British Museum. by JANES YATE JOHNSON, C.M.Z.S. NO. LIII

1899.1 ON THI ILNTIPAl'HARIAN CORALS OF MADBIBA. 813

11. Notes on the Antipatharian Corals of Madeira, with Descriptions of a new Species and a new Variety, and Remarks on a Specimen from the West Indies in the British Museum. By JANESYATE JOHNSON, C.M.Z.S. [Received May 22, 1899.1 The marine objects popularly called Black Corals are zoophytes which constitute the group of Antipatharia in systematic z001og~. Some of them are much branched and resenible bushes that occasionally reach the height of four or five feet. Others extend their branches almost in one plane in a fan-like manner ; others, again, are simple unbranched stems, slender and wire-like, that are sometimes found with a length of seven or eight feet. All are attached, when living, to submarine rocks or stones by a thiu spreading base. All have Ihard horny axis of a black or brown colour, and that axis is seen, on examining a section, to consist of concentric layers. Further examination will show that it has tt fibrous structure. Stem and branches are frequently armed with' minute spines arranged in longitudinal or spiral series, but sometimes the stem and inain branches nre amooth and polished. The hard axis is Necreted by the soft polypiferous ccenenchyma which clothes it. The polyps in the Madeirnn forms have six (in one species twenty- four) simple tentacles. Spicula are not anywhere present, and thus the Antipatharia are easily distinguished from the Alcyouaria. Eight species of Black Coral belonging to six genera have beeu found at Madeira, more than one-thirteenth of the total number of known species. In the late George Brook's excellent Report on the Antipatharia of the ' Challenger ' Expedition (1889) ninety- eight species were dealt with, but these included not only the forms collected by the naturalists of that expedition but all those previously described. The Report iq therefore a Monograph of the group. Until the publication of that work much difficulty was experienced in comiug to a conclusion with regard to the discrimination of species and the identification of specimens, owing to imperfect description and confusion of nomenclature ; and even now, notwithstanding that author's efforts, much remains to be clone, especially in regard to our knowledge of the polyps, before satisfactory definitions are possible and the classification placed on a trustworthy basis. All the species of Madeira come from depths below 40 fathoms They are brought to the surface by becoming entangled now and then in the lines of the fishermen. Of the eight species of Black Coral bere treated of, five have not hitherto been found elsewhere, and one of these is now described for the first time (Leiopathes eqansci). Another of the five species, having been confused with a West Indian species, is here distinguished by a fuller description, whilst a new name has PBW. ZOO&. soCr.-1899, NO.LIII. 53 814 MR. J. Y. JOHNSON ON WPl [June 20, been necessarily imposed on the tropical form. Furthermore, another of Gray’s species (Antipathes setacea) has been removed from the position it occupied of a queried synonym and is reestablished as distinct by a more detailed description. Two of the eight species here dealt with (Xavqlia Inmarchi and Leiopathes glaberrirna) are known in the Mediterranean, which great sea possesses apparently a smaller number of species than the sea immediately surrounding the diminutive island of Madeira, for it seems that not more than six forms can with certainty be attributed to the former. These are (in additiou to the two already mentioned as common to the two seas) Antipat7ies clichotonza Pall., A. ntecliterranea Brook, Anti- pathella subpinaata (E. & s.), and Parantipatlies larix (Esper). The existence in the Mediterranean of any elongate unbranched form like Cirripathes requires confirmation.

Gen. SAVAQLIANardo, 1843. Corallum horny, without spines ; polyps with 24 tentacles in alternate rows of twelve each.

SAVAGLIALAMABCKI (Haime). Savaglia lamarcki, Brook, Antipatharia of the ‘ Challenger,’ p. 79. Uerardia lamarcki, Lacaze-Duthiers, 1864. Leiopathes lamarcki, Haime, 1849. Moderately branched, furcately, in one plane. Axis black or brown ; trunk and branches very sinuous, not cylindrical, but compressed laterally so that the anterior and posterior faces are narrower than the intervening sides and the angles are rounded off. There is no groove on any of the faces or sides. The surface of the stem and main branches is seen under the lens to be minutely wrinkled and finely punctured, ilie punctures being numerous and irregularly scattered. The branches are elongate and tapering, very seldom fusing together. Only two specimens of this species have come under my observation. Neither has a base. The smaller example has a height of 50 centim. (278 in.) and a spread of 45. It has beenentirely stripped of its polyps. The other is 85 centim. (32; in.) high, and the longer axis of the lower part of the stem measures 15 millim. It has lost moRt of its branches ; there are remains of the ccenosarc and polyps on the highest ones ; their oval mouths, with thickened lips marked by radiating grooves, are large and conspicuous, the longer axis measuring 4 millim. There is some doubt as to the true zoological position of this organism. The horny branched axis has all the appearance of being antipatharian, but the polyps with their 24 tentacles are closely allied to the Zoanthids, es ecially to the genus Parazoanthw Haddon & Shackleton. Carygren therefore has advocated the re- mopal of Uerardia lamurcki from the Antipathad to the Zmthidta 1899.1 ANTIPATEARIAN UOEALS OF MADEIRA. 815 (‘ Ueber die Qattung Gkardia,’ 1865), and PourtalBs in 1871 took the same view. In the ccenosarc of Mediterranenn examples of this species Lacaze-Duthiers found a peculiar cirripede which he named Laura gwardicP. This has not been observed in Madeiran specimens. Hab. Madeira ; Mediterranean.

