitic mode of life or to a marital stage of existence, or falling under any other exiilauation — so far isolates the species now to be described that it may have to be placed alone in a separate family, Anamixidae, which would for the present bear the characters of the genus. Anamixis Hanseni, n. sp. (Plate 11 ) The head is longer than any one of the perseon-segments, and is narrowed distally, Avith rounded apex, with no distinct lateral angles. The second pleon-segment has the postero- lateral angles acute, slightly produced ; those of the third are a little blunted. The fifth segment is very short, scarcely distinguishable from the fourtli ; the sixth projects on either side of the telson, in a way to suggest that the missing third pair of uropods may be of a fairly large size. The eyes are round, placed in the middle of the sides of the head, consisting of about eighteen short ocelli. First antenme. The first joint long and rather stout, the second nearly three-quarters as long as the first, but much more slender, the third about two-fifths of the second ; the flagellum eleven-jointed, a little longer than the first joint of the peduncle. A sensory filament is present on several of the joints. These antennae dejiend from the head, being inserted just below its apex. Second antenntB. These are inserted at the other extremity of the head. The first two joints arc very small, the third is little more than half as long as the first of the upper antennae, the fourth is slightly longer than the second in the other pair, and the lifth is a little longer than the third joint of its own pair. The slender flagellum is four- iointed, less than half as long as the penultimate joint of the peduncle. Mouth-organs. The underside of the head is slightly carinate, and apparently attached to the keel there projects from between the second antennae a vertical plate, which may be called the oral lamina. Its truncate front edge has some minute microscopic teeth. With this curious and abnormal exception, no mouth-parts could be detected other than the pediform maxillipeds. Maxillqjcds. The second joint bulges slightly on the upper or inner side, while on the outer or lower side it has a shallow cleft, between two smoothly rounded apices, which just overlap the bases of tlie third joint. The third joint has no lobe or lamina, but in appearance is the basal joint of a five-jointed palp. Of the actual palp the first and second joints arc about equal in length, tlie second the wider, carrying some setse on the lower margin, the third joint is longer than either of the preceding, armed with several setue, and having its surface minutely furred. The fourth joint or finger is slender, strongly curved, nearly as long as the third joint. Owing to the absence of j)lates from the joints of the protopod the palps are in close contiguity. The appearance is rather that of legs than of mouth-organs. In dissection the head came easily away, leaving the maxillipeds very firmly attached to the first gnathopods. THE COPENHAGEN MUSEUM AND OTHER SOURCES. 37 First gnathopods. The side-plates ave sinall, triangular, with the apex to the rear. The second joint is slender, narrowest at tlie base. The small third joint is larger than the fourth The fifth joint is very much Avider and much longer than the second joint. It may he regarded rather as the hand tlian the wrist, its long and slender curved apical proces«i, tipped Avith a slightly curved needle-like spine, forming the immovable finger of the chela, while the sixth and seventh joints form the movable finger. The sixth joint is slender, i-ather straighter than the process of the fifth, but otherwise very like it, and tipped with a similar spine, which must be regarded as the seventh joint. It might be sujiposed that tlie fifth joint in this remarkable form rejiresented a coalescence of the wrist and liand, did not a com2)arison with the first gnathopods iu Leucotho'e make it reasonably certain that the chela is composed in the way just described. Second gnathopods. The side-plates are very large, rounded in front, produced beyond the first segment. The second joint is narrow, distally somewhat widened, a little curved. The third joint is of stouter build than usual, apparently articulating Avith all the three folloAving joints. The fourtli joint is small, most of it lying flatly on the inner surface of the fifth joint, with which it seems to be in coalescence. The fifth joint is subequal to the second, liut In-oadest at its base and apically pointed. The hand is broad and between three and four times as long as its breadth, the liind margin distally cut into three teeth. The finger, more than half the length of the hand, has a curved acute tip, and two slight projections on the inner margin. The complex clasjier is formed by its impinging against the apex of the Avrist and the denticulations of the hand. The peculiar arrangement of the third and fourth joints may be explained by the extraordinary massiveness of tliese limbs and especially of the hand, Avliich is in striking contrast to that of the preceding pair as well as to the general structure of the other limbs. First 2)erceopods. The side-plates are as deep as the preceding pair, but of much less Avidth. The branchial vesicles are all of remarkably small size. The second joint reaches a little beloAv the side-plate. Of the other joints the fourth is the widest, the fifth is sliorter than the sixtli, tin; nail is slender and curved. Second percBopods. The side-plates are rather Avider than the preceding paii-, with a faint emargination at the farther end of the loAver margin and a very shallow excaA^ation at the upper part of the hind margin. The limb differs but little from the preceding. Third ferceopods. The side-plates are bilobed, Avith the hinder lobe the larger. The second joint is not much longer than wide. The fourth joint has the hind margin slightly produced. The rest of the limb is missing. The spiny armature in these and the other limbs is insignificant. Fourth per ceopods. The side-plates are lobed behind. The second and fourth joints are rather larger than in the preceding and foUoAving pairs. Fifth perceopods. The side-plates are small, rounded behind. The second joint is rather more strongly spined along its front margin than is the case in the other limbs. Fleopods. The rami are small, five- or six-jointed, shorter than the peduncles. JJropods. The first pair are the longer. In both, the inner ramus is a little longer and the outer considerably shorter than the peduncle. Tlic third pair are missing. 38 KEV. T. R. E. STEBBING ON AMPHTPODA FEOM Telson. This is a little longer than broad. The broadly rounded apex does not reacli so far back as the projecting points of the preceding segment, which are in turn overlapped by the peduncles of the second uropods and these by the peduncles of the first pair. The length of one specimen, with tail folded in, was less than a tenth of aa inch, of the other rather more than a tenth. Hab. West Indies. From Goniastrcea varia Dana. The name is given in compliment to my friend, Dr. H. J. Hansen, the accomplished naturalist by whom it was obtained. Pam. PAEDALISCID^. Pardaliscoides Stebl)ing. 1888. Pardaliscoides Stebbing, 'Challenger' Amphipoda, p. 1725. 1893. Pardalisca Delia Valic, Gammarini, p. 691. Pirst antennae longer than second, second joint of the peduncle longer than the first, both flagella many-jointed. Mandibles with broad cutting-edge ; tliat on the left side having two blunt teeth above and two that are acute below, one of them small, the other large, a rather broad accessory plate with crenulate edge, and two plumose spines ; that on the right having a similar cutting-edge miuus the smaller acute tooth, no accessory plate, but three plumose spines ; the three-jointed palp slender, fringed with setiform spines. Pirst maxillte with one seta on the small inner plate, seven very unequal spines on the inner plate, one of them plumose; the second joint of the j^alj) broad, with many spinules about the apical margin. Second maxilbne with the plates slender, the outer carrying three, the inner seven sctse. Maxillipeds with inner plates vei-y small, outer plates short and narrow, fringed with spinules, the supporting joint not very large; j^alp with first joint short, second robust and long, but scarcely longer than the third, the finger long and slender with minute setules on the inner margin. Tlio tritiu-ating organs (anterior lateral gastric lobes) are armed with six long spines. The gnathopods are similar, simple, with the fifth joint robust, fusiform, the sixth and seventh slender, the seventh fringed with minute setules. The peraiopods are slender, elongate. The second uropods have the rami unequal, the third have the rami foliaceous. The telson is deeply cleft. Pardaliscoides tenelltjs Stebbing. (Plate 12.) 1888. Pardaliscoides teneUiis Stebbing 'Challenger' Amphipoda, p. 1~.'25. 1893. Pardalisca abtjssi Delia Valle, Gammaiini, p. G92. Head with acute rostrum. There appears to be a small dorsal denticle on each of the last two or three segments of the pleon. First antenncB. The first joint stout, the third half the length of the second ; of the flagellum thirteen joints remain, the first much the longest ; of the accessory flagellum seven or eight joints remain, the first as long as that of the primary. Second antennce. Last tAvo joints of the peduncle elongate, the last shorter than the preceding; the flagellum half the length of the peduncle, twelve-jointed. THE COPEXHAGEX MUSEUM AND OTHER SOUKCES. 39 Fh'st and second gnathopods. Tliere are numerous spines of vai'ious sizes on and near the margins. On the hind margin of the wrist and hand, commencing at the narrowed distal end of the wrist, is a series of short spines which are plumose. The setules on the inner margin of the fingers are exceedingly small. First and second perceopods. The fifth joint longer than the fourth or the sixth; the second pair rather longer than the first, and further distinguished by having the hind margin of the sixth joint fringed with about a dozen short blunt spines. Third perceopods. Second joint very slightly expanded, fourth joint the longest ; finger slender, acute. Fourth perceopods like the third, but with rather longer joints. Fifth perceopods considerably longer than the fourth, the second joint expanded above, narrowing dovrnward, fourth joint very long. Branchial vesicle small, narrowly oval. Telson much longer than broad, cleft thi-ee-fourths of the length, dehiscent, with a spinule at each apex and a setule on each lateral margin. Length not quite a third of an inch. Kah. South Pacific, lat. 37° 29' S., long. 83° 7' W. Taken by H.M.S. ' Challenger,' in tow-net at trawl, from a depth of 1775 fms. Fam. EUSIRIDJi;. EusiROPSis, n. g. Head distinctly rostrate ; body Avithout dorsal projections ; side-plates of perteon shallow. Antennte of male with calceoli on peduncle and flagelluui in both pairs ; first antennae shorter than the second, secondary flagelluui one-jointed : the second antennae of male with the last joint of the peduncle very elongate. Mouth-organs nearly as in Eitsirus, but the mandibles have the molar feebly developed and the first maxillse have the second joint of the palp narrower and scarcely longer than the first. Gnathopods nearly as in Eitsirus, but with the backward projection of the carpus almost obsolete. Eirstand second peraeopods slender, with the finger ending obtusely and tipped with long setoe. The three following perajopods slender, elongate, plumose, with the fingers acute. The uropods of the first two pairs with outer branch much shorter than inner, the third pair very plumose in the male, the outer ramus shorter than the inner. The telson narrow, apically incised. EusiROPSis RiiSEi, n. sp. (Plates 13, 14.) Head Avith triangular rostrum longer than broad ; second, third, and fourth the shortest of the peraeon-segments ; first three segments of pleon large, with the postero- lateral angles rounded, hind margin not serrate. Eyes to all appearance entirely wanting. First antennce. Eirst joint rather bulky, carrying eight tufts of setules on the lower margin ; second joint rather shorter and much narrower than the first, with calceoli along the lower margin ; third joint almost like a joint of the flagelluui ; the latter rather 40 REV. T. E. E. STEBBING ON AMPHIPODA FEOM stout, carryiug a calceolus on each of tlie thirty-one joints rernainiug, the end being broken ofT. The secondary flagelluin is minute and quite unobtrusive. Second antennce. The first three joints very short ; the fourth stout, not so long as the first of the upper antennae, furnished on the upper margin with nine tufts of setviles ; the fifth slender, longer than the whole peduncle of the other pair, armed above with many calceoli ; the flagellum longer than the peduncle; of the forty-two joints twenty-seven carry each a calceolus on the upper side, the terminal fifteen, which are more slender and elongate than most of the others, having each a sensory filament. It may be supposed that the abundant armature of the antennse compensates the animal for its want of eyes. Tipper lip. The apical margin is rounded, and the usual fringing moustache is strongly developed. Mandibles. As in the neighbouring genus Emirus the cutting-plate is scarcely dentate, the secondary plate on the left mandible is quinque-dentate, on the right it is more spine- like ; the spine-row contains five or six small spines ; the molar is very feeble and unob- trvisive ; the second joint of the palp is robust, tlie third is narrower but rather longer, fringed with many spinules. Lower Up. The inner lobes are smaU. First maxillie. The inner plate is feeble, short, and seemingly not armed with any set£e ; the outer plate carries ten spines, of which the two outermost are the largest, the rest are denticulate ; the palp has a stout first joint, but the second is weak and tapering, scarcely longer than the first, tipped with five setfc. Second maxillce. The inner plate is broader than the outer, but its apical armature is shorter and very scanty. Maxillipeds. The inner plates reach only to the base of the palp, and are not strongly armed, but have the usual apical teeth ; the outer plates reach scarcely to the middle of the palp's second joint, and are fringed with not numerous sctules. The palp's second joint is broad and carries a conspicuous row of setae near the apex ; the third joint is similarly armed ; and the finger, which is long and strong, has a few small setules on its inner margin. First gnathopods. The side-plates are very shallow, covering no part of the limb's second joint. The fourth joint is subequal to the third, its broad apex reaching almost to the base of the hand, and having the wrist attached to its front margin. The dista end of the wrist is attached to the front margin of the hand, not to the side as might be supposed from the drawings made from mounted and somewhat flattened dissections. On the inner side of the wrist there is a small process, probably homologous with the laro-e one in Eusirns which intervenes between the fourth joint and the base of the hand. The hand is massive, distally widened, with long convex palm carrying setules, and endin"- in a spinigerous pocket, into which the long curved finger inserts its tip. Second ynathopods. The side-plates are shallow, longer above than below. The branchial vesicles attached to these and the five following pairs of limbs have accessory lobes. The joints of the second gnathopod show saircely any difference in shape from those of the first. First percBopods. The side-plates have a curious sort of axe-head shape, the front THE COPENHAGEN MUSEUM AND OTHER SOURCES. 41 corner rounded, the hinder acutely produced. The whole limb is very slender, and, like all the other perseopods, is adorned with long plumose setse. The second joint is elongate, with four or five setoe on the hinder margin and two on the front. The fourth joint has four setfe on the front margin; the fiftli has one on the front and two on the hind margin ; the sixth, which is rather longer than either of the two preceding, has two setae of great length on the hind margin and two or three on the front. The seventh joint is not finger-like, less than half the length of the sixth, its apical margin not acute, fringed with six ])lumose sette, mostly of great length. Second perccopods. The side-plates are almost oblong, twice as long as deep, with the upper margin slightly excavate. The limbs are in near agreement with the preceding pair. Third per(E02)ods. The side-plates are bilobed, the hinder lobe the deeper. The second joint is ex23anded, rather deeply notched on the hind margin. The third joint is short. The remaining joints are all very elongate, armed with numei'ous spines and long plumose seta; on both margins, except in the case of the finger, which has the settc only on the hind margin. The fourth, fiftli, and sixth joints are severally much longer than the second, which is scarcely, if at all, longer than the straight acute finger. Fourth percBopods. The hind lobe of the side-j)lates is rather long, the front one evanescent. The limb is like that of the preceding pair, but with a larger second joint, and the fifth decidedly shorter than the sixth. Fifth percropods. The side-plates are not bilobed. ils already noticed, there are branchial vesicles. Tlie second joint is rather longer than in the preceding pair, while the fifth and sixth are not quite so long. As in the two preceding pairs, the setae fringing the margins of the sixth joint are of great length, and the hinder apex carries a remark- able tuft of these elongate appendages, which, together with those on the fingers, produce a striking eff"ect. Fleopods. There are three or four cleft spines on the first joint of the inner ramus, this ramus consisting of seventeen joints, while the outer, which is slightly longer, has nineteen. TJrop>ods. In the first pair the peduncle is nearly as long as the inner ramus, and carries some small plumose setaB on its outer, and spines on its inner margin. The outer ramus is a little more than half the length of the inner. Both have many spines alouo- the margins, those on the inner margin of the inner ramus being small but very close- set. The peduncle of the second pair is about as long as the outer ramus, which is less than half the length of the inner. These rami are armed as in the preceding j)air, which they a little exceed in length. In the third pair the pedimcles rather exceed in length those of the preceding pair. The outer ramus is a little shorter than the inner of tlie second pair; it has spines at six points of the outer margin, and the inner fringed witli spines and many long plumose setae. The inner ramus is rather longer than that of the second pair, and is fringed on both margins with spines and long plumose setae. Telsoii longer than the peduncles of the third m-opods, distally cleft for about a quarter of its length ; a little way above the acute apices a notch on either outer margin carries a long seta, and there is another a little higher up than the top of the cleft. SECOND SERIES. — ZOOLOGY, VOL. VII. 6 42 EEV. T. E. E. STEBBING ON AMPHIPODA FEOM Length, not including the antennae, in the slightly curved position of the sjiecimen figured, two-fifths of an inch. Hab. Tropical Atlantic. The label accompanying the specimen figured contained the words " Eiisei 55. Stud. Sanil. ded. 1892." A second specimen, in less satisfactory condition, was labelled " 22° N.B. 36° V.L. Hygom. Stud. Saml. ded. 1892." Copen- hagen Museum. The specific name was given in MS. by Professor Lutken in compliment to Herr Riise. Pam. undetermined. Sancho, n. g. liostral point small. Perseon depressed, very broad at the centre; fir.st segment short, seventh unusually long. Pleon narrow, much of it flexed. Eyes on the top of the head, separate. First antennae with principal flagellum longer than peduncle; accessory flagellum small, one-jointed. Second antennse with last joint of peduncle longer than the preceding joint. Upper lip not emarginate. Lower lip without inner lobes. Mandibles Avith cutting-edge, accessory plate, and spine-row small, but molar and three-jointed palp powerful. Pirst and second maxillae normal. Maxillipeds with outer plates smaller than, and scarcely reaching beyond, the inner ; fourth joint of palp small. Gnathopods subchelate, the first pair in both sexes feeble ; the second also feeble in the female, but in the male very long with very bulky sixth joint. Pera^opods normal, the fifth pair the longest. Second uropods with peduncle scarcely shorter and rami much longer than those of the first, third with short peduncle but very long inner ramus; in all three pairs the outer ramus is shorter than the inner. Telson short, triangular. The name is taken from a character famous in fiction. Sancho platynotus, n. sp. (Plate 9 A.) The second and third pleon-segments have the postero-lateral angles minutely produced ; at the second segment the pleon is bent and the remainder closely adpressed to the ventral surface of the trunk. Eyes round, separated by more than the diameter of either, composed of numerous ocelli, of which those of the outer ring appear to be larger than the rest. Specimens preserved in spirit have a light pinkish pigment. First antennce. Pirst joint stout, about once and a half as long as broad, second much shorter and thinner, third about half the length of second ; flagellum in male of forty- one joints, the first longer than the last of the peduncle, followed by many joints not longer than In'oad, to which succeed several much longer than broad. Second antennce. The basal joints short, the last joint of the peduncle rather long, longer and thinner than the penultimate, which has an apical tooth ; the flagellum in the male similar in structure to that of the first antenna, but perhaps shorter ; thirty- three joints in an example not quite perfect. Upper lip wath broad front, the margin little curved. THE COPEXHAGEN MUSEUM AND OTHEE SOURCES. 43 Ilatidibles. The cutting-edge small, little dentate, the molar massive; the third joint of the jmlp hroad, distally fringed with spinules. First ma.rillcB. Inner plate oval, tipped with two setse, outer plate with the eleven apical spines very small, inconspicuously denticulate ; second joint of the palp broad, with some apical setules. Second maxillj Sec, Zool. Ser.2 Voi,.VjI.Pl.8. ^^ O prp.5. V '^ Vf TRRSlcbhinij. 'Id prp -i. JTRReid hth ANDAWQTEK G0RPULEKTU5. (THOMSON) CTi: BEING A TfansLinn Soc Zool Ser S.Vol.YII I'i prp5 mji. iOo.'slp. TRRSlebbmg del Pardaliscoides tenellus Stebbing, JTRReidLith OTEDEING, Traws. Ljnn Snc.-ZooL Ser 2 Vol'^/11 Pl 13 -'- Stebbina. del JTRRieihil: Eus.iROPsio RirsEi Nov gen et sp Stebbino Trans Link .joc z.ool 6er c VolVU Pi, H' ft 4 I \ i pi^pi d^J EUSI.ROPSIS KlISEI NOV- GEN ET SP LINNEAN SOCIETY OF LONDON. MEMORANDA CONCERNING TRANSACTIONS. Thel'irst Scries of the Transactions, containing both Botanical and Zoological contributions, has been completed in 30 Vols., and a few entire sets are still for sale. Onl}- certain single volumes, or parts to complete sets, may be obtained at the original prices. The price of the Index to Yols. 1-25 is 8s. to the public, and 6s. to Fellows; to Vols. 26-30, 4s. to the public, and 3.?. to Fellows. The Second Series of the Transactions is divided into Zoological and Botanical sections. The prices of the Zoological parts of these which have been published are as undermentioned (for the Botanical parts see Botanical wrapper) : — When Price to the Volmne. p„bli,sh«l. Public. £ s. d. Secosd Seiues. — Zoology. I. Part I. 1875 1 4 0 . Part 11. 1875 0 6 0 . Part III. 1876 1 8 0 . Part IV. 1877 0 16 0 . Part V. 1877 0 18 0 . Part VI. 1877 1 2 0 . Part VII. 1878 116 0 . Part VIII. 1879 1 0 0 . II. Part I. 1879 1 4 0 . Part II. 1881 0 15 0 . Part III. 1882 1 8 0 . Part IV. 1882 0 7 6 . Part V. 1882 0 3 0 . Part VI. 1883 1 0 0 . Part VII. 1883 0 5 0 . Part VIII. 1883 0 3 0 . Part IX. 1883 0 3 0 . Part X. 1884 0 4 6 . Part XI. 1884 0 10 0 . Part XII. 1885 0 6 0 . Part XIII. 1884 0 6 0 . Part XIV. 1^85 0 6 0 . Part XV. 1885 0 4 6 . Part XVI. 1885 0 5 0 . Part XVII. 1886 0 3 0 . Part XVIII. 1888 0 2 6 . III. Part I. 1884 1 14 0 . Part 11. 1884 1 12 0 . Part III. 1885 1 10 0 . Price to Fellows. 0 18 0 .046 ..110 . .. 0 12 0 . .. 0 13 6 . .. 0 16 6 ...17 0 . .. 0 15 0 . . . 0 18 0 .. 0 11 6 ..110 . , 0 5 6 .. 0 2 3 .. 0 15 0 . . . 0 3 9 ... 0 2 3 . .. 0 2 3 . . . 0 3 6 . . . 0 7 6 . . . 0 4 6 . . . 0 4 6 ..046 . . . 0 3 6 ... 0 3 9 ... 0 2 3 . . . 0 2 0 . . . 1 5 6 . .. 1 4 0 . .. 1 2 6 When Pi'ice to tlic \olunie. Published. P.iblie. £ s. il. Second Series. — Zoology (c(»iiin III. Part IV. 1885 0 8 0 . Part V. 1887 0 8 0 . Part VI. 1888 0 6 0 . IV. Part 1. 1886 1 4 0 . Part II. 1887 1 8 0 . Part 111. 1888 0 16 0 . V. Part I. 1888 0 12 0 . Part 11. 1888 0 5 0 . Part III. 1889 1 7 0 . Part IV. 1890 0 12 0 . Part V. 1890 0 6 0 . Part VI. 1891 0 12 0 . Part VII. 1891 0 6 0 . Part VIII. 1892 0 8 0 . Part IX. 1892 0 12 0 , Part X. 1893 1 8 0 , Part XI. 1894 0 2 6 . VI. Part I. 1894 2 0 0 . Part II. 1894 1 11 0 . Part III. 1894 0 10 0 . Part IV. 1896 1 4 0 . Part V. 1896 0 10 0 . Part VI. 1896 0 8 0 . Part VII. 1896. ... 0 12 0 , Part VIII. 1897 0 2 6 . VII. Part I. 1896 0 10 0 . Part II. 1897 0 12 0 . Price to Fellows. d. ued). .. 0 6 0 .. 0 6 0 .. et 4 6 . . 0 18 0 .. 1 1 0 .. 0 12 0 .. 0 9 0 .. 0 3 9 .. 1 0 0 .. 0 9 0 .. 0 4 6 .. 0 9 0 .. 0 4 6 .. 0 6 0 .. 0 9 0 .. 1 1 0 .. 0 2 0 .. 1 10 0 . . 1 3 3 . . 0 7 6 . . 0 18 0 . . 0 7 6 .. 0 6 0 .. 0 9 0 . . 0 2 0 .. 0 7 6 .. 0 9 0 2nd Ser. ZOOLOGY.] [VOL. VII. PART 3. THE ^^/Z TRANSACTIONS OP THE LINNEAN SOCIETY OF LONDON. THE ORIGIN OF THE CORPUS CALLOSUM: A CO:\IPARATIVE STUDY OF THE HIPPOCAMPAL REGION OF THE CEREBRUM OF MARSUPIALIA AND CERTAIN CHEIROPTERA. BY G. ELLIOT SMITH, M.D., Ch.M. (SycNEY), ' JAMES KING " RESEARCH SCHOLAR OF THE UNIVERSITY OF SYDNEY ; ST. JOHn's COLLEGE, CAMBRIDGE. {Communicated hij Prof. G. B. Howes, Sec. Linn. Sue.) LONDON: PRINTED FOR THE LINNEAN SOCIETY BY TAYLOR AMD FRANCIS, RED LION COURT, FLEET STREET. SOLD AT THE SOCIETY'S APARTMENTS, BITRLINGTON-HOUSB, PICCADILLY, W. AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW. Jinir 1897. r 4.7 ] III. T/ie Oricfin of the Corpus Callostim : re]iaratioiis stained by Weigert's method, all the observations concerning JS'i/ctojihiltis which are here recorded. ORIGIN OF THE COEPUS CALLOSUM. 51 Perameles nasiita at about the junction of body and descending limb of the lateral ventricle. This corresponds to a situation almost immediately behind the commissures of the liemispheres, as we shall see later. In this section the 'tween-brain — a somewhat square-shaped mass with rounded corners — will be seen to occuj)y the ventral mid-region. The basal part of the thalamic region lies upon the optic tract {o.tr.), which may be seen extending into the furrow between the hemisphere (tlie region of the nucleus amycjdahv, n.a.) and the 'tween-brain. The two halves of the ojitic thalamus are connected by means of a broad bridge of grey matter — the commissura moUis (cm.) — and above and below the latter portions of the third venti'icle may be observed. The small recess of the third ventricle lying above the soft commissure is roofed by a somewhat dome-shaped epithelial fold, which is attached on either side to a ridge containing a longitudinal band of meduUated fibres — the sti'ia medullarls thalami (s.m.t.). Upon its lateral aspect the thalamic region fuses with the corpus striatum of the cerebral hemisphere, and a mass of medulltited fibres, cut either transversely or ol)Iiquely, indicates the place of union. The corpus striatum consists of two parts, which are separated one from another by the internal capsule (ci.). A smaller grey mass, lying above the internal capsule and projecting into the lateral ventricle, is the iiucleus caudatus (n.c). A larger grey mass lying below the internal capsule is the nucleus lentlcularis {u.L). The ventral part of the nucleus lenticularis [n.a.) is the homologue of the grey mass known in human anatomy as the nucleus mnygdalce. It is directly continuous with a cortical area of distinctive structure known as \hQ liyr'iform lobe [i^y.l.), from which the nucleus lenticularis in its upper part is separated by a layer of meduUated fibres (the external capsule, c.e.). Along its dorsal border the pyriform lobe is separated from the general cortex (which Turner calls "pallium," p.) by the shallow rhinal fissure {f.r.). Above the corpus striatum, where the lateral wall and roof of the ventricle are formed by this general cortex ov pallium, the inner lining of the latter is formed by a mass of meduUated fibres — the corona radlata {c.r.) of writers. If the cortex be traced in a mesial direction its edge wUl be found to present those peculiarly distinctive features which w^e at once recognize as hippocampus. Thus if the mesial waU of tlie hemisphere be traced down- ward in the section (fig. 1, PL 15) we reach a deep indentation — the hippocamptal fissure , and corresponding to this fissure a marked prominence into the lateral ventricle forming the hippocampus {hip.). Below the hippocampal fissure the characteristic features of the fascia dentata (fd.) (Avhich we shall consider subsequently) are immediately recognizable. And extending into the deep depression between the nucleus caudatus {n.c.) and optic thalamus is the prominent _^mir/ff (fi-), spur-like in transverse section. [In order to avoid an luinecessary confusion in the figure, the choroidal fold of the lateral ventricle which is attached to the fiml)ria has not been represented.] Having thus seen how the " hippocampal formation " constitutes the edge of the cortex cerebri, I propose to carefully examine the series of changes whicli this peculiar formation undergoes as we trace it forward in a series of coronal sections of the cerebrimi of the bandicoot (Perameles). 52 DK. G. ELLIOT SMITH ON THE A coronal section tlirongh the cerebrum, just in front of that represented in the first figure, presents a marked change in the appearance of the hippocampal region, Avhich is represented upon an enlarged scale in fig. 2. The hijipocarapus itself, as well as the fascia dentata, have undergone practically no change, but a huge bundle of medullated fibres from the upper part of the fimbria is now exposed in the act of crossing the middle line. This is the posterior exti'emity of what — to avoid confusion of terms — will for the present be termed the commissura dorsalis (c.i).). It would be foi-eign to the scope of this work to descril)e in detail the histology of the hippocampal formation, seeing that the minute structure of the Metatherian hippocampus is identical with that of the Eutherian, which Sala, Schaff'er, Ramon y Cajal, and von KoUiker*, among many others, have so carefully and minutely described Avithin recent years. But I wish to call attention to certain features which are intimately associated with the evolution of the structure we are primarily considering. Lying upon the upper surface of the " dorsal commissure " (c.d.) there is a complicated epithelial fold the outline of which is roughly triangular, and which projects upward into the great cleft between the two cerebral hemispheres. The fascia dentata {f.d.) lies in contact with this epithelial (choroidal) fold on either side, and may be easily recognized by certain distinctive features. Thus it possesses a very thickly -packed column of small cells — the stratum gramdosum (s.fjr.) — the nuclei of which stain very deeply, lying parallel to the surface, from which they are separated by a clear layer almost devoid of cells — the stratum molecuJare. The fascia dentata — consisting of stratum granulosum and stratum moleculare — is folded in a peculiarly characteristic manner. Its mesial extremity is completely bent upon itself, so that the stratum moleculare or superficial layer rests upon the dorsal aspect of the dorsal commissure. The lateral extremity of the fascia dentata undergoes a corresponding bending, the concavity of which is directed toward the mesial plane. It would appear that the fascia dentata had continued to increase in breadth whilst its mesial and lateral extremities were relatively fixed, and, to accommodate itself to tliese conditions, it bad bulged out into a pouch, as it were, and at the same time become flattened by the resistance of the general cortex lying immediately above it. Such factors, in reality, seem to have been at work, as I have recently pointed out t. The fissure which separates the fascia dentata from the general cortex where their resj^ective superficial layers come into contact is ihefissura hipiMcampi (f.h.). The concavity of this pouch of fascia dentata is occupied l)y scattered nerve-cells, many of which give origin to the brush of medullated fibres which are to be seen in the figure, collecting near the inner margin of the fascia dentata, to lose themselves in the huge mass of medullated fibres lying immediately below. The hippocampus proper projects as a large rounded swelling into the lateral ventricle. This swelling is covered by a thick layer of medullated fibres — the so-called ulceus (alv.). These fibres almost all pursue an obliqiie course, so that in a coronal section only short pieces of ^fibre are seen massed together. In the first figure most of the fibres of the * ' Gewebelehre des "Munsohen,' Bd. ii. zwciter Hiilt'te, IS'JG. t '■ The Fascia Dentata,"' Auatomischer Aiizeiger, xii. Bd. Nr. 4 iind .5, 1895, p. 119. OEIGIN OF THE COEPUS CALLOSUM. 53 alveus are tending towards the fimbria, but in the section under consideration many fibres of the alveus, as well as fibres coming from the " pouch " of the fascia dentata, proceed directly into the commissura dorsaUs along with many fibres of the fimbria which have come from more caudally situated regions of the hippocampus. It will be noticed, however, in the figure that as yet only the upper part of tlie fimbria is contributing fibres to the commissure. The tip of the fimbria, which contains fibres coming from the most distant parts (temporal pole) of the hippocampus, is yet undisturbed and has given no contriluition to the dorsal commissure. Between the alveus and the curved lateral part of the hippocanipal fissure there is a broad, more or less definitely stratified region. Slightly nearer to the alveus there is a very uniform and clear layer (in specimens stained by Weigert's method). This is the layer of " ^lyramidal cells," and consists of a closely-packed column of large cells, which stain moderately with ordinary stains. Between the layer of pyramids {i^yr.) and the alveus {ctlo.) there is a stratum containing scattered polymorphous cells and very abmidant medullated nerve-fibres — the layer of poli/morphous cells. The space between the hippocamjjal fissure and the layer of pyramids is usually divided by writers into three zones — a broad clear layer next to the layer of pyramids [stratum radiatum, s.r.), a thin superficial layer [stratum zonale), and an intermediate layer very rich in medullated nerve-fibres [stratu^ii lacunosum, s.l.). A large number of fine medullated fibres (.r) extend from the region of the layer of pyramids through the stratum radiatum into the stratum lacunosum. These fibres are very distinct in the region bordering upon the fascia dentata [clde fig.). In the stratum lacvmosum many of these fibres end by means of free branching in relation to the processes of the pyramidal cells, which extend through the stratum radiatum into the more superficial regions. But a considerable number of medullated fibres assume a longitudinal direction (/. e. in the long axis of the hippocampus) as soon as they reach the stratum lacunosum [s.l.). These fibres, which may 1)e called " longitudinal association fibres of the hipjwcamj^us," appear in transverse section [s.l.) as rounded bundles immediately surrounding the lateral or deeper part of the hippocampal fissure. In the general cortex [j).) the cells are scattered throughout the whole thickness of the cortex, whereas in the hippocampus (leaving out of account the fascia dentata) the great majority of the nerve-cells are crowded together into the regular row of pyramids. The transition-region from typical cortex to typical hippocampus is generally known as the sub/culuv/, cornu animonis [sub.). AVhether the " subiculum" ought to be grouped with the general cortex or with the hippocampus is a question to which Ave shall have to refer su.bsequently. The deej)est layer of the general cortex consists of a mass of medullated fibres — the corona, radiata. It will be noticed at a glance that while the fibres derived from the hippocampus and " subiculum " proceed into the alveus, those derived from the general cortex proceed away from the alveus, i. e. towards tlie external capsule. At the point of meeting of alveus and corona radiata there is obviously a commingling of fibres, but it is equally clear that any passage of fibres from the alveus into the " corona '' or vice versci is either non-existent or extremely slight and practically negligible. In the adjoining regions there are groups of longitudinal fibres belonging respectively to the two distinct formations, and it is quite possible that an interchange between these two 54 DE. G. ELLIOT SMITH ON THE series may take place. The longitudinal series belonging to the margin of the general cortex or pallium is the so-called cingulum of higher forms (figs. 6 & 11, cing., clngJ). There are, however, definite bonds of union between hippocampus and general cortex which are found in their junction. Thus from the stratum zonale and stratum lacunosum of the hippocampus fibres extend not only into the " neutral ground," so to speak, of the " subiculum cornu ammonis " but well into the molecular layer of undoubtedly true cortex. This is a definite association bundle between hippocampus and pallium, which may be called tractus hippocamfi ad pallium, (fig. 2,?/). As the hippocampal region is successively examined in a series of coronal sections, the main features of the hipj)0campal formation itself remain imchauged for some time, but its commissure — the commissura dorsalis — undergoes marked changes. Thus the depth of the commissure rapidly increases at the same time that the prominence of the fimbriae is diminishing (fig. 3). The meaning of this is that an increased number of fibres from the fimbria are crossing in the dorsal commissure. The scheme (fig. 1) of the commissures as they are seen ia sagittal section explains the meaning of these changes. In figure 5 a new factor for the first time becomes evident. Upon the lateral aspects of the commissure, more especially upon the right side of the figure, a mass of grey matter {s) will be observed to enclose the dorsal commis.sure and to become continuous ventrally with a mass enclosing a second commissu.re — the commissura ventralis. This grey mass enclosing the two commissures will for the jiresent be distinguished as the " commissure-bed." On either side it will be noticed to become continuous by a narrow bridge of grey matter with the corjms striatum (est.). In this narrow bridge there are a number of longitudinally-coursing nerve-fibres (s.t.). These are the uncrossed fibres of the stria ter'minalis, the other fibres of which may l)e seen entering the ventral commissure {s.t!.) in figure 3 to cross to the other side. In figure 5 the dorsal commissure has reached its greatest depth. Figure 6 shows the appearance which is presented by a section after all the commissural fibres of the fimbria have crossed. The thin commissura dorsalis (c.d.) is derived from the corresponding region of the hippocampal formation. The huge commissura ventralis (c.v.) appears to become continuous with the capsula externa [ce.], which, after being joined by fibres of the capsula interna (cL), becomes the corona radiata. The dorsal and ventral commissures lie in the large mass of grey matter which forms the commissure-bed. A thin lamina (?') of this grey mass may sometimes be distinguished upon the dorsal as^jcct of the commissura dorsalis. This may be distinguished as the " indusium vertim." The " commissure-bed " (s) contains large numbers of uncrossed fibres (d.f.), which are derived from the alveus of the hippocampus (which is situated above them) and from the fimbria. These fibres are divided into two groups by the ventral commissure. The p)ostcommissural fibres (columna fornicis, c.f.) collect upon the dorsal aspect of the ventral commissure and incline backward and downward to enter the thalamic region. The precommissural fibres proceed downward and forward in front of the ventral commissui'e. OEIGIN OF THE COEPUS CALLOSUM. 55 As we proceed forward in tlie examination of coronal sections, the dorsal commissure will be found to become gradually thinner and finally disappear altogether, while the desceudmg foniix-fibres remain unchanged. The ventral commissure disappears in about the same coronal plane. The " commissw^e-hed " extends a slight distance further forward before the hemisplieres become separated one from another. Pigure 7 represents (about four times the natural size) the appearance of a coronal section immediately in front of the place where the hemispheres become separated from each other. In this figure the hippocampus is again easily recognizable, although its conformation has become greatly simplified. The fissura Idppocampl (f.h.) is much shallower, and the hipj)Ocampal projection into the ventricle less prominent. The fascia dentata is much narrower, and consequently much less folded. The thin alveus may be seen proceeding directly into the mass of desceudinfj forwix-fihres which occupy the superficial portions of the grey mass which we may call the " corpus prcecominissiiralc" *, but which is really only the anterior continuation and homologue of the grey mass which further back has been called " commisstire-bed.'" The other features of the section do not concern us in considering the hij)pocampal formation. The only other point to which I wish to call attention is tliat the " corpus prcecommissurale " is directly continuous below the lateral ventricle with the corpus striatum, where the latter is capped by the peculiar cortex which Ganser calls the " cortex of the head of the corpus striatum," and which corresponds to the " tubei'culnm olfactorimi " [t.o.) — such a prominent feature upon the ventral aspect of this brain. If we were to trace these various structures forward in a series of coronal sections, we should find that the hippocampal fissure would become gradually shallower until it eventually disappeared, the prominence in the ventricle disappearing pari passu. The fascia dentata, now placed entirely upon the surface, becomes narrower and less clearly differentiated, until at last we fail to recognize any distinguishing feature of a hippo- campus. But by the time this has occurred we have arrived almost at the olfactory pedvtncle. In certain Marsupial brains {Dasyuriis clverrinus) it is even possible to distinguish the hippocampal formation as far forward as the olfactory peduncle, so that here wo have the "precommissural area " completely cut off from the general cortex or pallium {p.) by a hippocampal band, as semi-schematically represented in figure 8 {hip." & f.d.). This is the primitive condition of the hippocampus, which is found (although recognized with difficulty) in the Reptile, and which I have already described in the fcetal OrnithorMjnclms (Qu. Join-. Micr. Sci. vol. xxxix. pp. 181-206). Pigure 9 represents the appearance which this region presents in Perameles when viewed from its mesial aspect. In this figure the optic nerve (opt) is seen extending towards the lamina terminaHs, at the inferior extremity of which is the oval section of the optic tract (o.tr.). Extending upward from the optic tract is the thin ventral portion of the lamina terminalis' — the lamina cinerea {I.e.). At its upper part this lamina expands to enclose the large * By " corpus pnicommissunde " is meant that grey mass the mesial surface of which is the "precommissural area " (p.a.). SECOND SERIES.— ZOOLOGY, VOL. VII. 9 56 DK. G. ELLIOT SMITH ON THE commissura veiitralis (c.v.), oval in section and placed obliquely. Extending upward from the ventral commissure there is a thick mass of grey matter — the "com- missure-bed'" {s) — Avhich contains in its upper part the commissura dorsalis (c.D.), shaped somewhat like an inverted obliquely placed V Avith the dorsal limb nearly horizontal. Above the dorsal commissure the fascia dentata {f.d.) is easily recognizable and is separated from the general cortex (jk) by the fissura Mppocamjpi {f.h.). Anteriorly the hippocampal fissure gradually becomes shallower until it can no longer be traced, so that the fascia dentata, the stratvim zonale of the hippocampus, which now of course lies upon the surface (compare figure 8, hip.'"), and general cortex {p.) are indistinguishable. Upon its ventral side the fascia dentata becomes directly continuous with the extensive pale region which constitutes the " frecommissnral area" (p.a.). In the forward direction the "• precommissural area " is directly continuous with the mesial wall of the extremely short olfactory peduncle, which connects the hemisphere with the olfactory bulb {o.h.). Upon its ventral side the " precommissural area " becomes continuous with the grey boss generally known as the " tuherctihtm oJfactorinm " {t.o.). By means of this representation of the anterior portion of the mesial wall of the hemisphere of Ferameles (fig. 9), and the semi-schematic representation of the corre- sponding region in Dasyurus (fig. 8), we obtain a good general idea of the anatomy of the commissural region. In Dasyurtis it will be observed that the dorsal commissure is not so distinctly bilaminar as it is in Pe/'flrtK'/fs : and in Phascolarctos (PI. 16. fig. lOj it can hardly be called bilaminar. Thus we have in the Marsupial all the intermediate stages between the rounded dorsal commissure of Oriiithorhynchiis and Echidna and the distinctly bilaminar form found in most mammals. There are certain important features of this region which can only be demonstrated by sagittal section, which we will consider almost immediately. But before leaving the consideration of figure 8, I wish to call attention to the possible consequences of elongation of the commissura dorsalis. The " commissura dorsalis " is contained in the mass of grey matter which I have called the " commissure-bed," the separate anterior prolongations of which in the two hemispheres are formed by the '^corpora j^'^ffcommissuralia'" — the grey masses corresponding to the surface-region of each hemisphere, which is distinguished as "precommissural area." Now this combined grey mass — " commissure- bed" and " corpus prnecommissurale "■ — is separated from the general cortex (pallium) above it by tlie hippocampus ; and the dorsal limb of the commissure separates the " commissure- bed " from what we may call the " supracommissural hippocampus," while the " corpixs prsecommissui'ale " becomes directly continuous with wdiat we may call tlie " pre- commissural hippocampus." Now it is clear that if the dorsal commissure elongates in the forward direction it must cut otf jiart of this " precommissural " body from the precommissural hippocampus, which becomes " supracommissural." The structure lying upon the ventral aspect of such an elongated dorsal limb of the commissura dorsalis must be either " commissure-bed " or " corpus pra^commissiu'ale." ORIGIX OF THE CORPUS CALLOSUM. 57 Figure 10 represents the appearance of a sagittal section through the corresponding region in the brain of Phascolarctos clnereus, after stciining hy the Weigert method. Between the small commlssiim dorsalis (c.n.) and the large elliptical commissura ventmlis (c.v.) a large number of non-crossing fibres will be noticed in the " commissure-bed " and "precommissural area." Of these the postcommissural form the columna fornicis (c.f.). The precommissural area contains various sets of fibres. The largest group (a) connect the alveus hippocampi and the basal region of the brain just behind the tuberculum ■olfactorium (f.o.). In Oniif/ior/n/ncf/iis I distinguished these fibres as the " hippocampo- basal association bundle" *. Other fibres (/3) pass backward below the ventral commissure to enter the subthalamic region, wliere they become lost. Fibres entering into the constitution of these tAVO series may be seen in the jjrecommissural area (in the region marked jj.«.), proceeding from the most anterior i)art of the hijipocampus. Fibres may also be seen proceeding directly from the olfactory peduncle (and probably bulb) to enter the fascia dentata (f.d.). This is the olfactory bundle of the fascia dentata (y). Other fibres belonging to the same series (o) may be seen proceeding backward above the dorsal commissure as a well-defined tract, probably homologous with the nerve- fibres of the stria medialis Lancisli of the higher mammalian brain. In the Marsupial, a typical hippocampal formation lies upon the dorsal aspect of the commissura dorsalis. In tbis hippocampal formation there are two important series of longitudinal fibres, Avhich are displayed most clearly in a horizontal section. Such a. section of the cerebral hemisj^here of Trlchosurns culj)ecula is represented in fig. 11. In this figure (as also in fig. 2) a large series of longitudinal association bundles are visible in the stratmii lacunosum. Secondly, in the region of the subiculum (and possibly in other regions of the hippocampus) large numbers of longitudinal fibres are contained in the alveus, but in transverse section are not recognizable in the mass of oblique fibres. Upon the ventral aspect of the hippocampus in its anterior part we have had occasion to refer to a large mass of grey matter, the anterior paired region of which we have known as the " corpus prsecommissurale," and its posterior region as the " commissure-bed." This region has been the great source of confusion to comparative anatomists, more especially those working at rej)tilian and amphibian forms, and it therefore demands a very careful study. Most writers call the whole mass the " septum." A study of figs. 5, G, and 9, which are all from Perameles, shows that the " commissure- bed," or matrix of the commissures, extends across the median plane uninterruptedly, becomes continuous upon its dorsal aspect with the hippocampus of both sides, and upon its ventral aspect with that thin band {I.e.) of the lamina terminalis which extends from the ventral commissure to the optic tract (o.tr.). It is connected also in the median line above with the thin epithelial roof of the third ventricle (a structure not sho^^ n in the figures). It is clear, therefore, that this matrix, which includes the two commissures, occupies the situation of the dorsal part of the lamina terminalis, and in part bomids the third ventricle anteriorly. * Journal of Anatomy and Physiology, vol. xxs. p. 480. 58 DR. G. ELLIOT SMITH ON THE If we examine a series of transverse and sagittal sections of a reptilian brain, it will be found that both ventral and dorsal commissni-es lie in the easily recognizable lamina terminalis. This lamina contains in the median j)lane a minimal amount of grey matter. But on either side of the median plane this thin bridge of grey matter becomes continuous with a huge mass of grey matter which we readily recognize as the backward continuation of the " precommissural body." Thus in the Reptile the " commissure-bed " consists of the posterior extremities of the " precommissural arese " united by the thin grey lamina formed by the lamina terminalis. In the mammal it is evident that either the grey mass of the " precommissural area" has invaded, and thus thickened, the lamiua terminalis, which liecomes a continuous grey mass across the mesial plane, or, on the other hand, the mesial surfaces of the two posterior regions of the precommissural area may have become " glued " together. One or other event must have occurred ; and the evidence, I believe, points to a gradual thickening of the lamina terminalis by the invasion of grey matter from the adjacent " precommissural area." However the thickening is brought about, we know that the two cei'ebral commissures of the Marsupial lie in the " commissure-bed," which is morphologically a thickening of the lamina terminalis. Yet we are equally sure, both from its topographical relations to surrounding structures and from its minute structure, that the " commissure-bed " is morphologically part of the same sheet of grey matter as the " precommissural area." If the hippocampal region be studied where it bulges in the descending limb of the lateral ventricle in any Eutherian brain, a considerable variety of the form and relative sizes of its constituent parts will be apparent. In the Marsupial brain equally great variations are found in the corresponding region of the hipj)ocampus. But if we examine the supracommissural part of the hippocampus in a series of Marsupials, the greatest variety and most varied degrees of complexity are demonstrated. In Notoryctes, and less markedly in Dcmjiimis, the formation is very simj)le. In Ilacropus it is much more complex. Perameles is an example of a highly-developed type. Both Monotremes present in the supracommissural region a more complicated form of hippocampus than is fovmd in any other mammal. But the anterior extremity of the hippocampus undergoes the same process of unfolding and simplification in all Marsupials and Monotremes which I have so briefly described in Ferameles. Hence, should we find a difference in the relative degree of development of various parts of the hippocampal formation in the brains of two different animals, we must not lay much weight upon such differences as an indication of the systemic position of their possessor. The Mippocampal Region in the J3at. The cerebrum of the bat may now be studied in the same manner as that of the bandi- coot. Beginning with a coronal section through the hippocampal region of Miniopteriis immediately behind the region of tlie commissures, one cannot fail to notice the great similarity between the two forms (figs. 12 and 1). What differences do occur, apart from those of al)solute size, point to a less complex form of structure in Miniopterus (fig. 12) than in Perameles (fig. 1). This simpfication consists of a diminished depth of OEIGIN OF THE CORPUS CALLOSUM. 59 the hippocampal fissure [f.h.), a less prominent hippocampal eminence in the lateral ventricle, and a less broad and slightly less convoluted fascia dentata {f.d.). These variations in the form of the hippocampus are of little moment, for, as I have before remarked, great variations are found in any order. Moreover, it is noteworthy that in very small Marsupials like Notoryctes one finds a simplicity of hippocampal formation similar to that which the miniite Miniopterus presents (compare fig. 6 of my paper on Notoryctes*). In the next section, whicli passes through the posterior extremity or splenimn of the dorsal commissure (PL 16. fig. 13), we have a condition which recalls that presented by the corresponding section through the brain of Perameles (Pi. 15. fig. 2). But if we compare the hippocampus in figs. 12 and 13, a distinct simplification will be noticed in the foremost section (fig. 13). The fascia dentata (f.d.) is narrower, the layer of pyramidal cells {pyr.) is shorter, and the hippocampus as a whole is distinctly flatter. In fig. 12 there is a very sharp angle between the alveus and tlie corona i-adiata, while in fig. 13 the angle has become rounded otF. It is possible that fibres coming from a wider cortical field may round off the angle, but the carmine stain does not permit us to definitely decide this. But in the next section (fig. 11) there is no question, even with the carmine stain, of the much wider field of origin of the dorsal commissiu'e. If we are undecided whether any fibres other than hippocampal are passing to the dorsal commissure in fig. 13, we have in fig. 11 a very distinct demonstration of the existence of a verv considerable bundle of fibres arisina: from the sjeneral cortex and forming part of the superior commissure. In fiict, the insignificant flattened band of grey matter which now represents the hippocampus can contribute but a very small share towards the formation of the thick commissural tract (c.D.) upon which it lies. The hippocampus is now much more insignificant and more flattened than it was in the last figure. The regular row of nuclei immediately above the dorsal commissure belongs to the layer of pyramidal cells of the hippocampus {pyr.), and the group of closely- aggregated smaller nuclei which cap the mesial end of the layer of pyramids represents the stratum granulosum of the fascia dentata. In fig. 15 the region of the fascia dentata from fig. 11 has been rei)resented upon a larger scale, and in figs. 12, 13, 11, and 15 the stratum moleculare of the fascia dentata has been shaded in order to render its recognition more easy. It will be noticed, if we comj^ai-e figs. 13 and 11, that the commissural fibres of the general cortex pursue their new course through the alveus of the hippocampus. In other words, the alveus of the " dorsal hippocam^nis " of the bat consists of a mixture of fibres derived from the hijijiocampus, which we might call " true alveus," and commissural fibres derived from the general cortex — "false alveus." Our "false alveus " is the corpus callosum proper. Prom the examination of this series, therefore, we may infer that the fibres of the corpus callosum proper reach the mesial platie by passing through the alveus of the dorsal part of the hippocampus. * " The Cerebrum of Notoryctes tupldops,"' Transactions of the Eoyal Society of South Australia, vol. xix. pi. viii. (1895). 60 DE. G. ELLIOT SMITH ON THE As we proceed forward the supracommissural hippocampus becomes more and more rudimentary. Had we not been tracing- these structures in a series of sections, we should never recognize in the few cells situated above the commissures in fig. 16 the representatives of the fascia dentata {f.d.) and the rest of the hippocampus i])^)'.). If we were to examine a series of sections further forward we should find, after the anterior extremity of the dorsal commissure had been passed, that at the junction of the precommissural area and general cortex (pallium) a few small cells would be found to represent the hippocampus which is present in the marsupial (fig. 7). Now that we have seen how the hippocampus is disposed in transverse sections, it will complete oiu* ideas concerning the commissural region if we study its disposition in sagittal section. The transverse sections wliich I have figured were those of Jliniojpterus. The two series of sections of the brain of Nyctophilus, which would have served our purpose even better, were not exactly transverse, and therefore might be confusing. It may be remarked, however, that the " iion-hippocampal " element in the dorsal commissure is even less developed in Nyctophilus than it is in the single specimen of Minioptcrus which I have examined. I have made several series of sagittal sections of the brain of Nyctophilus, and, for illustration, a section slightly to the side of the middle line has been chosen in order to show the relation of th(; hippocampus to the " splenium " of the dorsal commissure, A section passing through the mesial plane presents no difi'erence in the appearance of the commissures from that presented in fig. 17. In this section (fig. 17) we cannot fail to be struck with the resemblance to the Marsupial. The relatively enormous size of the ventral commissure (c.v.) and the shape and proportions of the dorsal commissure (c.D.) are strongly suggestive of the Metatherian structures. Yet in no Marsupial are the two limbs of the commissure so broadly expanded. For the ventral limb (c.D.") is nearly vertical, and the dorsal limb (c.D.') nearly horizontal, so that the angle included between the two limbs is not far short of 00". Of the two lim1)s the ventral is the larger. Both limbs become thicker as they are traced towards their place of union, where they form a massive " splenium." Immediately behind the splenium in this section we see the fascia dentata {f.d.) cut very obliquely, and extending forward from the concavity of the pouch formed by the stratum granulosum the layer of pyramidal cells of the hijjpocampus {hip.) immediately above the dorsal commissure. The " commissure-bed " (s) enclosing both commissures fills up the great interval between the two limbs of the commissura dorsalis and tlie commissura ventralis. In the dorsal limb of the commissura dorsalis (c.i).') I have represented by means of dots the position which a study of transverse sections leads me to regard as approximately those occupied by the commissural fibres which do not spring from the hippocampal formation, i. e. by the fibres of the corpus callosum proper. From this we see that the anterior extremity of the dorsal limb is almost purely " callosal " — if we may use such a term. The transition from pvirely " callosal " to purely " hi2)i)ocampal " parts of the commissure is a very gradual one, " hippocampal " and " non-hippocampal " fibres being intimately mixed one with another at the place of junction. ORIGIN OF THE CORPUS CALLOSUM. 61 Tims it icould appear that in this bat cominissural fibres arising from a very restricted area of the dorso-mesial cortex above the supracommissural region of hippocayrtpus, instead of passing by the circuitous route via the external capsule and through the ventral commissure, pass through the alveus of the supracommissural hip)p)Ocampus to reach the opposite hemisphere through the dorsal limb of the dorsal commisstire. In other icords, fibres tohich do not c. /. F/c.Z F/a Fig. 1. — Scheme to show the arrangement of the hippocampus in the mesial wall of the hemisphere of a Marsupial. The darkly-shaded part represents the first corpus eallosum, such as is found in Miiuojjtenis, and the arrow indicates the Une of its backward extension in the higher Mammals. Fig. 2. — Scheme of the corresponding region of a more highly developed cerebrum, such as that of Pteropus. Fig. 3. — The " spherical " region from fig. 2 on a larger scale. remains of the hippocampus which cover the dorsal aspect of the corpus eallosum are stretched, and thus further attenuated, by the lengthening commissure. The backward growth of the commissure, owing to its obliquity, soon indents the hippocampal formation just at the junction of the atrophic " supracommissural " segment with the SECOND SERIES. — ZOOLOGY, VOL. VII. 10 64 DE. G. ELLIOT SMITH ON THE uaaltered " postcommissural " segment. This indentation soon increases and becomes a very extensive bend, the portion of the hippocampus which surrounds the posterior extremity of the dorsal commissure being carried backward and greatly stretched (fig,. 2). In this way the obliquely-directed commissure fi.rst of all pushes upward and backward the hippocampus at the junction of supra- and postcommissural parts, stretchiug and carrying back the part of hippocampus that invests it above the rest of tlie j^ostcommis- sural hippocmnpus. The " postcommissural hippocampus " (fig. 3, hip) thus appears tO' become subcallosal and to be separated in the vertical plane from the splenium of the dorsal commissure {s})!) and the circumsplenial portion of the hippocampus {hip') by a process of pallium (figs. 2 and 3, *) ujjon which Zuckerkandl has laid so much stress under the confusing name " Balkenwiudvmg." (It will be remem1)ered that it was the lack of this feature which, in part, led him to group a bat with Marsupials and Monotremes). As a result of the operation of the above-mentioned factors, therefore, a somewhat reversed S-shaped bending is produced in tlie hij^pocanipus at the jimction of its middle and posterior segments. The upper part of the S with its concavity looking forward is formed by the attenuated hippocampus (fig. 3, hip') surrounding the splenium (fig. 3, spl), while the lower part of the S, which is convex in front, is formed by a " sul)splenial " bending of a plumper hippocampal region, whicli I have distinguished as the " hippo- campal flexure " {fl^c)- Throughout all these changes tlie main mass of the ventral limb of the dorsal com- missure (CD.") has remained unchanged. In the Marsupial and in a large number of lowlier Eutheria the plump ventral limb extends obliquely upwards to meet the dorsal limb in a thick splenium. But when the dorsal limb extends further l)ackward, the main mass of the ventral limb is left in its old position (fig. 2, c.d."), its postero-supcrior extremity becoming greatly stretched and correspondingly thinned by tiie backwardly-extending corpus callosum. The fimbria maintains its position unchanged (figs. 1 and 2,_7^") and always lies upon the posterior or inferior aspect of the corpus callosum. In its backward growth the corpus callosum carries back its matrix or " commissure-bed," in which it always lies. But although the elongation of tiie corpiis callosum is most obvious in tlie backward' direction, it also groA^ s forward towards the anterior extremity of the brain. In the Amphibia and certain Reptiles, Avhere the olfactory bulb is placed in front and is not overlapped by the hemisphere, the " corpus prgecommissurale " extends forward as a horizontal baud from the lamina terminalis to the mesial wall of the olfactory peduncle.- It is bomuled upon its dorsal aspect by the homologue of the hippocampus. In the Marsupials with small pallia the upper margin of the precommissural body is slightly oblique (PL 15. figs. 8 and D). As the pallium increases in extent it bulges over the olfactory bulb (figs. 1 and 2, p. 63) more and more, and consequently the " precommis- sural " hippocam])Us {hip") becomes more and more oblique, because it always extends towards the olfactory peduncle. Thus the hippocampus comes to bend downward iu; front of the commissures. Now it is obvious that the dorsal limb of the commissura dorsalis (fig. 1), if it extends forward, must indent this precommissural segment of the OKIGIN OF THE CORPUS CALLO.SUM, 65 hippocampus as I have represented in fig. 2. This anterior portion of indented hippo- campus is the "gyms grnicuW' of Zuckerkandl. It is ai:)parent from this that the dorsal comiiiissure, which in. the Ilarsupial liesroithin a well-defined arc formed by the hippocampus (fig. 1), never breaks through this hippo- campal arc even in its much more extended and reconstructed form as a corpus callosum, but pushes the dorsal part of the arc before it as it elongates in. both directions (fig. 2). We have seen that in tlae Marsupial a large mass of grey substance, containing both commissures and formed by the "corpus prfecommissurale " and "commissure-bed," is situated i;pon the ventral aspect of this hippocampal arc. Since the hijipocampal formation may be recognized surrounding the corpus callosum even in the brains of the ox, the monkey, and man, it is clear tliat the matrix of this commissure can be formed of no other structure than the " precommissural body " and the " commissure-bed." This introduces us to the problem of the nature of the " septum pellucidum." We have already seen that the first " callosal " fibres rej)lace the hippocampal fibres of the dorsal limb of the " commissura dorsalis," and in every respect joresent the same relations as the fibres, the places of which they usurp. The first " callosal " fibres, therefore, must lie in the structure which is exactly homologous to what we have called, in the Marsupial, the " commissure-bed." In discussing the natiu"e of the " commissure-bed " we came to the conclusion that with the increasing dimensions of the commissures the dorsal part of the lamina terminalis became thickened or invaded by the posterior parts of the " corjjora prte- commissuralia," so that a bridge of grey matter — the " commissure-bed " — was formed, connecting the two hemispheres, and forming a matrix for the two cerebral commissures. This secondaiy thickening appears to take place gi-adually in the ontogeny of each individual, and cannot be regai'ded as a mere mechanical fusion of the precommissural areoe of the two hemispheres. We may therefore safely say that the dorsal commissure of the Metatherian and (for the reasons stated aljove) of the early Eiitherian cerebrum is contained in a matrix which is formed by a thickening of the dorsal part of the lamina terminalis at the expense of the " corpora praecommissuralia." As new commissural fibres crowd in aud the commissure increases ia length and depth, this matrix becomes extended, but I believe it always continues to enclose the dorsal commissure, even in its most extended form. Paul Martin says tha in the cat the cephalic extremity of the corpus callosum or genu is formed by the apposition of fibres in front of those which are situated in the thickened lamina terminalis. It must be evident (if we glance at fig. 1) that if these apposed fibres grow in a " secondary fusion of the walls of the hemisphere," such a fusion must consist of a " soldering " of the two "precommissural hippocampi" {hip"). But such a " soldering " does not take place, because in the gyrus genicuU we have seen the representative of the " precommissural " hippocampus pushed before the advancing corpus callosirm. There seems to be no other .alternative which could be for a moment maintained but that the corpus callosum is contained ui its original matrix or " com- missure-bed," which has become enormously but gradually elongated. This conclusion is strongly forced upon us by comparison, in the order mentioned, of such a series of 10* 66 DR. G. ELLIOT SMITH ON THE cerebra as those of the Iguana, Ornithorhynclnis^ Phascol- arctos, Perameles, Nyetophilus, Dasypus, and Talpa, with more highly-developed forms. In this process of forward extension the dorsal com- missure has carried forward its oAvn matrix, derived from the original " commissure-bed." It is obvious that if this takes place withovit a corresponding growth of the bridge of grey matter {s) which lies between the dorsal and ventral commissiu'es, we shall have pai't of the great longitudinal fissure (botmded on either side by the precommissural area) roofed in by the corpus callosum (contained, of course, in its proper matrix). This roofed space will be bounded posteriorly by the ventral limb of the " commissura dorsalis." This enclosed part of the great longitudinal fissure is the so- called "fifth ventricle^'' or " cavitm septi pellucUli." By means of the accompanying five schemes (in which the commissure-bed is shaded) T have graphically represented different stages in this process. From these schemes it will be readily recognized how the backward extension of the dorsal commissure (fig. 8) stretches not only the matrix in which it lies but also the sheet of the " precommissiiral body " {p.a.) which fills up the large angle between the two limbs of this commissure. Thus a portion of the pre- commissural body of each hemisphere becomes one of the leaves of the septum pellucidura. In liis valuable memoir Paul Martin speaks of a fusion taking place (in the brain of the cat) between the opposed walls of the hemisphere on the ventral side of the corpus callosum. In tliis manner, in the more highly-developed macrosmatic Eutherian cerebra, the massive corpora praecommissuralia below the corpus callosum may meet and fuse in the median plane so as to obliterate the cavum septi. But the commissure itself is not situated in anv such secondary fusion of the hemisphere walls, so that it does not aff'ect the main argument advanced above. But though the coi-pus praecomraissui'ale thus takes an important share in the formation of the septum pellucidum, part of it always remains distinct in the higher mammalian cerebrum as a vertical band, lying parallel to the lamina terminalis and separating the latter from the general pallium. This band is the "gyrus subcallosus " of Zuckerkandl. In order to show the last stage which this process reaches, I have represented in PI. 16. fig. 18 this region of the cerebmm /="/c. -?- Fie. 5. nc.e. Fic. 7. F'<^ Figs. 4-8. — Schemes to explain the evolution of the " septum pellu- cidum." Fig. 4 is a type of a Reptile. „ 5 ,, „ Monotreme. „ 6 „ „ Marsupial. OEIGIN OF THE COEPUS CALLOSUM. 67 of an adult man. The anterior extremity of the corpus callosum has become so bent that its rostrum (r.) becomes continuous with the remains of the " commissure- bed," which still surrounds the ventral commissure (c.v.). In this way the septum pcUucidum {sept.) becomes surrounded by a complete ring of the matrix of the commissiu-es and cut off from the gyrus subcallosus, which is the remainder of the precommissural area. This gyrus subcallosus extends down to the base of the cerebrum immediately in front of the ventral commissure and lamina cinerea {I.e.). Its line of separation from the pallium is indicated in the figure by a dotted line {a.l.). The gyrus subcallosus is traversed by a well-marked vertical depression — the Jissm-a jirima {fiss.])!'.), — Avhich has no important morphological significance. With the knowledge which we have acquired of the process of phylogenetic development of the region of the commissures, it is extremely instructive to study the beautiful figures with which Marchand, in man, and Paul Martin, in the cat, represent {op. cit.) the process of ontogenetic development. In the Marsupial we have a fissura arciiata or hippocampi, extending from the tip of the temporal pole right round the mesial wall of the hemisphere towards the olfactory pedimcle ; so, in the foetal child or kitten, we find the Bogenf urche (wliich we might, with Mihalkovics, appropriately call " Ammonsfurche ") following a similar course and shading away towards the cephalic pole of the hemisphere. And it is necessary to remark, in passing, that the so-called part of the " vordere Bogenfurche," which His calls " fissui'a j)rima," has noticing whatever to do with the true Bogenfurche or fissura arcuata, if we regard the latter as the primitive fissura hippocampi. In the early stages of the cat, the lamina terminalis becomes thickened and invaded by the commissural fibres of the fornix (Paul Martin), so in phylogcny we have the corresponding stages in the adult Monotremes and Marsupials. Then, as the commissural fibres increase in number, the grey mass or thickening of the lamina terminalis is invaded by so many white fibres that the grey substance seems to disappear, but it is in reality being gradually extended by the swelling commissure. The backwardly-extending commissure produces exactly the same series of changes — the same hippocampal flexure — in our phylogenetic series as it does in the developing brain of the cat and man (see Martin and Marchand's figures). Thus, for oil the stages in the developing brain of the cat, we can find almost exact prototypes among the more lowly-developed mammals. The great feature which far more than any other distinguishes the mammalian brain from that of all submammalia is the possession of a definite pallium — -distinct alike in its histological features and its morphological relations — giving rise to a definite internal capsule of projection-fibres and well-defined and fully-medvillated commissural fibres. At first, in the Monotremata and Marsuj)ialia, this pallium (like the parent mass of the basal ganglion from which it ajipears to have sprung) is united to its homologue of the opposite hemisphere by means of the " commissura ventralis " — " the commissure of the cerebral hemisphere " par excellence. But the rapid growth in extent and complexity of this general cortex or pallium is accompanied by a richer and more abundant commissural system. This growing commissural system from the dorsal part of the enormous pallium not only finds in the 68 DK. a. ELLIOT SMITH ON THE " commissura clorsalis " a shorter path, but a clearer scope for longitudinal extension than the ventral commissure provides. And tlms we have pallial fibres invading and subsequently superseding the dorsal limb of what was previously the hippocampal commissure. " Pallial " commissural fibres probably first make their appearance in Eeirtiles and form a very insignificant constituent of the ventral commissure. In Monotremes and Marsupials these fibres become extremely abundant, and swell the proportions of the ventral commissure enormously. But in Eutheria a rapidly increasing proportion of these fibres forsake the commissura ventralis and form the new " dorsal commissure of the pallium " — the corpus callosum, — which throws the " parent " commissure into insignificance. The remnant of the commissura ventralis is known in man by the somewhat misleading name " anterior commissure." EXPLANATION OE PLATES 15 & 16. Fig. 1. Coronal section througli the forebrain of Peranicles nasuta. Stained by Weigert's method. X4. The section passes just behind the commissures. 2. Coronal section through the hippocampal formation and the posterior extremity of the " com- missura dorsalis^' in Peramehs. Wcigert stain. Fascia dentata represented as it appears after nuclear staining, x 20. Figs. 3 & 5. Coronal sections of the " commissura dorsalis " of Perameles, a short distance in front of fig. 2. X about fi. Fig. 4. Representation of the commissures of Pernmek's (as seen in a mesial sagittal section) to show the planes in which the sections represented in figs. 2 («), 3 {b), 5 (c), and 6 (d) w-erc cut. 6. A coronal section of the cerebral hemispheres of Peram,eles, to show the general arrangement of the commissures. Wcigert stain, x about 6. 7. Coronal section of the right cerebral hemisphere of Perameles immediately in front of the cerebral commissures. Weigert-Pal and carmine stains, x 4. 8. Semi-schematic representation of the anterior part of the mesial wall of the cerebral hemisphere of Dasyurus vivervimis. X about 4. To show that, as a result of the unrolling of the hippo- campus anteriorly, the whole of its morpholoyicully superficial layer now actually forms part of the visible surface. Thus the surface of the hippocampus proper [hip.'") appears above the fascia dentata [f.d.'). 9. Anterior portion of the forebrain of Perameles, exposed in a mesial sagittal section. x 4^. The bulbus olfactorius only represented in part and in outline. 10. Sagittal section through part of the anterior portion of the mesial wall of the cerebral hemisphere of Phascolarctos cinereus. Weigert stain. X G. 11. Horizontal section through the left hemisphere of T/'icAosf^n/s ?;M/yje'c«/«, a short distance above the commissures, x 2. ant.mcs., the antero-mesial angle. 12. Coronal section through the right hippocampal formation of a bat [Miniopterus Schreibersii) , to compare with tlie hippocampal region of fig. 1. Lithinm-carniine stain. x about 20. 13. Coronal section o? Mi/iiopterus. Compare fig. 2. 14. Coronal section of right hemisphere of Miniopterus. X about 20. Compare with fig. 6. 15. The fascia dentata and adjacent parts. Enlarged from fig. 14. 16. Coronal section of the liippocampal region of Miniopterus, a short distance in front of fig. 14. 17. Sagittal section of the commissiires and adjacent parts of brain of Nyctophilus near the mesial plane. X about 40. 18. Drawing of the " gyrus subcallosus " of a human bram. OEIGIN OF T]IE COEPUS CALLOSUM. 69 ExplaHation of reference-lettei The significance of tlic reference-letters a, h, c, d, a, ^, y, a. I. Anterior limit of the " gyrus subcallosus.''' /. ali\ Alveus hippocampi. ciny. Cingulum. I.e. ciny' . Longitudinalfibresin the alveus hipjjocanipi l.v. adjacent to the cingulum. n.a. C.I). " Commissura dorsalis." n.c. CD.' Dorsal limb of the " commissura dorsalis." n.f.d. c.d". Ventral limb of the iMetatherian type of n.l. " commissura dorsalis." o.b. c.e. Capsula externa. o/;/. c.f. Columna fornicis. o.tr. c.i. Capsula interna. p. cm. Commissura mollis. p.a. c.r. Corona radiata. est. Corpus striatum. PJ/-f- c.v. " Commissura ventralis." pyr. d.f. Uesceudiug hippocampal or fornix-fibres. e. Epithelial fold roofing the third ventricle. r.o. e.o.t. External olfactory tract. s. f.d. Fascia deutata. sept. f.h. Fissura hippocampi. s.^r. fi. Fimbria. s.l. fi. The remains of the fimbria intermingled s.m.t. with fibres of the commissura dorsalis. spl. fix. The " subsplenial hippocampal flexure." s.r. f.r. Fissura rhinalis. s.t. g.h. Ganglion habenulte (and stria meduUaris s.t'. thalami). hip. Hippocampus. s.t". hip' . " Supracommissural hippocampus." sub. hip". "Precommissural hippocampus." t. hip'". Stratum zonale of the " precommissural hippocampus" lying upon the surface im- t.o. mediately above the stratum zonale [vd v. 3. molecularej of the fascia dentata [f.d.'). x. i. " Indusium vermn." i.l.s. Internal limiting sulcus of the tuberculum y. olfactorium {t.o.). •s in tlie Jigitfcs. h is explained in the text. " Longitudinal association bundles " of the hi[i2i{icampus. Lamina cinerca. Lateral ventricle. Nucleus amygdalae. Nucleus caudatus. Nucleus fasciae dentatae. Nucleus lenticularis. Olfactorj' bulb. Optic nerve. Optic tract. General cortex or pallium. "Precommissural area" or "corpus prm- commissurale." Pyriform lobe. Layer of pyramidal cells of the hippo- campus. Recessus ojiticus. " Commissure-bed." " Septum pellucidum." Stratum granulosum of the fascia dentata. Stratitm lacunosum of the hippocampus. Stria medullaris thalami. Splenium of the corjjus callosum. Stratum radiatum of the hippocampus. Stria terminalis. The crossing fibres of the stria terminalis entering the " commissura ventralis." The uncrossed part of the stria terminalis. Subiculum cornu ammonis. "Temporal pole" of the cerebral hemi- sjjhcre. Tiiberculum olfactorium . Third ventricle. MeduUated nerve-fibres stratum lacunosum. " Tract us hippocampi ad pallium." going to the Elliot Smith Trans Linn Soc Zool See 2,Yol.V1I,Pl 15 e.o I. rHiitK.Lith'Edin' HIPPOCAMPAL REGION OF BRAIN IN PERAMELES AND DASYURUS. Elhol Smith Trans Linn Soc Zool Ser 2,Vol VII.Pl 15 FHuth,Lilh'Edii!' HIPPOCAMPAL REGION OF BRAIN IN MARSUPIALS AND CHEIROPTERA. LINNEAN SOCIETY OF LOiNDON. MEMORANDA CONCERNING TRANSACTIONS. The First Series of the Trausaotious, coutaining both Eotauical and Zoologieal contributions, has been completed in 30 Vols., and a few entire sets are still for sale. Only certain single volumes, or parts to complete sets, may bo obtained at the original prices. The price of the Index to Vols. 1-25 is S^. to the public, and Us. to Fellows ; to Vols. 26-30, 4s. to the public, and '.is. to Fellows. The Second Series of the Transactions is divided into Zoological and Botanical sections. The prices of the Zoological parts of these which have been published are as undermentioned (for the Botanical parts see Botanical wrapper) : — When Price to the Volume. published. Pi.blic. £ s. d. Second Series. — Zooi.oor. I. Part I. 1S75 1 4 0 . Part II. 1875 0 6 0 . Part III. 1876 1 8 0 . Part IV. 1877 0 10 0 . Part V. 1877 0 18 0 . Part VI. 1877 1 2 0 . Part VII. 1878 1 16 0 . Part VIII. 1879 1 0 0 . II. Part I. 1879 1 4 0 . Part II. 1881 0 15 0 . Part III. 1882 1 8 0 . Part IV. 1882 0 7 6 . Part V. 1882 0 3 0 . Part VI. 1883 1 0 0 . Part VII. 1883 0 5 0 . Part VIII. 1883 0 3 0 . Part IX. 1883 0 3 0 . Part X. 1884 0 4 6 . Part XI. 1884 0 10 0 . Part XII. 1885 0 0 0 . Part XIII. 1884 0 6 0 . Part XIV. 1S85 0 6 0 . Part XV. 1S85 0 4 6 . Part XVI. 1885 0 5 0 . Part XVII. 1886 0 3 0 . Part XVIII. 1SS8 0 2 6 , III. Part I. 1884 1 14 0 Part II. 1884 1 12 0 Part III. 1885 1 10 U . Price to Fellows. £ s. d. 0 18 0 Vol 1111 When Publishetl. Price to the Public. Price to Fellon-9. 0 4 6 1 1 0 0 12 0 0 13 6 0 16 6 1 7 0 0 15 0 0 IS 0 0 11 6 1 1 0 0 5 6 0 •) 3 0 15 0 0 3 9 0 2 3 0 2 3 0 3 6 0 7 6 0 4 6 0 4 6 0 4 6 0 3 6 0 3 9 0 2 3 0 2 0 1 5 6 1 4 0 1 2 0 £ s. d. Second Series. — Zoology [confimicd). III. Part Part Part IV. Part Part Part Part IV. 188.5. V. 1887. VI. 18t^8. I. 1886. II. 1887. Part III. 1888. I. 1888. 11. 1888. Part III. 1889. Part IV. 1890. Part V. 189U. Part VI. 1891. Part VII. 1891. PartVril. 1892. Part IX. 1892. Part X. 1893. Part XI. 1894. VI. I. 1894. II. 1894. III. 1894. IV. 1896. V. 1890. VI. 1896. Part VII. 1896. Part VIII. 1897. Part Part Part Part Part Part VII. Part I. 1890. Part II. 1897. Part III. 1897. 8 0 0 0 0 1 4 1 8 0 16 0 12 0 5 1 7 0 12 0 6 0 12 0 0 8 0 12 18 0 0 2 8 0 6 0 4 0 8 0 0 0 0 0 0 0 0 6 0 0 0 2 0 1 11 0 10 1 4 0 10 0 0 8 0 12 0 2 6 0 0 0 0 (I 10 0 0 12 0 0 0 0 0 6 0 0 0 0 0 4 6 0 18 0 1 1 0 0 12 0 0 9 0 0 3 9 1 0 0 0 9 0 4 0 6 0 9 0 0 4 6 0 6 0 0 9 0 1 1 0 0 2 0 1 10 0 13 3 0 7 6 0 18 0 0 7 0 0 6 0 0 9 0 0 2 0 0 7 6 0 9 0 0 4 6 2nd Ser. ZOOLOGY.] [VOL. VII. PART 4. AUG 17 1893 THE io^lL TRANSACTIONS OF THE LINNEAN SOCIETY OE LONDON. ON THE MUSCULAR ATTACHMENT OF THE ANIMAL TO ITS SHELL IN SOME FOSSIL CEPHALOPODA (AMMONOIDEA). BY GEORGE CHAELES CRICK, F.G.S., F.Z.S., OF THE BRITISH MUSEUM (NATURAL HISTORY). {Communicated by the President, Dr. A. Gunther, M.A., F.R.S.) LONDON: PRINTED FOR THE LINNEAN SOCIETY BV TAYLOR AND FRANCIS. RKI) LION COURT, FLEET STREET. SOLD AT THE SOCIETY'S APARTMENTS, BURLINGTON-HOnSE, PICCADILLY, W. AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW. ^ June 1898. AUG 17 1893 r 71 IV. On the Mnscnlar Attachment of the Animal to its Shell in some Fossil Cephalo])oda [Ammonoidea). By G. C. CmcK, F.G.S., F.Z.S., of the British Iliiseum [Natural Sistory). [Communicated l»j the Piiesident.) (Plates 17-20.) Read 3rfl February, 1S!)8. INDIC^ITIONS of the muscular attachment of the animal to its shell have been tigured in not a few Nautiloids, but comparatively few Ammonoids have been recorded in which what has been believed to be the remains of this attachment has been iigured and described. Of these the best known are Oppel's figures of three examples of Ammonites steraspis* from the Lithographic Stone of Bavaria, iiublished in 1863, in which the body-chamber of each is shown to be traversed by a fine curved line, the relation of which to the animal was not, however, indicated by the author. In 1870, Trautschold f figured a specimen of Ammonites hicurnatus, exhibiting what he considered to be the impression of the muscular attachment of the animal, but this figure differs considerably from Oppel's figiu'cs. In the following year, Dr. W. Waagen % published his important paper, " Ueber die Ansatzstelle der Haftmuskeln beim Nautilus und Ammonoiden," in which he expressed his opinion that the " shell-muscle " in the Ammonoidea was attached to the inner (umbilical) portion of the lateral area of the whorl. He believed the curved line figured by Oppel on the body-chamber of Ammonites steraspis to be a trace of the " annulus," and probably also of the shell-muscle, and, reproducing one of Oppel's figures, he completed by a dotted line what he considered to be the form of the shell- muscle. This interpretation of Oppel's figures has been accepted by most authors. In 1879, Eck § figured and described a small septate fragment of Ceratites semipartitus from the Upper Muschelkalk of Schwielierdingen, near Stvittgart, Avhich he thought showed not only the appearance of the surface of the mantle (the so-called " epidermids "), but also the impression of the annulus. The specimen was merely an internal cast of five chambers, and on the surface of the cast of each chamber there was a depressed zone with a finely-pitted surface, occupying on the siphonal region the middle two-thirds, and becoming much narrower on the side of the whorl ; in two of the chambers a groove was also present on the antisiphonal area. * Pal. Mittheil. p. :2.j1, pi. Ixis. figs. 1, 2, & 6 (1863). t Bull. Soc. Nat. Moscou, vol. xliii. pp. 301-306 (1870). t Pateontographica, vol. xvii. pp. 185-210, pis. xxsis. & xl. (1871). § Zeitschr. deutsch. geol. Gesell. vol. xxxi. p. 276, pi. iv. figs. 5 a-d (1879). SECOND SERIES. — ZOOLOGY, VOL. VII. 11 72 MR. G. C. CRICK OX THE MUSCULAR ATTACHMENT OF THE In his lai'ge works on the Triassic Cephalopoda, Dr. E. von Mojsisovics * has figured several species of Ammouoids bearing on the internal cast of their body-chamber a groove or grooves (extending in some examples from the umbilicus on one side, over the peripliery, to tbe umbilicus on the other side), vrhicli he considered to be the impression of the homologue of tiie muscular attachment of the recent Nautilus. Dr. O. Jaekel t, in 1889, figured a Ceratite from the Trias of Riidersdorf, near Berlin, bearing a groove precisely similar to those figured by Dr. E. vou Mojsisovics, and although he douljted this indicating the position of the homologue of the annukis and of the shell-muscle in the recent Nautilus, he was unable to give any explanation of its nature. After an examination of Oppel's original specimen, this author says the line which Oppel figured and which has been regarded as indicating the position of the anterior boundary of the anuulus and of the shell-muscle can scarcely be folloAved with certainty, and he is inclined to doubt tiie correctness of the interpretation. In his ' Vorlaufige Mittheilung liber die Organisation der Ammoniteu ' +, Dr. Stein- mann evidently does not agree with Dr. Waagen's interpretation of Oppel's specimens (although he seems to make no special reference to them), a fact which is clearly brought out in the ' Elemente der Palaontologie ' (1890) by himself and Doderlein, where (p. 351, fig. 402) one of Oppel's figures (Pal. Mittheil. pi. Ixix. fig. 2) is reproduced, and tbe curved line on the body-clianiber completed in the manner suggested by Waagen, but in the explanation of the figure this line is thus described : " (?) vordere Greuze des Haftbandes." At the meeting of the Geological Society of London which was held on March 25th, 1891, a communication was read from Mr. S. S. Buckman, entitled " Notes on Nautili and Ammonites." Only an abstract of the paper was published §. Nautili and Ammonites were exhibited in illustration of the paper, and, according to the abstract, " Two specimens exhibited show long spatulate depressions more or less parallel to the periphery for about half the length of the body-chamber. It was suggested that these impressions indicated the position of the shell-muscles." So far, then, as I have been able to ascertain, no satisfactory examples exhibiting the form and position of the musctdar attachment of the Ammonoid animal to its shell have yet been figured and described either to confirm or to contradict the explanation whicli AVaageu gave of Oppel's figures. The Ammonite animal may have been, and prol)ably a\ as, attached to its shell at the edge of the last septum, as in the living Nautilus, but from my own observations it is * Abhaudl. d. k.-k. geol. Eeichsanst. Wieii, Bd. vi. Theil i. Heft i. (1873) pi. xvi. f. 3 {Phi/lloceras occidfum) ; pi. xix. f. 1 (Pinucoceras transiens) ; pi. xix. ff. '2, 3, 4 & pi. xx. ff . 8 &: 9 {Pmacoceras Jiumile) ; pi. xx. ff. 2, 3, 5, & 7 (Pivxicoceras insectum) ; [il. xxii. ff. 7, 8 {Pi nacoceras mi/ophorum) ; ibid. Bd. vi. Hiilf'te ii. (1893) pi. cxxxiv. f. 1 {Choristocei'fis amiiwnh'ifornu); ihUl. Bd. x. (1882) pi. liii. f. 2 (Mec/aj}hi/lKtes sandalinus); pi. liii. f. ;5 {Mega- phyllites oholus). t " Ueber eineu Ceratiteu aus dem Schaumkalk vou lliider.sdorf uud iiber gewisse als Haftring godeutete Eindriicke bei Cephalopoden," Ncues Jahrb. 18S9, ii. p. 19, pl.i. + Berichtc der naturforsehendcu Gesellschaft zu Freiburg, Bd. iv. Heft 3, jip. 31-47 (1889). § Abstr. Proc. Geol. Soc. Loudou, Session 1890-91, p. 105 (Quart. Journ. Geol. Soc. vol. xlvii.). ANIMAL TO ITS SHELL IN SOME POSSIL CEPHALOPODA. 73 quite clear that it was furnished with shell-muscles and an annulus like the recent Nautilus, and it is the object of the present communication to record the indications of these structures in various Ammonoids. It is not proposed here to record every Amnionoid in which these imj)ressions have been observed — this I hope to be able to do subsequently — but to describe the form and position of these impressions so far as I have been able to observe them in the different forms which the Ammonoids assume, e.g. in BacnUtes, Hamitcs, Crloceras, Ancylooeras, Macroscapliites, Scaphites, and Tiirr/lHes, the group of the Ammonites (ranging from very evolute to almost entirely involute forms), as Avell as in Chjmenia and the group of the Goniatites. It may be well at the outset to refer to the indications of the shell-muscles and of the annulus as they exist in the shell of tlie recent Nautilus, and for this purpose it will probably suffice among the many figures which have been published of the m.uscular attachment of the I'ecent Nautilus to call attention to the figures accompanying Dr. Waagen's paper already alluded to (p. 71)*. I may, however, here remark that in the recent Nautilus the shell-muscles are ear-shaped and situated upon each side of the animal ; they are connected both on the dorsal and on the ventral side by a narrow band — the annulus. The shell-muscles are not inserted into the shell-substance, but are merely applied to the innei* surface of the test with the intervention of a thin layer of conchiolin ; and all that is usually preserved in the interior of the shell to indicate the form and position of the muscular attachment is a fine, generalh' raised line, corre- sponding to the anterior boundary of the annulus and of each shell-muscle; it is only rarely that there are any indications of the posterior boundary of these structures. On an internal cast this raised line would be represented by an incised line, and since in fossils the remains of the muscular attachment are preserved almost always on internal casts, they therefore exist as incised lines. Such structures have been described and figured in not a few fossil Nautiloids, including the genus Nantilns itself. In the fossil forms any recoi'd of the form and jiosition of the muscular attachment would, when present, usually be preserved iipon the surface of the natural internal cast of the body- chamber, and hence raised lines on the inner surface of the original test would appear on the internal cast as incised lines, and clce versa. In order to observe in the recent Nautilus the exact position of the muscular attach- ment with relation to the edge of the last septum, an artificial cast of the shell ot Ndiitilvs pompilius w^as made by filling a sagittal section of a recent shell with paraffin wax, and then dissolving aw'ay the shelly matrix with hydrochloric acid. The anterior boundary of the muscular attachment alone was indicated by a very finely incised line. To show the usual position of the muscular attachment in Ammonoids, and for the better understanding of the less perfectly preserved examples, it is proposed first to describe the impressions of the " muscular scars " in the specimen on which they have been most clearly seen, and then the specimen in which the remains of the annulus have been most clearly observed. The former is an example of Crloceras from the * See also the receiitlj'-published paper by L. E. Griffin, " Notes on the Anatomy of N(n(tilu.s pomjnlncs," Zoo]. Bull. vol. i. no. 3, pp. 147-161 ; with bibliography. 11* 74 MR. G. C. CKICK ON THE MUSCULAR ATTACHMENT OF THE Speeton Clay, and the latter an Oxfordian Ammonite, now referred to the genus Cardioceras. Then it is proposed to descrihe the form and position of the scars and of the annulus (when seen) in various forms of Ammonoids, ranging from the straight (in the adult) from Bacidites, through Uamites, Crioceras, Ancyloceras, IlacroscapMtes^ Scaphites, and Tnrril'des to the group of the Ammonites ; then in Clymenia and the group of the Goniatites. The A^ery fragmentary character, and consequently imperfect determination, of some of the specimens must he mentioned; hut, hearing in mind the position of the shell- muscles in the Ammonoid, viz. on the inner portion of the whorl, it will he at once evident that the traces of the muscular attachment are more likely to he found in fragmentary specimens, principally on natural internal casts of the body-chamber. Muscular Scars. Tlie muscular scars have been best observed in a fragment of Crioceras qnadratnm *, n. sp., which I obtained from the Speeton Clay of Yorkshire (PL 17. figs. 10, 11). It consists merely of the base of the body-chamber, and is about 38 mm. long. The greater part of the test, which was in a very soft, white, friable condition, has been removed by a stiff brush, so as to expose the surface of the internal cast of the body-chamber without scratching it. The section of the whorl is subquadrate, the outer area being somewhat narrower than the inner ; at the last septum the height of the whorl is 29 mm., and its greatest thickness 27 mm. On the inner (dorsal) area of the internal cast, and near the posterior extremity of the body-chamber, there are two oval areas, one on either side of the median line, the major and transverse diameters of etich being 14 mm. and 11 mm. respectively. The longer diameter of each is placed transversely, but not quite at right angles to the median line, the inner end of this diameter being slightly in advance of the outer extremity. The two impressions ai*e nearly in contact, being only about 05 mm. apart at the inner or anterior end of their respective longer diameters ; their posterior borders are only slightly in advance of the last septum. Each impression is very slightly roughened, and consequently distinct from the surrounding and somewhat polished siu'face of the cast. An incised line forms its boundary on the inner, antei'ior, md outer sides, being deepest on the anterior side, while its posterior boundary is marked by a faint, somewhat irregular and imperfectly-defined line. No trace of the portion of the annulus surroimding the base of the body-chamber, and connecting the muscular impression on the one side Avith the muscular impression on the other side, can be seen. In the triangular space between the two impressions, and 3 mm. posterior to the point where they are nearest together, there is near each incised line a veiy small, shallow, double pit, each pair of pits being disposed longitudinally ; these doubtless were connected with the muscular attachment of the animal to its shell. * This is really a Bean MS. name. Some of the examples of this species in the British Museum Collection bear this name in Bean's handwriting, and I have adopted tlic name, since, so tar as I know, the species has not j'et lieen described. ANIMAL TO ITS .SHELL IN SOME FOSSIL CEPHALOPODA. 75 Animlus. The impression of the annuius lias been observed in several specimeDS, but in the British Museum Collection (No. C. 6801) there is an example of Cardioceras excavatimi (J. Sowerby) * from the Oxford Clay, but the locality is not recorded, in whicli the form of the miiscular attachment, and particularly that of the annuius, is remarkably well displayed (PI. 18. fig. 8). The specimen is apparently complete, and shows the aperture of the shell, which is provided with a narrow, ventral (or peripheral) apophysis. Its dimensions are : — Diameter of shell 120 mm. ; width of umbilicus 21 mm. ; height of outer whorl 58 mm. ; ditto above preceding- whorl 40 mm. ; thickness of outer whorl about 43 mm. The body-chamber occupies the last half of the outer whorl, its base being 44 mm. high and 29 mm. wide or thick. The whorl is subsagittate in transverse section ; its inner margin slightly overhangs the umbilicus, so that the width of the umbilicus, measured at the umbilical margin, is less than that measured at the suture of the shell. A greater part of the test is preserved, and this is mostly in a soft friable condition. Some of it had been removed from the left t side of the specimen, so as to expose a portion of the internal cast of the base of the body-chamber. By means of a small stiff brush the rest of the test was carefully cleared from both sides of the base of the body-chamber, and the form of the muscular attachment of the animal was displayed very clearly, especially on the left side, the attachment on the right side being jjreciselv similar, but a little less distinct. On the left side the impression of the anterior boundary of the shell-muscle crosses the umbilical margin 5 mm. in advance of the last septum, and passes as a very narrow l)and of a. thin film backward and outward in a flat forwardly- convex curve as far as the middle of the second lateral lobe, at which point, and 1"5 mm. posterior to it, there is seen to be another similar band Avhich passes from the mnbilicus immediately above the saddle on the margin of the umbilicus ; these two bands, con- tinuing at abou.t the same distance apart, pass over the lateral saddle into the first lateral lobe and, then diverging slightly, pass over the peripheral or external saddle forward and outward towards the jjeriphery in a flat forwardly-convex curve, but becoming very indistinct before reaching the periphery. These two very narrow filmy bands appear to l3e the remains of the anterior and posterior boundaries of the annuius respectively. Remains of the annuius are visible also on the right side, but much less distinctly ; they are, however, sufficiently clear to confirm the structures which have been mentioned as existing on the left side. Having described the usual form of the muscular scars and of the annuhis, I now proceed to describe their remains in various forms of Ammonoids, commencino" with the genus Bactilites and proceeding in the order already indicated {ante, p. 73). * J. Sowerby, Min. Con. voL ii. p. 5, pL cv. (1815). t The terms " right " and " left " are used in a strictly morphological sense, the periphery (or siphoual area) of au Amnionoid being ventral, and the anti])eripheral (or autisiphonal) area dorsal. 76 MR. G. C. CRICK ON THE MUSCULAR ATTACHMENT OF THE Order of Descriptions. Page Baciilifes, Lamarck 77 Samites, Parkinson 79 Crioceras, Leveillo 79 Aiicyloceras, d'Orbigny 80 Macroscaphifes, Meek 81 Scaphites, Parkiuson 81 Turrilites, Lamarck 83 [Metcroccras, d'Orbigny) 84 AMMONPrES 85 AmaltheidtE, Pischer 86 Oxi/)wticeras, Hyatt 85 Amalthens, Montfort 86 Cardioeeras, Neumayr & Uhlig 86 Neumayria, Nikitin 89 Tissotia, Douville 90 Lytoceratid^e, Neumayr, emend. Zittel. Lytoceras, Suess 91 -Sgocekatid^, Neumayr, emend. Zittel. Arietites, "Waagen 93 ^goceras, Waagen, emend. Zittel 95 Sonninia, Bayle 98 Harpoceratid^, Neumayr, emend. Zittel. Secticoeeras, Bonarelli 98 (Ecotranstes, Waagen 99 DisticJioceras, Munier-Chalmas 100 Stephanoceratidje, Neumayr, emend. Zittel. Stephanoceras, Waagen 101 Pei'isphi notes, Waagen 102 Aspidoceratid-E, Zittel. T?eltoceras, Waagen 102 Aspidoceras, Zittel 103 CoSMOCERATIDiE, Zittel. I'arkinsonia, Bayle 103 Prionotropip.e, Zittel. Schlcenbachia, Neumayr 104 Clymenia, Miinster . 105 GONIATITES. Glyphioceras, Hyatt .... 106 ANIMAL TO ITS SHELL IN SOME FOSSIL CEPHjVLOPODA. 77 Baculites, Lamarck. Baculites ovatus, Say. — An example of this species in the British Museum (No. C. 5415 f/) from the Fox Hill beds (Upper Cretaceous) of Horse-head Creek, South Dakota, U.S.A., displays the impression of the shell-muscles very clearly. The specimen consists of the natural cast of nearly the whole of the body-chamber, the cast of the last locvilus and of a portion of the penultimate loculus. The test has been almost entii'ely removed from the posterior part of the body-chamber (PI. 17. figs. 1, 2, ifc 3). The shell in this species is laterally compressed, and tapers very slowly. The length of the specimen is 290 mm. or abont 11^ inches ; its transverse section is oval, the greatest thickness being a little nearer the antisiphonal (dorsal) than the siphonal (ventral) area. Of this length the body-chamber occupies 215 mm., but the aperture is not preserved. The ventro-dorsal and transverse diameters of the base are 38 and 25'5 mm. respectively. The two muscular impressions are at the base of the body-chamber and on the antisij)honal (dorsal) area ; the inner, antei'ior, and outer portions of their botmdaries can be distinctly traced as a faint, shallow, depressed (not sharply-incised) line. On the antisiphonal area the suture-line has a small antisiphonal lobe separating the two portions of a broad saddle, each portion corresponding to the second lateral saddle in an ammonite ; each of these halves is followed by a rather broad lobe (the second lateral) which separates it from the saddle occupying nearly the middle of the lateral area, L e. the first lateral saddle. The median line of the antisiphonal area is occupied by a very shallow longitu- dinal groove. The inner boundaries of the two muscvilar impressions seem to meet in the middle line at about 2'5 mm. in advance of the most anterior part of the saddle, adjoining the antisiphonal lobe. Starting from this point, the boundary of the impression passes forward and outward until it is S"5 mm. from the same j)art of the suture-line ; then, turning backward and still maintaining its course outward, it passes along the outer side of the second lateral lobe, close to the inner side of the first lateral saddle. The boundary of the other impression has a precisely similar course. The impressions are somewhat oval, their longer diameters making an angle of about 45 with the median line of the antisiphonal (or dorsal) area. At a point slightly below the level of the most anterior part of the adjoining saddle (the first lateral) the outer boundary of each impi'cssion is a little angular. On one (the right) side of the specimen no trace of the annulus can be seen, but on the other (the left) side a line is seen to pass from the angular portion of the impression outward and upw'ard over, and about 1 mm. distant from, the adjoining saddle ; although it can be traced only for a short distance, owing to the roughness of the siuface here, it doubtless represents the anterior boundary of a portion of the annulus. At a distance of 5 mm. in front of the boundary of the impression on the right side, there is another line having precisely the same curvature. It is not nearly so distinct as the one just described, but most probably indicates the anterior boundary of the same shell-muscle. Possibly the posterior line denotes the position of the shell-muscle diu'ing the formation of a septum, and therefore dui'ing a period of rest, when the muscle would 78 ME. G. C. CRICK ON THE MUSCULAE ATTACHMENT OF THE become firmly attached to the shell. The anterior faintly-impressed line may have been the last attachment of the anterior boundary of the shell-muscle during the gradual growth of the animal upward prior to the formation of a new septum. The corresponding line on the left side is obscured by the shelly matter adhering to the cast. On the median portion of the peripheral area the septum j^ossesses a rather broad (median) saddle, on either side of which is a small lobe — the two halves of the peripheral lobe. This is followed by the peripheral or external saddle. Anterior to the septum and on the central portion of the same area there is a feebly -incised line having the form of half an ellipse (PL 17. fig. 3) ; it arises close to the outer portion of the external (or peripheral) saddle, i. e. at about 7 mm. from the median line of the siphonal (ventral) area ; then curves upward and towards the median line, attaining its greatest height at about 8'5 mm. in front of the median saddle, or about 5'5 mm. from its commencement. The line seems then to pass on to the other side without interruption, but the surface of the cast is not sufiici- ently well preserved to enable this to be stated w4th certainty. Its anterior portion is about on the same level as the anterior portion of the boundary of the muscular impression. About 6 mm. in front of this curve there is a similarly-curved feebly-impressed line, w'hich inost probably is comparable with the faint curve on the opposite (antisiphonal) surface. Although the annulus is not well shown in this example, there is another specimen [C. 5415 Jj] in the British Museum, from the same horizon and locality (PI. 17. fig. 4), which displays the muscular impression less distinctly, but clearly shows the anterior boundary of a portion of the annulus. This leaves the muscular impression at the " angle " mentioned in the previous description, and, passing upward over the adjoining saddle at a distance of about 1*5 mm. from the siiture-line, crosses the next lobe in a shallow depression, arid again rises over the next saddle iit about the same distance from it as before. This depression, however, is seen only with difficulty and by turning the specimen about in a fairly good light. There can, I think, be no doubt that the annulus was in the form of a simply-w^aved band, being elevated at the saddles and very feebly depressed in each lobe just as in the example of Cardioceras excavatum already described (p. 75). The total length of the specimen is 245 mm., the body-chamber (the aperture of which is not preserved) occupying 165 mm., the diameters of the base of tlie body-chamber being 44 and 31"5 mm. respectively. The other side of the cast of the body-chamber is obscured by fragments of the test. BacHlites vagina , Porbes. — A portion of the muscular impression has also been seen in an example of this species in the British Museimi Collection [one of the specimens numbered 83624] from the Upper Cretaceous of Pondicherry, India (PI. 17. fig. 5). It is merely the internal cast of the greater portion of the body-chamber, about 60 mm. long, anteriorly incomplete, but fairly perfect posteriorly. Its transverse section is oval, the diameters of its antei-ior end being 24 and 145 mm. ; those of the posterior end being 19 and 12-5 mm. The antisiphonal surface is broad and slightly flattened, the siphonal being narrow, flattened, and with suliangular borders. On the broad antisiphonal svu-face the feebly convex boundaries of the two muscular scars (indicated by a feebly-incised line ANIMAL TO ITS SHELL IN SOME FOSSIL CEPHALOPODA. 79 on the internal cast) meet nearly in the middle line in an obtuse backwardly-directed point whicli is 3*25 mm. in advance of the saddle on either side of the antisiplional lobe, these two saddles occupying a large portion of the autisiphonal area. From this point each Ijoundary passes forward and outward for a short distance, then turns backward and sweeps over on to the lateral area in a broad anterioi'ly-convex curve whicli gradually disajipears before reaching the last septum ; if continued to the septum it would meet the large lateral lobe on its antisiphonal side. It would seem tlierefore that, just as in the Baculites already described, the muscular scar on either side occupied the space between the central line of tiie autisiphonal area and the antisiphonal side of the large lateral saddle. There is no trace of the annulus in this example. Hamites, Parkinson. Samites viaximus, J. Sowerby *. — In the British Museum Collection (No. C. 6S02) there is an example of this species, from the Gault of Folkestone, which displays the form of the muscular scars exceedingly well (PI. 17. figs. 0, 7). It is only a fragment, /. e. one portion (the larger) of the terminal hook, and consists of the internal cast of the last loculus and of a portion of the body-chamber ; its length, measured on the pcrij)hery, is about 55 mm. The whorl is subcircular in section, only slightly compressed, its ventro-dorsal diameter (excluding the ribs) at the base of the body-chamber being 16 mm., and its transverse diameter (also excluding tlie riljs) 15 5 mm. Tlie impressions of the muscular scars are seen on the inner (dorsal) surface of the cast of the base of the body-chamber, one on either side of the median line and disposed souieAvhat oldiquely, just as in the exara2:>le of Crioceras quadraium already described in this paper {ante, p. 74). Each scar is elliptical in outline, and enclosed by a very faintly-incised line ; its longer axis is inclined to the median line of the dorsal surface of the body-chamber at an angle of about 45°; its posterior boundary is only about 1"5 mm. anterior to the antisiphonal saddle of the last septum. Anteriorly the impressions are 1 mm. apart. The longitudinal and trans- verse diameters of each scar are 12 and 10 mm. respectively. At about the middle of the outer portion of the boundary there is a slight angularity, which doubtless indicates the position of the commencement of tlie anterior border of the annulus, but no other traces of the annulus have been observed. In another example which is figured (PI. 17. tigs. 8, 9) each muscular scar bears a small roughened patch, and there are indications of a narrow band connecting the scars. Crioceiias, Leveille. Crioceras quadratumf, n. sp. — The muscular impressions in this genus have been best observed in two examples of Crioceras quadratam from the Speeton Clay, Yorkshire. One of these has already been described (see p. 74). The other forms part of the British Museum Collection (No. 89102). It consists of the natural cast of the body-chamber with * J. Sowerby, Min. Con. vol. i. p. 138, pi. lxii.(1814). t See footnote, ante, p. 74. SECOND SERIES. — ZOOLOGY, VOL. VII. 12 so ME. G. C. CEICK ON THE MUSCULAE ATTACHMENT OF THE only fragments of the test adliering to it. At the base of the body-chamber the whorl is subquadratein section, its height being 35 mm., and its greatest width 31 mm. (PI. 17. figs. 12, 13). The imi:)ressions are not so complete as in the example just described. The anterior margin of each is indicated by a sharply-incised line ; both lines almost meet each other in the median line of the inner (dorsal) area at about 10 mm. anterior to the summit of the innermost portion of the saddle on either side of the antisiphonal lobe. Dis- appearing posteriorly, each passes forward and outward for a length of about 10 mm., when each is about 5 mm. in advance of its origin ; then, turning abruptly backward, each gradually disappears at a distance of about 9 mm. from its point of origin and about 14 mm. from the median line. The posterior extremity of each has a tendency to turn outward, and at a short distance from the extremity the line gives off a very faint branch wdiich turns inward, but can be traced only for a short distance. Without doubt the faint line turning inward represents the boundary of the impression of the shell-muscle, and that continuing in an outward direction indicates the position of the anterior border of the adjacent portion of the annulus. Ancyloceeas, d'Orbigny. Ancyloceras Matheronimmm, d'Orbigny. — In the genus Aucyloccras what is be- lieved to be the impression of the mviscular attachment of the animal has been best seen in an example of Ancyloceras Matlieronimmm in the British Museum Collection (No. 0. 5322). It is a natural internal cast from the Neocomian of Cassis (Bouches-du- Pi,h6ne), France (PL 17. figs. 14, 15, 16). The coiled part of the septate portion is wanting, but the rest is fairly well preserved, although much waterworn. The body-chamber, measured along the centre of the periphery, is 380 mm. long ; the cross section of its base is oval, the dorsal (or inner) portion being somewhat wider than the ventral (or outer) ; the veutro-dorsal diameter (or height) of the base is 75 mm., the transverse diameter {i. e. the thickness) 54 mm. (excluding the tubercles). What I regard as the impressions of the boundaries of the two muscular scars are seen on the inner or dorsal area. The two scars are not quite symmetrical, that on the right side extending about 15 mm. further forwards than that on the left. These boundary-lines are incised, but they are much coarser than usual. This, however, is probably due to the -waterworn condition of the specimen, for the septal sutures, instead of being iinely -incised lines as usual, are fairly deep and wide, and indeed have quite the appearance of having been considerably water- worn ; and this condition of the septal sutures suppoi'ts the opinion that the lines about to be described are really the impressions of the boundaries of the muscular scars. That on the right commences at about 3 mm. to the right of the median line and about 20 mm. in front of the large saddle adjoining the antisiphonal lobe ; passing forward for about 35 mm. and approaching nearer the median line, it then turns outward, but still continues forward until at about 50 mm. from the last septum, when it turns still more outward and then disappears. That on the left appears to arise about 43 mm. in front of the miost anterior portion of the saddle adjoining the antisiphonal lobe, and about 6 mm. from the boundary on the right side ; passing forward and outward for nearly ANIMAL TO ITS SHELL IN SOME FOSSIL CEPHALOPODA. 81 20 mm., it is then at about 58 mm. from the last septum ; then with a broad curve it turns backward until it is about 40 mm. from its commencement, when it appears to rapidly die out. Although the impressions of these boundaries may perhaps be more properly termed grooves, their form and position agree so well with what I have observed in somewhat similar forms that I think there can be no doubt as to their nature. I have not been able to observe any trace of the annulus in this example. Mackoscaphites, Meek. Ifacroscophites fjigos, J. de C. Sowerby, sp. — In the genus Ilacroscaphites what 1 regard as the anterior boundary of the muscular scar has been observed in an example of Ilacroscaphites gigas, J. de C. Sowerby, sp., in the British Museum Collection (No. 32008). The specimen is stated to be from the Lower Greensand of the Isle of Wight, but its matrix and state of preservation suggest rather the Kentish Rag (Lower Greensand) in the neighbourhood of Maidstone (PI. 17. figs. 17, 18, 19). It is a much-compressed internal cast ; the greatest diameter of its sej)tate jjortion is 180 mm. ; the length of its body-chamber measured along the centre of the periphery and over the coarse ribs is 350 mm., the height of the base of the body-chamber is 68 mm., the thickness (including the ribs) being reduced by compression to 22'5 mm. On the right side, at about 15 mm. above the most anterior part of the last septum and 20 mm. irom the inner margin of the whorl when viewed laterally, a very fine incised line arises and passes thence as a flat arc forward and towards the inner margin, which it crosses at about 45 mm. in front of tlie last septum ; it then curves backward and comes to within about 10 mm. of the septum, w^here it is in the middle line of the compressed and somewhat distorted dorsal area. From this point another line, making an acute angle with the line just described, j)asses forward and outw^ard for rather more than 30 mm., when it curves outward still more and then disappears. These two curved lines we take to be the anterior boundary of the right and left muscular scars respectively ; no traces of the annulus have been observed in this example. The material of this natural internal cast is very coarse, and the specimen is so much crushed that the very faint lines bounding the muscular scars can only be followed with difficulty ; the boundary is preserved partly as an incised line and partly as a line of colour. ScAPHiTES, Parkinson. Scctphites biuodosus, A. Roemer. — The muscular impression can be traced in an example of this species contained in the British Museum Collection (No. C. 5182). It is from the Lower Senonian (Granulaten-Kreide) of Broitzen, near Brunswdck. The specimen is a fairly well-preserved, but somewhat distorted internal cast. Its greatest length is 43 mm., and the greatest diameter of the septate portion is 26-5 mm. (PI. 17. figs. 20, 21). At the posterior end of the body-chamber the whorl is 13'5 mm. higli and 10"5 mm. thick (excluding the tubercles). The whorl is somew'hat crushed obliquely, so that its sloping- inner area is much more clearly seen on one (the left) side. On this side a feebly-incised (partly double) line arises almost close to the last septum, and at a distance of 5 mm. 12* 82 MK. G. C. CRICK ON THE MUSCULAR ATTACHMENT OF THE from the inner edge of tlie Avliorl ; passing fovAvard in a broad, flattened, forwavdly- convex cui've, it graduall}^ approaches and finally crosses the inner edge of the whorl at about 7*5 mm. in front of the last septum. Passing on to the dorsal (or concave) portion of the whorl, it turns backward, and in a rather broad forwardly-convex curve nearly reaches the centre of this area, where it appears to be joined by the corresponding impression on the opposite side. The boundary of the impression on the opposite side is not quite so distinct, and on the sloping inner area of the whorl it appears to be a little nearer the inner edge of tlie whorl than on the opposite side, but this is doubtless due in great measure to the oblique crushing to which the specimen has been subjected. At the base of the body-chamber it is only 2-75 mm. from the inner edge of the whorl; passing forward it gradually approaches and finally crosses the same at about 7'5 mm. in advance of the last septum. I'assing thence in a forwardly-convex carve, it joins its fellow at about the centre of the dorsal surface of the whorl. The lines here described are probably the boundaries of the muscular scars. I have not seen any indications of the annulus in this specimen. ScapJdtes ceqiialis, J. Sowerby. — The muscular attachment of the animal has also been observed in an example of Scaphites (eqiialis, J. Sowerby, in the British Museum Collection (larger of the two specimens, No. S9113). The specimen is an exceedingly well-preserved and almost perfect natural internal cast from the Lower Chalk of Doi'set. The greatest length of the specimen is 21) mm. ; the greatest diameter of the sejjtate part (which is at the base of the body-chamber) is 15-5 mm., the whorl here being 9 mm. high and 11 mm. wide (excluding the tubercles). The mouth-border is well preserved. The anterior border of the muscular attachment is preserved on the lateral area as a very faint line arising (on the leit side) almost close to the last septum and at about 1'5 mm. from the edge of the whorl, and passing forward and towards the edge of the whorl crosses this edge somewhat obliquely at about G'5 mm. in advance of the last septum. Fortunately a portion of the body-chamber can be removed so that the course of this line can be satisfactorily traced. Continuing still forward it passes into a fairly wide shallow groove 4 mm. in front of the most anterior portion of the septate part of the shell, the groove being disposed across the floor or concave portion of the body-chamber in a flat forwardly-convex curve, with a very slight backward depression at the centre of the area. A similar fine line can be somewhat less distinctly traced on the right side of the specimen, passing into the same groove on the floor of the body-chamber. About 3 75 mm. posterior to the groove just mentioned there is another similar but narrower groove. From my examination of other specimens it is not unlikely that this will prove to be the posterior boundary of the shell-muscles, which seem to have been united on the concave portion of the body-chamber into a fairly wide band. The shell-muscles then appear to have been almost confined to the concave area of the body-chamber, only a narrow portion extending on to the lateral area. In the British Museum Collection (No. C. GSOO) there is also a fragmentary example of the same species (PI. 17. fig. 22), from the concave portion of the body-chamber of which the matrix has been removed as carefully as possible. There is a similar line on the lateral ANIMAL TO ITS SHELL IN SOME FOSSIL CEPHALOPODA. 83 area, passing into a shallow groove on tlie concave portion of the body-chamber, but the part of the latter between the groove and the anterior ini]n"essed line produced by the ornaments of the septate part of the shell is covered with a thin, yellowish, powdery layer, quite unlike the rest of the concave portion of the body-cbamber. This in all probability indicated tlie place of attachment of the shell-muscles. I bave not observed tlie remains of tbe annulus iji any example of this species that has come under my notice. TuKKiLiTES, Lamarck. Tm'7'ilites tuberculatus, Bosc. — An example of this species in tbe Museum of Practical Geology (No. 6372) from the Lower Chalk of Cliffe Anstey, Wilts, displays tbe anterior border of the muscular scars exceedingly Avell (PI. 18. ligs. 1, 2). Tbe specimen is a natural internal cast of about live Avliorls, the smallest of whicb is somewhat imperfect, and the rest fairly well preserved. Two of the upper wliorls show that the sipbuucle [si) is situated at about one-sixth of tbe height of the whorl below the suture of the shell. Unfortunately the suture-line is not well shown, so that it is not possible to say exactly where the body-chc^mber commences. Tbe aperture is fairly well preserved. The height of the last whorl (from the lowest of the three rows of small tubercles to tbe suture of tbe shell) is about 52 mm. ; the Avidth of the whorl (/. e. the distance across the specimen) being 1085 mm. Tbe anterior border of the muscubu- attachment is well shown on the last whorl as a well-marked narrow groove. VicAving the specimen with the aperture away from you and directed downward, tbe course of this groove may be thus described : — Commencing somewhat below the centre of the whorl it passes for a very short distance (about 2 mm.) downward and to the right ; then, turning still more to the right, it passes for a distance of about 13 mm. with a bold convex curve to within 12-5 mm. of the lowest row of smaller tubercles ; then turning again still more to the right it glasses under the uppermost row of small tubercles. Just beneath the tubercle to the left of the one under which this groove passes there are indications of a septal suture ; this is believed to be the last septum, and therefore to indicate the commencement of the body- chamber. Passing still to tbe light and down\Aard in a broad feebly-convex curve, and at the same time becoming deeper, the groove crosses the two lower rows of small tubercles at about 30 mm. from the last-mentioned bend ; it is continued on to the surface below the tubercles until at about 10 mm. below the lowest row it curves round in a broad curve and passes to the left for a short distance; then at about 21 mm. below the lowest row of tubercles it makes an angular bend, passes downward and to the left for a distance of 11 mm., when it curves upward and passes into the umbilicus ; this upward portion can be traced for about 2'1< mm., when it is obscured by matrix. With the exception of the 15 mm. first described, probably the whole of this incised line represents the anterior boundary of the two muscular scars. Turi'ilites Mantelli, Sharpe. — An imperfect example of this species in the Museum of Practical Geology (No. 0373), from the Lower Chalk of the Isle of Wight, also shows a portion of the anterior border of one of the muscular scars (PI. 18. fig. 3), The specimen 84 jNlE. G. C, CETCK ON THE MUSCULAE ATTACHMENT OP THE consists of the natural cast of only two wliorls. The aperture is not preserved, but a portion of the last whorl without doubt formed part of the body-chamber. The larger whorl is 30 mm. high (/. e. from the lowest row of tubercles to the suture of the shell), and measures 65 mm. from side to side. The anterior border of the muscular impression occurs as an impressed line very near the anterior end of the specimen. Viewing the specimen with the large whorl downward, this line commences a little below the middle of the whorl, and, crossing the uppermost of the three rows of small tubercles, turns somewhat aln'uptly to the right, becomes more deeply impressed, and Avith a bold anteriorly-convex curve crosses the other two rows of small tubercles and passes on to the sloping surface below the tubercles, being intercepted at about 11 mm. below the lowest row of tubercles by the broken anterior end of the specimen. Although the impression in this specimen is very incomplete, it serves to confirm the structure which has just been described in Turriliies tuberculatus. Subg. Heteroceras, d'Orbigny. Turrilites (Heteroceras) polyplocus, Roemer. — It is exceedingly interesting to have been able to recognize any indications of the muscular attachment in this subgenus of Ttirrilites, since here the shell is coiled in the opposite direction to that of the genus Turrilites, and hence a corresponding inversion of the muscular attachment was to be expected. The examj)le of Turrilites {Heteroceras) ^joIijjUocks, Eoemer, in which the attachment has been seen forms part of the British Museum Collection (No. 461-51) and is from the Upper Cretaceous rocks of Haldem, Westphalia (PI. 18. fig. 4). The specimen consists of about one and a half whorls. The aperture is exceedingly well preserved, but the suture-line is not shown, so that the base of the body-chamber cannot be thereby recognized. The form of the right side of the attachment is seen in fig. 4. Viewing the specimen as represented in the figure, the line indicating the anterior boundary of the muscular attachment is seen as an impressed line to commence on the lateral area of the whorl just to the riglit of the tubercle, which is above and somewhat to the left of the aperture ; passing downward it turns to the left immediately under the tubercle, and continues to pass downward for rather more than 20 mm. ; having turned upward very slightly, it continues as a line of colour in a broad sweep, nearly parallel to the outer curve of the whorl, to within about 60 mm. of the border of the aperture, when it again becomes a finely-incised line ; then, turning upward and still passing forward for about another 20 mm., it curves somewhat abruptly backward, continuing for about 20 mm. ; this last portion of the line enclosing an oval area, open posteriorly, the diameters of which are 20 and 12 mm. respectively. This area I regard as one of the muscular scars, and the line leading up to it as being partly the boundary of tlie shell-muscle, and partly the boundary of the annulus. It is much to be regretted that in developing the specimen its surface was somewhat scratched and rubbed, so that it is not possible satisfactorily to foUow the course of the impression on the other side of the whorl. ANIMAL TO ITS SHELL IN SOME FOSSIL CEPHALOPODA. 85 AMMONITES *. Amaltheid ,e, p. Fischer. OXYNOTICERAS, A. Hyatt. Oxynoticeras^ sp. — The example (PL 18. figs. 5, 6) apparently referable to this genus, in Tivhich the form of the muscular attachment has heen observed, is from the Great Oolite (Stonesfield Slate) of Stonesfield, Oxfordshire, and forms part of the British Museum Collection (No. 36710). It has been labelled "Ammonites discus, .7. Sowerby," but it is jjrobably not referable to Sowerby's species f , although its state of preservation does not allow an accurate determination. It is a crushed, poorly-preserved, internal cast lying loose on a piece of matrix. The greater part of the specimen is septate ; the body- chamber has been broken across obliquely, but fortunately the posterior portion is preserved. When complete the specimen probably was quite 145 mm. in diameter, and its umbilicus about 22 mm. wide. The last half-whorl at least was occuj)ied by the body-chamber, the base of which is o2"5 mm. high and about 10 mm. thick. The anterior border of the muscular scar and of the anuulus can be traced across each side of the specimen, but it is more distinct on that side (flg. 5) lying upon the matrix. Here the anterior border arises from the suture of the shell at a distance of 115 mm. in advance of the last septum ; after passing forward and outward for a short distance it turns backward in a rather broad curve, and passes uninterruj)tedly across the body-chamber in a i'airly straight line, which is almost parallel to the general direction of the last septum, being in advance of the septum 15'5 mm. on the inner portion of the lateral area and 21 mm. near the periphery, where it seems to turn backward a little, but this appeai'ance may be due to the much-compressed state of the fossil ; it has no depressions corresponding to the lobes of the suture-line. The portion of the border to a distance of about 13 mm. from the inner edge of the whorl is indicated by a well-marked depressed line, and the rest of the border is indicated by the anterior boundary of a band of colour about 4-5 mm. wide, the posterior boimdary of which is not sharply defined. The inner portion, bounded anteriorly by the depressed line, doubtless rejn-esents the muscular scar, and the broad band of colour the anuulus ; the scar exhibits distinct lines of growth. On the opposite side the muscular scar and annulus are less clearly defined. The muscular impression appears to be somewhat nearer the last septum, for its anterior boundary arises as a faintly-incised line at a point only S'5 mm. in advance of the last septum, and passes outward and backward as far as the septum, wliich it meets at a distance of 8'5 mm. from the inner edge of the whorl. Before reaching the septum, i. e. at a distance of 3-5 mm. from the septum, this boundary gives off a branch (indicated by a line of colour) which can be traced as a slightly- waved line across the body-chamber to the periphery, where it meets the anterior boundary on the opposite side. The portion of the annulus adjoining the muscular impression ou this side is only about * The nomenclature and grouping of the Ammonites here adopted are those given by Prof. Dr. K. A. v. Zittel in his ' Grundzuge der Palaontologie," 1895. t J. Sowerby, ilin. Con. vol. i. p. 37, pi. xii. (1813). 86 MR. G. C. CRICK ON THE MUSCULAR ATTACHMENT OP THE 1'5 mm. wide for a lengtli of about 8'5 mm. ; it then expands somewhat rapidly to a width of 4'5 mm., a width which it maintains across the rest of the body-chamber, joiniog the band on the opposite side, the portion near tlie periphery being, however, somewhat less distinct than the rest. The discrepancy in the position of the two impressions is doubtless due in a great measure, if not entirely, to the crushing which the specimen has undergone during fossilization. Amaltheus, Montfort. Amaltheus spinatus, Bruguiere, sp. — The muscular impression has been observed in an example of this species in the British Museum Collection (No. C. ''919), the locality of which is not recorded. The specimen consists of the well-preserved septate portion of the shell, together with a natural cast of one side of the dorsal (or inuer) part of the posterior portion of the body- chamber (PI. 18. fig. 7). The dimensions of the sjieciraenat the base of the body-chamber are : — diameter of the shell 57 mm. ; width of umbilicus 24 mm. ; height of outer whorl 19 mm. ; ditto above preceding whorl about 17 mm. ; thickness of outer whorl (excluding ribs) 19 mm. ; ditto (including ribs) 22 mm. A portion only of the anterior boundary of the muscular impression is preserved. Tliis, Avhich appears as an impressed line, arises almost close to the suture (of the shell) 9 mm. in front of the anterior part of the small saddle belonging to the last suture-line that is situated on the edge of the umbilicus ; it curves outward and backward for a distance of about 5 mm. where it is about 4 mm. from the suture (of the shell) ; then, passing backward nearly parallel to the inner edge of the whoid, it appears to divide just before meeting this small saddle, one part passing on the inner (dorsal) side of the small saddle situated on the edge of the umbilicus and then disappearing, the other passing on the outer side of the same saddle and then also disappearing. Posterior to this distinctly-impressed line there are several very faint lines coacenti'ic with it. Cardioceras, Neumayr & Uhlig. Cardioceras excavatum, J. Sowerby, sp. — In the description of the annulus, which has already been given, I have described the muscular attachment in an example of this species (see ante, p. 75). Another example of this species is also figured which shows the course of the anterior l)0imdary of the muscular scar on the dorsal or impressed portion of the whorl (Pi. 19. figs. 1, 2). Cardioceras aff. excacato, J. Sowerby, sp. — The muscular impression is also shown in an example of Canlioccras in tlie Britisli Museum Collection (No. 50098), but its locality is not recorded (PI. 18. figs. 9, 10). It is, however, allied to Card, excavatum, but is more comjiressed and more widely umbilicated than that species. The dimensions of the specimen are : — diameter of shell 54 mm. ; width of umbilicus 18'5 mm. ; height of outer whorl 20 mm. : thickness of outer whorl 12 mm. The test has been removed ANIMAL TO ITS SHELL IN SOME FOSSIL CEPHALOPODA. 87 from tl)e greater part of the body -chamber, and the latter can fortunately be removed from the rest of the specimen. The whorl is carinate, subquadrangular in transverse section, its sides mnch flattened and nearly jiarallel to each other ; at the base of the body-chamber it is 13'5 mm. high and 10 mm. thick. On the left side of the specimen (PI. 18. fig. 9), near the base of the body-chamber a feebly-impressed (on the internal cast) line arises from the suture of the sliell [i. e. the inner edge of the whorl) 4'5 mm. iu front of the last septum ; it extends outward and backward almost as far as the septum, nearly reaching the same at the inner side of the second (or inferior) lateral lobe ; it then turns outward, as if to pass over the adjoining saddle, and soon disappears. When the body-chamber is detached (PL 18. fig. 10) this incised line is seen to be continued on to the impressed zone, being also indicated by a difference of colour, the portion posterior to the boundary being much lighter than the rest of this surface. Prom the edge of the Avhorl this boundary passes a little backward in an /"-shaped curve nearly to the central line of the im2:)ressed area, when, turning abruptly backward, it passes nearly parallel to the median line of this area as far as the small saddle adjoining the antisiphonal lobe. The corresjjonding line on the other side of the impressed area iias the same form, so far as it can be traced, but the edge of the whorl on this side is somewhat imperfect. A line of coloiu" apj)ears to indicate that the median space between the two boundaries was bridged over anteriorly. Tlie form of the muscular attachment just described on the impressed area of the outer whorl is confirmed by its appearance on the inner surface of the dorsal portion of the test of the body-chamber, which remains attached to the l)receding whoii. The muscular scar on the right side of the body-chamber is not preserved, the inner portion of the base of the body-chamber having been broken off. On the middle of the left lateral area there is a linguiform space, 8 mm. long and ■1 mm. broad, enclosed by a very faint line, open behind and convex anteriorly (PI. 18. fig. 9). Its lateral boundaries are nearly parallel to the inner and outer margins of the whorl respectively, the inner being l-'o mm. distant from the inner margin, and the oiiter at the same distance from the periphery of the whorl. The inner boundary passes just into the inferior lateral lobe along its outer side, and the outer boundary just passes into the superior lateral lobe along its inner side. There is a similar linguiform area on the middle of the right side of the body-chamber, 95 mm. long and 6 mm. wide, oj)en behind, and Avith a convex anterior boundary. Its inner boundary passes just into the inferior lateral lobe at its outer side, whilst the outer boundary passes just into the superior lateral lobe along its outer boundary, and not its inner, as on the opposite side of the body-chamber. Hence the linguiform area on this side is somewhat broader than that on the left side, the inner boundary being at about the same distance from the inner edge of the whorl, while the outer is nearei the periphery than on the opposite side. Only in this one specimen have these linguiform areas been observed ; but, as they are so distinctly displayed on both sides of the fossil, it is just possible that they are connected with the attachment of the animal to its shell. This example seems to support the figure given by Trautschold. SECOND SF.KIES. — ZOOLOGY, VOL. VII. 13 88 MR. G. C. CEICK ON THE MUSCULAR ATTACHMENT OF THE Cardiocerm Sidherlandice, J. cle C. Sowerby, sp. — The position of the anterior boundary of the sbell-muscle is exceedingly well shown in a large example of this species from the Oxfordian of Scarborough, Yorkshire. It forms part of the British Museum Collection (No. 82369). It is a large internal natural cast, about 370 mm. (or nearly 14f inches) in diameter, the umbilicus being 85 mm. in diameter, and having almost perpendicular walls. Nearly one-half of the outer whorl is occupied by the body- chamber, the base of wbich is obtusely cordate in section, 132 mm. (or nearly 5| inches) high and 210 mm. (or rather more than 8J inches) wide, the inner area being 45 mm. wide and nearly perpendicular to the plane of symmetry of the shell. The anterior boundary of the muscular impression, represented by a well-marked groove, crosses the umbiUcal margin 60 mm. (or about 2| inches) in front of the last septum, and passing outward and backward becomes much less distinct at about 52 mm. from the last septum, and at about the same distance from the umbilical margin. It appears, however, to be continued across the whorl by a faint, somewhat irregular groove, which, originating some 6 mm. posterior to the groove just described, crosses the whorl in a flat forwardly-concave curve which is only 15 mm. from the anterior portion of the lateral saddle, and nearly touches the anterior part of the external saddle ; this we regard as the anterior boundary of the annulus. Passing inward from the umbilical margin, the boundary of the shell-muscle is indicated by a distinct groove which curves somewhat backward, but can be traced onlj'^ about halfway across the inner area of the whorl. Cardioceras funifermn, J. Phillips, sp. — A similar boundary of the shell-muscle has been observed in several examples of this species which are contained in the British. Museum Collection. It is well shown in an example (No. 50447) from the Oxford Clay" of Trowbridge, Wiltshire. This is 167 mm. in diameter and 68 mm. thick, the umbilicus being almost closed and its margin rounded. About one-half of the outer whorl is occupied by the body-chamber, the base of w^hich is acutely cordate in section, 64 mm. wide and 72 mm. high. On the left side of the specimen the anterior border of the shell-muscle is indicated by a groove which crosses the umbilical margin at 9 mm. anterior to the last septum ; passing backward and outward as a fairly wide groove for about 10 mm., it then becomes shallower and wider. The posterior boundary of the muscular scar is represented by a slightly raised ridge, which, passing from the umbilical margin just above the last septum outward and forward, meets the lower boundary of the anterior groove at a point 12 mm. from the umbilical margin. The anterior groove appears to be continued across the whorl {i. e. to within 5 mm. of tiie periphery) as an exceedingly faint groove, which indicates the position of tlie anterior border of the annulus ; this is 8 mm. anterior to the lateral saddle, 10 mm. anterior to the external saddle, and is depressed at each lobe. On the outer haK of the lateral area of the whorl this boundary is partly indicated by remains of the test. On the right side of the specimen there are similar indications of the anterior boundary of the muscular scar, but uo traces of the annulus. ANJMAL TO ITS SHELL IX SOME FOSSIL CEPHALOPODA, 89 Cardioceras sj). — Indications of the muscular attachment are well displayed in a small Ammonite from the Upper Jurassic of Kintradwell, Sutherland, which forms part of the British Museum Collection (No. C. 4389). Although the specimen is fairly well pre- served, I have not been able satisfactorily to determine the species, but it seems to be referable to the genus Cardioceras (PI. 19. figs. 3, 4). Its dimensions are as follows : — diameter of shell 235 mm. ; greatest thickness 8'5 mm. ; width of umbilicvis 7 mm. ; height of outer whorl 8 mm. The last two-fifths of the outer whorl are occvipied by the body-chamber ; the test having been removed from the left side and from the periphery of this portion of the whorl, the internal cast is weU displayed. Unfortunately only a small portion of the last suture-line can be made out, the rest being obscured by the test. The portion of the last septum adjacent to the suture of the shell is obscured, but on the internal cast of the body-chamber, at a short distance anterior to the last septum, a very fine incised line arises from the suture of the shell, passes inward in an almost radial direction for about 2'5 mm., then turns I^ackward for about 0-5 mm., and again resumes its radial direction across the whorl, being feebly depressed as it passes over each lobe, and slightly raised in passing over each saddle; it is about 0"5 mm. above the lateral saddle and almost touches the most anterior portion of the external saddle. As it approaches the periphery it lu^rns forward to join a somewhat peculiar-shaped roughened scar represented in fig. 4, the posterior inflated portion of which is rather rougher than the rest. Unfortunately the opposite side is obscured by matrix. The perijihery of the posterior portion of the body-chamber seems to be somewhat deformed, and to possess a feeble keel with a shallow sulcus on either side ; this deformation may account for the median division of this scar, which probably was originally horseshoe-shaped, as observed in several otlier Ammonoids. The portion of the incised line near the suture of the shell doubtless represents the position of the anterior border of one of the shell-muscles, the rest of the line indicating the position of the anterior boundary of the annulus, there having been, in addition, a firm attachment at the centre of the 2:)eriphery. Neumaykia, Xikitiu. Nenmayria catenulata, Fischer, sp. — In the British Museum Collection there is an example of this species exhibiting the form of the muscular attachment. It is a badly- cruslied internal cast from the Portlandian of Choroschowo, near Moscow, Russia. The specimen is 119 mm. in diameter, the width of the umbilicus (from suture to suture) being 18 mm., the height of the outer whorl 58 mm., and its thickness 26 mm. The inner area of the whorl slopes towards the umbilicus and at the base of the body-chamber is 6 mm. wide, the height of the whorl here being 43"5 mm., and its greatest thickness, which is at a short distance from the umbilical margin, 15-5 mm. Tlie last two-thirds of the outer whorl are occupied by the body-chamber. The whorl is sagittate in cross section. The muscular impression is well shown on one side, but tlie crushed condition of the specimen has obliterated it on the other side. Arising from the suture at about 7 mm. from the base of the l)ody-chamber, the boundary of the impression, in the form of a faintly-indented 13* 90 MR. G. C. CRICK 0]S THE MUSCULAR ATTACHMENT OF THE line, passes outward and backward in a curve, which nearly meets the last septum at about ll'Onim. from the suture, and thus encloses a subtriangular area on the inner side of the base of the body-chamber. Just before this outer line reaches the last septum it gives off a very faint branch which passes outward towards the periphery, but can be traced for a very short distance only. This is, without doubt, an indication of the anterior boundary of a portion of the annulus. Within and parallel to the outer boundary of the impression there are one or two very faintly indented lines. TissoTiA, Douville. Tissotia Uwaldi, v. Buch, sj). — The muscular impression is shown iu an example of this species in the British Museum Collection (No. C. 4810 a) from the Chalk (Senonian) of Mezab-el-Mesai, Algeria (PI. 18. fig. 11). The specimen is a natural cast, 8S'5 mm. in diameter ; rather more than one-fourth of the outer whorl is occupied by a portion of the body-chaml)er, but this latter is somewhat imperfect on one side, so that the dimensions of the specimen can best be taken at the base of the body-chamber. Here its dimensions are as follows : — diameter of shell 84'5 mm. ; width of umbilicus 8'5 mm. ; height of outer whorl 42 mm. ; greatest thickness of same (at a short distance from the umbilicus) 47 mm. The whorl is obtusely cordate in transverse section, rather wider than high, and indented to rather more than one-half of its height by the preceding whorl. The anterior border of the muscular impression is represented by a feebly-incised line. It cannot be traced quite as far as the suture of the shell, but on the narrow inner area of the whorl it is 4 mm. from the last septum ; passing thence forward and outward on to the lateral area, it attains its greatest distance from the last septum (viz. 8"5 mm.) at 4 mm. from the edge of the whorl. Then, curving backward, it passes on the vimbilical side of, and at a distance of 1'5 mm. from, the first (counting from the umbilicus) large lateral saddle, and, turning forward, sweeps round in an anteriorly-concave curve immediately about the small saddles occupying the margin of the umbilicus and soon disappears. The muscular scar thus bounded appears to have been an oval about 10 mm. by 7 mm., its longer axis being placed nearly in a radial direction but slightly inclined backward. Prom the anterior portion of the boundary of this impression a feebly-depressed line passes backward and outward in a broad f or wardly- convex curve to within about 2*5 mm. of the outer portion of the first (counting from the umbilicus) large lateral saddle, then in a forwardly-concave curve crosses the adjacent lobe, and disappears just above the next lateral saddle. This line may indicate a portion of the anterior boundary of the annulus. The opposite side of the body-chamber has been so much eroded that the muscular impression is not preserved. A^'IMAL TO ITS SHELL IN SOME FOSSIL CEPHALOPODA. 91 Lytoceratid^, Neumayi", emend. Zittel. Lytoceras, Suess. Lijtoceras cornucopia;. Young & Bird, sp. — The impression of one of the slicll-muscles has been observed in an example of this species in the British Museum (No. 37851) from the Upper Lias of Whitby, Yorkshire (PI. 19. figs. 15, 16, 17). The dimensions of the specimen are : — diameter 22 mm. ; width of umbilicus 8 mm. ; height of oviter whorl 8 mm. ; greatest thickness of the outer whorl 8 5 mm. Rather more than one-half of the last whorl is occupied hj the body-chamber, the base of which is nearly circular, its dorso-ventral and transverse diameters being each G mm. The amount of indentation by the preceding whorl is very small. At the base of the body-chamber, and on the inner area of the whorl, tliere is an oval impression (fig. IG), truncated on the inner side by the ed-TC of the " impressed zone." The diameters of the portion that can be seen are 1-75 mm. and 2'0 mm. respectively, the longer diameter being placed transversely on the whorl. The anterior and the posterior boundaries are both visible, the latter l)eing almost close to the first auxiliary saddle ; but no trace of the annulus can be seen. On the peripheral area there is an elongated linguiform impression (fig. 17), which is rounded and submucronate anteriorly, but is open posteriorly, each limb being situated in each half of the siphonal (periphei'al) lobe. Its anterior portion is a little in advance of the most anterior part of the siphonal saddle ; its posterior portion is slightly contracted. Its length is 3'25 mm. and its width (between the limbs) '75 mm. In the British Museum Collection there is also a portion of the natural internal cast of the body-chamber of another example of this species (PL 19. figs. 13, 14), from the Upper Lias of Whitby, which exhibits the impression of the boundary of the right muscular scar surrounded anteriorly and laterally by a dark-coloured, longitudinally-elongated, oval area, a portion of which is slightly rougher than tlie rest. Its form is shown in fig. 13. Only the right side of the base of the body-chamber is preserved, the heiglit of the whorl here being 35 mm. At about 4'5 mm. anterior to the last septum, the incised line indicating the position of the boundary of the muscular scar passes fi-om the suture of the shell — /. e. the edge of the impressed zone — in a radial direction for about 7 mm. ; then, turning abruptly backward nearly at right angles to its former course, it is continued as far as the last septum. At the suture of the shell the boundary of the oval dark-coloured area is 3'5 mm. anterior to the incised line just mentioned ; after proceeding forward in an anteriorly-concave curve for about 2 mm., it passes in an anteriorly-convex curve forward and outward for about 7 mm. ; then, after curving gently backwai'd for about 5 mm., it can be traced nearly as far as the last septum, having a direction nearly parallel to the edge of the impressed zone. Lytoceras fimbriahmi, J. Sowerby, sp. — The muscular impression is also shown in a fragmentary example of this species in the British Museum (No. 20837) from the Middle Lias of Kilsby Tunnel, Northamptonshire. The specimen consists of nearly half a whorl, and is a natural cast of part of the body-chamber of a shell having approximately the 92 MK, G. C. CBICK ON THE MUSCULAR ATTACHMENT OF THE following dimensions : — diameter 120 mm. ; width of umbilicus 49 mm. ; height of outer whorl 44 mm. ; greatest thickness of ditto 41 mm. The posterior portion of the body-chamber is preserved ; one side is much crushed, but the rest is fairly comjilete. A portion of the test still adheres to the cast. The sjjecimen, measured along the median line of the periphery, is 165 mm. long ; its transverse section is ovate, the greatest thickness being at about two-fifths of the height of the whorl from the inner edge ; at its posterior end the height of the whorl and its thickness (allowing for the crushing) are 80 mm. and 27 mm. respectively. The impression of one shell-muscle is well shown, but only the inner portion of the other (PI. 19. figs. 11, 12). The whorl is very slightly imj^yressed by the preceding whorl. In shape the impression is truncated-oval, its longer diameter being transverse and tlie truncated end towards the median line of the dorsal surface. Arising a little in advance of the saddle on the side of the antisiphonal lobe, the boundary of the impression passes forward for about 5 mm. nearly parallel to the median line of the dorsal or antisiphonal surface, and at 1'5 mm. from its fellow on the opposite side ; then, curving outward, it extends for about 10 mm. ia a direction almost at right angles to its previous course; bending rather abruptly backward, it jiasses for a short distance nearly parallel to the boundary of the impressed zone ; tehn, curving inwai'd and backward, it speedily disappears. The greatest width (transverse) of the impression is 12 mm. There is no trace of the annulus. The boundary of the impression is indicated by a depressed line, Avhich is rather deejily incised on the inner and outer portions, but somewhat less so on the anterior portion. The portion of the surface of the cast bounded by tlie impression is a little rouglier than the rest. Owing to the crushed condition of the whorl, only the inner portion of the other impression is preserved. The impression is so situated that in a lateral aspect of the whorl only a very small portion of it can be seen (fig. 12). Lytoceras qiuidrisulcatimi, d'Orbigny, sp. — One example of this species, collected by Dr. J. W. Gregory in East Africa, exhibits a portion of the muscular impressions. It is a crushed fragment of a natural cast of the posterior pai"t of the l)ody-chamber. The inner portion of the whorl is badly crushed, but exhibits on either side a part of the muscular scar. At the base of the body-chamber the whorl is 30'5 mm. high and 80 mm. thick, the greatest thickness being a little within the central portion of the lateral area ; tlie transverse section is therefore nearly circular. The muscular scars are somewhat darker than the surrounding surface of the cast, and present a slightly roughened appearance. That on the left is the better shown, but its inner portion is obscured so that the "whole of its anterior border is not visible. Its bovmdary is usually a feebly- indented line, but sometimes a slightly-raised line. Tiie anterior boundary of this impression appears to commence at about 7'5 mm. in advance of the most anterior part of the last septum, and about 8 ram. from the central line of the impressed zone; it passes thence slightly forward and outw^ard for a distance of about 2 mm. ; then, curving backward at about 10*5 mm. from the median line, it passes for a short distance nearly parallel to the median line ; then, at about 4'5 mm. in advance of the last septum, ANIMAL TO ITS SHELL IX SOME FOSSIL CEPHALOPODA. 93 curving rather abruptly inward but still passing backward, it disappears at about 5'5 mm. from the median line, exlni)iting, however, a tendency to again turn forward. The portion of the scar which is visible has a pyrifortn shape ; its axis is at an angle of about 45° with the median line of the dorsal area of the whorl, the '"stalk" being- directed outward and backward ; its length is about 8'5 mm., and its gi'catest width about 5 mm. The other scar is truncated by a fracture crossing it obliquely, so that only the posterior portion of the impression is seen. It is somewhat darker than the surrounding surface, and bounded anteriorly and laterally by a very slightly-raised line. Commencing from the fracture at about 9 mm. in advance of the most anterior part of the last septum, and at the same distance from the median line of the dorsal area, the anterior boundary passes backward and very slightly outward ; then, curving round, comes to within 2 mm. of the suture-line ; then, curving inward and forward, it is again intercepted by the fracture at about 4 mm. from the median line. The anterior and outer boundaries of the impression are sharply marked by a feebly-incised line, but the posterior boundary is not quite so sharply defined. There is a very shallow and rather broad depression on the peripheral area about 5 mm. in front of the summit of the siphonal (peripheral) saddle ; on either side of the median line of the periphery, and at a distance of 3 mm. apart, a very faint line curves forwai'd and towards the median line of the periphery, nearly meeting its fellow at about 3 mm. anterior to the depression just mentioned. This possibly represents the anterior border of the peripheral portion of the annulus. JiIgoceratid.'E, Neumayr, emend. Zittel. Arietites, Waagen, emend. Zittel. Arietites ohtusus, J. Sowerby, sp.* — The muscular impression is well shown in an example of this species from the Lower Lias of Lyme Regis, that was kindly lent to me by Mr. G. F. Harris, F.G.S. The specimen is a fairly-complete internal cast, almost entirely denuded of the test. Its dimensions are : — diameter of shell 101 mm. ; width of umbilicus 43 mm. ; height of outer whorl 32 mm. ; ditto above preceding whorl 29 mm. ; greatest thickness (excluding ribs) 30 mm. ; ditto (including ribs) 33 mm. The dimensions of the base of the body-chamber are : — height of whorl 26 mm. ; thick- ness (excluding ribs) 22 mm. ; ditto (including ribs) 25 mm. The body-chamber is complete and occupies the last half-whorl. The muscular impression is situated at the base of the body-chamber and occupies the inner area of the whorl (PI. 19. figs. 8, 9, 10). Its anterior boundary, represented by a fine incised line, is convex, its anterior portion being 16 mm. in front of the anterior portion of the second lateral saddle of the last septum. Passing backward its outer boundary is nearly parallel to, and at a distance of 6 mm. from, the suture (of the shell). It can be traced backward for a distance of * Min. Con. vol. ii. p. 151, pi. clxviii. (l'6\'l)= Asteroceras ohtusum (J. Sowerby): fide Hyatt, 'Genesis Arietidse,' 1889, p. 201. 94 MK. G. C. CRICK ON THE MUSCULAE ATTACHMENT OF THE 11 mm., beyond which point the surface of the cast is fractured. About 2 mm. behind this line, and nearly concentric with its anterior portion, tliere is another very faint line, evidently marking a former attachment of the shell-muscle. On the inner area of the Avliorl there are three or four fine, faint, longitudinal grooves, extending over three-fourths of the length of the body-chamber; one of these, 2 mm. from the suture (of the shell), extends the whole length of the body-chamber. On the periphery, on either side of the median line, there is a feebly-impressed sigmoidal line (fig. 10), tlie anterior portion of which is about 11 mm. in advance of the anterior portion of the external (peripheral or siphonal) saddle. Each line arises about 2 mm. from the median line of the periphery ; it passes away from this line, and backward to about 8 mm. from the most anterior portion of the external (peripheral or siphonal) saddle, and then disappears. These lines may indicate the position of the attachment of the animal to this part of its shell, but no connection can be seen between these lines and the impressions on the inner area. A line drawn from the centre of the shell to the most anterior part of these lines on the periphery crosses the inner edge of the outer whorl at a point 7"i3 mm. posterior to the anterior boundary of the impression which is seen on the inner area. Posterior to these lines on the peripheral area there is a horseshoe-shaped marking with nearly parallel sides, each of which passes posteriorly into each half of the sijjhonal lobe. It is rounded anteriorly, and crossed in its length by several lines parallel to the anterior border. Its length (above the small median saddle) is 11 mm., and its width 5 mm. This may, or may not, be connected Avith the muscular attachment of the animal, but similar impi'essions have been observed in other forms. (See jEgoeeras capi'icorn/iij/, p. 95.) Arietites rarlcostatus, Zieten, sp.*- — In the British Museum Collection there is an example of this species (No. 0.4882) showing the muscular impressions (PL 20. figs. 10, 11). The locality of the specimen is not recorded, but, judging from the matrix, it came pro- bably from the Lower Lias of Somersetshire. The specimen is a well-preserved natural cast with the following dimensions : — diameter 51 mm. ; Avidth of umbilicus 33 mm. ; height of outer Avhorl 10 mm. ; greatest thickness of outer whorl (excluding ribs) 14 mm.; ditto (including ribs) 16 mm. The body-chamber occupies the whole of the last whorl, and its posterior part {ah) can be removed from the rest of the fossil; its transverse section is transversely oval, its thickness (excluding the height of the ribs) being 10 mm., and its height 7 mm.; it was only slightly indented by the broad periphery of the preceding whorl. The greater part of the muscular impressions is on the concave dorsal area, i. e. on the impressed zone. On either side of the median line of this area (fig. 11) there is a subti'iangular impression, having its broadly-rounded apex directed forward. These impressions are about 1 mm. apart, that on the right side of the shell being the more nearly complete. Each is bounded by a faint double fine. Commencing * F. Zieteij, Vcrsteia. Wiirt. p. 18, \>\. siii. f. 4 = Caloceras rarkostatum {Ziitien), Jidi Hyatt, Gen. Arietidae, p. 144. ANIMAL TO IT.S SHELL IN SOME FOSSIL CEPHALOPODA. 95 at the most anterior part of the antisiphonal (or antiperipheral) saddle and at a point 0'5 mm. from the median line, the boundary of the more nearly complete impression 2)asses forward and outward until it is 3 mm. in advance of the last septum and about 2 5 mm. distant from tlie median line ; then, turning backward and passing still outward, it apj)roaches the last sejitum, and nearly (but not quite) meets the inner portion of the lateral saddle, where it disappears, its place of disappearance being 4*5 mm. from the median line of the dorsal area or of the impressed zone. The outer boundary of the muscular impression is on the edge of the umbilicus, and therefore in a lateral view of the Ammonite only the outer backwardly-directed portion of the impression can be seen. There is a corresponding and similar impression on the other side of the specimen, but it is not so complete, owing to a fracture of the fossil. There are no traces of the annulus. JiIgoceras, Waagen, emend. Zittel. JEcjoceras cajjricornum, Schlotheim, sp. — A portion of the im2")ression of the muscular attachment of the animal has been observed in an example of this species from the Lias of Cheltenham, that forms part of the British Museum Collection (No. 67929). The specimen is a well-preserved natural cast, bearing portions of the test in a soft, white, friable condition (PI. 20. fig. 2). By removing this with a stiff In'ush from near the base of the body-chamlier, the muscular impression is well shown on either side. The dimensions of the specimen are as follows : — Diameter of shell (including ribs) 47 mm. ; ditto (excluding ribs) 44 mm. ; width of imibilicus 20 mm. ; height of outer whorl (including ribs) 15 mm. ; ditto (excluding ribs) 14-5 mm. ; thickness of whorl (including ribs) 17 mm. ; ditto (excluding ribs) 14-5 mm. At the base of the body-chamber the height and thickness of the whorl are respectively 11 and 13 mm. including the ribs, or 10 and 10"5 mm. excluding the ribs. The outline of the muscular impression is seen on either side at the base of the body-chamber and quite close to the suture of the shell. It is convex anteriorly, its anterior boundary being 6 mm. in front of the most anterior jwrtion of the second lateral saddle ; its outer boimdary is nearly parallel to, and at a distance of 2 mm. from, the edge of the impressed zone, and passes jiosteriorly on to the outermost branch of the second lateral saddle. Its anterior boundary exhibits a tendency to curve forward at the suture of the shell. In the middle of the jieripheral area there is a tongue-shaped scar, which is rather more than twice as long as wide, has nearly parallel sides, and is rounded anteriorly ; each side of it just touches the boundary of the siphonal lobe. The scar seems to have been covered with a thin film of shelly matter, for j)art of this has been broken away from the anterior portion so as to expose a black layer beneath. This scar would be situated in the immediate neighbour- hood of the siphuncle, and may or may not have had anything to do with the muscular attachment of the animal. Its length above the extremity of the small median saddle is about 6 mm., and its width is 1"5 mm. A precisely similar scar has been observed in Arietites obtiisns*. * See ante, p. 93. SECOND SERIES. — ZOOLOGY, VOL. VII. 14 96 ME. G. C. CRICK ON THE MUSCULAE ATTACHMENT OF THE The outline of the musculav impression is also well shown in another example of this species from the Lower Lias, near Whitby, in the collection of F. L. Bradley, Esq., F.G.S., of Alderley Edge, Cheshire, to whom I am greatly indebted for the loan of the specimen. The impression has the same form and position as that already described in the example from the Lias of Cheltenham, which is in the British Museum Collection (No. 67929). The specimen is in the centre of a nodnlc ; its dimensions, so far as can be ascertained, are as follows : — diameter of shell 65 mm. ; width of umbilicus 31 mm. ; height of outer whorl 16 mm. ; ditto at base of body-chamber 13 mm. The last half-whorl is occupied by the body-chamber, the exposed portion of which is almost completely denuded of the test ; at its base and on the inner portion of the whorl the outline of the muscular impression can be seen. The anterior convex boundary of the impression is 5"5 mm. in advance of the last septum, the width of the impression from the suture of the shell being 2 mm. There is no trace of the annulus, and the perijiheral portion of the fossil is obscured by matrix. JEgoceras laqueolum, Schloenbach, sp.* — In an example of this species from the Lower Lias of Harbury, Warwickshire, in the British Museum Collection (No. C. 6619), one of the muscular scars is well displayed. The specimen consists of about three-fourths of a whorl 165 mm. in diameter, and is a natural cast of part of the body-chamber. The base of the body-chamber is complete : it is 24 mm. high ; 30 mm. wide including the ribs, or 29 mm. excluding the ribs. The muscular scar, which is clearly seen at the base of the chamber, is situated partly on the lateral area and partly on the impressed zone (PI. 20. figs. 3, 4). The faintly-incised line bounding the scar arises from the last sejitiim at a point 4"5 mm. from the inner edge of the whorl ; then, passing forward for a length of about 9 mm. nearly parallel to the inner edge of the whorl, it turns towards the impressed zone, becomes much more distinct, and crosses the inner edge of the whorl 13 mm. in advance of the last septum ; passing on to the impressed zone with a rather broad sweep, it gradually disappears, but the form of the scar is indicated by a difference in colour, the impression being a little darker than the surrounding surface. The impression did not quite reach the middle of the impressed zone ; it appears to have been elliptical in form, about 15 mm. long and 11 mm. wide, the larger portion being situated on the impressed zone. On the anterior portion of that part of the scar which is situated on the imjjressed zone there is an exceedingly faint line, concentric with the anterior border of the scar, which evidently marks a former attachment of the anterior border of the shell-muscle. Only a very small portion of the other scar is visible ; this is on the impressed zone. The greater part of the imjn-ession is obscured by matrix. There is no trace of the annulus. * raltcontogr. vol. xiii. pi. xxvi. f. \ = CaJoctras Uaskwn (d'Orbigny), _^c?e Hyatt, Geu. Arietida:, p. 139. ANIMAL TO ITS SHELL IN SOME FOSSIL CEPHALOPODA. 97 JEgoceras Portlocli'd, Wright. — An example of tliis species from the Boulder Clay of Ireland (exact locality iinkuown) was sent to me hy Dr. A. H. Foord for determination. It was in a reddish matrix. It is a natural cast of the posterior portion of the body- chamber, showing not quite the complete suture-line at its base, and bearing only fragments of the inner portion of the test. The length of the fragment measured on the periphery is just over 70 mm. ; the whorl is nuich comjjressed, nearly twice as high as wide ; the transverse section of the whorl is elongate-oval, its gi'eatest thickness being at about one-third of the width of the latei'al area distant from the suture (of the shell). The impressed zone on the inner side of the specimen shows that the indentation by the preceding whorl was very slight, and that the periphery of the jjreceding whorl was more acutely convex than that of the outer Avhorl. At the base of the body-chamber the height of the whorl is 33'5 mm., and the greatest thickness about 18 mm. Any indication of the muscular attachment that may be preserved on the surface of the impressed zone is obscured by portions of the test, but on each side of the inner portion of the lateral area of the posterior jioi'tion of the cast the impression of this attachment is to be seen. Its anterior boundary commences at a point on the edge of the impressed zone about 14 mm. in advance of the most anterior portion of the sutiire-line, /. e. the main or inner branch of the siphonal saddle ; it passes outward and backward for a length of about 5 mm., then for a short distance runs nearly parallel to, and at about 5 mm. from, the suture. After passing backward for a distance of nearly 10 mm. from its commencement, the line divides into two principal portions, one being continued almost parallel to, and only slightly approaching, the edge of the impressed zone, the other curving towards, and apparently reaching, the same edge at a distance of 11 mm. from its commencement. A fractui-e of the specimen prevents the former of the two lines being traced any further. About 1 mm. anterior to the incised line just described, there is a much fainter depressed line which, after extending backward for about 5 mm., nearly parallel to the incised line already described, appears to turn outward towards the periphery, and then quickly to disappear, vrhile posterior to the line already described, and also parallel to it, there are two or three much fainter lines. On the other side of the whorl there are lines corresponding to the principal incised line, and the line about 1 mm. anterior to it, but these can only be traced backward sonie G mm., owing to the crushed state of this side of the specimen. Near the base of the body-chamber, and slightly on the right side of the median line of the periphery, there is a horseshoe-shaped incised line having its convexity directed forward. Its anterior portion is 12 mm. in advance of the anterior portion of the outer branch of the siphonal (or peripheral) saddle. It is about 6 mm. long, 5 mm. wide in the anterior part, and 3"5 mm. wide in the posterior part. Each limb appears to be continued backward as far as the posterior end of the specimen, which is at about the level of the anterior portion of the outer branch of the peripheral (or siphonal) saddle, as a somewhat irregular, shallow, very faintly-impressed groove or grooves, each being almost parallel to the central line of the periphery. The posterior termination 14* 98 ]MK. G. C. CRICK ON THE MUSCULAK ATTACHMENT OE THE of each liinb is rather more deeply impressed than the rest. Within this horseshoe- shaped im])ressiou, and occupying- the median line, is a very fine incised line, which extends backward from about 1 mm. posterior to the anterior boundary of the impression for a distance of about 6'5 mm. Doubtless this horseshoe-shaped impression had to do with the muscular attachment of the animal to its shell, for it has been observed in other forms, but the present specimen does not enable us to trace any connection between this impression and the marks of the muscular attachment seen on the inner edge of the whorl. SoNNiNiA, Bayle. Sonninia sp. — In the British Museum Collection there is an example (No. C. 5188) of Sonninia sp. from the Inferior Oolite of Yeovil, Somerset, in which the impression of the shell-muscles can be seen (PI. 20. fig. 12). The test has been removed from the internal cast of the whole of the body-chamber, and the impression of each shell-muscle is displayed near the base of the chamber. The specimen has the following dimensions : — diameter of shell 91 mm. ; width of umbilicus (from suture to suture) 28 mm. ; ditto (from margin to margin) 37 mm. ; height of outer whorl 36'5 mm. ; greatest thickness (almost close to the umbilical margin) 24 mm., excluding thickness of test. The body- chamber occupies one-half of the last whorl ; at its base the whorl, or more correctly the internal cast of the whorl, is 27 mm. high and 19 mm. thick. The whorl is subquadrate in section, with nearly jjarallel sides and well-defined inner area, the latter at the base of the body-chamber being about 6 mm. wide and sloping towards the umbilicus, making an angle of about 140° with the lateral area. The boundary of the muscular impression is indicated by an incised line which appears on the inner area of the Avhorl at a point 10"5 mm. anterior to the last seiDtum; it jDasses outward and. backward, just crossing the subangular umbilical margin, on to the lateral area, being at about 6 mm. distant from the suture of the shell, and ajipears to almost touch the saddle of the last septum that is situated on the subangular umbilical margin. On the area enclosed by this incised line, and concentric with it, there are several very faint lines, similar to those seen in the impression of the shell -muscles in the recent Nautilus. The surface of the cast anterior and also exterior to this incised line bears a number of shallow and iiTegular indentations, which may have had something to do with the muscular attachment, but the incised line just described appears to have been the impression of the anterior and exterior boundary of the shell-muscle. A similar line and similar adjacent indentations are present on the opposite side of the cast, but no indications of the annulus have been observed in this specimen. H A R p o c E R A T I D M, Neumayr, emend. Zittel. Hecticoceras, Bonarelli. Meclicoceras hcnticnm, Reinecke, sp. — In the British Museum Collection tliere is an example (No. 22309 a) froiii the Brown Jura e, Dettingen, Wiirtemberg, which displays the impression of the muscular attachment on both sides of the body-chamber ANIMAL TO ITS SHELL IN SOME FOSSIL CEPHALOPODA. 99 (PI. 20. figs. 5, 6, 7). It is a natural cast, and has the following dimensions : — Diameter of shell 18'5 mm. ; width of umbilicus 7"5 mm. ; height of outer whorl G%j ; greatest thickness 4 mm. At the base of the body-chamber the whorl is 5 '5 mm. high and 3'5 mm. thick. The shell is somewhat comjjressed, the inclusion very slight, and the umbilicus wide and shallow. The body-chamber occupies rather more than one-quarter of the last whorl, its inner (dorsal) edge subtending a chord of 8 mm. on one side and a little less on the other side. Erom the fractured anterior end of the body-chamber a double feebly-impressed line passes backward at a distance of 1 mm. from the suture (tig. 6) ; then, diverging slightly, the two lines cross the whorl in a rather deep forwardly- concave curve, their greatest distance from each other being rather more than 1 mm. Approaching each other slightly, but without touching, they curve forward at the periphery, each forming a submucronate jioint before passing on to the other side. It is to be noted that these points are not in the median line of the periphery, but slightly ou one side (fig. 7). The posterior of the two lines crosses the lateral area about 1 mm. in advance of the anterior portion of the first lateral saddle. Ou the left side of the anterior end of the body-chamber (fig. 5) the double line just described curves slightly towards the periphery, but the opposite side, being a little shorter, does not exhibit this curvature. This double line probiibly represents a j^oi'tion of the anterior boundary of the shell-muscle, the two lines on the lateral and j^eripheral areas indicating the position of the anterior and posterior boundaries resijectively of the annulus. CEcoTiiAUSTEs, Waagen. CEcotrmistes crenatus, Bruguiere, sp. — Two examples in the British Museum Collection (No. 8968), from the Oxfordian of Doubs, France, exhibit what appears to be the muscular im2)ressiou. Both are natural casts. One (PL 20. fig. 9) of these (No. 8968 b) has the following dimensions (exclusive of the spines on the periphery) : — diameter 11 mm. : width of umbilicus 3 mm. ; height of outer whorl 5 ram. ; thickness of ditto 3 mm. At the base of the body-chamber the height of the whorl is 4 mm., and its thickness 2"5 mm. The septate part of the shell and the extreme base of the body-chamber are of a reddish-brown colour, while the rest of the body-chamber is of a much darker shade, a dark and very distinct line marking the division near the base of the body-chamber. This line appears to be the anterior border of the muscultir attachment of the animal. It commences from the suture only a short distance in advance of the last septum, and crosses in a radial direction the inner half of the lateral area until reaching the lateral saddle where it is 0*75 mm. in advance of the suture-line ; it then curves slightly backward, making a very shallow curve, uatil it reaches a point about 1'25 mm. from the median line of the periphery, where it turns somewhat abruptly forward, and reaches the median line of the periphery, 2*25 mm. \\\ advance of the summit of the outer (or peripheral) saddle, thus forming on the periphery a subtriangular projection. A precisely similar line is seen on the opposite side of the cast, but in tliis species there does not apjiear to be any forAvard prolongation near the 100 ME. G. C. CEICK OX THE MUSCULAE ATTACHMENT OE THE inner area of the whorl such as has been observed in Distichoceras Bcmgieri (see below) and Hecticoeeras hecticmn (see p. 98). The other specimen (No. 896S a) has the following- dimensions, exclusive of the periphei'al spines, which in this example are comparatively small : — diameter of the shell 11 mm. ; width of umbilicus 2 mm. ; height of outer whorl 5-5 mm. ; thickness of ditto 2'5 mm. At the base of the body-chamber the height of the whorl is 4"25 mm. (exchiding spines), and the thickness 2'5 mm. The state of preservation of this- specimen is quite similar to that of the one just described ; the dark line near the base of the body-chamber, which is most prol)ably the anterior boundary of the mxiscular attachment, is very like that in the previous specimen, but is slightly more waved on the lateral area. In this example also there does not appear to be any forward prolongation of the muscular impression on the inner {not impressed) area of the whorl. The lack of any forward prolongation in this species seems to be confirmed by a specimen (No. 19536 e) in the British Museum Collection, from the Drift of Braunstou, Northamptonshire, in which there is au indication of the anterior boundary of the muscular attachment near the su.ture of the shell, and in this example also it is only a short distance anterior to the last septum. Distichoceras, Munier-Chalmas. Bistichocoras Bangicri, d'Orbigny, sp. — In the British Museum Collection there are two examples (No. 5140 a, b) of this species which clearly display the form and position of the muscular attachment of the animal to its .shell. The species occurs in the Oxfordian, but the locality of these specimens is not recorded. In the larger example (No. C. 5140 a) rather more than one-half of the last whorl is occupied by the body-chamber (PI. 20. fig. 8). Its dimensions (not including the spines on the edges of the periphery) are : — diameter 17 mm. ; width of umbilicus 4'5 mm. ; height of outer whorl 8 mm. ; greatest thickness (which is at a very short distance from the umbilicus) 4*5 mm. At the base of the body-chamber the height of the whorl is 5 mm., and its thickness 3"5 mm., the transverse section being sublanceolatc. Measured along the curve of the inner portion of the lateral area, the anterior boundary of the muscular attachment can be seen as much as G'5 mm. in advance of the last septum. After passing backward at a distance of 1'75 mm. from, and parallel to, the inner edge of the whorl for a distance of about 5 mm., the boundary then passes outwards in a shallow curve towards the pcrijihery, just before reaching which it turns slightly forward. There does not appear to be a definite line at the anterior extremity of the forward prolonga- tion of the impression, but the area included between the boundary of the muscular attachment and the inner edge of the whorl is a little roughened, the surface of the specimen outside this line being smooth and polished. Where the boundary turns outward there are several fine incised concentric lines. It is to be observed that the boundary of the forward prolongation of the muscular attachment does not coincide Avith the longitudinal groove running along the middle of the lateral area, but is situated between the inner margin of the whorl and this groove. ANIMAL TO ITS SHELL IN SOME FOSSIL CEPHALOPODA. 101 The smaller example (No. C. 51iO h), wliich exliibits the muscular attachment, has the following dimensions (excluding both the height of the peripheral spines and the inflated terminal portion of the body-chamber) : — diameter 15 mm. ; width of umbilicus 4 mm. ; height of outer whorl 6'5 ; thickness 1 mm., the whorl at the base of the body- chamber being 4'5 mm. high and 3 mm. thick. In this specimen the outer boundary of the forward ijrolongation of the muscular attachment is slightly more distinct than in the example already described, and appears to be continued backward as a dark (not incised) line into the second lateral lobe on one side of the specimen and on to the second lateral saddle on the opposite side, but txt about 2'5 mm. from the last septum it gives off a branch which curves outward towards the peripheiy, where it is bent slightly forward. The width of the forward prolongation of the impression is 1"5 mm., the most anterior part of the impression being 5 mm. from the last suture-line. The form and position of the muscular attachment in this species are also exhibited by a specimen * (No. 22267) in the British Museum from the Brown Jura ^, Beuren, Wiirtemberg ; it agrees with that already described in the example C. 5110 a. In this species, then, the shell-muscle seems to have been attached to the long flattened siu'face on the inner area of the body-chamber. The anterior border of the shell-muscle and of the annulus agrees almost precisely wdtli that described and figured by Oppel in his well-known figures of Ammonites \z:=^Oppelia\ steraspis from the Lithographic Stone of Solenhofen, Bavaria, a fact which supports Prof. Zittel's arrangement of the genera Distlchocerus and Oppella in the same subfamily [OppdirKc). S T E p H A N o c E R A T I D ^, Ncuuiayr, emend. Zittel. Stephanoceras, Waagen. 8tep}ianoceras BanJcsii, J. Sowerby, sp. — There are indications of the muscular attach- ment of the animal in Sowerby's type-specimen, which forms part of the British Museum Collection. It is a natural internal cast from the Inferior Oolite of the West of England, but the 2^recise locality is not recorded. Its diameter is about 230 mm. (about 9 inches) and its thickness about 150 mm. (or nearly six inches). One-third of the outer whorl is occupied by the body-chamber, the base of which is 50 mm. high and 130 mm. mde. The inner area of the whorl is convex and slopes considerably towards the umbilicus. On the inner area, and arising from the suture of the shell at a point 43 mm. in advance of the last sejjtum, there is a faint impressed line which jiasses backward and outward across the inner area, crossing the umbilical mai'gin at about 30 mm. posterior to its point of origin and 25 mm. from the suture of the shell. Arising from the suture of the shell, and at a point 20 mm. posterior to the line cxlready described, there is another groove which is rather more distinct and passes backward (more quickly than the anterior line) and outward for rather more than 10 mm., then, with a forwardly-concave curve, passes * A peculiaritj- about this specimcu is that the chambers are arranged as it were in pairs, a large loculus being succeeded by a small loculus. This is certainly the case with the last sixteen chambers. 102 ME. G. C. CEICK ON THE MUSCULAE ATTACHMENT OF THE to about the middle of tlie inner area, and then turning backward again it disappears in the broken surface of the specimen. The anterior line probably indicates the position of the anterior boundary of the shell-miiscle, and the posterior line possibly the posterior l)0undary of the same ; I think the latter cannot indicate the position of a former attachment of the anterior boundary, for it is not parallel to the anterior line. The other side of the specimen is too badly preserved to show the muscular impression. I have not observed any indications of the position of the annulus iu this specimen. Perisphikctes, Waagen. Ferisphinctea Achilles, d'Orbigny, sp. — The lateral aspect of the muscular impression in this sj)ecies is clearly shown in an example in the British Miiseum Collection (the smaller of the two specimens. No. 37017) from the "White Jura of Randen, near Schalfhausen (PI. 19, fig. 5). The example is a natural internal c;xst, almost completely denuded of the test. Its dimensions are : — diameter of shell 69 mm. ; width of umbilicus 28 mm. ; lieight of outer whorl 24*5 mm. ; greatest thickness (almost close to the umbilicus) 16 mm. The whorl is much compressed, and not much indented by the preceding whorl ; the body-chamber, measured along the median line of the periphery, is about 80 mm. long. The incised line marking the boundary of the muscular iuipression is j)lainly visible at the base of the body-chamber. At the inner edge of the whorl it is 5 '5 mm. in advance of the most anterior part of the last septum ; passing backward and out- ward for a short distance, it is then continued backward as an exceedingly faint line as far as the last septum in a direction nearly parallel to the inner edge of the whorl. The width of the im^u'cssion from the suture of the shell is 2'25 mm., of which only about one-half is visible in the lateral aspect of tiic shell. The boundary of the anterior portion of the scar is more deeply incised than the rest. The impression on the other side is similar, l)ut not qviite so distinct. There is no trace of the annulus. AsPIDOCEKATID^, Zittel. Peltoceras, Waagen. Peltoceras sp. — The impression of both shell-muscles is well seen on an example of Peltoceras sp. that was found in the Ampthill Clay at Ampthill Tunnel, and that formed part of my late father's collection (PL 20. fig. 1). The specimen, a mere fragment of the natural cast of a large shell, consists of the posterior part of the body-chamber ; it is about 150 mm. long. It is quadrangular in transverse section, and at the base of the body- chamber is 73 mm. high, 65 mm. thick (excludiug the ribs), and 72 mm. thick (including the ribs). An antiperipheral, autisiphoual, or dorsal aspect of the fragment disjilays two longitudinally-elongated oval scars, somewhat pointed anteriorly and bounded by a faintly impressed line. The outer boundary of each area terminates ioimediately above the inner branch of the lateral saddle, while the inner boundary passes down close to the outer side of the antij^eripheral (or autisiphoual) saddle on the same side. The anterior extremity ANIMAL TO ITS SHELL IN SOME EOSSIL CEPHALOPODA. 103 of each impression is situated on the edge of the impressed zone, so that about one- third of the width of the impression is on the impressed zone, the other two-thirds being on the inner area of the whorl. The two impressions are nearest together at about half their length ; they are here 19 mm. apart, while the posterior extremities of their inner boundaries are 32 mm. apart. The length of each impression (measured from the anterior extremity of the inner branch of the lateral saddle) is about 50 mm. ; the greatest width being 17 mm. No traces of the annulus have been observed in this specimen. In the middle of the peripheral area, and extending a short distance into the peripheral lobe, there is a broad, faint, longitudinal depression about 25 mm. long and 3'0 mm. wide, slightly wider anteriorly than posteriorly. Possibly this had something to do ^^'ith the muscular attachment of the animal to its shell. On one side of this there is a narrower, shorter, and less distinct longitudinal depression. AsPiDOCERAS, Zittel. Aspidoceras sp. — Among the si^ecimens which Dr. J. W. Gregory collected from " below Chamgamwe, opposite Mombasa Island," East Africa, there is a fragment of an Aspidoceras which displays the muscular impression ; it is merely the natural cast of the posterior part of the body-chamber, and this is incomplete on one side (PI. 19. iigs. 6, 7). The section of the base is obtusely cordate, and measures about 43 mm. in width and 34 mm. in height. The impression is seen on the inner area of the whorl as a faintly- impressed line, which, ai'ising from the edge of the impressed zone (corresponding to the suture of the shell) at a point about 15 mm. in advance of the base of the body-cliamber, passes outward, describing a curve slightly convex forward, and when at a distance of 9'5 mm. from the edge of the impressed zone jiasses backward nearly parallel to this edge for a distance of at least 10 mm. It then seems to divide into several very faint grooves which are continued backward nearly to the base of the body-chamber. No indications of the annulus can be seen. Unfortunately the cast is very imperfect on the opposite side. C O S M 0 C E R A T I D .E, Zittel. Parkinsonia, Bayle. Parkmsonia Parkinsoni, J. Sowerby, sp. — The muscular impressions are well shown in a Avell-preserved example of this species in the British Museum Collection (No. 9) from the Inferior Oolite of Bayeux (Calvados), Prance. PuUy two-thirds of the outer whorl are occupied by the body-chamber ; the test having been removed from the greater part of this chamber, the internal cast is well displayed. The dimensions of the specimens are : — diameter of shell 213-5 mm. ; width of umbilicus 58 mm. ; height of outer whorl 49 mm. ; greatest thickness of ditto o2'5 mm. The body-chamber appears to be fairly complete ; its length measured along the median line of the periphery is 287 mm. ; and the dimensions of its base are : — height 31'5 ram., greatest thickness (which is at about one-third of the height of the whorl from the inner edge) 22-5 mm. The muscular impression on each side can be seen. At the suture of tlie shell the SECOXD SERIES. — ZOOLOGY, VOL. VII. 15 104 MR. G. C. CRICK ON THE MUSCULAR ATTACHMENT OF THE anterior border, an incised line, is 9'5 mm. in advance of that portion of the lateral saddle of the last suture-line which is close to the umbilical margin ; it passes thence outward and backward nearly parallel t6 the suture and almost as far back as the last septum, when it appears to turn ovitward and pass over this portion of the lateral saddle as a feebly-depressed rather than an incised line. This last-mentioned portion of the boundary may have been produced by the anterior border of the annulus. The greatest width of the portion of the impression that is visible is 5 mm., and it is confined to the somewhat sloping inner area of the whorl. Its surface is a little roughened, and the ira^^ression seems to have occupied the lol)e which is situated on the inner area of the whorl. An irregular, waved, depressed line passes over the siphonal lobe on eacli side of the body-chamber, and this most probably indicates the position of a further jDortion of the anterior border of the annulus. P R I o N o T R o P I D .E, Zittel. ScHLCENBACHiA, Neumayr. Schloenhachia GoodhalU, J. Sowerby, sp. — The remains of the muscular attachment are exhibited by an example of this species in the British Museum Collection (No. 46491). The specimen is from the Greensand of Blackdown, Dorset. It consists of a segment of two adjacent whorls, the outer of \\hich exhibits at about its mid-length the basal l^ortion of the body-chamber, which on the left side is entirely denuded of the test. The whorl is carinated, much compressed laterally, and its sides nearly parallel ; at the base of the body-chamber it is 80 mm. high and 45 mm. wide or thick, the inner area being concave, sloping towards the umbilicus, and 12 mm. wide. On the left side, on the internal cast of the body-chamber, the anterior boundary of the muscular scar is seen as an incised line arising at the suture of the shell at about 30 mm. in advance of the corresponding portion of the last septum ; passing obliquely backward and outward across the inner area, it crosses the umbilical margin at about 22 mm. in advance of the corresponding part of the last septum ; it continues outward and backward for a short distance, and then divides ; one part joasses backward and outward for a very short distance and then disapjiears ; the other passes backward nearly parallel to, and at a distance of 4 mm. from, the umbilical margin, and disappears shortly before reaching the suture-line of the last septum. The branch which passes outward is probably the com- mencement of the anterior border of the annulus, the other being the outer boundary of the muscular scar. Arising from the suture of the shell, 13 mm. posterior to the line already described, and crossing the inner area nearly parallel to the same line, there is another very feeble groove, which scarcely crosses the umbilical margin before reaching the last suture-line. Prom the appearance of the inner surface of the piece of the test which fits on to this portion of the specimen, it seems most probable that this feeble groove indicates the position of the posterior boundary of the muscular scar. The opposite side of the whorl is too imperfect to exhibit any trace of the muscular attachment. ANIMAL TO ITS SHELL IN SOME FOSSIL CEPHALOPODA. 105 At a point 29 mm. in advance of the small median saddle, the periphery is crossed by an incised line which on the left side passes inward and backward for about 8 mm., tlien passini? backward nearly parallel to, and at a distance of 8 mm., from the median line of the perijAery until it meets the external saddle. Traces of a precisely similar line can also be seen on the right side. This I believe to have also been part of tlie anterior boundary of the annulus. Clymenia., Miinster. CUjmeuia imdulata, Miinster. — The muscixlar impression in the genus Clymenia has been best observed in an example of this species in the British Museum Collection (No. 81826). It is from the Clymeuien-Kalk (Upper Devonian) of Scbiibelhammer, Bavaria (PI. 20. figs. 13, 14). This specimen, the outer wdiorl of which is somewhat imperfect, has the following dimensions : — Diameter of shell 48'5 mm. ; width of umbilicus 20'5 mm. ; height of outer whorllG mm. ; greatest thickness of ditto 12 mm. Nearly one half of the outer whorl is occupied by the body-chamber, its length, measured along the periphery, being 65 mm. ; its anterior portion is flattened for a distance of nearly 20 mm., the width of the flattened portion being 3 mm. It niay be remarked that in this genus the siphuncle is on the inner side of the whorl, and is relatively much larger than in the rest of the Ammonoidea. The internal cast of the posterior part of the body-chamber and of two or three of the succeeding loculi is denuded of the test. The base of the body- chamber is 12'5 mm. high and 10 mm. thick, the periphery here being quite rounded. One side of the basal portion is w^ell-preserved, but tlie other is somewhat imperfect. On the well-preserved side (fig. 13) a curved feebly-iucised line arises almost close to the suture of the shell and at a distance of 2 mm. from the last sejitum ; j)assing backward nearly parallel to the inner half of the lateral portion of the suture-line, it just clears the most anterior poi'tion of the latter and then passes straight on to the periphery, being at the centre of the latter about 2 mm. anterior to the last septum. It appears to cross the periphery without iuterruj)tion, but unfortunately the line here is somewhat obscured, and it is not impossible that the line here may be bent backward into a very minute V-shaped sinus. Another feebly-incised line proceeds from the suture of the shell and passes backward almost close to the last septum for some distance into the lateral lobe ; then turning forward it meets the anterior line, already described, at the most anterior point of the lateral portion of the suture-line. At a short distance from this junction and a little nearer the periphery, the anterior line gives off a bi-anch which gradually recedes from it and crosses the periphery about 1 mm. posterior to it. At the central portion of the periphery this line, like the anterior one, is somewhat obscured, but appears to cross the periphery without interruption, although it is not impossible that at the middle of the periphery this line may have had a very small V-shaped bend forward. It would seem, therefore, that the shell-muscle occupied the space between tlie suture of the shell and the outer side of the lateral lobe, and that the annulus extended from 15* 106 ME. G. 0. CEICK OX THE MUSCULAR ATTACHMENT OF THE tlie outer side of the lateral lobe across the perijihery as a narrow band having its greatest width at the centre of the periphery. Unfortunately, the opposite side of the body-chamber is not sufficiently well-preserved to show the muscular impression. Near the anterior end of the body-chamber the internal cast exhibits a wide and fairly deep groove, which crosses the lateral area in a curve which is concave forward, and projects considerably forward near the periphery. The lateral portion of the sroove is fully 0 mm. wide, but shallow, while the portion near the periphery is 4 mm. wide and deeper than the rest. The opposite side of the specimen is broken here, and does not show the groove. This groove, however, seems to have nothing to do with the muscular attachment of the animal, but is here mentioned for comparison with tlie groove which has been depicted in some Ammonoids (e. g. Pinacoceras*) as the anterior bonndarv of the muscle and of the annulus. GONIATITES. Gltphioceeas, Hyatt. Glyphioceras crenistria, J. Phillips, sp. — Among the Goniatites the form of the attach- ment of the animal to its shell is well seen in an example t of this species from Bolland, Yorkshire (PL 20. figs. 15, 16). It forms part of the British Museum Collection (No. C. 5080). It is a Avell-preserved example, having the following dimensions: — Diameter of shell 43 mm. ; width of umbilicus about 4*5 mm. ; height of outer whorl 21-5 mm.; thickness of ditto 27'5 mm. The body-chamber occupies five-sixths of the outer whorl, and, being denuded of the test, its internal cast is well displayed. The muscular attachment is so well preserved that it can be traced from the umbilicus on one side across the whorl to the umbilicus on the other side. As seen in a lateral aspect (fio". 15) the muscular scar is almost confined to the inner area of the whorl, and is depressed (on the internal cast) a little below the sui'rounding surface. The most anterior portion of its anterior boundary is a little more than 2 mm. in advance of the last septum. Passing backward and outward, this boundary just crosses the umbilical margin and reaches to within about 1 mm. of the septum, when, turning abruptly outward, it is continued as a very fine raised line at this distance from the septum, until passing into the lateral lobe, when it recedes very slightly from the suture-line. It crosses the lobe in an obtuse point 3"75 mm. from the point of the lobe, and, then passing up the other side of the lobe, it gradually approaches the suture-line and, just clearing the pointed external saddle, passes into the external lobe at about 1 mm. from the septum (fig. 16), at the same time separating into two very minutely -waved lines; at 5 mm. from the apex * Dr. E. V. Mojsisovics, ' Das Gebirge um Hallstatt," i. pi. xx. figs. 2 & 3 (^Pinacoceras insectum) ; pi. xx. figs. 8 & 9 {P. liumilc). t This is the specimen from which the suture-liue figured in Cat. Eoss. Ceph. British Museum, pt. iii. p. 101, fig. 76, was drawu. ANIMAL TO ITS SHELL IN SOME FOSSIL CEPHALOPODA. 107 of the external saddle these lines diverg-e, assume a direction at right angles to the plane of symmetry of the shell, and cross the external lohe with only a very sliglit V-shaped depression at the median line of the periphery. jS'early at the centre of the peripliery there is a small circular pit on the posterior lino. On the opposite side of the lobe, at about 1 mm. from the suture-line, these lines turn aliruptly forward, approach and coalesce, the resulting line having the same course as that on the opposite side. At the umbilical margin this line passes into the boundary of the muscular scar, which also has the same form as that on the opposite side. At the centre of the periphery and imme- diately behind the lines already described there is a feebly-depressed double line in the form of a horse-shoe, each side of which passes quite close to the outer side of each half of the small median saddle; the enclosed area is about 2 mm. wide. At a short distance anterior to the raised line already described there is a rather broad shallow groove, which first makes its appearance in the lateral lobe, becomes more distinct all around the external saddle, and disappears in the external lobe. Almost in the median line of the periphery, and slightly in advance of the apex of the external saddle, there is a minute, shallow, oval depression. A fairly-jirominent " normal line " extends some 7 or 8 mm. in advance of this depression, whilst the surface of the cast behind this depression and in front of the anterior boundary of the annulus bears numerous fine longitudinal raised lines. The depressed surface on the inner area of the whorl I regard as a portion of the muscular scar, and the faint line proceeding therefrom across the whorl I consider to be the impression of the annulus. The latter then seems to have been over a portion of its extent merely a line of attachment, but to have been 1 mm. wide on the periphery. In the present specimen the boundary of the muscular scar cannot be traced on to the inner or dorsal poi'tion ( = impressed area) of the whorl, Init this has lieen partially observed in another example of the same species in the British Museum Collection (the larger of the specimens Xo. C. 283). The diameter of the shell is 50 mm., the outer whorl (at the end of the septate part of the shell) being about one-half of the diameter. The specimen is entirely septate (PL 20. figs. 17, 18), tlie last septum being the base of the body-chamber. A portion of the test that formed the floor of the body-chamber still adheres to the preceding Avhorl, and on this the border of the muscular scar can be traced for a short distance. In this example the anterior border of the scar j^asses on to the floor of the body-chamber at about 3"5 mm. in advance of the last sej)tum ; for a short distance it is nearly parallel to the sejjtum, Imt at about the middle of the lateral saddle it turns towards the septum and seems to disapjiear ; unfortunately the test here has been broken away, and only a few fragments of the test remain on the opposite side of the whorl, so that its direction cannot be definitely traced further. On the floor of the body-chamber (fig. 18) there is also a strong raised line following the lobes and saddles of the last septum, and about 1 mm. in front of the same ; near the umbilicus it gradually approaches and then passes into the angle between the septum and the wall of the shell. This seems to have been a line of attachment of the body of the animal ; but, so far as can be made out from this example, it does not ajipear to have been continuous 108 MR. G. C. CRICK ON THE MUSCULAR ATTACHMENT OF THE with tlie anterior border of tlie muscular scar. It may have been the line of attachment of the j)osterior 2iortion of the body just prior to the formation of a new septum. Glypliioceras trimcatum, J. Phillij^s, sp. — An example of this species from the Carboniferous Limestone of St. Doulagh's, co. Kildare, Ireland, which has been lent me by Dr. A. H. Poord, T.G.S., also exhibits the anterior boundary of one of the muscular scars A^ery distinctly (PI. 20. figs. 19). The specimen consists of about five-sixths of a whorl, the whole constituting- part of the body-chamber, the base of which is fortunately preserved. Its dimensions are: — diameter 75 mm.; height of outer whorl 36 mm. ; ditto above jireceding Avhorl 20 mm. ; greatest thickness (which is at about the middle of the lateral area) 27 mm. ; width of umbilicus 10 mm. On one side of the specimen the anterior boundary of one of the muscular scars is preserved as an impressed line (on the internal cast of the body-chamber), which arises from tlie suture of the shell at about 4 mm. in advance of the last septum and curves gently outward and backward. It can only be traced for a length of about 6 mm., when it is obscured by the test ; at about one-half of this lencrth it is 45 mm. from tlie suture of the shell. Summary. As in the recent Nautilus, so in the Ammonoids, the shell was external and the animal was attached to its shell by means of " shell-muscles " and an " annulus." In the Ammonites and such allied forms as BacuUtes, Samites, &c. the shell- muscles Avere attached to the dorsal portion of the shell; they frequently either approximated or met each other in the median line of this region ; when they did not quite meet they Avere doubtless united by a more or less narrow band corresponding to the dorsal portion of the annulus in the recent Nautilus. My observations support the conclusion expressed by Dr. Waagcn that the line figured by Oppel on the body-chamber of Ammonites steraspis indicated the position of the anterior boundary of the annulus and of the shell-muscles, the latter being situated, as he supposed, upon the inner or umbilical portion of the lateral area of the AA^iorl. It is, howcAer, not a little sti*ange that in the species figured by Oppel the form of the muscular attachment differs somcAvhat considerably from that in the majority of the Ammonites which I have examined. It would seem that another intei'pretation is necessary for the figures Avhich have subsequently been given purporting to be the remains of the muscular attachment. In the earlier stages of develoj)ment and in the general form of the shell as well as in the aperture of certain species, affinities have been recognized betAA^een the Ammonoids and the Dibrauchiatcs. Prom the foregoing it is clear that the Ammonoid animal possessed a muscular attachment quite similar to that of the living Nautilus, the only recent genus of the Tetrabranchiates. Indications of the muscular attachment of the Ammonoid animal, instead of being rare, seem to be fairly common. There appears to be some ground for believing that its form is in j^art due to the shape of the transverse section of the Avhorl and to the ajN'imal to its shell in some fossil cephalopoda. 109 length of the hody-chamher, hut I venture to helieve that it will prove to ha clue to other causes, and also affoi'd an important character for the purposes of classification. Thus, among the forms of muscular attachment descrihcd in the present paper, the shell-muscles in Distichoceras and Oi^peUa seem to have been relatively much longer than those of the other forms herein described ; and hence one should expect these two genera to be closely related. Tliis affinity seems to be supported by their other characters, for they have been placed in the same subfamily, viz. the Oppelince*. Not only was the Ammonoid animal, like the Nautilus, at least at some periods, attached to its shell by means of the lobes and saddles of the posterior portion of the body (corresponding to those of the edge of the septum of its shell), but it seems from the foregoing descriptions that it was further provided with an annulus in addition to shell-muscles as in the recent Naiitilus. It would appear, therefore, as if the provision of an annulus were an absolute necessity to the animal in addition to the shell-muscles, and most probably Dr. Waagen's explanation of its occurrence is the correct one, viz., that the annulus and shell-muscles served not merely to hold the animal to its shell, but formed also an air-tight band around it, fastening the mantle to the shell. In conclusion, I desire to express my obligations to those who have in any way assisted me in my observations. To Dr. Woodward I am greatly indebted for valuable suggestions and much kind help, while to Mr. B. B. Woodward I am especially indebted for the advice which he has ever been most willing to give me. My thanks are also due to my colleague Mr. Edgar A. Smith for giving me every facility in the examination of the recent Cephalopoda under his charge. For the loan of specimens I have to thank Mr. F. L. Bradley, Mr. G. P. Harris, and the authorities of the Jermyn Street Museum. To Messrs. Sharman, Newton, and Allen, of that Museum, I am under deep obligations for all their kind help during the examination of the specimens under their charge. To Professor Howes, whose interest in my work has been a source of great encouragement to me, I desire to present my cordial thanks. EXPLANATION OF THE PLATES. The figures are of the natural size unless otherwise stated. X indicates tlie last septum. — * marks the remains of the muscular attachment. Plate 17. Baculites ovatus, Say (p. 77). Fig. 1. Dorsal aspect of fragment of body-chamber, showing anterior boundaries of muscular scars. 2. Right lateral aspect of same, showing on the left the boundary of the right muscular scar, and on the right the incised line on the ventral area. See Prof. Dr. K. A. v. Zittel, ' Grundziige der Palaoutologie,' p. 420. 110 ME. G. C. CEICK ON THE MUSCULAE ATTACHMENT OF THE Fig. 3. Ventral aspect of same, showing incised line on median portion. 4. Right lateral aspect of fragment of another example, showing anterior bonndary of right muscular scar and of a portion of the anuulus. Baculites vagina, Forbes (p. 78) . Fig. 5. Dorsal aspect of portion of body-chamber, showing near the base the anterior boundary of the two muscular scars. Hamites maxinms, J. Sowerby (p. 79). Fig. 6. Dorsal aspect of body-chamber, showing the two muscular scars near the base. 7. Left lateral aspect of the same. 8. Dorsal aspect of another example. 9. A portion of the same enlarged, showing the two scars, each bearing near its anterior border a small roughened patch, and also traces of a narrow band connecting the scars. X '{. Cr'wceras quadratum, n. sp. (pp. 74 & 79). Fig. 10. Dorsal aspect of a fragment of the body-chamber, exhibiting the two scars. 11. Left lateral aspect of the same. 12. Dorsal aspect of a frajjment of another example in which the anterior boundaries of the two scars nearly meet in the median line. 13. Left lateral aspect of the same. Ancyloceras Matheron'ianmii, d'Orbigny (p. 80). Fio-. 14. Left lateral aspect, showing portion of anterior boundary of left muscular scar, x \. 15. Right lateral aspect of a portion of the same, exhibiting the termination of the anterior boundary of the right muscular scar. X 5. If). Dorsal aspect of a portion of the same, showing the anterior boundaries of the two muscular scars, that on the riglit being much higher than that on the left. X \. Macroscaphites giyas, J. de C. Sowerby, sp. (p. 81). Fig. 17. Riglit lateral aspect. x\. 18. Left lateral aspect of a portion of the same, showing part of the boundary of the two scars. X 5. 19. Dorsal aspect of a portion of the same, also showing the anterior boundary of the two scars. X 5. Scaphites binodusiis, A. Roemer (p. 81). Fig. 20. Left lateral aspect. 21. Dorsal aspect of a portion of the same, with part of the body-chamber removed. Scaphites aqualis, J. Sowerby (p. 82). Fi"'. 22. Dorsal aspect of a fragment of the body-chamber. The curved lines at the lower part of tlie figure are the impressions of the ornaments of the septate part of the shell ; the anterior border of the muscular attachment is indicated by the arrow. ANIMAL TO ITS SHELL IN SOME FOSSIL CEPHALOPODA. Ill Plate 18. TurrUites tuberculatus, Bosc (p. 83). Fig. 1. Lateral aspect, two of the upper wliorls exhibiting the siphuncle {si), and tlic body-cliamber showing the muscular attachment, x ^. 2. Dorsal aspect of body-chamber of same, showing the muscular attachment. X \. TurrUites Mantelli, Sharpe (p. 83). Fig. 3. Lateral aspect of fragment, the lower whorl of which exhibits a portion of the boundary of the muscular attachment. x |. TurrUites {Heteroceras) polyplocus, Roemer (p. 84) . Fig. 4. Lateral aspect of terminal portion. X |. Oxynoticeras ? sp. (p. 85) . Fig. 5. Left lateral aspect of part of an example showing the muscular scar and the annulus. x 5. 0. Right lateral aspect of same, x f . Amaltheus spinatus, Bruguiere, sp. (p. 86). Fig. 7. Left lateral aspect of imperfect specimen, exhibiting on the inner portion of the whorl (opposite the arrow) the boundary of the muscular scar. Cardioceras excavutum, J. Sowerby, sp. (pp. 75 & 86). Fig. 8. Left lateral aspect, showing the muscular scar and the annulus. x h. Cardioceras aff. excavuto, J. Sowerby, sp. (p. 80). Fig. 9. Left lateral aspect, exhibiting the ar.tcrior boundary of the muscular scar on the inner portion of the whorl opposite the aiTow, and a linguiform impression on the middle of the lateral area. 10. Dorsal aspect of posterior portion of the body-chamber of the same, showing near the base the course of the anterior boundary of the muscular scar on the dorsal area of the whorl. Tissotia EwahU, v. Buch, sp. (p. 90). Fig. 11. Left lateral aspect, showing close to the umbilicus (opposite the arrow) one of the muscular scars, and, passing from it towaids the peripheiy, a portion of the anterior boundary of the annulus. x f . Plate 19. Cardioceras excaratnm, J. Sowerby, sp. (p. 86). Fig. ] . Right lateral aspect, x §. 2. Sagittal section of the natural internal cast of the body-chamber of the same specimen, showing anterior boundary of muscular scar on dorsal portion of whorl. X f . Cardioceras sp. (p. 89). Fig. 3. Left lateral aspect, showing anterior boundary of scar and of annulus. x 'j. 4. Ventral view of the same, exhibiting the impression on the central portion of that area. SECOND SERIES — ZOOLOGY. VOL. YII. 10 112 MR. G. C. CRICK ON THE MUSCULAR ATTACHMENT OF THE Perisphinctes Achilles, d'Orbigny, sp. (p. 102). Fig. 5. Left lateral aspect. The boundary of the muscular scar is seen on the inner portion of the outer whorl opposite the small arrow. Aspidoceras .«/*. (p. 103). Fig. 6. Dorsal aspect of a fragment of the body-chamber. 7. Right lateral aspect of the same. Arietites obtusus, J. Sowerby, sp. (p. 93). Fig. 8. Left lateral aspect of a portion of a specimen, showing the muscular scar on the inner portion of the outer whorl opposite the arrow. 9. Right lateral aspect of the same. 10. Ventral aspect of the same, showing the liuguiform impression on the central portion, and anterior to it an /-shaped incised line in the depression on each side of the keel. Lytoceras fimbriatum, J. Sowerby, sp. (p. 91). Fig. 11. Dorsal aspect of the posterior portion of the body-chamber, exhibiting the left and a portion of the right muscular scar. 12. Left lateral aspect of the same. Lytoceras cornucopite, Young & Bird, sp. (p. 91). Fig. 13. Dorsal aspect of the posterior portion of the body-chamber. 14. Right lateral aspect of the same. 15. Right lateral aspect of another specimen, showing the small oval scar on the inner part of the outer whorl opposite the aiTOw. 16. Portion of same enlarged, showing muscular scar. X f . 17. Ventral aspect of portion of same enlarged, exhibiting the linguiform impression, x "f . Plate 20. Peltoceras sp. (p. 102). Fig. 1. Dorsal aspect of fragment, showing the two elongated muscular scars. X ^. ^goceras capricorimm, Schloth., sp. (p. 95). Fig. 2. Right lateral aspect. The muscular scar is the small white patch on the inner portion of the outer whorl opposite the arrow. j^goceras laqueolum, Schloenbach, sp. (p. 96). Fig. 3. Dorsal aspect of fragment of body-chamber, showing the left muscular scar and part of the right. 4. Left lateral aspect of the same. ANIMAL TO ITS SHELL IN SOME FOSSIL CEPHALOPODA. 113 Hecticoceras hecticuni, Reinecke, sp. (p. 98). Fig. 5. Left lateral aspect, x if . 6. Right lateral aspect of same. x %. 7. Ventral aspect of portion of same. X i'. Distichoceras Baiigieri, d'Orbiguy, sp. (p. 100). Fig. 8. Right lateral aspect, x j. CEcotraustes crenatus, Bruguiere, sp. (p. 99). Fig. 9. Right lateral aspect. X f. Arietites raricostatus , Zieten, sp. (p. 94). Fig. 10. Left lateral aspect of entire specimen. 11. Dorsal aspect of the portion ab, which is a natural internal cast of part of the body-chamber showing at the base the remains of the two muscular scars. Sonn'uiia sp. (p. 98) . Fig. 12. Left lateral aspect, exhibiting the muscular scar on the inner portion of the outer whorl opposite the arrow, and also the adjacent indentations. Clymenia undulata, Miinster (p. 105.) Fig. 13. Right lateral aspect. 14. Peripheral aspect of the same. Glyphioceras crenistria, J. Phillips, sp. (p. lOG). Fig. 15. Left lateral aspect. The fine white line opposite the arrow is the boundary of the annulus ; the subtriangular patch into which it passes near the umbilicus is the left muscular sear, the black line posterior to it is the edge of the last septum, the lighter portion near it being the broken edge of the test. 16. Front view of the same, the arrow pointing to the line indicating the annulus. 17. Left lateral aspect of another example. The anterior septum is the base of the body-chamber. The muscular scar is indicated by the light patch (opposite the arrow) on the piece of the test, one layer of which is part of the floor of the body-chamber. 18. View of same from above ; the raised line on the piece of test wliicli is part of the floor of the body-chamber being indicated by the white line marked a. Glyphioceras tnmcatum, J. Phillips, sp. (p. 108). Fig. 19. Left lateral aspect. X j. Crick. TrAKS.LiNN .Soc . ZoOL .SER.2.VoL,.Vn .PI .17. z'^"'^'*^, -^ r^i;?:sf"^"'1 ?i"? «\rf ^.^/r^fWi '"t*),,.-. I' J .-& PL -.- f (A frt r ') \^ l '^1 ^^ »-4^ c ^ '^*-l ^~^^^/ -=''<«lii:4^ •*- -aa^iij^ J Green, del etlitb.. -. V rr::^?;^,. — ^ j_ — — ^tt ^ — ^isSi ManternBros irtLp . MUSCULAR IMPRESSIONS IN FOSSIL CEPHALOPODA. Cricls. . Trans .Linn Soc Zool . Ser.2 .Vol .VII, PI .13. i > //> 4P?>- J Greeiv del et litlr\ . MUSCULAP. IMPRESSIONS IN FOSSIL CEPHALOPODA MirLterruBros irap. Crick. Tr-Ans Likn Soc . ZooL . Ser . 2 .Vol, VII. PI . 20 . J Green del etlith IvIiTLterrL Bros nVp ■ MUSCULAR IMPRESSIONS IIJ FOSSIL CEPHALOPODA. LINNEAN SOCIETY OF LONDON. MEMORANDA CONCERNING TRANSACTIONS. The First Series of the Transactions, containing both Botanical and Zoological contributions, has been completed in 30 Vols., and a few entire sets are still for sale. Only certain single volumes, or parts to complete sets, may be obtained at the original prices. The price of the Index to Vols. 1-25 is 8s. to the public, and Cs. to Eellows ; to Vols. 26-30, 4s. to the public, and 3.?. to Fellows. The Second Series of the Transactions is divided into Zoological and Botanical sections. The prices of the Zoological parts of these which have been published are as undermentioned (for the Botanical parts see Botanical wrapper) : — When Price to the Vohnne. Published. Public. £ S. d. Second Seeies. — Zoologt. I. Part I. 1875 1 4 0 . . Part II. 1875 0 6 0 . . Part III. 187G 1 8 0 . . Part IV. 1877 0 16 0 .. Part V. 1877 0 18 0 . . Part VI. 1877 1 2 0 . . Part VII. 1878 1 16 0 .. Part VIII. 1879 1 0 0 . . II. Part I. 1879 1 4 0 . . Part II. 1881 0 15 0 .. Part III. 1882 1 8 0 . . Part IV. 1882 0 7 6 . . Part V. 1882 0 3 0 .. Part VI. 1883 1 0 0 . . Part VII. 1883 0 5 0 . , Part VIII. 1883 0 3 0 . . Part IX. 1883 0 3 0 .. Part X. 1884 0 4 6 . . Part XI. 1884 0 10 0 . . Part XII. 1885 0 6 0 . . Part XIII. 1884 0 6 0 . . Part XIV. 1885 0 6 0 . . Part XV. 1885 0 4 6 . . Part XVI. 1885 0 5 0 . . Part XVII. 1886 0 3 0 . , PartXVIlI. 1888 0 2 6 .. III. Part I. 1884 114 0 . , Part II. 1884 1 12 0 . , Part III. 1885 110 0 . . Part IV. 1885 0 8 0 . . p rice to Fellow 7a. £ s. d. .. 0 18 0 . 0 4 6 .. 1 1 0 .. 0 12 0 .. 0 13 6 . 0 16 6 .. 1 7 0 .. 0 15 0 .. 0 18 0 . 0 11 6 .. 1 1 0 , 0 5 6 . 0 2 3 . 0 15 0 .. 0 3 9 .. 0 2 3 .. 0 2 3 .. 0 3 6 .. 0 7 6 .. 0 4 6 .. 0 4 6 .. 0 4 6 .. 0 3 6 .. 0 3 9 .. 0 2 3 .. 0 2 0 .. 1 5 0 .. 1 4 0 .. 1 2 6 .. 0 6 0 When Price to the Voliiine. rt 1 I- I 1 I) I I- Piihhslied. rublic. £ s. d. Second Series. — Zoolouy (continued). III. Part V. 1887 0 8 0 .... Part VI. 1888 0 6 0 IV. Part I. 1886 1 4 0 .... Part II. 1887 1 8 0 . . . . Part III. 1888 0 16 0 . . . . V. Part I. 1888 0 12 0 Part II. 1888 0 5 0 . . . . Part III. 1889 1 7 0 Part IV. 1890 0 12 0 . . . . Part V. 1890 0 6 0 Part VI. 1891 0 12 0 . . . . Part VTI. 1891 0 6 0 .... Part VIII. 1892. ... 0 8 0 .... Part IX. 1892. . ..012 0 . . . . Part ' X. 1893 1 8 0 Part XI. 1894 0 2 6 . . . . VI. Part I. 1894 2 0 0 , , . . Part II. 1894 1 11 0 .... Part III. 1894 0 10 0 . . . . Part IV. 1896 1 4 0 . . . . Part V. 1896 0 10 0 . . . . Part VI. 1896 0 8 0 Part VII. 1896. . . . 0 12 0 . . . . PartVIII. 1897 0 2 6 . . . . VII. Part I. 1896 0 10 0 Part II. 1897 0 12 0 . . . . Part III. 1897 0 6 0 . . . . Part IV. 1898 0 10 0 . . . . Price to Fellows. d. 0 6 0 0 4 6 0 18 0 1 1 0 0 12 0 0 9 0 0 3 9 1 0 0 0 9 0 0 4 6 0 9 0 0 4 6 0 6 0 0 9 0 1 1 0 0 2 0 1 10 0 1 3 3 0 7 6 0 18 0 0 7 6 0 6 0 0 9 0 0 2 0 0 7 6 0 9 0 0 4 6 0 7 6 2nd Ser. ZOOLOGY.] [VOL. VII. PART 5. THE ^y/ / 9 TRANSACTIONS OF THE LINNEAN SOCIETY OF LONDON. OBSERVATIONS ON THE STRUCTURE AND MORPHOLOGY OF THE CRANIAL NERVES AND LATERAL SENSE ORGANS OF FISHES; WITH SPECIAL REFERENCE TO THE GENUS GADUS. BY FllANK J. COLE, DEMON.STRATOn OF ZOOLOGY, UNIVERSITY COLLEGE, LIVERPOOL. {Comviunicnied by Pro^. W. A. Herdman, D.Sc, F.Il.S.) LONDON: PRINTED FOR TRE LINNEAN SOCIETY BY TAVLOli AND FRANCIS. RED LION COURT, FLEET STREET. SOLD AT THE SOCIETY'S APARTMENTS, BURLINGTON-HOUSB, PICCADILLY. W., AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW. ''■■' Ontoh(^r 1898. [ 115 ] V. Observations on the St7'nctiire and Moiyhology of the Cranial Nerves and Lateral Sense Organs of Fishes ; tcith special reference to the Genus Gadiis. By Frank J, CoIj'e, Demonstrator of Zoology, University College, Liverpool. [Communicated by Prof. W. A. Herdman, D.Sc, F.B.S.) AUG 20 >899 (Plates 21-23.) Read 3rd March, 1898. Contents. Page A. Introduction 116 B. Historical 117 C. Sensory Canals in General 120 Conclusions 120 D. Sensory Canals of Gadus morrhtia 122 (1) Supra-orbital Canal 122 (2) Lifra-orbital Canal 123 (3) Hyomandibular Canal 126 (4) Lateralis Canal 127 Supra-temporal Portion 127 Lateral or Body Portion 127 E. Gadus morrlMa and Gadiis virens compared . . 129 F. The Sense Organs on the Lateral line Canals . . 129 G. Other Sense Organs belonging to the Lateral line System 131 H. Eelation of the Sensory Canals to the SkuU . . 131 Conclusions 132 I. The Trigemino-facial Ganglionic Complex .... 133 J. Morphology of the Facial Ganglion 136 (1) Conclusions 142 (2) Typical Branchial Nerve 144 (3) Facial Nerve of Man 144 K. Morphology of Jacobson's Anastomosis 145 Conclusions 147 L. Metamerism of the Lateral line System .... 148 (1 ) Conclusions 152 (2) Value of the Lateral line Nerves 154 M. Innervation of the Sensory Canals in Gadus . . 156 (1) Supra-orbital Canal 156 Superficial Ophthalmic Trunk 156 SECOND SERIES. — ZOOLOGY, VOL. VII. Page (2) Infra-orbital Canal 158 Buccal Trunk 158 (3) Hyomandibular Canal 161 Hj'omandibular Trunk 161 (4) Lateralis Canal 164 Laterahs Trunk 164 N. Structure and Morphology of the R. lateralis accessorius (" trigemini '") 166 (1) Condition in Gadus 17.5 (2) General Conclusions 177 0. Morphological Value of the Lateralis Nerve of Petromyzon . 178 Conclusions 179 P. Gadus compared with other forms 179 Amia 181 Q. The Lateral Sense Organs of Vertebrates and Invertebrates 185 Conclusions 186 Iv. The Phylogeuy of the Sensory Canals 187 Conclusions 192 S. The Lateral Sense Organs and the Auditory Organ 194 Conclusions 197 T. Special Considerations 199 (1) Friant (1879) 199 (2) Morphology of the Facial Nerve of Fishes 200 Summary 203 Bibliography 20U 17 116 MB. F. J. COLE ON THE STRUCTURE AND MORPHOLOGY OF A. Introduction. 1 HIS investigation was first suggested to me as a desirable study by my former teacher, Professor J. C. Ewart, F.R.S., some years ago. I had at the time contemplated worknig at a Pleuronectid fish^ — allured by the problems which the asymmetrical head affords ; but it was pointed out that I should be better equipped for such difficult work were I to study a modern symmetrical Teleostean first, and that in fact it were almost an impossi- bility to understand the former without having first investigated the latter. I therefore abandoned my first project, and decided to work out the anatomy of the lateral line system and its nerves on the common Codfish. That this work was necessary, and indeed essential to the proper understanding of this complicated system, is very obvious from a perusal of the literature. In spite of the enormous bulk of the latter, and the ponderous theses and memoirs tlmt have been written on the subject of the present communication, we still know very little about the fine anatomy of the lateral line system, and very few authors have taken the trouble to grapple with the literature and to conduct their investigations on a strictly logical and scientific basis. To Mr. Edward Phelps Allis, who must be regarded as the pioneer in this work, vertebrate morphology owes much. He was not only the first to make a comjilete study of the topographical development of the lateral canals, and to give a correct and detailed account of their innervation, but many important morphological deductions have been directly inspii-ed by his work. This author has recently completed his work on Jmia. It is a completion upon which I may be permitted to cordially congratulate the author, and to express the hope that he will continue on other forms the studies he has pvn-sued with such conspicuous success on Amia. Ewart has followed Allis with an account of his skilled dissections of the very complex lateral line system of Lcemargus and Rata, and Pollard has added some details on the ancient Siluroid Teleosteans. Pinkus has very largely filled in a gap by an able account of the nerves of Protoptertis, and the writer has published a description of the system in Chimcera. The most pressing want, therefore, was an investigation of the details in a modern specialised Teleostean, and the pi-esent memoir is largely an attempt to supply this deficiency in our knowledge of this interesting apparatus in Gadus. The experience gained whilst working under Professor Ewart, and afterwards in investigating Chimcera, has been most valuable in the interpretation of the nerves of the Codfish. That the skeleton of this creature represents a very specialised condition is a palaeontological fact, and it is hence not surprising to find its nerves in an equally advanced stage. My previous work on the simpler cartilaginous fishes (most of which has yet to be published) has prevented me from going astray in the interpretation of the nerves, and has enabled me to see in the specialised Cod the disguised form of the primitive cartilaginous fish. The separation of the components in the Cod has necessarily been effected with the microscope, and one cannot urge too strongly the importance of this instrument in the elucidation of cranial nerve morphology. Investigations based solely on microscopic or naked-eye anatomy olten contain gross but pardonable errors ; and whilst, on the other hand, a combination of the two methods docs not indeed remove THE CRAISIAL NERVES AND LATERAL SENSE ORGANS OF FISHES. 117 the possibility of error, it certainly reduces mistakes to within a somewhat reasonable limit. Had I relied entirely upon sections, I should (as Pollard actually did) have described the lateralis accessorius as a modified lateral line nerve ; whilst, had I confined myself to naked-eye dissection, the very interesting condition of the " facial ganglion" would have entirely escaped observation. Pigure 1 (PL 21) and the description of the sensory canals themselves (pp. 122-129) are based on the dissection of a single specimen, which, as shown in the Plate, had its anterior and posterior narial openings abnormally fused together. The right side, how- ever, was normal in this respect. This specimen measured 237 mm. from tlie snout to the anterior extremity of the dorsal tin *, 210 mm. from the snout to the dorsal attacliment of the opercular fold, and 80 mm. from the barbel to the apex of the mid-ventral ano'le formed by the fusion of the two opercular folds. Its girth from the anterior extremity of the dorsal fin to the dorsal border of the base of the pelvic fin was 195 mm. The young Gadus vire/is used in the microscopic part of the work were obtained at St. Andrews, where, on this as on many other occasions. Professor W. C. Mcintosh, P. U.S., very kindly placed a table in his admirable new laboratory at my disposal. I also have to thank him for collecting the material and keeping it alive in the tanks when the exigencies of teaching left little time at my disposal. The smallest specimen I brought away from St. Andrews was 35 mm. long, but those sectioned were larger than that. They may therefore be described as young adults. To Professor G. B. Howes, P.Pt.S., I am indebted for much kindly assistance and encouragement, and indeed both to him and to my valued chief. Professor W. A. Herdman, F.R.S., I owe " many acts of kindness that I can never repay, and have no wish to, for I prefer continuing the obligation." The preparation of the bibliography, of which a section is printed at the end of this paper, and which will be published separately wdien completed, has necessarily laid me under obligations to many librarians, as well as to Professor Howes. Of these I should like to mention Mr. A. W. KajJiiel, of the Linnean Society ; Mr. P. H. Waterhouse, of the Zoological Society ; Mr. A. H. White, of the Ptoyal Society; and Mr. B. B. Woodward, of the Natural History Museum. To these gentlemen, who freely placed their knowledge and time at my disposal, I must express my heartiest thanks, B. Historical (G^i-f^s) t- The first reference I find to the sensory canals and cranial nerves of the Cod is contained in the work of Monro (144) published in 1785. Monro considered the lateral canals glandular oi'gans (i. e. " mucous canals "), and gives a large but superficial figure of Gadus in plate v., which shows the supra-orbital commissure and the median unpaired tubule, but which overlooks the supra-temporal canal. This author was the * The point of extension forward of the dorsal fin is extremely variable even in the same species, as may readily be seen by comparing common Cod placed side by side in a fishmonger's shop. t I purposely omit references to text-books that are in constant use, such as T. J. Parker's ' Zootomy,' &c. [Since this section was written, several other Oadus references have been encountered, but these will be fully treated in my complete Bibliography.] 17* 118 MR. F, J. COLE ON THE STRUCTURE AND MORPHOLOGY OF first to discover the nerve-supply of Lorenzini's ampuUse in Selachians. In 1825 Desmoulins and Magendie (60) gave a figure of the brain and roots of the nerves of the Cod, whilst Giltay in 183i (88) figured and described the sympathetic nervous system, of wliich more presently. Leuret and Gratiolet, in a fine work (1839-1857, 119) treat of some Invertebrate nervous systems, and give a brief account of the central nervous system and cranial nerve roots of a few fishes, including the Cod, as also does Alcock (3, 1839). In 1842 Stannius (198) commenced his studies on the peripheral nervous system, and published a careful description of the nerves of "■ Gudus callarias" {=G. mor/'hua). The results obtained on this form were included in his well-known general treatise published in 1849, which will be discussed elsewhere. In the same year he wrote a short note on the lateral canals of fishes (197), in which he briefly but accurately describes the lateral canals of " Gadus callarias^ He corrects Monro's mistake re tlie supra-temporal canal, and devotes a few lines to the nerve-supply, remarking that the supra-temporal lateral line nerve is homologous with the Ramus auricula ris vagi of higher Vertebrates. This homology can, of course, be maintained no longer. Bonsdorff', in 1846 (30), gives us one of an admirable series of papers on cranial nerves, and treats of the trigeminus and cephalic sympathetic of " Gadus lota " [=Lota vulgaris). He gives a good figure of the brain and nerves, but in his interpretation of the latter was greatly misled by comparing them in too much detail with the Mammalian nervous system. Stannius's monograph (1849, 199) is concerned largely with Gadus, and will be discussed further in the text. Besides giving a fine figure of the lateralis accessorius, and showing further its connection with the spinal nerves (in " Gadus callarias "), he gives a further figure of the cephalic sympathetic and arteries in the same animal. Leydig (1850, 120) was the first to describe the histology of the lateral sense organs in Lota vulgaris, and shoAvs the lateral line ossicle with its sense organ and perforating nerve. In Swan's beautiful atlas (1804, 205) there are several good figures of our type, but tlie accompanying text is unfortunately very meagre. The figure of the Cod's brain and nerves here given has been copied into several text-books, and may be seen in the first volume of Owen's 'Vertebrates,' p. 298. In Traquair's work on the asymmetry of the Pleuroneotidse (1865, 207), the first accurate account of the topography of the sensory canals of Gadus and their relations to the bones of the skull is given. A figure (tab. 32. fig. 1) is published indicating the approximate position of the various canals (omitting details), and Dr. Traquair puts to brilliant use the lateral line system in sustaining his views on the Pleuronectid skull, Owen (1866, 149) makes some references to the nerves of the Cod, but his statements and figures are compiled from the works of Swan, Alcock, and Cuvier and Valenciennes. The histology of the brain of " Gadus lota " has been treated of by Stied;i (1868, 201), who examined tiie nerve-tracts in the brain and spinal cord, and devotes a few notes to the roots of the nerves. A very important work, historically, which has hitherto been overlooked, and to whicii I shall refer in some detail later on, was published in 1879 by A. Friaat (73). This author describes and figures the brain and nerves of many bony fishes, including THE CRANIAL NERVES AND LATERAL SENSE ORGANS OF EISHES. 119 " Oadus carbonarius'' {=G. vlrens) (pi. v.) and " Gadus lota" (pi. vi.), and his work is remarkably accurate for the time it was published, anf'^ must certainly be regarded as a pioneer research. Cunningham (1890, 55) says (p. 77) : "In the Cod, which may be taken as exhibiting pretty nearly tlie original condition of the cephalic tubes of the Sole and other tlat flshes, tliere are no superficial sense-organs on the head, and the tubes of the two sides are symmetrically arranged." Omitting the last, I cannot acquiesce to any of these statements. The lateral sense organs have xicdoubtedly advanced purl passu A\ith the development of the skull and the evolution of the asymmetry of the Pleuro- nectidge. Cunningham's figure of the sensory canals of the Cod is copied " with slight modifications " from the figure given by Traquair in his Pleuronectid paper. Traquair's figure, as far as it goes, is perfectly accurate, and I must confess that Cunningham's modifications are somewhat unfortunate — especially the one that connects the liyo- mandibular with the lateralis canal. Further, the statement as to the absence of superficial sense organs is a curious oversiglit, especially as the mandilnxlar line of pit oi'gans is more obvious than these organs are in any other fish I am acquainted with. Allis, in his last Amia paper (1S97, 6), makes some passing references to Gadus, which he states are based on work being done in his laboratory by Dr. Dewitz and Mr. Samuel Mathers. On p. 631 he says : — " In Gadus there is a line of surface organs on the outer surface of the opercular bones immediately behind the preoperculum. They are all innervated by a special branch of the mandibularis externus facialis, and are therefore of the character of pit organs, and unquestionably represent in Gadus one of the cheek- lines of Amia or a similar line not found in Amia." Again, on pp. 032-633: — '' h\ Gadus a line of surface organs is found along the lower edge of the mandible, parallel to the mandibular canal, and it is innervated by a long branch of the externus facialis, which first runs forward through the adductor mandibuUe, to tlie hind edge of tiie infraorbital canal behind the eye, and there turns downwards and reaches the mandible. A nerve in Esox corresponding in position to this nerve in Gad/is innervates a line of surface organs lying on the upper jaw immediately below the infraorbital canal." Respecting the innervation of the latter organs lie says (p. 637) : — " Tiiis nerve [/. e. a branch of the externus facialis] in Silurus may l)e a l)ranch cori-esponding to the one which in Amia innervates the mandibular line of pit organs, and in Gadus innervates a mandibular line of the slit like organs peculiar to that fish." Finally Goronowitsch (1897, 90) has given an admirable account of the trigemino- facial complex of JLoia vulgaris, in which tlie histology of the medulla and the nerve tracts of the ganglionic complex are ably elucidated and described. It is thus seen that our knowledge of the lateral sense organs and their associated nerves in Gadus is still of a vei'y imperfect description, and in fact all the work of any pretension was done before the lateral line system w'as properly understood, and l)efoie the researches of Allis, Ewart, and Strong had placed our knowledge of this interesting- system on a secure basis of carefully ascertained fact and logical deduction. It is the purpose of the present communication to supply this defect, to add some new facts 10 our knowledge of the lateral organs, and jJerhaps to elucidate many points that are still somewhat obscure. I find with Allis that " as the work has j)rogressed it has repeatedly 120 ME. F. J. COLE ON THE STRUCTURE AND MORPHOLOGY OF been found necessary to study details, and to include in the investigation whole subjects not contemplated in the beginning." The more important additions are the sections on the lateralis accessorius (which I trust is here correctly understood for the fii'st time), the phytogeny of the lateral canals, the relations of the latter to the auditory organ, the metamerism of the lateral sense organs, and the morphology of the " facial " ganglion and Jacobson's anastomosis. Hence the somewhat comprehensive title of the present communication. C. Sensory Canals in General. The sensory canals of the Cod differ in no essential respect from those of the typical form. Indeed the uniformity that exists in this connection, even amongst fishes of widelv different classification, is one of the most remarkable incidents in the history of this remarkable system. The same fundamental plan may be seen in all, whether we take an Elasmobranch, a Holocephalan, a Dipnoan, a "Ganoid," or a " Teleostean." Even in the Palaeozoic Coccosteaus and Asterolepids, we see, thanks to the admirable and laborious researches of Traquair, the same generalised type. In Coccosteus, for example, we may consider the horizontal canal traversing the anterior dorso-lateral, ihe external occipital, and the central plates, as the lateral or body canal. This, after a break, gives rise to a supra-orbital canal, traversing the central and pre-orl)ital plates, and an infra- orbital, situated on the central, post-orbital, and maxillary ossicles. Tliere is also a supra-temporal commissure situated on the two centrals, and this probably marked the anterior extremity of the lateral canal. The hyomandibular series is represented by canals on the post-orbital, marginal, and maxillary, and perhaps by the other canals in this region. In PtericJithi/s, again, the sensory canals are very simplified, being reduced to representatives of the lateral and iafra-orbital canals, the former being connected anteriorly on the median-occipital plate by the supra-temj)oral commissure. Here the practical absence of a supra-orbital canal is perhaps correlated with the curious condition of the orbits in the Asterolepid fishes. As I have elsewhere maintained (46, p. 181), and as the pioneer researches of Allis must be held to have established, the sensory canal systems of fishes, both recent and fossil, may be reduced to a common type. Pending the production of weighty evidence to the contrary, this must be considered as one of the best-established conclusions attained by the careful study of these organs by many able observers. This common type may be summarised as follows : — (1) A supra-orbital canal passing, as its name implies, over the eye. Innervated by the superficial ophthalmic lateral line nerve. Anteriorly may anastomose with the infra-orbital canal or end blindly ; posteriorly anastomoses either Avith the infra-orbital or lateral canal, but more often with the former. (2) An infra-orbital canal coursing underneath the eye. Innervated by the buccal -l-otic lateral line nerves. Anteriorly may anastomose with the supra-orbital canal or end blindly ; posteriorly anastomoses either with the lateral canal only or with both the supra-orbital and lateral canals (usually). THE CRANIAL NEKVES AND LATERAL SENSE ORGANS OF FISHES. 121 (3) A hyomandibular or operculo-mandibular canal. Developed in connection witJi the lower jaw, and may be represented by one or more canals. Innervated by the external mandibular lateral line nerve. Ventrally may anastomose with its fellow or fellows of the opposite side, but usually ends blindly ; dorsally usually anastomoses with the infra- orbital canal, may in some cases do so with the lateral canal, may anastomose with botli these canals, or may remain independent of any. Figure of Gadus virens. half natural size, showing the sensory canals and their innervation — the latter imlicatefl by the different kinds of shading. D.F. anterior extremity of dorsal fin; Z,'. supra-temporal portion of lateralis canal ; L\ lateral portion of lateralis canal ; iV\ anterior and posterior narial openings ; 0. opercular fold ; PcT.F. pectoral fin; Plv.F. pelvic fin; P.O. pit orgms (innervated by the lateral line nerves) ; S.O.C. supra- orbital commissure. (1) Supra-orbital canal, S.O. Innervated by superficial ophthalmic lateral line nerve — eroBS-hatched. (2) Infra-orbital canal, F.O. Innervated by buccal + otic lateral line nerves — dotted. (S) Hyomandibular or Operculo-mandibular canal, H. Innervated by external mandibular lateral line nerve — longitudinal shading. (4) Lateralis canal, D and L'^. Innervated by lateralis lateral line nerve — oblique shading. (4) A lateral canal situated at the side of the body. Innervated by the lateralis lateral line nerve. Anteriorly anastomoses usually with the infra-orbital canal only, but may also do so with the supra-orbital canal ; posteriorly ends blindly at the tail. The canal known as the supra-temporal or occipital commissure ( = Commissural canal of Ewart) has somewhat variable relations, and is perhaps not homologous in all fishes. 122 ME. F. J. COLE ON THE STRUCTURE AND MORPHOLOGY OF Strictly speaking, and arguing from its condition in those fishes in which its innervation has heen carefully ascertained, it should form a dorsal connection hetween the anterior extremities of the lateral canals. It may, however, interpolate itself, as it does in Cliimcera and other recent fishes, and perhaps also in Pterichthys, hetween the two orhital canals, so that these do not anastomose posteriorly, hut hoth arise from the lateral canal. But, however this may he, no canal can he considered a supra-temporal canal which is not innervated by the lateralis nerve, and which therefore is not a portion of the latei-al canal. I hence propose to describe the sensory canals of Oadus morrJiim in the following order : — (1) Supra-orbital canal ; (2) Infra-orbital canal ; (3) Hyomandibular canal ; (4) Lateralis canal. All these canals will be described from before backwards. D. Sensory Canals of Gabus morrbua. (PI. 21, fig. 1.) (1) Supra-orbital Canal. The supra-orbital canal commences blindly at the anterior extremity of the nasal bone, and therefore of course at the extremity of the snout. This blind extremity appears very late in the ontogeny of the system, and is not seen in sections of young Gadids — the canal here being directly continuous Avith the first dermal tubule (I propose to designate as " dermal tubules " those small tubes by which at intervals the main sensory canals communicate with the surface *). The supra-orbital canal on the nasal bone pursues a slightly sigmoid course obliquely inwards, upwards, and backwards, and measures whilst on this bone about 40 mm. long. The nasal is deeply grooved to receive the base of the canal, and its inner edge is jmrtly turned outwards to form a rough tube — imperfect, however, dorsally and externally. Anteriorly the bony nasal furrow is imperfect ventro- lateially, in order that the fii'st dermal tubule may enter the supra-orbital canal (see PI. 21). The second dermal tubule enters the supra-orbital canal 25 mm. from the posterior extremity of the nasal, and lies in a backwardly-directed oblique furrow situated on the lateral wing of the nasal. The canal now leaves the nasal and for a distance of about 5 mm. has no bony floor, but only a ligamentous one. An internal wall and roof are, however, partly supplied by a forward projection of the frontal. The third dermal tubule enters the canal at this region, but somewliat nearer the frontal than the nasal. The remainder of the supra-orbital canal on the frontal is more conveniently described from behind forwards. In front of its anastomosis with the infra-orbital it courses inwards and forwards for a distance of about 20 mm. It then enters by a lateral foramen a canal bored in the substance of the frontal. Tiie median portion of this canal forms a perfect tube, but posteriori}^ the roof is imperfect owing to the jiresence of a * The dermal tuLules and sense organs of the lateral canals are enumerated in order from before backwards in full knowledge of Allis's important proof of the relation between the sense organs and the tulnilcs. Any scientific enumeration of these structures in the ease of the Cod is, however, impossible until the development of both has been investigated, and any attempt therefore to do so here would be purely arbitrary and misleading. THE CEANIAL NEEVES AND LATERAL SENSE ORGANS OE FISHES. 123 long, laterally compressed, oval median fontanelle, which measured 12 mm. long in the present sjDCcimen. The posterior extremity of this fontanelle represents the nearest approach which the supra-orbital canal makes to the mid-dorsal line, the two supra- orbital canals being only separated at this region by a cylindrical bony basal projection or septum 3 mm. wide. It is here that the two supra-orbital canals communicate Ijy a well-marked transverse commissure, very obvious in sections of young Gadids, but apt to be overlooked in dissections of adults. In front of this region the canal at once liegins to course obliquely outwards, and continues to do so as far as its anterior extremity. The bony septum above I'eferred to becomes wide at the anterior extremity of the fontanelle, where it is nearly 5 mm. wide. The median closed portion of the tube is a little over 10 mm. long, in front of which the next 14 mm. of the canal are imperfect externally, becoming more and more so anteriorly. The course of the supra-orbital canal from the posterior extremity of the median fontanelle up to tlie anterior edge of the frontal has been slightly obliquely outwards and forwards. The fourth mipaired dermal tubule, which is the longest of those opening into the supra-orbital canal, enters it at the short transverse commissure. The surface pore in adults is, in all the specimens that I have examined, slightly to the left of the median dorsal line, whilst the tubule itself opens precisely at the centre of the commissure, although I have seen it also somewhat to the left. In sections of young Gadids, however, both the surface j)ore and the internal opening are exactly median — another instance of the slight disturbance of the symmetry of the lateral line organs which lakes place in the adult. It may be mentioned here that this is the only unpaired dermal tubule in the Cod. (2) Iiifra-orhital Canal. This will be described under the headings of the various bones which support it. First Suh-orhital or Lachrymal Bone *. — The infra-orbital canal arises blindly a little in front of the lachrymal near the mid-dorsal line at the extremity of the snout; The blind extremity exists from the very beginning, and is obvious both in sections of young forms and in dissections of the adult. The anterior extremity of the lachrymal is grooved to receive this part of the infra-orbital canal. The lachrymal itself is roughly triangular, with the apex directed forwards. Its length is about 60 mm., and breadth at the base over 35 mm. The base may be divided into two concave portions, separated by a median * The synonymy of the bones of the Teleostean skull is, it is hardly necessary for me to point out, a somewhat obscure subject. I have endeavoiu-ed as fur as possible to steer a middle course, and therefore hope that no objection can be taken to any of my terms. It is, perhaps, necessary to exjilain that Allis (6) and other authors consider the first sub-orbital and lachrymal to he two separate bones. No reasonable objection, however, can be brought against the view advocated by McMurrich (136), that the lachrymal is a modified sub-orbital plate, and I therefore describe it as the first sub-orbital. "We may note in this connection that the Teleostean lachrj-mal has been erroneously termed by some authors the pre-frontal. The correct homologue in the Teleosts of the Ganoid pre-froutal is, however, the lateral ethmoid (=the parethmoid). For the rest, sphenotic and pterotic are undoubtedly synonyms of post-frontal and squamosal respectively (but see p. 132). The confusion in the nomenclature of the bones of the fish's skull has arisen partly by the misuse of some terms and partly by the unnecessary introduction of others. SECOND SERIES. — ZOOLOGY, VOL. VII. 18 124 ME. F. J. COLE ON THE STRUCTUEE AND MOEPHOLOGY OF prominence, the ventral portion bearing a facet for the articulation of the second sub-orbital. The sensory canal passes along the middle of the bone, marking it into two almost equal halves. The dorsal border of the canal is slightly sigmoid, and anteriorly it aj)proaches the border of the upper jaw, whilst the direction of the whole canal is distinctly downward. A thin bony lamina (see below) projects downwards from above and forms a roof and an external border for this portion of the infra-orbital canal, leaving it open, however, ventrally. About 35 mm. from the posterior extremity of the lachrymal this bony lamina is seen to be l)roken, thus forming a sort of notch and leaving this portion of the infra-orbital canal wdthout any external boundary, and with only a portion of its roof. The first four dermal tubules open into the iDfra-orbital canal on the lachrymal, the first at the anterior extremity of the bone, and the tbird at the notch mentioned above. The articulation of the lachrymal with the second sub-orbital is somewhat peculiar and may here be descrilied. As mentioned above, and as is the case with the other sub-orbitals, the lachrymal sends down externally a bony lamina the function of which is to protect the external wall of the infra-orbital canal. With the posterior extremity of tbe imperfect tube thus formed the dorsal portion of the anterior convex extremity of the second sub-orbital becomes fitted or wedged in. Ventral to the external bony lamina the second sub-orbital simply overlaps the lachrymal, and the large facet thus formed at the posterior extremity of the latter bone measures 6 mm. antero-posteriorly. Second Sub-orbital. — The direction of this ossicle is backw^ards and downwards. Its length dorsally at the sensory canal is about 20 mm., and its greatest width 22 mm. Tlie anterior articular surface is convex, and is connected with the lachrymal in the way above described. Posteriorly there is an oblique articulation with the third sub-orlntal. The second sub-orbital is a large semi-cartilaginous plate, the dorsal border of which is turned over so as to form a stout support for this portion of the infra-orbital canal. As, however, the dorsal border does not fuse on to the large body of the plate ventral to the sensory canal, the tube is imperfect ventrally, and is, in fact, only comj)leted by the ligamentous sheath of the ossicle. The dermal tubules 5 and 6 open into the infra- orbital canal on this ossicle — the former at the anterior extremity, where it overlaps tlie lachrymal, and the latter near its posterior border. Third St(l)-orbital. — This ossicle continues the downward direction of its predecessor, but is somewhat more horizontal. The infra-orbital canal is still passing backwards and downwards. The length of the third snb-orl)ital is 15 mm., and its width 16 mm. The anterior and posterior articular surfaces are oblique and roughly parallel, but the anterior is much the longer. An imperfect tube for the lodgment of the sensory canal is formed in the same way as in the second sub-orbital. The seventh dermal tubule opens into the canal at the posterior extremity of the third sub-orbital. Fourth Sub-orbital. — The infra-orbital canal now begins to take an upward turn, and, in consequence, the direction of this ossicle is upwards and backwards. It is further midw^ay betw ecu the vertical and horizontal sub-orbitals, and measures 14 mm. in length and 7 mm. in width. The sub-orbitals thus narrow from before backwards. The sensory canal tube is formed similarly to those on the preceding sub-orbitals, but it must be noted THE CRANIAL NEEVES AND LATERAL SENSE ORGANS OF FISHES. 125 that as the sub-orbitals become narrower, so the edge turned over to form the tube approximates more and more to the ventral or posterior edge, the tendency thus being for the whole of the sub-orbital ossicle to be nsed up in the formation of the protective sensory tube. The eighth dermal tubule opens at the dorsal or posterior extremity of the ossicle, the latter tapering off to receive it. Fifth Sub-orbital or First Post-orbited. — The iufra-orbital canal has by this time taken a sharp upward curve, so that the fifth sub-orbital is not far removed from the perpendicular. Its length is 17 mm., and the anterior edge almost approximates to the posterior in the formation of the tube lor the iufra-orbital canal. The ninth dermal tubule is received by the infra-orbital canal betweca the fifth and sixth sub-orbitals. This tubule is not rej)resented in Gaclus virens. Sixth Sub-orbital or Second Fost-orbital. — This, the last and most posterior sub-orbital, articulates ventrally with the fifth sub-orbital, and dorsally with the postero-lateral region of the frontal, and also slightly with the post-frontal or sphenotic. Its direction is as near as possible perpendicular, and the portion of the infra-orbital canal that it lodges is of course directly continuous with that on the frontal. The length of the canal on the sixth sub-orbital is 14 mm., and it is protected in a precisely similar manner as on the fifth sub-orbital — the anterior and posterior edges not quite approximating. Frontal. — The infra-orbital canal is now on the frontal. Situated on the upper surface of the postero-external angle of the frontal is a triangular fossa with the apex directed inwards and forwards. The base of the triangle forms a portion of the posterior edge of the frontal. The infra-orbital canal comes first along the outer side of the ti-iangle (a distance of about 14 mm.), and arriving at the apex first of all anastomoses with the supra-orbital canal, and then turns sharply backwards, edging the inner side of the triangle (about 15 mm.). The canal, therefore, at this region turns first inwards and forwards, and then outwards and backwards. The two sides of the triangular fossa on the frontal are depressed so as to form slight furrows for the reception of the base of the infra-orbital canal. Further, the inner side of the triangle is raised up and arches outwards, thus forming a roof for this portion of the infra-orljital canal. As, however, elsewhere, the tube is imperfect externally. Post-frontal or Sphenotic. — The sphenotic now comes to the surface and lodges the next 5 mm. of the infra-orbital canal, the latter stUl pursuing the downward and backward direction of the portion immediately preceding it on the frontal. The sphenotic, however, only supports the floor of the infra-orbital canal, an inner wall and roof being here formed partly by a backward projection of the frontal and partly by a forward prolongation of the pterotic. Externally the sensory canal is only protected by ligament. The tenth dermal tubule enters the canal at about the middle of the sphenotic, and in front of the canal in that bone transmitting the otic branch of the outer buccal nerve. Squamosal or Pterotic. — The iufra-orbital canal, leaving the sj)henotic and still passing backwards and somewliat downwards, enters on the pterotic, which supports it for the last 30 mm. of its length, until, between the pterotic and the succeeding supra-temporal, the infra-orbital canal anastomoses with the lateral or body canal. A more or less perfect 18* 126 ME. F. J. COLE ON THE STEUCTUEE AND MOEPHOLOGY OF tube is formed on the pterotic by a latero-external process being sent down. Posteriorly this process only forms an imperfect tiibe, but anteriorly it usually almost completely encloses the canal, yet does not fuse ventrally with the floor of the ossicle. The tube is thus imperfect for the whole of its length, but less so anteriorly than posteriorly. The last or eleventh dermal tubule of the infra-orbital canal opens into it between the pterotic and the adjoining supra-temporal, but the greater part of the aperture is on the pterotic. (3) Hi/omandihula r Gmial. The hyomandibular canal arises on the dentary at its anterior extremity near the symphysis, its length, whilst on this bone, being about 80 mm. The direction of the canal depends, naturally, on the position of the lower jaw. When the jaw is closed the canal has a stioug upward incline. Its relations with the articular portion of the canal are, of course, fixed, and do not vary with the gape of the jaw. The hyomandibular canal fits into a deep groove on the ventral surface of the dentary which is continued almost to the anterior symphysis. This groove is widely open ventrally, and gradually tapers towards its anterior extremity. The first five dermal tubules open into the hyomandibular canal whilst on the dentary, the sixth opening at the junction of the dentary and articular, but rather on to the dentary. Leaving the dentary the hyomandibular canal passes on to the articular, and then begins the upward curve which carries the canal on to the pre-operculum. Strong dorsal and ventral ridges form a stout protection for the canal, which is only exposed externally. The hyomandibular canal is 15 mm. long on the articular. Owing to an inward shelving on the part of the articular and pre-operculum, the next 10 mm. of the canal have no bony support, but are protected only by a ligamentous sheath. This, of course, is developed in connection with the movement of the lower jaw. There is also a strong upward curve on the part of the canal to meet the succeeding portion on the pre-opei'culum. The seventh dermal tubule opens on to the canal at about the middle of this ligamentous section. Having reached the pre-operculum, the hyomandibular canal lies at first somewhat horizontally. It then takes a graceful curve upwards, so that the posterior portion of the canal is practically vertical. Whilst on the pre-operculum the hyomandibular canal has a length of about 80 mm. A thin bony lamella projects downwards anteriorly and backwards posteriorly, forming an external lateral wall for the canal, but, however, leaving it open ventrally in front and posteriorly behind. Commencing at about 11 mm. from the posterior extremity of the canal, a ridge of bone 8 mm. long is seen to be thrown up, which forms a posterior wall for the canal. At this region the anterior bony lamina forming the external lateral wall of the canal tapers down, otherwise the hyomandibular canal would in this region be completely enclosed in bone. The tube then is here also incomplete externally, but is inore perfect than it is at any other part of the pre-operculum. The hyomandibular canal ends blindly 10 mm. below the dorsal extremity of the pre-operculum. The dermal tubules eight to twelve open on to the THE CRANIAL NERVES AND LATERAL SENSE ORGANS OF FISHES. 127 canal at the pre-opei'culum, the eleventh helow and the twelfth ahove the ridge of bone just mentioned. It is thus seen that the hyomandibuiai- canal is perfectly independent of any of the others, and does not anastomose either with the infra-orbital or lateral canals. (4) Lateralis Canal. First or Supra-temporal Portion. As with the infra-orl)ital canal, it will bo most convenient to describe the lateral canal under the headings of its various supjiorting ossicles. First Supra-temporal or Extra Scapula.— The supra-temporal canal commences blindly in both G. morj^hua and G. vij'ens at the anterior extremity of the first supra-temporal ossicle. Its direction is markedly inwards and forwards. The greatest length of the ossicle is 17 mm., and its greatest breadth 6 mm. It is free anteriorly, that is, it has no articulation with any otlier bone, but its extremity is attached by a ligament to the dorsal spiny process of the parietal overhanging the foramen transmitting the so-called " cutaneous branch of the Vth." * Posteriorly the ossicle articulates by means of an oblique convexity with tlie second supra-temporal. The first supra-temporal consists of an internal plate, having its dorsal edge turned over somewhat considerably to protect the supra-temporal canal. Anteriorly the portion turned over tapers down to admit of the entrance into the canal of the first or proximal dermal tubule. Second Supra-temporal or Extra-scapula. — Supports the distal or basal portion of the supra-temporal canal, and consists of an internal plate, the two edges of which have more or less become opposed to form a tube. Posteriorly this apposition is complete, so that the tube is here perfect, thougii the two edges in question have not fused, such not being a characteristic of the Cod. Anteriorly, however, the tube, as in most of the other lateral line ossicles, is imperfect, but this is to enable the second dermal tubule to eater the canal. The upper posterior border of the ossicle bears a semicircular notch, which fits into the upper or epiotic limb of the post-temporal. The internal face of the second supra-temporal also rests on this limb of the post-temporal. The greatest length of the second supra-temporal is 17 mm., and its greatest width about 11 mm. Anteriorly it presents a concave border for articulating with the first supra-temjooral, whilst posteriorly it articulates with two ossicles : a, by a long, slightly convex, ventral facet with the third supra-temporal ; b, by a backwaixlly directed posterior facet (slightly concave) with the fourth supi"a-temporal. Posteriorly the supra-temporal portion of the lateralis canal takes a dowuAvard curve and becomes confluent with the latei-al or body portion. Second or Lateral Portion. The first 20 mm. of this section of the lateral cjinal are supported partly by the third supra-temporal, biit also partly by the ventral jiortion of the second (see PI. 21. fig. 1). This fact must be taken into consideration when reading the description of the former ossicle. * =the R. lateralis acoessorius (sec Section N, p. 16G). 128 MB. F. J. COLE ON THE STEUCTURE AND MORPHOLOGY OF Third Supra-temporal or Extra-scapula. — The third supra-temporal is a somewhat long bone supporting tlie anterior extremity of the lateral canal as it passes downwards and backwards from the pterotic. Its greatest length is 22 mm., and its greatest width 9 mm. It consists of a plate slightly convex intei*nally, the dorsal edge of which is turned over, but not very markedly, to form an imperfect tube. Its ventral border is quite straight, whilst its dorsal border is somewhat irregular. Anteriorly it articulates by means of an obliqvie and well-marked concavity with the pterotic, the dorsal border of the third supra-temporal being prolonged forwards so as to fit into a corresponding shelving on the pterotic. Ventrally the latter presents a large triangular facet 10 mm. in length from base to apex, on w^hich the ventral portion of the anterior extremity of the third supra-temporal rests and articulates. This articulation is very compact and strong. Posteriorly the articulation is with two ossicles : a, by a long dorsal and slightly concave facet with the second supra-temporal ; b, by an oblique concavo-convex facet with the fourth supra-temporal. The third dermal tubule opened into the lateral canal near the posterior extremity of the third supra-temporal, and slightly posterior to the entry of the supra-temporal canal into the lateral. This tubule was present on the left side of the fish, but not on the right. Fourth Supra-temporal or Extra-scapula. — This consists of a flat internal plate 11 mm. across at its widest part, and with a maximum length of 16 mm. Its dorsal edge is turned over so as to form an external jirocess protecting the sensory canal. Anteriorly it has a close ligamentous articulation with two ossicles : «, by a convexo-concave ventral facet with the third supra-temporal ; h, by a slightly convex dorsal facet with the second supra-temporal. The latter facet is internal to the first and passes back at an angle to it. Posteriorly the fourth supra-temporal rests on and overlaps the angle formed by the two forks of the post-temporal. The fourth dermal tubule opens into the lateral canal near the centre of the fourth supra-temporal. Post-temporal or Supra-scapula *. — The angle formed by the union of the two limbs of the post-temporal now comes to the surface of the skull and supports the next 9 mm. of the lateral canal. It is here that the lateral canal commences to approacli the surface of the body and to take tip that position in which it is to be found immediately under the skin. The fifth dermal tubule opens just at the posterior extremity of the post- temporal, and almost between that bono and the first lateral line ossicle. Behind the post-temporal the lateral canal loses the support of the bones of the skull and pectoral girdle, and is only protected at intervals by the lateral line ossicles, which themselves diminish posteriorly as the canal approaches the skin and decreases in size. At about the middle of the fish the lateral canal lies immediately under the skin, and is situated at the side of the body, its position being indicated by the modified scales lying immediately over the canal. In the specimen from which the above description has been written, it was about 55 mm. ventral to the anterior extremity of the dorsal fin. It is supported behind the post-temporal by a series of imperfect bony cylinders, the * The suin'a-elavicle of some authors (t'. (/. McMurrich, 136, and Gegeubaur). It is imiiossible, however, to homologiso the boues usually termed post-temporal and supra-clavicle. THE CEANIAL NERVES AND LATERAL SENSE ORGANS OF FLSHES. 129 so-called "lateral line ossicles." The anterior extremity of each "lateral line ossicle" is 2:)lacecl just posterior to the internal opening of a dermal tu1:)ule, so that if we regard the lateral canal as composed of segments defined by the internal openings of the dermal tubules and the intermediate occurrence of sense organs, then the lateral line ossicles will lie at the anterior extremities of these segments. Each cylinder is imperfect externally, the space being obliterated by the periosteal membrane. The length of one taken from about the region of the pectoi-al fin was 7 mm. Anteriorly they increase in size, the first, just posterior to the post-temporal, being an irregular imperfect tube 11 mm. long. E. Gadus MoitiiHUA a:nd G. vibess compared. The differences between these two species are few but decisive. The sense and pit organs of both are the same {roughly as regards the pit-organs) both in number and position. The only variations I could find were in the dermal tubules. These will be at once noticed if a comparison be made between figs. 1 and 2, PLs. 21 & 23 (representing G. morrlma and G. virens respectively), and are as follows :— 1. The ninth dermal tubule on the infra-orbital canal of G. morrlma is absent in G. virens. This, as far as I am aAvare, is a perfectly constant difference. I have never failed to find it in G. morrlma, nor have I ever seen it in G. virens. 2. In G. virens the hyomandibular canal opens anteriorly on to the surface, but does not do so in G. morrlma. This, I believe, is another constant variation. 3. In G. morrlma the third dermal tubule of the lateral canal is of variable occurrence. In the fish from which fig. 1 was drawn it was present on the left side, but absent on the right. It will be noted that the nares also were abnormal in that fish on the left side, being fused there, but quite normal on the right side. In G. virens the fourth dermal tubule of the lateral canal of G. morrlma is, I have found, invariably absent, nor have I ever seen any fusion on the part of the anterior and posterior nares. E. The Sense Organs on the Lateral line Canals. (Plates 22 & 23.) These have been carefully worked out, as well as the number and position of the dermal tubules verified, from sei'ial sections of G. virens. This is the only method which ensures perfect accuracy in this connection. The sense-organs themselves not being obvious to the naked eye, the dissector has to rely on finding the nerve twigs supplying them. If he should chance to overlook one or more of these twigs, as often happens, the corresponding sense organs are, for the time at least, also overlooked, and this may easily nullify the results of the investigation. I accordingly paid a visit to St. Andrews and collected a large number of young " Green Cod" {G. virens), many of which have since been cut into serial sections. The heads of the living animals were cut off behind the pectoral fin and fixed in Hermann's platinum acetic osmic mixture. The sections (10 ticks thick, Cambridge rocking microtome) were stained on the slide with Mann's methyl-blue — eosin* — a most admirable method, giving perfect pictures even of the brain * Jouru. Anat. & Phys. vol. xsis. (N. S. vol. ix.), p. 100 (1894). 130 ME. F. J. COLE ON THE STEUCTUKE AND MOEPHOLOGY OF and nerve tracts. The sense organs were then systematically mapped out from the sections, so that it is hoped any possible chance of error has been removed. Suxtra-orhitul Canal. — There are five sense organs on the supra-orbital canal as against four dermal tubules. As a rule, one sense organ is placed midway between the internal openings of two dermal tubules, so that there is usually a rough correspondence between the number of sense organs and that of the tubules. The first sense organ on the supra- orbital canal is placed between the openings of the first and second dermal tubules, but nearer the second ; the second between the second and third tulmles, but nearer the third (just over the anterior extremity of the nasal sac) ; the third and fourth anterior and slightly posterior to the supra-orbital commissure respectively ; and the fifth near the posterior extremity of the canal, and in the young forms * exactly dorsal to the optic chiasma. Infra-oi'hital Canal. — This canal has eleven sense organs and ten dermal tubules. The first two sense organs are situated between the first and second tubules, and just under the first sense organ of the supra-orbital canal. There is a slight overlapping on the part of the first two sense organs, the j^osterior extremity of the first being above the anterior extremity of the second, and the two together extending practically from the opening of the first dermal tubule to the second. The next six sense organs, the third to the eighth, lie between their respective tubules as shown in the figure ; the ninth and the tenth, however, are not separated by a tubule, and this seems to imply that the ninth dermal tubule has degenerated. This tubule is indeed present in G. morrhua, and in such a position as would separate the ninth and tenth sense organs of G. vwens. It is therefore possible that in the latter species the tubule has existed and disappeared. The eleventh sense organ lies between the ninth and tenth tubules, but somewhat nearer the former. Hyomandihular Canal. — Twelve sense organs and thii'teen dermal tubviles were found in connection with this canal. The first sense organ is slightly anterior to the opening of the second tubule, and the second has the same relations with the succeeding tubule. The remaining ten sense organs are situated between the openings of the tubules as indicated in Plate 22 the last being placed on the narrow portion of the canal at its posterior extremity. Lateral Canal, — The supra- temporal segment of the lateral canal has two sense organs and two dermal tubules, the first sense organ being placed slightly posterior to the opening of the first tubule, and the second lying between the ojjening of the second tubule and the junction of the supra-temporal portion of the canal with the lateral portion. The third sense-organ is found just anterior to the junction of the two canals mentioned above, whilst the fourth is placed betweeu the third and fourth dermal tubules. The remainder of the sense organs of the lateral canal are situated as usual between the openings of the dermal tubules. The condition of the lateral canal in young Cod behind the operculum will be described and discussed elsewhere. * See note on p. 150. THE CEANIAL NERVES AND LATEliAL SENSE ORGANS OF FISHES. 131 G. Other Sense Organs belonging to the Lateral line System. (Plate 21.) Belonging undoubtedly to the lateral line system, since they are innervated by the same nerves, are the pit oi'gans or sensoi'y follicles, described by Merkel, Fritsch, AUis, and Ewart & Mitchell *. These in the Cod are somewhat laiimerous, and do not occur in the well-defined series that they do in the Elasmobranchs and Ganoids. Their structure has already been fully discussed by the authors above mentioned, so that only a few words are necessary as to their distribution. As will be seen by reference to tlie figure of G. morrhua, the pit organs cannot be separated into groups, being too irregularly scattered. It is equally obvious, however, that most of them occur in connection with the sensory canals, and this is most marked in the case of the hyomandibular canal. In spite of their apparent irregularity and number, they are very constant both in number and position, as will be ascertained by a careful examination of an adult G. morrlina and a comparison with the figure. The pit organs are usually conspicuous on account of their pigmented lips, and even where this is slight or absent a contrast in colour usually betrays their existence. As is well known, the pit oi'gans are excavated papillae formed by an invagination of the skin, and lodging a sense-organ innervated by a twig from one of the lateral line nerves. My figure (PL 21) of these organs is the result of a minute examination of the heads of several large specimens of G. morrhua. Contrasting with the numerous pit organs is tin; complete absence of any traces of Lorenzini's amiiulUe and the associated ampuUary canals. Tliis fact is duly commented on elsewhere. H. PiElation of the Sensory Canals to the Skull. That the lateral line system has no definite relations with the primitive skull is quite evident even on a priori grounds. An epiblastic structvu'e, necessarily connected with the skin, and actually found on the surface in Chim metameric significance. On p. 523, referring to Van Wij he's views above, he remarks : — "The arrangement of the sense organs and nerves of the lateral system, the regular occurrence of primary tubes between consecutive dermal bones of the head, as well as between consecutive scales of the lateral line, and the singular correspondence between the infra-orbital and opercular canals is further evidence in this same direction " [*. e. of metamerism]. I am not inclined, however, to attach much weight to these considerations. In the first place, there are too many exceptions to the regular occur- rence of the primary pores between consecutive dermal bones for this consideration to have much value *, and in the second place, although in the case of the Cod there is a * Allis says (p. 537) : — " Although some of the prima'')' tubes issue through the bone, one always issues between every two cousecutive bones along each liuo." 21* 150 MK. F. J. COLE ON THE STRUCTURE AND MORPHOLOGY OF certain correspondence between the infra -orbital and byomandibular lateral canals, it does not occur, as far as I am aware, in any other iish. And surely, in any case, the corresjiondeuce has absolutely no significance. The byomandibular canal is situated on a region morphologicaUi/ posterioi' to the infra-orbital canal, and cannot belong to the same segments or segment of the head. If the correspondence had been l)etween the supra- and infra-orbital canals (and these do not correspond either in Ainia, Gadus, or any other fish that I know of), it might have had some significance, although, I am inclined to think, very little. An interesting addition to the developmental litei-ature was made in 1891 by Wilson {225), who was the first, I believe, to discover tliat the lateral line organs and auditory organ may arise from a common sensory anlage. This seems to be oj)posed to the metameric view, bnt Wilson himself believes it to be a physiological adaptation and to have no phylogenetic significance. He favours Beard's views with regard to the head, but does not think the lateral organs of the body were primitively segmental. Cunningham (1890, 55) draws attention to the fact (p. 75) that in the Sole, " corresponding to eacli scale of the lateral line, there is a pore in the skin which leads into the dermal tube of the lateral line." On p. 81, however, he says : — " There is not a sense organ to every scale of the lateral line ; in the middle of the body there is a sense organ on every third scale : that is to say, there are two scales bearing no sense organs between two scales which bear them. The position of the sense organ in relation to the scale on which it is situated is always the same." Fig. 6, pi. xiv., shows this somewhat anomalous condition. Without impugning the accuracy of Cunningham's statement, I may point out that it is a remarkable exception to the almost universal law that there should always be at least one sense organ between the openings of two dermal tubules *. Mitrophanow (1890, 141) confirms Wilson's statement re the common anlage of the lateral line and auditory systems, and in 1892 (142) disagrees with Wilson and states his belief that the lateral organs are not metameric. In a full paper published in 1893 (143) Mitrophanow repeats his former statements, Imt adds that tlie common anlage gives rise to the auditory organ, the lateral line organs, and Beard's branchial sense organs. He considers this suflB.cient ground for maintaining that the lateral organs were not primi- tively metameric, and says this conclusion is based on the study of all the Ichthyopsid types he has investigated. Houssay (1891, 103), in an interesting review of Mitrophanow's second work above, is inclined to accept Eisig's invertebrate origin of the lateral organs, and combats Mitrophanow's statement that they are not metameric. Ayers (1892, 7) endeavours to show that the auditory organ is not supplied by a discrete cranial nerve but by the branches of two cranial nerves, /. c. the facial and the glossopharyngeal ( = tbe vagus: be considers the lateralis lateral line nerve to be a branch of the IXth). He thus favours tlie metameric view of the lateral organs. On p. 314 he says : — " As Froriep has shown, the ectodermal thickenings which Beard described as giving rise to the lateral Hne organs have in fact another fate. The genuine lateral line organs escaped Beai'd's observation, and in consequence Beard's conclusions * Sense organs may exir.t withoul dermal tubules (cp. Ewart & Mitchell, p. 10(»), but. mt viva verm. THE CRANIAL NERVES AND LATERAL .SENSE ORGANS OF FISHES. 151 as to the laomology of the vertebrate avicUtory organ are incorrect "*. Although Avers' views as to the innervation of the auditory organ cannot any longer be maintained, it is possible that his latter statement is to some extent sufficiently near the truth to require a reopening of the whole question of the " branchial " or " epiln-anchial " sense organs. In his L(cmargi(s j^aper (1892, 68), Evrart supports the view of the innervation of the lateral sense organs taken by Triant and AUis. He does not consider that Beai-d's scheme applies to the adult Elasmobranch, nor does he consider the lateral canal as composed of ontogenetic metanieric sense organs (p. 79). In his paper with Mitchell we find (69, p. 100) : — " But v/hile the sense organs and tubules have a nietamerlo arrcmgement in the trunk, there is no relation between the sense organs and segments In the head region; and, as already pointed out, some portions of the cranial canals, though possessing numerous sense organs, have no tiilnxles connecting them with the exterior. In all the cranial canals, l^cjth dorsal and ventral, there are far more sense or"-ans than segments ; e. g. in the supra-orbital canal there are nearly ninety sense organs, and in the infra-orbital there are over ninety." Pollard (1.892, 161), referring to the relation between the lateral canals and the dermal bones, in which connection we should remember Allis's statements above, says (p. 527) : — " In Clarias it is by no means a rule that pores should open at sutures " ; but on p. 539 we find the absolutely contradictory statement that " as the dermal bones are much reduced in Auchcnaspis the close relationship of pores to sutures, which exists in Clarias, is not seen." Willey (189 i<, 223) makes a somewhat remarkable statement. He says (pp. 41-i5) : " It seems certain that at first the sense organs of the lateral line must have been iuncrv^ated by spinal nerves. This follows both from ti priori considerations and also from the condition in Amphioxus, where the ectoderm of the metapleural folds is innervated by the Rami ciitanel ventrales of the dorsal spinal nerves. Under these circumstances it is necessary to supjiose with Eisig that the lateral line nerve {Ramus lateralis vagi) arose as a collector." I am unacquainted with any evidence in support of this statement. Eisig's views on the subject I have already dealt Avith, but I may mention that his explanation of the lateralis nerve has never been accepted by vertebrate morphologists. Bashford Dean (1895, 58) evidently considers the metamerism of the lateral canal a secondary modification. He says (p. 51), after considering the probable phylogeny of the system : — " The sensory cells are no longer scattered evenly along the floor of the canal ; they now occur in metameral masses supplied with a distinct nerve branch, located in the region immediately below the external tubules." Again on p. 52 he remarks : — " The original significance of the lateral line system as yet remains undetermined. As far as can be judged from its development, it appears intimately, if not genetically, related to the sense organs of the head and gill region of the ancestral fish : in resj^onse to special aquatic needs, it may thence have extended fu.rther and further backward along the median line of the trunk, and in its later differentiation acquired its metameral characters." Locy (1895, 130) agrees ndth other observers that the branchial sense organs of Beard and Froriep are not the lateral sense organs, but may perhaps corre.spond * When Ayers wrote this he must have forgotten what lie had previously written on n. 213 I 152 ME. r. J. COLE ON THE STBITCTURE AND MOEPHOLOGT OF to the segmental sensory papillre of Annelids. On p. 577 et scq. he confirms for Squalns the observations of Wilson on Set^raims and Mitrophanow on Acanthias and otlier Elasmobranclis as to the common anlage of the lateral sense organs and auditory oro-an, and states : — " MitroiAanow departs from the nsual point of view that the organs of the lateral line are metaraeric, and in that particular, I think, I should be inclined to follow him." Strong (1895, 204) is more emphatic on the point. He says (p. 197) : " It is also evident that the lateral line system has no specially segmental character, and that it cannot properly be used in the manner in which it has been attempted to use it, as a general guide in determining the segmentation of the head." Miss Julia Piatt (1896, 158) argues from the opposite point of view*. She states (pp. 502-503) : — " Since Mitrophanow claims as the result of his study that the segmenla- tion of the lateral-line system is entirely secondary, I shall be interested to discovei; when I again have my Acanthias material with me whether traces of primitive segmentation so evident in Necturus cannot also there be found, for it is dilficult to believe that the great similarity which exists in the position and direction of the main lines of sense organs in Necturus and Acai/thlas should not be the result of a similar coiirse of development." Miss Piatt also describes four of the sense organs of the infra-orbital line as being partly innervated by the ophthalmicus profundus, and concludes (pj). 530- 531) : — " I do not, for this reason, include the trigeminus among the lateral line nerves, bvit should nevertheless hesitate to say that the ' trigeminus proper does not participate in the innervation of the lateral line system.' " With regard to the ultimate fate of Beard's branchial sense organs, Minot (1897, 140) says (pp.706 & 709) : — "We have further to emphasize those traces which have been discovered of long series of sense organs, of Avhich the nose, eye, and ear are probably derivatives, in the ancestors of the vertebrates, although in all known vertebrates most of these series have become rudimentary or lost. The serial sense organs I designate under the comprehensive name of ganglionic sense organs. There are probably two, and only two, series along each side of tlie body : one series, the ujiper, corresponds to the lateral line of comparative anatomy, the other to the epibranchial line." After pointing out the diiferences between the two series, he concludes : — " The sense organ above the gill cleft [i. e. branchial or epibranchial sense organ], though differentiated, is a larval structure only, and disappears in the adult." Finally Wilson and Mattocks (1897, 226) confirm for Salmo the discovery already made by the former author in Serranus of a common anlage of the lateral and auditory sense organs, and state further that the portion in front of the auditory saucer gives rise by bifurcation to the supra- and infra-orbital canals, whilst the portion posterior to the saucer grows backwards and forms the lateral canal. We thus see that the metamerism of the lateral line nerves and their associated sense organs has for a long time been a problem, the solution of which has been attended with considei-able difficulties. The older anatomists considered the nerves to be branches of the trigeminus, facialis, and vagus, until it was shown by Priant in 1879 that the * But cp. ])p. 4!l2 and 50], which go to show thai Beaid"s branchial sense organs do not belong to the lateral line system. THE CEANIAL ]\'ERVES AND LATERAL SENSE ORGANS OF FISHES. 153 nerves usually considered to be branches of the trigemiuus really belonged to the facial. To Friaut, therefore, belongs the credit of being tlie first autlior to arrive (approximately) at the truth with regard to the innervation of the lateral sense orgaas. His work was confirmed by Marshall and Spencer in 1881 and by Mlis in 1889, and it was thus established that the superficial oplithalmic and buccal lateral Hue nerves belonged to the facial, rather than to the trigeminal, nerve. The work of Beard and Froriep in 1885, imjiortant as it is, has added considerably to the confusion wiiich exists on the subject. There can be no doubt that if the branchial or epibranchial sense organs ultimately develop into the lateral sense organs of the adult, then the latter must have been (primitively) segmental structui'es. But we have seen there is good evidence against the iilentity of the two series of sense organs, and even before this evidence transpired it is significant that the innervation of the adult lateral sense- organs could not be harmonised with the supposed embryonic conditions (cp. especially Allis and Ewart). And added to this the fact that the lateral organs and the auditory organ have been found to develop from a common sensory anlage, which discovery has been extended to several forms and has been found to apply to Elasmobraiiclis as well as to specialised Teleosts, we have an opposing view of the development of the lateral organs which has the advantage of being easily reconcilable with the facts of adult structure. This leads us to the latter view of the question. In his 1889 paper, Allis, independently following Friant, and working on the lines laid down by Marshall and Spencer, described the lateral line nerves as branches of the facial, glossopharyngeal, and vagus nerves. In this he was largely followed by Ewart (1892) and Pinkus (1894)), until it was considered completely established that the lateral nerves were undoubtedly branches of the facial and vagus, and perhaps also of the glosso- pliaryngeus. In the meantime Mayser (1882), Pollard (1892), and Strong (1895) had been working out a totally different hypothesis, the tendency of which was in the first place to regard all the lateral line nerves as morphologically branches of one trunk, and in the second to associate that trunk with the auditory irerve and system. This view was developed by me in my Cliimcera paper, in which I endeavoured to show that the lateral line system was an independent system of sense organs innervated by a distinctive and independent series of nerves of characteristic size of fibre, and further that the whole was the morphological equivalent of the auditory organ, with which it should therefore be associated. I further stated the belief that the branch of the glosso- pharyngeus innervating sense organs of the lateral line would be found on investigation to be a branch of one of the lateral line nerves, and finally, pending further investigation, provisionally associated the lateral line nerves with the facialis. These views have been entirely confirmed by recent investigations, such as tliose in 1897 by Kingsbury, Herrick, and Allis, so that the j)i"esent position, from the point of view of the comparative anatomist, and setting aside for the moment purely developmental evidence, may bo summarised as follows : — (1) The lateral line system is an indejoendent series of sense organs, differing histo- logically from any other cutaneous sensory system, such as the terminal buds described by Merkel and other authors, and not innervated by the same nerves. 154 MK. 1', J. COLE ON THE STEUCTURE AND MOEPHOLOGT OF (2) That this system is innervated by a series of lateral Hne nerves which are un- doubtedly independent of miy of the recognised cranial nerves, and are in fact mori^ho- logically branches of one trunk. (3) That the lateral line brancli of the glossopharyngeal belongs to the lateralis lateral line nerve. (4) That the lateral sense-organs and nerves are precisely comparable to the auditory organ, with which they should be associated — the lateral and auditory nerves arising from a common centre in the brain which is further peculiar to this system of nerves. It is also possible that the lateral sense organs, together with the auditory organ, in most forms arise from a common sensory anlage from the skin. We have seen that the developmental evidence in favour of the metamerism of the ]ateral line system is too conflicting to be made the basis of any discussion, but there is no room for doubt that the sense organs of the body canal are often metameric in the adult, as has been described above. It is hence necessary to enquire into the nature of this regular occiu-rence of the body sense organs. I have previously pointed out that the lateral sense organs primitively belong to the head, and have only secondarily extended on to the trunk. This can be proved both by the development and adult structure of the lateralis canal. Even in young adults of Gadiis the body canal posterior to the shoulder girdle is still very imperfectly formed, and in fact is somewhat difficult to detect in sections. In all known cases of development the body canal grows from before backwards, and its most posterior part is the last to be formed of any of tlie sensory canals. When the sensory canals are quite perfect in the head and have fully reached their adult condition, the posterior section of the body canal is still nothing more than a mere rudiment or anlage. It is hence impossible to regard any condition of the body canal as representing a primitive condition, since it is itself only a secondary structure. The innervation of this canal, however, settles the point. The nerve whicli supplies it is in all fishes the lateralis lateral line nerve, which, as it is not a collector, and is absolutely independent of the metameric nerves, must be held to disprove the essential metamerism of the body canal, and indeed shows that such a condition of this canal is purely superficial. Now it is obvious that the independent character of the lateral line system, which is such a characteristic feature of the post-embryonic conaition of existing fishes and amphibians, may be either primitive or acquired. It may be, either that the lateral nerves were originally branches of the various cranial nerves, and that the present apparent relations with the latter are the vestiges of that connection, or that the lateral nerves were primitively independent and have already commenced to fuse with, and to form an essential part of, the true cranial nerves. It seems to me that the latter conclusion, although it was advocated by myself, is being hastily and injudiciously adopted, and that further embryological investigations Avill have to be concluded before we liave sufficient data to arrive at an accurate appreciation' of the subject. The question may be viewed from the thi'ee standpoints of Embryology, Comparative Anatomy, and PalEContology. With regard to the former, in spite of the numerous and bulky memoirs that have been written on the subject by such naturalists as Dohrn, Goette, Balfoui', THE CRANIAL NEEVES AND LATERAL SENSE ORGANS OF FISHES. 155 Van Wijhe, and Beard, much has yet to be done. From an embvyological standpoint, the first alternative above is still a possible, if not a probable, explanation of the facts, and imless the evidence from Embryology is to be rejected altogether, which is out of the question, it is necessary to wait until further investigation from this standpoint definitely vipholds one view or the other. Whatever result is finally arrived at, it should be one in which both embryologists and anatomists concur, and it seems to me that this result will most probably be one that is adverse to the metamerism of the lateral organs. Comparative anatomy helps us but little. It is true that in CUmcBra, which shows us the innervation of the lateral sense organs in its simplest known condition, all the lateral nerves, except the external mandibular, arise separately from the brain, and are not in any way connected with the true cranial nerves. It is also true that as we go higher in the vertebrate scale we get every possible gradation between this comparatively independent condition and the very complex one found in the highly specialised recent Teleosteans — where the mingling between the lateral and cranial nerves is at its maximum. It is now a question of whether it is permissible to argue from this that the primitive condition must have been that in which the nerves were concentrated and independent, and therefore not metameric, since the cartilaginous fishes are simpler than the Teleostean forms, and fossil Ichthyology teUs us that they have departed less from the primitive type. Such a contention of course accepts as granted that the soft parts have advanced pari passu with the specialisation of the skeleton. It is evidence of perhaps little or doubtful value, but it may, I think, be used to confirm a view already resting on a more solid basis of fact. From Palaeontology we learn still less, and it is most valuable when it is able to throw light on the ancestry of a type that we have other reasons to believe is primitive. It shows us also that the lateral line system is an extremely archaic structure, and further that as regards the geograpliy of the sensory canals it has remained in a largely unmodified condition for untold ages. What bearing this has on the innervation of the system it would be hazardous to conjecture. Briefly, the question at issue between embryologists and anatomists is whether the lateral line system is metameric or not. The most valuable evidence that has been advanced from the latter point of view is that in which the lateral nerves have been microscopically traced both to their central origin and peripheral distribution. Considering this evidence, and pending agreement among embryologists, I take up the provisional position (which, however, I believe to be a very strong one) that it is not. I shall therefore in the present communication describe the lateral nerves as purely independent structures having no connection, other than purely secondary, with the true cranial nerves. The following description of the innervation of the lateral canals of Gadus is based partly on dissections and partly on an examination of serial transverse sections of young adult Gadus virens (="(?. carbouarms"). SECOND SERIES. — ZOOLOGY, VOL. VII. 22 156 ME. F. J. COLE ON THE STRUCTURE AND MORl'HOLOaY OF M. Innervation of the Sensory Canals in Gadus. (Plate 22.) (1) Supni-orhital Canal. Superficial Ophthalmic Trunk. — This consists of two portions, which are {a) the supei-ticial ophthalmic of the trigeminus, and {b) the superficial ophthalmic lateral line nerve. The latter arises from the posterior dorsal edge of the proximal portion of the lateral line ganglion, and at its origin lies between the ventral edge of the cerebellum and the anterior extremity of the auditory capsule. Owing to a certain amount of over- lapping on the part of the facial and auditory nerves, the origin of the lateral oplithalraic nerve is somewhat difficult to determine, owing to the close apposition of its root to the anterior edge of the auditory ganglion. It would in fact be difficult to say whether some auditoi'y fibres did not mingle with the lateral ophthalmic. There are a few ganglion cells at its base, but otherwise no cells whatever in the course of the nerve, and the few above belong, strictly speaking, to the lateral ganglion. The lateral ophthalmic now passes horizontally forwards over the trigemino-facial ganglion and soon commences to pass upwards. As, however, it leaves the ganglion, it receives a conspicuous root containing a few ganglion cells from the anterior dorsal edge of the ganglion, and this is the root of the trigeminal ophthalmic, since it is from the direct continuation of this portion of the trunk that the general sensory fibres arise. The root of the trigeminal ophthalmic is at first very flat, and closely opposed to the ventral surface of the lateral ophthalmic root, but no mingling of the fibres coidd be detected either at this region or any other. As the trunk passes gradually upwards towards the posterior poi'tion of the infra- orbital canal, the trigeminal portion becomes narrower and is almost completely separated from the lateral portion by a small blood sinus or vessel. Fui'ther forwards, however, the two nerves become approximated by passing under this vessel. By this time the nerve has entered what appears to be a rudimentary eye muscle canal, and lies opposite the ventral edge of the alisphenoid over the posterior portion of the eye, and (iu the young forms but not in the adult *) somewhat close to the brain opposite the dorsal border of the optic thalami and the origin of the optic nerves. The twig to the fifth sense organ of the supra-orbital line is given off from the dorsal or lateral portion of the trunk at this region. It perforates the wall of the eye muscle canal, passes upwards and forwards to the outside of the alisphenoid, finally piercing the frontal to reach its distribution. Continuing its course, the ophthalmic trunk passes obUquely upwards and assumes a position between the dorsal border of the eye and the supra-orbital canal. Soon after supplying the twig to sense organ 5, the lateral ophthalmic gives off a somewhat large branch [S.O.^) which, after coursing parallel with the main trunk for a short distance, * The rc4atioiis are by uo meaus the same. Fur example, a trausverse section may pass through both the optic lobes and a portion of the eye iu the young forms, whereas in the adult no such section could pass through a»y part of the lirain and eye at the same time. THE CRANIAL NEKVES AND LATERAL SENSE ORGANS OE FISHES. 157 divides into three twigs whilst still lying near the brain in the eye muscle canal and opjjosite the edge of the inner ventral process of the frontal. The inner of the three twigs passes inwards and upwards, perforates the frontal, and lies between the Irontal and the sensory canal. In this position it passes forwards for some little distance and innervates sense organ 4 of the supra-orbital line. The remaining two twigs doubtless innei'vate pit organs in this region. Opposite the point where the twig to sense organ 4 perforates the frontal, the ophthalmic trunk divides into two conspicuous nerves, a larger dorsal lateral portion, and a small ventral trigeminal portion — as is shown by the nerves arising from them, the lateral-line twigs arising from the dorsal nerve and the general cutaneous fibres from the ventral iierve. As the trunk passes forwards the division becomes more mai'ked, until two rounded nerves are distinctly differentiated. The lateral portion soon gives off a twig which passes uj)wards and enters the frontal slightly anterior to the supra-oi'bital commissure. It passes obliquely through the substance of the bone inwards and forwards, perforates it dorsally, lying between it and the sujira-orbital canal, and finally innervates sense organ 3 of the supra-orbital line. Anterior to the region where this twig enters the frontal, the two divisions of the ophthalmic trunk begin to approximate and once more continue their course together. At this region too, a small blind sac is seen in the sections to open into the supra-orbital canal external and opposite to tlie supra-orbital commissui-e and partly opposite the fourth sense-organ. It contains no sense organs and no lateral line twigs could be traced to it, and seems to me to correspond precisely to a much larger but otlierwise similar structure in the same position described by Hyrtl in Lota. It therefore possibly represents a degenerate or modified dermal tubule. In the region of sense organ 3 the supra-orbital trunk shows a tendency to split up again, and can clearly be resolved into its two constituents with the higher power of the microscope. This tendency, however, is soon lost and the nerves become inseparable as before. It is here, moreover, tliat the ojilithalmic trunk leaves what I take to be the eye muscle canal, and becomes for the first time perfectly round in transverse section. It immediately enters the frontal at the junction of the frontal and pre- frontal (= lateral ethmoid or parethmoid), and courses obliquely inwards and forwards in the spongy sub- stance of tlie frontal. After leaving the latter bone it enters a large space bounded above and internally by the frontal and below by the pre-frontal. Anterior to this sjmce the trunk begins to pass ujiwards and inwards towards the Ligamentous portion of the supra- orbital canal, which it accompanies, and an examination of it with the high power at once reveals the distinctiveness of its lateral and trigeminal portions. These again separate, and the ventral portion or trigeminal ophthalmic gives off a large nerve, which could not be satisfactorily traced in the sections, bvit seemed to be a cutaneous sensory nerve *. In front the two ophthalmics again more or less approximate. Before leaving the space in the frontal mentioned above, the lateral ophthalmic gives off dorsally a twig {S.O.^) which passes inwards and forwards, curves upwards round the * The fibres of this nerve undoubted!}' came from the trigeminal ophthalmic. 22* 158 MK. F. J. COLE ON THE STRUCTURE AND MORPHOLOGY OF anterior edge of the frontal, between the latter and the ligamentous portion of the supra- orbital canal near the mid-dorsal line, to the inner side of the sensory canal, and was traced almost on to the skin and then lost. It most probably supplied the pit organs in that region. Coincident with the ligamentous portion of the supra-orbital canal, the ophthalmic trunk separates into its two components, quite obvious with the low power, the lateral constituent being dorsal and the larger of the two. The latter soon afterwards gives off a twig which perforates the posterior edge of the nasal, lies between the latter bone and the supra-orbital canal, and Anally passes obliquely inwards and forwards to supply sense organ 2. • The ophthalmic trunk has now completely split up, both components at first lying close together at the posterior extremity of the nasal, ventral to the latter bone and just below the supra-orbital canal. Opposite the second sense organ the two components commence to run alongside again, but do not mingle. Anteriorly the lateral, together with the greater portion of the trigeminal, ophthalmic pass inwards and upwards along the inner face of the nasal, and whilst there a trigeminal twig is given off (S.O.^) which, passing outwards and upwards along the outside face of the nasal, is distributed to the skin in that region. Whilst the ophthalmic nerves are passing along the inner face of the nasal the relation between them is for the first time changed, the smaller trigeminal curling round the larger lateral ophthalmic and assuming the dorsal position. Opposite the anterior extremity of sense organ 2 the two components finally separate out and do not for the remainder of their course come into contact again *. The smaller dorsal trigeminal ophthalmic (S.O.^) passes rapidly upwards, dividing into two, is distributed to the skin of the dorsal region of the snout, and, although coursing with the lateral line nerve described in the footnote, is not in any way connected with it. The large ventral lateral ophthalmic, on the other hand, jiasses forwards, perforates the nasal opposite the opening of the second dermal tubide, and terminates in sense organ 1. (2) Infra-orbital Canal. Buccal Trunk. The buccal trunk, leaving the ophthalmic and buccal ganglion, passes downwards and outAvards through the " trigeminal " portion of the trigcmino-facial complex, and whilst passing through this ganglion divides into the nerves (1 and 2 below) which issue from the ganglion at different levels. (1) Outer buccal branch. — Issues from the trigeminal portion of the V-VIIth ganglion dorsal and posterior to the second branch. It is smaller than the latter, and gives off', * Just at. this region a long nerve from the upper ramus of the inner buccal lateral line nerve passed straight up, crossed the tvro ophthalmic nerves internally, but was not connected with either, curved outwards and round the dorsal edge of the nasal, and finallv passed downwards to innervate a pit organ opposite the anterior extremity ot sense organ 1. This curious nerve is not represented in other fishes, and probably consists of lateral superficial ophthalmic fibres following a buccal course. THE CRANIAL ISTERVES AND LATERAL SENSE ORGANS OF FLSHES. 159 immediately after leaving the ganglion, tlie otic nei've, being thereafter continued on to the orbit as the posterior or outer buccal nerve of the buccal trunk. Both these nerves are described below. Coursing for a time alongside the outer buccal branch and issuing from the "trigeminal" ganglion immediately ventral to it, in fact from the maxillo-mandibular portion of the ganglionic complex, is one of the nerves described by Strong as " accessory branches of the trigeminus," and also mentioned by Allis (6, p. 605, &c.). The nerve now described undoubtedly corresponds to AUis's nerve " c " from the maxillo-man- dibular trunk, and is distributed mostly to the tissues behind and below the eye. Also issuing from the ganglion at the same level, but posterior to it, and passing at first outwards and then downwards, is a twig which seems to tally with AUis's branch of the inferior maxillary " r.lap.do.''' (6, p. 610), since I traced some of its fibres to the dilator- operculi muscle. (2) Inyier buccal branch. — Leaves the ganglionic complex ventral and anterior to the first nerve. As it issues from the ganglion and passes into the orbit it is joined and accompanied by the maxillo-mandibular trunk, but lies dorsal and somewhat posterior to it. There is, however, never any connection between them, and sections show the inner buccal branch to be perfectly distinct, from its origin at the ojihthalmic and buccal ganglion onwards. The maxillo-mandibular trunk soon begins to separate into maxillary and mandibular nerves respectively, and as it does so the inner buccal branch passes upwards so as to lie dorsal to the superior maxillary, when the separation of the maxillo- mandibular trunk is complete. (I may here remark that the relative positions of the inner buccal and trigeminal nerves in the orbit are subject to some variation. The con- dition just described is shown in my second figure. In the sections, however, and in. other dissections the inner buccal and its two rami were situated below the supeiior maxillai'V throughout the whole orl^it.) In front the inner buccal crosses over the superior maxillary so as to lie ventral to it. As it crosses the orbit, and just over the division of the maxillo-mandibular trunk into its two derivatives, the inner buccal branch divides into a smaller inner and dorsal portion, and a larger outer and veutral one. These will be described respectively as the upper and lower rami of the inner buccal branch of the buccal trunk. Outer buccal branch. Otic nerve. — This nerve arises from the outer buccal branch immediately the latter emerges through the pro-otic notch, and then passes straight upwards but slightly backwards, accompanied by an arterial twig, over the pro-otic and the external face of the post-frontal, until it reaches a backwardly-directed dorso-ventral canal bored in the substance of the post-frontal. Having passed through this, it emerges on the dorsal surface of the skull between the post-frontal and the anterior overlapping portion of the squamosal. It then passes straight backwards over the squamosal, lying dorsal to the latter bone and ventral to the infra-orbital canal, until it reaches the last or 11th sense organ of the infra-orbital line which it suj)plies. Outer buccal branch. Outer buccal nerve. — On leaving the ganglion, this nerve passes forwards and downwards. It soon gives off a twig which, travelling upwards and for- wards, pierces the sixth suborbital and supplies sense organ 10 of the infra-orbital canal. 160 MR. ¥. J. COLE ON THE STKUCTURE AND MOKPJIOLOGY OF Beyond this the outer bviccal takes a sharp turn downwards, but before doing so crosses internally the vertical portion of the infra-orbital canal almost at right angles, and takes up a position in front of and above it. Shortly after giving off the twig to the 10th sense organ, and as the outer buccal enters the posterior portion of the orbit, another twig is given off, which, passing downwards and forwards, perforates the fifth sub-orbital and siipplies sense organ 9 of the line. In front of this the outer buccal is seen to sej)arate into three nerves. One of these is the main portion of Allis's nerve " c " (described above), which, having hitherto accompanied the outer buccal, now passes upwards to its distri- bution, whilst of the two ventral twigs the outermost passes outwards and downwards, jjerforates the fourth sub-orbital, and innervates sense organ 8 of the infra-orbital canal. The i-emaining one, in front of the twng to sense organ 8, takes a ventral curve, and after giving off a branch ventrally which was traced to a pit organ situated ventral to the canal, pierced tlie third sub-orl)ital and supplied sense organ 7 of the line. Tpnei^ hnccal branch. — This, the larger of the two divisions of the buccal trvmk, passes downiwards and forwards, as previously described, across tlie orbit, and consists of two l^arts. The lowei- and larger part soon gives off a twig which, after a long course down- Avards and forwards (lying internal to the sclerotic), perforates the second sub-orbital and supj)lies sense organ 6 of the infra-orbital canal. Almost opposite this sense organ another twig is separated off from the lower ramus of the inner buccal, and in this region the first and second sub-orbitals are seen in the sections to begin to overlajj. The twig above pursues a course somewhat parallel to its predecessor, obliquely perforates the lachrvnial, and supplies sense organ 5 of the line. At about the region of this sense orean the two rami of the inner buccal begin to separate, tlie ventral one passing down- Avards and approaching the infra orbital canal, whilst the dorsal ramus passes upwards. Tlie ventral ramus opposite sense organ 5 gives off another twig, and, joassing rai^idly downwards, lies near the inner face of the lachrymal, slightly dorsal to the upper border of the infra-orbital canal. The twig above perforates the lachrymal at the doi'sal border of the canal, and innervates sense oi-gan 4. After supplying this sense organ the ventral ramus comes to lie immediately internal to the inner face of the lachrymal, and exactly opposite the infra-orbital canal. It then gives off a branch which suj)plies the pit organs ventral to the canal, and immediately afterwards the whole of the nerve perforates the lachrymal and lies between that bone and the upper and lower borders of the canal. Tliere are thus three perforations in the lachrymal for twigs of the ventral ramus of the inner buccal to sense organs of tJie infra-orbital line. Before reaching the third sense organ the ventral ramus divides into two — the ventral twig sujiplies sense organ 3, whilst the dorsal passes forwards and supplies sense organ 2 only. To return to the upjier ramus of the inner buccal. After separating from the ventral ramus it passes upwards and ventral to the nasal sack, and is joined by a branch of the superior maxillary division of the trigeminus. The two nerves do not, however, mix. The ujiper ramus then gives off its first branches, which terminate in the numerous pit organs in the region of the snout between the supra- and infra-orbital lines. It now passes downwards and becomes opposed to the inner face of the lachrymal at a spot immediately dorsal to the place where the ventral ramus passes through its most THE CRANIAL NERVES AND LATERAL SENSE ORGANS OF ELSHES. 161 anterior foramen in the lachrymal. It then courses forwards parallel and dorsal to the ventral ramus, and, after giving off dorsally more twigs to the pit organs of the snout, passes slightly downwards, perforates the lachrymal opposite the dorsal horder of the infra-orbital canal and slightly posterior to sense organ 2, and JiuaUu supplies sense oi'gan 1. (3) Kijomandibiihif or Operculo-mandibiilar Canal. Hyomandibular Trunk. This large nerve (^), just at its origin from the trigemi no-facial ganglion, is divided liy the orbital vein into two portions. The anterior ventral portion consists of two parts representing the morphological palatine and pre- and post-spiracular nerves or the "facial proper," whilst the posterior dorsal portion constitutes the "lateral line root of the facial." These constituents run alongside one another external to the vein, but are soon quite separable by the microscoj^e into an anterior bundle (iZ"'), composed of the pre-spiracular and main trunk of the external mandibular branch, and a posterior bundle (if"), which is mostly formed by the post-spiracular nerve, but also includes a branch of the external mandibular. In dissections it was noticed that the pre-spiracular separated off before the hyomandibular trunk divided into the two preceding bundles. External mandibular nerve. — The posterior branch above of this nerve arises as already described, and then passes straiglit outwards, downwards, and backwards. It is closely applied to the postero-external border of //\ and receives some lateral line fibres from it ventral to the orbital vein. As the entire hyomandibular trunk passes down- wards, on approaching the hyomandibular bone the two bundles E} and E- become separated so as to form two distinct nerves. The posterior, or smaller of the two (i?^), which is composed of a branch of the external mandibular -f the hyoide;in nerve, or post- spiracular division of the facial proper — the latter being the larger bundle, passes through the oblique backwardly-directed facial canal in the hyomandibular bone, whilst the larger bundle (_£f') traverses a foramen in the hyomandibular larger, just ventral, and somewhat anterior and internal to it. Distally the apertures are somewhat widely separated owing to the backward direction of the facial canal. The hyoidean nerve + the posterior branch of the external mandibular, on emergino- from the facial canal, immediately take a sharp turn downwards and slightly backwards, coursing near and parallel to the posterior edge of the anterior downward process of the hyomandibular. On leaving the facial canal the two parts of the nerve are seen to be separated, in the sections the external mandibular being situated externally to the facial proper component. The two nerves pass downwards and slightly backwards together, and pass under the preoperculum at the base of the deep notch at its upper extremity. The external mandibular fibres {W) leave the hyoideus near the ventral extremity of the facial canal, and, passing outwards, do^vnwards, and backwards, soon collect into two bundles. The upper passes downwards, perforates the preoperculum, and supplies the 12th sense-organ of the hyomandibular line. The lower, after a some- what long course downwards, obliquely tunnels the preoperculum some little distance 162 ME. F. J. COLE ON THE STKUCTUEE AND MORPHOLOGY OF above sense organ 11 of the line, which it supplies. The motor hyoideus {S*), the post- spiracular nerve, passes downwards and becomes related to the anterior face of the hyoid arch. With the distribution of this nerve we are not concerned. The anterior or larger division of the hyomandibular trunk (M^) consists, as already described, of two parts : (a) the main trunk of the external mandibular lateral line nerve {S^) ; and (b) the " internal mandibular" of Allis (i?^). I shall show later on that Allis was wrong in his interpretation of this nerve, and that it really corresponds to the modified pre-spiracular nerve, or chorda tympani, of the cartilaginous fishes, and as such cannot be the internal mandibular, which is a post-spiracular nerve. The internal mandibular or post-spiracular nerve is really the hyoideus. Sections show that this trunk consists of two parts, an anterior chorda tympani and a posterior external man- dibular. The chorda soon separates off, passes downwards, and becomes related to the posterior face of the lower jaw. On leaving the hyomandibular bone the external man- dibidar immediately passes straight downwards over the anterior downward process of the hyomandibular, and accompanied by a blood-vessel which in spirit dissections looks astonishingly like a nerve *. Shortly after emerging from the hyomandibular foramen a long nerve is given off from the external mandibular (H"') which passes forwards through the adductor mandibulse to innervate some pit organs in the neighbourhood of the infra- orbital canal. It then courses downwards and forwards to supply certain of the pit organs lying ventral to the hyomandibular canal. Before passing downwards, however, it sends down a long fine nerve which innervates the jiosterior set of pit organs in connection with the hyomandibular canal. These nerves are mentioned by Allis (1897, p. 632). They also midoubtedly correspond to Allis's nerves mef.vl. and Tiief.mdl. innervating the vertical cheek and mandibular lines of pit organs. Soon after leaving the canal in the hyomandibular a twig is given off, which, passing downwards and backwards over the external face of the anterior process of the hyo- mandibular, perforates the preoperculum ventral and somewhat internal to its anterior upward process, and just dorsal to the point of the backward process of the quadrate. It supplies sense organ 10 of the hyomandibular canal. After giving off the above twig the external mandibular nerve passes downwards over the outer process of the hyo- mandibular and begins to take a marked turn forwards. Leaving the hyomandibular it passes over the symplectic, and whilst on this bone gives off the twig to sense orgau 9 of the line. This twig jiasses forwards over the symplectic, perforates the preoperculum internal to the facet for the quadrate and at about the anterior third of the facet, and so reaches its sense organ. The external mandibular now begins to assume a horizontal position, and arriving at about the middle of the posterior edge of the symplectic, where the body of the quadrate overlaps it, passes almost horizontally, but still somewhat downwards, in a rough canal * Several of the more important nerves in the Cod are aeeompanied by blood-vessels, which often appear, as in the case of the superficial ophthalmic trunk, to arise directly from the nerve. (They really perforate it.) In specimens that have been along time in spirit these vessels are remarkably like nerves, ami Jiiive been taken for such hi/ several natiiralists ! THE CKANIAL NERVES AND LATERAL SENSE OEGANS OF PISHES, 163 formed by the superposition of the quadrate on the symplectic. Whilst passing between these two bones it gives off the twig to tlie 8th sense organ, which, passing downwards on the internal face of the quadrate, perforates the preoperculum near the anterior extremity of its dorsal border (close to the quadrate facet) and thus reaches its destination. Having passed between the symplectic and the quadrate, the external mandibular courses downwards over the inner face of the latter internal to its head, and then by a downward curve reaches the inner face of the articular, where it lies in a deep horizontal groove situated at a level somewhat dorsal to this section of the hyomandibular canal and opposite the head of the articular. Whilst on the head of the articular the twig to sense organ 7 of the hyomandibular line is given off. This arises somewhat anterior to the sense organ, passes backwards, enters the articular oi3posite and ventral to the anterior edge of the facet for the quadrate, and passes transversely through it to its sense oi-gan. At this point a branch of the inferior maxillary division of the trigeminus comes down and runs alongside the external mandibular, lying immediately dorsal to it; but although they may be very closely opposed there is no real connection, the ventral branches of trigeminal branch curving round the inner and outer surfaces of the lateral line nerve to reach their destination. The external mandibular nerve now commences to pass slightly downwards, and first lies internal and somewhat above Meckel's cartilage, and afterwards, in front of the 6th sense organ, is situated directly internal to it. The branch to sense organ 6 is separated off somewhat in front of the twig to the 7th sense organ. It courses forwards and out- wards, obliquely perforates the dentary, and after running alongside the sensory canal for some little distance reaches and supplies its sense organ. The mandibularis externus is now opposed to the inner face of Meckel's cartilage, and there gives off the twig to sense organ 5. This proceeds downwards and forwards, and, obliquely piercing the dentary, reaches its respective sense organ. At about this region a nerve is seen to pass down- wards under Meckel's cartilage from its outer surface and then turn upwards to become opposed to, but never connected with, the outer face of the external mandibular. It is at first situated between the latter nerve and the cartilage, but finally comes to lie on the top of the lateral line nerve at the place where the twig to the fourth sense organ perforates the dentary. This nerve is undoubtedly Allis's branch " r.ghi.'' * springs from the inferior maxillary division of the Vtli, and constitutes a part of the innervation of the geniohyoid muscle. Of the t^\ags to sense organs 3 and 4^ nothing need be said further than that they passed through separate perforations in the dentary, and that each gave off a very small bundle of fibres which were traced ventral to the hyomandibular canal and then lost. They doubtless innervated the pit organs of this region. The external mandibular is still situated internal to Meckel's cartilage. Anterior to sense organ 4 the lateral line nerve commences to pass downwards so as to occupy a position ventral and somewhat internal to Meckel's cartilage. Here also the anterior termination of the main trunk of the inferior maxillary division of the trigeminus * Cp. AUis, 6, p. 639. SECOND SEKIES. — ZOOLOGY, VOL. VII. 23 164 ME. F. J. COLE ON THE STRUCTURE AND MORPHOLOGY OF is seen lying internal to and above Meckel's cartilage *. In front the external man- dibular passes still further downwards, finally taking up a position ventral to Meckel's cartilage and slightly dorsal to the hyomandibular canal. It is still accompanied by trigeminal fibres, which, however, pass to their destination anterior to the termination of the external mandibular. Tlie twigs to sense organs 1 and 2 pass tiirough the same foramen at the anterior extremity of the dentary. (•i) Lateralis Canal. The lateralis lateral line nerve j^ursues a course somew^hat analogous to that of the buccal. It gives off a supra-temporal branch, which corresponds to the otic, and like it supplies one of the termini of its canal, and then splits into a dorsal and a A'entral ramus which may be said to coincide with the two divisions of the buccal trunk. The dorsal ramus innervates the anterior or abdominal lialf of the lateralis canal with its associated pit organs, and the ventral ramus does the same for the posterior or caudal moiety of the canal. Both rami are of course typical lateral line nerves, and as such are composed of special sensory and not of typical somatic sensory fibres. They may be united by one or more commissures, and together form wdiat are referred to in the text-books as the " cutaneous branches of the vagus." Lateralis Trunk. This large nerve emerges from the medulla at a high level immediately behind the lateral line ganglion of the trigemiuo-facial complex, in front of the root of the glossopharyngeus and some distance in front of and dorsal to the root of the A^agus. The latter nerve arises by a double root from the medulla distinctly ventral and posterior to the lateralis. On leaving the medulla the lateralis passes downwards and backwards, and soon swells into the large lateralis ganglion. It still continues the same course and passes external to the root of the glossopharyngeal, to which, as it passes, it gives a very small twig which accompanies the pre-branchial division of the IXth. The lateralis ultimately joins the vagus, Avhich it accompanies, lying external to it. Hitherto there have been ganglion cells along almost the w hole of the length of the lateralis root, but as soon as it reaches the vagus the ganglion cells largely disappear, and only a few are to be seen. Coincident wdth the disajjpearance of cells in the lateralis root is their appearance in the root of the vagus, in which they multiply as the nerve recedes from the brain. Both nerves pass throiigh the same canal in the exoccipital, the vagus occupying the anterior and internal position, and the lateralis stiU possessing a few ganglion cells. No mixing of the two nerves was observed at any part. Outside the exoccij)ital foramen the vagus root immediately enters the large compound (?) vagus ganglion, and there is also a I'urther collection of ganglion cells on the lateralis in connection with, and at the base of, its supra-temporal branch. * AUis (6, p. GIO) saj".s that this part of the ini'orior maxillary accompanies the " r. buccalis facialis," but this is obviously a slip for " r. maudibularis eiternus facialis." THE CRANIAL NERVES AND LATERAL SENSE ORGANS OF FISHES. 165 Supra-temporal branch. — This branch arises from a clump of ganglion cells (which also extend into its proxixnal portion) from the lateralis immediately the latter emerges from the skull. It curves round the auditory capsule, and passes at first in a vertical plane outwards and then upwards. The first bi-anch to be given off is a small posterior one. It arises from the ventro-external portion of the supra-temporal a little distal to the ganglion cells, and passes straight backwards, pierces supra-temporal 4, and supplies the fourth sense organ of the lateral or body canal. It also gives off two branches : first, a somewhat obvious one, which passed forwards and somewhat downwards, bifurcated, and was lost on the skin (this was separated off near the origin of the posterior branch of the supra-temporal) ; second, a bundle of fibres which was given off as the parent nerve approached the fourth sense organ, and which passed backwards and was also lost on the skin. Neither of these nerves ended in the vicinity of pit organs. The larger part of the supra-temporal now takes an upward and forward turn, but before reaciiing the lateral canal divides into two — a dorsal and an anterior division. The former {U) passes upwards and crosses the lateral canal internally and posterior to the entry of the supra-temporal canal, ^vhilst the latter (i*) passes at first straight forwards just below the lateral canal. JO'' soon divides into two branches. The lower of these perforates the second supra-temporal and innervates sense organ 2 of the supra-temporal canal; the upper courses upwards and forwards, perforates the first sujira-temjioral, and supplies sense organ 1 of the line. L* passes forwards and slightly upwards, passes in between the opposed horizontal edges of supra-temporals 2 and 3, and thus reaches sense organ 3 of tlie lateral canal. Just as it passes in between these ossicles it gives off a twig which jiassed forwards and was lost on the skin above the lateral canal. Here also there was an absence of any pit organs. It seems to me probable that this branch, as well as the similar ones first described, arc composed of fibres of the vagus which have accompanied the supra-temporal nerve and its branches *. The main trunk of the lateralis, after giving off the supra-temporal branch, passes downwards, outwards, and backwards, and in the sections is seen to be situated between the dorsal border of the pseudobranch and the kidney. As it passes backwards it divides to form two conspicuous nerves, one of wliich is distinctly smaller than the other. The smaller division passes upwards and outwards over the top of the pseudobranch, and then straight backwards, but before doing the latter gives off' externally a largish nerve, which itself separated off a very small twig and then passed straight downwards in the inner wall of the branchial chamber. I was unable to trace this branch. The parent nerve, ho\\'cver, then coursed outwards and upwards, pierced tlie first lateral line ossicle, and innervated sense organ 5 of the lateral or body canal. After giving rise to the nerve above described the smaller or dorsal ramus of the lateralis trunk continues to pass upwards and backwards approaching the lateralis canal, and as it does so it sends off a branch dorsally which passes first upwards and then forwards for some little distance and markedly dorsal to tlie lateral canal. It was finally lost on the skin. Almost immediately afterwards another small branch was given oft', * Cp. particularly description of posterior root of accessory lateral nerve, p, 176. 23* 166 MR. F. J. COLE ON THE STEUCTUEE AND MORPHOLOGY OF which coursed backwards for an interval with the parent nerve, and then passed upwards, crossed the lateral canal internally, and innervated a pit organ just dorsal to the canal. By the time the parent nerve has reached the surface tubule 5 it has come to lie immediately ventral to the lateral canal. Beyond this tubule the nerve gave off a dorsal branch which pa«;sed upwards, pierced the second lateral line ossicle, and innervated sense organ 6 of the line and its two closely associated pit organs. Posterior to the origin of this twig a smaller one was given off which passed straight up and supplied a pit organ dorsal to the body canal. Behind this again the branch arises which perforates the third lateral line ossicle and is connected with sense-organ 7 of the line and its two neiglil)0uring pit organs. The ventral or larger ramus of the lateralis trunk coursed backwards and steadily downwards until reaching the pectoral fin, from which point it passed almost straight backwards somewhat dorsal to the fin and following the line separating the dorsal from the ventral musculatiu'e. I was unable, however, to detect any branches from its anterior moiety. N. The Structure and Morphology of the Ramus lateralis accessorius. Synonymy. (1) Nervus lateralis accessorius (Weber, 1820). (2) Nerf pterigo-dorsal (Desmoulins & Magendie, 1825). (3) Ramus lateralis trigemini (Weber, 1827, and most other authors). (4) Ramus quartus s. lateralis nervi trigemini (Bonsdorff, 1846). (5) Dorsals Schedelholilenast (Stannius, 1849. Cp. also Hoffmann & Pollard). (6) Ramus recurrens trigemini et facialis (Stannius, 18i9). (7) Ramus recurrens facialis (Siluroids. Stannius, 1849. Cp. Pollard). (8) Ramus cutaneus quinti (T. J. Parker, 1884). (9) Nervus Weberi (Goronowitsch, 1897). (10) Nervus accessorius Weberi (iu part. Haller, 1897). Under the name of " ramus lateralis accessorius " (the reasons for retaining which are given below) I propose to describe those curious and interesting nerves familiar to all students who have dissected a Codfish, and which I at first thought were modified lateral line nerves. A minute investigation of the roots, however, and, further, an examination of the literature, convinced me not only that this conchision was absolutely erroneous, but also pointed to the correct view, as I think, of theii- morphological value. This small investigation also led to an examination of the essential meaning of the so-called lateralis nerve of the Lamprey, the results of which, however, are given in the succeeding section. That the so-called " cutaneus quinti " of Gadus corresponded to the " recurrent facial " nerve described by Stannius and Pollard, was pointed out independently by Allis and the writer — the former m his second Amia paper (6, p. 628), whilst a note to that effect was inserted in the second English edition of Wiedersheim's "Comparative Anatomy" (221, p. 187) at my suggestion. As, however, we were both anticipated by earlier THE CRANIAL NERVES AND LATERAL SENSE ORGANS OF FISHES. 167 writers, the point is of no importance, but may serve to emphasize the correctness of the homology. The first description of the accessory lateral nerve is to be found in E. H. Weber's famous work on the auditory organ of fishes (1820, 216). He here describes and figures {i. e. root only, tab. v. fig. 30) the nerve afterwards called the " recurrent facial " of Silurus glanis. He says : — " Ramus primus nervi trigemini, per tectum cranii exiens, (in apicibus processuum spinosorum ad caudam usque progrediens, ibique cum nervis spinalibus ramisque nervi lateralis magni plexum complicitum nervorum componens,) nervi lateralis accessorii nomine appellandus." There can be no doubt as to the homoloo-v of this nerve, and it is to be noted that Weber considered it a bi'anch of the trigeminus and not of the facial as his successors did. Desmoulins & Magendie (1827, 60) refer to the nerve as the " pterigo-dorsal " (Part II. p. 869), and were the first to recognise the homology between it and the superficial nerve of tlie higher Teleosts. They give a fairly long description of its peripheral distribution in the latter fishes, which is, however, not as accurate as that published later by Stannius. Weber himself in 1827 (217J devotes further attention to the matter, and successfully homologises his first nerve with the elaborate cutaneous system in " Gadus lota " ( =Lota vtdgaris), thus confirming independently the conclusion of Desmoulins & Magendie. He renames the system the " Ramus lateralis trigemini," which is the name by which it has usually been known to anatomists, and describes its anastomoses with the spinal nerves. In 1830, Cuvier & Valenciennes devote a section of their work on the Natural History of Fishes (57) to a consideration of the accessory lateral nerves. They are both figured and described (t. i. pp. 440-44'l), and erroneously considered to be largely motor. The anterior and posterior roots are described, and so also are the anastomoses of the dorsal ramus with the spinal nerves. They found the branches to the pectoral and anal fins, but not, however, to the pelvic, and clearly recognised the homology between the recurrent nerve of the Siluroids and the better develojied system in the specialised Teleosts, believing that representatives of the system were probablv to be found in all tishes. Their figure of the nerves of the Perch, showing the accessory lateral nerves, has been copied by Owen (149, vol. i. p. 304), Nuhn (146, p. 558), and into several other text-books. Biichner (1835, 36) states that the accessory lateral is characteristically developed in the Cyprinoids, and describes it in " Ci/printis barbus " {^zBarbus vulgaris), but fails to distmguish it from the lateralis lateral line nerve. He, however, correctly compares it with its more specialised form in the modern Teleosts. Alcock (1839, 3) says (p. 268) : " Lastly, in many [fishes] the nerve [trigeminus] is distributed in a manner and to an extent for which there is no analogy among other animals, the fins being throughout furnished with branches from the fifth. Hence in fish, in which the distribution of the nerve is so much more extended than in other animals, both the size of it is proportionately greater, and it consists of a greater number of divisions ; these, which in the three other classes of vertebrate animals are only three, amounting with them to from three to six." A figure of the brain and " fifth " nerve of the Cod is given on p. 276, which shows the anterior root of the accessory lateral nerve 168 MR. F. J. COLE ON THE STRUCTURE AjSD MORPHOLOGY OF and its double origin. A brief description of it is also to be foviud in Bonsdorff's work published in 1846 (30), in which the two roots of the accessory lateral and tlieir course through the parietal are described in " Gadus lota'' but little is said about its peripheral distribution, whilst it is again renamed the " Ramus quartus s. lateralis nervi trigemini." The description which Stannius gives of the accessory lateral nerves (1849, 199) is concerned rather with their peripheral distribution, which, as far as the description goes, is perfectly accurate. Gadus and other Teleosts are carefully described and figured, and one may mention the figure of the former as being particularly admirable, whilst an account is also given of the homologous nerves in the Siluroid Teleosts. In establishing experimentally the somatic sensory nature of the nerves, and confirming the connection which they have in some of the modern Teleosteans with the dorsal, pectoral, pelvic, and anal fins, and also with the dorsal bi'anches of the spinal nerves, Stannius went a long way towards the elucidation of their true morphology, which subsequent observers failed fco follow up *. Hoffmann (1860, 100) briefly refers to the accessory lateral of Ci/pn'ims carpio as the "dorsale Schadelhohlenast " (one of Stannius's terms), but only devotes a few lines to it. He mentions the connection with the " trigeminus," and figures the posterior root in Cyprinus. Swan (1864, 205) also mentions it in a few lines, points out the connection of the dorsal ramus with the dorsal branches of the spinal nerves, and states that the system is not represented in the Skate. He gives two figures which show the origin and distribution of the accessory lateral nerves (which he calls " posterior branches of the filth "), on of the brain and roots, and another illustrating its peripheral distribution to the dorsal, j)ectoral, pelvic, and anal fins. I believe this figure, which appeared in the first edition of this work, published in 1835, was the first to accurately show the peripheral distribution of the accessory lateral system. Baudelot (1870, 15) briefly mentions and figiu'es it (figs. 2 and 6) as the recurrent branch of the " trigeminus," whilst Fee, who describes and figures in a number of Teleosts both the true lateral and accessory lateral systems, agrees with the erroneous conclusion of Weber that both these systems are perfectly homologous," although he had noticed that the true lateral nerve was never connected with the spinal nerves. As the result of numerous experiments Fee concluded that neither the triie nor the accessory lateral nerves were motor in function. Owen (1866, 149) says (p. 303): — " A branch of the vagus ascends forward to join the fifth in forming the dorsal division of the ' nervus lateralis,' which escapes by a foramen in the parietal bone," by which it will be seen that both the anterior and posterior roots of the accessory lateral were well known at that date. Baudelot (1868, 11) takes a somewhat bold step when he compares a recurrent branch of the pathetic nerve distributed to the pia mater with the ramus lateralis trigemini ! In the same volume (12), after an investigation of Leuciscus riUilus, he concludes that the true and accessory lateral nerves are perfectly homologous, and after referring to several authors who had * As 1 have been working from the second edition of Swan's Atlas, published in 18G4, I am taking it after Stannius's work. The first edition, however, was published before Stannius's, and anticipated it in many important respects, especially as regards the branches to the fins and the connections with the spinal nerves. THE CEANIAL NERVES AND LATERAL SENSE ORGANS OF i'lSHES. 1G9 discussed the supposed anastomosis of the tnie lateral nerve with the spinal nerves, he decides in favour of the connection between the two. It seems to the writer, from a careful examination of the various descriptions of this supposed connection, that the writers concerned have been dealing with the lateral ramus of the accessory lateral nerve, and have indeed found the connections they described, but mistake one of the nerves they were dealing with. However this may be, it is certain that the true lateral nerve never anastomoses with the spinal nerves. The "opercular branch" mentioned by Eaudelot (1869, 14) in various Cyprinoids seems to belong to the accessory lateral system, since it arises partly from the vagus and partly from the " trigeminus," :i3 LATERAL SEXfSE OKGANS OF FISHES. 181 This occurs particularly between the frontal and nasal on each side, and between the upper and lower ends of the preojicrculum and the squamosal and angular [=articular] respectively." In S. 409) says " forty," but this is doubtless a miscalculation, since the numbers he gives total 47. 25* 182 ME. F. .1. COLE ON THE STRUCTUEE AND MOEPHOLOGY OF (1) Supra-orhital canal. — In Amia there are three on the nasal and four in the frontal, of which the first nasal and the last frontal are those perhaps missing in Gadus. In hoth types the sense organs are fewer than in any of the other canals. The second dermal tubule of Amia is missing in Gadus, the fourth undoubtedly corresponds to the 3rd of Gadus, whilst the 6th, 7th, and Sth of Amia are again missing. Indeed, the Gadus supra-orbital is, for perhaps some functional reason, exceptionally abbreviated, (2) Infra-orhital canal. — I must first pi'otest against Allis's definition of this canal, which is very nnphilosophical. He considers i-ightly the innervation of the lateral canals to be the best criterion of their bovmdaries, and applies this method to all hut the infra- orbital canal, yet ofi'ering no justification for the exception. Behind the operculo- mandibular anastomosis the main canal is innervated by the lateralis lateral line nerve (including the lateral line branch of the IXth, which belongs to the lateralis), and this should determine the boundary of the body canal : in fact the operculo-mandibular anastomosis with the main canal is situated between the infra-orbital and lateral canals. Ewart (1892, 68) also criticises Allis's definition of this canal, and considers, Avith the writer, that its posterior segments belong to the lateral canal, whilst Miss Piatt (1896, 158) agrees Avith Ewart and correctly limits the infra-orbital canal to the buccal nerve. I hence consider Allis's sense organs 17-21 infra-orbital to be situated on the lateral or body canal. The ethmoid and ant-orbital or j^re-orbital ossicles of ^m«« are unrepresented in Gadus. Both are probably lateral line ossicles and do not belong to the skull sensu slricto — the ant-orbital being doubtless a modified sub-orbital plate. This diff'erence makes it difficult to homologise the sense organs of the two forms, but assuming that the first post-orbital of Amia represents two fused ossicles, as seems probable, then sense organs 9 . 10 . 11 . 12 . 13 Amia = 6 . 7 . 8 . 9 . 10 Gadus. Sense organ 14 of Amia in the sjihenotic is absent in Gadus, but the course of the otic nerve in the latter form is marvellously suggestive either of its having once existed there, or, what is perhaps more probable, that it has shifted backwards and now forms the 11th sense organ on the j)terotic. The otic branch of Amia innervates two sense organs, both on the pterotic, of which the 16th seems to be the one missing in Gadus, since the 16th dermal tubule is absent. Dermal tubules 13 and 14 Amia are also wanting in Gadus virens, but 13 is present as No. 9 in G. morrhua. The absence of 13 in G. virens is somewhat remarkable, but I have not seen it in any of the specimens I have examined. The outer buccal nerve is represented in Amia by the branches to sense organs 11 . 12 and 13 infra-orbital, with perhaps the branch to sense organ 14. (3) Ryomandihular canal. — The dentary of Amia bears 7 sense organs as against 6 in Gadus, but it would be impossible to say which was absent in the latter genus. The canal then enters in Amia what AUis calls in his first paper the " angular," but is some- what doubtful about the point. The bone undoubtedly corresponds to the Teleostean articular, which homology Allis has recognised in his last paper. Bridge's small ossicle and Allis's ossicle a (1897, 6, figs. 1 & 2) are doubtless comparable to the Teleostean angular. There are three sense organs in the Amian articular as against one in Gadus, and six in the preoperculum as against five. In the latter case there can be little doubt THE CRANIAL NERVES AND LATERAL SENSE ORGANS OF FISHES. 183 tbat it is the organ 16 of Allis that is wanting in Gadas, and further that dermal tuhule 11 Amia corres])ouds to No. 8 Gadus. (4) Lateralis canal.— lleve it is at once obvious tliat the tirst organ of Amia situated in the squamosal ( = the glossopharyngeal organ No. 17 infra-orhital Allis) is absent in Gadus. The next four [Amia) are situated in the single supra-temporal or extra-scapula, and these doubtless correspond respectively to the first four sense organs and the four supra-teniporals of Gadus. It is possible that facts like this (and there are many others) show that each sense organ had originally its own supporting ossicle. The i^ost-temporal or supra-scapula of Amia has one sense organ, but none in Gadus, whilst a further difference is that the scapula or supra-clavicle of Amia supports a portion of the body canal and two sense organs, but takes no part in the support of the canal system of Gadus. It is true that some authors consider the supra-clavicle a post-temporal ossicle, but it is impossible to satisfactorily homologise the two bones, and hence the difference between Amia and Gadus must remain. Behind the region of the skull and shoulder girdle both forms agree in that the lateralis or body canal is supported at intervals by small lateral line ossicles. We thus see that the two forms agree in a very remarkable nianner. This will be seen if a comparison be made between Allis's tig. 49 and fig. 2 of this paper. Except that Amia has more sense organs, and omitting differences in detail, the Ganoid form does not differ as much from the Teleostean form as many of the latter do among themselves. In fact the resemblances between the two are much more remarkable than their points of difference. These facts of course strongly support the view originally maintained by fossil Ichthyologists that the Ganoids cannot be ordinally separated from the Teleosts. Guitel (1890, 94) describes the lateralis lateral line nerve in Cyclopterids exactly as it is in the Cod. There are the dorsal and ventral rami — the dorsal innervating the anterior part of the body canal, and the ventral the posterior moiety. The same author (1891, 95) has given us a somewhat full paper on the lateral line of Lophlus. He has unfortunately made an insufficient study of the literature, and hence misnames the nerves. He states, for example, that the supra-orbital caual is innervated l)y the " ophthalmique du trijumeau," and also that a part of the operculo-mandibular line is supplied by " le nerf operculaire superficiel du pneumogastrique qui echange des fibres avec I'operculaire superficiel du facial." Further, the infra-orbital canal is said to be supplied by the " maxillaire superieur du trijumeau." Guitel mentions Allis's first Amia paper, but omits to state on what grounds he rejects the latter's nomenclature and adheres to the old views. His errors are perhaps largely explained by the fact that he did not follow out the components of the nerves with the microscope. A supra-orbital commissure is described as in Gadus. In Elasmobranchs we find the lateral line system in its most complex form. Beyond in a very general way any comparison between the sensory canals of Elasmobranchs and bony fishes is somewhat out of the question. This is very largely due to the absence of dermal bones, which afford a very valuable guide in determining the homologies of the individual sense organs and portions of the cinals. Ewart (1892, 68) describes infra-orbital and supra-temporal anastomoses in Lcemaryus, the former lodging two sense organs. The infra-orbital ;ilso anastomoses with the supra-orbital and hyo- 184 ME. F. J. COLE ON THE STRUCTURE AND MORPHOLOGY OF mandibular, and the otic nerve innervates the free anterior extremity of the lateralis canal, which extremity must therefore be considered to belong to the infra-orbital line. In the Skate {Mala hatis) the remarkable reduction of the dermal tubules in several places is a noteworthy feature, and enjoins an investigation into the development of these portions of the canal system (Ewart & Mitchell, 1802, 69). Pollard (1892, 161) states that the superficial ojihthalmic and buccal lateral line nerves arise from a single trunk in Clarias (p. 529) and Auclieiiaspls (p. 532). This is interesting when we remember the description of the development of these canals given by Wilson, Mitrophanow, and Locy. In Callicthys, however, the otic nerve arises separately from the ganglionic complex and not from the buccal nerve (j)p. 534-535). In Clarias sense organ 1 of the lateralis canal [-— s. 0. 7 main canal, Pollard) is stated in the text (p. 529) to be suj)plied by the glossopharyngeal, whilst the figure of this form shows an innervation from the lateralis lateral line nerve. On p. 534 in Callicthys the ampullation of the " main canal " is described as a " remarkable phenomenon, hitherto unrecorded in any other animal." It is in fact described by several authors, notably by Traquair (18G5, 207), Hyrtl (186G, 105), and Sappey (1880, 175). In Clarias, Auchenaspis, and Choitostomns tb(> operculo-mandibular canal communicates with the body-canal, but in the Siluroids, where the canal system is reduced, it is unconnected with any of the other canals. In comparing the lateral line system of Siluroids with Coccoslens, Pollard apparently overlooked until too late the accurate description by Traquair published two years before, and makes use of the somewhat erroneous figure by Pander. The lateralis lateral line nerve of Ferca, according to Cuvier and Valenciennes (1830, 57;, Vogt and Yung (1894, 215), and other authors, is in much the same condition as in Gadus, dividing into an upper ramus to the anterior portion of the body canal, and a lower ramus to the posterior portion of the same canal*. Herrick (1897, 98) finds in the highly specialised Acanthopteran Meuidia that although macroscopically the nerves are fused and their relations disguised, yet microscopically the whole of them may be reduced to the common tyj)e, of which Lmmargus, Raia, Chimcera, Protopterus, Amia, Gadus, and the Siluroids are other examples. The vagus ganglion can be broken up into five ganglia — four of which belong to the branchial nerves and one to the visceralis. Herrick finds the body canal of Menidia occupying a superficial position posteriorly just as it does in Gadus, and further points out that the lateral line branch of the glosso- pharyngeus is really a branch of the laterahs. The common origin of the lateral line and avxditory nerves is confirmed, and the V-VIIth nerves arise by a double root as in Gadus, of which the ventral one is the motor root of the facial and is connected with the hyomandibular trunk only. All the lateral Line nerves, except the lateralis, arise from the dorsal root of the trigemino-facial complex, but in two handles — a dorsal bvmdle for the superficial ophthaluuc + l^^i^ccal, and a ventral l)undle for the external maudibidar. The otic nerve innervates three sense organs, whilst two aj^parently derive * In ri-otoji>tt:rvii thfic are four rami to the lateralis, and nil (lislliiciivih/ laUrcil line nerves (I'inkus, 1894, 157), whilst the aceessorv lateral system is eompletely absent. THE CRANIAL NERVES AND LATERAL SEXSE ORGANS OF FISHES. 185 their innervation from the glossopharyngeus. There is thus a general agreement hetween Gaclus n,ndi Ilenidia, nx\A the differences hetween them are certainly not greater tlian hetween Gcuhis and Amia. Q. The Lateral Sexse Ohgans of Vertebkatks and Invertebrates. Before proceeding to discuss tlie phylogeny of the lateral sense organs, it is first necessary to enquire into the evidence relating to their supposed homology with the lateral organs of Invertehrates, as maintained by Eisig, Whitman, and other authors. The bulk of the evidence is against such an homology, hut the position of some morplio- logists with regard to tlie question renders its discussion somewhat necessary. Leydig (1868, 126), who was one of the first authors to conclude tliat Savi's vesicles, Lorenzini's ampullae, and the sensory canals all Ijelonged to the same system, /'. e. the lateral line system, was also the first to connect the Ititeral sense organs with the sense organs of Annelids. This view was, however, first elaborated in detail by Eisig (1879, 64, and 1887, 65), who in his Naples monograph devotes considerable attention to the question (pp. 501-517). He endeavours to show that the two series of organs are homologous in all essential respects, and even goes to the length of compai'ing a trans- verse section through the branchial region of a Vertebrate with an inverted transverse section of a Capitellid worm. Tlie resemblances to which be draws attention in this connection are of a very superficial character, and certainly not sufficient to establish the relation of the two sets of sense organs on a reliable foundation. Eisig seems to have been misled by mistaking the lateralis nerve of the Lamprey for a true lateral line nerve, for his theory demands that the lateral organs of Vertebrates should be metameric, and the lateralis -hqvyq oi Petromyzoii, it 11 lateral line nerve, would strongly support that view. We have ah-eady seen, however, that the bulk of the evidence is against the primitive metamerism of the vertebrate lateral line system, and this consideration in itself is sufiicient to disprove the theory that Eisig seeks to maintain. Balfour, writing however, before the publication of Eisig's mature views, says (1881, 9, vol. ii. p. IIS) : — " The organs which resemble those of the lateral line are the remarkable sense organs found by Eisig in the Capitellida3 ; but I am not inclined to think that there is a true homology between these organs and the lateral line of Vertebrata." In his first paper. Whitman (1881, 219) compares the segmental sease organs of the Leech with the vertebrate latei-al line, and believes that both may be traced back to a common origin. In his next work (1889, 220) he considerably develops this view, and considers further that the lateral organs have served as a starting point not only for the taste organs, nose, and ear, but also for the eye. He assumes, what is indeed possible, that the vertebrate lateral organs were derived from invertebrate sense organs of some sort, but it is only adding assumption to assumption to further postulate a metameric arrtingement for tliese sense orores, may be retained in the adult, but most of them undergo a repeated dichotoinous division, thus giving rise to groups of surface jiores and to corres^'oudiug dendritic systems of * Wright, however, was wiitiug without access to Merkel's work. 190 ME. F. J. COLE ON THE STRUCTURE AND MORPHOLOGY OF canals." Before this can take place, however, the anlage of the system must be formed, and Allis finds (ji. 535) tliat " each group [/. e. line of sense organs] develops from the cord of cells lying iix the deeper layers of the eiiidermis, and each cord from a special sensory thickening, which when first seen from the surface in S2:)ecimens liardened in chromic or picro-sulphuric acid, appears as a large, whitish, and slightly raised spot." From this anlage all the canal organs arise, and " in the early stages of their development lie below the surface, but they soon push through the overlying epidermal cells, and their upper central portions become exposed. Each pit organ subsequently sinks slightly below the surface, and a little epidermal pit is formed above its central portion " (p. 536). The canal organs are develojied as above described. Combining Allis's results first with the work of Balfour and the eai'lier embryologists and then with the recent observations of H. V. Wilson, Mitrophanow, and Locy, we are enabled to draw up the following scheme * of the development of tlie lateral canal system, which is known to apply to some, if not to all, fishes : — (1) A small sensory anlage arises in the neighbourhood of tlie ultimate position of the auditory organ, the differentiation of which gives rise both to the auditory organ and the system of sensory canals. (2) This anlage grows forwards and backwards as a non-segmental cord of cells. In front this cord bifurcates and forms the anlagen of the supra- and infra-orbital canals. Behind it forms the anlage of the lateral or body canal. (3) Along the non-segmental cord arise series of sense organs, which, in the case of the body canal are segmental, but are not known to be such in the other canals. (4) Each sense organ sinks down, carrying \he skin with it, so that it ixltimately lies at the bottom of a groove. The lips of the groove fuse — each sense organ therefore lying in a short tube opening at each end on to the surface by a pore ( = "half pore," Allis). (5) The short tube extends both ways and fuses with the two adjacent tubes, adjacent half pores fusing to form a single " primary " pore which in most fishes forms the external aperture of a dermal tubule. In the Ganoid fishes the primary pore, by a process of repeated dichotomy, produces what is known as a " dendritic system." A lateral canal is therefore formed by the end-to-end fusion of pieces or segments, and not by the formation of a primitive continuous furrow, and it further follows that there must have been primitively a dermal tubule between every two adjacent sense organs. Bateson (1889-90, lo) fails to distinguish between terminal buds and pit organs, and although pointing out that these two sets of organs have not the same histological structure, seems to have overlooked the previously described fact that the pit organs belong to the lateral line system. According to Coggi (1891, 42), Savi's vesicles develop in the same Avay in Torpedo as the sense organs of the lateral canals, but his description differs somewhat from Balfour's observations on Scyllinm. Ayers (1892, 7), who occupies an isolated position on some aspects of the question, believes Savi's vesicles to have been produced retrogressively from canal organs, and considers that they are " without doubt descendants of the canal organs " (p. 164). He bases this opinion on the observations of Fritsch, which he thinks show that the Savian vesicles * This scheme ignores the opereulo-mandibular caual, of the development of which our knowledge is still very imperfect. I sec uo roa-^on, however, wh}' it should not develop in just the same way as, for example, the infra- orbital does, i. e. by branching from the main trunk. THE CEANIAL NERVES AND LATERAL SENSE ORGANS OF FISHES. 191 " are transformed from the canal type during ontogeny," and concludes that " the Torpedoes possess a system of organs which is lacking in all other Batoids, but which is represented in these forms by the more primitive canal type." It cannot, I think, be admitted that there is any foundation for the view that Savi's vesicles are deirenerate canal organs. The observations of Allis, Coggi, Boll, Merkel, and numerous other authors, show that the pit organs, Savian vesicles, and other more or less superficial sense organs, develop as the canal organs do up to a certain stage, but that there they stop. There is no evidence of retrogression, and none in fact that any of the lateral line organs ever acquire a more central position than that which they ultimately occupy in the adult. The fact that a sensory canal of one form may be represented by a line of pit organs in another, may be more reasonably explained by precisely the opposite hypothesis, i. e. that the one type represents a more advanced condition than the other. This is the view adopted by most morphologists, and the one that has the sanction of the facts *. The pit organs oi Amia are correctly homologised by E wart & Mitchell (1892, 69) Avith the sensory follicles they described in the Skate; and the '' Spalt-papillen " of Fritsch. Bashford Dean (1895, 58), in referring to the pbylogeny of the lateral canals, says (p. 49) : — " The ancestral condition of the lateral line of Sharks appears to have been represented in an open continuous groove, lined with ciliated sense cells, and protected only by an overlapping margin of sliagreen denticles. In this condition it at least exists in the ancient Sharks and in CJiimcera. That the canals of the head rea-ion were also primitively of this character appears cvceedingly probable; they are thus retained in the adult Chumera." Ley dig (1895, 128) gives a brief expression to a very important and significant fact when he states that there are so many varieties of lateral sense organs, from the superficial organs to the canal organs, that it is impossible either to divide them into groups or even to draw sharp lines between individual forms. A passage in Miss Platfs work on Nectariis {I'fi^Q, 158) is interesting in this con- nection. She says (p. 526) : — " At each side of a mid-dorsal fold in the skin a row of mucous glands is found, composed of a few cells invaginated from the deeper layer of the ectoderm, and now lying below the surface, tiny balls of cells surrounding a central cavity that opens to the surface by a small pore. Similar glands are found on the ventral surface of the body between the fore-limbs, and on the tail. Although these glands are about the size of sense organs, nothing in their structure or in the manner of their development suggests the genetic affinity of sense organ and mucous gland on which Leydig [1895, 128] insists." Whilst it is true, as Miss Piatt maintains, that there is no genetic affinity l)et\veen the lateral sense organs and mucous glands, it is also true that a lateral sense organ consists of sensory and secretory portions, and it is indeed probable, as Eisig suggests, that certain dermal mucous organs have become incorporated into the lateral line system. In his last Amia paper (1897, 6) Allis says (p. 627) : — " In early stages of development they [i. e. canal organs and pit organs] closely resemble * Ayers' views on the morphology of the semicircul;i,r canals compel him to take up the position that he does (see particularly pp. 218-220;. Nevertheless I submit that tJicre is no Icuowii fact in the Jdstori/ of the lateral Jim system tending to sJiow that there is any differentiation in the canals after thsy are once enclossd. 192 ME. F, J. COLE ON THE STRUCTUEE AND MOEPHOLOGY OF each other, but in later stages they differ greatly, the pit organ retaining its embryo- logical form and individuality, while the canal organ first increases greatly in size, and then, by the independent growth of other similar organs immediately adjoining it, gives rise to a large sensory patch, or nerve ridge (Merkel), in which the separate organs lose to a greater or less extent their individuality." Again, on p. 629, lie says : — " A line of pit organs represents ajoparently the possibility of a canal ; with the disappearanee of the organs the possibility of the canal, even in rudiment, would certainly disappear." A striking fact in the comparative anatomy of the sensory canals that has hitherto been overlooked is the relative abundance of the mucus secreted. Starting with the Cyclostome fishes, we find most elaborate mucous organs and a very imperfectly developed lateral line system. In Myxine the mucous sacs occupy the position of the lateral or body canal in the higher fishes, and I think it must be agreed with Eisig and Dohrn that these sacs represent very archaic structures which must have existed before the lateral canal system itself. In the Elasmobranch fishes we get a condition ia which mucous and sensory elements are combined, but in which the sensory elements as it were have developed at the expense of the mucous elements. In other words there is a com- bination of sensory and secretory cells, but the latter now exercise a subordinate (though prominent) instead of a predominant function. Above the cartilaginous fishes the m.ucous elements become more and more reduced, until the minimum is reached when tliey are reduced to a few secretory cells at the base of each sense organ. Further, in the recent TeleostSj the sense organs themselves are unmistakably reduced in number — in some cases very considerably. The lateral sense organs were, as is well known, for a long time regarded as secretory organs. On the discovery, however, of sense organs in the canals, opinion veered romid to the other extreme, and completely ignored the undoubted occurrence of glandular structures in the lateral canals and ampulhe. This is the further surprising since the existence of such structures is to be expected. The other sense organs of the head — the nose, the eye, and the ear — have all accessory glands connected with them, and the physiological necessities of the lateral sense organs postulated the existence of glandular cells liere also. Now the glandular organs ai)pcar to have played at first a predominant, and then an accessory part in the history of the lateral line organs, and it consequently follows that if they were at first predominant they were also pre-existent and independent structures. There is therefore, in my opinion, some justification for the view that the sensory and secretory portions of the lateral line system were at first independent of each other, and that the incorporation of the latter into the former necessarily followed on the adoption of the canal in preference to the superficial type of sense organ. The rapidly evolving lateral line system, as it sank below the surface, seized on the glandular organs and subordinated them to its own use. But what is this use ? I think it is sufficiently obvious. From tlie structure and consistency of the lateral canal mucus, it follows that it has considerably departed from its original function of lubricating the surface of the body, and can no longer perform that function. In the lateral canals it forms a delicate jelly, the "shivering" of which agitates the sensory hairs and thus couvc) s a message from witiiout. It thus corresponds precisely to the endolymph of the ear, is physiologically unnecessary in a superficial THE CRANIAL NEKVES AND LATERAL SBN8E ORGANS OF FISHES. 193 sense organ, but became an obvious necessity when the superfical organs, increasing in number, size, and sensitiveness, were sinking below tlie surface. The later process, whilst it afforded the desirable protection to the sense organs, necessarily removed them further from the sphere of their activity and demanded the physiological compensation supplied by tlie introduction of mucus into tlie canals. That the superficial type of sense organ preceded the central canal organ, or, in other words, that tlie sensory canals arose jirimitively on the surface and then sank into their usual subdernial position as definite canals, is to my mind amply proved both by fossil Ichthyology and Emlnyology. The former shows us that in the most ancient fishes known, such as the Ostracoderms, Artlirodira, and the older Sharks {Cladoselache, Cladodus), the lateral canals existed as superficial grooves, which must therefore be held to be the most primitive known condition. The development of the canals iu existing fishes is further evidence in the same direction, and tells us that superficial grooves arise first by a process of decentralisation, whilst the canals themselves are formed in segments after- ■viards /i/ situ. These two considerations enable us to form a general idea as to the lines ujion which the lateral canals were laid down. Turning to the comparative anatomy of the system, we can not only confirm the view suggested by the above evidence, but are enabled to follow the process iu some detail. Whilst I fully concur in Lsydig's contention previously mentioned, we may for the purposes of convenience classify the lateral sense organs under three heads, as follows : — / (a) Pit organs sensu striclo, as iu Tcleosts. Most superficial. (1) Superficial ij/pc ■! (b) Saviun vesicles [Torpedo] and nerve-sacks (Ganoids). lutermeJiati! ^ between 1 and 2. (2) Intermediate type. AmpuUffi of Lorcnzini (Elasniohi'anchs). (3) Central type. The canal sense organs (almost all fishes). Here we have a continuous series connecting the superficial sense organs on the one hand with the canal organs on the otlier, and indicating the manner in which the canals and their organs may have been produced. In nearly all fishes the lateral line system is in a very unstable condition. In Elasmobranchs the pit organs have sunk below the surface, and have been converted either into Savian vesicles or ampulkB of L u-enzini, principally the lattei', with the result that pit organs sensu stricto are largely absent. In Teleosts, where Lorenzmian ampullae are absent, we find, as we should expect, the pit organs on the surface, and existing in large numbers, perhaps as com- pensation for the more or less reduction of the canal organs themselves. In Ganoids, where the canal system, as far as our limited information goes, still flourishes, the pit organs are comparatively few and have in some cases Ijcen converted into what Merkel calls nerve-sacks. Iu many cases we find undoubted evidence of a canal in one form being represented by a line of pit organs in another, showing tliat the latter represent the possibility of a canal. Thus there seems to be a constant migration of sense organs from the surface to the interior — a constant stream of superficial organs inwards forming canals and organs in some cases and reinforcing them in others. It is easy to see, for example, from what we know of the development of the canal organs, how a line of ampuU-ae might form a canal. The insinking would only have to form a furrow 194 MR. F. J. COLE ON THE STEUCTUKE AND MOKPHOLOGT OF instead of an ampuUary tube, and every condition for the development of a canal would be satisfied, since the latter, as we know, is formed in segments. This brings us to the final question as to whether the canals were originally formed as continuous furrows or by the fusion of contio-uous pieces. On this point the evidence is unanimous and conclusive. The comparative anatomy of the different classes of sense organs and canals, and especially the development of tlie latter where known, prove beyond doubt that the lateral canals were formed, not by the closure of continuous furrows, but by the fusion end to end of adjacent tubes. The precise processes, as far as can be judged, would take place in the following order : — (1) Formation of cords of indifferent cells along the lines of the future canals ; (2) differentiation of sense organs along these cords ; (3) sinking down of each sense organ to form a furrow, and the fusion of the lips of the furrow to form a tube open at each end ; (4) fusion end to end of series of these tubes in such a way that continuous canals are formed, and there is an opening on to the surface between each two sense organs. Such, I believe, has been the phylogeny of the sensory canals, and the question must now be left for further investigation either to substantiate or to disprove this view. S. The Lateral Sense Organs and the Auditory Organ. Thomas Willis (1664, 224) may be said to have suggested the morphology of the auditory organ to his better informed successors, when, in the same year that the lateral line system was discovered, he described the auditory nerve as the dorsal branch of the VIIth,and considered both nerves to constitute the seventh pair of cerebral nerves. Leydig, however, was the first to derive the auditory organ from lateral sense organs (1850, 120), but he unhappily recanted in 186S (126), and described the lateral oro'ans as forming an organ of 6th sense, and having no morphological connection with the auditory organ. In 1870 (182) Schulze, who erroneously thought that water flowed through the sensory canals, regarded the latter as accessory auditory organs ; whilst Dercum (1880, 59) went still further, and minutely compared the histology of the lateral sense organs with the maculae of the ear, concluding that they resembled each other in every essential respect. Emery (1880, 66), who agrees with Schulze that the lateral organs represent an accessory auditory system, first discovered that the lateral oro-ans possess a well marked cupula, which he believed to be foi'med of successive cuticles secreted by the peripheral cells of the sense organ, and correctly compared with the cupula terminalis of the ear. This important and interesting discovery has been extended to several other forms besides Fiei^asfer, and I find a very well marked cupula in young Gadus virens (see PL 23. fig. 4). The important work by Mayser (1882, 135) is perhaps the first scientific contribution to the question, and formed the basis upon which all further work was conducted. This author discovered in Cyprinus that the fibres of the lateral line nerves and also the fibres of tlie auditory nerve arose from a common centre in the brain — the tuberculum acusticum. He was consequently led to regard the lateral line system as a low form of auditory organ, and indted describes the lateralis lateral line nerve as the " hintere THE CRANIAL NEEVES AND LATERAL SENSE ORGANS OF EiSilES. 195 Acusticuswurzel," and regards the semicircular canals as modified lateral tubes. I entirely concur with Ayei-s that " Mayser was mucli nearer the truth tlian he ever realised when he arrived at the conclusion above quoted " (7, p. 112). Mayscr's discovery of the common origin of the two sets of fibres has been fully confirmed and also extended by many subsequent authors, such as Strong, Kingsbury, and Herrick, so that it must now be considered as an established fact. Bodenstein (1882, 24) briefly discusses (p. 137) Mayser's conclusions, and largely concurs. He compares the semicircular canals with the canals of the lateral line, and considers the endolymph of the ear to correspond to the mucus in the lateral canals, but is somewhat inconsistent in regarding the auditory organ as phylogenetically older than the lateral line system. Beard (1884, 17) looked at the question irom a new aspect when, instead of regarding the lateral system as an accessory auditory organ, and therefore discounting the value of the comparison by im2:)lying that tlie two systems are only homoplastic and not homologous structures, he derived the auditory organ from the lateral line system, and concludes that '• the auditory organ of Vertebrates is fundamentally a specialised portion of the system of sense organs of the lateral line." This view, that the ancestral Verte- brate possessed a system of lateral sense organs which gave ri'^e lioth to the auditory organ and to the modern lateral line system, has been abundantly supported by further evidence, and now stands as one of the most probable views of vertebrate cephalogenesis. Beard subsequently developed his hypothesis (1885, 19), and considered, with Gegeubaur, Marshall, Balfour, and other embryologists, that the auditory nerve was a dorsal branch of the facial. In 1884, Wright (227) followed Bodenstein in comparing the endolymph of the ear with the mucus in the lateral canals, and in a further publication in the same year (228) independently confirms Mayser by describing a common origin in the brain for the lateral line and auditory nerves of Amiurus. He, however, still perpetuated the mistake of his predecessors in classing most of the lateral line nerves with the trigeminus, and further states, what m.ust be regarded as doubtful, that the buccal nerve " contains fibres other than those derived from the tuberculiim aciisticmn." Eisig (1887, 65) discusses Beard's derivation of the auditory organ from lateral sense organs, and is favourably disposed towards the view (pp. 711-712) ; whilst Fritsch (1887, 75), who was unable to find a cupula to the lateral sense organs of Ilalaptenii-us, advances the extraordinary explanation that it may have been washed away by the passage of sea water through the canals. The cousins Sarasin (1887, 177), who favour the derivation of the auditory organ from lateral sense organs, and further consider that the latter function as accessory auditory organs, enter into an elaborate comparison between the histology of the auditory and lateral sense organs, and endeavour to establish that both are identical in all essential respects. Cunningham (1890, 55), discussing the cupula of the lateral sense organs, is " inclined to think the cupula is, during life, of a mucous nature, and therefore semi-liquid. It seems certain that the sensory hairs are imbedded in the cupula. It is difficult to understand how such cells as those of the sense organ should secrete mucus or form a cuticle ; perhaps the cupula is nothing more than the ordinary mucus of the dermal tube, SECOND StUlIES. — ZOOLOGY, VOL. VII. 27 196 MR. F. J. COLE O'S THE STRUCTURE AND MORPHOLOGY OF which keeps a coustant position in prejoarations because it is retained by the numerous sensory hairs " (p. 81). Here Cunningliam has overlooked the gland cells at the base of the sense organ, which are doubtless responsible for the secretion of the cvipula. There is therefore no necessity to suppose that it is secreted by sensory cells. Fritsch (1890, 77) was again unable to find a cupula in the Torpedoes, but afte r careful investigation admitted that the lateral sense organs agreed very closely in their minute histology with those of the auditory organ. In 1892 Ayers published his elaborate and important memoir on the morphology of the vertebrate ear (7). In this work he endeavours to show that the cupuloe terminates of the lateral and auditory sense organs are " artifacts produced in the main from the auditory hairs of the sensory cells " (p. 314). He says (p. 163) : — " Now, while there is no question that some mucus is present in all cases (all in which exact analyses have yet been made), there is no evidence to show that in a state of nature it exists as a dome- shaped mass covering the sense organ. On the other hand, it is clearly proven that the most typical cupulae yet studied are products of the action of reagents, and that so far as fibrous structures in the channels occupied by sense organs (whether car organs or those of the lateral line system) are concerned, they have in every thoroughly studied case shown themselves to be hairs more or less unchanged." I have previously referred to the existence of cupuloe over the lateral sense organs as evidence that the latter and the auditory organ are genetically related. Whether the cupulye are artifacts or not is a question that does not affect this argument, since the conditions producing the artifact must be the same in both cases, and the ai'tifact therefore open to be used in argument as denoting equally important and similar conditions in both systems of sense organs. The following quotations from Ayers' valuable monograph briefly summarise his line of argument : — " The whole process of the formation of this [auditory] cup is an exact repetition of the formation of a canal organ and its canal as seen in Amia, or tlie formation of an ampulla and its canal in the Salmon " (p. 175). "There could not be a more complete agreement between two developmental processes, occurring in diff^erent animals, than there is between the formation of the canals on the surface of the body in Amia and the formation of the canals of the internal ear in the Herring and Salmon, as given by Von ISoorden. Of the correctness of the account for the latter I am certain, for I have observed the same process both in the whole embryo and serial sections. From uur present knoAvledge of the development of tlie labyrinth in the bony fishes we have only agreement with the type of growth of the surface organs " (p. 181). " The so-called semicircular canals develop as more or less curved structures, because they are formed out of a portion of the external surface of the body, which has been enclosed within the head. Since on the sxu-face of the body, in the typical and usual process of develojiment, the canals always open at both ends on the surface, so inside the auditory vesicle they open at both ends " (p. 222). " It is proven that there is no essential difference between the sense organs of the internal ear of the vertebrate group and the superficial sense organs of the Ichthyopsida as they exist in the surface canals of these forms. Both consist of canals containing sense organs lying below the surface of the body, and they may (as in ordinary canals and the ear in some Elasmobranchs) or they may not (as in THE CR.\.NIAL NERVES AND LATERAL SEN.SE ORGANS OF FISHES. 197 Savi's vesicles [!] and the ear in the majority of Vertebrates) comrauuioate with the surface. Both forms may or may not be entirely enclosed in cartilage or in bone. The semicircular canals of the ear are simply remnants of the canal system of the surface, and although bent into shapes more or less semicircular, they still retain their communi- cation Avith the (morphological) exterior by means of their terminal (/. e. surface) pores or ojaenings in the walls of the utriculo-sacculus, which in its turn often communicates with the surface of the head by the endolymphatic duct or surface canal Thus it is true that the develoi^ment of the semicircular canals in the ear is due to phylogeuetic and mechanical causes, and is not in the least a response to physiological necessities or requirements, and conversely the canals have no active part in the auditory function. They serve merely as chambers to hold the liquid necessary to tloat the auditory sensory hairs " (p. 308). Ewart & Mitchell (1892, 69) remark (p. 100) :— " Restiug on the top of the hillock there is often what Solger terms the ' Cujmlabildung.' This seems to consist of mucin. In some cases we have seen long threads of mucin extending from the hillock into the cupula or across the canal, the threads having frequently leucocytes entangled between them." A similar condition of the cupula is described by Emery (18S0, 66). Willey (189-4, 223) and Bashford Dean (1895, 58) agree that the auditory organ represents a specialization of the lateral line system, and belongs to the same category as the sense organs of the remainder of the lateral line. Locy (1895, 130) goes further, and regards the ear, the nose, and probably the eyes as derived from the lateral line system, and states on p. 51'7 that " there is now general agreement that the ears belong to tlie lateral line series." Again, on p. 579, he says : — " The history of the auditory vesicle in sharks has been worked out beyond this period by Ayers, and it is very clear from its mode of growth that it is directly related to the canal organs of the lateral line." Finally, Miss Piatt (1896, 158) remarks (p. 492): — "The ear has developed from the dorso-lateral thickening in the hyobranchial intersegment, and it will be noticed that although the ear undoubtedly belongs in [?to] the lateral line system, and is in fact the centre of that system, it is not proj)erly a ' branchial sense organ,' as Beard suggests, for this term cannot be acciu^ately applied to sense organs above the epibranchial line." The general fact that the auditory organ is a modified portion of the lateral line system must hence be considered established on a secure basis of fact, and it accordingly follows that the lateral system of the ancestral Vertebrate must have existed before differentiated sense organs, and then diverged in at least two directions (omitting mention of the nose and eye, the phylogeny of wdiich is doubtful) — one producing the lateral canals of recent fishes and the other the vertebrate auditory organ. The associa- tion of the two sets of organs is completely justified on the evidence of the innervation alone, since ]\Iayser and Strong have estal:)lished (1) that the lateral line and auditory nerve fibres constitvite a system by themselves, and are quite independent of the other cranial nerves ; (2) that this system arises from a common central origin in the brain, which is farther distinct from the fibres of any of the other cranial nerves. The argument from the nerves is well backed up by the argument from the minute structure and development of the two series of organs, and I have already drawn attention to the fact 27* 198 MR. F. J. COLE ON THE STRUCTURE AND MORPHOLOGY OF that in both these respects (?'. e. histology and embryology) the agreement between the lateral and auditory organs is so remarkable as to admit of only one explanation, and that favourable to the close geuetic relationship of the two series. When, however, we descend to details, and enquire into the phylogeny of the semicircular canals, the way is not so clear. Ayers contends that a lateral line canal may differentiate further after being cvit off from the surface. I have already joined issue with him on this point, and have urged that there is no direct evidence favouring this statement. The lateral canals are jirimitively laid down in the form they are found in the adult, and the only possibility of svich differentiation is perhaps to be found in the blind diverticula of the canals described by Hyrtl, Sappey, Pollard, the writer, and others. Ayers' postulate is by no means impossible, but is at present mere speculation, and not founded on fact. The bearing of this on the question of the semicircular canals is obvious. The latter develop as evaginations of the wall of the auditory sac after the sac has become cut off from the surface *, and is in a condition corresponding to a, lateral line canal. It is true that the lumina of the semicircular canals are pieces cut off from the outside world, and one can admire Ayers' comparison of a semicircular canal with a single segment of a lateral canal still opening on to the surface (= cavity of sac) by its two terminal pores. These con- siderations, and the others advanced by Ayers, do not discount the cardinal fact that the semicircular canals of all Vertebrates develop in a way for which there is no jiarallel in 0 r knowledge of the anatomy and development of the lateral line system. At one time 1 thought that the ear with its semicircular canals was formed by a concentration of certain lateral canals, and that the aqueeductus vestibuli or ductus endolyniphaticus corresponded to a dermal tubule connecting an internal canal with the exterior. Both these contentions are held and ably advocated by Ayers, and the latter undoubtedly holds good. The question that remains is whether the auditory organ has arisen by the diflfereutiatioa of an organ such as an ampulla of Lorenzini, or whether it was formed by the concentration of certain primitive canals of the ancestral Vertebrate. In spite of the wealth of argu- ment employed by Ayers, I cannot pass over the developmental difference described al)ove, and must therefore hold to the former view, which by the way explains why there shoiild be only one aquseductvis vestibuli in the history of the auditory organ. I thei-efore follow Beard in regarding the ancestral auditory organ as represented by the primitive auditory sac of the embryo, and that the semicircular canals arose later in the ancestral history of the organ, and after it had entirely lost connection with the lateral line system. Such a view of course involves the corollary that although the semicircular canals bear many points of resemblance to the sensory canals of the lateral line, yet they must be ren-arded as only homoplastic and not homologous Avith them. Ayers' view, on the other hand, implies that the auditory organ was from the first a complicated structure and possessed semicircvilar canals. * [n some of the lower Vertebrates, of course, this never comjihteh/ bai)pens. THE CRANIAL NERVES AND LATERAL SENSE ORGANS OF FISHES. 199 T. Special Considerations. Schulze (1870, 182) was the first author to state that the lateral canals arose in sections with a sense organ in each section, and that the dorsal tubules were formed by the fusion of the adjacent extremities to form a single peripheral tubule and pore. This was confirmed by Solger (1880, 193), Bodeustein (1882, 24), and AUis (1889, 4) — the latter's work in tliis connection having been previously referred to. Friant (1879, 73) indepen- dently discovered with Schwalbe that tlicre were two siqjerficial oplithalmic nerves, and further has the credit of being the first author to approximately arrive at a correct appreciation of the trigeminal and facial nerves of fishes, being the first writer to hold that the superficial ophthalmic and l)uccal lateral line nerves were not branches of the trigeminus*. He says (p. 103): — "Le nerf facial nait isolement des cotes du bulbe, au-dessus de la racine posterieure du trijumeau, se distribue aux meninges, a la peau et aux canaux muqueux du sommet de la tete et de la region sous-orbitaire, aux teguments d'enveloppe de Freil et de I'orifice nasal, a la peau et aux teguments fibro-musculaires du museau, aux muscles peaueiers de la joue, et chez la Perclie, a la jieau de la reo-ion dorsale et aux muscles des nageoires dorsales." It will lie seen that for the time it was written this was a very accurate description, and its historic imjjortance makes it extraordinary that Friant's work should hitherto have been overlooked. The fishes are classified by Sappey (1880, 175) into four groups according to the development of tlie lateral line system : — (1) the Plagiostoma, in which it is greatest developed ; (2) " L'Ange et les Squales," not so well marked ; (3) the majority of the bony fishes, where it is more reduced than in 2 ; (1) fishes having no lateral line system at all[!]. Such a classification, based on so very variable a structure (even though its general anatomy is identical in nearly all fishes), is necessarily artificial, and has little or no taxononiic value. It ^^■ould involve, for example, placing Jmia alongside Gadus. Baudelot (1883, 16) states that the facial nerve is absent in many bony fishes {e.g. the " Cariae ") and that in these forms it is represented by a branch of the trigeminus. Such a statement must be due to imperfect observation, and if by " Carpe " we may under- stand Cyprinus, a true facial nerve was described by several authors in that form long before the time at which Baudelot was writing. Sagemehl (1881, 172) emphasizes the fact that the two dorsal lateral canal commissures are not homologous, and that a parietal commissure cannot be considered to represent a supra-temporal commissure where in any genus the former is the only one found. Eamsay Wright (1885, 229) considers (p. 491) that in Lepidosteiis a portion of the R. oticus facialis represents the prsse-trematic or prge-s2)iracular branch of the facial, since it supj)lies the anterior sense organ of the spiracular cleft, and states that a similar condition is found in Amia. It is impossible that the two nerves mentioned can be homologous, since the otic is a true lateral line nerve and cannot, therefore, be compared with a nerve consisting of visceral sensory fibres. Wright, therefore, failed to find a prae-spiracular nerve in Lepidosteus and Amia. * Confirmed independently in 1881 by Marshall and Spencer, and in 1889 by Allis (4, pp. 513, 514). 200 ME. F. J. COLE ON THE STRUCTUKE AND MORPHOLOGY OF Whilst on the subject of the facial nerve it is necessary that AUis's views on tliis nerve in Amia should be discussed. He says (1897, 6, pp. 615, 616) : " After issuing from that [=facial] canal the truncus facialis continues dowiiward, outward, and backward, lying in the depression on the outer surface of the hyoniandibular, and separates almost immediately into its two main j)ortious, the truncus mandibularis facialis and the 11. hyoideus facialis. The former lies anterior to the latter, turns downward, and soon separates into its two portions, the 11. mandibularis exteruus facialis and the R. mandibu- laris internus facialis." On p. 618 the "internal mandibular" is described as entering into relations with the lower jaw. Before going further and discussing the homology of the latter nerve with the chorda tynipani, it is necessary to point out that Allis has misnamed his nerves. As I have already pointed out (18S)6, 46, p. 657 et neq.), the terms internal mandibular and Jiijoidetis as first used by Stannius are not only synonymous hut ap2)ly to a motor post-spiracular nerve related to the anterior face of the liyoid arch, just as the pre-spiracvilar nerve should be related to the posterior face of the mandibular arch. On this ground alone (and there are others) Allis's "iuternal mandibular" must be a morphological pre-spiracular nerve, and as sucli is wrongly iiamed by him. As the whole question, however, involves also the question of the homology of the chorda tympani, it is first necessary to show what Allis's views are on the latter question. He says (6, pp. 638, 639): "As the nerve [i. e. the "iuternal mandibular"!, in Amia, lies behind the spiracular canal, it is a post-trematic branch of the facitilis, and cannot, therefore, be the chorda tympani, for the course of that nerve in man through the upper portion of the tympanic cavity and then downward anterior to that cavity certainly indicates that it is a pre-spiracular nerve. That this nerve, in Amia, is the homologue of the nerve of the same name described by Ewart, Pollard, and Strong, in other Ichthyopsida, and considered by them as the homologue of the chorda tympani, is hardly open to qitestion. The nerve in Amia is probably to be compared to the branch which, on each side of the branchial arches, runs downward over the anterior face of the arch on to the inner surface of its ventral portion. Its position in Amia, along the inner surface of the mandible, could be easily derived from that in Selachians as given by Vetter. In Heptanchus what seems to be the nerve is shown lying along the posterior edge of the mandible ; tVom this position, as tlie hyoideo-mandibular fold of Amia was formed, the nerve could as naturally come to lie along the inner sui-face of the mandible as along the lateral surface of the hyoid." An examination of this passage reveals several flaws. It is true that the chorda tympani is a pre-spiracular nerve,'as I have previously shown (1896, 46), but is Allis's " iuternal mandibular " a pos^-spiracular nerve ? We have seen that Ramsay Wright failed to identify the homologue of the pre-spiracular in Lepidostens and Amia, and Allis himself (18S9, 4, \)\). 501 et seq.) says nothing about it. The fact is, nothing is known of the relation of the "internal mandibular" nerve of Amia to the spiracular cleft, since the embryonic condition hns not yet been worked out, and the cleft itself degenerates in the adult. We are, therefore, not in a position to say whether this nerve is pre- or post-spiracular. Further, what is the definition of a pre-spiracular nerve ? It should fulfil three conditions : (1) it should rim in frcmt of the spiracle : (2) it should pass akaig the posterior face of the mandibular arch ; and (3) it should consist of visceral sensory fibres. With regard to the first THE CRANIAL NERVES AND LATERAL SENSE ORGANS OF FISHES. 201 condition, there are several cases on record where on the disappearance of the spiracle the pre-spiracular accompanies for a time the post-spiracular nerve and thus becomes a topographical, but not a morphological, post-spiracular nerve. It seems to me that when the early development of the nerves of Amia has been investigated it will be found that the " internal mandibu.lar " nerve is morphologically pre-sj)iracular, though occupying a post-spiracular position in the adult. This is what we know has happened in Raiia (cp. Strong's " iuternal mandibular "), and what has douljtless also happened in Chliiiceni, and Gadtis. The degeneration of the spiracular cleft naturally leaves the nerves unaffected, since they ai'C concerned rather with the sensory and motor supjDly of the arches, and tli*; latter are untouched by the disaj^pearance of the spiracle. With regard to the second condition above, that is practically i'ulfilled, and so doubtless also is the third. The "internal mandibular " nerve is certainly not a motor nerve in Amia, in which aise it must be a visceral sensory nerve, and indeed AUis says tliat it " is distributed to the inner surface of the liyoid and mandibular arches" (6, p. 745). The "internal mandibular" nerve of Amia therefore is, as far as our knowledge of it goes, a morphological pre- spiracular nerve, and in any case is excluded from l)eing the true internal mandibular nerve, since this should be distributed to the muscles of the hyoid arch, and is undoubtedly represented in Amia by the E-r. hyoideus and opercularis. In further criticism of the passage quoted above I may remark that it is very much " open to question " that the internal mandibular of Amia corresponds to the nerve of the same name described by Ewart. The nerve of the latter author is a motor nerve partly to the muscles of the hyoid arch, and as such differs essentially from the nerve in. Amia. The nerves described b}' Pollard *, Gaupp (1893, 84), and Strong, are correctly homologised, Init should not have been named "internal mandibular," as Miss Piatt (1890, 158, p. 534) and the writer have pointed out. Finally, an anterior brancliial ramus is related to the posterior and not to the anterior face of its arch, as Allis himself mentions is the cise with the " internal mandibular " of Ilcptaiichus. Pinkus (1891, 157) introduces further confusion into the synonymy of the facial nerve of fishes. He correctly homologises his " inferior palatine VII " with the chorda tympani (but did not I'ecognize that this nerve represented the pre-spiracular nerve of other fishes), and divides the morphological post-spiracular nerve ( = , + the lateral line element, the hyomandilnilar trunk) into three parts— (1) internal mandibular ; (2) hyoideus f ; and (3) motor Vllth. Pinkus is of course wrong in using two synonyms (i. e. internal mandibular and hyoideus) to describe two different branches, and it seems to me that 1 and 3 together represent the internal mandibular of other fishes. The motor Vllth branches correspond undoubtedly to the ramus oper- cvilaris facialis of the bony fishes. There is hence some confusion in the terminology of the facial nerve of fishes, which it is desira])le should be removed. I have therefore drawn up a scheme of the constitution of the facial or Vllth cranial nerve in a typical fish, with the synonyms of the three cardinal branches. It is to be hoped that future authors will, before naming a nerve, first ascertain whether it is somatic or splanchnic, and then whether it is sensory or ■^ * See 1S92. 160, pp. '.VJl and :i9b (table). It will fie seen that Pollard's nerve is the one I have id(>ntified as the chorda tympani in Gadtis. f As this is a sensory nerve it must he unrepresented in other fishes. 202 MR. F. J. COLE ON THE STRUCTURE AND MORPHOLOGY OF motor. I may mention here that there is a nerve in Gad us exactly corresponding to the " internal mandibular " of Allis in Amia. I have for the above reasons described it as the pre-spiracular or chorda tympani division of the facial. The following scheme of the constitution of the facial nerve in a typical fish takes no account of the lateral line constituent, which only accompanies the facial and is not a branch of it : — Eacial nerve sensu stricto (minus external mandibular lateral line nerve). Forks over spiracular cleft into — (1) Palathnis or viscer alls facialis. Is joined by Jacobson's anastomosis (^palatinus or visceralis glossopharyngei), and is a visceral sensory nerve distributed to the mucous membrane of the roof of the mouth in front of the spiracle. (2) A visceral sensory nerve arising from the base of (1) and forming part of the same bundle of fibres. FrcB-spiracular, frce-hranchial, prcB-trcniatic, cr chorda tjjnipani division of facial. May be continued ventrally on to pharynx, in which case it should become related to the posterior face of the mandibular arch, and be distributed to the mucous membrane of this arch. (3) A visceral motor nerve = main trunk of facial. Forms external to spiracle a portion of the nerve known as the truncus hyomandibularis facialis, of which the remainder and larger part here=the external mandibular lateral line nerve. JPost-spiracular, post-brancliial, post-tremafic, internal mandibular^ or hyoidean branch of the facial. Continued ventrally on to pharynx, where it becomes related to the anterior face of the hyoid arch, the muscles of which absorb most of its fibres (Allis, 6, p. 745). Gives off posteriorly the nerve known as the Ramus operctUaris facialis to the muscles of the hyoid arch (Allis, loc. cit.). Allis (4, p. 472), in seeking to prove the independence of the supra-orbital canal (which is of course independent as far as innervation proves anything), goes somewhat too far in advancing the developmental independence of the supra-orbital canal from the infra- orbital as proving this point. As the lateral canals, according to his own description, develop in indejiendent segments, each containing a sense organ, his argument proves nothing, especially as there is a dermal tubule at the point of anastomosis between the two canals which involves their independence as a developmental necessity. It will have been noticed that I have not described an ophthalmicus profundus nerve in Gadus. The question of the existence of this nerve in Teleosts is fully discussed by Allis (1897, 6, pp. 538 et seq.), and lie concludes : " In no Teleost, w^ith the single recorded exception, so far as I can find, of Trigla, is there a sej)arate profundus ganglion and root. Both ganglion and root are apparently always completely fused with the ganglion and root of the trigeminus." He further points out that the nerve identified by Pollard as the profundus in certain Siluroids is really the ophthalmicus superficialis VII, and, I think,, rightly doubts any fusion between the profundus and superficialis as explaining the apparent absence of the former, believing it to be " more probable that the ophthalmicus profundus of Elasmobranchs is entirely wanting in Teleosts." Goronowitsch's and other authors' nerves, therefore, named the ophthalmicus profundus, represent the ophthalmicus superficialis V, and his ophthalmicus superficialis is the lateral line nerve of that name THE CRANIAL NEEVES AND LATERAL SENSE ORGANS OF FISHES. 203 (1897, 90) *. In his later work (1897, 6, p. 625) Allis states, what I had previously- suggested, that the so-called lateral line branch of the IXth of Amia is really a branch of the lateralis lateral line nerve. A rather curious mistake is made by Ayers (1892, 7), who describes the lateralis latei-al line nerve as the "anterior (dorsal) branch of the IXth." Pollard makes a statement which 1 had overlooked in my previous work, and which is closely paralleled by the condition sometimes found in Chimcera {46, p. 616). He says (1892, 160, p. 395) " In a young specimen of Poli/ptents about 20 cm. long, which was cut into sections' there appeared to be a commissure from the ganglion of the trigeminus to the ophthalmicus superficialis of the facial. The material was, however, not well preserved, and consequently the existence of this commissure -which would represent the superficial portion of the trigeminus could not be determined with absolute certainty." The buccal and superficial oplithalmic lateral line nerves of Polypterus arise from a common trunk, and the former accompanies the superior maxillary division of the Vth. Miss Piatt (1896, 158) describes the IXth nerve of Necturiis innervating sense organs of the lateral line, and considers the operculo-mandibular line of organs to be composed of two distinct portions. Kingsbury (1897, 114) states that the lobi trigemini of Elasmobranchs and Ganoids are directly homologous and constitute the lateral line centre, but that in Teleosts tiie lobus trigeminus is another structure and belongs to the fasciculus communis system. I had originally intended writing a section on the function of the lateral line system, but this must be deferred for the present. In the meantime I may refer those interested in the question to the works of Ayers (1892, 7), Dercum (1880, 59), Eritsch (1888,76, and 1890, 77), Fuchs (1895, 78), de Sede (1885, 129), Ilichard (1896, 167), Savi (184.4, 178), Stahr (1897, 196), (Willey 1891, 223), and Ramsay Wright (1884, 227). Summary. 1 . The present communication contains the first description of the lateral sense organs and their nerves in the genus Gadus. 2. The sensory canal system of Fishes, both recent and fossil, is resolvable into a common type, to which the lateral canals of all forms can be reduced. This type includes a lateral canal on the body, a canal over the eye, another underneath it, and finally one related to the lower jaw. All these canals may be connected by median commissures with their fellows of the opposite side. 3. The sensory canals of Gadus may be described as having diverged but slightly from the typical form. The canal sense organs are somewliat reduced in number, which involves a similar reduction in the dermal tubules. There are no Savian vesicles or Lorenzini's ampuUse, but pit organs of the usual constitution are fairly common. 4. The lateral line systems ofGadusmorrhua and Gadus vii^ens exhibit specific differences. 5. The sensory canals, by their skeletal support, become secondarily related to the * He says (p. 30) : " Der N. ophth. superf. ist wie bei Knochenfisclieu ein Ast des Trigeminus II. [=lateral line component]. Der N. ojihthahnicus jirofinidus von Acipenser ist aiisschlie^slich cin Ast des Trigeminus I. tiiid ist dcninach kein uoUl-ommeiics ffomoloijoii des (flcichiiamii/en Nervcii voa Lota, wcleher FAemcnte des Facialis ciUhdJi. Der R. oticus ist boi Acipenser oin Ast des Facialis und des Trigeminus II. und kanu demnach aucb iiicht mit dem N. ophthalmicus prof, der Knochentischc direkt vergiichen wcrden." SECOND SEUIES. — ZOOLOGY, VOL. VII. 28 204 ME. F. J. COLE ON THE STEUCTUEE AND MOEPHOLOGY OF skeleton of the head, and on their disappearance left a legacy to the skull in the shape of certain of the lateral line ossicles, which either on account of additions to their original function were retained as new elements, or were fused on to, and came to form a part of, hones previously existing. 6. The ganglia of the trigeminal and facial nerves have fused to form an elaborate complex, in which the individuality of the ganglia has been lost, and which arises from the brain by two roots. The ganglion in connection with the lateral line nerve is, however, distinct. The sympathetic trunk is connected with the vagu.s, glossopharyngeal, and trigemino-facial ganglia. 7. A portion of the trigemino-facial complex which is connected with visceral sensory fibres, and which I have called the facial ganglion, is in such a condition as to strongly suggest the view that it is migrating from the complex and is in process of being converted into a sympathetic ganglion. It is in connection with the sympathetic trunk, and has been described by many authors as a symjiathetic ganglion. The condition of this ganglion favours the view that at least the greater part of the facial, glossopharyngeal, and vagus nerves of Fishes belongs to the sympathetic system. 8. A typical branchial nerve has the constitution tabulated on p. 144. 9. The facial nerve of Man and Mammals can be derived from the branchial facial nerve of a fish. 10. The anastomosis of Jacobsou is the visceral sensory palatine or visceral branch of the IXth, and connects the glossopharyngeal ganglion with the ramus palatiuus facialis (also visceral sensory). It is neither a somatic sensory nor a lateral line anastomosis, and cannot be homologiscd with either of these. 11. The lateral line system of Fishes is, as far as our present knowledge goes, not metameric. This conclusion is supported by the whole of the anatomical and part of the embryological evidence. The observations under the latter head that have hitherto been considered to support the metameric view of the lateral organs are certainly in part, if not entirely, concerned with another series of sense organs, and do not relate to the lateral sense organs at all. 12. The nerves supplying the lateral sense organs (both superficial and canal) form a separate series in themselves, and are not connected with any of the other cranial nerves , They may be named the Superficial Ophthalmic, Buccal, External Mandibular, and Lateralis Lateral line Nerves. The innervation of the lateral sense organs of Fishes and Amphibia is remarkably constant, and should be taken as the guide in determining the limits of a canal. The sensory canals and pit organs of Gadiis are innervated by the branches of these nerves — the glossopharyngeus taking no part in the innervation of the system either in Gaclus or morphologically in any other form. 13. The system of nerves known as the ramus lateralis trigemini is a ganglionated system of somatic sensory nerves formed typically of the dorsal branches of the Vth, Vllth, IXth, and Xth cranial nerves, with a certain number of the same branches of the spinal nerves, all more or less fused together, and forming the sensory nerve supply of all the fins of the body. Its common name, therefore, should be abandoned in favour of its oriu'inal name of ramus lateralis accessorius. I THE CRANIAL NERVES AND LATERAL SENSE ORGANS OF FISHES. 205 14. The so-called lateralis nerve of Fetromyzon is not a lateral line nerve at all, but its anatomy sliows it to belong vindoubtedly to the lateralis acccssorius system. 15. The lateral line system of Gadus is in remarkable agreement with that of Ainia, except that the latter form exhibits less reduction in the number of its sense organs. 16. The sense organs of the lateral line of Pishes and Amphibia are not lioiuologous ■with the lateral sense organs of Annelids {e. g. Ca})itellidte). 17. The sensory canals were probably represented in the ancestral Vertebrate by non- segmental sujierficial sense organs. These sank below the sui-face, forming a series of tubes, by the end-to-eud fusion of which the sensory canals were formed. As the sense organs sank below the surface, certain lateral glandular organs were also included which secreted the mucus tilling the canals. These glands have been subsequently much reduced in importance, and their function now is an accessory one, l. e. to secrete a substance corresponding precisely to the endolympli of the ear. Before the sensory canal system extended forwards and backwards it was probably confined to the region now occupied by the auditory organ. 18. There can l)e little doubt that the primitive lateral line organs gave rise both to the recent lateral system and also to the vertebrate auditory organ. It is doubtful whether the semicircular canals have arisen by a concentration of primitive sensory canals or by the decentral development of a primitive simple sac. The author holds to the latter view, and believes that the semicircular <*.anals are homoplastic and not homologous with tiie sensory canals. 19. The author's previous view that the chorda tympani nerve of Mammals has been derived from the prae-spiracular division of the facial of the Piscine ancestor of the Vertebrates is confirmed. The nerve referred to as the " internal mandibular " by Allis and some other authors is wrongly identified, and is really the prae-branchial or chorda tympani division of the facial. 20. A typical facial nerve of a fish has the constitution tabulated on p. 202. Zoological Department, University College, Liverpool. March 2ud, 1898. Postscri2)t. — Since writing Section H, I have examined a large number of Gad/is skulls, and have no doubt now that the description does not represent the normal condition, which in the vast majority of the specimens examined was that of complete fusion between the dermal and cartilage pterotic elements, although the point of fusion could always be easily detected. The nomenclature of these two l)ones is lience in urgent need of revision. The terms post-frontal and squamosal should be reserved for the separate dermal lateral line elements, and sphenotic and pterotic for the separate cartilage auditory bones. A comj)ound term is also necessary for the condition in which the lateral and cartilage elements have secondarily fused to form one bone. Where the lateral line bones are never known to be separate from the underlying cranial bones in the adult there is no pressing need in the meantime for any revision of the nomenclature. — Sept. 27, 1898. 28* 206 MR. r. J. COLE ON THE STRUCTURE AND MORPHOLOGY OF Preliminary Bibliography *. 1. Ahlborn, F. — " Untersuchuiigen iiber das Gehirii der Petromyzonten," Zeits. f. wissen. Zool., Bd. xxxix. pp. 191-294., Taf. xiii.-xvii., 1883. Theil I. published as an inaugural dissertation, pp. 1-47, Leipzig, 1883. 2. Ahlborn, F. — " IJbor den Ursprung und Austritt der Hirnnerven von Petromyzon," u. " tJber die Segmentation des Wirbelthierkilrpers," Zeits. f. wissen. Zool., Bd. xl. pp. 286-330, Taf. xviii., 1884. 3. Alcock, B.— " The Fifth Pair of Nerves," in Todd's ' Cyclopaedia of Anatomy and Physiology,' vol. ii. pp. 268-310, London, 1839. 4. Allis, E. P., Jr. — " Tlie Anatomy and Development of the Lateral line System in Amia calva," Jour, of Morphology, vol. ii. pp. 463-566, pi. xxx.-xlii., 1889. 5. Allis, E. P., Jr. — " The Cranial Muscles and Cranial and First Spinal Nerves in Amia calva," Jour. of Morphology, vol. xi. pp. 485-491, 1895. Preliminary to 6. 6. Allis, E. P., Jr. — " The Cranial Muscles and Cranial and First Spinal Nerves iu Amia calva," Jour. of Morphology, vol. xii. pp. 487-808, pi. xx.-xxxviii., 1897. 7. Ayers, H. — " Vertebrate Cephalogenesis. II. A Contribution to the Morphology of the Vertebrate Ear, with a Reconsideration of its Functions," Jour, of Morphology, vol. vi. pp. 1-360, pi. i.-xii. 1892. 8. Baer, K. E. von. — " Uebcr den Seiteukanal des Stiirs," Meckel's Archiv f. Anat. u. Physiol. T. ix., pp. 376 and 377, Jahr. 1826. 9. Balfour, F. M. — " A Treatise on Comparative Embryology." Vol. ii. Vcrtebrata, pp. 443-446, London, 1881. 10. Bateson, W. — " The Sense Organs and Perceptions of Fishes ; with Remarks on the Supply of Bait," Jour. Marine Biol. Assoc. United Kingdom, vol. i., n. s. pp. 225-256, pi. xx., 1889-90. 1 1 . Baudelot, E. — " De la Determination homologique d'une Branche du Nerf pathetique chez le Merlan [Gaelics merlangus]," Bull, de la Soc. des Scien. Nat. de Strasbourg, Ann. i. pp. 81-83, 1868. * Tho above list simply oontains those papers consulted by the author in writing the present work. A complete Bibliography of the anatomical, enibryological, physiological, recent and fossil literature, directly or indirectly treating of the Brain, Cranial Nerves, and Lateral Sense Organs of Fishes and Amphibia, is iu active preparation, and will be publislied separately. Nevertheless, although the present list has already been more than doubled, the author would be extremely grateful for any references not included in it, either to the sender's own papers, or to any others which should have been recorded. The objects of the forthcoming Bibliography (by no moans a mere list) are : (1) to provide an absolutely complete and accurate catalogue of all the works ever published more or less relating to the subject; (2) to be a guide to an author (by means of abstracts, indexing, &c.) newly taking up the subject to the literature immediately affecting his sjtecial purpose, and also to those general works it were expedient that he should consult ; (3) to supply the pressing need of an Index Expur(jatorius, With regard to the latter, and perhaps most important, object, it is obvious that an author, however conscientious, cannot be expected (unless, indeed, he is writing an extensive monograph) to wade through at least 700 papers before writing his own. It is equally obvious that many of these papers, either on account of the antiquity of their contents, or other equally potent but less excusable reasons, may very well be consigned to the " silent tomb," and piously allowed to remain there. The want of such an analytical Bibliography has, even in the confined experience of the author, over and over again imjjaired an otherwise useful piece of work. It is respectfully hoped, tliercfore, that a diligent and unprejudiced search through the literature will have the general effect of clearing the air, and thus proving of some service to those subsequently taking up the work. THE CRANIAL NERVES AND LATEEAL SENSE ORGANS OF FISHES. 207 12. Baudelot, E. — " Considerations sur le Tronc Lateral du Pneumogastrique chez les Poissons," ibid., pp. 109-112, 1868. 13. Baudelot, E. — "Observation relative :i une Branche anastomotique des Nerfs trijumeau et pneumo- gastrique cliez le Merlan [Gadus merlangus]," ibid., pp. 114-116, 1868. 14. Baudelot, E. — " Observation sur les Origines de la Branche operculaire du Nerf lateral du Pneumogastrique cliez quelques Poissons," ibid., Ann. ii. pp. 22-24, 1869. 15. Baudelot, E. — " Etude sur I'Anatomie comparee de I'Encepbale des Poissons," Mem. de la Soc. des Sci. Nat. de Strasbourg, t. vi. livr. ii. pp. 51-128, plan, i., 1870. 16. Baudelot, E. — ' Recherches sur le Systeme nerveux des Poissons,' pp. 1-178, pi. i.-x., Pans, 1883. Reviewed by E. Dubrueil in Revue d. Sci. Nat. (Montpellier), ser. iii. t. iii. pp. 633-638, 1883-84. 17. Beard, J. — "On the Segmental Sense Organs of the Lateral Line, and on the Morphology of the Vertebrate Auditory Organ," Zr)ol. Anz., Jahr. vii. pp. 123-126, 140-143, 1884. Preliminary to 19. 18. Beard, J. — "On the Cranial Ganglia and Segmental Sense Organs of Fishes," Zool. Anz. Jahr., viii. pp. 220-223, 1885. Preliminaiy to 19. 19. Beard, J. — "The System of Branchial Sense Organs and their associated Ganglia in Ichthyopsida. A Contribution to the Ancestral History of Vertebrates," Q. J. M. S. n. s. vol. xxvi. pp. 95-156, pi. viii.-x., 1885-6. 20. Beard, J. — "The Ciliary or jNIotor-oculi Ganglion and the Ganglion of the Ophthalmicus Profundus in Sharks," Anat. Anz. Jahr. ii. pp. 565-575, 1887. 21. Beard, J. — "Morphological Studies. II. The Development of the Peripheral Nervous System of Vertebrates. Part 1. Elasmobranehii and Aves," Q. J. M. S., n. s. vol. xxix. pp 153-227, pi. xvi.-xxi., 1888. 22. Beck, B. — " Anatomische Untersuchungen iiber einzelne Theile des VII. und IX. Hirnnerven- paares," pp. 1-69, Taf. i.-iii., Heidelberg, 1847. 23. Bendz, H. — "Bidrag til den sammenlignende Anatomic af Nervus glossopharyngeus, vagus, accessorius Willisii og Hypoglossus hos Reptilierne," Det K. Danske Videns. Selsk. Naturv. og Math. Afhand. Deel x. pp. 113-152, Tab. i.-x., 1843. 24. BoDENSTEiN, E. — " Dcr Seiteukanal von Cotitts gobio," Zeits. f. wissen. Zool., Bd. xxxvii. pp. 121- 145, Taf. X., 1882. See criticism by B. Solger, " Bemerkung iiber die Seitenorganketten der Fiscbe," Zool. Anz., Jahr. v. pp. 660-661, 1882. 25. Boll, F. — " Die Lorenziui'scheu Ampullen der Selachier," Archiv f. mikros. Anat., Bd. iv. pp. 375- 391, Taf. xxiii., 1868. 26. Boll, F. — " Uber die Savi'schen Bliischen von Torpedo," Monats. der Konig. Preuss. Akad. der Wissens. zu Berlin, Jahr. 1875, pp. 238-241, 1876. Preliminary to 27 and 28. 27. Boll, F. — " Le Vescicole di Savi della Torpedine," Atti della R. Accad. dei Lincei, ser. ii. vol. ii. pp. 385-392, tav. i., 1875. See German translation below (28). 28. Boll, F. — " Die Savi'schen Blaschen von Torpedo," Archiv f. Anat. Phys., u. wissen. Medicin, pp. 456-468, Taf. xi., 1875. Identical with the Italian memoir (27). 29. Bonnier, P. — See J. Richard (167). 30. BoNSDORFF, E. J. — " Disquisitio Anatomica, Nervum trigeminum partemque cephalinam Nervi sympathici Gadi Lotae, Linn., cum nervis iisdem apud Hominem et Mammalia comparans," pp. 1-52, tab. i., Helsingforsise, 1846. 208 ME. F. J. COLE ON THE STKTJCTUEE AND MORPHOLOGY OF 31. BoNSDOKFF, E. J. — "Symbolic ad Anatomiam compai-atam Nervorum Animalium vertebratorum/J Nervi cerebrales Coni cornicis (Linn.)," Acta Soc. Sci. Fennicic, t. iii. pp. 505-569, pi. vi.-vii., 1852. 32. BoNSDORFF, E. J. — Ibid., " 11. Ncrvi cerebrales Grids cinerece, Linn.," ibid., pp. 591-624, pi. x., 1852. 33. BoNSDOEFF, E. J. — "Jemfcirande anatomisk Beskrifning af Cerebral Nerverna hos Raja clavata" Acta Soc. Sci. Feunicse, t. v. pp. 1X5-227, tab. ii.-iv., 1856. 34. Bridge, T. W.- — " On the Morphology of the Skull in the Paraguayan Lepidosiren and in other Dipnoids," Trans. Zool. Soc. London, vol. xiv. part v. pp. 325-376, pi. xxviii.-xxix., 1898. 35. Brinton, W. — " The Seventh Pair of Nerves," in Todd's ' Cyclopaedia of Anatomy and Physiology/ vol. iv. pp. 543-556, Loudon, 1852. 36. BtJCHNER, G. — " Meoioire sur le Systeme nerveux du Barbeau (Cyprinus barbus, L.)," Mem. de la Soc. du j\[us. d'Hist. nat. de Strasbourg, t. ii. pp. 1-57, plan i., 1835. 37. Bunker, F. S. — " On the Structure of the Sensory Organs of the Lateral Line of Ameiurus nebulosus, Le Sueur," Anat. Anz., Bd. xiii. pp. 256-260, 1897. 38. BuRCKHARDT, li. — " Das Ceutralncrvensystem von Protopterus annectens. Einc vergleichend- anatomische Studie," pp. 1-64, Taf. i.-v., Berlin, 1892. 39. BuRCKHARDT, E. — " Beitrag zur Morphologic des Kleiuhirns dcr Fische," Archiv f. Anat. u. Phys., Anat. Abth., Supplementband, pp. 111-136, Taf. vii., 1897. 40. Chevrel, R. — " Sur le Systeme nerveux grand sympatbique des Poissons osseux/' Comp. Reud. Acad, des Scien. Paris, t. cvii. pp. 530-531, 1888. Preliminary to 41. 41. Chevrel, E. — "Sur TAnatomic du Systeme nerveux grand sympatbique des Elasmobranches et des Poissons osseux," Arch, de Zool. Exper. et Gen., ser. ii. t. v. bis, Mem. v. pp. 1-196, pi. i.-vi., 1887-90. Published as separate volume, Poitiers, 1889, and Paris, 1891. See abstract. Rev. Scientif. ser. iii. t. xvi. ( = t. xlii.), p. 409, 1888. 42. CoGGi, A. — " Le Vescicole di Savi egli Orgaiii della Linea lateraie nellc Torpedini," Atti d. E. Accad. dei Lincei, Eendiconti, vol. vii. ser. iv. pp. 197-205, 1891. See French translation below (43). Reviewed by S. Gai'man, ' Science,' vol. xix. p. 128, 1892. 43. CoGGi, A. — " Les Vesicules de Savi et les Organes de la Ligne lateraie cbez les Torpilles," Arch. Italiennes de Biologic, t. xvi. pp. 216-224, plan, i., 1891. Identical with the Italian memoir (42). 44. Cole, F. J.— "The Cranial Nerves of Chimmra monstrosa," Proc. Roy. Soc. Edinburgli, vol. xxi. pp. 49-56, 1896. Preliminary to 46. 45. Cole, F. J. — " On the Sensory and Ampullary Canals of Chimara," Anat. Anz., Bd. xii. pp 172- 182, 1896. Criticism of 54. 46. Cole, F. J. — " On the Cranial Nerves of ChimcBra monstrosa (Linn., 1754) ; with a Discussion of the Lateral line System, and of the Morphology of the Chorda tympani,'' Trans. Roy. Soc. Edinburgh, vol. xxxviii. pp. 631-680, pi. i.-ii., 1896. 46 a. Cole, F. J. — " Reflections on the Cranial Nerves and Sense Organs of Fishes," Trans. Liverpool Biol. Soc., vol. xii. pp. 228-247, 1898. 47. CoLLiNGE, W. E. — "Note on the Lateral Canal System of Polypterus," Proc. of the Birmingham Philos. Soc, vol. viii. pp. 255-263, pi. i.-iii., 1891-93. 48. CoLLiNGE, W. E. — " The Lateral Canal System of Ltpidusteus osseus," ibid., pp. 263-272, pi. iv.-ix., 1891-93. 49. CoLLiNGE, W. E. — " The Sensory Canal System of Fishes. Part I. — Ganoidei," Quart. Journ. Micros. Soc, n. s. vol. xxxvi. pp. 499-537, pi. xxxix.-xl., 1894. THE CRANIAL NERVES AND LATERAL SENSE ORGANS OF FISHES. 20i) 50. CoLLiNGE, W. E. — " On the Sensory Canal System of Fishes/' Report British Assoc. Nottingham, pp. 810-811, 1894. 51. CoLLiNGE, W. E. — '•■ The Morphology of the Sensory Canal System in some Fossil Fishes," Proc. of the Birmingham Nat. Hist, and Philos. Soc. vol. ix. pp. 58-70, pi. i.-ii., 1894-95. 52. CoLLiNGE, W. E. — " Some Researches upon the Sensory Canal System of Ganoids," ibid., pp. 77-83, 1894-95. Abstract of work previously published. 53. CoLLiNGE, W. E. — " On the Sensory Canal System of Fishes. Teleostei — Suborder A, Physo.itomi," Proc. Zool. Soc. London, pp. '271—299, pi. xviii.-xxi., year 1895. 54. CoLLiNGE, W. E. — "On the Sensory and Ampullary Canals of C/dmcera," Fvoe. Zool. Soc. London, pp. 878-890, pi. li.-liii., year 1895. 55. Cunningham, J. T. — "A Treatise on the Common Sole (Solea vw/jram), considered both as an Organism and as a Commodity," pp. 1-147, pi. i.-xviii., Plymouth, 1890. 56. CuviER, G.— ' Lecons d'Anatomie comparee,' t. ii. pp. 166-172, 193-242, Paris, 1799. 57. CuviER, G., & V.iLENCiENNEs, A. — '• Illstoire naturelle des Poissons," t. i. pp. 415-445, plan, iv.-vi., 1828 ; also t. v. plan, cxl., Paris, 1830. 58. Dean, Bashforu. — ' Fishes, Living and Fossil. An outline of their Forms and probable Relation- ships,' Nevi' York and London, 1895. 59. Dercum, F. — "The Lateral Sensory Apparatus of Fishes," Proc. of the Acad, of Nat. Sei. of Philadelphia, year 1879, pp. 152-154, 1880. 60. Desmoulins, a., & Magendie, F. — " Anatomic des Systemes nerveux des Animaux a vertebres, appliquee k la Physiologic et a la Zoologie," part i.-ii.. Atlas, plan, i.-xiii., Paris, 1825. 61. DixoN, A. F. — " On the Development of the Branches of the Fifth Cranial Nerve in Man," Sci. Trans, of the Roy. Dublin Soc, ser. ii. vol. vi. pj). 19-76, pi. i.-ii., 1896. 62. DoHRN, A. — " Studieu zur Urgeschichte des Wirbelthierkorpers. XIIL Uber Nerven und Gefasse bei Ammoca'tes und Petromyzon planeri," Mittheil. a. d. Zoolog. Stat, zu Neapel, Bd. viii, pp. 233- 306, Taf. x.-XT., 1888. See C. JuLiN (i 08-1 12). 63. DuBRUEiL, E. — See E. Baudelot (16). 64. EisiG, H. — " Die Seitenorgane und becherformigen Organc der Capitelliden. Zweiter Auszug aus einer Monograpliie der Capitelliden," Mittheil. a. d. Zoolog. Stat, zu Neapel, Bd. i. pp. 278- 342, Taf. vii., 1878-79. Preliminary (in part) to 65. 65. EisiG, H. — " j\ionographie der Capitelliden des Golfes von Neapel und der angrenzenden Meeres- abschnitte, uebst LIntersuehungen zur vergleiclienden Auatomie und Pliysiologie," Fauna u. Flora d. Golf. v. Neapel, xvi. Monographic, Berlin, 1887. 66. Emery, C. — " Le Specie del Genere Fierasfer nel Golfo di Napoli e Regioni limitrofe," Fauna u. Flora d. Golf. v. Neapel, ii. Monographic, pp. 1-76, tav. i.-ix., Leipzig, 1880. 67. EwART, J. C. — " On the Cranial Nerves of Elasmobrauch Fishes. Preliminary Communication," Proc. Roy. Soc. London, vol. xlv. pp. 524-537, 1888-89. 68. EwART, J. C. — •" The Lateral Sense Organs of Elasmobrunchs. I. The Sensory Canals of Lmmargus," Trans. Roy. Soc. Edinburgh, vol. xxxvii. pp. 59-85, pi. i.-ii., 1892. See abstract, Zool. Anz., Jahr. xv. pp. 116-118, 1892. 69. EwART, J. C, & Mitchell, J. C. — ■" On the Lateral Sense Organs of Elasmobranclis. II. Tlie Sensory Canals of the Common Skate (Raja batis)," Trans. Roy. Soc. Edinburgh, vol. xxxvii. pp. 87-105, pi. iii., 1892. See abstract, Zool. Anz., Jahr. xv. jjp- 118-120, 1892. 70. EwART, J. C. — " Supplementary Note on the Cranial Nerves of Elasmobranchs," in " Papers on Electrical and Lateral Seu'^e Organs and on the Cranial Nerves of Elasmobranchs," No. 9, pp. 1-3, Edinburgh, 1893. 210 ME. F. J. COLE ON THE STRUCTUEE AND MOEPHOLOGY OF 71. EwART, J. C.J Se Cole, F. J. — "On the Dorsal Branches of the Cranial and Spinal Nerves of Elasmobranchs/' Proc. Roy. See. Edinburgh, vol. xx. pp. 4.75-480, 1895. 72. Fee, F. — " Recherclies siir le Systeme lateral du Nerf Pneumo-gastrique des Poissons," Mem. de la Soc. des Sci. Nat. de Strasbourg, t. vi. livr. ii. pp. 129-201, plan, i.-iv., 1870. 73. Friant, a. — " Recherches anatomiques sur les Nerfs Trijumeau et Facial des Poissons Osseux," Bull. d. 1. Soc. des Sci. de Nancy, ser. ii. t. iv. fasc. ix. pp. 1-1U8, plan, i.-vi., 1879. Also published as separate volume, Berger-Levrault et Cie., Nancy, 1879. 74. FiiiTSCH, G. — ' Uutersuchungen iiber den feiueren Bau des Fischgehirns, mit besonderer Beriick- sichtigung der Homologien bei anderen Wirbelthierklassen,' pp. 1-91, i.-xv., Taf. i.-siii., Berlin, 1878. 75. Fritsch, G. — 'Die elektrischen Fische. Erste Abtheilung, Malopterurvs electricus,' pp. 1-90, Taf. i.-xii., Leipzig, Veit & Comp., 1887. 76. Fritsch, G.— " Uber Bau und Bedeutung der Kanalsystemc unter der Haut der Selachier," Sitz. d. Konig. Preuss. Akad. der Wisseu. zu Berlin, Halbbd. i. pp. 273-306, Jahr. 1888. Reprinted and expanded in 77. 77. Fritsch, G. — ' Die elektrischen Fische. Zweite Abtheilung, Die Torpediueen,' pj). 1-1 16, Taf. i.-xx., Leipzig, 1890. 78. FccHS, S. — " Ueber die Function der unter der Haut licgeuden Caualsysteme bei den Selachiern," Pfluger's Archiv f. d. ges. Phys. d. Mens. u. d. Thicre, Bd. lix. pp. 451-478, Taf. vi., 1895. Reviewed by F. Schenck, Zool. Central., Jahr. ii. i)p. 120-121, 1895. 79. FiJRBiNGER, M. — "Ueber die spino-occipitalen Nerven der Selaehier und Holocephalen und ihre vergleicheude Morphologic," in Festschrift liir Gegenbaur, Bd. iii. pp. 349-788, Taf. i.-viii., Leipzig, 1897. 80. Garman, S. — " On the Lateral Canal System of the Selachia and Holoccphala," Bull. Mus. Comp. Zool. Harvard, vol. xvii. pp. 57-119, pi. i.-liii., 1888-89. 81. Garman, S. — See A. Coggi (42). 82. Gaskell, W. H. — "On the Structure, Distribution, and Function of the Nerves which innervate the Visceral and Vascular Systems," Jour. Pliys., vol. vii. pp. 1-80, pis. i.-iv., 1886. 8'?. Gaskell, W. H. — "On the Relation between the Structure, Function, Distribution, and Origin of the Cranial Nerves ; together with a Theory of the Origin of the Nervous System of Vertebrata," Jour. Phys., vol. x. pp. 153-211, pis. xvi.-xx., 1889. 84. Gaupp, E. — " Beitriige zur Morphologie des Schiidels. — I. Primordial-Cranium und Kieferbogen von Ra/ia fusca," Schsralbe's Morph. Arbeiten, Bd. ii. pp. 275-481, Taf. xiii.-xvi., Jena, 1893. 85. Gegenbaur, C. — " Ueber die Kopfnerven von Hexanchus und ihr Vcrhaltniss zur ' Wirbeltheorie ' des Schiidels," Jen. Zeits. f. Med. u. Natur. Bd. vi. pp. 497-559, Taf. xiii., 1871. 86. Gehuchten a. van. — " Contribution a I'Etude du Systeme uerveuxdes Tel^osteens. Communication preliminaire," La Cellule, t. 10, pp. 253-295, plan, i.-iii., 1894. 87. GiLTAY, C. M. — " tiuseritur Descriptio ueurologica Esocis Lucii, figuris illustrata et cum reliquis vertebratis animalibus, prsesertim quod ad cerebrum et uervos encephali attinet, comparata," Comment, ad Qusest. Zool., ab ord. discip. Math, et Phys. in Acad. Lugduno-Batava, pp. 1-59, tab. i.-v., Leyden, 1832. 88. GiLTAY, C. M. — " Dissertatio anatomico-physiologica inauguralis de Nervo sympathico," pp. 1-165, tab. i.-ii., Lugduni Batavorum, 1834. 8q. Goronowitsch, N. — " Das Gehirn und die Cranialnerven von Acipenser ruthenus," Morphol. Jahrb., Bd. xiii. pp. 427-454, Taf. xvii.-xxiii., 1887-88. 90. Goronowitsch, N. — " Der Trigemino-facialis-Koniplex von Lota vulgaris," in Festschrift fiir Gegenbaur, Bd. iii. pp. 1-44, Taf. i.-ii., Leipzig, 1897. THE CEANIAL NERVES AND LATERAL SENSE ORGANS OF FISHES. 211 91. GuiTEL, F. — " Sur le Systeme de la Ligne laterale des Lepadogasters/' Comp. Rend. Acad, des Scien. Paris, t. cv. pp. 687-690, 1887. 92. GuiTEL, F. — " Sur les Canaux muqueux des Cyclopterides," ibid. t. cix. pp. 648-651, 1889. 93. GuiTEL, F. — " Sur la Ligue laterale de la Baudroie [Lophius piscatorius, h.)," ibid. t. ex. pp. 50-52, 1890. Preliminary to 95. 94. GuiTEL, F. — " Sur le Nerf lateral des Cyelopterides," ibid. t. cxi. pp. 536-539, 1890. 95. GuiTEL, F. — " Recherches sur la Ligne laterale de la Baudroie {Lophius piscatorius)," Archiv. Zool. Exper., 2 ser. t. ix. pp. 125-190, plan, vi.-viii., 1891. 96. Haller, B. — " Uber das Centralnervensystem, iusbesoudere iiber das Riickenmark von Orlhayo- riscus mola," Morpholog. Jalirbuch, Bd. xvii. pp. 198-270, Taf. xiii.-xv., 1891. 97. Haller, B. — " Der Ursprung der Vagusgruppe bei den Teleostiern," in Festschrift fiir Gegen- baur, Bd. iii. pp. 45-101, Taf. i.-iv., Leipzig, 1897. 98. Herrick, C. J. — "The Cranial Nerve Components of Teleosts," Anat. Auz., Bd. xiii. pp. 425-431, 1897. 99. Herkick, C. L. — "Additional Notes on the Teleost Brain," Anat. Anz., Jahr. vii. pp. 422-431, 1892. 100. Hoffmann, C. E. E. — ' Beitrage zu Anatomic uud Physiologic des Nervus vagus bei Fischen,' pp. 1-34, Taf. i., Giessen, 1860. loi. Hoffmann, C. K. — " Klassen und Orduungen der Amphibieu," in H. G. Bronn's 'Die Klassen und Ordnungeu des Thier-rcichs.' Wirbelthierc, Bd. vi. Abth. ii. pp. 190-227, 367-377, Taf. xv.-xvii., xxi.-xxii., xxix., Leipzig u. Heidelberg, 1873-78. 102. Hoffmann, C. K. — "Zur Ontogenie der Knochenfischc," Archiv f. mikroskop. Anat., Bd. xxiii. pp. 45-108, Taf. iv.-vi., 1884. 103. HoussAY, F. — " Sur la Signification metamerique des organes lateraux chez les Vertebres, par P. Mitrophauow, analyse et critique," Archiv. Zool. Exper., 2 ser. t. ix. Notes et Revue, pp. iv.-viii., 1891. Review of work by P. Mitrophanow (142). 104. HoussAY, F. — "Theorie vertebrale du Crane et Theorie metamerique de la Tete et du Corps," in " Notice sur les Travaux scientifiques de M. Frederic Houssay," pp. 25—46, Paris, 1894. 105. Hyrtl, J.- — " tJber den Seitencanal von Lota" Sitzungs. der Kais. Akad. der "Wissensch., Math.- naturw. CI., Bd. liii. pp. 551-557, Taf. i., 1866. See abstract, Ann. Mag. Nat. Hist., 3rd ser. vol. xviii. p. 264, 1866. 106. Jackson, W. H., & Clarke, W. B. — "The Brain and Cranial Nerves of Echinorhinus spinosus, with notes on the other Viscera," Jour. Anat. & Phys., vol. x. pp. 75-107, pi. vii., 1876. 107. Jherin'g, H. von. — 'Das peripherische Ncrvensystem der Wirbelthierc, als Grundlage fiir die Keuntniss der Regioncn-bildung der Wirbelsiiule,' pp. 1-238, Taf. i.-v., Leipzig, 1878. 108. JuLiN, C. — " Le Systeme nerveux grand sympathique de I'Ammocoetes {Petroinyzon planeri)" Anat. Anz., Jahr. ii. pp. 192-201, 1887. Revised and reprinted in 112. 109. Julin, C. — " De la Valcur morphologique du Nerf lateral du Petromyzon," Bull, de I'Acad. Roy. des Sci. de Belgique, ser. iii. t. xiii. pp. 300-309, 1887. Revised and reprinted in 112. 1 10. Julin, C. — " Les deux premieres Fentes branchiales des Poissons Cyclostomes, sont-elles homologues respectivement il rEvetit et a la Fente hyobrauchiale des Selacieus?," Bull, de I'Acad. Roy. des Sci. de Belgique, ser. iii. t. xiii. pp. 275-293, plan, i., 1887. Preliminary to 112. SECOND SERIES. — ZOOLOGY, VOL. VII. 29 212 MR. F. J. COLE ON THE STEUCTUEE AND MOEPHOLOGT OF • 111. JunN, C. — " Reclierches sur I'Auatomie de I'Ammocoetes. — I. Les deux premieres Fentes branchiales des Poissons Cyclostomes, sont-elles homologues respectivemeut k I'^vent et a la Fente hyo- branchiale des Selaciens?," Bull. Sci. du Nord de la France et d. 1. Belgique, t. xviii. ser. ii. ann.x. pp. 265-295, plan, iv., 1887. Fuller than preceding, but preliminary to ii2. 112. JuLiN, C. — " Recherches sur I'Appareil vasculaire et le Systeme nerveux peripherique de rAmmocoetes (Petromyzon planeri) . Horaologie entre la premiere fente branchiale definitive de cet animal et I'^vent des Selaciens — valeur morphologique du corps thyroide," Archives de Biologic, t. vii. pp. 759-902, pi. xxi.-xxiii., 1887. See criticism by A. Dohrn, 62, and also reply by Julin, Zool. Anz., Jahr. xi. pp. 567-568, 1888. 113. Kingsbury, B. F. — "The Lateral line System of Sense Organs in some American Amphibia, and Comparison with the Dipnoans," Trans, of the Amer. Micros. Soc, vol. xvii. pp. 115-154, pi. i.-v., 1895. 114. Kixr.sBURY, B. F. — "The Structure and Morphology of the Oblongata in Fishes," Jour. Comp. Neurology, vol. vii. pp. 1-36, pi. i.-v., 1897. 115. Klaatsch, H. — " Ueber die Bedeutung der Hautsinnesorgane fiir die Ausschaltung der Sklero- blasten aus dem Ektoderm," Verhand. d. Anat. Gesellsch. a. d. ix. Vers, in Basel, voni 17-20 April, 1895 (Anat. Anz. Ergaiizungshcft z. Bd. x.), pp. 122-134, 1895. 116. Lenhosskk, M. v. — 'Beitrage zur Histologic des Nervensystems und der Sinnesorgane,' pp. 1-190, Taf. i.-iii., Wiesbaden, 1894. 117. Lenhosskk, M. v. — " Die Nervenendigungen in den Endknospen der Mundschleimhaut der Fische," Verhand. d. naturf. Gesellsch. in Basel, Bd. x. pp. 92-100, Taf. ii., 1895. 1x8. Leroux, — "Recherches sur le Systeme nerveux des Poissons," Revue Scientif. t. xliv. — ser. iii. t. xviii. pp. 278-9, 1889. Abstract. Not by author. No record o£ anything further published. 119. Leuret, F., & Gratiolet, P.— 'Anatomic comparee du Systeme nerveux considere dans ses rapports avec ^Intelligence,' t. i.-ii. Atlas, plan, i.-xxxii., Paris, 1839-57. 120. Leydig, F. — "Ueber die Schleimkanale der Knochenfiscbe," Miiller's Archiv f. Anat., Phys. u. wissen. Med., pp. 170-181, Taf. iv. figs. 1-3, Jahr. 1850. 121. Leydig, F. — " Ueber die Haut einiger Siisswasserfische," Zeits. f. wissen. Zool., Bd. iii. pp. 1-12, Taf. 1. figs. 1 & 2, 1851. 122. Leydig, F. — "Zur Anatomie und Histologic der Chimcera monstrosa," Miiller's Arch. f. Anat., Phys. u. wissen. Med., pp. 241-271, Taf. x. Jahr. 1851. 123. Leydig, F. — ' Beitriige zur mikroskopischeu Anatomie und Entwicklungsgeschichte der Rochen und Haie,' pp. 1-127, Tab. i.-iv., Leipzig, 1852. 124. Leydig, P. — ' Anatomisch-histologische Untersuchungen iiber Fische und Reptilien,' pp. 1-120, Taf. i.-iv., Berlin, 1853. 125. Leydig, F. — ' Lehrbuch der Histologic des Meuschen und der Thiere,' pp. 1-551, Hamm, 1857. 126. Leydig, F. — " Ueber Orgaue eines sechsten Sinnes, zugleich als Beitrag zur Keuutuiss des feineren Baues der Haut bei Amphibien und Reptilien," Verhand. der Kais, Leop.-Carol. dent. Akad. der Natur., Bd. xxxiv. Abhand. v. pp. 1-108, Taf. i.-v., 1868. 127. Leydig, F. — "Neue Beitrage zur anatomischcu Kenntniss der Hautdecke und Hautsinnesorgane der Fische," in " Festschrift zur Feicr des hundertjalirigen Bestehens der natur. Ges. in Halle," pp. 129-186, Taf. vii.-x., Halle, 1879. 128. Leydig, F. — "Intt^gument und Hautsinnesorgane der Knochenfiscbe," Zool. Jahr. Abth. f. Anat. u. Ontog., Bd. viii. pp. 1-152, Taf. i.-vii., 1895. 129. LiKoux, P. DE Sede, DE. — " Recherches sur la Liguc laterale des Poissons osseux," in "Theses presentees h. la Faculte des Sciences de Paris," pp. 1-114, pi. i.-iv., Paris, 1884. See criticism, 'American Naturalist,' vol. xix. pp. 612-613, 1885. THE CEANIAL NERVES AND LATERAL SENSE ORGANS OF FISHES. 213 130. LocY, W. A. — " Contribution to the Structure and Development of the Vertebrate Head," Jour. of Morphology, vol. xi. pp. 497-594, pi. xxvi.-xxx., 1895. 131. Malbranc, M. — "Von der Seitenlinie und ihren Sinnesorganeu bei Amphibien," Zeits. f. wisseu. Zool., Bd. xxvi. pp. 24-86, Taf. i.-iv., 1876. 132. Marcusen, J. — " Die Familie der Mormyren. Eine anatomisch-zoologische Abhandlung," Mem. d. I'Acad. Imp. d. Sci. d. St. Petersbourg, vii"" ser. t. vii. no. 4, pp. 1-162, tab. i.-iv., 1864. 133. Mayer, F. — "Das Centralnervensystem von Ammocwtes," Anat. Anz., Bd. xiii. pp. 649-657, Taf. i., 1897. 134. Mayer, F. J. C. — " Ueber den Bau des Gehirns der Fisehein Beziehung auf eine darauf gegriindete Eintheilung dieser Thierklasse," Verhand. der Kais. Leop. -Carol, deut. Akad. der Natur., Bd. XXX. Abhand. vi. pp. 1-40, Tab. i.-vii., 1864. 135. Mayser, p. — " Vergleichend anatomische Studien liber das Gehirn der Knochenfische mit besonderer Beriicksichtigung der Cyprinoiden," Zeits. f. wissen. Zool., Bd. xxxvi. pp. 259-364, Taf. xiv.-xxiii., 1882. 136. McMuRRiCH, J. P. — "The Osteology of Aminrus catus (L.), Gill," Proc. of the Canadian lustit., Toronto, n. s., vol. ii. pp. 270-310, pi. ii., 1883-84. 137. M'DoNNELL, R.— " On the System of the 'Lateral Line' in Fishes," Trans. Roy. Irish Acad., vol. xxiv. pp. 161-187, pi. iv.-vii., 1860-71. 138. Merkel, F. — ' Uber die Endigimgen der sensiblen Nerven in der Haut der Wirbelthiere,' pp. 1-214, Taf. i.-xv., Rostock, 1880. See criticism by B. Solger (194). 139. Miklucho-Maclay, N. VON. — ' Beitrage zur vergleichenden Neurologic der Wirbelthiere. I. Das Gehirn der Selachier. II. Das Mittelhirn der Ganoiden und Teleostier,' pp. 1-74, Taf. i.-vi. u. i., Leipzig, 1870. 140. MiNOT, C. S. — ' Human Embryology,' New York & Loudon, 1897. 141. MiTROPHANow, P. — " Uber die erste Anlage des Gehororganes bei niederen Wirbelthieren," Biolog. Central., Bd. x. pp. 190-191, 1890-1891. 142. MiTROPHANOw, P. — " Note sur la Signification metamerique des Nerfs craniens," Congres Internat. dc Zool., Sess. II., Moscou, part i. pp. 104-111, 1892. See criticism by Houssay (103). 143. MiTROPHANOW, P.- — " Etude embryogenique sur Ics Selaciens," Arch, de Zool. Exper. et Gen., ser. iii. t. i. pp. 161-220, pi. ix.-xiv., 1893. 144. Monro, A. * — 'The Structure and Physiology of Fishes explained, and compared with those of Man and other Animals,' pp. 1-128, tab. i.-xliv., Edinburgh & London, 1785. 145. MiJLLER, H. — "Den nervosen FoUikel-Apparat der Zitterrochen und die sogenannten Schleim- kaniile der Knorpel-Fische," Verhand. der Physik.-medicin. Ges. in Wiirzburg, Bd. ii. pp. 134- 149, 1852. 146. NuHN, A. — 'Lehrbuch der vergleichenden Anatomic,' pp. 527-570, 644-645, Heidelberg, 1878. 147. OsBORN, H. F. — "A Contribution to the Internal Structure of the Amphibian Brain," Jour, of Morphology, vol. ii. pp. 51-96, pi. iv.-vi., 1889. 148. Owen. R. — ' Descriptive and Illustrated Catalogue of the Physiological Series of Comparative Anatomy contained in the Museum of the Royal College of Surgeons in London,' vol. iii. part i. 'Nervous System and Organs of Sense,' pp. 1-208, pi. xxx.-xlii., London, 1835. 149. Owen, R. — 'On the Anatomy of Vertebrates,' vol. i., 'Fishes & Reptiles,' London, 1866. 150. OwsiANNiKow, Ph. — "Uber das sympathische Nervensystem der Flussneunauge, nebst einigen histologischen Notizen iiber andere Gewebe desselben Thieres," Bull, de I'Acad. Imper. des Sci. de St. Petersbourg, t. xxviii. pp. 439-448, 1883. ' Monro sectmdus." 29^ 214 ME. r. J. COLE ON THE STEUCTURE AND MOEPHOLOGY OF 151. Parker, T. J.^ — 'A Course of Instruction in Zootomy (Vertebrata)/ pp. 86-129, London, 1884. 152. Parker, W. K. — "On the Structure and Development of the Skull in the Salmon [Sahnosalar, L.)," Bakerian Lecture, Phil. Traus., vol. elxiii. pp. 9o-145, pi. i.-viii., 1873. 153. Parker, W. K., & Bettany, G. T.— ' The Morphology of the Skull,' pp. 332-336, London, 1877. 154. Parker, W. N. — "On the Anatomy and Physiology of Protopterus annectens,'" Trans. Roy. Irish Acad., vol. XXX. pp. 109-230, pi. vii.-xvii., 1892. 155. Philpeaux, J. M., & VuLPiAN, A. — " Memoire sur la Structure de I'Encephale des Poissons cartilagineux et sur I'Origine des Nerfs craniens chez ces Poissons, accompagn^ de Planches explicatives," Comp. Rend. Acad, des Sci. Paris, t. xxxvii. pp. 341-344, 1853. See also for report and discussion by Dumeril, Flourens, Duvernoy, Serres, and Prince Bonaparte, ibid. t. xxxviii. pp. 336-345, 366-376, and 421-428, 1854. 156. Pinkus, F.— " Ueber einen noch nicht beschriebenen Hirnnerven des Protopterus annectens," Anat. Anz., Bd. ix. pp. 562-566, 1894. Preliminary to 157. 157. Pinkus, F. — " Die Hirnnerven des Protopterus annectens," Morpholog. Arbeiten hrsg. v. G. Schwalbe, Bd. iv. pp. 275-346, Taf. xiii.-xix., 1894. Abstract, Zool. Anz., Jabr. xviii., Litteratur, p. 40, 1895. 158. Platt, Julia B. — " Ontogenetic Difl'erentiatious of the Ectoderm in Necturus. Study IL On the Development of the Peripheral Nervous System," Q. J. M. S., n. s., vol. xxxviii. pp. 485-547, pi. xx.xvi.-x.xxviii., 1896. 159. Plessen, J. von, & Rabinovicz, J. — ■' Die Kopfnerven von Salamandra maculata im vorgeriickteu Embryonalstadium/ pp. 1-22, Taf. i.-ii., Munich, 1891. See criticism by Strong (204, p. 162 et seq.). 160. Pollard, H. B. — "On the Anatomy and Phylogcuetic Position of Polypterus," Zoolog. Jahrbiich., Abth. f. Anat. u. Ontog., Bd. v. pp. 387-428, Taf. xxvii.-xxx., 1892. 161. Pollard, H. B. — "The Lateral line System in Siluroids," Zoolog. Jahrbiich., Abth. f. Anat. u. Ontog., Bd. v. pp. 525-550, Taf. xxxv.-xxxvi., 1892. 162. Pollard, H. B. — "The ' CirrhostomiaP Origin of the Head in Vertebrates," Anat. Anz., Bd. ix. pp. 349-359, 1894. Preliminary to 163. 163. Pollard, H. B. — "The Oral Cirri of Siluroids and the Origin of the Head in Vertebrates," Zool. Jahr., Abth. f. Anat. u. Ontog., Bd. viii. pp. 379-424, Taf. xxiv.-xxv., 1895. 164. Rabl, K. — " Uber das Gebict des Nervus facialis," Anat. Anz., Jahr. ii. pp. 219-227, 1887. 165. Ransom, W. B., & Thompson, D'A. W. — " On the Spinal and Visceral Merves of Cyclostomata," Zool. Anz., Jahr. ix. pp. 421-426, 1886. Sec criticism by Dohrn (62). 166. Rex, H. — " Bcitrage zur ^lorphologie der Hirnveneu der Elasmobranchier," Morpholog. Jahrbuch, Bd. xvii. pp. 417-466, Taf. xxv.-xxvii., 1891. 167. Richard, J. — -"Sur les Fonctions de la Ligne laterale du Cyprin dore," Comp. Rend. Hebdom. des Sean, et Mem. de la Soc. de Biol. ser. x. t. iii. pp. 131-133, 1896. See criticism by P. Bonuier, " Sur le Sens lateral," ibid. pp. 917-919, 1896. 168. Roux, Le. — See Leroux (118). 169. RuuE, G. — " Ueber das peripherische Gebiet des Nervus facialis bei Wirbelthieren," in Festschrift fiir Gegeubaur, Bd. iii. pp. 193-348, Leipzig, 1897. 170. Sagemehl, M. — " Beitrage zur vergleichendcn Anatomic der Fische. I. Das Cranium von Amia calva, L.," Morphol. Jahrb., Bd. ix. pp. 177-228, Taf. x., 1883-84. English translation by Shufeldt (q.v. — 187). THE CEANIAL NERVES AND LATERAL SENSE ORGANS OF FISHES. 215 171. Sagemehl, M. — " Beitriige zur vergleichenden Anatomie der Fische. II. Einige Bemerkungen iiber die Gehirnhaute der Knochenfische," ibid. pp. 457-474, Taf. xxiii., 1884. 172. Sagemehl, M. — " Beitriige zur vergleichenden Auatomie der Fische. III. Das Cranium der Characinideii nebst allgemeinen Bemerkungen iiber die mit einem Weber'schen Apparat versehenen Physostomenfamilien," ibid. Bd. x. pp. 1-119, Taf. i.-ii., 1884-1885. 173. Sagemehl, M. — " Beitrage zur vergleichenden Anatomie der Fische. IV. Das Cranium der Cyprinoiden," ibid. Bd. xvii. pp. 489-595, Taf. xxviii.-sxix., 1891. 174. Sanders, A. — "Contributions to the Anatomy of the Central Nervous System iu Vertebrate Animals. Part I. Ichthyopsida, Section I. Pisces, Subsection I. Teleostei," Phil. Trans, vol. clxix. pp. 735-776, pi. Iviii.-lxv., 1879. 175. Sappey, p. C. — ' fitudes sur I'Appareil raucipare et sur le Systeme lymphatique des Poissons,' pp. 1-64, pi. i.-xii., Paris, 1880. 176. Sarasin, P. & F. — "Einige Puncte aus der Entwicklungsgeschichte von Ichthyophis glutinosus {Epicrium gl.)," Zool. Anz., Jahr. x. pp. 194-197, 1887. Preliminary (in part) to 177. 177. Sarasin, P. & F. — "Zur Entwicklungsgeschichte und Anatomie der ceylonesischeu Blindwiihle Ichthyophis glutinosus, L.," in ' Ergebnisse naturwissenschaftlicher Forschungeu auf Ceylon,' Bd. ii. pp. 4.3-56, Taf. iv.-vii., Wiesbaden, 1887-1893. 178. Savi, p. — " Etudes anatomiques sur le Systeme nerveux et sur I'Organe electrique de la Torpille," in C. Matteucci's ' Traite des Phenomenes electro-physiologiques des Animaux,' pp. 272-348, pi. i.-iii., Paris, 1844. 179. ScHENCK, F. — See Fuchs (78). 180. ScHEPiLOFF, Catherine. — " Recherches sur les Nerfs de la viii""' paire cranienne et sur les Fonctions du Cerveau et de la Moelle chez les Grenouilles, avec un Aper9U comparatif des Fonctions du Systeme nerveux central dan.s la Classe des Batraciens," Mem. de la Soc. de Phys. et d'Hist. nat. de Geneve, t. xxxii. no. 6, pp. 1-115, plan, i., 1894-97. 181. Schulze, F. E. — " Ueber die Nervenendigung iu den sogenannten Schleimkanalen der Fische und iiber entsprechende Orgaue der durch Kiemen atlimenden Amphibien," Arch. f. Anat., Physiol, u. -wisscn. Medicin, pp. 759-7G9, Taf. xx., Jahr. 1861. 182. Schulze, F. E. — "Ueber die Sinnesorgane der Seitenliuie bei Fischen und Amphibien," Archiv f. mikros. Anat., Bd. vi. pp. 62-88, Taf. iv.-vi., 1870. 183. Schwalbe, G.- — "Das Ganglion oculomotorii. Ein Beitrag zur vergleichenden Anatomie der Kopfnerven," Jen. Zeits. f. Natur. h. v. d. Med.-natur. Ges. zu Jena, Bd. xiii. pp. 173-268, Taf. xii.-xiv., 1879. 184. Sede, p. de. — See P. de Sede, de Lieoux (129). 185. Shore, T. W. — "The Morphology of the Vagus Nerve," Jour. Anat. & Phys., Norm. & Path., vol. xxii., u. s. vol. ii., pp. 372-390, 1888. 186. Shore, T. W. — "On the miuute Anatomy of the Vagus Nerve in Selachians, with remarks on the seg- mental value of the Cranial Nerves," ibid. vol. xxiii., n. s. vol. iii., pp. 428-451, pi. xx.-xxi., 1889. 187. Shufeldt, R. W. — "The Osteology of Amia calva, including certain special references to the skeleton of Teleosteans," from the 'Annual Report of the Commissioner of Fish and Fisheries for 1883,' pp. 747-878, pi. i.-xiv., Washington, 1885. Contains translation of Sagemehl's paper (170, q. v.). 188. Solger, B. — " Zur Kenntuiss der Seitenorgane der Knochenfische," Central, f. d. Medicin. Wissen., Jahr. XV. pp. 657-658, 1877. 189. SoLGERj B.- — " Zweite Mittcilung iiber Seitenorgane der Knochenfische," ibid. pp. 817-820, 1877. 190. SoLGER, B. — "Ueber die Seitenorgane der Fische," Leopoldina, Kais. Leop.-Car. deut. Akad. der Natur., Dresden u. Halle, Hft. siv. pp. 74-80, 1878. 216 MR. F. J. COLE ON THE STEUCTUEE AND MORPHOLOGY OF 191. SoLGER, B. — " Neue Untersuchungen zur Anatomie der Scitenorpane der Fische. I. Die Seiten- organe von Chimara," Arcliiv f. mikros. Auat., Bd. xvii. pp. 95-113, Taf. viii., 1880. 192. SoLGER, B.- — "Neue Uutersucbungen zur Auatoraie dcr Scitenorgane der Fische. II. Die Seiten- organe der Selachier," ibid. pp. 458-479, Taf. xxxix., 1880. 193. SoLGER, B. — " Neue Untersuchungen zur Anatomie der Seitenorgane der Fische. III. Die Seiten- organe der Knochenfischc," ibid. Bd. xviii. pp. 3G4-390, Taf. xvii., 1880. 194. SoLGER, B.— "Ueber den fcinercn Bau der Scitenorgane der Fische," Bericht ii. d. Sitzung. d. naturf. Gcs. zu Halle, pp. 105-109, Jahr. 1880. Criticism of F. Merkel's work (138). 195. SoLGER, B. — See E. Bouenstein (24). 196. Stahr, H. — "Zur Funktion der Seitenorgane. Eine Beobachtung an chinesischen Zierfischen," Biolog. Central., Bd. xvii. pp. 273-282, 1897. See abstract. Jour. Roy. Micros. Soc., August 1897. 197. Stannius, H. — " Ueber die Knocben des Seitencanales dcr Fische," Froriep's 'Neue Notizen a. d. Gcbicte der Natur- uud Ilcilkunde,' Bd. xxii. s. 97-100, 1842. 198. Stannius, H. — "Ueber das peripheriscbe Nervensystem des Dorsch, Gadus callarias," Midler's Arcbiv f. Anat. u. Physiol., pp. 338-366, Jahr. 1842. Revised and incorporated in 199. 199. Stannius, H. — ' Das peripheriscbe Nervensystem der Fische, anatomiseh und physiologisch uuter- sucht,' pp. 1-156, Taf. i.-v., Rostock, 1849. See criticism by F. Ley dig (123), pp. 117-119. 200. Stannius, II. — " Handbuch der Anatomie der VVirbeltbiere. Zweite Auflage," Fische u. Amphibien, Berlin, 1854-56. 201. Stieda, L. — " Studien uber das eentrale Nervensystem der Knochenfische," Zeits. f. wissen. Zool., Bd. xviii. pp. 1-70, Taf. i.-ii., 1868. 202. Strong, O. S. — "The Structure and Homologies of tlie Cranial Nerves of the Amphibia as determined by their Peripheral Distribution and Internal Origin," Zool. Anz., Jahr. xiii. pp. 598-607, 1890. Preliminary to 204. 203. Strong, O. S. — "The Structure and Homologies of the Cranial Nerves of the Amphibia as determined by their Peripheral Distribution and Internal Origin," Part 2, Anat. Anz., Jahr. vii. pp. 467-471, 1892. Preliminary to 204. 204. Strong, O. S. — " The Cranial Nerves of Amphibia. A Contribution to the Morphology of the Vertebrate Nervous System," Jour, of Morphology, vol. x. pp. 101-230, pi. vii.-xii., 1895. 105. Swan, J. — ' Illustrations of the Comparative Anatomy of the Nervous System,' Second edition, pp. 1-250, pi. i.-xxxv., Loudon, 1864*. 206. Thane, G. D. — "The Nerves" in ' Quain's Elements of Anatomy,' vol. iii. part ii., 10th ed., London, 1895. 207. Traquair, R. H. — " On the Asymmetiy of the Pleuroneetidse, as elucidated by an Examination of the Skeleton in the Turbot, Halibut, and Plaice," Trans. Linn. Soc. London, vol. xxv. pp. 263- 296, pi. xxix.-xxxii., 1865. 208. Traquair, R. H. — ' The Ganoid Fishes of the British Carboniferous Formations. Part I. Palseoniscidse,' printed for the Palseontographical Society, pp. 1-60, pi. i.-vii., London, 1877. 209. Traquair, R. H. — " On the Structure and Classification of the Asterolepidte," Proe. Roy. Phys. Soc. Edinburgh, vol. x. pp. 23-46, pi. i.-ii., 1889. * First published 1835. THE CEANIAL NERVES AND LATERAL SENSE ORGANS -OF FISHES. 217 210. Tr AQV AiK, n. U.—" Homosteus, Asmuss, compared with Coccosteus, Agassiz/' ibid. pp. 47-57, pi. iii., 1889. 211. Traquair, R. H.— " On the Structure of Coccosteus decipiens, Agassiz," ibid. pp. 211-324, pi. xi., 1890. 212. Tkaquair, R. H.— "On Phlijctcenius [ = Phli/ctainaspis], a new genus of Coccosteidie," Geol. Mag., decade iii. vol. vii. pp. 55-61, pi. iii., 1890. See also Ann. & Mag. Nat. Hist., ser. vi. vol. xiv. pp. 369-370, 189i. 213. Vetter, B. — " Untersuchungen zur vergleichenden Anatomic die Kiemen- uud Kicfermusculatur der Fische." Theil I., Jcu. Zeits. f. Natur. h. v. d. Med.-natur. Ges. zu Jena, Bd. viii. pp. 405- 458, Taf. xiv.-xv., 1874. 214. Vetter, B.—Ibid. Theil II., ibid. Bd. xii. pp. 431-550, Taf. xii.-xiv., 1878. 215. VoGT, C, & Yung, E.— "Traite d' Anatomic comparee pratique," t. ii. pp. 339-552, Paris, 1894. 216. Weber, E. H.— 'Dc Aure et Auditu Hominis et Auimalium. Pars I. De Aurc Animalium Aquatilium,' pp. 1-134, Taf. i.-x., Lipsiae, 1820. 2 17. Weber, E. H. — " Ueber vier Liingeniierven bei einigeu Fischen, von denen zwei von dem Trigeminus und zwei von Vagus entspringeu, die die ganze Lange des Rumpfes durchlaufen," Meckel's Arch. f. Anat. u. Physiol., pp. 303-308, Taf. iv. fig. 25, Jahr. 1827. 218. Westling, Charlotte. — " Beitrage zur Kenntniss des peripherischcn Nervensystems," Bihaug till Kongl. Svenska Veteas.-Akad. Handliugar, Bd. ix. Hft. ii. Inn. viii. pp. 1-48, Taf. i.-iii. 1884. 219. Whitman, C. O. — " The Segmental Sense organs of the Leech," The American Naturalist, vol. xviii. pp. 1104-1109, pi. xxxiii., 1884. 220. Whitman, C. O. — " Some new Facts about the Hirudinea," Jour, of Morphology, vol. ii. pp. 586- 599, 1889. 221. WiEDERSHEiM, R. — ' Elcmeuts of the Comparative Anatomy of Vertebrates,' adapted from the German by W. N. Parker, pp. 1-488, London, 1897. 222. WiJHE, J. W. van. — " Ueber das Visceralskelett und die Nervea des Kopfes der Ganoiden und vou Ceratodus," Niederland. Archiv f. Zool., Bd. v. pp. 207-320, Taf. xv.-xvi., 1879-82. 223. WiLLEY, A. — ' Amphioxus and the Ancestry of the Vertebrates,' pp. 21 & 42-45, New York & London, 1894. 224. Willis, T. — 'Cerebri Anatomc : cui accessit Nervorum Descriptio et Usus,' pp. 1-456, Tab. i.-xiii., London, 1664. Reprinted in ' Opera omnia Thomie Willis,' 1682. 225. Wilson, H. V.—" The Embryology of the Sea Bass {Serranus atrarlus)," Bull, of the United States Fish Comm., vol. ix. pp. 209-277, pi. Ixxxviii.-cvii., 1889-91. 226. Wilson, H. V., & Mattocks, J. E. — "The Lateral Sensory Anlage in the Salmon," Anat. Anz., Bd. xiii. pp. 658-660, 1897. 227. Wright, R. R. — "On the Skin and Cutaneous Sense Organs of Amiurus," Proc. Canadian Inst., Toronto, n. s., vol. ii. pp. 251-269, pi. i., 1884. 228. Wright, R. R. — " On the Nervous System and Sense Organs of Amiurus," ibid. pp. 352-386, pi. iv.-vi., 1884. 229. Wright, R. R. — " On the Hyomandibular Clefts and Pseudobranchs of Lepidosteus and Amia," Jour. Anat. & Phys., Norm. & Path., vol. xix. pp. 476-499, pi. xxiv., 1885. 2] 8 ME. F. J. COLE ON THE STRUCTURE AND MORPHOLOGY OF EXPLANATION OF THE PLATES. The figures illustrating this paper have been carefully drawn, exact size, from nature, and, it is hoped, are something more than mere diagrams. The various structures and bones figured will be found to occupy their natural positions — every care having been taken to ensure this. Plate 21. Fig. ] . Dissection from the left side of Gadus morrhua, showing pit organs, sensory canals, and the skeletal support of the latter. Natural size. A. Articular. Ao. Angular. D. Dentary. D.F. Anterior extremity of dorsal fin. Fr. Frontal. Fr^. Triangular fossa on frontal lodging the bend of the infra-orbital canal. H. Hyomandibular canal. 7.0. Infra-orbital canal. U. Supra-temporal portion of lateralis canal. ") , ,. ^ [- = Lateralis canal. L^. Lateral portion of lateralis canal. ) Lg.H. Ligamentous portion of hyomandibular canal. La.S. Ligamentous portion of supra-orbital canal. L.O. First lateral line ossicle. N. Anterior and posterior narial openings abnormally fused together (the specimen was normal in this respect on the right side). Na. Nasal. 0. Opercular fold. PcT.F. Pectoral fin. P.F. Post-frontal or sphenotic t Foramen for otic branch of buccal nerve. Plv.F. Pelvic fin. P.O. Pit organs (= sensory follicles). P. Op. Pre-operculum. P.T. Post-temporal or supra-scapula. P.T^. Upper or epiotic limb of same. P.T^. Lower or parotic limb of same. Sb.O^. First sub-orbital or lachrymal. Sb.O^'^. Sub-orbitals 2 to 6 (5 and 6 = first and second " post-orbitals "), S.O. Supra-orbital canal. S.O.C. Supra-orbital commissure. Sq. Squamosal or pterotic. S.T^'*. Supra- temporals ( = extra- scapulae) 1 to 4. The numbers denote the dermal tubules of the respective sensory canals. THE CKAMAL NEKVES AND LATERAL SENSK ORGANS OF FISHES. 219 Plate 22. Fig. 2. Dissection from the left side of Gadns virens, showing the innervation of tlie sensoiy canals. The latter are coloured in segments, to indicate the positions of tlic skeletal supports. The lateral line nerves are shaded. Natural size. A. Articular portion of hyomaudibular canal. B. Buccal lateral line nerve. B'. Outer Imccal nerve. B". Otic branch of same. B\ Branch of outer buccal to pit organ. £'. Inner buccal nerve. B". Upper ramus of same. 5". Branches of upper ramus to pit organs of snout. B^. Lower ramus of inner buccal. B". Branch of above to pit organs below infra-orbital line. Cb. Cerebellum. C.H. Cerebral hemisphere. Cn. Outline of cornea. D. Dentary portion of hyomaiidibuhir canal. D.F. Anterior extremity of dorsal fin. Fx. Frontal portion of suiira-oi'bital canal. Fj?'. Frontal portion of infra-orbital canal. H. Hyomandibular nerve trunk. H\ Anterior or larger division of same. //^ Posterior or smaller division of same. H^. Chorda tympani nerve (= prte-branchial or pra;-spiracular — facial). H\ Hyoidean nerve {= internal mandibular, or post-branchial, or post-spiracular) . H^. External mandibular lateral line nerve. H^. Posterior branch of same to sense organs 11 and 12 of the hyomandibular canal. H\ Anterior branch of same to facial and mandibular pit organs. H.C. Hyomandibular sensory canal (violet). I.e. Infra-orbital sensory canal (red). L. Lateralis lateral line nerve. L\ Supra-temporal branch of above. L-. Posterior division of supra-temporal branch to sense organ 4 of the lateral canal, &c. L^. Dorsal division of supra-temporal branch to sense organs 1 and 2 of the lateral canal, &c. L'. Anterior division of supra-temporal branch to sense organ .'3 of the lateral canal, &c. L''. Doisal ramus of lateralis nei've. L^. Dorsal forward branch of same. U. Ventral ramus of lateralis. L.C\ Supra-temporal portion of lateralis canal. ) T /-.. T + 1 I 1-- 1 i- r 1 r = Lateralis sensory canal (brown). L.t . Lateral portion or lateralis canal. j . J \ i- L.G. Giinglion of superficial oplithahnic, buccal, and external mandibular lateral line nerves. L.l. Lobus inferior. L.O' ". Lateral line ossicles 1 to 3 supporting the lateralis canal. Mb. Medulla oblongata. Mif. Third or mandibular or inferior maxillary division of tlu^ trigeminus. Mx. Second or maxillary or superior maxillary division of the trigeminus. SECOND SERIES. — ZOOLOGY, VOL. VII. 30 220 ME. F. J. COLE ON THE STRUCTURE AND MORPHOLOGY OF Mx.Mn. Maxillo-mandibulav trunk (trigeminus). N. Autei'ior and posterior narial openings. Na. Nasal portion of the supra-orbital canal. 0. Opercular fold. O.L. Optic lobe. Olf. " Olfactory lobe." PcT.F. Pectoral fin. P.F. Post-frontal or sphcnotic portion of the infra-orbital canal. Pzr.F. Pelvic fin. P.O. Pit organs or sensory follicles. P.Op. Pre-opercular portion of the hyomandibular canal. P.T. Post-temporal or supra-scapular portion of the lateralis canal. Sb.O\ First sub-orbital or lachrymal portion of infra-orbital canal. Ss.O-''. Posterior sub-orbital portions of infra-orbital canal. S.C. Supra-orbital sensory canal (blue). S.Cm. Commissure between the two supi'a-orbital sensory canals. S.O. Superficial ophthalmic nerve trunk (= lateral line and trigeminal portions). /S.O'. Superficial ophthalmic lateral line nerve. S.O". Superficial ophthalmic or first division of the trigeminus. S.O^. Branch of -S.O.' to sense organ 4 of the supra-orbital canal. The two twigs probably supply pit organs. S.OK Branch of S.O.' to pit organ (?). S.Q\ Branch ol S.O.' to skin. S.O^. Anterior termination of S.O.- which has become separated from the lateral line nerve. Sq. Squamosal or pterotic portion of the infra-orbital canal. S.T^~^. Supra-temporal or extra-scapular portions of the lateralis canal. V. Vagus nerve with compound (?) ganglion. 1. Olfactory tract, peduncle, or "nerve." II. Optic nerve. V.-VII. Trigemino-facial ganglionic complex (adult condition), with portions of its two roots. The numbers denote the sense organs and dermal tubules of the respective sensory canals. Plate 23. Fig. 3. Transverse section of the head of a young adult Gadas virens, passing through the eye posterior to the lens, x 27. Drawn with camera. 1. Supra-orbital sensory canal. 2. External aperture of the 4th dermal (com- missural) tubvile of supra-orbital canal. 3. Superficial ophthalmic lateral line nerve. 4. Branch of above [=8.0.' of Fig. 2) to sense organ 4 of supra-orbital canal (note its condition on the other side). 5. Ophthalmicus superficialis trigemini. 6. Infra-orbital sensory canal. 7. Upper and lower rami of inner buccal lateral line nerve. 8. Superior maxillary Vth. 9. Inferior maxillary Vth. 10. Allis's nerve " c " (see text). Distributed to tissues behind and below eye. 11. Hyomandibular sensory canal, showing sense organ 8 of this line. 12. External mandibular lateral line nerve. 13. Mandibular pit organ. 14. Optic chiasma. 15. Posterior extremity of cerebral hemispheres (no lateral ventricles). 1(5. Frontal bone. THE CEANIAL NEEVES AND LATEEAL SENSE OEGANS OF FISHES. 221 17. Fifth sub-orbital. 18. Anterior extremity of pre-operculum. 19. 1 Primordial cranium. 20 = the " supra-orbital 20./ bands" of Parker (152, p. 129). 21. " Great fontanelle" (Parker, loc. cit.) 22. Fused trabeculse ( " inter-trabecula ") . 23. Parasphenoid. 24. Metapterygoid. 25. Syraplectic. 26. Quadrate. 27. Cerato-hyal. 28. Brancliiostegal rays. 29. Gill arches, with attached muscles, blood vessels, and gills. 30. Fragments of thyroid gland. 31. Phaiyngeal chamber. 32. Anterior extremity of opercular chamber. 33. Retina. 34. Sclerotic. 35. Superior rectus -1 36. Inferior „ I eye muscles. 37. Internal „ J Fig. 4. Transverse section of sense organ 3, supra-orbital canal, from a young Gadus virens. X 240. Details filled in with Zeiss Apochr. Hom. Imm. 2'0 mm., apert. r30, oc. compens. 4, 45 mm. 1. Sensory cells of sense organ. 2. Glandular cells of same (note periphery of organ consisting entirely of gland cells, 2', and centrally the secretion of the latter passing between the sensory cells). 3. Cupula terminalis (apparently contracted from its original attachments). 4. Nerve of sense organ. 5. Supra-orbital canal in transverse section. 6. Blood vessels. 7. Lateral frontal [i. e. lateral line ossicle secondarily fused on to frontal) . 8. Frontal. Cole. Trans Linn, Soc Zool. Ser SVol.VII, PI 21. F J Cole dc! K'Faj-Unc A. HrsViac LitV Ec; NERVES AND LATERAL SENSE ORGANS OF FISRES Cole Trans Linn Soc- Zool. Ser 2 VolYII. PI 22. ti- GADUS VIRENS. nat. s F J Cole, del M'FarlaJie 4- Er:Kme. I.nJi''' Edir.' NERVES AND LATERAL SENSE ORGANS OF FISHES Cole. Trans Linn Soc Zool, Ser 2 VolVII, P1.23. V\i 4. < 240 Fi^. o. X 27 M'FiriaTie A.Erskir.c Luh'^' Ea. NERVES AND LATERAL SENSE ORGANS OF FISHES. LINNEAN SOCIETY OF LONDON. MEMORANDA CONCERNING TRANSACTIONS. The First Series of the Transactions, containing both Botanical and Zoological contributions, has been completed in 30 Vols., and a few entire sets are still for sale. Only certain single volumes, or parts to complete sots, may be obtained at the original prices. The price of the Index to Vols. 1-25 is 8s. to the public, and (!«. to Fellows; to Vols. 26-30, 4s. to the public, and 3s. to Fellows. The Second Series of the Transactions is divided into Zoological and Botanical sections. The prices of the Zoological parts of tliese which have been published are as undermentioned (for the Botanical parts see Botanical wrapper) : — Volume. When Publi^hwl. Price to the Public. Price to Fellows. £ s. d. £ s. d. Secosd Series. — Zoologt. I. Part I. 1875. . . . 1 4 0 .. .. 0 18 0 Part II. 1875 0 6 0 .. .. 0 4 6 Part III. 1876 1 8 0 .. .. 1 1 0 Part IV. 1877 0 16 0 .. .. 0 12 0 Pait V. 1877 0 18 0 .. .. 0 13 6 Part VI. 1877 1 2 0 .. .. 0 16 6 Part VII. 1878 1 16 0 .. .. 1 7 0 Part VIII. 1879 1 0 0 . . .. 0 15 0 II. Part I. 1879 1 4 0 .. .. 0 18 0 Part II. 1881 0 15 0 .. .. 0 11 6 Part III. 1882 1 8 0 .. .. 1 1 0 Part IV. 1882 0 7 f) . . .. 0 5 6 Part V. 1882 0 3 0 .. .. 0 2 3 Part VI. 1883 1 0 0 .. .. 0 15 0 Part VII. 1883 0 5 0 .. .. 0 3 9 Part VIII. 1883 0 3 0 .. .. 0 2 3 Part IX. 1883 0 3 0 .. .. 0 2 3 Part X. 1884 0 4 0 .. . . 0 3 6 Part XI. 1884 0 10 0 .. .. 0 7 6 Part XII. 1885 0 6 0 .. .. 0 4 6 Part XIII. 1884 0 6 0 .. .. 0 4 (i Part XIV. 1885 0 6 0 .. .. 0 4 6 Part XV. 18S5 0 4 6 .. .. 0 3 6 Part XVI. 1885 0 5 0 .. .. 0 3 9 Fart XVII. 1886 0 3 0 .. .. 0 2 3 Part XVIII. 1888 0 2 6 . . .. 0 2 0 II. Part I. 1884 1 14 0 .. .. 1 6 i; Part II. 1884 1 12 0 .. .. 1 4 0 Part III. 1885 1 10 0 .. .. 1 2 6 Part IV. 1885 0 8 0 .. .. 0 6 0 When Price to llie Price to Volimu-. Pulilishcd. I'liblic. Fellows. .£ s. d. £ s. d. Second Series. — ZooLotiv {cuniinued). III. Part V. 1887 0 8 0 0 6 0 Part VI. 1888 0 6 0 0 4 0 IV. Part I. 1886 1 4 0 0 18 0 Part II. 1887 1 8 0 1 1 0 Part III. 1888 0 16 0 .... 0 12 0 V. I'ar-t I. 1888 0 12 0 0 9 0 Part II. 1^88 0 5 0 0 3 9 Part III. 1SS9 1 7 0 .... 1 0 0 Part IV. 1890 0 12 0 0 9 0 Part V. 1890 0 6 0 0 4 6 Part VI. 1891 0 12 0 .... 0 9 0 Part VII. 1891 0 0 0 .... 0 4 6 PartVfll. 1892 ....0 8 0....0 6 0 Part IX. 1892. ...0 12 0....0 9 0 Part X. 1893 1 8 0....1 1 0 Part XI. 1894 0 2 6 .... 0 2 0 VI. Part I. 1894 2 0 0 ... .'''1 0 Part II. 1894 Ill 0....1 3 3 Part HI. 1894 0 10 0 0 7 6 Part IV. 1896 1 4 0 0 IS 0 Part V. 1896 0 10 0 0 7 6 Part VI. 1896 0 8 0....0 6 0 Part VII. 1896. . ..0 12 0....0 9 0 PartVIII. 1897 0 2 6 0 2 0 VII. Part I. 1896 0 10 0 .... 0 7 G Part I'. 1897 0 12 0 0 9 0 Part llf. 1897 0 6 0 0 4 6 Part IV. 1898 0 10 0....0 7 6 Part V. 1898 0 18 0 .... 0 13 6 ^ 2nd Ser. ZOOLOGY.] [VOL. VII. PART 6. THE AUG 29 ^aSS TRANSACTIONS OP THE LINNEAN SOCIETY OE LONDON. A CONTRIBUTIOX TOWARDS OUR KNOWLEDGE OF THE MORPHOLOGY OF THE OWLS. BY W. P. PYCRAFT, A.L.S, M.B.O.U. LONDON: PRINTED FOK THE LINNEAN SOCIETY BV TAVI.OR AND FRANCIS, RED LION COURT, FLEET STREET. SOLD AT THE SOCIETY'S APARTMENTS, BURLINGTON-HOUSE, PICCADILLY. W. AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW. November 1898. [ 223 VI. A Contribution towards our KnoiuJedge of the Morplioloqy of the Owls. By W. P. Pycraft, A.L.S., 3LB.0.U. (Plates 24-29.) AUG '29 '■tjyt^ Head L'lst April, 1898. Part I.— PTERYLOGUAPHY. I. Introductory. IHE foimdation of the j)i'esent paper was laid by Prof. Newton, when, at tlie close of the year 1895, he sent me the remains — in the shape of a trunk — of a Sceloglnux alhifacies which had but recently died, M'ith a request that I would try to " make something of them," inasmuch as the bird "was exceedingly rare and verging on extinction. This trunk was that of a bird which liad for some three years been in the possession of Sir Fraccis Boileau, who had brought it to England from New Zealand. Unfortunately it had been sent immediately after deatli to a taxidermist, who had, however, been instructed to save the body for Mr. J. H. Gurney, from whom, it came, through Prof. Newton, to the Department of Comparative Anatom.y of the University Museum of Oxford, where I was then woi'king. Thus it happens that certain points concerning the external anatomy are probably lost for ever. The general pterylosis, and details of the form of the external aperture of the ear, ought to have been carefully examined in comparison with that of other species ; but as it is, the only information on the subject we possess we owe to Mr. J. H. Gurney (7) and the kindness of Prof. Newton, who generously lent me a stuffed specimen and gave me permission to relax the skin and make what I could of an examination of its pterylosis. This I liave done, and the results will be found embodied in this paper. It struck me that the best method of learning something about Sceloglaux was to be attained by first of all gathering as much information as possible about all the other Owls, as by this means we might hope to gain som.e insight as to its probable systematic position. The present paper represents the first fruits of an attempt to realize this end, and deals with the external topography only. Later I propose to describe the Osteology, Myology, Syrinx, Sacral Nervous System, and Visceral Anatomy of all the species whose joterylosis is now described, and of as many additional species as can be obtained. The pterylosis of Asio accipitrimis has lieen taken as the type of the paper and is described in some detail, wliile only the salient features of the remaining species have been touched upon. In this first part the pterylosis of some 20 species is described. I have been liberally supported by fi'iends interested in the subject, and wish here to record my deej) sense of indebtedness to those who have contributed material. SECOND SERIES. — ZOOLOGY, VOL. VII. 31 224 ME. W. P. PYCEAPT ON THE Ri'st of all my tlianks are due to Prof. Newton, who started tlie work, provided me with a valuable collection of skeletons, and furnished me with an unlooked-for chance of examining the pterylosis of Sceloglaux. Moreover, he has helped me throughout the preparation of this essay by kindly sympathy and advice. To Dr. Sclater and Mr Beddard I owe much ; they have furnished me with valuable material from the Zoological Gardens, amongst which I would mention Bnho lacteus, B. mrgiiiianus, B. tovqnatus, Speotyto, and the only Barn Owl I have been able to examine. Mr. Jesse, of La Martinifere College, Lucknow, sent me some beautifully preserved embryos o^ Athene hrama and other material. Mr. Allan Wilson, of Wadham College, Oxford, procured me the Tawny Owl from which the description which follows was taken, and since then I have had one other from the Zoological Gardens. Lastly I have to record the share contributed to this paper by one who has since passed away. I allude to the death of one of my kindest and best of friends, Mr. Daniel ^leinertzhagen. His loss Avill be keenly felt, not only by those to whom he had endeared himself in a thousand different ways, but by ornithologists at large. Young, enthusiastic, an exquisite draughtsman, a most careful worker, and imbued with a deep love of Ornithology, he seemed in every way one from whom we might expect great things, and had he lived we should not have been disappointed. He took the keenest interest in the progress of this work, and furnished me with most of the forms herein described from liis aviaries at Mottisfont Abbey. Amongst the rarest are the Tengmalm's Owl, Ural Owl, Snowy Owl, Ninox, Hawk Owl, and Scops. To his pen we owe the very beautiful and accurate figures which form the Plates illustrating this part. These were drawn in Prof. Lankester's Laboratories at Oxford, where we were both working, and they represent the last work he ever did before taking his farewell of the University. II. Historical. Almost all that we know concerning the pterylography of the Owls we owe to Nitzsch (13). Regarding them as closely allied to the Accipitres, he yet realized that they differed pterylologically from them in many important particulars, svich as the absence of an altershaft to the feathers, the absence of down-feathers on the pterylae, and the nude oil- gland. In their general pterylosis, as he pointed out, they bear what is certainly a striking resemblance to the Falcons ; but, as will be shown later, the points which he selected as distinguishing the Owls from these break down when applied to a large series of examples. His study of the pterylography induced him to divide the Owls into two sections by cutting off the Barn Owls as a gronp by themselves apart fi-om the rest. Later researches, taking into account the osteology and myology of the two groups, have confirmed his action. According to Nitzsch the pterylosis of all the Owls, save Strix, was exactly the same as in Bubo (Stria; bubo, 'L., = B. ignavus, Forst., was tlie type figured), save in one or two minute points, the various species being distinguished one from another by the relative MOKPHOLOGY OF THE OAVLS. 225 lengths of the primaries and the amount of the emargination of their inner vanes. In Speolyto, save for the spinal tract, he tells us that " the pterylosis differs in no respect from that described for S. bubo.'" In the Barn Owls the ventral tract is said to " agree precisely with that of Cathartes, only differiug in that, in Utjhris { = Slria; Jlammea), the contour- feathers are more numerous aud stand closer together." He was the first to draw attention to " two minute feathers seated at the apex of the oil-gland." The Ibrin and nature of the facial disc are very carefully described, but he does not seem to fully apj)reciate the difierenciis in the form of the external aperture of the ear and of the opercular fold in the various Owls. I have been unable to confirm his description of the disc-feathers of S. otiis { — Asio otus). According to him, " they consist of a large tube perfectly open above, upon which a very small, sparsely-barbed shait is seated." As will be seen later, my own investigations bear out the main conclusions of Nitzsch's work most completely. The wonder is that he failed to notice the numerous small but very real ditiercnces by w^hich not only genera but even species may be distinguished. It is also a matter for some surprise to find liim so closely associating Syrnium and Scops, the pterylosis of which is said to be "as in Bubo.'' The resemblance between Bubo and Scops is undoubtedly very great, as he indicates, but Sp'uium can be very readily distinguished from either. It is somewhat strange to find that Kaup (8), in his " Monograph of the Strigidse," published in the Trans. Zool. Soc. 1862, makes no mention of Nitzsch's work. He takes into account only such facts as the presence or absence of '• horns," the form of the disc — which he divides into a central " veil " surrounded by a " wreath," — the length of the primary remiges, and whether or not they are eniarginate, and the presence or absence of featiiers on the acrotarsium and acropodium. He draws attention to the fact that the feathers seated in the posterior membranous fold surrounding the external aperture of the ear are so arranged as to slope in opposite directions, meeting at an obtuse angle. The feathers in the upper half of the fold are directed downwards, those in the lower half are directed upwards. The feathers in this fold constitute the " wreath." Besides these, he also employed such external features as the form of the nostrils, the development of the " cere," and the size of the external aperture of the ear. The points noticed in the latter were (1) the size, measured with relation to the long axis of the eyelids, its symmetry or asymmetry, and (2) the presence or absence of an operculum or " anterior ear-lap " and of a posterior or " hinder ear-lap." The somewhat com- plicated relations which, obtain in the asymmetrical ear-apertm*es of Asio are fairly well described. The systematic arrangement resulting from his labours *, however, is decidedly inferior to that of A'itzsch. He divided the Owls into two subfamilies, Syrniinse aud Striginse. * The osteology of the skull also found a place in his diagnoses. He employed such characters as the form of tho occipital region and the pncumaticity of the bones. He was, I believe, the first to point out the swollen spongy nature of the interorbital septum of the Barn Owl. 31* 226 ME. W. P. PYCEAFT ON THE The former contained tlie genera Olaucidium, Ntjctala, Athene, Syrnia, and leroglawx, the latter Scops, Otus, Bnho, and Strix. In the earlier editions of Yarrell's well-known ' History of British Birds,' the scheme of classification which was followed for the Owls was that of dividing the group into " horned " and " hornless." This jilan was discarded some six-and-twenty years ago hy Prof. Newton when he undertook the task of re- writing this work (12). He adopted in the main that of Messrs. Sclater and Salvin. Internal characters apart, they were divided into two groups, the one, representing the Barn Owl and its allies, characterized hy the disposition of the feathers of pt. ventralis, the form of the operculum, the straightness of the heak at the hase, and a serrated middle claw. Tlie other was further subdivided into " the Owls which possess an operculum to the ear and the Owls which do not;" .... " to the first of these belong the Tawny, Tengmalm's, the Long-eared and Short-eared Owls, and to the second the remainder of the species." Whether the subdivision adopted for this latter group is destined to prove the right one or not I will not venture to say ; for my own part I have some doubt as to its power to hold its own. The form of the operculum and cavernum (p. 259) in Asio and Syrnium have really very little in common, whilst that of Ni/ctala can hardly be called an operculum at all. The minor characters utilized in the formation of genera were such as the form of the facial disc, length of the remiges, presence or absence of " ear "-tufts, feathering of the legs, and the form of the external aperture of the ear, whetlier symmetrical or not. Pour years after the publication of Prof. Newton's work Dr. Sharpe brought out his 'Catalogue of the Striges ' (16). Measurements, and such points as the size of the external apertures of the ear, measured with relation to the length of the horizontal axis of the eyelid, the presence or absence of an operculum, and the development of the facial disc, are the characters which he adopted for the purpose of breaking up the group into genera and species. Pollowing Sclater and Salvin, he divided the Owls into two groups — the Striyidce, containing the Barn Owl and its allies, and the JBubonidce. The only external characters adopted for the purposes of this di\dsion were the relative lengths of the middles toe and the presence or absence of a serration along the inner margin of the claw of that digit. The Buhmiidce were further subdivided into the Buhoiiina;, with the " ear-couch not larger than the eye (? eyelid), without an operculum; facial disc unequal, the portion below the eye being always much greater than the area above the latter ;" and the Syrniince, with the " ear-conch much larger than the eye, with very large operculum shutting in the ear ; facial disc always distinct, and extending as far above the eye as below it." These two sub-families agree with Prof. Newton's — those which possess an operculum to the ear, and those which do not. In addition, however, to the question of the presence or absence of an operculum, Sharpe takes into account the size of the aperture of the ear ; this w as a step in the right direction, but it was not enough, since even now Owls are brought together which ought to be placed in separate sub-families. i MOEPHOLORY OF THE OWLS. 227 Thus, Stjrnia, Carine, Speotyto * are placed in the Bubonlnce, to wliich I do not think they rightly belong, whilst Aslo, Sijniium, and Nyckda probably represent as many sub-families. Dr. Gadow (2-4) accepts Nitzsch's diagnoses of the general pterylography of the Owls. In one point, however, he is confident where Nitzsch wavered, inasmuch as he states that down-fpathers occur and are confined to the apteria. Nitzsch was doubtful on this point, but observed that he had never fou]id them when he searched for them. As a matter of fact, they occur only in ih.o, pt. ularis (PL 24, d.f.). In the structure of the neossoptiles Gadow considers the Owls to approach the Goat- suckers. He remarks that they are double as long as the neossoptiles of the Accipitres, and are further cliaracterized by their extreme softness (" weichlieit "). As will be pointed out later (p. 253), the " nestling-down " of the Accipitres is of two kinds — one pi-eceding the future definitive contour-feather, and one preceding tlie future definitive down-feather ; and it is probably with this last that the " nestling-down " featlier of the Owl has been compared, instead of with the former, with \^■llich it really corresponds. Moulting takes place once annually, in July and August. Although the curious separation of the external aperture of the asymmetrical ears of Aslo into two chambers (p. 259) is briefly described, no mention is made of the " oper- culum " or of the posterior membranous fold fencing in this aj^jcrture from behind. Meijere, in his papers on the distribution of the hair in the Mammalia (9) and of the feathers amongst Birds (10), has brought to light some most interesting and valuable facts. In his first paper he sets himself the task of proving that there is considerable evidence to show that the arrangement of the Mammalian hair indicates the presence of an earlier clothing of scales f . On the loss of the scales the hairs remain, and thus is explained their peculiar distribution in groups of three or more, as the case may be. The author, in bis second paper (10), points out the close similarity in the arrange- ment between hairs and feathers, and remarks that, though he does not wish to imply an homology between these structures, yet their close resemblance in this particular would be a not unimportant argviment in favour of such homology if it were supported by other evidence. But our knowledge on this point is as yet very meagre. According to Meijere, at a certain stage in the development of the Avian type the liody-clothing consisted of alternating groups of similar feathers, whose long shaft bore a biserial system of branches — Rami. These, in their turn, supported a biserial system of smaller branches — Radii. The latter consisted of a row of cells, provided distally with a pair of fila (" wimperchen"), which probably here and there took the form of booklets. Later, the middle feather of each group developed at the expense of the rest, which now assumed the form of the diminutive feathers which we know as Filoplumes. The middle * I am only takiug into account such genera as I have had the opportunity of studying ; Hetero'jlaux and Gi/ninasio have not yet come under my ken. t The nature of the combined covering of scales and hairs can bo studied on the scaly tails of such animals as Mas decumanus. Here they arise in groups of three from behind each scale, the middle one being the strongest. In some cases the number of hairs associated with each scale is greater than three. In Castor fiber, for instance, there are eight such hairs. In cases of this kind the hairs seem aU to be of equal calibre. 228 ME. W. P. PrCEAFT ON THE feather became either a true contour-feather, in which the Fila became more and more perfectly specialized to form " Hooklets," at the same time as the shaft, rami, and radii increased in strength, or it became a true down-feather. In the latter case the change might be brought about in two ways : — (1) the radii became relatively weaker and more numerous, and the fila of the radii became metamorphosed in various ways, or (2) the shaft became very considerably shorter and weaker. Umbelliform (" doldenformige ") dowu-feathers represent reduced long-shafted feathers. The same, he thinks, may be said of umbelliform neossoptiles. As to whicli is the older of these two primitive types of long-shafted feathers we know at this time nothing for certain. In the course of his paper on the distrihution of feathers Meijere briefly discusses the arrangement of the feathers covering the tarso-metatarsus of a young Barn Owl. These he describes as seated in groups of three, a median and two lateral, under the anterior margin of the scales of the podotheca. Such an arrangement recalls, he reminds us, that which obtained on the scaly tail of Mns decumanus. The median feather was much the largest ; the two lateral repi*esented degenerate contour-featiiers sunk to the level of filoplumes. On the tibio-tarsus the feathers are also described as arising in groups of three, but the median feather is here mucli more developed, possessing a distinct though seraiplumous vane. In an adult Barn Owl the lateral feathers of the scales of the podo- theca were wanting or occurred sparingly. Down-feathers and degenerate filoplumes are described as occurring on the dorsal apterium. On the breast filoplumes are stated to occur sjjaringly. I find myself vmable to confirm one or two small points in this paper, inasmuch as after careful search I failed to find more than a single filoplumu to each feather on the tibio-tarsus, nor could I succeed in finding any filoplumes associated with the down- feathers of the Owls examined by me. As touching matters of identification, it seems to me that Meijere has failed to distin- guish between true down-feathers and semiplumes. The latter occupy the position of and represent contour- feathers which are in process of degeneration and final disappear- ance— the modus ojjerandi of the formation of pterylJB. Down-feathers occupy the spaces between contour-feathers. This makes it rather difficult to accept the hypothesis that thev are degenerate contour-feathers. III. Pteroltsis of the Adult. Asio ACCIPITRINUS, Pall. (PI. 25. figs. 1-3.) PterylcB : — Fteryla capitis. — For the purposes of convenience in description this tract will be described throughout under the following subdivisions : — Fronto-parietal are a. — This extends from the base of the beak to the crown of the head. It is clothed by a narrow median band of feathers, expanding later to cover the occipital area. On either side of this median band, in the region corresponding with the level of the superior limb of the crescentic membranous folds surrounding the MORPHOLOGY OF THE OWLS. 229 external aperture of the ear, lies a single longitudinal row of some 6-7 feathers distinctly longer than those in the neighbourhood, whose shafts are directed outwards and back- Avards; they form the " ear "-tufts or " horns." Running along the outside of the bases of these '• ear "-tufts is an oblong apterium. Occipital are a. — This is a backward continuation of the parietal area, terminating at the occiput. Tlie feathers radiate from the middle line outwards and somewhat upwards. As they approach the free edges of the post-aural ibid they gradually decrease in size, and become at the same time more thickly planted. Loreal area (PI. 25. fig. 2). — This is occujned by a somewliat cresceutie and densely-packed group of stiff-shafted feathers (p. 25G). Its upper limb runs backwards to terminate in the region above the middle of the eye. The lower limb runs downwards and backwards behind the gape to join the feathers of the operculum. Ocular are a. — In consequence of the fact that the eyes, instead of being sunk entirely within the orbit, are directed outwards and forwards, a considerable portion of the sclerotic ring is brought into close relation with the outer world. This portion is protected by the eyelid, which is on this account of greater relative size than in other birds. It is clothed by a more or less regular series of semicircular rows of feathers, decreasing in size as they approach its free edge. Circum aural area. — As its name implies, this area concerns the region around the external aperture of the ear. In the present species, and in some o^her Owls, this region is characterized by the considerable development of membranous folds of skin — the pre- and postaural folds. When the former is largely developed, it is known as the " operculum." Operculum. — The feathers seated along tlie free edge of the operculum (PL 25. fig. 2) are aU elongated, with narrow vanes ; they form a series of from 3 to 4 rows liaving an outward and upward direction. Hunning parallel with the feathers on the free edge, but separated by a considerable apterium, is a single row of feathers, differing only from those of the free edge in that they are of a more slender character. Postaural fold. — The feathers seated along the free edge of this fold form the periphery of the disc (PI. 25. fig. 2). They belong respectively to the fronto-parietal, occipital, and ramal areas, and are continvied downwards and forwards to unite in the middle line near the syrapliysis of the mandil^le. They are compai-atively short, slender, and nai'row-vaned, so densely packed that their embedded bases form a thickened rim to the fold. The feathers occupying the lower segment of this " rim " are directed upwards, and those in the lower segment downwards ; as a consequence, where the two series meet a notch is formed (PI. 27. fig. 3). This point was apparently first noticed by Kaup (p. 225). Ramal area (PL 25. fig. 2). — This may be divided into two sections — an anterior, extending from the gape on each side of the ramus to the junction of the pre- and post aural folds, and a posterior strongly-marked triangular section, the base of which merges with the feathers of the postaural fold. Inter ramal are a. — The space from the symphysis of the mandible Ijackwards to the point where the inferior limb of the crescentic peripheral disc-feathers converges is occupied by a broad truncated-conical patch of elongated, slender, forwardly-directed 230 ME. W. p. PYCEAFT ON THE feathers. It is bounded on either side and in front by an aptevium. The remainder of the interramal area lies behind the convergent ends of the peripheral disc-feathers, but is almost directly lost in the pteryla colli ventralis (PL 25. fig. 2). It forms a narrow tract of some two feathers in width. In one case, however, this expanded to merge on either side with the mandibular area. The facial disc is formed by the feathers of the opercular and loreal areas afl.d those of the postaural fold. These last form the periphery of the disc. Ft. colli dorsalis (PL 25. fig- 3). — A densely -feathered tract, not closely investing the neck, but seated upon extensive lateral folds of skin, forming the T-shaped expansion of a vertical median membrane. The tract is widest a little below its origin at the occiput, then suddenly contracting, so that, at the point where it emerges with the pf. s])in(ilis, it is reduced to an obliquely-transverse row of some 5 feathers. A membra- nous fold investing a retractor muscle runs from the underside of the lateral neck- membrane at a point a little below its greatest width, in an obliquely-downward direction, to the pt. colli ventralis, and thence to the shoulder (PL 25. fig. 2). Ft. spinalis (PL 25. fig. 1}. — It is not pos.sible to draw a hard-and-fast line separating this at its upper end from the i^t- colli dorsalis. It is divided into aa upper A-sliaped and a lower Y-shaped portion. The former will henceforth be called the interscapular fork, and the latter the lumbar fork. The arms of the interscapular fork arise in the upper third of the interscapular region — iu one sj)eciinen at the root of the neck, where it took origin from the pt. colli dorsalis, there being no stem — and terminate on a level with the free end of the scapula. The feathers are seated in obliquely-transverse rows of not more than 4 feathers in each row. The branches of the lumbar fork are wide apart and continued upwards on to that portion of the axillary membrane lying between the humeral tract and the arms of the interscapular fork. A double row of semiplumous feathers enters into the formation of each branch, and these converge into the common stem some distance in front of a line drawn through the acetabular region. The stem is made up of transverse rows of feathers arranged in triplets (thus — . * .). It is bifurcated posteriorly, terminating just in front of the oil-gland, after somewhat increasiug in width. Pt. caudcv. — There are 12 rectrices. Ft. colli ventralis. — This comes into existence in the form of a backward coutinuation of the interramal tract. At first made up of a treble (transverse) row of feathers, it gradually increases in width to the middle of the neck, where it forks, each branch rimning to join the pt. ventralis at the shoulder. Ft. ventralis (PL 25. fig- 3). — Three branches mu.st be recognized in this tract — an outer, a median, and an inner. The outer is very distinct and runs backwards from the shoulder across the pectoral muscles to the preaxial border of the patagium. The median and widest of the three arises at the shoulder and runs backwards to within a short distance of the level of the posterior border of the sternum ; here it turns abruptly upwards and forwards to join the hypopteron. This upw^ard limb is the "hook" of Nitzsch. The area between the hook aud its main stem is more or less thickly MORPHOLOGY OF THE OWLS. 231 besprinkled with feathers. Tlie inner branch arises from the inner border of the median at a point rather in front of the level of the anterior end of the carina stcrni, and runs backwards and inwards to join its fellow of the opposite side in front of the cloacal aperture. It is widest on the breast, and contracts suddenly on the abdomen, to form a narrow band at some distance from the middle line. Its anterior end is ()l)li([uely trun- cated ; the lower and outer angle receives the feathers forming the preaxial border of the pt. marginalis ;. the vipper and inner angle is formed l)y a single row of feathers of the pt. veutralis at the shoulder. Fi. femoralis (PI. 25. fig. 2). — The limits of this tract are not sharply dehued. It is continued forwards to form a pt. Imnhalis. Along the posterior border of the thigh there runs a strongly-marked band of stout feathers, which is continued forwards to terminate below the knee-joint. This band I shall call {hafemoro-cruralh^xiix. Ft. cruralis (PI. 25. fig. 2). — The feathers of this tract are slender and do not invest the leg completely, but leave the inner surface bare. This tract is continued downwards over the acrotarsium and acropodium, to terminate over the penultimate phalanx of each digit. The planta is bare. Pt. alar is (PL 24) :— Metacarpo-digitals (primaries) 11; 9th longest (reckoning from within outwards) ; outer vane of the 10th serrated, with the tip of the inner vane emarginate ; distal end of the 9th with the outer vane serrated; 11th reduced to the condition of a remicle. Cubitals (secondaries) 14, the innermost small, not easily distinguished from their major coverts. The wing is aquintocubital. Tectrices : — T. major es. — The major coverts of the dorsal surface of the manus large; the lltli much longer than its remex (the remicle), but least of all the coverts of the hand belonging to this row. Carpal covert and remex large, closely attached by their base one to another and to the base of the 1st metacarpo-digital. The covert is larger than the remex. The dorsal major coverts of the cubital remiges are of imiform length throughout ; sometimes, however, there is a distinct break in the continuity of the line formed by the distal ends of these coverts. In such cases the break is found to be due to the fact that the coverts 1-5 are slightly longer than the rest of the row, whilst no. G is slightly shorter *. The row terminates at the elbow-joint, and is not carried inwards on to the humerus. The overlap is distal. On the ventral surface the feathers of the manus, with those of the cubital remiges, gradually increase in length from without inwards. Overlap distal. T. medicc. — On the dorsal surface of the manus this row commences at the 3rd meta- * As will be shown later, in the Owls the cubital major coverts of the dorsal surface are usually of uniform length throughout, locquality in the length of the major coverts is never found except in aquintocubital wings. The 5th major covert is distinctly longer than the 6th in Lc[itoiitilus javanicns, Griis (ntstraliensis, (J. ciiurea, Ardea ciiicrea, Palamcdea cornuta, Ccryle torquata, and Conurus. The 5th major covert is shorter than cither the 4th or (ith in Megapodius n'uvliariensis. SECOND SERIES. — ZOOLOGY, VOL. VII. 32 232 MR. W. P. PYCRAFT ON THE carpal rem ex. The overlap is distal. On the forearm the coverts 1-5 are very consider- ably the longest in the row, gradually increasing in length from without inwards. The series terminates at the elbow and is not continued inwards on to the arm. The overlap is proximal. On the ventral surface of the manus this row terminates at the base of the 6th remex. The feathers of the cubital series are fairly long and uniform in length, but concealed by the 1st row of the t. minores, which are slightly longer and closely approximated at their bases. T. minores. — On the dorsal surface these coverts are wanting on the manus. There are about 3 cubital rows, all with a proximal overlap. With the 1st row, as with the median series, coverts 1-5 are considerably the longest in the row. The same is true of the two succeeding rows, preaxially, but in the case of each row the differences of length decrease rapidly. On the ventral surface of the manus they are moderately long, replacing the t. medice after the 6th metacarpal remex. Those of the cubital remiges are slightly longer, and therefore conceal the t. medice. The 2ad row is made up of much smaller feathers than the 1st. It is only represented on the manus by 3 feathers lying at the bases of the metacarpal remiges 1-3. The 3rd row is not repi'esented on the manus. On the forearm it follows the course of tlic ulna. In texture the feathers of this row are small and semiplumous. The patagial membrane is clothed with 2 rows of degenerate semi- plumous contour-feathers. T. marginales. — On the dorsal surface there are 2 rows running along the preaxial border of the manus. The greater part of the surface of the patagium is clothed by some 4-5 rows, which arc continued inwards along its margin as a closely-set triple row of feathers, which ultimately join the ^9^. humeralis. On the ventral surface there are from 2 to 3 rows. The coverts of this scries are seated in closely-packed obliquely transverse rows. The postaxial row, seated on the patagium, are very long, and overlap the semiplumous t. minores. Parapteron. — This is made up of some 8 obliquely-transverse rows of 3 feathers in each row. Of these, the middle feather is the longest. The most preaxial feathers at the distal end of the arm rest on the humerus, the remainder on the muscles of the arm. Viewed as a whole, the parapteron is of a rounded conical form. It is connected by a double row of small feathers with the tectrices on the forearm. Whether these feathers can be brought into serial relation with the tectrices majores and medics of the forearm is a point not easy to determine. Hypopteron. — Elongated, delicate, semiplumous feathers, running along the under surface of the arm in some 6 transverse rows of 3 feathers in each row, make up the BJijpopteron. Proximally they merge with the recurved end of the median branch of the ■pt. vcntralis. Ala spuria. — Four strong feathers take part in the formation of the ala spuria. It is a point worthy of remark that the feathers of the dorsal surface of the pt. alaris, instead of being set in quincunx, form obliquely-transverse rows sloping from without MOEPHOLOGT OF THE OWLS. 23-5 nwards. The only other birds that I can recall in wliicli a similar arrangement obtains are Caprhmilgtis and Rhea. Semiplumoi : — Somiplumes * occur in the interscapular fork of the pt. spinalis, and compose tlu; branches of the lower Y-shaped portion of this tract. The femoral tract is largely made up of semiplumes. The inner border of the pf. humeralis is bounded by semiplumes, so also is i\io.iit. colli ventralls ; they also cover the axillary patagial fold. If the wing of a frcsbly- killed Owl be raised, a large bare sj)ace will be seen entirely liounded by semiplumes. They run along the recurved limb of the median branch of ventral tract, downwards, or, rather, backwards, from the origin of this to the pt. femoralls, the anterior border of which forms its jiosterior boundary, and hem in the space dorsally through the branches of the luml)£ir fork of the pt. spinalis, finally joining the j)Osterior end of the pt. humeralis. Thus, this portion of the apt. trimc. laterale (p. 231<) is entirely shut off from that lying between the pt. humeralis and pt. spinalis, and that between the inner and median branches of the pt. ventralls and the inner branch of the pit. ventralis and the thigh. PlumulcB : — Plumulaj or down-feathers are entirely absent on the trunk ; they occur onli/ on the wing, and are distributed as follows : — On the dorsal surface they are absent along the preaxial border of the wing ; behind this they occur singly between the contour-feathers m such a way as to form a quincunx with them (thus — ' • )' ^'^^ down-featlier occupying the centre. Two are found at the l>ase of every major covert, and two at the base of every remex (PI. 27). On the manus there is but one doAvn-feather to each major covert. Their distribution on the under surface of the wing is much as on the upper. Filoplumce : — The filoplumes are very slender and not easy to find. One or more occur at the base of every contour-feather on the body. Rhamphothcca. — The reniform external nares pierced through the anterior margin of a soft slightly tumid membrane at the base of the beak, resembling the cere of the Accipitres. For convenience' sake this will in future be called the cere. Podolheca. — Clothed with feathers save the base of the terminal phalanx, which is covered by 3 scales. Claivs. — There are no claws on the wing ; those of the foot not much curved, rounded, the inner border of the middle claw forming a sharp edge, decreasing in size from the base downwards. JJropygiwnt napiform. It is not tufted. Moulting. — Moulting takes place once a yeai', in July and August. {Gadow.) * It must be remembered that Semiplumoc are nothing more than degenerate coutour-feathers, and can always be distinguished from down-feathers in consequence of their position on the body, which is, necessarily, serial with the contour-feathers. Down-feathers arc always placed between contour-feathers when they occur on pterylic. Semiplumes always occur along the margins of pteryla) — if the whole pteryla is not made up of semiplumes — aud indicate a further restriction of a once wider tract. 32* 234 MK. W. p. PYCEAFT ON THE Apteria : — Ajjferinm cajnih.— The apteria of the liead are represented by the bare spaces surrouudiug the eyelids, the spaces on the operculum, and the space on the outer side of the feathers forming the " ear "-tufts (PL 25). JpL colli laterale (PI. 25. fig. 2). — Extends from the base of the neck, as a continua- tion of the apt. trntici lat., forwards to the base of the head, where it branches into au upper limb, running to the postaural fold, but terminating some distance from its free edge (PI. 25. tig. 3), and a lower, which serves to separate the ramal from the inter- ramal areas of the pt. capitis. Apt. trunci laterale (PI. 25. fig. 2). — Arising at the shoulder, it serves to divide the humeral fi-om the spinal tract, and is continued baclovards to the tail, and thus separates the spinal from the femoral tracts. At the free edge of the axillary fold it runs down- Arards to embrace the whole side of the bodv. Its extreme ventral limit is bounded bv the inner border of the pt. ventralis. Apt. spinale (PL 25. fig. 1). — This is the space enclosed by the brandling arms of the upper and lower portions of the pt. spinalis. Apt. mesogastrcei (PL 25. fig. 3). — The dimensions of this space can be seen at a glance in the figure. Apt. crurale. — This is the space at the l)ack of the leg (p. 231), and is continued downwards on the tarso- metatarsus. Aft. alee superioris. — The space dividing the humeral from the wing-tract. Apt. al(B infei'ioris. — Embraces the bare region of the arm and patagium. Asio OTUS, Linn. Pt. capitis : — Fronto-parietal are a. — The feathers forming the " ear "-tufts are distinct from the general j)lumage of the head, and consist of a longitudinal row of some 7-8 feathers, bounded on either side by an apterion. Loreal are a. — Sharply defined; upper limb free,more slender than in A. accipitrinus. I n t e r r a m a 1 are a. — The anterior symphysial portion shorter antero-posteriorly than in A. accipitrimiH ; the posterior portion joining the feathers of the ramal area on (.'ither side. Pt. colli dorsalis. — The lateral neck-folds gradually tapering backwards, and not suddenly expanded just behind the head. With a retractor muscle. Pt. spinalis. — Interscapular fork commencing at the confluence of the j^t- colli dorsalis and p>t. spinalis. The free ends of the intersaipular fork are received into the arms of the lumbar fork, which are continued outwards on the posterior patagial membrane as in A. accipitrinus. The tract l)ifurcates posteriorly, embracing, but not surrounding, the oil-gland. Pt. colli veniralis.—Yerj broad at the point of bifurcation. Pt.fehioralis. — Femoro-crural band terminating at the base of the patella. Pt. ventralis. — Inner branch somewhat nearer the middle line than in A. accipitrinus. MOEPHOLOGY OF THE OWLS. 235 Pi. alaris. — Tectrices : — T. majores. — The dorsal major coverts show an interesting departure from the general rule, in that, instead of the break in the uniformitj' of the length of these coverts occurring after the 5th covert, it occurs after the 4th, the 5th being slightly shorter than the 1th, but, as usual, longer than the 6th. T. medue. — The break in the uniformity of the length of the coverts is here very marked, the 6th covert being much sliorter than the 5th. T. minores. — The remarks apply to the first two rows of these coverts as to T. medics ; in the 3rd row the differeace is not so marked. Such other points as are unnoticed are indistinguishable from A. accipitrinns. Bubo ignavus, Torst. Pt. capitis : — F r 0 n t o - p a r i e t a 1 arc a. — The two areas included under the name can here be readily distinguished. The frontal area is represented by a broad median band of feathers running from the base of the beak to a point corresponding with a line drawn across the head behind the eyes, where the tract rapidly expands to form the parietal area. This distinction is not so readUy seen in the case of B. viaculosus on account of the smaller size of the feathers. The feathers forming the ear-tufts, in a single row, l)ounded posteriorly by an apterium, the anterior exti'emity of the ((pt. colli laterale. C i r c u m a u r a 1 area : — P r e a u r a 1 f o 1 d. — This is but slightly developed ; there is a double I'ow of feathers along its free edge, the bases of Avliich form a " rim." Postaural fold. — The upward and downwardly directed feathers along its free edge converging at about the middle of the fold, the bases of which form a " rim." The peripheral disc-feathers terminating on the mandible, just in front of the gape. Ocular area. — The rows of feathers encircling the region of the outer wall of the eye can be well studied in this species. Loreal area not sharply defined, the lower limb continued backwards to join the feathers at the confluence of the opercular and circmuaural folds. Ramal area not extending to the proximal end of the jaw, but turning inwards in the form of a narrow baud on the throat. Interramal area. — In the form of a broad tract, bounded on either side by a wide apterium. A^it. colli laterale. — Continued upwards from the neck on to the proximal end of the mandible ; Jrom thence it can be traced upwards and forwards on to the crown, where it serves to isolate the row of feathers forming the " ears." The continuity of the space is interrupted, however, by two tracts of feathers, one running from about the middle of the postaural fold, the other from the region immediately above the fold, to the feathers of the occipital area. 230 MR. W. P. PYCRAFT ON THE Ft. colli dorsalis. — Vertical and lateral neck-folds moderately well developed. No retractor muscle. Ft. sinnalis. — Branches of the interscapular fork arising- at the commencement of the interscapular region ; arms long, hut not extending to the end of the scapula. Branches of the lumhar fork arising about midway between the anterior end of the ilium and the region over the acetabulum, running forwards to embrace the ends of the interscapular fork, and terminate over the proximal f of the free end of the scapular. Stem forked posteriorly, branches widely divergent, terminating some distance to the outer side of the oil-gland. Pt.ventralis. — Outer branch strongly developed, arising at the summit of the shoulder. Median branch broad, very strong, with an upward trend at its free end, where it gives off the " hook." Inner branch wide, not sharply defined at its upper end ; the breast between the median branch and the carina being almost continuously feathered. Pt. humeraUs not continued forwards to johi the Ft. ventrcdis. Ft.femoralls. — The femoro-crural band terminating on the crus some distance below the knee. Pt. cruralis. — Continued downwards on the acrotarsium and terminating on the acro- pcdium. The base of the ungual phalanx clothed with scales. Proximal end of the planta bare. Bubo virginianus, Sw. & Rich. Pt. capitis : — Parietal are a. — " Ear "-tufts very distinct, the stumps of the feathers appreciably larger than those of the neighbouring feathers. They are seven in number, arranged to form a longitudinal row, the inner border of which is bounded by a large lenticular apterium. Loreal are a. — Upper limb not continued vipwards beyond the level of the anterior corner of the eyelid. Circum aural area. — Preaural fold feebly developed, a double row of feathers along its free edges ; a broad apterium separating these from the ocular area. Postaural fold.- — The convergent feathers along the fi'ec edge of the postaural fold meeting near its middle. The feathers behind the free edge arranged in the form of a triangular patch. Peripheral disc-feathers arising on the crown of the head just behind the eye, terminating on the mandibular area. Mandibular area. — Extending beyond the articular end of the mandible, very broad, at the level of the gape turning sharply inwards to merge with the feathers of the interramal area. Inter ramal area. — Very broad, at the level of the gape merging with the ramal ai*ea. Apterium colli laterale. — Extending on to the postaural fold and skirting the articular area of the lower jaw. Passing Ijehind the triangular patch of the postaural fold iipwards, it attains the level of its superior angle, but is cut off by a broad band of feathers from the lenticular apteria to the inner side of tlie " ear "-tufts. MOEPHOLOGY OF THE OWLS. 237 Ft. colli dorsalis. — Lateral neck-folds considerable ; with a retractor muscle as in Asio. Ft. spinalis. — Arms of the interscapular fork arising- about the middle of the inter- scapular region ; branches short, running almost parallel, not widely divergent, extending to the end of the scapula. Lumbar fork. — Arising behind the middle of the preacetabular region of the ilium; well developed, extending forwards on to the axillary membrane. Stem cleft posteriorly into two long arms, not extending to the oil-gland. Pt. colli veutraUs. — As in the other si^ecies of Bubo, branching about the middle of the neck to join the Pt. ventralis at the shoulder. Pt. ventralis. — Outer branch strong, arising low down from the outer border of the median branch. Median extending backwards as far as the posterior f of the carina sterni. Inner branch broad, but more sharply defined than in B. maximus or B. lacteus. Pt. alaris : — Bemiges. — Metacarpo-digital 11 ; cubitals 19 ; the innermost 2-3 cubital remiges scarcely differing in size from their coverts. Tectrices : — T. nuijores. — The major coverts of the dorsal surface vmiform in length. T. medicc. — With a distinct break in the uniformity of the length of the series, the 6th being much shorter than the 5th. T. minores. — With a l^reak in each row, similar to that seen in the T. medice. Parapteron. — Made uj) of some 10 obliquely-transverse rows sloping from without inwards. Each row composed of 4 feathers, the most postaxial a semiplume, and over- lapping the next, a true contour -feather, and much longer ; this in turn overlaps the one next above and longest of the series, which is overlapped by the most preaxial feather. Proximally these humeral feathers rest upon the muscles of the arm; distallv, upon the humerus ; running along the upper surface of the humerus is a row of semiplumes, all with the tips directed towards the tip of the wing. Between each of the transverse rows just described, and on the extreme postaxial border of the humerus, are two semiplumes, one lying a little above the other. The uppermost one runs parallel with the lai-ger feathers of the parapteron; the lower is pressed close to the arm, Avith its tip directed towards tlie end of the wing. Whether these feathers are really serially homologous with the dorsal and ventral major coverts and the dorsal row of median coverts, and a ventral row of minor coverts, as represented by the semiplume, or not, remains to be proved. It will be noted that in this case the x-emiges are wanting. The coloration of the feathers of the parapteron is not to be distinguished, save in minute points of detail, from those of the fore arm. Hypoptei'on. — Three parallel rows along the arms, somewhat semiplumous in texture. Pt.femoralis. — Femoro-crural band terminating considerably below the knee. Pt. eruralis. — Toes thickly feathered. Terminal phalanx covered by 3 broad scutes. 238 ME. "W. P. PTCEAPT ON THE Bubo MACULosrs, Bp. Ft. capitis : — iPronto-parietal area. — The feathers formmg the " ear "-tufts in a single longitudinal row on the crown of the head, with a w ell-markcd apterium on either side. Loreal area. — Not sharply defined; vipper limb scarcely extending beyond the anterior corner of the eyelid. Circumaural area. — The preaural fold is but feebly developed, and does not Ibrm an operculum. It bears but a double row of feathers along its free margin, the bases of which form a " rim " to the margin. The feathers of the postaural fold seated along the free edge of the upper part are, as usual, all directed downwards, those of the lower part upwards, the two series meeting in the middle of the fold. There is no " rim " to this fold. The peripheral disc-feathers apparently terminating below on the mandible, near the base of the rhamphotheca. Ramal area. — Truncated posteriorly, not extending backwards to the proximal end of the jaw. Interramal area. — Wide anteriorly, gently tapering joosteriorly to join the ft. coll. ventralis. Apt. colli laterale. — Occupying the lower portion of the circumaural fold, and running upwards and forwards behind a triangular patch of feathers clothing the upper part of the Ibid, to cut off the row of feathers forming the " ear "-tufts from the rest of the parietal area. Pt. colli dorsalis. — Vertical and lateral neck-folds of slight extent ; the posterior end of the tract where it joins the^^. spinalis 4 feathers in width. Pt. spinalis. — Branches of the interscapular fork arising in the middle of the inter- scapular region ; arms not extending to the free end of the scapiila. Branches of the lumbar fork arising about midway between the anterior end of the ilium and the acetabular. Strongly developed, embracing the ends of the interscapular fork, crossing the free end of the scapula, and terminating on the axillary membrane. Stem strong, broad, bifurcated posteriorly, but not reaching the oil-gland. Pt. ventralis. — Outer branch broad, arising at the summit of the shoulder. Median branch not extending backwards beyond the level of the upper ^ of the carina sterni. It is very sharply defined, with an upper trend at its free end. Its great distinctness seems to cut it off from the recurved branch or " hook " characteristic of tliis portion of the tract. Inner branch arises rather above the middle region of the clavicle, and follows the covirse of that bone up to the carina sterni ; running along the upper end of this for a short distance, it turns rapidly outwards and backwards to the region over the posterior lateral process of the sternum ; from this point it runs backwards to join its fellow in the middle line a short distance in front of the cloacal aperture. Pt. humtralis not running forwards to join the ft. ventralis. Pt. alaris. — Eemiges too abraded and broken to be of use for comparison as to length, &c. MOEPHOLOGT OF THE OWLS. 239 Tectrices : — T.majores. — Dorsal major coverts of tlie cubital remiges unfortunately moulting. The 6th covert not fully developed in one wing and absent in the other. In both wings, however, the old 5th and 7th coverts remain, and furnish a point of interest in that the 5th is considerably shorter than the 7th. The coverts from the 7th inwards gently decreasing in leng th T. medics. — The dorsal median coverts increase rapidly in length from 1-5, and are the longest in the row. The 5tli is considerably longer than the 6th. T. minores. — The dorsal minor coverts resemble the median in that 1-5 in each row are longer than the feathers lying proximally ; the difference is less marked in each row from behind forwards. Parapteron and Hypopteron, se-e Bubo virginianus (p. 236). The description is from a fresh specimen from the Zoological Gardens, and therefore better for study than in the present species — a spirit sjiecimen. Pt.femoralis. — Femoro-crural band terminating some distance below the knee. Ft. cruralis. — Continued downwards on to the acrotarsium and acropodium. In the latter region the feathers are not sufficiently large to conceal the skin. Base of the terminal phalanx clothed with scales. Jxliamphotheca. — The cere having the anterior margin sharply defined against the culmen, much as in Accipitres, slightly swollen immediately around the external aperture of the nostrils. TJropygium. — In form of an elongated cone, bearing 3 apertures at the tip. Bubo lacteus, Steph. Fteryla capitis : — Fronto-parietal are a. — Resembles in general conformation that of B. ignavus, but not so sharply defined. The cut bases of the " ear "-tufts not easily distinguishable from the rest of the feathers of the head. Ramal are a. — Very broad posteriorly, encroaching upon the throat ; not continued backwards so far as the proximal end of the mandible. Inter ramal are a. — Connected with the ramal by two separate lateral branches, one some distance behind the region of the gape, the other near the articulation of the jaw. Best of the feathering of the head almost identical with that of B. ignavus. Apt. colli laterale. — Extending upwards and forwards in the form of a long narrow space behind the postaiu'al fold, terminating about its ixdddlc. It is cut off by a broad baud of feathers from the small apterium lying to the inner side of the bases of the ear-tufts. Ft. colli (lorsalis.— With, vertical and lateral neck-folds slightly developed, less than in B. maximus. Ft. spinalis. — Interscapular fork arising in the upper fourth of the interscapular region ; the arms extend to the end of the scapula. Lumbar fork arising rather below the middle of the preacetabular part of the ilium ; SECOND SERIES. — ZOOLOGY, VOL. VII. 33 240 MR. W. P. PYCEAFT ON THE its arms extend outwards on to the axillai-y memhrane, bet\yeen the x>t- liumeralis and the scapula. Stem forked posteriorly, so as to embrace Imt not surround the oil-gland. Ft. ventral is.— Outer branch very strong, arising low down, over the dorsal third of the furculum. Median branch narrower and more sharply defined than in B. maximus, extending backwards to the level of the posterior third of the carina sterni. Not recurved at its posterior end, where it passes into the " hook." Inner branch arising from the median rather about its middle ; more sharply defined than m B. maxlmus. Pt. alaris : — Bemiges. — Metacarpo-digitals 11 ; cubitals 19. Tectrices : — T. majores. — The dorsal major coverts of the cubital remiges liaving the 5th slightly shorter than 4th, ))ut longer than the 6tli. 5th and 7th equal. T. medi(B. — 1-5 longest in the row ; 6 abruptly shorter than 5 ; from 6 inwards gradually increasing in length. T. minores. — The same remarks apply to these feathers as to the t. medicB. There are three rows of minor coverts. Pf.femoralis. — IFemoro-crural band terminating below the knee. Pt. cruralis. — Toes bare, but beset with stumps of old feathers worn down to surface of skin. Terminal phalanx having the base covered with scales. Bubo toequatus, Daud. {Syrnium iJerspicillatum, Lath., of Sharpe's Catalogue, p. 277.) Pt. capitis : — Pronto- parietal are a. — Frontal area in the form of a broad well-defined band, rapidly expanding above and rather behind the eyes to merge with the parietal area. Loreal area. — Ill-defined. Inter ramal area expanding in the region of the gape to join the ramal area, which does not extend backwards to the end of the jaw. Periphery of the disc ill-defined. Pt. colli laterale. — As in Bubo. Pt. colli dorsalis. — Lateral neck-folds very slight, tract narrow. Pt. colli ventralis. — Divides about halfway down the neck. Pt. spinalis. — Interscapular fork arising in the upper third of the interscapular region ; arms narrow, extending to the level of the end of the scapula. Branches of the lumbar fork arising near the anterior border of the ilium ; short, running out on to the axillary membrane. Pt. ce^itralis. — Outer branch well defined, arising at the summit of the shoulders ; median long, narrow, not curved terminally (as in Syrniuni) ; inner branch broad, sharply defined from the median, and, rising near tlie summit of the shoulder, joins its fellow of the other side just in front of the cloacal aperture. Pt. femo)'alis. — Pemoro-crural tract terminating at the knee-joint. MOEPHOLOGY OP THE OWLS. 2il Rhamphotheca. — Cere of considerable size, slightly inflated at the base ; aperture of the nostrils circidar, looking slightly downwards. Podothcca. — Acrotarsium feathered. Acropodium covered with warty protuberances. Terminal phalanx with the base sheathed by three scales. Uropygium. — The oil-gland is conical, as in Bnho- The pterylosis closely resembles that of Buho maculosus, but the vertical and lateral neck-folds are much less in extent than in this species. Bubo torquatus * and Syriiium aluco compared together prove conclusively the affinities of the former with Bubo rather than with Syniium, with which it appears to have been generally associated. Bubo nyctekus, L. {Nyctea scandiaca, PI. 26.) Fteryla capitis : — P r 0 n t a 1 area shorter and wider tlian in Bubo ; it is also much arched. Parietal area. — Feathers of the " ear "-tufts in a single longitudmal row, some 6 in number ; their bases conspicuously larger than those of the surrounding feathers, and hounded on the inside by an elongated lenticular apterium. Loreal area merging above with the frontal are;i. Circumaural are a. — Preaural fold feebly developed, with a double row of feathers along its free edge. Postaural fold as in Bubo, e. g. B. igiiavus. The periphery of the disc, as demarcated by the cut Ijases of the feathers, arising just above the postaural lold, rather helow the level of the toj) of the eye, runs dowmvards and forwards along the postaural fold on to the mandible, terminating midway between the gape and the rhamphotlieca. Ramal area extending backwards to the articular end of the jaw; bi-oad poste- riorly, exteading downwards to join the interramal tirea. Inter ramal area. — The anterior end giving off a narrow lateral branch on each side, at right angles to the main axis of the tract, from the region rather in front of the angle of the gape ; behind this the tract is constricted, expanding again immediately after to join the hinder end of tlie ramal area. Apt. colli latenUe extending forw ard on to the base of the postaural fold, but entirely cut oh' from the apterium at the base of the feathers of the " ear "-tufts. Pt. colli dorsalis very broad anteriorly. Vertical and lateral neck-folds considerable. Pt. spinalis. — Interscapular fork arising in the upper fifth of the interscapular region. Arms long, wide apart, extending to the end of the scapula, and almost joining the arms of the lumbar fork, so as to enclose a space. Lumbar fork. — Arms sharply defined (PL 26), arising in about the middle of the preacetal)ular region of the ilium ; they extend forwards on to the axillary membrane, meanwhile expanding so as to assume a conical form. Stem sharply defined, broad posteriorly, forked, the arms wider apart than the width of the oil-gland and terminating in front of this. * Cf. footnote at bottom of p. 270. 242 ME. W. P. PYCEAFT ON THE Pt. ventralis. — Outer branch distinct, arising rather below the summit of the shoulder. Median broad, terminating at a jioint corresponding with a line drawn through the carina sterni at its middle. Inner branch broad, arising about halfway down the inner side of the median branch. Nowhere sharply defined on the breast, but very distinct on the abdomen. Pt. alaris : — Bemiges. — Metacarpo-digitals 11. Cubitals 18. Tectrlces : — T. majores. — Coverts 1-5 of the dorsal surface longest, 6 slightly shorter than 5, 7 equal in length to 5. T. medice. — Damaged. T. minores. — Three rows. All showing an abrupt break in the length of the feathers in each row. 1-5 always considerably the longest. Pt.fcmoralls. — Fcmoro-crural band terminating some distance below the knee. Pt. cruraUs. — Continued downwards over the acrotarsium on to the acropodium, termi- nating behind the ungual phalanx, the base of which is clothed by scutes. The feathers on the acrotarsium and acropodium very long, half concealing the claws. Planta bare in region of tarsal joint only. Tlropyglmn. — In the form of an elongated cone, slightly swollen at the base. All the tracts are very narrow, but the feathers are of great length ; a humeral feather measured 8 in. from lower umbilicus to tip, and one from the median branch of the pt. ventralis 7 in. c OPS LEUCOTIS, Temm. Pt. capitis : — Fronto-parietal area. — Feathers of the " ear "-tufts or "horns" in 2 longi- tudinal rows of about 5 feathers in each. Their massed bases are surrounded by an apterium, except anteriorly, where they merge with the broad band of feathers occu- pying the median line of the head, thus tending to form short brandies to this band. Loreal area. — Moderately well defined, crescentic in form, the upper limb longest and terminating above the region of the middle of the eye. Feathers of great length. Circumaural are a. — The preaural fold feebly developed, not forming an oper- culum. A double row of feathers runs along its free edge, the bases of which form a " rim " ; in front of this is a single row set in an apterium, as in Asio accipitrinus (p. 229). The bases of the feathers along the free edge of the postaural fold also form a "rim," as in Asio &c. Interramal area. — Very wide posteriorly, merging with the mandibular area from the gape backwards, rapidly contracting on the upper part of the neck, where it becomes the pt. colli ventralis. R a m a 1 are a. — Broad, extending backwards beyond the end of the mandible. Lateral cervical apterium terminating at the base of the lower half of the circumaural fold (PI. 26). MOEPHOLOGY OF THE OWLS. 243 Pt. colli dorsalis. — With a slight lateral skin-fold : no retractor muscle. Pt. spinalis. — Ai'ms of inter sea j)ular fork ai-ising- rather ahove the middle of the inter- scapular region, not widely divergent, but running almost parallel, terminating at the end of the scapula. Branches of the lumbar fork arising midway between tlie anterior end of the ilium and the acetabulum ; widely divergent ; base of the stem forked poste- riorly, so as to embrace the oU-gland. Pt. ventralis. — Outer branch feeble ; median strong, recurved ; inner relatively broad and strong, arising just in front of the region of the anterior end of the carina of the sternum, terminating in front of the cloacal aperture. Pt. Inmieralis continued forwards to form the jd. ventralis at the summit of the shoulder. Pt. alaris : — Tectrices majores. — The 5th cubital covert of the dorsal surface slightly shorter than the 4th, but equal in length to the Gth. T. media;. — The 5th cubital covert of the dorsal surface much the longest in the series, the feathers rapidly lengthening from 1-5 ; Gth much shorter than 5, slightly shorter than remaining proximal coverts. The proportions of the rest of the coverts much as in A. accijntriiius. Pt.femoralis. — Ecmoro-crural baud terminating over the distal end of the femur above the knee-joint. Pt. o'uralis. — Acrotarsium covered with relatively short feathers. Feathers on the acropodium scanty, degenerated, confined strictly to the upper surface ; sides of toes bare, covered with granulated skin ; base of ungual phalanx covered by three scales. Rhamiihotheca. — Cere very short, much inflated. Nostrils pierced in its anterior border; aperture circular, looking directly forwards, and not laterally, as in other Owls. UrojJi/giiim napiform, with a tendency to a conical form, as in Buho. Scops giu, Scop. Pteryla capitis : — Loreal area not qviite so well defined as in S. leucotis, and the feathers not elongated as in that species. Interramal area with a narrow branch running obliquely backwards to join the ramal area behind level of the gape. E, a m a 1 area narrow, not extending backwards beyond the end of the mandible. Pt. colli dorsalis -with a lateral skin-fold and a short retractor muscle crossing the neck obliquely. Pt. s/j/»a//s.— Interscapular fork as in S. leucotis. Lumbar fork with the branches rising near the anterior border of the ilium and dividing at the base into two rows of semipluiues, one of w^hich runs out on to the axillary membrane, and one is continued upwards to join the free ends of the interscapular fork, enclosing an oval apterium. Pt. alaris : — Tectrices majores. — The 5th cubital covert slightly shorter than the 4th, but longer than the Gth. 244 ME. W. P. PYCBAFT ON THE Pt. femora J is. — Femoro-criiral band terminating below tbe knee-joint. Rhamphotheca. — Cere somewhat longer antero-postei'ioi-ly than in S. leucotis, and with a slight lateral swelling. Podotheca. — Acropodium covered Avith small round scales. TJropygium napil'orm, with a long tubular extremity resembling that of Qarine. NiNOX NOV^-ZEALANDI^, Gm.* Pteryla ccqntis : — L o r e a 1 area ill defined. C i r c u m a u r al area : — Postaural fol d. — The feathers along the free edges of this fold seated in a double row, the massed bases of which form a " rim." Interramal area suddenly contracting in the region of the gape, to pass back- wards into the p)t- coll. ventralis. Apt. colli laterale terminating at the base of the postaural fold. Pt. colli dorsal/s with a distinct vertical membranous fold; lateral folds very small. Pt. spinalis. — Branches of the interscapular fork arising in the middle of the inter- scapular region, short, terminating at the end of the scapula. Branches of the lumbar fork arising over the anterior ends of the ilium, not widely divergent, but received between the branches of the interscapular fork. The stem is forked posteriorly, but the branches are not sufficiently wide apart to embrace the oil-gland. Pt.femoraUs. — Femoro-crural band extending on to the ceres and turning upward to terminate at the femoro-tibial articultition. Pt. cruralis. — Acrotarsium thickly feathered ; acropodium covered with warty pro- tuberances bearing degenerate feathers, the shafts only of which remain. Filoplumes absent. Base of the ungual phalanx clothed with three scales. Jlhaniphotheca. — Cere laterally inflated, resembling that of Scops. Podotheca. — Claws short and blunt. Tylari of Digit II. large. Vropycjlum napiform, with conical tendency. ScELOGLAUX ALBIFACIES, Gray t. Pt. capitis : — Loreal area not extending upwards above the eye. Interramal area broad, expanding to merge with the ramal area at the gape. Cir cum aural area. — Preaural and postaural folds with a rim; no "notch" in the rim of the postaural fold. Apt. colli laterale terminating on circumaural fold, not extending on to the head. ? No parietal apterium. A large apterium above upper eyelid, as in other Owls. * This description is taken from a plucked specimen. t Examined frum a skin kindly lent by Prof. A. Newton, from tbe Cambridge University Museum. MOEPHOLOGT OF THE OWLS. 245 Pt. sjmialis.— Arms of interscapular fork long, arising higher up, not widely divergent. Stem of lumbar fork veri/ wide. Seven feathers, forked posteriorly. Arms short, Fis. 1. .if. Portion of the upper end of the lumbar fork of SceIo(/lau.r alhifiicies, to show the extreme breadth of the stem and short arms. i.f. Interscapukir fork. l.f. Lumbar fork. closely approximated, running almost parallel, at their distal ends expanding, so as to clothe the axillary membrane. Ft. veil trails.— Inner branch not sharply differentiated from median, distinguishable by reason of their more sparse distribution and semiplumous character. They form a narrow band on the abdomen. Apteria generally thickly sprinkled with semiplumes. Perhaps the most characteristic feature of the pterylosis of Sceloglaux is the great width of the stem of the lumbar fork. Mhamphotheca. — Cere slightly swollen ; nostril pierced in its anterior margin. Podotheca. — Feathers on the acropodium reduced to bristles. Scales on the acropodium more distinct than on the acrotarsium. Claw^. — The inner side of the claw of the middle toe with a flange. Syrnium alitco, L. (PI. 26.) Pteryla capitis : — Loreal are a. — Upper limb continued backwards to join the confluent feathers of the pre- and postaural tracts. C i r c u m a 11 r a 1 are a. — The preaurai fold is well developed and forms an operculum (see p. 229 and PL 28- figs. 1-2). It supports numerous strong, well-developed feathers, arranged in 4 jmrallel semicircular rows ; rows 1 and 2 are bound closely together, their bases forming a rim to the free edge of the fold ; rows 3 and 4s are placed further forwards. At some distance from the most anterior row, and separated by an apterium, is a single row running from the gape, backwards and iipwards, to the anterior corner of the base of the oiwrculum. This row is with difficulty distinguished anteriorly from the feathers of the ocular area. Postaural fold almost as cxtens.ive as m A. acci2)itrinHS. Likewise the feathers seated along its free edge are densely packed, and their bases form a rim to the free edge of 246 ME. W. P. PTCEAFT ON THE the fold. These feathers help to form the periphery of the disc, which is continued down- ward and forward to join its fellow of the opposite side heneath the jaw^ (PI. 26). Ramal area extends from the rhamphotheca of the lower jaw backwards to the level of the region where the opercular and postaural folds coalesce. The peripheral disc-feathers divide it into an upper and a lower portion (PL 26). Interramal area broad, sending off on either side a single row of feathers to join the H a m a 1 area in the region where it merges with the pt. colli ventralls. Apt. colli laterale extending on to the circumaural fold as far as the peripheral disc- feathers ; from its upper border, bounding the pt. colli dorsalis, it is continued as a very- slender space on to the pteryla capitis, terminating at a point somewhat behind the level of the region where the anterior ends of the pre- and postaural folds coalesce. Pt. colli dorsalis tapering from the head backwards. Lateral neck-folds of sUght extent ; wdthout retractor-muscles. Pt. spinalis. — Branches of the interscapular fork arising about halfway dowTi inter- scapular region ; arms short, 3 feathers in width, not extending to the end of the scapula. Branches of the lumbar fork represented ])y a single row of feathers arising far forward on a level with the ends of the scapula ; they embrace the arms of the interscapular fork. The stem of the fork bifurcated posteriorly, but terminating short of the oil-gland. Pt. ventralis. — Outer branch joined to the median by a single row^ of feathers, abruptly expanded near its middle to 3 feathers in width. Median branch broad, with a strong recurved limb springing from its free end — the " hook." Inner bi-anch broad, arising from the median at the point where the latter passes over the acrocoracoid region from the neck ; it follows the course of the furculum for some distance, then runs backwards to join its fellow of the opposite side in front of the cloacal aperture. Pt. humeralis. — Not connected anteriorly witli tlie pt. ventralis. Pt. alaris : — Metacarpo-digital remiges 11 ; the 11th a remicle, very small, little more than half as large as its dorsal major covert ; the 6th remex the longest, giving the wing a short, rounded appearance: the 5th to the 9th remex having the distal end of the outer vane more or less emarginate ; outer vane of 10th serrate. Cubital remiges 15. Tec trices : — T. major es. — Those attached to the dorsal surface of the cubital remiges of uniform length, save the 6tli, which is distinctly shorter than the 5th or 7th. T. medicr. — 1-5 of the cubital remiges of the dorsal surface gradually increasing in length, and distinctly the longest in the row. T. minores. — 1-5 of each row of the dorsal minor coverts markedly longer than the remainder of the coverts of their respective rows. Pt.femoralis. — AVell developed; femoro-crural baud strong, terminating on the crus, a short distance below the knee. Pt. criiralis. — Continued downwards over the acrotarsium and acropodium ; feathers long and thickly placed. Plauta having the proximal end bare. Penultimate phalanx of toes scale-covered. MOEPHOLOGY OF THE OWLS. 247 'Rhami^hotlieca. — Cere of considerable extent, very slightly inflated. Podotheca. — Claws rounded, not much curved. TJropyglmn. — Napiform, with a very marked tendency towards the elongate-conical form of the Buboniuas. Strnium uralense, Pall. The pterylosis of this species differs only in the following details from S. aluco. Pteryla capitis : — Inter ramal area. — With a lateral branch rimning at aright angle to join the ramal area in the region of the gape, as in S. aluco. A short distance behind this lateral branch it forms a broad expansion merging with the ramal area. Apit. colli laterals. — Extends upwards and forwards to the level of the angle of the exoccipital wing of the tympanic. Ft. spinalis. — Branches of the lumbar fork rising near the anterior border of the ilium, not widely divergent, running up to, but not quite joining, the free ends of the branches of the interscajmlar fork, thus enclosing a median s^iace. The stem not extending backwards so far as in ^S*. aluco, but terminating some distance in front of the uropygium. Pt. alaris. — The 6th major covert of the ciibital series shoi'ter than the 5tli ; the remainder of the other coverts moulting. " Carpal remex " penuaceous, as large as its covert, thus differing from that of S. aluco, in which this feather is semiplumous and markedly smaller than its covert. External aperture of the ear symmetrical, its vertical axis slightly longer than the longitudinal axis of the closed eyelid. Uropygium. — Napiform, with a long terminal tube. Ntctala tengmalmi, Gm. Pteryla capitis : — Loreal area. — Not sharply defined, with an apterium at its base, almost dividing the area into two portions, an interior and a posterior, the former including the greater part of the area. The apterium, which is very narrow, arises in the region in front of the eye, on a level with its anterioi- corner, and runs downward and then upward so as to partly encircle the eye, terminating rather above the level of its posterior corner. This apterium differs from that of Speotyto in that it fails to divide this area into two, in that it is less wide, and in that the posterior semi-detached patch is much less in size C i r c u m a u r a 1 are a. — The bases of the feathers seated along the free edge of the preaural fold form a rim ; lying in front of, and parallel with the rim, is a single row of feathers — as in Asio — banded on either side of the apterium. The feathers running along the free edge of the postaural fold are densely packed, their bases of insertion forming a " rim " to the fold. From the base to within a short distance of the superior limb of the fold the feathers seated on this free edge are directed second series. — ZOOLOGY, VOL. VII. 31 248 ME. W. P. PTCEAFT ON THE obliquely upwards ; the rest ai*e directed downwards ; heace the notch in the rim of the fold (PL 27. figs. 5-6). The j)eripheral disc-feathers (p. 236) are continued downwards and inwards to meet in the middle line a short distance behind the symphysis of the mandible. 1 n t e r - r a m a 1 are a. — Cut into two by the meeting of the peripheral disc-feathers ; the posterior portion somewhat triangular in form, its base planted against the confluent peripheral disc-feathers, its tapering apex merging with tlie pt. colli nentralis. Apt. colli laterals terminating at the base of the circiimaural fold. Pt. colli dorsalis with considerable A'ertical and lateral neck-folds. The tract decreasing rapidly in width from before backwards. Pt. sjniialis. — Interscapular fork bai'ely percejotible (PL 26). The lumbar fork, so far as my specimens are concerned, has the branches very feebly developed and disconnected at their bases from the main stem. This last is continued forwards so as nearly to join the interscapular fork (PI. 26). Pt. colli ventralis. — Very narrow at its upper end, where it joins the interramal area ; sends off two branches at the middle of the neck to join the pt. ventralis at the shoulder. Pt. ventralis. — Outer branch well developed ; median strong, recurved ; inner strong, not meeting its fellow in the middle line in front of the cloaca, but terminating over the free end of the pubis. Pt. Immeralis. — Not continued forwards to join the pt. ventralis. Pt. alaris : — Tectrices : — T. majores. — The dorsal coverts of the cubital remiges of uniform length, save the 6th, which is much shorter than the coverts on either side. The same proportions appear to obtain in the t. medice and t. minores. Pt.femoralis — Femoro-crural band turning sharply upwards to terminate over the distal end of the femur. Pt. crtiralis. — Aerotarsium and acropodium densely and thickly feathered ; penultimate phnlanx clothed with 3 scales. Phamphotheca. — Cere short, inflated, closely resembles that of Speotyto. TJropyginm. — Small, napiform, but the pointed extremity very short, making the gland almost triangular in form. Stjrnia ulula, Linn. Pteryla caintis : — Interramal are a. — Expanding at the gape to join the ramal area. Apt. colli laterale. — Terminating on the jiostaural membrane. Pt. colli dorsalis. — With an oblique membranous fold, from the pt. colli dorsalis to the centralis, similar to that of Asio, but more feebly developed. Pi. spinalis. — Interscapular fork barely perceptible ; the tract, which is broad, termi- nating midway down the interscapular region. The posterior moiety of this tract has a very broad anterior end, not differentiated into branches. MORPHOLO&T Oi^ THE OWLS. 249 Ft. ventralis. — The median branch is very broad posteriorly. Pt. alaris : — Tectrices. — The major coverts were moulting. T. medice. — The 6tli median cubital covert was much shorter than the 5th, as usual. T. minores — The same remarks apply to this series as to the t. medice. Pt.femoralis. — Femoro-crural band terminating at the knee-joint. Bhamplwtheca. — Cere short, slightly inflated. Podotheca. — Densely feathered. TJropygium. — Napiforin. Speotyto cunicularia, Molina. Pteryla capitis : — Loreal area. — Divided by an ajitei'iuin into two portions: an anterior, running round the base of the bill, and a posterior, encircling the anterior region of the eye. This latter is continuous at its lower posterior angle, immediately above the gape, with a single row of feathers running round the base of the lower eyelid and upwards to its posterior corner. The loreal apterium separates this row from the feathers of the ramal area. O c u 1 a r a r e a. — The single row of feathers above described as forming a continuation of the lower limb of the loreal area forms the lowermost boundary of this area. Above it, separated by a broad apterium, is another single row of feathers ; above this are two other rows, each separated by an apterium. The upper eyelid is clothed witli numerous rows of small, downy feathers. A. very considerable crescentic apterium sejiarates the uppermost rows from the featliei's of the fronto-parietal area. The rims of the eyelids are fringed with some 3 rows of feathers, decreasing rapidly in size as they approach the free edge. The hindmost are characterized by the production of the shaft outwards beyond the vane. In the outermost, nothing but the shafts remains, and these form " eyelashes." Circumaural fold : — Preaural fold (PL 28. fig. 5). — This fold is very feebly developed. Along its free edge is a single row of elongated narrow-vaned feathers. They are somew^hat curved, the concavity being turned towards the head so as to form a feathery operculum. Postaural fol d. — Of greater extent than the preaural. It supports, along its free edge, a row of feathers precisely similar to those on the opercular fold. They are over- lapped by the feathers of the preaural fold. (PL 28. fig. 5.) Ramal area not extending backwards as far as the angle of the jaw. Inter ramal area. — Anteriorly, from the region at the level of the gape to the symphysis of the mandibles, surrounded by an apterium, the posterior portion expanded so as to merge with the ramal area. Apt. colli laterale terminating at the base of the lower half of the circumaural fold, and embracing the region of the angle of the mandible. Pt. colli dorsalis. — Borne upon lateral folds, not closely investing the neck. Pt. sjmialis (PL 26). — The arms of the interscapular fork, arising in the middle of 31* 250 MR. W. P. PTCRAFT ON THE the intersca]iular region, very short, terminating a short distance from the free end of the scaj)ula. The arms of the lumhar fork represeated hy a single row of feathers arising over the anterior end of the ilium and which, crossing over the free end of the scapula, terminate on the axillary memhr'ane. The stem of this portion very distinct ; its free end is bifid, and terminates some distance from the oil-gland. Pt. ventralis. — Outer branch feebly developed, represented by a single row of feathers ; median liranch broad, short, not extending beyond the middle of the sternum, the recurved branch or " hook " very slender ; inner branch, arising in the middle region of the furculum, follows the course of the clavicle and the carina, tlien turns inwards and upwards to the thigh, whence it takes an almost straight course backwards to join its fellow of the opposite side in front of the cloacal aperture. Pt. femoralis. — Pemoro-crural band terminating at the knee-joint. Pt. criiralis. — Feathers on the acrotarsium short. Planta bare. Acro[)odium covered by numerous warty protuberances, each of which bears a degenerate feather ; little more than the shaft, with perhaps — here and there — one or two rami, remaining ; radii absent. Filoplumes absent. The base of the ungual j)halanx clothed with scales. Pt. alaris : — Bemiges. — No carpal diastema, the space between the metacarj)al and cubital remiges being very much less than that separating the individual cubital remiges. Tectrices -. — T. mujores. — The major coverts on the dorsal sui'face gradually increasing in length from 1-5. This last the longest ; 6th shorter than 5th. T. mediae. — 1-5 gradually increasing in length, but the 5th abruptly longer than 1-4. 6th about as long as 4th. The 5th feather more closely bound to its covert — 5th major covert — than the remaining feathers to their respective coverts. T. minores (row no. 1). — Covert 5 abru2)tly long(>r than 1-4, but of the same length as 6. Disproportion in the length of the coverts of this group, above the 1st row, less marked. Mhmivphotheca. — Cere very short antero-posteriorly, much inflated, forming two pisiform swellings on each side of the culmen. Aperture of the external naves circular, dkected more forwards than outwards. Rhamphotheca compressed. Podotheca. — Claws short, very little curved. TJropygium. — Napiform. Carine noctua, Scop. Pt. capitis. — Fronto-parietal and occipital areas densely feathered. Loreal area indistinct. Interramal area expanded to join the ramal area on either side, from tlie level of the gape backwards. Apt. colli laterale terminating at the level of the posterior limb of the external aperture of the ear. Pi. colli dorsalis borne on a deep vertical fold ; no lateral folds. MOEPHOLOGT OF THE OWLS. 251 Ft. spinalis with a short interscapular fork, the branches arising below the middle of the interscapular region, short, widely divergent, tapering, terminating rather below the free end of the scapula *. The stem of the posterior portion not forked, but running sti-aight up to the convergent bases of the arms of the interscapular fork. Widest posteriorly, forked terminally ; the branches short, not wide enough apart to embrace the oil-gland. Pt. colli veniralis. — Forks almost immediately after its origin. Pt. centralis. — Outer branch very narrow ; median broad and strong, terminating in the middle region of the sternum. The recurved branch (" hook ") very feeble. Inner branch arises in the region of the middle of the furculum and runs in the dii'ection of the carina of the sternum ; over the middle of the sternum it trends outwards and upwards, again converging on the abdomen, each branch terminating separately some distance in front of the free end of the pubis and not meeting its fellow in the middle line in fi'ont of the cloacal aperture. Pt. Immeralis. — Connected with the pt. veiitralis by a single row of feathers. Pt.femoraUs. — Femoro-crural band terminating over the proximal third of the crus, not running upwards to the femoro-tibial articulation. Pt. cruralis. — Feathers of acrotarsium barely concealing the skin ; on acropodium degenerate, shafts only remain. No filoplumes. Planta feathered ; the base of the terminal phalanx of the toes covered by 3 scales. Uliampotlieca. — Cere short, inflated, forming two pisiform swellings on either side of the culmen as in Speotyto. TIropygium. — Napiform, the tubular extremity relatively long. Strix plammea, Linn. Pteryla capitis : Fr onto- parietal area. — In the form of a narrow, median tract of feathers, bounded on either side by an apterium. Occipital are a. — Occujoies the whole width of the skull, extending outwards on to the postaural fold. Loreal are a. — The base almost obliterated by a semioval apterium lying along the margin of the tomium from the region of the nostril backwards to the gape. The inner boundary of the apterium is formed by a single row of feathers running from the loreal area backwards below the eye to join the feathers of the operculum. C i r c u m a u r a 1 area (PI. 28. tig. 1, p. 229).— The preaural fold is well developed and forms an operculum. It is clotlied by some I vows of delicate, elongated feathers, having narrow, discontinuous vanes. Postaural fold. — Its upper portion may be considered to merge into the upper limb of the loreal area above the region of the middle of the eye; below it passes insensibly into the ramal area (see " Periphery of the Disc," p. 252). * 111 a young iudividual the interscapular termination of this tract was very narrow — out littlo wider than tie stem. The branches were very short, and parallel, not divergent. 252 MR. W, p. PTCRAFT ON THE Interramal are a. — The only specimen of a Barn Owl tvhicli I have been able to procure I owe to the kindness of Dr. Sclater. Unfortunately it is so damaged in this region that Mr. Meinertzbagen and I have been obliged to fall back upon Nitzsch's inter- pretation and restore the figure after that given by him. P e r i ph e r y o f t h e D i s c. — This is more sharply demarcated than in any other Owl so far examined. The feathers taking part in its formation are seated in a deep fold of skin which stands out abruptly from the general contoiir of the head in the form of a sharp ridge. Its anterior portion is contribvited by the superior border of the loreal are;i ; running backwards into the postaural fold, and downwards along the interior border of the mandible, it finally dies out in the region corresponding to the level of the gape. The feathers in the immediate region of the external aperture of the ear are very long, witb narrow, discontinuous vanes, and all much curved forwards, so as to form a hollow around the ear, and a very considerable lateral extension of the postaural fold. Apt. colli lutcrale. — Terminating at the base of the jjostaural fold, some distance below the level of the external aperture of the ear. PL colli dorsalis. — Not borne upon lateral neck-folds, but closely investing the neck, therein differing from all other Owls so far examined. It forms a diamond-shaped expansion in the middle of the neck. Tt. spinalis. — Interscapular fork arising within the interscapular region. Arms narrow, 3 rows of closely-packed feathers, not extending as far as the free end of the scapula. The arms of the lumbar fork arise over the anterior end of the ilium, cross the free end of tlie scapular, and terminate on the axillary membrane. The stem is truncated posteriorly, terminating some distance in front of tlio oil-gland. PL colli veiitralis. — Bifurcating on the lower \ of the neck, branches very narrow; feathers set in oblique rows of 4 feathers in each row. PL ventralis. — Outer branch moderately broad. Median branch narrow, not free posteriorly, but continued backwards to join the inner branch over the region of the posterior lateral process of the sternum. The "hook" is given off just before this junction takes place. Inner branch very distinct ; coming off from the median opposite a point corresponding with the region of the ventral \ of the furculum, it is continued backwards to join its fellow of the opposite side over the free ends of the pubes. PL alaris : — Bemiges. — Metacarpo-digitals 11, cubitals 15, the innermost not readily distinguished from their coverts. Teetrices. — For i\\Q most part missing in this specimen. Pt.femoralis. — Femoro-crural band terminating below the knee. PL cniralis. — Feathers on the acrotarsium degenerate, the shaft only remaining. They occur in groups of 3, of which ,the centre one, as described by Meijcre, is much the longest; the two lateral shafts being in most cases minute, and sometimes absent. Each group of 3 is seated under the anterior border of a soft, swollen scale. On the aero- podium the scales become firmer and the feather-remnants fewer. BliampJiotheca. — Cere closely investing the base of the beak, nostrils pierced in its lower anterior border. MORPHOLOGY 0¥ THE OWLS. 253 Podotheca. — Planta feathered, save the proximal end. The feathers on the posterior border of the acrotarsium directed upwards. Scales oa the toes more distinct than on the acrotarsium. Pectination of the claw of the middle toe very slightly developed. IV. The Distribution of the Neossoptiles or Nestling-Down. I hoped to liave been enabled to describe tbe nestling-down of Aslo, but after all failed to procure a specimen of this genus. By the kindness, however, of my friend Mr. Meiuertzliagen, a substitute has been provided in a nestling Speotijto cnnicularia, which was one of a clutch of 5 bred in the Mottisfont aviaries this spring (1897). There are two kinds of nestliug-dowu to be distinguished : (1) the down-feathers ■which precede the future definitive contour-feathers, and (2) those which precede the future definitive down-feathers. This is a j)oint not to be overlooked, inasmuch as the latter, as I have already shown in the case of Op'tsthocomus (14 a), may, from the greater length and number of the rami, bear the greater share in clothing the young bird. Just as in Opistlwcomus, the nestling-plumage of tlie Common Kestrel, Falco tiiiimn- culus, is mainly contributed by the down-feathers belonging to the second kind, the down-tufts preceding the contour-feathers being relatively insignificant in size. One might naturally expect to find that this two-fold down-clothing would obtain wlierever coutoiu-- and down-feathers were to be found associated together in the adult. This does not, however, appear to be the case, inasmuch as I find that the nestling-down of the Common Duck, for instance, is entirely composed of the down-feathers of the first order — those preceding the contour-feathers. I would now suggest that these two kinds of nestling-down should in future be carefully distinguished when describing the plumage of nestling-birds. The feathers preceding the contour-feathers might be called pre-ioenncB, whilst those preceding the future definitive down-feathers might be called pi^e-plumulte. In the nestling of Speotyto the feathers investing the trunk are pre-pennae ; pre-plumulae are absent. As is shown elsewhere, the corresponding plumulse are absent on the trunk of the adult also. The clothing of the wing differs from that of the trunk in that pre- plumulse occur amongst the pre-penna?, hui they are very small and contribute no great part towards the general covering. Whether these last are really pre-plimiulae or the actual definitive down-feathers I am unable to decide. Tlie question can be settled only by the examination of younger specimens, the one under consideration being half-tledged. Save the remiges and their dorsal major coverts, and the rectrices, the true contour- featliers of the adult do not seem to be devc-loped till after the first moult. All the nestlings which I have been enabled to examine — ranging up to birds nearly full-grown — have been clothed with a peculiar loose semiplumous covering, sprinkled here and there, in some cases, with typical adult contour-feathers. 254 MR. W. P. PTCEAFT ON THE V. Ptertlosis or the Embeyo. Asio OTTJS, Linn. My embryos of this species may be divided into 3 stages, tbe last corresponding to that of a 10-day chick. In this, only the rudiments of the contour-feathers have made their appearance in the form of the characteristic bead-like papillae. Rio'ht side view of an embryo of Asio oins, showing the form of the pterylse. At this stage the pteryla colli dorsalis closely invests the neck as in Sirix Jimnmea. The external aperture of the ear has not yet reached its full size, and there are no pre- and postaural folds. The beak has not yet assumed its characteristic curved shape. Stage III. FterylcB : — Ft. ccqdtis. — The feathei'ing of the upper part of the head, as indicated by the feather-rudiments, is very uniform, giving no sign of the differentiation into areas — loreal, fronto-parietal, &c.— which appear in the adult. I n t e r r a m a 1 area as in the adult. Ft. colli dorsalis. — There are no lateral folds, the skin closely investing the neck, as in StrLv Jlammea. Ft. spinalis as in adult. Ft. ventralis. — No outer branch is yet visible. Ft.femoralis. — As in adult, femoro-crural band strongly marked. Ft. cruralis. — No feather-rudiments have yet made their appearance on the podotheca. Ft. alaris. — The position of the future remiges, and nearly all the tectrices, are now indicated, and the wing is seen to be aquintocubital. The external aperture of the ear has not yet reached its full development. There are no opercular or postaural folds. Bhamphotheca. — Note the straight form of the beak and the absence of any indication of a " cere." MOEPHOLOGY OF THE OWLS. 255 Clmos. — There is a small claw on the pollex ; there is also an indication of a claw in Digit II. Stage II. Pt. coj^ntis. — Only the feathers of the loreal area and the region about the eye yet indicated. The position of the trunk-feathers is almost as plainly indicated as in Stage III. In the pteryla alaris, however, only the remiges and dorsal major coverts are yet suggested. Stage I. The skin as yet perfectly smooth. Carine brama, Temm.* This embryo almost exactly corresponds to Stage III. of Asio accipitrUms. Pterylte : — Pf. capitis. — ^The feathering of the upper -part of the head not differentiated into distinct areas. I n t e r r a m a 1 are a. — Expanding cit the level of the gape to join the ramal area. Pf. spinalis. — The interscapular fork differs markedly from that of C. noctua (p. 250), in that there are well-marked interscaj)ular branches present as in Asio, Scops, &c. ; they terminate over the region of the free end of the scapula. In the adult mentioned the interscapular fork is barely perceptible. There is a well-marked lumbar fork, the branches of which arise near the anterior border of the ilium, are widely divergent, and extend on to the axillary membrane. In this also this species differs from the adult C. noctua, in which there is no lumbar fork. Pf. femoralis. — Pemoro-crural band terminates at knee-joint. Both digits, I. and II., of the wing are furnished with claws. Nyctala Tengmalmi, Gm. Stage IV. — Ai*eas of feather-distribution as in the adult. There is a claw on both poUex and index digits. Tlie external aperture of the ear is oval in form ; its vertical axis is nearly twice that of the longitudinal axis of the closed eyelid ; it is entirely closed by the sclerotic ring, which comes quite to the surface. Stage III. — The form of the interramal tract can can be weU studied here. It is very broad on the throat, and narrows suddenly on the upper part of the neck. Pt. spinalis. — The lumbar stem appears to be continued forwards, in the form of two single parallel rows of feathers, to join the free ends of the interscapular branches. It is cleft in front of the oU-glaud. Stage II. — The papillsje are just appearing ; their distribution appears to be the same as in Staa;e III. ^o^ Strix flammea, Linn. The embryos of this species represent Stages III. to V. That corresponding to Stage III. is most instructive, and best preserved ; and from this the following description is taken, the remaining stages being compared with this. * For this most beautifully-preserved specimen I have to thank my friend Mr. .Jesse of Lucknow. SECOIsD series. — ZOOLOGY, VOL. VII. 35 256 ME. W. P. PTCRAFT ON THE FterylfB : — Pt. cajyft/s. — The papillne which will give rise to the peripheral disc-feathers are fairly sharply defined from the rest of those covering the head, as a narrow baud of closely-set raised points. The aperture of the ear is nearly circular ; its vertical axis is less than that of the longitudinal axis of the closed eyelid. There is no trace of an operculum. Ft. spinalis. — Interscapular fork with the branches very long, and joining those of the lumbar fork so as to enclose a space. At the junction of the free ends of the lumbar fork with those of the interscapular fork, a side branch is given off from each to the posterior end of the Pt. humeralis. Stages IV. and V. agree, so far as the distribution of the feathers is concerned, with Stage III. In Stage V. the first trace of the operculum appears. VI. Structure op the Feathers. Contour-feathers. — These, in the Ow^ls, are of considerable length, and soft and some- what loose in texture. There is no aftershaft. The remiges in many, e. g. Asio, Bubo, present on the upper surface of the vexillum a very characteristic velvety pile, due to the enormous elongation of the free ends of the distal radii, which extend forw^ards over as many as three rami, not including the ramus from which they arise. The feathers composing the facial disc present a few points w^orthy of notice. Round the free edge of the operculum they have an elongated and rigid shaft, bearing numerous moderately long rami, which form a discontinuous vexillum. The most distal rami break up into about three short branches, all of which bear fila, thus giving the tips ■ — under the microscope — a curiously frayed appearance. The radii are very short, set almost parallel with the ramus, and showing no more than traces of fila. They also occur on the shaft in the internodes between the bases of the rami. A feather from the lower limb of the postaural fold shows the following points : — The shaft is bent upon the calamus at a considerable angle in an upward and lateral direction. The calamus is abruptly distinguishable from the scapus by its greater thickness. The inner vexillum is roughly triangular in form, the base extending from the superior umbilicus to the top of the stem ; the outer vexillum is much smaller and also triangular, its base running from the upper umbilicus to the middle of the stem. The rami are entirely disconnected and wadely separated. The radii are very short ; those of the proximal series lie parallel with the ramus ; the distal radii are longer. Fila are absent. Short radii run along the scapus in the interval between the rami. The feathers of the Loreal area are long, with fairly stiff, rigid axes. Rami very short, degenerate, showing neither fila nor booklets, and decreasing in length on each side of the ramus from the base upw'ai'ds, finally disappearing. They also run along the internodes on each side of the shaft from the base of one ramus to that of the next in front. Plmmilce. — In these the rhachis is relatively short as compared with the rami, which are of great length. The radii are long, transparent, and divisible into segments — nodes and internodes — by the deposition of pigment in the region of the fila. The fila MOEPHOLOGY OF THE OWLS. 257 occur in pairs (fig. 3, G) at the proximal nodes ; beyond this thej^ suddenly disappear, but the corresponding" region is marked by a gentle swelling containing a dark pigment, which in masses appears black. The adult down-feathers of Asio, Speotyto, and otlier Owls closely resemble those of the Kestrel {Falco t'mmmculm, fig. 3, E) ; but they may be distinguished by the following points : — In the latter the nodal areas of tlie radii are more numerous, the Fig. 3. B --^rr/^^- 1.. -'W- - J»iTf??r?r?? ^^^^^m^ ^^^ — -— A, portion of a rndius from a defini- tive down -feather of Capri- multjus eurupa'us. The swelling of the nodes is somewhat ex- aggerated in the figure. B, por- tion of a radius of a pre-peniia of C. e.uropcms. C, portion of a radius from a pre-plumula of Falco tiiinunculus. D, portion of a radius from a pre-penna, and E, portion of a radius from a definitive plumule of F. tin- nuncxlus. F, portion of a radius from a pre-penna, and G, por- tion of a radius from a definitive plumule of Spxotyto cunicularia. f, filum. ,/ rhachis is much longer, and there is a vestigial aftershaft. They may be very readily distinguished from those of Caprimulrjus (fig. 3, A) by the greater length of the radii in the latter, the long black nodes — longer than the internodes and very little thicker, — and the absence of fila. Like the Kestrel, the down-feathers of Caprimulgus possess a long rhachis and a vestigial aftershaft. Filophmies. — In the adult these present no special features of interest. The rhachis is long, bearing from two to three rami (fig. 4). The radii are of moderate length, and bear, here and there, a few fila. The only case in which I failed to find filoplumes in the adtdt was that of Speotyto cunicularia. 35* 258 ME. W. P. PTCEAFT ON THE In the nestling, however, they were readily visible in the form of somewhat conspicuous packets, one at the base of every contour-feather. Their examination under the microscope revealed some very interesting facts concerning the nature and history of these curiously modified feathers. As will be seen in fig. 5, each at this stage consists of a sheath investing a number of rami, and a rhachis, clustered around a column of " pith-" cells. Before the full growth of the feather is attained everything save the Fis. 4. Fig. 5. Fig. 4. — The distal end of a filojjlume of jS'co^js leucotis. Fig. 5.-Filoplume of a nestling Siuotuto mnicularia. The part marked / is all that remains in the adult, and forms the filoplume ; T, the transitory radii ; j*, pith-cells. main axis is thrown off. This remains as the typical filopkune of the adult. Dr. Gadow (4} remarks, "their development shows them to he degenerate, and not primitive feathers." Neossoptiles.—Ks is pointed out in another part of this paper (p. 253), two kinds of neossoptiles must be distinguished — (1) pre-pennse and (2) pre-plumulae. Pre-penn£e * can be readily distinguished from pre-plumulce only under the microscope. The component rami of these umbelliform down-tufts, being continuous with those of the definitive contour-feathers, are broken up and spread out horizontally as soon as the tips of these latter have burst through the sheath by which they were enclosed. * The pre-pcnna3 described arc those of Sptotyto mniculana. MOEPHOLOaT OF THE OWLS. 259 The radii — like the rami — are transparent. Fila are generally but not always present. The rami are laterally compressed. Structurally they can scarcely be distinguished from those of Falco tinnunculus ; but in this latter the fila are longest and regularly arranged along the radius (fig. 3, D). The pre-pennse of Capriniulgm can readily be distinguished from either Speotyto or Falco tinminculus. As in these, they are umbelliform, but the rami are produced con- siderably beyond the most distal radii into long filamentous processes, and the radii bear minute fila at regular intervals (fig. 3, B). The pre-plunuda; differ from the pre-pennae, both in Speotyto and the Kestrel, in that the radii are divided np into segments by the collection of pigment in the region from which the fila arise. In this respect they resemble the definitive down-feathers, but can always be distinguished from them. At least this is true of the Kestrel. Thus, in this species, the pigment, though in the form of a dense black deposit in the region of the fila, shows a tendency to be distributed all along the radius (fig. 3, C). Eurther, the pre-plumulse are umljelliform, whilst the definitive plumulae have a rhachis and after- shaft. From the fact that the feathers which I have doubtfully described as pre- plumulae in Speotyto have a distinct rhachis, and the radii are indistinguishable from those of the adult, it is more than probable that these feathers are really definitive down-feathers. Both the pre-plumula3 and the definitive down-feathers of Speotyto — if these, in this case, are not one and the same — can be distinguished from the plumulae of the Kestrel in that, in the latter, the segments of the radius are more numerous (fig. 3, E) and the rhachis is longer and supports an aftershaft. VII. The External Ear. (Plates 27-28). The external aperture of the ear of birds is usually concealed by feathers. If these be raised, a small round or oval opening will be seen leading into a more or less spacious chamber — the outer portion of the auditory meatus. This chamber is continued inwards in the form of a tunnel to the tympanum — the boundary-Avall between it and the middle e;ir. It will be helpful, for clearness' sake, in the present paper to call the cavern-like chamber communicating directly with the outer world the Cavernum, and the tubular inner continuation the Cavornulum. In Athene noctua we have apparently one of the least specialized meatuses of all the Owls. The external opening is almost circular in form and very small, its vertical axis being about half the height of the horizontal axis of the closed eyelid. The cavernum is spacious, and leads backwards and downwards to pass by insensible gradations to the cavernulum. In other words, it represents the typicalimspecialized form of the external ear in birds. Erom this ue can pass by a series of slight gradations to forms of con- siderable relative complexity, as will presently be shown. "What has been described for Athene will answer almost equally well for Speotyto. Bubo lacteus leads us a step further ; at the same time it represents the simplest form of the external ear to be found in the whole genus save that of B. nycteus. The 260 ME. W. P. PYCEAFT ON THE aperture is almost circular ; its vertical axis is rather less than the horizontal axis of the eyehd. The cavernuui gradually deepens from aliove downwards and backwards. Its floor is divided into two deep pits by a vertical fold of skin investing the second portion of the temporalis muscle which runs from the postorhital process to the corouoid region of the lower jaw. In Bubo nycteus the external aperture is pyriform in shape, the base being directed upwards and sloping oliliquely back^vards ; its vertical axis is equal to the horizontal axis of the eyelid. The cavernuDi is sjjacious, and closely I'esembles that of B. macnlosits. It differs, however, in this respect, that the eye projects more laterally, and has given rise to a shallow depression immediately behind it, at the bottom of which lies the mouth of the cavernulum. In Bubo ignavus the aperture is oval, its vertical axis equal to the horizontal axis of the eyelid. The depression beliind the eye just described is here much more marked, so that the cavernum is divisible into an anterior and a posterior portion. It is further- more continued upwards between the skin and the skull to the crown of the head, and backwards between the postaui'al fold and the skull : thus it has gained a considerable increase in size. B. virginianiis differs from B- ignavus in that the aperture is somewdiat smaller, and the anterior and posterior divisions of the cavernum still more sharply marked. B. maculosus agrees with B. ignavus in the' size of the aperture, but differs from this species and agrees with B. vlrgiuianus in that the cavernum is divisible into two portions, the posterior division being even deeper than in this species. Bubo torqimhis agrees with B. virginiaims in the size of the aperture, but more nearly approaches B. muculosus in the form and size of the cavernum. Scops and Bubo appear to closely resemble one another in the form of the external ear, so far as can be judged from an examination of two species of the former genus. The posterior division ditfers from that of Bubo in that it extends the whole vertical heirfit of the cavernum instead of being confined to its lower third. In S. leucotis the aperture is semilunar, and its vertical axis is equal to the horizontal axis of the eyelid. There is a considerable extension of the cavernum iipwards to the crown of the head, between the skull and the skin, and between the postaural fold and the skull. In S. giu the vertical axis of the aperture is only half the horizontal axis of the eyelid. The anterior portion of the cavernum is very shallow, being almost filled by the eye. The transition between the anterior and posterior portions is sudden, aud the latter very deep. The floor is spacious. Ninox noccs-zealandice has an almost circular aperture, the vertical axis of which is nearly equal to the horizontal axis of the eyelid. The cavernum closely resembles that of Scops. In Sceloglau.i: albifacles, so far as I could make out from a relaxed skin (p. 244). the aperture of the ear was semilunar, equal to, and probably greater than, the horizontal axis of tIic eyelid, iiazardiug a guess, I should say that the cavernum probably most closely resembled that of Bubo. Mr. J. H. Gurney (7), after the examination of a MOEPHOLO&T or THE OWLS. 261 specimen soon after death, gives the measurements of the external aperture as i~2 inch, and describes it as of an elongated shape. " The ears," he writes, " are very low in the head, in fact placed beneath the eyes, and appear, from external examination, to be quite symmetrical ; but after the bird was skinned I could see, by looking at the skull from the back, that the hind angle of the inferior mandible was lower on one side than the other." Sijrnia ulula has a nearly circular apertnx'e, the vertical axis of wliich is equal to the horizontal axis of the eyelid. The cavernum is spacious, divisible into anterior and posterior portions, sharply defined ; the latter very deep, and passing insensibly into the cavernulum. There is a considerable upward extension of the cavernum between the skin and skull as far as the crown of the head. So far we have been discussing forms of the external ear which are perfectly symmeti"ical on the two sides of the head. In the forms presently to be described there is a marked asymmetry, with exceptions, however, which will be noticed in their place. The first and most remarkable of these is that of Nyctala Tencjmalmi. The mem- branous aperture is oval in form ; its vertical axis is twice the length of the longitudinal axis of the eyelid. The preaural fold — the loose flap of membrane formino- the anterior border of the apertm-e — forms an imperfect operculum (PL 27. figs. 5-6). The cavernum, as in the forms just described, is divisible into an anterior and a posterior portion, the latter being in this case by far the larger. By a considerable devcloijment of the lateral wing of the squamosal and exoccipital bones, the form of the meatus has been considerably modified, and is, as just hinted, asymmetrical on the two sides of the head. On the left side this wing takes the form of a thin bony plate jutting out from the wall of the cranium from behind, outwards, downwards, and forwards to the level of the post- orbital process, so as to form a complete bony outer wall to the posterior portion of the cavernum. It is continued downwards and forwards to overlap the articulation of the lower jaw. The bony floor of the cavernum is incomplete, being hollowed out or " cut away " to expose the quadrate from behind. Thus the posterior portion is converted into a deep recess roofed over by bone, and looking outwards and forwards on the anterior portion. On the right side the squamosal Ming does not extend along the cranium as far forwards as the base of the postorbital process, but trends outwards and forwards to join the outer border of its lower end, instead of being continued downwards to overhang the articulation of the jaw as on the left side. Thus the posterior diAasion of the cavernum is converted into a deep recess, differing from that of the left side in that it is open above and that it is shut oiT below from the outer world and from a portion of itself by the junction of the squamosal wing with the postorbital process. The portion cut off, just referred to, is represented by the deep hollow over the superior border of the articular end of the jaw, and is well shown in fig. 6, PL 27. 262 ME. W. P. PTCEAFT ON THE In Asia the asymmetry does not extend to the bones of the skull, hut is confined to certain membranes uithin the cavernum. In A. accipitrinus (PI. 27. figs. 3— i) the membranous aperture is crescentic in form and of great size, extending from the middle of the lower jaw, from a point corresponding to a line drawn from the base of the cere downwards, backwards, and upw^ards to a point over the region of the middle of the eye. Thus the cavernum has gained an enormous superficial area, such as has been liinted at in other Owls, in the extension of the cavernum over the skull, and backAvards round to tlie occiput. The pre- and postaural folds are very A'olumincnis, the former being kno\\Ti as the opercidum. That portion of the cavernum which lodges the eye, and the cavity behind, corresponds to the anterior and posterior divisions of the cavernum in other species. The latter division is here much increased, and, by reason of the disposition of certain membranes, has come to be asymmetrical as regards the two sides of the head. On the right side this will be seen to be divided into an upper and a lower chamber by means of a membranous horizontal partition running from the squamosal wing to the eye, and forward to the operculum. The upper chamber is blind, and may therefore be called the diverticulum. It is possible that it may act as a sort of resonator. The lower, which should, I would suggest, be called the via recta, is very spacious and leads to the caverntdum. It is bounded above by the floor of the diverticulum; behind by the squamosal wing ; below by a thin membrane from the postaural fold to the operculum, running immediately above the superior border of the lower jaw ; and in front by a thin membrane from the lower jaw to a bony tul^ercle on the postorbital process. On the left the diverticulum lies below the via recta. The membrane dividing the two is placed nearer the lower jaw, and sloj^es oljliquely downwards and inwards. It is also partly attached to the tubercle of the postorbital process just referred to — which tubercle, by the way, occurs o)ily in this species. The ear of ^. otus differs from the species just described in that there is no postorbital tubercle, that the via recta of the left side is without the anterior boundary-membrane, and in that the diverticulum of the cavernum is narrow and deep instead of broad and shallow. Syridmn furnishes iis with a third type of modification of the external ear. In .S. aluco it is asymmetrical, but the asjTnmetry is confined to the membranous aperture, which is semilunar in form. The vertical axis of the aperture of the right side is equal to the horizontal axis of the eyelid ; that of the left aperture is less. The cavernum is divisible into anterior and posterior portions, the former larger and shallow'er, the latter deep, but confined to the lower half of the posterior corner of the cavernum, the floor of which is fairly spacious. The upward continuation of the cavernum over the skull is but slight. The pre- and postaural folds are large, the former, constituting the operculum, overlapping the free edge of the postaiu-al fold wdien the ear is closed. S. uralense is generally believed to be asymmetrical as regards the form of the external ear, and, moreover, the asymmetry is further supposed to extend to the bones of the skull. A specimen in the flesh from my late friend Mr. Meinertzhagen's aviaries MORPHOLOGY OF THE OWLS. 268 certainly does not contirm these beliefs — at least so far as the soft parts are concerned ; which differed from those of S. ahtco only in that the vertical axis of the aperture was slightly greater than the longitudinal axis of the eye, and that there was no asymmetry, the size of the aperture being the same on each side of the head. Strix Jlammea concludes the list of those which 1 have been enabled to examine. The, aperture of the ear is almost square ; its vertical axis is equal to half the longitudinal axis of the eyelid. The preaural fold constitutes an operculum. The cavernum is not divisible into anterior and posterior divisions, but forms a deep cavity leading almost directly into the caverntdum. The operculum, arising a short distance behind the gape as a low ridge, rapidly expands into a square flap terminating some distance above the free edge of the aperture on a level with the top of the eye. The base of the operculum measures about twice the length of the aperture of the ear. The loose membranous postaural fold supporting the peripheral disc-feathers of Asio otus is here represented by a low dermal ridge standing up distinctly from the surface of the head. It arises on the inferior border of the lower jaw, and running backwards, upwards, and forwards, terminates at the base of the beak. This ridge, as previously intimated, serves to support the peripheral disc-feathers. These are much elongated, and narrow-vaned, standing out at right angles to the head, and curved so as to present a deeply concave surface forwards, and a corresponding convex sru'face looking backwards. From their shaj)e and disjjosition these feathers afford an admirable apparatus for the collection of sound. It is not surprising to find that the operculum varies slightly in form. In one specimen (of two examined) the lateral border was shorter than that described above ; and in consequence the superior border slopes gently upwards, giving the operculum a truncated- conical rather tlian a square outline. It would seem, from these facts, that the form of the external ear was at some time subject to considerable variations, the most successful of which have become more or less fixed by selection. Results. That a detailed study of the Pterylography of the Owls would lead to any very startling results was hardly to be expected, and, as the " Keys " at the end of this paper show, it does little more than confirm wliat has alreadv been done. It must not be forErotten, however, that these results have been obtained by a deliberate desertion of the beaten track into the neglected highways and byways opened up by Nitzsch, but sinc<^ allowed to sink almost into oblivion. I have been enabled to push Nitzsch's methods somewhat further than lie himself attempted, for it will be remembered that he founded his genera rather upon the data obtainable by the time-honoured methods of measurements and the relative proportions in the length of the remiges, their number and general outline, and so on, rather than upon pterylological characters, which seemed to him to differ little, if at all, from his type Bubo. SECOND SERIES. — ZOOLOGY, VOL. VII. 3(1 264 .MR. AV. P. PYCEAFT ON THE That the common resemhlances which pervade the whole group are many, no one mil deny, but that nevertheless every species possesses some slight departure of its own from the main tyj)e, which, judiciously combined with other external characters, such as the form of the external ear and of the cere, will absolutely distinguish it from every other, is a fact which T believe I have been the first to point out ; it now awaits the test of criticism from other quarters. On the strength of the evidence brought to light during the coarse of the present investigation, I have felt justified in removing certai" genera from their moss-grown pedestals and placing them elsewhere, or even suppressing their claims to generic rank altogether. Sometimes this has been but an act of restoration, sometimes of innovation. This spiriting away of genera has led to the complete isolation of forms generally closely associated : thus, Asio, Syrninm, and Nyctala have been regarded as closely allied by such authorities as Prof. Newton, Dr. Sharjje, and Mr. Ridgway for instance, on the strength of the possession of an operculum and the size of the external aperture of the ear ; nevertheless, I ventixre to think that this relationship is only apparent. The genus Syrnium of Dr. Sharpe's Catalogue, again, probably includes several species which will have to be similarly separated. I infer this from the fact that his Syrnium 2iers2ncillatum — the Bubo torqnahis of the present paper — belongs undoubtedly to the genus Bubo, since, apart from the absence of an operculum and the small size of the external aperture of the ear, its pterylosis alone shows it to be much more closely allied to this genus than to Syrimmi. The real position of Sceloglaux and Ninox has proved a bard nut to crack, the former especially so. In its general pterolysis it closely resembles Bubo, in the form of the cere and external aperture of the ear it resembles Scops, while it differs from both these genera in the great width of the lumbar division of the spinal tract ; but whether on this account it is entitled to generic rank is open to question. Certain points in the osteology of the skull — to be discussed in my next paper — would have inclined me to place it in the Nyctalincc ; and it remains to be-seen whether, when the rest of its anatomy has been worked out, these suspicions will be confirmed. Bubo nycteus, again, has proved rather a stumbling-block. Is this, or is it not, \\ orthy of generic rank ? To jilace Carine and Speotyto in the same genvis would, I fear, rouse some adverse criticism. I am sorry, but at present this seems necessary. After I have completed my study of the two forms it is jiossible that I shall find that, after all, they are worthy of separate generic rank. The skull and the aperture of the ear of Syrnium, uralense and S. lapponicum are generally described as asymmetrical. Whatever may be the case in the latter sj)ecies, in the former this region is most certainly symmetrical, as is proved by a specimen in the flesh generously furnished me by my much-lamented friend Mr. Meinertzhagen. I have besides a skull of this species kindly lent to me by Prof. Newton, which is also quite symmetrical. That the nestling-down of ptilopsedic birds may consist of two distinct kinds of feathers is a point of some interest, and which may yield some useful facts if carefully MOEPHOLOGY OF THE OWLS. 265 studied. It is probable, as I liave pointed out in an eaiiier part of this paper, that " pre-pennae " only are present in the young Owl, and that the " pre-plumulge " are not represented in the very young nestling. When one reflects on the gi-eat range of grada- tion in complexity of structure that obtains in the pre-pennae of different groups, from the highly complex featliers of the Tinairiou, through the less complex feathers of the GalIiform.es and Anserufonnes to the exceedingly simple hair-like structures of tlie Columhce, it is pretty obvious that a great deal of work yet remains to be done before this study is exhausted. JS'itzsch was apjoarently the first to notice the twofold nature of the neossoptiles. He wrote concerning the " nest-plumage of Diurnal Raptorial Birds " that it " consists as usual of downy barbs which are attached to the first perfect barbs of all the contour • feathers, even the remigcs and rectrices, and disappear as the bird becomes fledged . Moreover, the whole body, with tlie exception of the axillary cavity and lateral neck- spaces, is covered with true permanent down-feathers, which bear no evanescent tips.' In this last point, that the absence of " evanescent tips " to the feathers in question proved them to be permanent down-feathers, he is evidently mistaken, inasmuch as they are quite distinguishable from the defioitiv(; down-feathers, as was pointed out in the earlier part of this paper. Dr. Gadow has stated that down-feathers occur on the nestling wherever they are found on the adult. This is certainly not the case in the Common Duck, and there are probably mauy other exceptions. The position of the femoro-crural band, which is peculiarly well developed in the Owls, if it is not confined to this group, was a point upon which I was for a time inclined to place considerable hope, and even yet it may prove of no little value to the often hard- pressed taxonomist. For the moment, however, with me its trustworthiness is under a cloud. The point, it must be remarked, is to note whether its outer end terminates above or below the knee. After examining a considerable number of spirit-specimens with regard to this point, an uncomfortable suspicion crossed me tliat this might dej)end upon the position of the legs when finally fixed by the liardening influence of the spirit : that is to say, it may depend upon the amount of rotation of the femur towards the spine, since the skin in this region seems to have a certain amount of " play," in Avhich case, of course, this band might appear to terminate a little above or a little below the knee, just as this was near to or far from the vertebral column. Certainly the rather large series of Asio accipltrinus all agree in having this band terminate at the knee- joint ; in Bubo it appears to terminate below the knee, and in Carine very much below this region. It was over this very species, however, that my faith received its first shock, inasmuch as in a second example of C. noctua, which I examined before placing it in spii'it, i was able to move the distal end of this tract slightly up or down at will. The amount was slight, but still appreciable ; of course my test may be at fault, as I pushed the leg upwards and the skin downwards at the same time. Now this particular specimen had been dead some days, so tliat the delicate attachments may a'l have broken away directly my somewhat rough experiment was begun; it is to be hoped that someone will try to settle this point soon. 36* 266 MR. W. p. PYCRAFT ON THE A comparison between the pterylosis of the adult and the embryo could not be expected to lead to any very startling contrasts ; nor does it. It is interesting to notice, however, that in none of the embryos of Jsio which I have examined is there any trace of the voluminous neck-folds of the adult, but tlie feather-rudiments are seen to closely invest the neck, as in Strix Jlannnea. Again, the external aperture of the ear in this species is marked by a shallow depression whose long axis is not more than twice that of the longitudinal axis of the eyelid ; the opercular and postaural folds are not even hinted at. In Strix flammea the operculum, as in Asio, does not appear till later in development ; its peculiar four-sided form seems not to be assumed until after the development of this fold. The branches of the interscapular fork in the embryo of Strix appear to be longer than in the adult ; but since I have only had the opportunity of examining a single adult specimen, little importance can be attached to this. Similarly the interscapular fork of Carine hrama differed, and this time markedly, from what one would have expected, inasmuch as its branches were of considerable length, and not, as in the adult of C. noctua, short and barely perceptible. This may mean one of three things : (1) the embryo is not that of C. brama ; (2) that C. hrama may not belong to the genus Carine, but some other genus of the Btihonidcc ; or (3) that this may represent an earlier stage in the history of the development of this region of the tract. In all the embryos there was a distinct claw both on the pollex and index digits. The form of the beak in all the embryos agreed with that of the embryo and adult of Strix in being relatively long and straight, and contrasted with the short curved beak of the embryo of Falco or of the adult Owls of the Asionid group. We come now to a question of wider and therefore of more general interest — the nature of the evidence that a study of the Pterylography of the Owls affords as to their systematic position. Nitzsch (13) years ago, approaching the subject from this point of view, came to the conclusion that they were most nearly related to the Accipitres. They formed his Aecipitrin(B nocturnce. He tells us that " there are two points which essentially distin- guish this group pterylogra])hically from the preceding (Accijritres dkirnce), namely, the constant absence of the aftershaft and the tcant of the circlet of feathers at the apex of the oil-gland.'" And, further, " In their form the tracts agree in general with those of the Ealcons, although they present noticeable differences. Among these the chief is that in the Owls the region of the throat betAveen the rami of the lower mandible is never continuously feathered, but the inferior tract issues from the angle of the gonys in the form of a narrow band, and divides on the middle of the neck or close to the front of the furcula " The ^;/? < UJ 1 < EC. t- T) Mfinertiiliageii del West Newma^i Pljotn iith MOB.PH0L0GY OF THE 0V7LS. LINNEAN SOCIETY OF LONDON. MEMORANDA CONCEKNINfi TRANSACTIONS. TlieErst Series of the Transactions, containing both liotanical and Zoological contributions, h;i.s been completed in 30 Vols., and a few entire sets are still for sale. Onl)- certain single volumes, or parts to complete sets, may be obtained at the original prices. The price of the Index to Vols. 1-2.5 is Ss. to the public, and (is. to Fellows; to Vols. 26-30, 4s. to the public, and 3s. to Fellows. The Second Series of the Transactions is divided into Zoological and Botanical sections. The prices of tlio Zoological parts of these which have been published are as undermentioned (for the Botanical parts see Botanical wrapper) : — ^'oliiine. Part Part Part Part Part Part Part Part \Vli,-ii Publislipd. Price ti) tlie Pul.lir. II. Part Part Part Part Part Part Part Part Part Part Part Part Part Part Part Part Part Part J.' .V. r/. SECoNn Skkiks. — Zcjolooy. I. 1S75 1 4 0 II. 1875 0 6 0 HI. 1876 1 8 0 IV. 1877 0 10 0 V. 1877 0 18 0 VI. 1877 1 2 0 VII. 1878 1 10 0 VIII. 187!) 1 0 0 0 I. II. III. IV. V. VI. VII. VIII. IX. X. XI. XII. XIII. XIV. XV. XVI. XVII. XVI IT. 1879. .. 1881. .. 1882. .. 1882. . . 1882. .. , 1 883. . . 1883. .. 1883. . . 1883. .. 1884. .. 1884. . . 1885. . . 1884. . . 1885. . . . 18S5. . . 1885. . . , 1886. . . , 1888. .. . 1 4 0 15 0 1 8 0 0 7 0 (f 3 (1 1 0 0 0 0 0 3 0 0 3 0 0 4 0 0 10 0 0 6 0 0 6 0 0 0 0 0 4 0 0 5 0 0 3 0 0 2 6 III. Part I. 1884 1 14 0 Part II. 1884 1 12 0 Part III. 1885 1 10 0 Part IV. 1885 0 8 (» Price tn Fellows. £ s. (I. 0 18 0 0 4 6 1 1 0 0 12 0 0 13 6 0 10 6 1 7 0 0 15 0 Volume. \Miei. Piililislied. Price to lh( Public. Price to Fellows. 0 0 0 0 0 0 0 0 0 0 0 0 0 18 0 oil 6 1 1 0 0 5 0 0 2 3 0 15 0 1 5 6 1 4 0 1 2 6 0 0 0 ,i .*. honis, unaccompanied bj' any figures. Alessandrini. — " Ceuni suU' Anatoniia del Dasipo minimo, Desm." Mem. Pi. Accad. delle Se. dell' 1st. di Bologna, t. vii. 18-36. Contains very brief and unimportant notes on Tatuna. W. H. Elower. — " On the Commissiu-es of the Cerebral Hemispheres of the Marsvipialia and Monotremata." Philosophical Transactions, 1865. In this classical memoir the mesial surface and a coronal section through the cerebrum of Choloqnis didaciylus have been represented in figures with an exactness whicli is lacking in the eai'lier memoirs. W. Turner. — " On the Anatomy of the Brain of Dasi/jnis sexcinctus.'' Jom-nai of Anatomy and Physiology, vol. i. 1867. The most complete account we possess of the Ijrain of Dasijpus scJ-'cinctus, illustrated by three figures. 284 DE. G. ELLIOT SMITH ON Georges Potjchet.— " Memoire sur FEncepliale des Edeutes." Journal de I'Anatoinie et de la Physiologie, tome v. 1868 aud tome vi. 1869. In this important work figures of tlie actual brain in the genera Tamandua, Ci/doturus, Chlamydopliorm, Dasypus, Manis, and foetal specimens of Bradypus, Cholcepus, Tatasia, and Orycterojnis, have been given in addition to cranial casts of Cholcepus, Orncteropus, Mylodon, and Glyptmhn. I have been unable to consult Pouchet's earlier " Memoires sur le grand Fourmilier." Patil Gervais. — "Memoire sur les Formes cer^brales propres aux Edentes vivants et fossiles." Nouvelles Archives du Museum d'Histoire naturelle de Paris, tome V. 1869. Contains an excellent series of figures of cranial casts of all the representative existing genera of Edentates as well as most of the extinct forms. In addition there are figures of the actual brain of Mynnccophaya, Manis, Dasypus se.vcinctus, and a young Orycteropus. A. H. Gabeod. — " Notes upon tlie Anatomy of Tolypeutes tricinctus, with remarks upon other Armadillos." Proceedings of the Zoological Society of London, 1878, pp. 222-230. The earliest description of the brain of Tolypeutes : in addition the author describes, with an excellent illustration, tlie brain of Xenurtis. W. A. EoBBES. — " On some Points in the Anatomy of the Great Ant-eater {Myrmecophaga jubata)." Proceedings of the Zoological Society of London, 1882. The best account of the brain of Myrmecopliaya yet published. W. TuRNEB. — "The Convolutions of the Brain." Journal of Anatomy and Physiology, vol. A-xv. 1890. Professor Turner adds brief notes and two figures of C'holoepus Iloffmanni to his previous contribution dealing with Basypns. H. Eabl-Puckhabd. — " Einiges iiher das Gehirn der Edentata." Archiv fiir mikro- skopische Anatomic, Bd. xxxv. 1890. This brief memoir represents all that has hitherto been accomplished towards the elucidation of the histology of the brain in Edentata, and concerns itself merely with the anterior commissure and the presence of a longitudinal ependymal fold in the roof of the aqumductus Sylvii of a foetal Xenurus. The most valuable feature of this memoir is the excellent series of figures of sections stained by the method of Weigert. Max Webeb. — " Beitrage zur Anatomic und Entwickclung des Genus Manis." Zoolog'ische Ergebnisse einer Kcise in Niedcrlandisch Ost-Indien II., Leyden, 1892. Contains the best description of the brain of Manis hitherto published, with illustrations. At the time when this investigation was undertaken the literature included in the above list represented all that had hitherto been published (so far as the writer is aware) concerning the anatomy of the brain in the Edentata. But since the present memoir was announcedj Professor Howes has kindly called my attention to a brief note concerning a monograph upon this subject, which unfortunately has not been publislied *. Quite recently Ziehen has published some comparative notes upon the brain of Manis in the first part of a large monograph upon the brain in Monotremata and Marsupialia t . * George Hu.xtington. — " Contributions to the Visceral Anatomy of Myrmecophaga jubata, Tamandua hivittata, Arctopitheciis didactylns, Dasypus sexcincins, Tatusia novemcincta, and Manis lotiyicaudata." [Abstract.] Transactions of the New York Academy of Sciences, vol. xv., January 1896, p. 98. t TuKODOK Ziehen. — " Das Centralntrvensystem der Monotremen und Marsupialier." 1 Theil. Semen's 'Zoologischc Forscl.ungsreiscn in Australien und dem Malayischen Archipel," 1897. THE BRAIN IN THE EDENTATA. 285 There is much in this important memoir which calls for criticism, but at present I will merely take cognizance of the references to Jlau/'s. In a future memoir I sluili discuss some of the statements in reference to the Monotremata and Marsupialia, as well as some of the more general questions, the consideration of which is precluded here by limitations of space. Several investigators have recorded the weight of the brain in various Edentates, the most noteworthy contribution to this suliject being Max Weber's monograpii (Jam. cit. p. 282). If we sum up this literature we find tliat, apart from observations upon the general form and size of the brain, it contains little information of sufficient exactness to be of value to the comparative anatomist. Almost the whole of the data collected in the memoirs concern the cerelnul cortex, and even with regard to this important region of the brain we are provided with only the most meagre information, often very vaguely expressed. Almost all writers are silent concerning the basal regions of the brain, which are of especial interest in the lowlier mammals. Heviewing the literature of the different families, we find that nothing is known of the adult brain of Orycteropus, except such information as the examination of cranial casts has provided. The brief notes of Pouchet and Gervais upon the brain of a foetal Orycteropus add little, if anything, to our knowledge. Concerning the Myrmecopliayidce, we have, in the recent memoir of Forbes and the earlier contribution of Gervais, a very good account of the cerebral hemisphere of MyrmecopJiaga, although the base of the hemisphere and the rest of the brain receive very scanty treatment. Practically nothing beyond the brief and imperfect notes and unsatisfactory figures of Pouchet has hitherto been made known of the brain of Tamandua. Por our knoAvledge of Ci/cloturm \yq are indebted mainly to Pouchet and the early memoir of Tiedemann. The knowledge of the brain in the Sloths is even more unsatisfactory, as no minute description has yet been given of the brain in either Brudi/pus or CJioloepus. Gervais' beautiful figures of cranial casts in these two orders have made us familiar with the exact sliape and size of the brain, and the series of figures of Pouchet and the more recent contriljution of Turner have given us a much fuller view of the brain in this order than the earlier works of Tiedemann and Rapp conveyed. But even now our knowledge of the brain in this family is very deficient, and I am able to supply the required information only very imperfectly. I have endeavoured to supplement and discuss as a whole the data concerning the Armadillos, for which we are mainly indebted to the memoirs of Turner, Pouchet, and Garrod, and which relate to all the genera except Priodon, our only knowledge of which is derived from the figure of a cranial cast, which Gervais has contributed. The interesting brain of ChlamdyopJioriis, concerning which Hyrtl published a few notes, which were supplemented by a fuller descri^jtion by Pouchet, has been carefully studied in a series of sections stained with carmine. This has enabled me to srive SECOND SERIES. — ZOOLOGY, VOL. VII. 4-0 286 DE. G. ELLIOT SMITH ON a mncli fuller and more accurate account of this peculiar brain than Pouchet has provided. Max Weber has recently contributed an admirably illustrated account of the brain in Manis, which is a most valuable supplement to Pouchet's imperfect work I have been able to add slightly to Weber's account, and upon comparative grounds have interpreted certain features in a somewhat different manner. The Ventral Surface of the Cerebrum. The most convenient landmark with which to begin the description of a brain for the purposes of accurate comparison is probably the optic chiasma. The optic nerve, taking its origin from the eye lying in the orbit, enters the cranium and extends toward the base of the brain at the situation wliere the brain-stem joins the cerebral hemispheres. In Orycferopiis * this point is situated about midway between the cephalic and caudal extremities of the brain, or, to be exact, 52 mm. behind the apex of the hemisphere (which is formed by the olfactory bulb), and 56 mm. in front of the place where the medulla oblongata merges into the spinal cord. (These measurements, as all the figures given in this memoir, refer to material which had been kept in alcohol for jirolonged periods.) The optic nerves enter into relationship with the base of the brain at this mid-point, and effect an intercrossing of fibres, the oj^dc chiasma, from each side of which a rounded bundle of fibres, the oj^fic trad, arches laterally, and with a slight inclination backward, to disappear (at a distance of 6-5 mm. from the mesial plane) under an overlapping fold of cerebral cortex, which we shall subsequently recognize as part of the j^yi'/form lobe. Por convenience of description we may regard the areas in front of the optic chiasma as part of the cerebral hemisphere, and the parts behind it as the brain-stem. Immediately in front of the optic chiasma, on each side of the mesial plane, we find a flat depressed quadrilateral area of about 8 mm. diameter. This is called by different writers a variety of names, of which loctts 2ic>'foratus {ant lens) is as convenient as any. As I shall not refer to any other ferforated space, I may omit the qualifying adjective. In front of the perforated space we find a large oval area of grey substance, which is slightly raised above the surrounding regions. This is the tnherculum olfuc- toriuiii. Its major axis, which is sagittal, measures 18 mm., and its maximum breadth 13 mm. It is not confined to the ventral surface, but also extends on to the mesial surface of the hemisphere, and there presents a fusiform outline below the precom- missural area. The lateral border of the iuherculum olfactorium is clearly defined by a furrow in which a A-ery compact strand of nerve-fibres, which constitutes the olfuctonj trad, is found. In a fresh bi^ain the pure white colour of the latter would present a marked contrast to the brownish-grey colour of the olfactory tubercle. TJie tradus olj^actoriiis is situated upon the surface of a cortical area, which we shall call the anterior l}art of ihe, jjyriform lobe. This jiart of the pyriform lobe is visible along the lateral * As practically nothing is known of the brain of Oryderopus, I describe it in some detail and point out the features in other forms b}- contrast. THE BRAIN Ii\ THE EDENTATA. 287 margin of the olfactory tract as a narrow band of grey substance with an irregular tuberculated margin. The pyriforni lobe, the anterior part of which is formed by this tuberculated worm-like process, will be found to expand posteriorly and form a large rounded prominence, which is often known as the natiform eminence. This expanded portion of the pyriform. lobe we may distinguish as the lobus injriformis posticus, to distinguish it from the lohus pt/riformls aniiciis, which is merely the tapering cephalic extremity of the same histological formation. These two parts of tlie pyriform lobe are separated the one from the other by a depression, the vallecula S/jlvii, which extends obliquely in a mesial direction toward the perforated space. Fig. 1. bulb, olfact. lob. pyriform. ant loc. perforat lob. pyriform. post. nerv. III. . pons Varol.. pyramid. - ped. olfact. ■ tubercul. olfact. • tract, olfact. vail. Syl. tract, opt. crus cerebri Corp. geaiculat. po6t^ ■gangl. interped. nerv. V. •fiss. flocculi ■paraflocc. dors, parafioco. veatr. trapezium medulla oblong. Ventral surface of brain of Ofijcicropus. Xat. size. The terms " lohus pynformis " and " tractm olfactorius " have been applied in a sense slightly different from that which the majority of writers attach to them. This attempt at a greater precision in description needs some sliglit explanation. It is customary to restrict the term " lohus pijriformis" to that expanded posterior area of the lobe (as we understand it) which is situated behind the vallecula S//lcii, and for which the terms " natiform eminence," " hippocampal lobule," and many other variants of these terms are regarded as alternative names. The anterior tapering part of the same histological formation, which we haA^e already 40* 288 DR. G. ELLIOT SMITH ON iucliuled in tlie pyi-iform lobe {lobus pyriformis anticns), is generally referred to as the " tracfns olfactorlus." This confusion may be avoided if we restrict the term olfactory tract to its literal meaning as the group of nerve-fibres lying on the surface of a cortical area, Avhich. is the cephalic extension of the pyriform lobe. [To be strictly accurate we should say that the lobiis pijriformis postictts is the greatly expanded caudal extension of the lobtis 2^!/f'ifor7nis anticns, but as the name " pyriformis" is usually associated exclusively with the former, I have spoken of extending it (the name) to the latter.] I have deliberately avoided the use of the term " hippocanipal lobule," or any of the many vai'iants of this term, because such expressions introduce a most undesirable confusion with the true "hippocampus," a structure quite distinct from the pyriform lobe, but one which in literature is frequently mistaken for it as a result of the confusing nomenclature in common use. This is an interesting example of the widespread misconception to which such loose nomenclature can lead in comparative anatomy : for the student of the literatvxre x'elating to the fore-brain in Reptilia and Dipnoi (more especially Burckhardt on Protopderiis) has had but too clear a demonstra- tion of the utterly chaotic state of the morphology of the fore-brain which the confusion, of pyriform lobe with hippocampus can produce. If we consider tlie base of the hemisphere as a whole, it will be found to present a pyriform shape. It is broadest at its posterior extremity, and contracts gradually as it extends forward, so that it passes insensibly into the large rounded stalk of the huge olfactory bulb which caps the anterior extremity of the heniispliere. The hidhns olfactorius is a large conical cap of grey substance of about the size and shape of a Albert nut. The contour of the ventral surface is not unlike the conventional heart-shape, with the apex pointing forward. This surface is indented by a deep Fis. 2. cerebellum Bss. prima ped. olfact. bulb. 9lfact.1 i fiss. flocculi parafloco. dors. medulla oblong. lob. pyriform. ant. ' tract, olfact. tnbercul. olfact. Corp. geniculat. ant.;crus. : ,?°'^^, ■ parafiocc. veutr. I cerebri area "arol. , *^ Corp. tegmenti flocculus geniculat. post. Left lutenil hurt'ace of biaiu of Uri/ctcio^m.s. .Xat. size. The cerebral hemisphere has been separated from the rest, by cutting the jmictiou of the optic tbahimus and corpus striatum so as to expose the Literal aspect of the thalamic aud mid-brain. sagittal fissure which is placed nearer the lateral than the mesial border. This fissure indents the whole thickness of the wall of the olfactory bulb, so that when we examine THE BRAIN IN THE EDENTATA. 289 the large cavity or ventricle of the bulb we find a very prominent longitudinal ridge in the floor, Avhich is tlie result of this indentation (fig. 19), The olfactory bulb is connected with the rest of the licmisphere by a large rounded peduncle, which is visible upon the base of the brain in the interval between the tuberculum olfactorium and the buJbus olfactorms. Even in a view of the dorsal aspect of the brain the peduncle is visible, because the apes of the cerebral cortex does not overhxp it to so great an extent as in most mammals. On the lateral aspect of the hemisphere the massive olfactory peduncle extends backward and becomes directly continuous with the cephaUc extremity of the j^yriform lobe. The medullary covering which whitens the surface of the peduncle converges to form the compact olfactory tract, and immediately above this we find- the vermiform anterior pyriform lobe. The narrow Fi-. 3. ■ - bulb, olfact. —ped. olfact. -lob. pyriform. ant. tisa. rhinal. . fiss. prima — area A '-^ area B ■■ area C fis3. Beeunda medulla oblong. Doreiil surfuci-^ oi' ))raiii of Orjjcteropus. Nat. size. Based on a pencil sketch by tlie late Rt. Hon. T. H. Huxley. band of grey substance which constitutes this part of the pyriform lobe is deeply constricted in many places, so that it appears to consist of a series of round knolis. These knobs are so prominent that they produce correspondingly deep depressions on the cranial wall, and hence we can readily recognize them in cranial casts. A beautiful 290 DE. G. ELLIOT SMITH ON demonstration of this is givea Ijy Poucliet's figure of a cranial cast of Ori/cteropus*. In a dorsal view of the actual brain the tuberculated lobtts 2}yi"iformis anticus is very distinctly seen at the lateral border of the anterior part of the cerebral hemisphex*e (fig. 3). In the specimen of OrycteTopus which we have so far been considering, the upper boundary of the pyriform lobe is very definitely indicated in its whole extent by a deep and clearly-cut rhinal fissure, Avhich begins anterioidy in the cleft between the apex of the hemisphere and the olfactory peduncle, and proceeds backward in a direction Avhich is practically horizontal (fig. 2). The posterior half of the fissure descrilies a very slight arc whose convexity is directed ventrally. Upon being traced backward the rhinal fissure extends on to the caudal surface of the hemisphere and proceeds transversely inward, as vre may observe in a view of the ventral surface of the brain (fig. 1). The fissure ultimately reaches tlie mesial surface of the hemisphere and then ceases abruptly (fig. 4). In the -whole of its course the fissure lies about midway between the upper and lower surfaces of the hemisjihere. The term ^ssitra rhhudis may be most conveniently employed to designate the whole of this extensive fissure, as I have applied it in the above description, and not only to the anterior part as Krueg uses it f . The anterior jmrt of the fissure, wdiich forms the upper boundary of the lohus 'pyrxjormis anticus, may be distinguished as the fissura rhinalis anterior, and for the posterior part of the fissure we may adopt Krueg's name "Jissura rhinalis jiosterior." The term "rhinal fissure" may be used to apply to the combination of these two fissures. The dee}) cleft between the olfactory peduncle and the apex of the hemisphere is sometimes regarded as part of the rhinal fissure. If we admit this, the rhinal fissure begins and ends on the mesial surface of the hemisphere in Oryctero2n(>s (fig. 4). In the brain Avhich I have hitherto been describing, the features of the extensive undivided rhinal fissure were clearly defined in its whole extent, and presented an exactly similar appearance in both hemispheres. But both of the brains in the Museum of the lloyal College of Surgeons present a peculiarity in the disposition of the rhinal fissure, in which, while they agree the one with the other in both hemispheres, both diff"er from the type specimen Avhich we have been so far considering. This peculiarity consists of a separation of the anterior from the posterior rhinal fissure. The anterior rhinal fissure begins anteriorly in the ordinary manner, and forms the upper boundary of the anterior jiyriform lobe in the whole of its extent, but then ceases without joining the l^osterior rhinal fissure. The posterior rhinal fissure extends forward above the anterior rhinal and pursues a course for some distance parallel to the anterior fissu.re. Thus the cortex of the pyriform lobe is connected to the cortical area lying above the rhinal fissure, now generally known as imllium, by a narrow process of cortex lying between the overlapping extremities of the two parts of the rhinal fissure. This condition has been somewhat schematically represented in fig. 5. * G. Poucbet, op. cii., ' Journal de TAnatomie,' &c. torn. vi. pt. t. fig. 3. t Julius Krueg, " L'eber die Furchen auf der Grosshirnrinde dor zonojjlaccntak'ii SiiugetLiere," Zeitschr. f. wissensch. Zoologie, Bd. xxxiii. ISSU, p. 610. THE BKAIN IN THE EDENTATA. 291 hippocamp. vestigia ex. Fig. 4. psalterium dors. I Corp. / callos. bulb, olfact I fiss. hippocampi f^^ fascia dentata -fiss. rhinal • . , , ,, . , I '^ \ N fimbria / tubercul. olfact. | \ \_ \ \ Ilea, oiiact. commiss. ant. \ ,„. » ., , '. ^ \lob. pynform. psalterium ventr. \ tubercul. hippocampi comm habenuL Corp. pineale a fiss. prima i =°!'P- ,' cerebellum I quadrigem. / fisa. secunda foramen Monr. "ventr. IV. vel. medullar© pons Varol. commiss. moU. ; ( I Corp. mammill. | gangl. interped. aquaeduct. Syl. Mesial surface of the brain of Orycteroims, exposed by mesial sagittal section. Xat. size. The hemisphere above separated from the rest of the brain, and bdo^r in part attached to the brain-stem. Fig. .5. fiss. rhinal. ant. y ^ bulb, olfact. : ^ ; : ■ lob. pyriform. i fiss. rhinal. post. tract, olfact. tubercul. olfact. lob. pyriform. post. Scheme of the lateral aspect of the cerebral hemisphere of Oryctcroinis, to show a variation in tlie arrangement of its fissures. 292 DK. G. ELLIOT SMITH ON The I'liinal fissure thus forms a very clear boundary in the whole extent of the lateral and caudal extremities of the hemisiohere between the pyriform lobe and the histo- logically and morphologically distinct area of pallium. But ujjon the mesial surface of the hemisphere (fig. 4) the upper margin of the pyriform lobe blends with the pallium without any clear external indication of the line of junction. In this part of the liemisphere the jiyriforra lobe extends around the postero-inferior angle, and is partly limited in front by the loAver joart of an arcuate cleft — the ^/issar a hippocampi. But below the hippocampal fissure the pyriform lobe seems to be continuous with a peculiar oval body, which, for reasons to be subsequently explained, will be distinguished as the " tuherculuni hippocampi " or " hippocampal tubercle." The height to which the lobiis pyriformis extends upon the mesial surface is indicated by the indentatiou \\\)o\\. the posterior margin of the hemisphere, which is the termination of the rhinal fissure (fig. 4). If Ave compare these basal regions of the cerebral hemisphei"e of Orycteropiis with those of the Great Ant-eater, Ilyrmecophaga juhala, which in size and general mode of life presents some resemblance to the African representative {Orycteropiis), we find a general agreement bstween the two forms, but at the same time a number of interesting points of diff"erencc. Although ihe figures and descriptions of Forbes * have added much in clearness and exactness to the earlier observations of Gervais t, we still lack a faithfid representation of these basal regions of the hemisphere of Myrmecophaga. The bulhus olfactorius is relatively almost as large as that of Orycteropus, but it is not so pointed, nor is it indented by any fissure such as we have seen in the latter. It is also attached to the hemisphere by a large rounded peduncle, whose relations are analogous in the two forms. The luherculum oJfactoriiim is not so elongated as it is in the Aard-vark {Orycteropus), and presents a somewhat quadrilateral form. The tractus olfactorius presents the same general features, but in Myrmecovhaga ends in a prominent elliptical nodule, which is situated in the vallecula Sylvii immediately in front of the emineutia natlformis and at the lateral margin of the locus perforatns (fig. 6). This Fig. 6. loc. perforst. nerv. II tnbercal.' ol/act, tract, opt. ,■ ^ ,, . \ : . . tract, olfact. VJ, • y » -- Alob. prrifonn, rXmf^imlfmK /fsnoia dcntata f^^'^^ll/ Irl /""f- "'■Mill. m\Wk//////if """ °°"''" %vyJ^U-tpfJ gaogl. interped. Parh of tlie ventral surface of the brain of Myrmecojihaga juhata. Sligbtlj' reduced. little nodule, which has been indicated in the figures by the name tuherculum tractus olfactorii, has been described in the Hedgehog {Erinaceus) by Ganser $, and is found in * Op. cii. Proc. Zool. Soc. London, 1882. t Op. cit., Nouvelles Archives, t. v. t S. Ganser, " Uebcr das Gcbiru des Mauhvurfs," ilorphologiscbes Jahrbuch, Ed. vii. 1882. THE BKAIN IX THE EDENTATA. 293 a well-developed form in Tamandua (fig. 9), the Dasypodidte and Bradypodidce, and possibly in other representatis^es of the Edentata. A very well-defined fissura rJmiaUs is present to indicate the upj)er bouudary of the pyrit'orm lobe, but it differs very distinctly from that of Orycteropus in that the posterior rhinal fissure joins the anterior rhinal, with which it is continuous, at an angle (fig. 7), wliile the anterior rhinal fissure Fia bulb, olfact. ped. olfact. fiss. rhinal. ant tnbercul. olfact. paraflocc. pons Varol. lob. pyrifonn. S fiss. rhinal. post Lateral aspect of a braiu oi 21yrmecophaya julattt. S, Fossa Sylvii. Very slightly reduced. is still horizontal ; the posterior rhinal inclines obliquely downward and backward, so that it reaches the ventral margin at a point about midway between the vallecula Sylvii and the posterior extremity of the hemisphere ; it then curves inward upon the inferior surface, and ultimately ends upon the jDostero-inferior part of the mesial surface at a point just behind the lower part of the hippocampal fissure. This mode of termination of the rhinal fissiu^e in Myrmecophaga has l)een well figured by Forbes *, and presents a close analogy to the disj)Osition which is rejiresented in Tamandua in this memoir (fig. 10). The rhinal fissure in Tamandua (fig. 8) presents a less acute bend than is the Fip ^lob. ant. y ! area A area 6 ^ cerebellum ; paraflocc. bulb ■ olfftpt ■ ' '""> I flocculus pulD. oaaot.jj.^gj olfact.; Pjnform.; fiss. rhinal. pons Varol. Lateral surface of brain of Tanniitdua titradacff/h(. Nat. size. case in the larger Ant-eater, and as a result it approaches much nearer the posterior margin of the hemisphere, but otherwise its disposition closely agrees with that of * Op. dt. Proc. Zool t^oc. London, 1882, fij;. 4. SECOND SERIES. — ZOOLOGY, VOL. VII. 41 294 DE. G. ELLIOT SMITH ON Myrmecoi^haga. In both representatives of the Myrmecopliacjidce the rhiual fissure is nearer the base_.of the In-ain than it is in Orijcteropus, so that in a view of the ventral surface practically the whole of the fissure is visible (tig. 9). Fis. 9. fiss. rhinal. ' loc. perforat tract, opt. Corp. mammill.- nerv.V. trapezium bulb, olfact. ..ped. olfact. -lob. pyrifonn. tuber, olfact. ■ nerv. II. tuber, tract, olfact. infundib. crus cerebri pons Varoi. .|;>-~ parafloec. .. ^flocculus nerv. VIII. pyramid. Ventral surface of brain of Tamandua. Nat. size. It follows from the above remarks that the shape of the pyriform lobe is very different in the American Ant-eaters, and esjiecially the Great Ant-eater, from that which we have found in the African Aard-vark. Thus in 3Iyrmecoplictga, and to a less degree in Tamandua, there is a very distinct flexure in the pyriform lobe which is wanting in Orycferopus. The pyriform lobe is besides relatively smaller in the American forms. In these brains also the anterior limit of the pyriform lobe upon the mesial surface of the brain is more distinct, because the peculiar hijapocampal tubercle which we found in Orycteroptts is lacking in the Ilyrmecophagidce, and the hippocampal fissure, as a consequence, extends much lower (fig. 10). As a result of the flexure of the pyriform lobe in Ilyrmecopliaga, the vallecula Sylvii is deepened and a number of deep clefts or puckerings indent the surface of the pyriform lobe in this region. They are probably mere mechanical results of the bending (fig. 7). In Tamandua the tuhercuhmi olfactorium presents the same quadrilateral shape which we noted in Ilyrmecophaga. The olfactory bulb is relatively smaller than it is in the larger brains, and its peduncle is not so prominent because its dorsal surface is completely hidden by the anterior pole of the hemisphere. My knowledge of tlie conformation of the brain in the small arboreal Ant-eater Cycloturus has, in the al)sence of any material, been derived mainly from the imperfect notes and illustrations which Pouchet (op. cit.) has contributed, and the even less complete observations of Tiedemann {op. cit.) and Gervais {op. cit.). It is not clear from Pouchet's work whether any rhinal fisstire is present in the small arboreal Ant-eater. He says in the text that the brain of Cycloturus (which he calls Dionyx) is smooth, and he does not represent a rhinal fissure in his figure. But since he does not represent THE BRAIN IN THE EDENl'ATA. 295 the clearly defined rliiual fissure in Tamandua, tlie non-delineation of a corresponding fissvire in his figure of Cycloturus cannot he regarded as any indication of its absence. The illustrations which the earlier observers (Tiederaann, Leuret, and Rapp) have given of the brain in the Sloths convey practically no information concerning those regions of the brain which we have just been considering in the Aard-vark and Ant-eaters, and the figm-es which Pouchet [op. cit.) has given of the Brady podiclce do not represent even a feature so fundamental as the rhinal fissure. Professor Turner's representation * of the brain of Choloepns Hoffmanni, so far as I am aware, is the only figure wiiich represents this feature Fig. 10. psalterium dors. psalterium ventr. / '^corp. callos. / euJc. Umitans pallii bulb, olfact. fimbria hjppocamp. inversus fascia dentata -fiss. hippocampi fisB. rhinal. ped. olfact. -' ; taberoul. olfact. i area praeoommiss. > \ tract, opt. lob. pyriform. i_ tuber, tr^ot. olfact. .commiss. ant. lamiua terminalis Mesial surface of right central licmisphere of Tamancliut. Enlarged 2. diam. in the brain of the Sloths. Of the two brains of Bradypus tridactylus at my disposal, one had almost reached the adult state, whereas the other was much younger, since it was taken from an animal measuring only 34 cm. from the nose to the anus. In the former the anterior rhinal fissure begins in the ordinary manner and extends obliquely upward and backward, and when it reaches a point midway between the anterior and posterior poles of the hemisphere it joins the posterior rhinal fissure alino.st at a right angle (fig. 11). The posterior rhinal fissure forms an arc, the postei"ior extremity of which crosses the posterior border of the hemisphere at about the same horizontal level as the cephalic extremity of the anterior rhinal fissure. In the younger brain the two parts of the rhinal fissure approximate much more closely to the horizontal, so that their angle of junction is very obtuse. The posterior rhinal fissure is represented merely by a very shallow furrow. In Cholcepus didactylus and, jvidging from Turner's figure [op. cit.), also Choloepiis Hoffmanni, the angle of meeting of the two parts of the rhinal fissure is quite as obtuse * AV. Turner, op. cit., Journal of Anatomy and Physiologj-, vol, xxv. 41* 296 DR. G. ELLIOT SMITH ON as it is in the younger Bradjipiis. The anterior rhinal fissure is very ohlique, hut the posterior rhinal fissure is almost horizontal (tig. 12). The conformation of the base of the brain in the Bradypodidce is so like that of Tamandua that it is unnecessary to give special illustrations, nor is it requisite to describe these features in detail. The general shape of the base of the brain in Brady pus is very bulb, olfaot. tnbercnl. olfact. fiss. rhinal, paraflooc. lob. pyriform. Lateral surface of brain of lii-ddiiiinn ilnlitclijlus. Nat. size. bulb, olfact... fisa. rhinal. paraBocc. dors. . paraflooc. ventr. tract, olfact • nerv.V. ■ lob. pyriform. tnbercoL olfact. : flocculns pons Varol. trapeziom . pyramid. Lateral surface of Imiin of CholceiiHs didactj/lvs-. Slif;htly enlarged. much like that of Tamcmdua (fig. 9), but in the former more of the pallium is visible along the lateral margin of the anterior part of the pyriform than is the case in the Ant-eater. This means that the proportion of pyriform lobe to iiallium is less in the Sloth than in the partially arboreal Ant-eater. In Cholcepns the base of the brain appears to be relatively broader than it is in B)-adijpns, but otherwise it presents a close resemblance. In none of the Armadillos (with the possilile exception of Priodon) do we find a complete rhinal fissure. Nothing is known of the brain of Priodon, except such information as we can gather from the examination of cranial casts, such as Gervais has represented {op. cit. pi. ii. fig. 12). But it is quite possible and even probable that in such a large brain a complete rhinal fissure will be found, because we find among mammals that an increase in bodily dimensions is accompanied by an increased extent THE BRAIN IN THE EDENTATA. 297 of pallium : in other words, an increasing disproportion between the sizes of the pallium and the pyriform lobe. It is such inequalities in the rate of growth of these funda- mentally distinct cortical areas that give rise in all probability to the rhinal fissure. In Chlami/dophonis there is uo rhinal fissure whatever, but in all the other Dasypodidae there are small anterior and generally also posterior rhinal fissures, w^hich are separated by a varying interval in different genera. In Tatusia noremcincta, two specimens of which I have had the opportunity of ex- amining, we see the arrangement of these fissures in a typical manner. A short anterior rhinal fissure begins in the cleft between the olfactory bulb and the apex of the pallium, and extends horizontally backward for less than a third of the length of the hemisphere. A posterior rhinal fissure, which is very shallow, extends from just below the middle of the posterior border of the hemisphere and passes forward for less than a third of the length of the hemisphere. In the figure which Pouchet gives of a young Tatusia {op. cit. pi. vi. fig. 1), these fissures are only indicated very faintly. As both of the specimens which I examined were distorted, I have modified this figure of Pouchet so as to indicate the features present in my specimens (fig. 13). Fig. 13. fiss. rhinal. post. fiss. rhinal. ant. o ...paraflooc. bulb, olfact. ! tubercul. olfact. \ lob. pyriform. Scheme of lateral surface of brain of Tatuf'ui [Hha. In the pallium, just above the wide gap between the anterior and the posterior rhinal fissures, we find a short oblique fissure which may be distinguished as 0. Fig. U. 0 S ■ \ fiss. rhinal. post. fiss. rhinal. ant. ,j)arafloco. bulb, olfact. ; ; '"''• pyriform. flocculus : : tuBer. tract, olfact. . ,. ,'f . tubercul. olfact. tract, olfact. Lateral surface of brain of U'lsiqu's si.rcinrttia. Nat. size. In Dasypus sea'cinctus (fig. 14) and Basi/pm vUIosks we find practically the same arrangement of the rhinal fissures, with the difi'ereace that the more extensive posterior rhinal fissure enters into uninterrupted continuity with the pallial fissure /3. 298 DK- Gr. ELLIOT SMITH ON In the figures which Turner (oj). (?//•)' Gervais (ojj. cit.), and Pouchet {op. cit.) give of the hrain of Dasypus we find the same ai-rangement as I have just described in my specimens. In Tolijpentes and Xemirm we find essentially the same arrangement, hut the caudal extremity of tlie posterior rhinal fi.ssure may fail in both these forms. Garrod {op. cit. Proc. Zool. Soc. London, 1878) gives figures of the brain in both of these Armadillos. In all of the Armadillos the olfactory bulb and all the parts of the brain which are intimately associated with it reach relatively enormous dimensions. Visual acuteness is a matter of subsidiary importance to animals which burrow in the ground for their means of subsistence, and hence the importance of a sense like that of smell, which enables these animals to detect objects which they cannot see, becomes considerably enhanced. Thus the macrosmatism which characterizes all mammals, except those whose pallium has become sufficiently specialized to be something more than a mere sensorium, becomes very much exaggerated in these burrowers, who are guided in most of their pursuits mainly by olfactory impressions. The large olfactory bulb varies considerably in shape in the different genera of this family. As the soft bulb is more liable to distortion than any other part of the brain, it is safer to rely upon the evidence of cranial casts than upon the examination of actual specimens when these have not been seen in the fresh state. Gervais has fortunately provided us with a complete series of cranial casts of the Dasypodidae {op. cit. pi. ii.). From the shape of the cranial casts, as well as from all the figures of Gervais, Pouchet, Turner, and Garrod, and from my own observations, it is evident that the large olfactory bulbs of both species of Dasypus and also of Tohjpeutes assume a conical shape such as I have already had occasion to note in the Ant-eaters, Slotlis, and Aard-varks. This form of bulb is also found in the extinct Ground-Sloths Mi/lodon, Scelidotherium, and probably also in Ilegatherium. It is equally found in the great extinct ArmadUlos Glyptodon (in which the olfactory bulb assumes enormous proportions), and probably Eutatus. In Xenurus and Tutusia the projection of the olfactory bull)s is still a feature of the brain, hut it is not so developed as it is in Dasypus. There is a certain amount of antero-posterior flattening of the lower part of the bulb in these forms. In Chlamijdoplwrus there is a very distinct antero-posterior flattening of the bulb, just as there is in the larger brain oi' tlie Pangolin {Manis). But the shape of the olfactory bulb is a matter of slight importance. We can better appreciate the high degree of macrosmatism of the Armadillos if we compare their brains with that of the little arboi-eal Ant-eater Ci/oloturus, the shape and size of which is admirably shown by the figures of Tiedemann, Pouchet, and Gervais. "When Ci/cloturus relinquishes a terrestrial life, such as its near relative Mi/nnecophaya leads, the olfactory sense diminishes in importance, and hence we have a reduction in the size of the olfactory bulbs and the rest of the apparatus pertaining to the function of smell. In the brain of tlie Armadillos the predomijiance of the sense of smell manifests itself in other ways in addition to its direct effect upon the olfactory bulb. The tnberculwni olfuctoriimi is uniformly very large and of relatively greater prominence than it is in the THE BHALX JX THE EDEXTATA. 299 Sloths and smaller Aut-eaters. The i)yriform lohe also becomes relatively much laro-er, and frequently constitutes as much as half the lateral wall of the hemisphere. In spite of this increase in size, the posterior part of the pyriform lobe does not form such a prominent projection (" natiform eminence ") upon the base of the brain as it does in the Ant-eaters. In the latter the growth of the pallium contributes toward the thrusting down, as it were, of the posterior part of the pyriform lobe. In most of the Armadillos the pyriform lobe does not project downward to a much lower level than the tuberciUnm olfactonum, and in the peculiar little distorted brain of Chalmycloj)lion(s the pyriform lobe does not descend nearly as low as the ftibercultmi olfacforium, which is much the most dependeat part of the hemisphere (figure 15). bulb. oltMt. lob, pyritorm. tabor:iul. olfact. Lateral surface of brain of Chlamtjihphorus truncaius. Enlarged 2 diam. ■v. Deep groove formed b}- semicircular canals upon base of hemisphere. In Chlamijdophorns there is no rhinal fissure, but histological examination demonstrates that the junction of pyriform lobe and pallium is placed very high upon the lateral wall, so that the pyriform lobe forms more than half of this aspect of the hemisphere. It is unnecessary to do more than refer to the peculiar, flattened, cake-like, sessile olfactory bulb and the elliptical tuherculum olfactorinm wdiose major axis is transverse. The receding surface of the pyriform lol^e in Chhimndophoms is seen to even better advantage in a view of the mesial surface of the hemisphere (figure 16). This, as well as Fig. 10. hippocamp. vestigia Corp. oailoa. flex, liippocampi bulb, olfact psalterium ventr. fiss. hippocampi fascia dentata oommiss ant tubercul. olfact. area praecommisB. fimbria Mesial surface of right cerebral hemisphere of Chhiiiii/diqiJ'oriis. Enlarged 3 diam. .r. As in fig. 15. the preceding figure, shows part of the peculiar depression (a;) which we find upon the ventral receding surface of the pyriform lobe, and to which Hyrtl {oj). cit.) and Pouchet 300 DR. G. ELLIOT SMITH ON {op. cit.) have alroady called attention as resulting from the upAvard projection of the bony semicircular canals in the periotic hone. Pouchet has already given a ligure of tlie ventral surface of the brain, which indicates tlie situation of this depi'ession. The imperfect descriptions of the brain of Manis which Pouchet {op. cit.) and Gervais {op. cit.) had given have recently been supplemented by the concise and well-illustrated description by Mas Weber {op. cit.). I have been able to confirm the accuracy of his brief description by the examination of two specimens of this genu.s. The basal regions of the hemisphere present in a typical manner the characteristic features which we have met in all the other forms. The base of the brain is very broad. The large olfactory bulb forms a thick buffer-like pad upon the flattened ceplialic extremity of the hemisj^here. The large tuberculum olfactorium presents an almost circular outline. The posterior part of the pyriform lobe is most markedly dependent, and descends to a much lower level than the rest of the hemisphere. This feature is much more pro- noimced than it is in any other individual in this heterogeneous group of mammals, and is probably a result of the obvious restriction to tin: expausiou of the brain in the longi- tudinal direction. The rhinal fissure is interrupted in a manner which suggests an analogy to the arrangement in the Armadillos. Tiiere is a very short horizontal anterior rhinal fissure, which is quite independent of the posterior rhinal fissure. The latter begins at the posterior margin of the hemisphere and arches forward, but just before reaching the anterior rhinal fissure it extends up\^ard into continuity with a pallial fissure, in a manner not unlike the posterior rhinal fissure in Dasyj^us. The fundamental features of the regions of the base of the brain which we have just been considering vary within relatively narrow limits in the Mammalia, although both Mono- tremes exhibit features in common which readilv distino;uish them from all other mammals. Such variations as we have found among the different representatives of this order are characteristic of the changes which the basal region of the hemisphere undergoes in the wider range of the Mammalia generally, and are in the main associated witli the habits of life rather than witli the systemic position of the individuals. For we find among other mammalian orders, more especially the Marsuj)ialia, Insectivora, Kodentia, Carnivora, and Ungulata, parallel modifications Avhich indicate the more or less direct causal relationship to the mode of life. All the areas which we have been considering are intimately associated with the olfactory apparatus, and, Avhatever other functions they may subserve, there can be no question that tlieir predominant r61e is to constitute a receptive area for incoming impulses of smell. The importance of the sense of smell varies considerably in the different members of this order, although in all it reaches a very high degree of acuteness, as the great development of the olfactory bulb and its associated cortical areas demonstrates. But in this macrosmatism the Edentata are like all other lowly-organized mammals, in which the sense of smell has a larger function than in higher mammals, in which an intelligence dispenses with the necessity of any such predominant and guiding sense as the impressions of smell convey. In Edentates like (Jvycteropus and the Dasypodiilce, which S])end their Jives mainly in digging and bru-roA^iug in the ground foi' t)bjects as to the position of which the sense of THE BKAIN IN THE EDENTATA. 301 smell is the main determinant, it is not surprising to find that the olfactory areas of the In-ain reach a relatively larger development than in the arboreal Bradyjiodida' and Cycloturiis, in \yliich the smell-sense is not of such vital importance. And in such a form as Chlamydophorus, which lives mainly underground and depends to a very slight degree upon its feeble visual organs, it is natural that the importance of the olfactory ajiparatus should be still further enhanced, as finds expression in a relatively enormous olfactory bulb and tuherculimi olf actor iitm, and a pyriform lobe which includes more than half the lateral sui'face of the cerebral hemisphere. In a short memoir upon tbe basal regions of the hemisphere, Gustav Retzius last year called attcjition to certain features of the pyriform lobe and the adjoining areas in some Edentates, among other mammals *. He makes special mention of the little " tubercle of the external olfactory tract " in Marsupials and Edentates. He dignifies this little body Avith the name (jyrus iufermedms rluueiicephuli (p. 107), but although he carefully describes its situation, he does not give us any information concerning its structiu'e. It will prove a source of considerable confusion if the term " gyrus " is applied to such structures as tliis, involving as it does a considerable distortion of the generally accepted idea of the meaning of this term. Retzius goes on to state that in the same animals {Bidelphys, Ilacroptis, Myrmecoplmgu, Basypus among others) two or three sagittal " gyri " are found behind the "gyrus intermedins." He calls these the "gyrus lunarls '* and " gyj-usambiens {medialis and laterulis),'" in conformity with a nomenclature he had previously apialied to the human brain f. He further mentions that in the Insectivora {Erinaceus) there Is a very large "gyrus lunaris " and a " gyrus ambiens.'" Retzius does not make any mention of the fact that, in Erinaceus and many of the Edentates, the fascia dentata (as we shall explain subsequently) extends on to the base of the brain in the region of his " gyrus lunaris.'' Its relation to the latter structure is well shown in the figure of the base of the brain in MyrmecopUaga (fig. 6). The Cerebral Commissures. In all vertebrates, series of nerve-fibres proceed from one cerebral hemisphere to the other, and serve to bring into functional association brain areas which are otherwise quite separate the one from the other. These fibres, whether they connect strictly homologous areas or constitute a symmetrical bond between heterologous parts of the two hemispheres, are generally known under the comprehensive title of " commissures." In all excej)t the lowliest vertebrates there are two compact and \Aell-defined bu.ndles of these crossing fibres forming ventral and dorsal commissures of the cerebral hemisphere. Within the class of mammals the constitution of these two commissures becomes rearranged in order to more readily accommodate the enormously increasing number of fibres which * Gustav Eetzius, " Zur Kenntuis der Wiiiduugen der Riechhirns,"' Verhandl. Anat. Gesellsch. 1897, p. 10.5 et seq. t Gustav Eetzius, ' Das Menschenhirn,' 1896. SECOND series. — ZOOLOGY, VOL. VII. 42 302 DR. a. ELLIOT SMITH ON are proceeding from that essentially mammalian cortical area, the pallium. In all mammals except the Monotremata and Marsupialia (in which the pallial fibres cross the mesial plane in the veiitral commissurel a large and increasing proportion of these pallial fibres invade the dorsal commissure nnd form the corpus callosum. When we remember that this structure has been acquired in its jii-esent form within the mammalian class, we naturally turn with great interest to the examination of the commissiues in such an enigmatical order of mammals as the Edentata. Pouchet says * that the Edentates, considered as a whole, present a remarkable diminution in the size of the corpus callosum. He says that in a young Choloejjns, which was 19 cm. long, the corpus callosum was 7"5 mm. long and 0'5 mm. thick. la the specimen of Cholcppns Hoffmanni which Turner figures t the corpus callosum is repre- sented as being 11 mm. long and 1 mm. thick in a hemisphere whose maximum length (without the olfactory liulb) is 8G mm. In Elower's memoir there is an enlarged figure of the mesial surface of the brain of a Choloppus d/df/cti/lns J, but unfortiinately the degree of magnification is not indicated. In a hemisphere 70 mm. long, he represents a corpus callosum 23 mm. long and 2-8 mm. thick. Judging fi-om the average dimensions of the adiilt hemisphere of Cholcejms, these measurements are probably about twice the actual size of the objects. In my specimen of Cholcepns didactyhis (which had been in alcohol for some years) the cerebral hemi- sphere is barely 35 mm. long, and the corpus callosum measures 10'75 mm. loug and I mm. thick. We may safely conclude that the corpus callosum in the adult Two-toed Sloth is about II mm. long and about 1 mm. thick. We cannot speak with the same degree of certainty of the measurements in Bradyptis, because there is no information upon the subject in the past records, and my only specimen available for measurement is not fully grown. In the bi-ain of Bradypus tridactyliis which Professor Max Weber generously gave me, the cerebral hemisphere is 26 mm. long, and possesses a corpus callosum which is 6"75 mm. long, and 1 mm. thick. In the adult brain the cerebral hemisphere attains a length of about 32 mm. In this young Three-toed Slotli we have a corpus callosum which is much shorter, both absolutely and relatively (to the length of the hemisphere), than is the case in the Two-toed Sloth, but this shortness is compensated by a greater thickness, for in this small brain of Bi'adypus the corpus callosum has already attained to the same thickness which it readies in the much larger brain of the adult Clwloepns. Poiicbet §, after describing the corpus callosum in tlie Sloths, proceeds to descril^e that of the Ant-eaters. He says that in Cycloiurm aud Myrmecophaga it presents similar features to that of the Sloths. Tliis is a very loose and utterly misleading statement, for, as we shall see subsequently, there is a very marked contrast between the features presented by the two families in this respect. He gives the measurements of the corpus callosum in Myrmecophaga as * Op. at., Journal de rAuatomie et de la Physiol, tm. vi. p. 308 et seq. t Turner, o^<. cit., Journal of Anatomy and Physiology, vol. xxv. fig. 15. X Flower, 0^1. cit., Pbil. Trans. 18(55, pi. xxxvii. fig. 5. § Oj,. at. p. 308.; THE BK.AIN IN THE EDENTATA. 303 2G mm. in lengtli aud 2'5 mm. iii thickness. The corjjus callosum in tlie specimen which Forbes figures is about the same length, but not quite so thick. The cerebral liemisjihere in Ilijrmecophaga is usually about 54 mm. long, so that the corjnis callosum is almost half tlie lengtli of the hemisphere. In a specimen of Tamaudua, whose hemisjjhere is 34 mm. long, the corpus callosum increases 14 mm. in length, and its narrowest part 1 mm. in thickness. Pouchet * rejjresents the mesial surface of tlie cerebral hemisphere of a Cyclutaras, the corpus callosum of which is very narrow and 6-5 mm. long, the hemisphere itself being 17 mm. long. Thus we see that in the brains of the ]\Iyrmeco2)hagidse the corpus callosum is rela- tively shorter in proportion as the animal is smaller, the corpus callosum measuring respectively 48 per cent., 41 per cent., and 38 per cent, of the length of the hemisphere in the three genera. In all the Armadillos the corpus callosum is very small. Pouchet describes the corpns callosum in a young Dasypus, whose cerebral hemisj)here is about 23 mm. long, as being 0 mm. long and 1 mm. thick. Turner represents f in Dasyijus sescinct us the cerebral hemisphere 30 mm. long, possessing a corpus callosum 6 mm. long aud 1 mm. thick. lu the brain of Xenitrus, the cerebral hemisphere of which is 25 mm. long, we tind a corpus callosum measuring o mm. in length and barely 1 mm. in thickness (in a specimen preserved in alcohol). In ToUjpeutes the proportion is about the same. In a cerebral hemisphere of Chlamydophorus, 11-5 mm. in length, we find a diminutive corpus callosum of 2-25 mm. length and about '25 mm. thickness. In all the xirmadillos, in spite of the small dimensions of the cerebral hemisphere, the proportion between the lengths of the corpus callosum and the hemisphere is much smaller than it is in the Sloths, and, a fortiori, than in the Ant-eaters. In his figures of the brains of a Ilauis, Max Weber represents J a corpus callosum G mm. long and 1'5 mm. thick in a hemisphere whose maximum length is 24 mm. Pouchet represents ^ a corpus callosum 4'5 mm. ioug aud 0 75 mm. thick in the brain of a young Manis, the hemisphere of which is 20 mm. long. In a cerebral hemisphere of the same length as Weber's specimen I have found a corpus callosum 6 mm. long, but barely 1 mm. thick. The disparity between these measurements of the thickness is probably explained by the fact that Max Weber has not figured the supracallosal vestige of the hippocampus, but represented it, together with the corpus callosum, as one structure. In Orycteropus I have found that the corpus callosum measures 21 mm. in length and 2 mm. in thickness in a hemisphere whose maximum length is 60 mm. Although the corpus callosum becomes reduced in some families, especially in the Armadillos, to very diminutive proportions, it is very misleading to state as a general rule that the corpus callosum is small in the Edentata. To appreciate the full significance * Op. fit. pi. iv. fig. 4. t O^). cit. vol. XXV. fig. 13. J Max Weber, o^j. cit. tab. ix. fig. 69. § Pouchet, 0/1. cit. pi. iv. fig. 10. 42* 304. Dll. G. ELLIOT SMITH ON of the evidence to be gained by a study of the cerebral commissures, it is of much greater impoi'tance to investigate their constitution and morphology than to judge them on a purely quantitative basis. In separating the two cerebral hemispheres the one from the other by means of a mesial sagittal section, we cut throvigh the large white masses of the cerebral commis- sures, which will be seen to consist of a small ovoid ventral commissure and a large elongated dorsal commissure of a peculiar shajie. This dorsal commissure in Orycteropus is composed of two divaricated limbs, which meet and fuse posteriorly. The larger dorsal limb is the corpus callosum. Arching downward and backward from the ventral surface of the caudal extremity of the coi'pus callosum, there is to be found a peculiarly modified cortical area known as the hippocampus. The greater part of this peculiar structure is submerged below the general level of the cortex, and its position is indicated upon the surface by the hippocampal fissure. The cephalic margin of the hippocampus is fringed by a band of white fibres — which is known either as fimbria or forma- — and which is plainly visible upon the mesial surface of the hemisphere. After these preliminary remarks we are in a position to ajjpreciate the features of the cerebral commissures. If we examine the mesial surface of the brain of Orycteropus which is exposed by a mesial sagittal section (fig. 4, p. 291), we find the lamina terminalis extending obliquely upward and forward from the oj)tic chiasma to reach the ventral or anterior commissure, which consists of a large mass of transverse fibres presenting an oval outline in section and measuring 4-5 mm. by 3"5 mm. Proin the dorsal aspect of the ventral commissure the upper part of the lamina terminalis proceeds obliquely backward and uj)ward to the inferior extremity of the ventral limb of the great bilaminar dorsal commissure. The ventral limb of this dorsal commissure consists of fibres which are proceeding from the hippocampus of one hemisphere through the fornix to the other hemisphere, and it is custoniaiT to call these crossing fibres of the fornix the psaltcrimn *, from a fancied resemblance in the arrangement of their homoloo-ues in man to a stringed instrument. The psalterium in Orycteropms consists of a large ventral fusiform mass of fibres, which form the psalterium ventrale, and a slightly narrower upper part, the psalterium dorsale, placed immediately below the dorsal limb of the great commissure, which latter is best known by the designation corpus callosum. The psalterium dorsale joins the caudal extremity of the corpus callosum, which consists of a large rounded mass of fibres known as the spleninm. The corpus callosum is a larger structure than the psalterium, and, unlike the latter, is of fairly uniform thickness, except at its cephalic and caudal extremities, which are thicker than the rest of the commissure. It is placed obliquely, and its posterior part is parallel to fhe psalterium dorsale; h\xt n^ the psalterium ventrale l)ends downward toward the ventral or anterior commissure, a triangular interval is left l^etAveen the j)salterium and corpus callosum. This interval is filled by a mass of grey substance which will be subsequently described as the corpus pai-acommissurale, and which in part represents the septum lucidiim of human anatomy. * FsnJtc)-iuin,^lifr(i, = coiniiiissur(t forn(cis.=^commis^Hrit hi/ipocaiii/ii. THE BRAIN IX THE EDENTATA. 305 If we compare the commissures iu O/'i/cterojms \Yith those of Ilyrinecophacia, we find that the ventral (anterior) commissure is approximately of the same size or even slightly smaller in Ilyrmecophaga, hut the dorsal commissure presents a marked contrast. The fsalterinm is much longer and more markedly attenuated than it is in the African genus, while the corpus callosum is greater, more especially in length, and more nearly horizontal than is the case in Orycteropus. The anterior commissure in a representative of the latter genus raeasui'es 4"5 x 3'5 mm., and in a specimen of Ilyrmecophaga 3-5x3 mm. ; while the respectiA^e measure- ments of the psalterium are about 12x2 mm. in Orycteropus and about 17 x less than 1 mm. in Ilyrmecophaga. The corpus callosum of the latter, measuring about 26 mm. long and with an average thickness of about 2 mm., greatly exceeds in size that of Orycteropus, which is only 21 nmi. long, and about the same tliickness as that of Ilyrm ecoph aga. The condition of the commissures in these two geneva points to a most decided superiority in the brain of Ilyrmecophaga, when we recollect that the two brains are approximately equal in size. The anterior (ventral) commissure is larger in Orycteropus because the olfactory bulb and pyriform lobe, from which it is mainly derived, are larger than they are in Ilyrmecophaga. But the larger corpus callosum in the Great Ant-eater points to a marked superiority in pallial development, /. e. a distinctly higher state of cerebral organization, in this animal than in the African Orycteropus. The psalterium of Ilyrmecophaga is more attenuated than that of Orycteropus, partly because it has been subjected to a greater amount of stretching * by the larger corpus callosum, but partly because it is actually smaller in the former, a result of the smaller dimensions of the hippocampus. In the cerebral commissures of Tamaiidua we find a considerable family resemblance to Ilyrmecophaga,, making due allowance for the smaller size of the former (fig. 10). The anterior commissure, measuring about 1"5 x 1*25 mm., is placed in a vertical lamina termiualis, which is attached dorsally to the ventral extremity of the psalterium ventrale. The psalterium is elongated and attenuated as it is in Ilyrmecophaga, Ijeing about 9 mm. long, and reaching a maximum thickness of 0'75 mm. ia the plumpest part of the psalterium ventrale ; biit the psalterium dorsale cowsisXa of an extremely thin sheet of crossing fibres. The elongated corpus callosum is more oblique than it is in Ilyrmeco- phaga, and is nearly 14 mm. long and 1 mm. thick. The anterior and posterior extremities of the corpus callosum are distinctly thickened. As far as Pouchet's figure allows us to express an opmicm, Cycloturus also presents the typical elongated corpus callosum which seems to be the distinctive character of the family of MyrmecophagicUe among the Edentata. In brains of approximately the same shape the proportion between the lengths of the corpus callosum and the cerebral hemi- sphere may be taken as a rough estimate of the degree of functional perfection of the pallium. Increase in extent of pallium and an increasing richness of texture will find expression in an increased number of crossing fibres. * Concerning this stretching inflnence of the corpus callosum, see '.Journal of Anatomy and Physiologj%' vol. sxxii. 1). 41. 306 DE. G. ELLIOT SMJTIl UN In Myrmecophaga the leugth of the corpus callosum is about 48 per cent, of the length of the hemisjihere ; in Tamaudna it is about 41 per cent., and in Cycloturus it is about 38 per cent. In Orycteropus, an animal of similar habits , and approximately equal size to Myrmecophaga, the corpus callosum, which is about the same thickness as it is in Myrmecoplwga, is only 35 per cent, of the length of the hemisphere. The cerebral commissures in the Sloths present a marked contrast both in appearance and in size to those of their relatives — the American Ant-eaters. We have just seen that the characteristic feature of the corpxis callosum in the latter is its great length ; the effect of this elongation is most noticeable in the caudal direction, and as the upper extremity of the psalterium is attached to this extremity of the corpus callosum it follows that the psalterkmi dorsale becomes markedly attenuated. Fig. 17. psalteriam hippocamp. vestigia bulb, olfaot. ' area praecommiss. tubercul. olfact Corp. oallos. tela fisB. prima / Corp. qaadrigem. aquaeduct, Syl. vel. medallare '\ pone Varol. lob. pyriform. ooramisa. moll. oommiss. ant. / tract, opt'. ventr. III. Brain of Bradypvs tridactylua, jiiv. ; surface exposed by mesial sagittal secli Tela refers to roof of third ventricle. Enlarged 2 diam. In the Bradypodid(B we find a vei'v short corpus callosum ; in three specimens of Cholcepus tlie average length is 30 per cent, of the length of the hemisphere ; and in a young specimen of Bradypus the j)roportion is as low as 26 per cent. These proportions are all the more noteworthy if we recall that the hemisphere in the Sloths is itself relatively short compared with that of the Ant-eaters. The shortness of the corj)us callosum explains the peculiar shape of the psalterium. For, being free from the stretching influence of a caudally elongating corpus callosum, the psalterium remains as a small plump mass of fibres, with no division into p)saUeritmi dorsale and ventnde. Tiiis little mass of fibres is thickest at its junction with the corjjus callosum, and rapidly tapers to a rounded point as it extends downward and forward, forming an angle of about 30° with the corpus callosum. It is about half the length of the latter. Flower soys * that "the anterior end [of the corpus callosum] is simple Flciwer. Phil. Trans. 18G.5, ]). 630. THE BRAIN IN TH]'] EDENTATA. 307 and obtusely pointed, without a trace of tlie reflected rostrum." In my specimen of Cholrepits, as well as in the representative of Brmhjpus, the anterior extremity of the corpus callosum is fuller and not pointed as seems to have been the case in Plower's specimen. There is, as Flower remarks, no reflected rostrum or genu in Cholcepus, but in my specimen of Brad>jpuii there is a most decided genu, for the plump anterior extremity of the short corjius callosum j^resents a considerable curve in the ventral direction. The shape of the psalterium and its relation to the corpus callosum in the Sloths recall that peculiar reversion to a simpler type of cor2)us callosum which we have previoiisly met in Nyctophilus and other Bats *. The difference is a considerable increase in the size of the corpus callosum in the Bj-adypodidce. Ill the Sloths the anterior commissure is relatively smaller than it is in either Ilyrmecophaga or Orycteropus, probably because the pyriform lobe is relatively smaller in the former. In three schemes which I have prepared to illustrate the vai-iations in the hii^pocampus in the three American families {vide infra, tig. 23), the shapes of the commissures are clearly demonstrated and their marked contrast shown more forcibly than any description can picture them. In the Armadillos we meet with yet another type of dorsal commissure which markedly contrasts with both of the types — those of the Bradypodidie and Myrmecophagid(e — whicli we have just described. A representative of this type is seen in Dasypus viUosus. Here we find a dorsal commissure composed of two limbs — corpus callosum and psalterium — of approximately the same size, but of different shai^es. Both of these limbs are placed very obliquely and in contact one with the other for the greater part of their extent, only a very small part of the extreme anterior end of the corjous callosum being separated by any interval from the psalterium. As a result, there is practically no septum lucidum, in the sense in which that term is applied in human anatomy. The anterior extremity of the diminutive corpus callosum is pointed and depressed. The psalterium and corpus callosum each consist of a club-shaped mass of fibres of approximately the same size. The thick end of the club in the case of the psalterium consists of the cut edge of a vertical wall of crossing fibres, which are placed above the anterior commissure. This mass is the pjsalteriiim ventrale. At its upj^er extremity it tapers to the handle of the club, ^nliich is composed of the psalterium dorsale. This is short and is inclined slightly backward, and becomes continuous with the thick end of the club-shaped mass of the corpus callosum, which is placed upon the dorsal aspect of the psalterium dorsale. The corpus callosum is extremely oblique, and rapidly tapers to a point anteriorly aud below. A very slight interval is left between the anterior extremity of the corpus callosuui and t\\e pjsalteriuin ventrale. In Turner's description tiud figure of the brain of Dasypus sexcinctus t there is no indication of any distinction between Wxe ptsalteriuni dorsale and ventrale, and the whole * Cf. this Vol. p. 47. t Turner. «p. cil., 'JouriKil of Anatomy and Physiology,' vol. i. 308 DK- G. ELLIOT SMITH ON dorsal commissure is represented as an obliquely-placed, inverted, U-sliaped structure, with symmetrical limbs of approximately the same shape and size. It seems probable, however, that there is a very definite attenuation and bending of the upper part of the psalterium in both species of Dusyjnis. In Tolypeutes tricinctus. Garrod * represents a dorsal commissure of the same shape as that represented by Turner in Dasy^ous, but with a corpus callosum slightly larger than the psalterium. The specimen of Tolypeutes at my disposal was not sufficiently well-preserved to permit me to supplement Garrod's brief notes. The commissures in the brain of Xeimrus unicinctus have not hitherto been described or figured, so far as I am aware. In the well-preserved specimen representing this genus in the College of Surgeons, the dorsal commissure very closely resembles the appearance figured by Turner in Dasypus f . We can say that in the three genera Dasypus, Tolypeutes, and Xeimrus, the corpus callosum is very small and pointed, obliquely situated, and of apj)roximately the same size as the jssalterium. In the specimen of Tatusia in the College of Surgeons, the psalterium presents features similar to those which we have described in Dasypus villosm, but the corpus callosum is considerably larger than the psalterium, and is plumpei', longer, and of more unifoi'm thickness than the corpus callosum of the other Armadillos. In this specimen the dorsal commissru'e as a whole w^as not unlike that found in the Hedgehog {Erinaceus) and in the Manidfc. Such being the case, it is surprising to find the following observations of Poucliet % : — " Sur la coupe du cerveau dc I'Encoubert [Dasyj^iis] le corps cfilleux mesure 6 millimetres de long et 1 millimetre d'epaisseur; c'est chez un jeune Cachicame [Tatusia] qu'il nous a montre les proportions les plus exigues. Nous I'avons vu aussi, sur cet animal, nettement recevoir des fibres de la partie anterieure et de la partie posterieure de I'hemisphere qui viennent s'y jeter en longeant la scissure mediane." It is probable that the immediate ancestors of the Dasypodida' possessed a considerably larger corpus callosum than the existing types, for otherwise it would be difficult to understand the stretching of the psalterium. If this be so, the corpus callosum has undergone a retrogressive diminution in size, and the condition in my specimen of Tatusia, rather than that of the other Armadillos, would be nearer the j)rimitive one. In Chlamydopjhorits we find a very peculiar dorsal commissure conforming to the same type as Dasyims villosus, but in which the retrogressive changes resulting in the diminution of the corpus callosum have gone much further. We find a large psalterium divided in a typical manner into dorsal and ventral portions, and a cor2)us callosum so reduced in size that it is only slightly larger than the psalterium dorsale. This is one of the most extreme forms of reduction of the corpus callosum met with in the Eutheria. If this process of reduction were to be carried much further in Clilamydopliorus we should reach a state of cerebral modification wdiich we find in the Marsupial Notoryctes. * Garrod, op. cit., Proc. ZooL Soc. Loudon, 1S7S. t Unfortunately, in tlie jirocess of mounting, a glass rod bad l.ieen jushcd through the junction of the psalterium and corpus callosum, thus rendeiing imjitssible an acC'urate description of this region. J Pouchet, op. cit. torn. vi. p. 309. THE BKAIN IN THE EDENTATA. 309 There is a marked contrast between the fomily type of dorsal commissure met witli in the Brudijpodidie and the equally Avell-defined family type of the Dasupodidtc (fig. 23, p. 320). In the latter we find that specialized form of psalterium which we have seen in the Ilyrmecophagidce and Orycteropus, and which is found in the vast majority of Eutlieria —a form of psalterium which we associate with a fully-developed corpus callosum. The diminutive size of the corpus callosum in the Armadillos therefore sviggests a retro- gressive development from some form witli a large corpus callosum. Eig. 18. hippocsmp. vestigia Corp. callos. Corp. quadrigem psalterium bmbria faBCia deotata Jiippocamp. inversus tract, opt tobercul. olfact. ; lob! pytiform. poD8 Varoi. rorji, interped- commiaa. moll. Jytiform. vent, i fisB. hippocampi Brain of Cholapiis duhniiihis ; surface exposed by Ille.■^lal sagittal section. Enlarged .! diam. The brain-stem has been cut away from the hemisphere. In the Bradijpodidiv we tind a marked contrast in the shape of the commissures. Here we have a corpus callosum Avhich lias attained a much larger size than in the Dasypodida', though we must still regard it as relatively small and primitive. The psalterium presents a simple form in the Sloths, which a comparison with the marsupial would point out as a primitive type. In other words, we might regard it as the result of a progressive development from some simpler type rather than a retrogressive modifi- cation of a more highly organized form, as may have been the case in the Armadillos. In all the Armadillos the ventral or anterior commissure is relatively very large, just as we should expect from the large size of the pyriform lobe. In the brain of Manis the corpus callosum is very short, being barely 6 mm. long in the specimen at my disposal, and almost of imiform thickness throughout. It does not taper anteriorly, nor, on the contrary, is it thickened or bent anteriorly to form a genu. In my specimen the corpus callosum was not nearly so thick and plump i]i proportion to its length as that figured by Ma.^ Weber *. Max Weber has not figured the whole of the psalterium, but this structure in my specimen appears to be clearly composed of two parts — a ventral and a dorsal, neither of which is very plump. So that, on the whole, the dorsal commissure of Jlanis is mitcli more like that of Eriiiaceus than any of the Edentates ; for, while the psalterium resembles that found in the Dasypodldce, the corpus callosum is more like that of the BradypodidcB in shape, size, and position. The anterior commissure is moderately large. * Max Weber, op. cit., ' Zool. Ergebnisse,' ii. tab. i.v. fig. (j9 co. SECOND SERIES. — ZOOI.OGY, VOL. VII. 43 310 DE- G. ELLIOT SMITH ON In reviewing the features presented by the cerebral commissures in the Edentata, we find that, while each family presents a more or less distinct type of dorsal commissure, all of the commissvires conform to the type prevailing in the Eutberia generally, and an analogue for almost any Edentate type may be found among the Eutberia. There is no feature in the arrangement of the commissures which indicates a transition stage between the primitive Saiiropsidan -like type which prevails in the Prototheria and the Eutberian brain with a well-formed coi'pus callosum. It may be that the gradual dwindling of the corpus callosum which is obviously going on in the Armadillos may indicate the mode by which the Marsupials might have lost their corpus callosum, when their dorsal commissure assumed a resemblance to that of the Monotremes in being purely hippocampal. Eor the possibility of the Marsupials having once had a corpus callosum and subsequently lost it is forced upon our consideration by the dwindling of the anterior and dorsal parts of the hippocampal arc — changes which we associate causally in the Eutberia with the development of a corpus callosum. But the discussion of the question whether the ancestors of the Marsupialia originally had a corpus callosum must be postponed for a future memoir. There are certain general questions relating to the cerebral commissiu-es which may be discussed to better })urpose after we have considered the cerebral cortex. The Hippocampal Formation. We have seen elsewhere * that in the Monotremata and Marsupialia the hippocampus retains a peculiarly simple arrangement with relation to the margin of the hemisphere, which it has inherited from the ancestors of the Mammalia. In approaching the study of this region of the brain in the Edentata, I have deemed it important to enter with some detail into the exact arrangement of the hippocampus, in the hope that some indication of the affinities of the Order might be obtained. Extending downward and backward from the postero-iuferior aspect of the psalterium, we find upon the mesial surface of the hemisphere (fig. 4, p. 291) two or more peculiar arcuate bands, which represent all that can be seen upon the surface of the pecviliar hipjjo- campal formation — the homologue of the hippocampus major of human anatomy. If, before beginning the study of these ])eculiar surface areas of the hippocampal formation, we examine this structure from within by opening up the cavity (lateral ventricle) of the hemisphere, we shall gain a much clearer conception of the region than would be the case otherwise. If we dissect away the lateral wall of the hemisphere of Ori/cterojjus so as to expose the mesial wall of the lateral ventricle, we find a large crescentic white mass bulging in the posterior part of the ventricle, looking not unlike the pupa of a silkworm lying in its cocoon. This large curved swelling is the hippocampus. Its concave anterior border is fringed by a prominent ridge of compactly arranged fibres — the fimbria. The lower extremity of the swollen mass is pointed, and the upper extremity, which extends slightly * Cf. this Vol. p. 47. THE BRAIN IN THE EDENTATA. 311 further forward than the lower, is rounded and is placed upon the dorsal side of the broad anterior end of the fimbria. ventr. olfact, fiss. rhinal. : ped. olfact. bulb, olfact. lob. pyriform. .' ; tract, opt. ; liiPPOcamP- Corp. paracommisB. • fimbria commiss. ant. Left hemisphere of Orycteropus ; the lateral wall removed and the lateral ventricle opened. Nat. size. x. " Complete " invagination of the ventral wall of the olfactory bulb. If we make a horizontal section through this hippocampal swelling at about the middle of its extent, we gain a more exact idea of the extent of the prominence and the relationship between those parts of the hippocampal formation which make their fascia dentata inversus . fascia ; dentata: fimbria fiss. hippocampi pallium Semi-schematic representation of a transverse section through the hippocampal formation in Ori/ctcropus. Enlarged 4 diam. X. Hippocampo-pallial junction, i/. Corona radiata. appearance upon the surface and the deeper-lying parts which produce the large ventricular bulging. In such a section the hippocampal formation presents, roughly speaking, an elliptical 43* 312 DR. G. ELLIOT SMITH ON outline, and the greater part of its free surface is covered by a very definite layer of medullated nerve-fibres — the alveus. The anterior border of the formation is rounded, and from the antero-mesial part of the cui've a small triangular mass of fibres — -the fimbria — projects. This spur serves to indicate the line of demarcation between the ventricular and extrave utricular surfaces of the hippocampus. The whole of the ventricular and the greater part of the extra- ventricular surfaces (in this section) are covered with a A'ery well-defined layer of medullated nerve-fibres — the alveus. Extending in the lateral direction from the fimbria, the alveus extends outward and then sweeps in a large curve caudally and extends back- ward as far as the posterior aspect of the ventricle, of which it forms the mesial wall. From the extent and degree of curvature of this large alveus-coated sui"face in the figure, we can appreciate more readily the great size of the Ijulging of the hippocampus into the ventricle (fig. 19). The alveus gathers fibres from the whole of the deep or ventricular surface of the hippocampus and conveys them by an oblique course to the iimbria, which is composed of a large mass of such fibres gathered into a compact bundle. But the alveus is not confined to the ventricular aspect of the hippocampus. Eor we find an extraventricular alveus extending backward from the fimbria upon the mesial surface of the hemisphere for a considerable distance until it meets a grey band of peculiar constitution — the fascia dentata. The fascia dentata is morphologically the extremely specialized margin of the costa, which has undergone a peculiarly modified hypertrophy in its superficial layers, resulting in that curiously folded layer which is known in human anatomy as the fascia dentata. [It is not advisable to apply the term gyrus to this structure, as many writers, following- Huxley, are in the habit of doing, because it is not a gyrus in the ordinary sense of the term, but merely the peculiarly modified sujjerficial region of a cortical area.] An exceedingly small fragment of the fascia dentata makes its appearance upon the surface in the section we are considering, since almost the whole of its real surface is hidden from view l)y being opposed to the morphological sui'face of the hippocampus and the adjoining cortical area. In the section we find the superficial fragment of the fascia dentata at the posterior extremity of the extraventricular alveus ; from this point the fascia dentata extends for a considerable distance in the lateral direction and in apposition with the area of cortex which adjoins the hippocampus proper. The fascia dentata then bends forward, its surface now coming into contact Avith the morphological surface of the hippocampus ; in the rest of its course the fascia dentata pursues a course parallel to the alveus : that is to say, it pi'oceeds forward and then curves mesially and then slightly backward. The potential cleft which extends into the cortex in the interval between the fascia dentata and the geoeral cortex and separates t];e morphological surface of the fascia dentata from the surfaces of the hippocampus and the adjoining cortex is the fissura hippocampi. The deeper parts of this "fissure" are merely potential, because the surface of the fascia dentata becomes adherent to the opposed surfaces, especially that of the hippocamjjus. The hippocampal fissiu'e is unique in its nature and mode of formation, and cannot be classed with any other fissures of the brain. THE BRAIN IN THE EDENTATA. 313 Behind the liippoeampal fissure, in the section we are considering, we find the general cortex forming' the posterior lip ol" the fissure. If we trace this cortex toward the hippocampus, we find that its superficial layer becomes directly continuous with that of the hippocampus. The change from general cortex to hippocampus takes place opposite the bend in the hippocampal fissure, and consists essentially in the levelling down of the scattered cell-elements of the general cortex into a regular column of cells of peculiar shape, which characterizes the hippocampal formation. We may call this region the hippocampo-pallial junction. From this point the hippocampus proper extends forward, its deeji surface coated with alveus forming the mesitil wall of the ventricle, its real surface being in apposition Avith that of the fascia dentata. The hippocampus, still maintaining these relations, extends forward, then curves around so that its "deep" alveus-coated surface sweeps beyond the fimbria and thus makes its appearance upon the surface of the hemisphere. In this manner the morphologically deep as^iect of the hippocampus actually forms part of the surface area of the hemisphere ; in other words, part of the hippocampus becomes completely invested, with its " superficial " area excluded from the actual surface and its " deep " aspect exposed. This area of the surface of the brain, which is placed between the fascia dentata and tlie fimbria, and is covered by extraventricular alveus, may be termed the " hipjiocampus inversus." The whole of the " morj)hological surface" of the hippocampus proper is hidden from view in tliis section because it is submerged in the depths of the hippocampal tissui-e. After these considerations we may better appreciate the appearance of the mesial surface of the hemisphere. In a mesial view of the hemisphere of Orycteropus (fig. 4) we see the fimbria beginning below at a point just beliind the optic chiasma as a very narrow band and extending obliquely upward and backward, then curA'ing upward and ultimately horizontally forward to the situation of the psaltcrium ventrale. It rapidly increases in breadth during its course from below upward. If, instead of examining this region from its lateral aspect, we look at it from below (fig. 21), we find tliat for its upper half the fimbria follows a very oblique course forward and inward, so that anteriorly it approaches close to the mesial plane, and a large proportion of its fibres extend across to the other side of the Ijvain, thus forming the 'psalterlam ventrale. The fimbria presents analogous features in all the other Edentates, in common with most lowly mammals. Lying behind the fimbria, we find in Orycteropus the area of inverted hippocampus which separates the fascia dentata from the fimbria. As we trace these structures upward, we find that the area of inverted hipjDO- campus rapidly tapers, and at the same time the fascia dentata In-oadens and approaches the fimbria. The exact arrangement of the upper endings of these three bands will be discussed after the inferior' endings are considered. In Orycteropus the inverted hij)po- campus and the fascia dentata appear to end suddenly below in a deep arcuate furrow (fig. 1) which marks the upper limit of the peculiar hippocampal tubercle, to which a bj-ief reference has already been made. As this hippocampal tubercle exists among the Edentata in Orycteropnis only, we may witli advantage consider the mode of termination of the hippocampus in the other forms first. In all Edentates we find the three super- ficial bands— ;;^"/«Z//7Vr, hippocanqjits inversus, and fascia dentata — which we have noted 314 DE. G. ELLIOT SMITH ON in Oiycteropns. As we trace these bands downward and forward in any Edentate other than Orycteropiis, they will be found to taper and end simply just behi]id the optic chiasma. This mode of ending is shown in the figure of Tamanchia (fig. 10), and Max Weber's figure shows it in Mauls. In his account of the state of the hippocampus in Fis. 21. hippocamp. nudus fascia dentata fascia dentata f- hippocamp. inversus Corp. striat fimbria / . ' ^ psalterium l'"'^' P^aoonimiss. colum. fornicis fiss. hippocampi Dissection to expose the ventral surfaces of hinder parts of the cerebral hemispheres in Orifcterojnis. Enlarged ^ diam. .V. Purrow corresponding to hippoeampo-pallial junction. Cholopjms, Flower * mistakes the inverted area of hippocampus for part of the fascia dentata. In many of these brains the inferior extremity of the fascia dentata may be seen, upon the ventral sixrface of the brain, in direct continuity with that part of the pyriform lobe Avhich Gustav lletzius calls the " gt/rus lunaris." This is shown in our figure of 3fi/)-mecophaga (fig. 6), in which the antero-inferior extremity of the fascia dentata makes a peculiar bend toward the fimbria and almost at right angles to the rest of its course. In Orycteropus, however, we find, the region of the ventral extremity of the hippo- campal formation occupied by the peculiar large oval swelling which I have called tubei'culwn hiiiiJOcumxn. We may at a glance appreciate the significance of this peculiar structure if we remove the lower extremity of the hippocampal formation from the rest of the brain and examine it from the front, so that we may at the same time see part of the ventricular surface of the hip^iocarajms and the whole of the extra ventricular jiarts (fig. 22). We then see the fimbria descending and rapidly vanishing as a distinct ridge by the scattering of its fibres over the lower parts of the hippocampus. In front of the fimbria we can see the lower part of the ventricular surface of the hippocampus, which we haA^e already seen in profile (fig. 19). We see this alveus-coated surface of the hippocampus sweeping around the lower extremity of the fimbria and becoming extra- ventricular as the tuherculum Idppocampi. The latter structure is obviously nothing else * Flower, op. cit., Phil. Trans. 1865. THE BEAIN IN THE EDENTATA. 315 thau a promiuoiit boss of inverted hippocampus covered with, an exceedingly attenuated coat of extraventricukr alveus, and therefore serially homologous with the flattened .-v..^°-"'' KO"^''' 4*8 fimbria f ; ,• " ^.f^^^^j^ Jentata alveus W^ '■ fiss. hippocampi tubercul. ,.■%. J/r^^^ hippocampi xii^:^^^ 'ob. pyriform. "x Dissection to expose ventral extremity of the hippocampal formation in Ori/cterojnts. Nat. size. .!-. Attenuatcfl lower extremit_v of fascia dentata. band of inverted hippocampus which separates the fascia dentata from the fimbria. But the lower extremity of the fascia dentata has not entirely disappeared, for it is repre- sented by an extremely attenuated band which separates the hipjiocampal tubercle from the pyriform lobe just as the "band of Giacomiui " does in the human brain. The tuhercidum hippocampi is exactly analogous to the tip of the uncus in human anatomy, which E,etzius has recently called the gyrus iiitralimhicus* . Such an arrangement, so far as I am aware, exists nowhere else outside the Primates. As the hippocampal formation proceeds upward toward the psalteriiim and splenium of the corpus callosum it at the same time rapidly approaches the mesial plane, so that a perspective view, such as a representation of the mesial surface of the hemisphere affords, gives a very distorted picture of the hippocampal region. To gain an accurate idea of this region we must examine from below those large opercula-like cortical folds which form a dome-like roof al)ove the optic thalamus and corj)ora quadrigemina. Such a view may be obtained by making a horizontal section immediately above tlie anterior commissure and inverting the upper part of the cerebral hemispheres (fig. 21). In this view we may start from the typical section of the hippocampus which has been described above (fig. 20), and Avhich is now exposed once again. We see tlie fimbria extending obliquely forward and inward toward the columna fornicis, the elliptical section of which is seen on each side of the mesial plane. Many fibres of the fimbria enter the columna fornicis of tlte corresponding side ; many other fibres of the fimbria cross the mesial plane just behind the coliimncc fornicis and thus form the psalterium veiitrale. The area of inverted hippocampus which in the cut surface is placed immediately behind the fimbria may be observed to rapidly taper and soon disappear as the mesial plane is approached. But as the fascia dentata approaches the mesial plane it increases considerably in * Gustav lietzius, op. cU., ' Das Mensclieuhiru.' 316 DR. G. ELLIOT SMITH OiN" breadth : in other words, an increasing area of that part of the fascia dentata which is in contact Avith tlie cortex {ride figure 20) becomes exposed. Pari passu with this un- covering of the fascia dentata, the opposed surfaces of pallium and hippocampus, which hide it from view elsewhere, also hecome exposed. There is a general unrolling, as it were, of the hippocanipal formation. So that, as the fascia dentata becomes more and more exposed, we also find that the hippocampo-pallial junction approaches and ultimately appears upon the surface as a shallow furrow separ-ating the pallium, which lies behind it, from the exjjosed surface of the hippocampus, which now makes its appearance on the surface in front of the furrow (fig. 21, x). This exjwsed lup]iocampus {/dppocampus nuchis) is the true morphological surface of the hippocampus which elsewhere is submerged, hidden in the depths of the hipjiocampal fissure *. Near the mesial j)lane it appears to emerge from the hippocampal fissure, and under the name " Ballenioiiidung ," which Zuckerkandl introduced, it has given rise to much discussion. Zuckerkandl's " Ealkeuwiudung " is nothing else than our '■'■ hippo- campus nndusr As the fascia dentata becomes more ex2:)0sed and apparently broader it bends trans- versely inward, and when quite close to the mesial plane it bends suddenly backward and rapidly tapers to a point l)elow the splenium of tlie corpus callosum. These recurved portions of the fascise dentatte upon the two sides of the mesial plane are in close proximity the one to the other, and in many animals [c. (j. the Rabbit) they actually meet and fuse in the course of development f . But the two structures arc development- ally independent, and each is derived w holly from its own cerebral hemis2)here. If we now return to the consideration of the mesial surface of the hemisphere (fig. 4), we readily recognize the mesial border of this infrasplenial bent part of the fascia dentata as a pear-shaped body proceeding backward upon the imder surface of the psalterium dorsale and splenium, and appearing to be directly continued around the splenium on to the dorsal surface of the corpus callosum as a rovmded white cord. But the fascia dentata really ceases, as a definitely recognizable entity, upon the ventral surface of the splenium. The rounded white cord which surrovmds the splenium, and a2)pcars to be merely the attenuated fascia dentata, is more directly the upwai-d continuation of the hippocampus iiudus, i. e. the hippocampus proper. The fascia dentata, as its mode of develojiment indicates, is a comparatively late specialization of the margin of the hipjiocampus projier. As it develops it rolls itself, as it were, over the surface of the hippocampus and hides this from view. On the inferior aspect of the splenium the hippocampal formation unrolls, the hippocampus proper becomes once more exposed, the fascia dentata tapers and disappears almost entirely, and a remnant of the more primitive liif)pocampus proper surrounds the splenium and extends along the whole length of the upper surface of the corj)us callosum as a rounded w^hite cord. This cord-like remnant is a vestige of the anterior part of the more extensive hippo- * The true significance of this will be appreciated at a glance from fig. 21 of my short memoir in the ' Journal of Anatomy and Physiology,' vol. xxsii. p. 49. t Compare in this connection the comparative observations of Stieda in Zeitsch. f. wisscnsch. Zoologie, 1870. THE BRAIN IN THE EDENTATA. 317 campal arc which we find in Marsupials and Monotremes *. la Orycternpus this circum- callosal hippocampal vestige is a very ])rorainent rounded strand whicli we can readily follow with the naked eye around the spleuiura, along tlie whole length of the corpus callosum, and around its anterior extremity. But beyond this we cannot thus follow it. We shall see that in most of the other Edentates the mode of termination of this vestige of the anterior part of the hippocampal arc is more clearly exhibited than it is in Orycteropiis. Turning to the consideration of the brain of Tamandua (fig. 10), we find a typical hippocampal formation, three concentric bands of which — fimbria, inverted hippocamj)us, and fascia dentata — appear upon the surface. These three bands end simply below, i. e. without that peculiar modification which in Onjcteropus was described under the name hippocampal tubercle. At its upper extremity we find the same series of circumsplenial modifications of the hippocampal formation as we have already described above. But the infrasj^lenial bending of the fascia dentata is much further removed from the psa/terium ventrale than is the case in Orycteropus : in other words, the corpus callosum has grown relatively further backward, carrying the infrasplenial hipjiocamjial flexure {flexura hipp)ocampi) with it, and has extended the pi^^^ltermm dorsale to permit this separation from the psaUerimn ventrale. A large supracallosal vestige of the hippocampus may be recognized upon the dorsal aspect of the corpus callosum in Tamandua, but it is relatively smaller than the corre- sponding vestige in Orycteropus. The deep cleft which separates the hippocampal vestige from the overhanging pallium corresponds to the callosal fissure of human anatomy. It is not, as is often erroneously stated, the continuation of the hippocampal fissure, because the latter always ceases in the Edentates, as in most Eutheria, imme- diately below the splenium of the corpus callosum when the hippocampus unrolls. The hippocampo-pallial limiting furrow, Avhich emerges from the upper extremity of the hippocampal fissure, surrounds the splenium, and joins the fissura cnllosalls, is often mistaken for the hippocampal fissure itself, and hence the belief in the continuity of hippocampal and callosal fissures. Unlike the condition of affairs in Orycteropus, the Jissura callosalis in Tamandua becomes continuous anteriorly with a well-defined fissure which arches downward and forward to disappear in the deep cleft which separates the upper surface of the olfactory peduncle from the apex of the hemisphere. This is the ventral boundary of the pallium, and hence may be called fissura limitans p)allii. It serves to indicate to the naked eye the upper limit of the hippocampal vestige which we can discover along this line by histological examination. In Myrmecophaga wo find a close agreement with the condition just described in Tamandua. There is also a large and clearly-defined hippocampal vestige, which Forbes has failed to indicate in his figure of tlie mesial surface. However, he speaks of the fascia dentata " being continued, as described by Professor Turner in Dasypus, as a thin layer of longitudinally-disposed fibres over the corpus callosum to near its genu" f. * Cf. this vol. p. 50 ; also ' Journal of Anatomy and Physiology,' vol. xssii. t W. A. Forbes, op. cit., Proc. Zool. Soc. London, 18S-', fig. 4, and p. L'iJ4. SECOND SERIES. — ZOOLOGY, VOL. VII. 4,^ 318 DR- G. ELLIOT SMITH ON As a description of tlie mere appearance of the hippocampal vestige in Myrmecophaga this account is lucid enough, but if we literally iuterpret either it or the expression of Turner (Avhich it somewhat ti'avesties) as a statemeat of the actual constitution of the parts we shall gain a very erroneous conception. For, as tlie hij^pocamjjal formntion surrounds the splenium, the fascia dentata dwindles and practically disappears, leaving the diminutive remnant of the simple hippocampus proper upon the surface of the corpus callosum. This hippocampal vestige is very rich in longitudinal inedullated fibres, which in human anatomy are called the strice Lmicisii (as well as an immense variety of other names). This explains the statement of Turner that the fascia dentata is prolonged into tlie upper surface of the corpus callosum as a narrow band *, which, to the naked eye, it appears to be. It also explains the statement, which Eorbes wrongly attributes to Turner, that the fascia dentata is " continued forward as a thin layer of longitudinally- disposed fibres." Histological examination enables us to exactly interpret the puzzling macroscopic appearances and to state definitely that the hijipocampal formation, a complex of fascia dentata and hippocamj)us (in the narrow sense), extends up to the splenium of the corpus callosum and becomes continuous around tlie S[)lenium with a diminutive baud which represents the dwindled hippocampus from which the fascia dentata has practically vanished ; as this vestigial liippocampus contains a large number of me;lullated nerve-fibres, it is often mistaken for a purely fibrous structure. Although tbe figure of the mesial surface of the brain of Cycloturus which Pouchet gives + is lacking in detail, it suffices to show that a sulcus limitans pallii exists in a form exactly analogous to that of the other two representatives of the Myrmecopjhagldcc ; and the fact that his figure represents a vestigial hippocampus at all shows that this structure is more prominent than is usual among Eutlieria. Flower has given a useful figure and a descrij)tion of the hippocampal region in Choloepus X, but the corresponding region in Bradypus has not, so far as I am aware, been properly figured or described. In the brain of both of the Sloths the hippocampus presents the typical features in the greater part of its extent, such as we have already described in Tamandua, and which we find in most mammals. But in the upper part of its extent the hippocampus presents in both Bradypodidce certain features in common which distinguish it from that of the other Edentates. We have already observed that in this family the corpus callosum has not grown backward even to the same extent as it has in Orycteropus. As a result the psalterium, as we have already seen, is short and sim[)le, and in addition the upper part of the hippocampal formation has undergone very slight disturbance. In both Sloths the fascia dentata appears to extend directly toward the splenium without such a marked infrasplenial flexure as we find in the other Edentates. To the naked eye no area of exposed hippocampus makes its appearance behind the splenium, and the * W. Turner, op. cit., ' Journal of Anatomy and Physiology,' vol. i. t G. Poucliot, op. cit., ' Journal do I'Anatomie et do la Physiologic,' tome vi. pi. iv. fig. 4. J W. H. Flower, oj). cit, Phil. Trans. 18G5. THE BEAIN IN THE EDENTATA. 319 hippocampal fissure even appears to extend into contiuuity with the callosal fissnn;. Looking at the mesial surface of the hemisphere, the upper part of the fascia dentata appears to accompany the fimbria right up to the splenium and then to pass without any marked bending into the vestigial hippocampus which covers the dorsal surface of the corpus callosum. A large supracaliosal hippocampal vestige is found in both Sloths, and in my specimen of Bradijpus we have an interesting demonstration of the precallosal course of the hippo- campal vestige. Thus we find extending obliquely downward and forward from the front of the corpus callosum two shallow furrows which slightly diverge as they approach the olfactory peduncle (fig. 17). The upper of these furrows corresponds to the limiting fissure of the pallium in the 3Iijr7necophagid(B . The area included between these two furrows is the precallosal part of the hippocampal vestige. It appears broader than the supracallosal part of the vestigial arc because the latter is flattened horizontally upon the corpus callosum, while the precommissural vestige is vertical, being a constituent part of the mesial wall of the hemispheres. In representing the hippocampus from any one point of view it is impossible to convey an accurate idea of its exact disposition, because it bends about in various planes during its course so that in a perspective drawing it appears distorted. As it is a matter of great importance to clearly understand the behaviour of this importaut and peculiar part of the brain, I have drawn three schemes (fig. 23) : one of Tamandua, representing the Ant- eaters; one o? Choloepns, representing the Sloths, which Ave have already discussed; and one of Dasi/p'us villosus, representing the Armadillos, which we shall consider almost immediately. These schemes represent the actual relations of the hippocampal formation to the commissures in the three American families of Edentates. After the preceding descriptions these schemes are almost self-explanatory. In the Ant-eater (A) we see the three bands of fimbria (Ji-), inverted hippocampus (i.h.), and fascia dentata (f.d.) ascending in front of the hippocampal fissure (h.f.), just as they do in the Rabbit or any of the common Eutheria. As they ascend, the intermediate band (inverted hippocampus) disappears and the fascia dentata comes into contact with the fimbria. Tlie main portion of the latter continues its forward course to the p)salte)'lmn_ ventrale (;ps.v.), but a few of its uppermost fibres become widely scattered in tlie triangular interval bet^veen the subsplenial flexure of the fascia dentata (f.d.'} and the fibres going to the ventral part of the psalterium. Erom these scattered fibres the thin Taemhranous psalterium doi^sale (ps.d.) is formed. The fascia dentata (/.''/.) suddenly diverges from the fimbria and bends backward beneath the splenium (spl.) of the corpus callosum. As it does so it rapidly tapers and fades away. As this is taking place a part of the true hippocampus (u.h.) (using that term in the strict and exclusive sense) crops out of the upper part of the hippocampal fissure (hf). This little fragment of naked hippocampus {n.h.) becomes directly continuous with its atrophied serial homologue, the vestigial hippocampus {v.Ji.), which surrounds the corpus callosum (c.c.) and extends forward not only as far as the genu {(/.), but beyond this point toward the situation of the olfactory peduncle. U* 320 DR. G ELLIOT SMITH ON TiK. 2-3. Three schemes to represent the relation of the hippocampal formation to the commissures in the Mynnecoj^Jtcyidce (A), the Bradijpodidw (B), and the Basi/podida' (C). Reference Ictte e.c- .il)le to be in the presence of a corpus callosum. But the resemblance of the condition in the Sloths to that which we have found in certain small bats *, and which is obviously a reversion-type, suggests that the simplicity of the Sloth brain may possiljly not l)e primitive. At the same time Ave must recognize the possibility of the archaic simplicity of type being retained, just as it is, for instance, in the case of the auditory ossicles according to Kitchen Parker. In the Armadillos, again, we have the typical Eutherian plan exemplified. The relatively large vestigial hippocampus in these Edentata invited a minute study in the hope of finding some definite indications of its original structure; but here again * Cf. this vol. p. 58. 322 DK. G. ELLIOT SMITH ON we find tliat the structure is quite as atrophic as the relatively much more insignificant vestiges of the hippocanipus in the more highly-organized mammals. There is another reason of quite different natui'e for entering so minutely into the anatomy of this region in the Edentata. The clear demonstration of the hippocampal region which the Edentate brain affoi'ds rendei's it a particularly suitable object upon which to demonstrate the typical features of a region which has given rise to groattr perplexity, perhaps, than any other of the brain. Eor u.any years the question of the mode of ending of the dorsal extremity of the liippocampus has excited a great deal of controversy. Several investigators have from time to time given an acciu-ate account of its behaviour in different mammals. Thus Henle and Giacomini long ago demonstrated that the hippocampal formation extends on to the dorsal aspect of the spliMiium in Man. Other writers, on the other hand, have emphatically denied this and affirmed that the hippocampus is always subcallosal. Other writers, again (such as Zuckerkandl, who has a large following), have stated that the fascia dentata extends on to the upper surface of the corpus callosum. Then, again, Ganser says that in the Mole the fascia dentata ceases upon the ventral aspect of the corpus callosum, but that the layer of pyramidal cells {i. e. the hippocampus proper) becomes continuous with the rudimentary supracallosal ijidusium. We have, then, in the literature of this subject four mutually contradictory beliefs concerning the mode of termination of the upper extremity of the hippocampus, each maintained by its own supporters with much cogency. We have already seen in the different Edentates that the manner in which the hippo- campus terminates can be clearly demonstrated, and we can also see with the naked eye the cephalic extension of the hippocampal vestige, not only to the anterior extremity cf the corpus callosum, but also as far as the olfactory peduncle. No other writer upon this subject seems to have even suspected the presence of this cephalic extension of the hippocampus. Eor it has been possible, upon the evidence of comparative data, to definitely state that the indus'mm is not only continuous with the hippocainpus, but in itself atrophied hippocampus *. " The Pabacommissueal Body." In examining the mesial surface of the hemisphere we find, in front of the psalterium and ventral commissure, a region to which passing reference has been already made as the area piYrcommissitralis. This term was originally introduced by the writer as a pui-ely descriptive title for a region of doubtful identity in front of the cerebral commissures in Ornithorhynclms. This was subsequently shown to he the surface of a large ganglionic mass to which the name corpus prcecommissurale was given t. In the course of further investigations which were carried on after the introduction of the * For the evidence of this point see ' Journal of Anatomy and Phj'aiology,' irp, cit. vol. xxxii. t Cf. " The Origin of the Corpus Callosum," this volume, p. 47. THE BRAIN IN THE EDENTATA. 323 former term, I was able to clearly establish the identity of the area pra;commissuralin with those surface-areas corresponding to sejji/im lucidum and gyms snbcallosns (Zucker- kandl) of the human brain. Taking into consideration the fact that these two areas are merely parts of one and the same formation, the separation of which into two parts is purely arbitrary in most mammals, it seemed desirable to retain the term " area pra-- commissuralis " for the whole surface-region, even though the term may not be so apt elsewhere as it is in the case of Ornithorlnjuchus and the Submammalia. The term " precommissural area " sufficiently explains the locality of the region to which it is meant to be applied in all vertebrate brains. The term " ])recommissu7-at body''' however, is not nearly so appropriate a title for the ganglionic mass whose surface is the pi'ecommissural area, and in many cases it is singularly inappropriate. By substituting for it the name '' paracommissioral body " we shall obtain a term which may justly be applied to tliis important ganglionic mass in any vertebrate. Unlike the term "precommissural area," which was introduced much earlier, the term "pre- commissural body " has not yet been adopted by other writers, and hence there can be no objection to the substitution of the more apt title " corpus paracommissurale " for the ganglionic mass which I had previously called the " corpms prcecommissurale." The need for some definite name for this mass of grey matter which forms so prominent a constituent of the brain in the E,eptilia, Amphibia, and Dipnoi has been felt for a considerable time. The terms " septum lucidum" and "gyrus suhcallosus " are not onlv meaningless but inaccurate when applied to any other brains than the most bighly organized of the Mammalia, and their application to such brains as the Prototherian. Saurian, or Amphibian cannot fail to give rise to great misconception. Nor can these terms be applied with any more propriety to the Basypodidce. In the dissection of the lateral ventricle of Orycteropus (fig. 19) the paracommissural body makes its appearance in tlie interval bet^reen the anterior commissure and the corpus callosum as a very plump mass of grey matter w^hich forms a most marked contrast to the surrounding areas because they are all lined wdth white meduUated tibres. The posterior margin of this body is abrupt and rounded, and appears as though it were placed on the lateral aspect of tiie fimbria. Anteriorly the paracommissural body graduallv sliades away toward the tunnel in the olfactory peduncle. The paracommissural body is separated from the corpus striatum by the slit-like lateral ventricle. It is unnecessary to enter into a description of tbe corpus striatum or lateral ventricle in this place. These features of the brain do not appreciably dilfer from those of common Rodents such as Lepus. The essential features of the paracommissural body are similar in all the Edentates. The shape of the body, however, is modified considerably by the growth of a corpus callosum, and hence it is very variable in the ditferent families. These changes in shape are best appreciated from a study of the mesial surface of the hemispliere, Avhere the paracommissural body enters into the constitution of the w^all of the brain formin"- the precommissural area. In the Sloths (figs. 17, 18, and 23 B) we see the typical arrange.-ncnt of the precom- 324 DE. G. ELLIOT SMITH ON missural area. It forms a broad tract lying in front of the lamina terminalis (l.t.), which contains the anterior commissure : it extends forward to become continuous with the mesial surface of the olfactory peduncle; it is separated above from the pallium by the precommissural vestige of tlie hippocampus {v.h.'), and below it extends into continuity with the tuberculum olfactorium. The dorso-caudal angle of the paracommissural body has been drawn upward and backward as a long pomted process [ji'), which fills up the gap between the psalteriurn {j>s.) and corpus callosum {c.c). Tliis process is somewhat stretched, and is the homologue of the septum lucidum of human anatomy. With the corresponding lamina of the other hemispliere it encloses a narrow median slit — the cavum septi, which opens forward and downward. The only important difference in the Myrmecoijhagidce (fig. 23 A) is that the "septal" part {p.') of the paracommissural body has been stretched to a mucii greater extent. In the Armadillos (fig. 23 C) there is practically no septal part of the paracommissural body. In the Pangolins the condition most nearly approaches that of the Sloths. The Pallium. It is convenient for descriptive pui-poses to associate together all the surface-areas which we have so far discussed — bulbus and pedunctiliis olfactorins, tuberculum otfo.ctorium, lobus lyyriformis, lobus perforatus, hij^pocamjms, and area prcccommissuralis — in contradistinction to the rest of the surface of the hemisphere, \\ hich it is customary to call 2)allium.. The histological structure and the morphological relations of each of the structures which have just been enumerated are quite distinct, and present little, if anything, in common to those of any of the other areas. But while they present so little in common they are clearly distinguished from the rest of the cortex or pallium by the fact that they are phy logon etically older, and at the dawn of the mammalian epocli had reached the height of their histological difTerentiation and morphological importance. The pallium, on the other hand, is phylogenetically quite recent, seeing that it can be first definitely recognized as a very insignificant element in the cerebral hemisphere of rej)tiles. In the Mammalia for the first time it develops its distinctive features, and continues to increase in morphological importance and in histological elaboration. There is, therefore, a very clear line of demarcation between the behaviour of the pallium attd the rest of the sui'face of the brain which deserves to he clearly i*eflected in any system of nomenclature Avhich we may adopt. From the fact that all the surface-areas, apart from the pallium, are more or less intimately associated with the olfactoiy apparatus, it has become customary to apply the term " rhlnencephalon " to this group of surface-areas. Tliis application of tlie term " rhinencephalon " was suggested by Sir William Turner*, although at that time the exact limits of the region — more especially with regard to the precoii missural area — could not be mapped out. The suggestion of Schtifert to include * AV. Turner, op. ril., ' Journal of Anatomy and rbysiology,' voL xxv. 1890. t E. A. Schafer, Quain's 'Anatomy,' 10th cd. vol. iii. part 1, ISO'.i. p. IGU. THE BUMS IN THE EDENTATA. 325- the gyrus fornicatus in the rhinencephalon, in conformity with Broca's idea of a limbic lobe, would, if adoj^ted, destroy the whole value of the term " rhinencephalon " as a contradistinction to the term " pallium." In employing the term rhiuencejihalon we must always lieav in mind that this heterogeneous collection of siu'face-areas is not exclusively a " smell-brain," but that the association of all these parts with the olfactory ai)paratus is much more intimate than it is with any other sensory organs. This will explain the apparently paradoxical statement that smell-less animals, like certain Cetacea, possess a " smell-brain " or rhinencejDhalon, since the pyriform lobe and hippocampus do not entirely vanish in such animals. The pallium is the j)rogressive part of the cerebral hemisphere, and therefore the region to Avhich we must attach most importance in the study of the mammalian brain. Por in the lowest mammal the rhinencephalon has already reached as high a stage in the evolutionary process as it ever will reach, and its essential features change but little in mammals. The slight change that does occiu' is princij)ally a process of retrogression in the more highly-organized members. In the following description I have departed from the usual custom by employing the term " sulcus " lor all the pallial fui-rows instead of the more usual term " fissure," not only because the former is a much more acciu'ate name, but also in order to introduce a distinction between the strictly intrapallial furrows and the pallial boundaries, which we still call " rhinal Jissu/-e," " hqjpocampal fissure," and " callosal fissure." To this usual distinction the only exception will be made in the case of the Sylvian fissure, which in its fully-developed form deserves the name " fissure." The mesial sm-face of the pallium presents much greater ujiiformity in the arrangement of its sulci than the cranial surfaces, and hence w e may preferably begin an examination of the pallium in this reoion. Upon the inner aspect of the hemisphere of Orycteropus the pallium forms a broad strip of cortex which extends from the corpus callosum to the dorsal edge of the hemisphere. This strip of pallium bends downward in front of the corpus callosum, and becomes continuous along an ill-defined line with the precommissural area. Behind the corpus callosum it bends downward and becomes continuous with the pyriform lobe. The whole extent of this strip of pallium is divided into two ajjproximately equal areas, a central or circumcallosal and a peripheral or marginal respectively, by a single deep and well-defined sulcus (fig. 4). This sulcus, which we may at present distinguish as a, begins posteriorly in close proximity to the caudal extremity of the rhiual fissure : that is, at the extreme ventral Hmit of the mesial pallium. As it ascends it follows a coiu'se parallel to the peripheral margin of the hemisphere, and bends forward midway between the corpus callosum and the upper edge of the hemisphere vintil it I'eaches the anterior j)ole ; still following its course parallel to the peripheral margin, it bends downward in front, then after a short course suddenly bends again upward and backward, I. e. jjarallel to the palho-precommissural junction, and ultimately terminates in a triangular depression formed by the opening up of the anterior extremity of the callosal fissure. From the antero- superior bending of this sulcus one or two small but deep branches are usually SECOND SERIES. — ZOOLOGY, VOL. VII. 45 826 DE. G. ELLIOT SMITH ON given off, \^ liicli are probably the expression of a puckering of the cortex at the seat of flexure. The posterior ascending part of the sulcus is A^ery deep, and indents the whole thickness of the cortex, i. e. is a " complete fissure.'" The possible significance of this observation will be discussed later. The arrangement of sulci upon the mesial surface of the hemisphere in 3fi/rmecophaga has been well represented by Forbes*. The extensive sulcus a of Orijcteropus is represented by an analogous arrangement of three independent sulci: — a posterior vertical {cm" of Forbes), Avhich we may distinguish as a' ; a longitudinal {cm of Forbes), which we may call a" ; and an oblique anterior {cm of Forbes), which we may call a'". The longitudinal sulcus (a") extends beyond the corpus callosum at each end, and sejoarates the supracallosal pallium into t^^o longitudinal strips. The po.sterior vertical sulcus («') appears to l)e much more extensive than the cor- responding 2)art of the sulcus a in OriicteropHS. This apparent ditference serves to accentuate the constant relationship which the sulcus a maintains to the fissiira rliinalis. In Ilp'mecophaga the rhinal fissure terminates at the ventral border instead of high up on the posterior border of the hemisphere as it does in Orycteropus, and the sulcus a' descends to the ventral limit of the pallium, i. e. in close proximity to the fissura rhinalis. In Tamandiia there is no definite sulcus representing the a scries of ITi/rmecopha(ja,hxiii merely three deep pit-like depressions, which are probably the rudiments of the cor- responding three sulci in the Great Ant-eater. The sulcus a! is represented by a short deep sulcus, merely 3 mm. long, which rises as high as the level of the splenium of the corpus callosum. The sulcus a"' is represented by a deeply-incised triangular depression a short distance in front of the genu of the corpus callosum. Above the anterior extremity of the corpus callosum there is a short shallow furrow which is partially analogous to the sulcus a" in the Great Ant-eater. In the figure of the mesial surface of the brain of C//cloturus which Pouchet gives t there is no trace of any pallial sulcus. In the Bradi/podidce the mesial surface of the pallium presents an arrangement of sulci which is analogous to the a-system in the Ant-eaters. In addition the mesial extremity of a sulcus, which belongs more especially to the cranial aspect of the pallium, extends on to this surface of the hemisphere in both Cliohtpns and Bradi/pus. This sulcus, which we shall describe later under the designation {6, deeply notches the anterior border of the hemisphere. A sulcus, which we may distinguish as a, begins upon the mesial surface of the hemisphere in Cliolcepus didaciylus, just in front of the termination of the posterior rhinal fissure, and opposite the midpoint of the hippocampal fissiire. Its relation to the rhinal fissure is exactly similar to the sulcus we have similarly designated in Myrmecophaga. As it ascends it j)ursues a course which is approximately j)arallel to the upper half of * Forbes, op. cH., Proc. Zool. Soc. Londou, 1SS2, p. 293, tig. 4. t Pouchet, ojj. cif., torn. vi. pi. iv. fig. 4. THE BRAIN m THE EDENTATA, 327 the hijopocampal fissure, and upon reaching the dorsal part of the hemisphere it does not hend forward but continues its course as far as the dorso-posterior corner of the hemisphere, where it ceases just before reaching the dorsal surface of the hemisphere. In the right hemisphere this sulcus just crosses the upper edge to reach the cranial surface. In his description of the brain of Cholcepus Hoffmanni Turner says * that this sulcus, which he does not name, extends on to the cranial surface above. Flower f does not represent any sulcus corresjjondiog to a! in his specimen of Cholcepus dklacti/lus. In my young specimen of Bradijpus trldactylHS there is a very well-developed sulcus a', which presents an arrangement exactly analogous to that which Turner has described in Chola-pus Hoffmanni. It begins below just in front of the rhinal fissure, and ascends vertically ; in its upward course it is placed upon the caudal rather than ujion the mesial aspect of the hemisphere, as was the case in Cholaqjus, and crosses the postero-superior border of the hemisphere at a distance of about 3 mm. from the interhemisplieral cleft ; it, or more strictly a sulcus y with Avhich it is in uninterrupted communication, extends forward for a distance of 10 mm. on the dorsal or cranial surface of the hemisphere (left). On the right hemisphere of this brain the sulcus 7, w'hich is in continuity with a', is much shorter, while in the brain of Bradi/pKS in the College of Surgeons this sulcus y extends for more than half the length of the hemisphere. In the brain of Cholapus a longitudinal sulcus divides tlie area of j)alliuni which lies above the corpus callosum into two horizontal bands, of which the lower is slightly the narrower. In my specimen of Cholapus dldactylus this sulcus, which we may call a", begins slightly in front of the corpus callosum and terminates close to the ujiper extremity of the sulcus a' by bifvircating so as to form a short vertical sulcus parallel to 05'. In Flower's specimen of the same species the sulcus a'' begins slightly further forward and extends considerably further backward before it ends simply without bifu.rcating. In Turner's specimen of Hoffmann's Sloth this sulcus, which Turner calls by Krueg's designation of '' splenialls" begins far forward and with a bifurcated extremity ; posteriorly it bends downward into a direction parallel to the sulcus a', and, according to Tu.rner, is " apparently continuous with the hippocampal fissure." Sucli apparent junctions between the hippocamjial and other fissures are of absolutely no importance, because the ftssiu-a hipjJOCciDipl is quite sul generis, and of a nature entireh- diff'erent from all other fissures or sulci. In my specimen of Bradypus the sulcus a" begins far forward in a bifurcated extremity in much the same manner as it does in Turner's specimen of Cholcepus Hoffmanni. In the right hemisphere it ends simply at the posterior extremity, but in the left hemisphere there is a slight bifurcation, and in addition a small independent vertical sulcus midway between the adjacent extremities of a and a". In all the specimens representing the brains of Sloths we find a well-defined family type in the mode of disposition of the mesial surface of the jmllium. Among aU the Armadillos we find some arrangement of sulci analogous to the series a, except in Chlamydop)horus, which, like Ci/cloturus, has a smooth pallium. * Turner, op. cit., Journal of Anat. & Physiol, vol. xxv. p. 122. f Flower, 021. cit., Phil. Trans. 1865, fig. 5. 45* 528 DE. G. ELLIOT SMITH ON In Dasypus sexcinctus and Dasyptis villosus there is a sulcus wbich we may refer to as simply a. This consists of a small slightly arched sulcus, about 10 mm. long, which jmprints the pallium about midway between the corpus callosum and the superior border in sucli a manner that the centre of the arc is placed above the prominent splenium of the corpvis callosum. In Xenurus the sulcus a is much more extensive, and exhibits a jieculiar resemblance to the fissure a in Ori/cteropiis, excepting that it lacks the posterior descending i^axi, which we have called a' in Ilyrmecophaga and the Bradypodidse. It is not improbable, however, that this defect is more apparent than real. In speaking of MyrmecojjJiaga we had occasion to point out how much lower the sulcus a descends in this form than the corresponding part of the sulcus a does in Orycteropiis, and we associated this with the fact that the sulcus a presents a constant relationship to the rhinal fissure, which is placed higher up on the cerebral surface in the latter. The rhinal fissure is placed much higher still in the Armadillos, and hence it is not improbable that the apparently incomplete sulcus a in Xenurus may be analogous to the whole of the extensive sulcus a in Orycteropus. In 3£anis, Pouchet * has figured an extensive longitudinal sulcvis analogous to the sulcus a" in the Bradypodidse and Myrmecophaga. It is considerably longer than the •corpus callosum, but quite simple. More recently Max Weber has given an excellent figure of the mesial surface of the brain of Manis f , in which he represents under the title "fssura sjdenialis " a longitudinal sulcus which extends the greater part of the length of the mesial surface, and ends in front in a T-shape. In his figure there is a faint line proceeding from the neighbourhood of the hippocampal fissure around the posterior margin of the hemisphere on to the dorsal surface of this in a manner similar to the sulcus u in Bradypus. There is, however, no mention of any such sulcus in his description. Moreover, Ziehen, who has recently examined the brains of four representatives of this genus, makes no mention of any such sulcus. He simply states J that in Manis he found aJ/ssKva splenialis 15 mm. long, Avhich extended further forward than the genu of the corpus callosum, and ended in a T-shape, 4*5 mm. from the frontal pole. The specimen of Manis to which I had access in the College of Surgeons j)resented on the mesial surface a sulcus exactly resembling that which Weber figures and describes under the name " fssura splenutlis." Before passing on to the consideration of the cranial surface of the pallium we may briefly consider, in the light of the imperfect data at our command, the significance of this a-system of sulci, which in some form or other is present in all the heterogeneous representatives of this order, with the exception of the smooth-brained Chlamydophorus and Cycloturus. If we examine the mesial surface of the brain of the Sheep {Ocis aries) we shall find a I'learly-defined series of sulci presenting an arrangement analogous to the a series in the * Pouehet, op. cit., torn. vi. pi. iv. fig. 10. t Max Weber, Zool. Ergebnisse, ii. tab. ix. lig. 69. t Ziehen, ' Das CcntraluerTCiis^vstem tier Monotremcn und Marsupialicr,' Jena, 1897. p. 156. THE BRAIN IN THE EDENTATA. 329 Edentata. We find a deep sulcus beginning just in front of the termination of the hinal fissure, and following a course upward and then forward, pai'allel to the corpus callosum. Before reaching quite as far as the genu this deep sulcus stops as such, but it may be continued as a shallow furrow obliquely upward on the dorsal surface. To this Krueg has applied the name ^'fssura splenialis,'" from its relation to the splenium corj)oris callosi *. An independent sulcus pursues a course parallel to the anterior extremity or gemt of the corpus callosum, and is hence known as the fssura geii/mlls. In the important memoir of Krueg to which reference has just been made, the presence and essential uniformity in the arrangement of this splenial sulcus in all the Ungulata is clearly demonstrated by means of a beautiful series of illustrations and a lucid description. Two years later the same investigator earned our further gratitude by showing, in the same clear manner, the existence of an analogous sulcus in a large series of other mammals, including the important order of Carnivoraf. From these valuable series of observations we learn that in a very considerable number of widely-separated mammals we find a constant arrangement of sulci cor- responding to those which we have called «' and a" in Myrmecophaxja, and which together form a feature for which Krueg has introduced the term fissura splenialis. We also learn that a genual sulcus surrounds the anterior extremity of the corpus callosum, but its shallowness and inconstancy point to the fact that it is of considerably less importance than the splenial sulcus. The high importance of the splenial sulcus is shown not only by its depth, its constancy, and its well-defined features, but also by the fact that it is the first pallial fissure to make its appearance in the Cat and probably also in the Sheep %. In the foetal Cat it begins as a small arcuate sulcus behind the corpus callosum, which is analogous to the permanent condition of this sulcus in Dasj/pus ^. As the hemisphere grows, the sulcus splenialis extends forward. In the develo^wuent of this sulcus in the Edentata it is highly probable that a similar mode of development would be found in Myrmecophaga and Orycteropjus for the posterior vertical part of the sulcus which we have called a! is deeper than the rest, and in Orycteropus, if not in others, it is " complete." It is this postero- ventral part of the sulcus, m.oreover, which is the most constant. We have already had occasion to observe the constancy of its relation to the rhinal fissure not only in Orycterop'us and Myrmecophaga, but also in the JBradypodldcc and possibly in the Dasypodidcs. In many families of Marsupials we find a sulcus upon the mesial surface of the hemi- sphere, which presents the typical relationship to the fissura rhinalls which characterizes the sulcus splenialis. In many Marsupials (e. g. Phascolarctos) this sulcus,is quite short, and is confined to the situation corresponding to that in which the early splenial sulcus * Julius Krueg, " Ueber die Furchung der Grossliirnrinde der Ungulateu," Zeitsch. f. wiss. Zool. Bd. sxxi. p. 308. t Julius Krueg, " Ueber die Furehung auf der Grossliirurinde der Zonoplaccnlalen Siiugethiere," Zeitseh. f. wiss. Zoologie, Bd. 3xsiii. + Vide Krueg, o/j. cit., Zeitsch. f. wiss. Zoologie, Bd. xxxi. & xxxiii. § Vide Krueg, op. cit., Bd. xxxiii. Taf. xxxiv. 330 DE. G. ELLIOT SMITH ON first makes its apjiearance in the kitten's brain. But in other Marsupials (e. g. Macroptis, Thylacinus) the sulcus extends up^\'ar(;Is, and frequently crosses the ujDper border to reach the dorsal surface of the hemisphere. In the larger Chiroptera we find a very extensive and fully-developed splenial sulcus in a brain which lacks a genual fissure. Turner has figured it in Cynonycteris collaris and Pteropus medius *. I have examined and found practically the same arrangement in Pterojjus poliocejihalus. There is a general tendency, therefore, in most of the mammalian orders to the formation of a definite splenial sulcus upon the mesial surface of the hemisjihere. We find this tendency clearly expressed in the Edentata as well as in the Marsupialia, Ungulata, Cetacea, Carnivora, and Chiroptera. It is therefore the expression of a very general tendency among the Mammalia, and the fact that this fissure develops upon parallel lines in the different families of Edentates must not be considered as an indica tion of their mutual afiinity, but rather of their wider kinship with the general body of mammals. In the Monotremata, or rather in Echidna, which alone of the two genera possesses a convoluted brain, we find no definite splenial sulcus, and this absence seems rather to accentuate the closer interrelationship of the other mammals, in which the sulcus in question is never lacking in a convoluted brain. If we examine the brain in the large series of mammals which possess some representa- tive of this sulcus, we shall find in many and very diverse orders an extensive and regularly-arched sulcus extending forward in a " genual " manner such as we have met in Orycterojms and Xenurus. We have already noted the existence of such a splenial sulcus in Pteropus and other large Chiroptera; in many Cetacea we find a similarly extensive and regularly arcuate splenial sulcus, as for instance Turner's figure of BalcBHoptera f shows ; we find the same arrangement in many Ungulata, as for instance in Eqmis, and among the Carnivora we find an admirable example of the same arrangement in Proteles cristatus +. In the latter brain Krueg distinguishes the anterior part of the large arcuate sulcus as genual, and the rest of the fissure as splenial. The inconstancy and variability of the genual sulcus show^ that it is of very slight morpho- logical importance, and in many cases it may be merely a mechanical product. The splenial sulcus, howevei% is a much more definite and constant feature, and is obviously of great morphological significance. But the most constant part of the splenial sulcus is that portion which lies behind and below the splenium of the corj)vis callosum. In a series of mammalian brains this part of the sulcus is absolutely constant in all hemispheres which have any pallial sulci whatever ; it is also the deepest part of the sulcus ; it is, as we have seen in the Cat, the first part of the sulcus to appear in ontogeny, and from comparative studies we find that many brains (e. g. Phascolaretos) which possess no other true pallial sulci have a small splenial sulcus in this situation. The posterior or ventral part of the splenial sulcus is the constant element, but the anterior part is much more * Turner, op. cit., Journal of Anatomy and Physiology, vol. sxv. t Fig. 33. t W. H. Flower, " Anatomy of rrotdcs," Proc. Zoo), f-'oc. London. lS6i>, fig. 4. THE BRAIN IN THE EDENTATA. 331 Tariable, often insignificant or entirely lacking', and altogether of quite minor morplio- logical imjiortance. In the Sloths we have seen that these two elements in the splenial sulcus, the constant element («') and the varial)le element (a"), ai'e quite distinct the one from the other. If we turn from the consideration of the splenial sulcus in the majority of mammals to its condition in the Primates, we are surprised to find that the sulcus which is usually described as the " splenial " is lacking in its most essential part, namely in that retro- splenial j^ortion which we have just learned to regard as the only constant part of the sulcus. Thus Tm-ner *, in common with many other writers, uses the term " splenial fissure "as synonymous with " calloso-marginal fissure," tacitly implying a homology which is extremely questionable. We are quite prepared to admit that the calloso- marginal fissure of Man and the other Primates may represent the inconstant and mor- phologically unimportant {a") element of the splenial sulcus, but it obviously doas not represent the whole, nor in fact the most important part, of that fissure. Moreover, it seems highly improbable that the most constant sulcus of the pallium throughout the Mammalia, the earliest fissure to make its appearance in development and often the only " complete " sulcvis present, should entirely vanish in the Primates. One of the first, if not actually the earliest, sulci to make its appearance in the development of the human pallium is a short, oblique, and very deep farrow a short distance behind and slightly below the splenium of the corpus callosum. This important sulcus has been called bv Professor Cumiin2:bani f the " anterior calcarine.'" The features of this sulcus in its earliest stage are so analogous to the earliest stage of the " splenial fissure " of Krueg in the developing brain in the Cat that the question of their homology naturally suggests itself. When we remember that the anterior calcarine sulcus is a complete and a very constant sulcus, of obviously great morphological importance, and that it makes its appearance in a situation analogous to that occupied in most mammals other than the Primates by a very constant sulcus Avhich is sometimes complete and of undoubted morphological importance, there is sufficient evidence to justify us in suggesting, as a tentative working hypothesis, the morphological identity of the two structures, i. e. of the anterior calcarine fissure of Primates with the essential or caudal part of the s]jJ en ial fissure of most other mammals. Por it seems to me unlikely that such a deeply-imprinted feature, which, almost alone of all the pallial fissures, tends to make its appearance in all the mammalian orders, except the specialized Monotremes, should suddenly disappear in the Primates without leaving any trace behind. If we glance at such a brain as that of Lemur nigrifrons, which Turner has reproduced % from a memoir of Plower, the resemblance of the calcarine fissure to the a' element of the splenial sulcus is indeed striking, since the sulcus in question presents the characteristic relationship to the rhinal fissure. In Man and the Man-like Apes the similarity is not so striking, not * Turner, o^;. dl., Journal of Anatomy and Ph) siolog3% vol. xxv. p. 143. t D. J. Cunningham, " The Surface Anatomy of the Primate Cerebrum," Cunningham Memoirs of the Royal Irish Academy, No. vii., 1892. % Turner, op. a'(., Journal of Anatomy and Physiology, vol. xxv. fig. 38. 332 DE. G. ELLIOT SMITH ON only because the rhinal fissure has vanished, but also because the " calcarine fissure " appears to branch backward and upward into two other sulci, the posterior calcarine (in Man) and the parieto-occipital respectively. But, as Cunningham remarks *, these latter fissures are of v^astly less morphological importance. In many of the Carnivora we see the beginning of that backward extension of the occipital pole of the hemisphere wbich produces the occipital " lobe " of the Primates, and as a result of tliis growth the "calcarine limb" of the splenial sulcus (or in other words the a element) becomes acutely bent on the " culloso-marg'mal limb " or a" element of the splenial sulcus. It is not difiicult to understand that the continuation of such a process results in the separation of the calcarine from the less important and more unstable calloso-marginal element in the Primates. In this manner we wovJd have produced the extensive calcarine sulcus which Cunningham describes f in the Anthropoid Apes. [Since the above was written I have met Avith the following interesting remarks, which lend considerable support to the thesis I have independently put forward. In discussing the question " whether we are justified in saying that quadrupeds have a far less developed occipital lobe than the Primates," Professor Moriz Benedikt says % : — " The first argument for this view is that the quadrupedal classes of animals have no occipital fissure. This statement appears to me to be contrary to the real condition. The stem of the fork-shaped occipital fissure (the combined calcarine and parieto-occipital) of Man is characterized as an arc with its convexity directed towards the posterior j)ole, and it surrounds that part of the gyrus fornicatus which limits the splenium corporis callosi. When we search after this fissure in a great number of gyrencephalic animals, we find it, but only in connection with the fissura calloso-marginalis. " It may be remarked that in certain anomalous human brains the fissura calloso- marginalis does not come to an end, as in typical cases, by being curved upwards, so as to form the posterior limit of the paracentral gyrus of Betz. It is continued towards the posterior part, forms a limit between the prsecuneal gyrus and the corresponding part of the gyrus fornicatus, and unites with the stem of the fork-shaped fissure in such a manner that this stem becomes the most posterior part of the calloso-marginal fissure." This is most important and valuable corroborative evidence ; for what Benedikt calls "the stem of the fork-shaped fissure" is the same sulcus which Cunningham caUs " anterior calcarine " and which I have already regarded as the caudal extremity of the splenial sulcus. Speaking of Benedikt's earlier expression of this view §, Professor Cunningham says, " Upon his further statement that the calcarine fissure is also developed in brains below the Primates we are not in a position at present to oifer an opinion " || . * Cunningham, op. cit., Cunningham Memoirs, p. 41. t Cunningham Memoirs, oj}. cit. + Moriz Benedikt, " Some Points on the iSurface-Anatomy of the Brain, " Journal of Anatomy and Physiology, sxv. p. 211. § Moriz Benedikt, •' Der Hinterhaupts-Lappen der Saugethiere," Centralbl. f. d. med. Wissensch., 1877, No. 10. II D. J. Cunningham, " Complete Fissures of the Human Cerebrum," Journal of Anatomy and Physiology, vol. xxiv. p. 343. THE BEAIX IN THE EDENTATA. 333 It seems to me that the su;^gestive mass of evidence in favour of the view that the " retro-limbic tissure " of Quadrupeds is tlie homologue of the calcarine sulcus of Primates cannot be lightly gainsaid.] The great bulk of the litei'ature relating to the brain in the Edentata is mainly con- cerned with the cranial aspect of the pallium. But altliough this is the case, the information to be culled from a study of this mass of descriptive matter is of a very imperfect and in many ways an unsatisfactory nature, for our methods and ultimate aims in investigating the anatomy of the brain are now vastly different from those which inspired the authors of the records which we find so disappointing in their insufficiency. It would therefore be hardly justifiable to enter into a serious discussion of many aspects of tlie study of the surface of the brain which take a foremost place in the memoirs to which reference will be made. It will conduce to clearness in the description of the surface of the pallium in this order if we begin Avith the. consideration of Myrmecophaga, in which the fissures very clearly conform to a Avell-recognized type. In the memoir of Gervais * we iind excellent representations of the conformation of the pallium in a specimen of Myrmecophacja, as well as a figure of a cranial cast of another specimen. In the more recent memoir of Forbes f Ave find some very useful semi-diagrammatic figures of the brain in two specimens of Ilynneeojjhaga. My observations were made upon two specimens in the galleries of the Iloval Colleo^e of Surgeons ; but, as one of these brains AA'as still clothed Avith its membranes, practically only one Avas available for examination. A partially-dissected specimen of the brain of Myrmecophaga in the stores of the College enabled me to investigate many pioints in its internal anatomy. Figure 7 represents the left lateral aspect of one of the brains of Jlyrmecophaga in the College of Surgeons, and may serve as a type upon Avhich to base our description. The A^entral boundary of the pallium is formed by the rhinal fissure, Avhich consists of an approximately parallel anterior segment passing by a regular arcuate course directly into the posterior rhinal segment. The latter forms an angle of about 120^ with the former, and hence the posterior part of the pallium descends to the base of the brain as a large dependent process behind the pyriform lobe. Above the point of junction of the anterior and posterior rhiual fissures Ave find a triangular depressed area of pallium, Avhicli we may distinguish as the fossa Sylvii. Its posterior border is formed by a vertical lip which begins below at about the mid-point of the oblique posterior rhinal fissure. As it ascends this lip diverges from the lip formed by the pyriform lobe at the posterior rhiual fissure, and AA-hen it has reached a point about 7 mm. from the dorsal surface of the hemisphere (vicAved in profile) it curves backward for a short distance and terminates suddenly. Thus Ave find the fossa Sylcii limited below and behind by two prominent arcuate lips, the convexities of Avhich face one another. These lips meet below, but above and in front they are widely separated. In this broad interval the depressed area of * P. Gervais, op. cit. pi. i. figs. 3, 3a, and '3h, and pi. ii. fig. 3. t W. A. Forbes, op. cit., Proc. Zool. Soc. London, 1882. SECOND SERIES. — ZOOLOGY, VOL. VII. 46 334 DE. G. ELLIOT SMITH ON pallium gradually slopes up to the general level of the pallium without any definite lip to separate it from the rest of the surface. Toward the anterior extremity of the fossa, however, an ill-defined ridge of cortex or lip makes its ajipearance just above and parallel to the anterior rhinal fissure. In other words, a hoi-izontal depression about 1"5 mm. broad is found immediately above the posterior half of the anterior rliinal fissure ; from the anterior extremity of this depression a sulcus begins and extends obliquely upward and forward toward the antero-superior corner of the hemisphere, where it terminates at a point about 2 mm. from the apex and an equal distance from the mesial border of the pallium. We may distinguish this sulcus by the designation /3. Upon the right hemisphere of the same brain the fossa Si/lvii presents features which closely resemble those upon the left hemisphere, and the sulcus 3 arises in the same manner, but does not extend so near to the mesial plane. But in addition we find in this hemisphere a small oblique sulcus, 9 mm. long, upon the dorsal aspect of the apex, just in front of the termination of the sulcus /3. In the brain which Gervais figures the two hemispheres appear to be perfectly symmetrical so far as the fossa Si/lvil and the sulcus j3 are concerned. The lips which limit the fossa Si/lvil appear to be more prominent, and in addition there seems to be a definite though faint lip bounding ih.G. fossa Si/lvii in front. As a result of the existence of this anterior lip, the upper extremity of the fossa Si/lvii seems to be converted into a definite fissm-e, which arches obliquely upward and backward and then pursues a course of about 10 mm. in the caudal direction parallel to the great interhemispheral cleft (vide Gervais, pi. i. figs. 3 and 3 a). So far as we can judge from Gervais's figures, the sulcus /3 is quite independent of the fossa Sylvii and the fissura rhinalis, but begins at a distance of 2 mm. above the anterior rhinal fissure and extends obliquely forward to cross on to the mesial surface of the hemisphere at a distance of about 4'5 mm. from the apex of the pallium. Unlike the condition in our type-specimen, the sulcus /3 appears to be symmetrical in the two hemispheres. Forbes represents the sulcus /3 in one of his specimens extending into continuity Avith the rhinal fissure ; but he remarks that the area in front of the sulcus /3 (which, in accordance witii the extraordinary suggestion of Broca, he calls the " frontal lobe ") is connected to the fodsa Sylvii (Broca's " lobule sous-sylvien ") by a small, sometimes deep, bridging fold. At a distance of about 6 mm. from the mesial plane there is, upon the dorsal surface of each hemisphere, a longitudinal sulcus which we may distinguish as y. The anterior extreaiity bends obliquely outward and downward, and is 13 mm. distant from the apex of the hemisphere. The posterior extremity approaches quite close to the posterior border of the hemisphere, and bends outward for a short distance imrallel to it. In the specimen figured by Gervais the anterior extremity of this sulcus does not bend laterally, but otherwise it agrees Avith our type-specimen. In all cases the sulcus 7 approaches very close to the sulcus f3, but without actually joining it. The broad elliptical area which lies between the posterior lip of the ybss« ^_y/»w and the posterior margin of the hemisphere is divided into two approximately equal j^arts by a deep vertical sulcus, whicli begins above at a point about 1; mm. (viewed in profile) below the sulcus 7, and ends below at about an equal distance above the ventral margin of the hemisphere. Tiiis sulcus, which we may call I, is approximately symmetrical THE BKAIN IN THE EDENTATA. 335 upon the two liemispheres. From the postero-ventral corner of the hemisphere a shallow furrow ascends upon the left hemisphere for three or four millimetres, and upon the Yis;ht hemisphere a more definite sulcus in the corresponding place. We may call this the sulcus e. Upon the right hemispliere of our type-specimen there is a faint trace of a vertical furrow midway between the sulcus S and the fossa Sylvil. We may refer to this as the sulcus n. Thei'e is no sign of it upon the left hemisphere. In the figure of the lateral aspect of the brain which Gervais gives there are two extensive vertical sulci upon the postero-ventral area of ])allium, and a trace of a third. The exact homologies of these sulci are somewhat doubtful. It is possible that the most caudal represents an exceptionally well-developed sulcus e, and the more anterior the sulcus 5. Upon the left hemisphere of our type, as Avell as in the specimens figured by Gervais and For]:)es, there is a shallow horizontal furrow midway between the anterior part of the sulcus -y and the narrow tapering extremity of the fossa Sylcil. This may be distinguished as S'. If we compare the arrangement of sulci upon the cranial surface of the hemisphere of Myrmecophaga with the principal and more constant fissures which are found upon the pallium in the Carnivora, we must admit that the resemblance is so close that it suggests somethins' more than a mere coincidence. The similarity is so striking? that we do not hesitate to compare one by one the sulci of Myrmecophaga, Avith sulci which are obviously homologous in almost any Carnivore. The resemblance is perhaps more striking in the case of the small Carnivores, because soaallness implies a simplicity in the arrangement of the sulci. In the lai'ger Carnivores the pattern becomes complicated by the introduction of numerous subsidiary and morphologically unimportant sulci. But in such a brain as that of Proteles, of which Flower has provided its with admirable illustrations*, we find hemispheres devoid of all the subsidiary sulci and only the important sulci remaining. In such a type we find the homologues of the sulci of Myrmecophaga preserved, while most of those not represented in the brain of the Ant-eater are also lacking in the small Carnivore. For purposes of comparison, however, I shall refer to the better-known l)rain of the Dog, of which many admirable descriptions are aAailablc. In the Dog we find a fissure arising from the point of junction of the anterior and posterior rhinal fissures, and extending obliquely upward and backward. It is customary among writers to call this the " Sylvian fissure," but Ave must accept this term, which is borrowed from human anatomy, with some reservation. The so-called Sylvian fissure of of the Dog's brain is rather an early expression of those factors Avhich, in the human bi"ain, produce large folds of pallium that overlap an intermediate area of sunken cortex, and by their meeting form a fissure. The so-called Sylvian fissure in the Carnivora ought to be regarded as analogous rather than strictly Jiomologous to the true or human Sylvian fissure. In the Carnivora * W. H. Flower, " Anatomy of Proteles," Proc. Zool. Soc. Londcn, 1869. 46* 336 DE- (i- ELLIOT SMITH ON the downward beading of the posterior part of the pallium, which results in the production of the large post-rhinal tongue-like process, necessarilj' implies a restriction to, or rather a tension on, the growing pallium at the seat of the bending. It is not unlikely that this flexure produces a kink — the fossa Sylvil — at the place of bending, and that the tension of the growing cortex in this region is relieved by the outgrowtli of lips of iiallium at the margins of the kink. In the human brain both of these factors become greatly exaggerated, and a much larger area becomes involved, but the process is probably analogous to that which takes place in the Dog. In the Dog's brain the Sylvian fissure is formed by the meeting of two lips of the pallium, \\hich extend toward their place of meeting by overlapping a depressed area of cortex. If these lips be cut away, we expose a triangular depressed ai'ea of exactly the same shape as the fossa Sylvih in Ilyrmecophaga. The anatomy of this region of the Dog's brain has been most lucidly demonstrated by Dr. Langley *. In the dissection represented in his ligvire 16 we find a triangular depressed area ov fossa Sijlvii (which he calls the " Island of Ueil "), limited on the ventral side by tlie convex junction of the two parts of the rhinal fissure ; on the caudal side bounded by an almost vertical limiting furrow, which is formed by the outgrowth of overhanging lip (which has been cut away in the dissection) ; and bounded in front and dorsally by an oblique furrow which is formed by the overhanging anterior lip. The superior limiting furrow joins the posterior limiting furrow above, and the two are analogous to the sulcus Umitans insidce of the human brain. It is also noteworthy that a narrow horizontal depression extends forward from the fossa Si/Uni above the anterior rhinal fissure, just as we find in Myrmecophaga, but in the Dog this depression is overhung by the anterior extremity of the upper Up or operculum. Thus the representative of the sulcus /3 in the Dog's brain appears in a superficial view to join the anterior rhinal fissure. In Myrmecophaga there is no Sylvian fissure, even in the sense in wliicli this term is applied to the Carnivore's brain. But we have a fossa Sylvil, which is limited posteriorly by an operculum that has not yet overlapped the depression, and anteriorly by the very fjiintest trace of an anterior operculum. Before we can say that a Sylvian fissure exists, these two lips must increase considerably in size, overlap the depressed area, and ultimately meet to form a fissure. The exposure of the fossa Syloli in the Edentata is indicative of the small extent of pallium as compared with the Carnivore. The sulcus [5 is represented in the Dog by a sulcus which is called siqjraorbilal by FloAver f and others, or ^j/r - Sylvian by Owen and others. The sulcus y is represented in the brain of the Carnivora by a longitudinal sulcus knoflm as the coronal or sagittal sulcus. In the Carnivora, however, the cephalic extremity of the coronal sulcus is usually bent in the lateral direction to accommodate the crucial sulcus, which is lacking in the Edentata. The caudal extremity of this sulcus is * J. N. Laugley, " The Structure of the Dog's Brain," .loumal of Physiology, vol. iv. .See especially, in this connection, pi. vii. tig. 16. t W. H. Flower, " Anatomy of Protdes,'' Proc. Zool. Soc. Loudon, 1809, p. 479. THE BEAIN i:*^ THE EDENTATA. 337 often bent downward, and in Mtjnnecophaga it is quite possible tbat this occasional feature of the Carnivorous brain may occur, for in our type-specimen we find a faint depression y" in the situation where this more extensive sulcus occurs at times in the Dog. The only sulcus in the post-Sylvian region of the Dog's l)rain which proseats any constancy is one that Owen calls post-Si/lvian, while Krueg distinguishes it as posterior supru-Syhiun. It corresponds to the sulcus S in Myrmecophaga. But in the Dog it usually forms part of a large arc, the anterior extremity of which may possibly be represented in Myrmecophaga by the small but constant depression S'. In the Dog the growth of the Sylvian opercula covers up the Sylvian fossa and forms the fissure of Sylvius, and after that has taken place the tension of further growth is relieved by the formation of the large supra- Sylvian sulcus. In the Dog a sulcus, which is not nearly so constant as the posterior supra-Sylvian, is sometimes found between the latter and the Sylvian fissure. Tlie occasional furrow i; in Myrmecophaga may represent this sulcus, which in the Dog is called the "posterior ecto- Sylvian fissure " by Langley {op. cit.), who has modified Wilder's nomenclature. The furrow e in Myrmecophaga finds its analogue in a sulcus of the Dog's cerebrum, which "Wilder and others following him have called ecto-lateral. From this brief review we liave seen that every sulcus or depression upon the pallium of Myrmecophaga finds an analogue to correspond witli it upon the brain of the Dog. Moreover, we find that if a large number of Dogs' brains be compared, or if we compare the brains of a large number of different Carnivores, the constant and deepest sulci are those which we find represented in tlie brain of the Great Ant-eater. There is, however, one important and very significant exception to this generalization. In the Dog, as in all Carnivores, we find a short deep sulcus extending transversely outward from the interhemisphcral cleft near the cephalic extremity of the pallium. This sulcus, commonly known as the crucial, is a very characteristic feature of the brain in Carnivora, and is quite lacking in the brain of Myrmecophaga. This absence is very significant when we recall the fact that physiological experiment has sliown that the pallium which immediately surroimds the sulcus in the Dog is the only " excitable " or " motor " area of the cortex. It may be that the brain of the Dog shows a marked superiority over the brain of the Ant-eater in that the central area, which presides over skUlcd movements, shows a sudden increase in extent, resulting in the formation of a new sulcus, the crucial. On the other hand, the formation of the crucial fissure may be to some extent the expression of a general, rather than a local, increase in the extent of the pallium. For we find in the brain of the small Carnivore Geiietta tigrina, which Mivart has described, an absence of a true crucial sulcus, Avhereas all the sulci found in the Edentata are well developed *. In this small and very active Carnivore there are only some shallow pits to represent the crucial sulcus, and, as we have no reason to believe that skilled movements are less developed in this family than in the other Carnivores, it is possible * St. George Mivart, " Notes oii some Points iu the Anatomy of the .Elui-oidea,'' Proc. Zool. Soe. London 1S82, p. 510, fig. 11. 338 DE- G- ELLIOT SMITH ON that tlie general growth of tlie pallium may liave some influence in the production of the crucial sulcus. Apart from this apparent difference in the extent of the excitable areas of the cortex, there are other differences between the pallia in the contrasted brains. If we compare the size of the brain in Ilyrmecophaga and Caiiis, we shall find a striking contrast between the two organs. In his valuable monograph upon the weight of the brain in mammals, Max Weber* gives the weight of the brain in grammes as comjiared to the body-weight in four specimens of the Great Ant-eater as follows :— Percentage of Brain- Body- Braiii-woiglit weight, ■weight. to Botly-weight. Myrmecophmja juhuta, % 75 : 20,800 0-36% $ 84-5:25,500 0-33% ? 87 : 23,000 0-37 7o ? 84 : 28,086 0-29% }) From the same lists 1 select for comparison the brain-weights of four Dogs whose body-weights approximate to those of the Ant-eaters, and exj)ress them in a similar manner thus : — Percentage of Brain- Body- JJrain-weighfc weight, weight. to Body-weight. Canisfam.: Bernhard, 'i ] 16 : 28,000 0-41% Canis fum. -. Sat/ax venaticiis -f- Extrarius aquaticus ieme novtE, S I'^'T' : 27,500 0-39% Canis faw.: ditto, S 9« = 25,000 0-39% Can is fum. : Molonossus 95 : 2 1 ,000 0-45 % In all of these Dogs there is a most decided superiority in the brain-weight over that of the Ant-eaters. The brain-weight in Dogs is exceedingly variable, as Weber's tables amply demonstrate, but in all Dogs approaching the size of a Mijrmecophaga there is a very decided superiority in the weight of the Dog's brain over that of the Edentata. Among otlier sj)ecies and genera this superiority is even more striking. Thus in Cams juhatus, according to Weber, we find a brain Aveighing IGO grms. in an animal of only 23,G00 grms., i. e. a brain 0'7 7o of the body-weight, and in Felis panlas a brain of 130 grms. in an animal of 23,820 grms., or O'Si 7o of the body-weight. On the other hand, in a Hycena striata weighing 28,750 grms., or more than the heaviest of the four Ant-eaters, Weber found a brain weighing nearly 81 grms., or 0-28 7^ of the body-weight. This, however, must have been a very exceptional case, for in another specimen of the same species and of the same sex he found a brain of 89 grms. (i. e. heavier than the brain of any of the Ant-eaters), while the animal weighed merely 17,500 grms., thus giving a brain-weight which is 0-508 7^ of the body-weight. * Max Weher, " Yorstudien iiber das Hirngewicht der iSaugethicre," Separat-Abdruck aus Festschrift fiir Carl Gegenbaur, Leipzig, 1S96. THE BRAIN^ IN THE EDENTATA. 339 In the Dog the brain is, ronghly speaking, more than 10 \ of the body-weight heavier tlian in the Ant-eater, taking into consideration only animals of approximately the same size. "We find upon examination that tlie difference in size of the brain in Mijrmecopliaga and Cants (in animals of approximately the same size) is almost wholly a question of the extent of the cerebral cortex. Now in IJijrmecopluuja the rhinencephalon is both relatively and absolutely very much more extensive than it is in Canis, and hence the disproportion between the size of the pallium in the two brains is even greater than a comparison of the brain-weights might lead us to imagine. It becomes an interesting question to determine what effect this increase in extent has upon the configuration of the paUium. I have already explained the effect of local hypertrophy of the excitable cortex in the production of the crucial sulcus and the lateral displacement of tlie anterior ead of tlie coronal sulcus (7). I have also shown how two definite opercula form around the fossa Sylvii and ultimately expand so as to cover up that depressed area. In addition to this the homologues of the sulci which the pallium of Myrmecophaga presents become deeper and more extensive in the Dog, and large numbers of new^ sulci make their appearance to enable the more extensive pallium to accommodate itself in a minimum space. Let us notice also some general results of the larger pallium. In the Ant-eater the posterior border of the hemisphere is vertical, and the hemisphere itself entirely in front of the cerebellum. In the Dog the general expansion of the cortex is relieved by the caudal extension of the postero-superior part of the hemisphere over the cereliellum, so that the posterior border becomes oblique. This l)ackward or occipital extension has obvious effects upon the mesial surface, which need not be specifically detailed. This occipital extension of the pallivtm in the Carnivores is analogous to the much more pronounced growth in the caudal direction which we find in the Primates. I have discussed at some length the points of resemblance and of contrast in these two brains in order to explain more clearly the salient characters of the brain in the Ant-eater. "\Ve have thus been able to estimate those features wiiich constitute the superiority of the Dog's brain over that of the Ant-eater, and have been impressed with the important conclusion that the two brains are built upon essentially the same plan, or, in other words, that the same factors wliich are at work in mouldin"" the pallium in the Carnivora are fashioning the homologous part of the brain of Myrmecophaga in the same likeness. The significance of these facts will be considered later. At present I merely note the resemblance. The records of any observations which have hitherto been made upon the brain of Tamandua are of the scantiest nature. Gervais gives a figure representing the dor.«al aspect of a cranial cast *, and Pouchet gives figures of the dorsal and lateral aspect of the cerebral hemispheres t, but these figures are very unsatisfactory, because they lack all detail. Even a landmark so well-defined as the rhinal fissure is repre- sented only in a small ])art of its course. * P. Gervais, op. cit. pi. ii. fig. 2. f G. Pouchet, op. cit. pi. v. figs. 1 & 2. 340 DE. G. ELLIOT SMITH ON Forbes has given tis an exceedingly crude scheme of the dorsal aspect of one hemisphere of Tamandita *. My own observations were made upon these brains, one of which had been left to the lloyal College of Science by Professor Huxley, aud the other two are at present in the galleries of the Eoyal College of Surgeons. To illustrate ray description I have constructed a semischematic drawing (fig. 8) from one of the specimens in the Collesje of Surgeons. If we imagine an extension of the process of simplification which would be necessary to reduce the brain of Canis to the condition of Iliirmecopliaga, so that in a smaller braiu all the less pronounced features in the brain of the latter became eliminated, and the more deeply-impressed characters became more faintly marked, we should gain some idea of the appearance of the brain of Tamandna . We have seen that in Ca)iis the faintly-marked lips which serve to denote the boundaries of the fossa Sylvii of Jlyrmeco- 2)liaga have become definite opercula which cover up the fossa Sylvii. In Tamandua the lips entirely disappear, so that we find no trace of a fossa Sylvii, for the pallium above the regular arc formed by the rhinal fissure is smooth. The sulcus /3, which we have seen to correspond to the supraorbital sulcus of the Dog, is represented Ijy Fiff. 25. bulb, olfact. lob. ant. fis9. prima lob. flooculi medulla oblong area A i'" area B area 0 Dorsal surface of braiu of Tamandua ietradacti/la. Nat. size. a short oblique sulcus which does not apj^roach nearer than 4 mm. to the rhinal fisssure in our type-specimen. This sulcus is constant in the six hemispheres which I have examined, as well as in the figures of Pouchet and Forbes. It is even visible as a depression upon the cranial cast which Gervais represents. This is the most constant sulcus upon the cranial surface of the pallium in Tamandua. In this connection it is not withovit interest to note that, according to Krueg f, it is one of the earliest fissures to make its appearance in the development of the pallium in the Cat, as well as in the Sheep and Ox. * W. A. Forbes, q/). cit. fig. 5. t Julius Krueg, Zeitsch. f. 'wissensch. Zool. Bd. xxxiii. THE BEAIN I^" THE EDENTATA. 311 A sagittal sulcus (7) pursiics a course parallel to and at a distance of about 5 mm. from the interhemispheral cleft for about the middle two fourths of the margin of the hemisphere in our type-specimen. Just behind its caudal extremity there is another short sulcus which probably belongs to the same series. In a second brain this sulcus is also present in both hemispheres, but is interrupted and in parts very shallow on the riglit side. In the third specimen the sagittal sulcus is represented in each hemisphere merely by short sulci at the cephalic and caudal extremities of the position which the sulcus Avould occupy. These short sulci are united by a very shallow furrow. In Pouchct's and Forbes's illustratious continuous sagittal sulci are represented; while in the cranial cast figured by Gervais deep coronal grooves are indicated, and each apparently lodges a large blood-vessel. Just below the caudal extremity of the sagittal sulcus (y) there is in our type-specimen a shallow, almost vertical sulcus, which probably corresponds to the sulcus 0 in Ilurmeco- fliacju. In one of the other two braios this sulcus is even more faintly marked still, but in the third we find an exceedingly interesting condition. In this brain tliere are indications of an arcuate sulcus resembling the suprasylvian sulcus in the Carnivora_ A sulcus begins a short distance aljove the point of junction of the anterior and posterior rhinal fissures, and after arching upward and backward for a short distance it passes into a shallow furrow which ciu'ves backward and vertically downward in a situation exactly analogous to the sulcus S in MyrmecoplicKja. It seems as though this represented a more complete arcuate or siiprasi/Jvum sulcus than is present in Ilijrmecophaya. It may be that the depression S' in the latter represents the cephalic extremity of the arc. Upon the left hemisjihere of this l)rain of Tamanditu we find the uj^per part of the arcuate sulcus without the descending part. The cranial cast of Tanmndua which Gervais has represented * is considerably larger than any brain which I examined or of which there is any record. From this cast we might gather that the sagittal (y) and supraorbital (/3) sulci were deeper and more definitely maj)ped orit than they are in the brains we have examined. In the absence of any fossa Sylvii it would have been difficult to detect in the brain of Tamandua any resemblance to the Carnivorous type of bi-ain, but 3Iij rmecopha ga provides for us the needful connecting-link. For, while clearly exhibiting the family type which binds it to Tamandua, it shows us how the Sylvian fissure becomes obliterated, so that the transition from the Carnivorous to the simple Tamandua type is easy to follow. A notable demonstration of this resemblance is given by the developing l)raln of Felis, which Krueg has figured f. At a ]3eriod just before the Sylvian fossa makes its apjiearance, the configuration of the pallium of the foetal kitten presents a surprising resemblance to the adult Tamandua, with its supraorbital (/3), sagittal (7), and suprasylvian (g) sulci. At the same time Tamandua demonstrates the superiority of the Carnivore's brain For in the latter order a brain of the dimensions of that of Tamandua would certainly * P. Gervais, op. ell. pi. ii. fig. 2. t Julius Krueg, op. (it., Zeitsch. f, wissensch. Zool. lid. xxxiii. Taf. sxxiv. SECOND SERIES. — ZOOLOGY, VOL. VII. 47 31.2 DR. G. ELLIOT SMITH ON be more richly fuiTowed even than that of llyrmecophaga. We know that the pallium increases in extent in any order with the added hulk of the animal. A Carnivore as small as Cyclottirus would have a pallintu with as complex a pattern as the Great Ant-eater. I have not had an opportunity of examining- the brain of the small arboreal Ant-eater Cycloturus, Init, judging- from the figures of Tiedemann, Pouchet, and Gervais, its pallium must be quite smooth and devoid of sulci. The transition from Tamandua with its shallow and faintly-marked sulci to the smooth condition is quite simple, and amply justified by the difference in size between these two Ant-eaters. One cannot fail to be struck by the resemblance in the configuration of the hemisphere of Tamaiidna, and especially Cijdoturus, to the Rodent type. Moreover, the exact shape of the corpus callosum and the behaviour of the hippocampal formation, which we liave found so variable elsewhere, are closely reproduced in the Uodentia. When we compare the brain in the Rodeutia with that of the Myrmecophagidse, we can confidently afiirm that any superiority of type that may exist certainly does not favour the Rodent. Unfortunately no collection of brain-weights is available for comparing the two groups on the basis of relative size of brain. The solitary brain-weight of Tamandua which Max Weber gives* is much highei- than tliat of any Rodent of approximately the same body-weight. But a large mass of data is necessary before we can confidently compare the two groups. It is sufficient at present to note that there is an interesting series of transition stages in pallial modification in the Carnivora, Myrmecophagida;, and Rodentia. But while the evidence at our disposal conclusively shows the distinct superiority in cerebral organization of the Carnivore over the Ant-eater, it does not permit us to say that the latter is in any way infei-ior to the Rodent. All the information we possess clearly and decisively points to the undoubted kinship of the Ant-eaters with the Sloths and Armadillos. But for several reasons I propose to consider next the pallium of the Aard-vark. From the writings of Pouchet and Gervais one might conclude that there is a close resemblance between the brains of Mynnecojihaga and Orycteropas. I shall demonstrate that such a belief is utterly illusory and erroneous, and. shall discuss Orycteropzis in this place in order to show the strong contrast between the two forms, and in tliis way to accentuate the im])ortance of the agreement we have found between Mynnecophaga and the Carnivora. In our brief review of the literature relating to the brain in different families of Edentata, Ave found that practically nothing was known of the brain in the Orycteropidaj beyond its general shape. Leuret published some observations t upon the brain of Oi-ycteropm sixty years ago, but, acting under the mistaken idea that mammals might be classified according to the relative richness of their cerebral convolutions, he included in one group three such dissimilar brains as those of Orycteropus, Macropus, and Pteropus, without giving us any information of lasting value. * Max Weber, op. cit., Gegeubaur's ' Fe.st.schrit't,' lS9t5. t Leuret, ' Anatomie com2)aree du Systume nerveux,' Paris, 1838. THE BKAiX IX THE EDENTATA. 313 Writing thirty years after Leuret's original memoir, Pouehet says * tliat the Ant-eaters form with Oryeteropus a separate group, in which we find an agreement in the arrange- ment of the convohitions and in the general form of the brain. The same writer says (p. 1(3), in reference to a late foetus of Oryeteropus : — " Quant aux circonvolntions, leur analogic est complete avec celles du Tamandua." He proceeds to state that in tlie adult the resemblance to Ilyrmecopliaya becomes more marked. In giving details to support the latter statement, Pouehet gives a free rein to his imagination. Probably the culminating instance of this is his statement that the fissure of Sylvius is well marked (p. 17). There is not the faintest trace even of a, fossa Sylcil. Gervais also insists in equally decided terms upon the close resemblance between the hrain, and especially the cerebral cortex, in Myrmecophaga and Oryeteropus t. He concludes his remarks upon this suhject with the following words :—" Je lui [I'Orycterope] trouveune analogic incontestable aveccelui du grand Fourmilier, c'est a dire duTamanoir. L'un et I'autre ont daus leur forme quelque chose du cerveau des Carnivores, mais avec moins de circonvolutions, et le type en ]-este distinct a certains egards." And For])es adds his testimony to this fancied resemblance, wdiich Pouehet and Gervais had pre- viously described, in the following tei-ms : — ''Oryeteropus in its cerebral characters seems to approach Myrmecophaya more nearly than any other form, the sulci and gyri of the brains of the tw^o forms, as well as tlieir conformation, being very similar " %. In two animals which follow similar modes of life and whose dimensions are not widely different, it would be strange if there were not some points of resemblance, but there is not so much justification for pointing resemblances in the pallium of Oryeteropus and of Myrmecophaya as there would be in a comparisan of Ocls with Canis. It is perhaps only just to the authors of the remarkable statements quoted al)ove to mentioned that there is no evidence to show that any of them ever saw the brain of an adult Oryeteropus. In spite of a somewhat illusory resemblance in general shape, which iiaturally obtrudes itself in the examination of cranial casts, the configuration of the actual brain of Myrmecophaga presents a decided contrast to that of Oryeteropus. To begin with, the shape of the pallium is very different in the two forms. For in Oryeteropus the ventral boundary of the pallium, Avhich of course is formed by the rhinal fissure, is horizontal ; and hence there is no caudal downgrowth of pallium such as we find in the brain of the Ant-eater. There is no trace of a fossa Sylvii in Oryeteropus, for the pallial area immedi- ately above the mid-region of the rhinal fissure is quite smooth. In my specimen of Oryeteropus a sulcus, the arrangement of which is analogous to that of the supraorbital or presylc'uni sulcus of the Dog and to the sulcus 3 in Myrme- cophaga, springs from the mid-point of the anterior rhinal fissure, and extends obliquely forward and upward to wnthin a distance of about 5 mm. from the interhemispheral cleft. In the two brains of Oryeteropus in the College of Surgeons, this sulcus springs from the cephalic extension of the posterior rliiual fissure, which is independent of the anterior rhinal fissure. * G. Pouc-befr, 0^). clt. torn. vi. p. 15. t P. Gervais, op. cti. pp. 47 & 48. X W. A. Forbes, op. cit. p. 295. 47* 341' DR. »• ELLIOT SMITH ON From about the micl-poiut of the sulcus |S another sulcus takes origin, and after arching mesially and caudally, extends directly backward as far as the posterior extremity of the hemisphere. This sulcus is exactly analogous to the sac/lttal sulcus (y). In parts this sulcus is very shallow and sometimes interrupted. In my specimen the sulcus is more complete upon the left than upon the right hemisphere. The relations of the sulci /3 and y the one to the other are constant in the three brains to which I had access. This arrangement differs but slightly from that obtaining in Myrmecophaga, and which is generally prevalent among the quadrupedal Mammalia having convoluted brains. The appearance of the combined sagittal and supra- orbital sulci and the relation of the latter to the rhinal fissure present even greater resemldance to the arrangement prevalent in the Ungulate brain than that of the Carnivore. Midway between the sagittal sulcus and the rhinal fissure we find a series of frag- mentary sulci whose extent is variable in different brains. The horizontal sulcus of Avhich they constitute the rudiment is the homologue of that which Krueg calls the fissura su-prasylvia in the Ungulata *. This sulcus is the homologue of the suprasylvian sulcus of the Carnivora, and of the rudimentary representative (S) of this in My rmecop h aga . In many Ungulata the "fissura suprasylvia " is horizontal, just as it is in Orycteropus, but in others the more definite formation of a Sylvian fissure becomes associated with an arcuate arrangement of the suprasyMan sulcus. In the Carnivora and Myrmecophagidce the representatives of the suprasylvian sulcus always present a deep ventral concavity. At the same time we must not lose sight of the fact that in many Ungulata the supra- sylvian sulcus is decidedly arcuate. But we associate this arcuate condition of the sulcus in question with a relatively much higher degree of pallial development than is necessary to give rise to an arcuate condition in the Carnivora. In other words, an arcuate suprasylvian sulcus is a characteristic feature of the Carnivore's brain, whereas in that of the Ungulate it is an indication of a high state of pallial development. The fact that Ovycteropus jiossesses supraorbital (jS) and sagittal (y) sulci, which are analogous to those I have already described in Myrmecophaga^ is not to be considered as an indication of a close kinship, but merely signifies that both of these peculiar animals belong to the great group of mammals which also includes all the Ungulate and Unguiculate animals, as well as, in all probability, the Rodentia and Chiroptei-a. The fact that Orycteropus possesses a horizontal rhinal fissure, and consequently exhil)its no tendency towards the formation of a pallial downgi'owth at the posterior part of the hemisphere, distinguishes this peculiar mammal from the Ant-eaters and Unguicu- lata, and suggests a rajjprochement with the Ungulata. But, it may be argued, the absence of a fossa Sylvii in Orycteropus clearly distinguishes it from the Ungulata, in which a Sylvian fissure (in the same sense in Avhich this term is used in speaking of the Carnivora) is well developed. But there are distinctions between the Sylvian fissure of the Carnivore and the Ungulate. In the Ungulata it is * Julius Krueg, op. cit., Zeitsch. f. wissensch. Zool. Bd. xxxi. THE BEAIN IN THE EDEXTATA. 345 ixmisnal to find the fossa Sylvii, or tlie fissui e wliicli is formed l)y the meeting of its lips, so well defined in the neiohbourhood of the rliinal tissui-e as is customary in the Cariii- vora. Sometimes, it is true, the posterior lip is as well defined as it is in Caruivora, as, for instance, in the brain of the Pig (S/is) ; but very often the posterior lip, which we have already noted as the most important boundary of the fossa Sylvii, is not so well formed, and in mnny Ungulates tlie Sylvian fossa becomes reduced to insignificant proportions. In some Ungulates it becomes so far reduced as to be a mere horizontal depression ab:ve the caudal extremity of the anterior rhinal fissure, with no attempt at the formation of a fissure. The most extreme instance of this reduction with which I am acquainted is the Musk-Deer (Moschus moscJiiferns), excellent figures of which have been provided for us l)y Tlower *. In this animal we find a brain of peculiar simplicity, which presents a remarkalde resemblance to that of Orycteropus. If we compare Plovver's fig. 12 with the repre- sentations of the lateral aspect of the hemisphere of Orycteropus (figs. 2 and especially 5), this similarity will be at once apparent. In the two brains of Orycteropus in the College of Surgeons we have already observed that the anterior is distinct from the posterior rhinal fissure, and that the latter extends forward for some distance above and parallel to the former. This arrangement becomes more intelligible when we compare it with the condition in Moschus. In this interesting brain the posterior rhinal fissure also appears to extend forvvavd above and parallel to the anterior rliinal for a considerable distance. But the anterior rhinal fissure joins the posterior rhinal, before the apparent anterior extension of the latter overlaps the former ; and in addition the area between the two overlapping fissures is depressed, so tliat by com- parison with Myrmecopliaya or any of the common Carnivores or Ungulates we find in this depression the representative of the fossa Sylvii. In Orycteropus it is probable that the apparent cephalic extension of the posterior rhinal fissure is intrapallial and represents the upper boundary of the fossa Sylvii, the relation of which to tiic sulcus j3 is identical with that which obtains in Ilyrmecophaga (tig. 7)- The configuration of the pallium in Orycteropus is analogous to that which is found in the simplest form of Ungulate brain. The pallium in Orycteropus is relatively much smaller than it is in Ilyrinecophaga, and this fact probably explains the great difference in the size of the corpus caliosum in the two brains. It is also much smaller than the pallium in any Ungulate with which I am acquainted. Tiedemann, Pouchet, and Turner, in the memoirs to which I have frequently referred, have given figures and brief descrii^tions of the brain of the Two-toed Slotb, and in Gervais's excellent figure of the cranial cast of Cholcepus the shape and size of a brain of this genus are accurately represented. Leuret andPouchet have figured the brain of the Three-toed Sloth, and Gervais has represented a cranial cast of this genus. "While these contributions to our knowledge relieve me from the necessity of describing * W. H. Flower, ' Oii the Structure and Affinities of tlie 3Iisk-Deer,' Pi-oc. ZojI. Sao. London, March, 1875. 31.6 DR. G. ELLIOT SMITH ON tlie shape and size of the brain in this family, they unfortunately convey a very imperfect idea of the exact couJignration of the pallium of Chola'ims awdi Bradypas. This is all the more unfortunate because the state of preservation of my specimens does not permit me to describe the pallium with that amount of detail which is desirable. My observa- tions liave been made mainly upon two representatives of each of tlie genera Bradyjms and Cholcepus. In all the representatives of both genera to Avhich I "have had access, as well as nil those to which reference is made in the memoirs quoted by me, there is a very well- defined supraorl)ital sulcus (/3) presenting peculiar yet constant features. In Bradjipus iridaciylus the supraorbital sulcus (/3) appears to spring from the anterior rhinal fissure near its middle, and proceed almost horizontally forward. At its apparent origin from the rhinal fissure this sulcus is very shallou-, and may even fail (fig. 11), but it rapidly deepens as it extends forward. It proceeds transversely inward across the anterior pole of the hemisphere (fig. t^O), and deeply notches the mesial surface of the hemisphere at its anterior border ( fig. 17 /3). Fiff. 20. Dor.sal surface of brain of Byai/i/pus triclacli/liif:. licduced to fy nat. size. In Cholcepus didactyhis, and apjoareutly also in Cholcppus Hoffmanni (judging from Turner's figures), this sulcus presents a disposition exactly analogous to that I have described in Bradypus (tigs. 12 & IS /3). In Bradydiis the sagittal sulcus (-y) is variable. In the specimen I have represented (fig. 20) there is an extensive sagittal sulcus which is almost two thirds of the length of the hemisphere. It crosses the posterior border to become continuous with the sulcus a on the mesial surface. On the dorsal aspect of the anterior pole of the hemisphere there is another sulcus which probably belongs to the same (7) series. It begins just behind the sulcus /3, and extends backward for a short distance laterally to the main part of the sulcus y. The representatives of the y series of sulci present a slightly different arrangement on the smaller brain of Bradypus. On the left hemi- sphere the fissure 7 is divided into two parts, just as we have noted in the first brain. Of these the posterior, which is in continuity with the sulcus a', extends forward for only about half the length of the hemisphere. The anterior fragment of the sulcus THE BRAIN IN THE EDENTATA. 317 begins on the mesial side of the posterior fragment and extends forward towards the sidcus /3, bitt just liefore its termination it gives off a peculiar hook-like lateral branch. On the riglit hemisphere the sulcus y is split up into three parts, of which the anterior and posterior are very short and laterally placed. The intermediate segment, which occupies about tlie middle two fourths of the hemisphere, overlaps both of the other fragments on their mesial side. In the admiral)le representation of the dorsal aspect of a cranial cast of Brndypm which Gervais gives, the sagittal sulcus is represented by a single and very definite furrow which extends almost the whole length of the hemisphere. After comparing the condition of tlie 7 series of furrows in Bradypus with the analogous sulci in Cholcppus, it seems possible that the fragmentary sulci which we have found in the former may represent the two sagittal sulci which we find in the Two-toed Sloth. In one of my specimens of Chola'pns didaclylm a sulcus begins near the lateral margin of the posterior surface, and after pursuing a mesial course along the superior border of the hemisphere it bends forward, at a distance of 7 mm. from the mesial plane, and extends the whole length of the hemisphere till it meets the sulcus /3. At a distance of 9 mm. from its cephalic limit this sulcus takes a sharp bend toward the mesial plane and continues its course at a distance of about 4- mm. from the great inter- hemispheral cleft. Erom the caudal extremity of the more raesially placed part of the sulcus a much shallower sagittal sulcus arises and proceeds backward almost to the posterior border of the hemisphere, dividing the area between the sagittal sulcus (y) and the interhemispheral cltft into two equal areas. This condition is practically symmetrical in the two liemispheres. In the second brain there were also two sagittal sulci, but tliey were apparently not so complete (fig. 12) as those just described. In a figure which Pouchet gives of the dorsal aspect of the brain of a young Two-toed Sloth two very complete and independent sagittal sulci are represented, of which the lateral unquestionably represents the sulcus siKjiltalis of other forms. Again, in the cranial cast of Cholcepiis which Gervais figures there are two very distinct sulci. In the brain of Choloepits Hnffmanni which Turner has figured, we find fragmentary, sulci representing these two sulci. The proper sagittal sulcus crosses the posterior margin of the hemisphere and bends to the lateral side of the mesial or subsidiary sagittal sulcus, where it suddenly ceases. Further forward we find traces of the anterior portion of this sulcus. The question as to the na' ure of the mesial sagittal sulcus naturally suggests itself. It seemed at first not unlikely that this sulcus might represent the anterior part («") of the splenial sulcus, taking into consideration the course of the a' element of the sjilenial sulcus and the fact that it is not uncommon for this sulcus to extend on to the dorsal aspect of the pallium in many Ungulates and other mammals. But the existence of a very deep sulcus on the mesial surface in the usual position of a", and the extreme shallowness of the mesial sagittal sulcus, lead me to discard this suggestion and to regard the additional furrow as a mere mechanical product of the peculiar pallial growth of Choloepus. 348 DR- G. ELLIOT SMITH ON When we consider the close resemblance in shape between the brains of the Sloths and the gigantic extinct Ground-Sloths, and in view of the fact that increase in bodily dimensions in any group of mammals usually implies a more extensive pallium, we mio'ht expect to fiiid in the brains of Megatherium, Mylodon, and Scelidothermm some signs of a richer pattern of sulci and perhaps a definite medial sagittal sulcus. So far as can be judged from the figures which Gervais gives of the cranial casts of these extinct monsters, the brain of Mylodon appears to have been richly convoluted after the analogy of the brain of the rhinoceros. While we can readily recognize the sagittal sulcus, we cannot be sure of the presence of a definite mesial sagittal sulcus after the manner of Cholaepus. In Scelidotherinm, whose brain is just like a very much enlarged brain of Bradypus, there is a typical sagittal sulcus, but we can recognize no mesial sulcus. The same remarks might ajiply to the brain of Megatlierium ; but the observer cannot fail to be struck with the apparent simplicity of these two brains. In the brains of such large animals, when we consider the well-developed sulcal pattern in the minute Bradyjnis, we might have expected an extremely rich and complex pattern of sulci. There appears to be some indication of this in the cranial cast of Mylodon. On the lateral aspect of the sagittal sulcus there is, in the brain of BiHidypus, a strongly- arched sulcus whose features are very constant in all the brains which I have examined, and of "nbich there are records as well in the cranial casts. It is unquestionably the liomologue of the suprasylvian sulcus of tlie Carnivores, to which we have referred as S in the other Edentate brains. It is interesting to note that it is strongly arcuate, like the corresponding sulcus in the Carnivores and its rudiments in the Ant-eaters ; for it therefore contrasts in a marked manner with the corresponding sulcus in the lowlier Unffulates and the Aard-vark, in which it is horizontal. In both of the brains of Bradypus which I have examined, a deep furrow springs from the angle of junction of the anterior and posterior rkinal fissures and extends obliquely upAvard, with a slight inclination backward toward the central point of the concavity of the sulcus S. This furrow we may for the present distinguish as the sulcus /I. If we were considering the brain of a Carnivore or an Ungulate, there should be no hesitation in calling the sulcus n the fissure of Sylvius (as that term is generally applied outside the Primates), but there are reasons for some hesitancy in adopting this title in Bradypus, because no other Edentate whatever has a Sylvian fissure. There is a considerable resemblance between the brains of Bradypus tridactylns and the small Carnivore Genetta t'lgr'ma, of which Mivart has given some useful illustrations *. The cranial aspect of the pallium of Genetta, the brain of which is not unlike that of Bradypms in shape, is impressed by sasittal, suprasylvian, and supraorbital sulci, which in simplicity are comparable with those of the Three-toed Sloth. The shape and appearance of the suprasylvian sulcus and its relation to the simple oblique Sylvian fissiire are exactly comparable to the arcuate sulcus S and its relation to the enigmatical sukais n in Bradyprns. A comparison of my fig. 11 with Mivart's fig. 11 demonstrates this, but the similarity of our younger brain is much more surprising. The resemblance * St. George Mivart " On the JElnrmdin" Proc. Zool. Soc. London, 1882, p. 516. THE BRAIN IN THE EDENTATA. 319 between these two brains is rendered all the more patent from the fact that the crucial sulcus has almost disappeared from the brain of Genetta. In makin