Gen. STICEOPATHBSBrook. Axis forming a long, slender, flexible rod without branches. Polyps arranged in a longitudinal series on one side of the stem, not distributed on all sides as in Cirripathes ; tentacles six.

STICHOPATHESGBACILIS (Gray). Antipat7m (Cirripathes) yaracilis, Bray, P. Z. 8. 1857, p. 291. Stkhopathes gacilis, Brook, Antipatharia of the Challenger,’ p. 90. Jet-black ; the stem throughout armed with short conical spines at right angles to it, arranged irregularly in spirals (fig. 111. 1, p. 823). On the lower part of the stem there are about nine series. The base spreads thinly over the object to which it adheres, and is from 10 to 15 millim. in diameter. The lower part of the stem is usually from 3 to 4 millim. in diameter. This species is not of very rare occurrence. The individuals are commonly attached to a well-rounded stone or to masses of cal- careous sand cemented by shells, worm-cases, &c. One small specimen had seated itself on the spineless test of a dead sea-urchin (Avbmia). Two or more may sometimes be seen adhering to the same stone, and, indeed, I had once observed as many as twelve individuals on the same block. But great was my astonishment when a stony niass, 10 in. by 5, was shown to me upon which were seated more than 120 specimens, in two groups, some 20 being sepnrated from the rest, which formed a grove so thickly planted that it was difficult to count them correctly. Unfortunately t,he majority were broken, leaving stems only a few inches long; the length of the perfect ones was about three feet. Two and even three distinct stems may spring from the same basal expansion. It may have been that the bases were at first separate and afterwards coalesced as they extended, but there was no evidence to show that this had been so. The largest specimen that has been met with at Madeira had a length of 9 ft. 3 in. (2820 millim.). This has been placed in the Seminario Museum, Punchal. In contrast with specimena of this size, young ones 6 millim. long have been found, and two of these have been mounted in balsam on a slip of glass. Brook says that the stem is sinuous but not spiral. Two specimens, however, are in my possession which in their upper part form a few very loose irregular spirals. They are on the same mass of indurated And, shells, worm-cases, &c. Normally the individuals of this species are destitute of branchee 5Y* 816 YR. J. Y. JOHNSON ON THR [June 20,

but I have a specimen which has put forth a very slender spineless branch 150 millim. (nearly 6 in.) long. This abnormality was perhaps due to the fact that the upper end of the stem, four or five inches above the ramus, had been broken off by.some accident, and the branch may have been the result of an effort on the part of the colony of polyps to continue their growth. With regard to t,he spines, those of all the specimens I have seen are simple and conical, but Brook says that the majority of those on the older portions on the stems of specimens in the British Museum formed double spines. Hah. Madeira.

STICHOPATHESSETAOEA Gray. Antipathes (Cirmpnthes) setacea, Gray, Ann. & Mag. N. H. ser. 3, vol. vi. p. 31 (1880). ? dntipathes simplex, Alcide d’Orbigny in Webb & Berthelot’s Hist. Nat. des Iles Canaries, Zoologie : Polypiers, p. 151. Dr. Gray’s description of his A. setacea runs thus :-rc Coral simple, elongate, setaceous, straight, erect, closely covered with short conical spinnles. Length 18 in. Hab. Madeira.” He said further that it was straight, without the slighest tendency to assume a spiral form. Mr. Brook (Antipatharia of the ‘ Challenger,’ p. 90) stated that he had been unable to fmd Gray’s type of A. sctacea, and added--‘l As I have no means of ascertaining what form Gray did regard as A. setncsa, and as his description of the type contains no characacter not applicable to this species (i. e. StichopatlLes gi-acilis Gray), I have given A. setacea as a probable synonym.” In going over my collection of Antipatharia for the purpose of preparing this paper, 1 have found a small specimen, which, fragmentary as it is, proves beyond a doubt that Gray’s A. setarera is a good species, quite distinct from Sticliopathes graci7i~. The specimen referred to is a portion of a stem broken at both ends ; what remains has a length of only 14.5 millim. (58 in.), with a diameter at the thicker end of scarcely so much as *$5 millim. It tapers very gradually, and there is a wide central channel. It is bent into a semicirctilar form, and it has a brown colour. The spines are numerous and arranged in longitudinal rows, of which about six may be seen in one aspect. They are upright, high in comparison with the diameter of the stems, and more or less compressed. A few are simple and pointed, but most of them are bifid or notched irregularly at the tips (fig. 111.2, p. 823). Sometimes there are narrow longitudinal ridges bearing as many as five spikes. Measnring fiom tip to tip of the spines in the same row, the interval between my two is about equal to three or four times the height of each. Polyps are altogether absent. The specific name and Gray’s epithet (‘setaceous ” are suitable enough to this very slender form, but me quite inapplicable to Btichopathes graciZis. Alcide d’Orbigny (~oc. cit.), a reference not given by Brook, described a small Ant+athe.s from the Canaries, 1899.1 ANTIPA'l'HAEIAN UORALS OF MADIPIEA. 817 which may possibly have been of this species or a young St. gracilk, in these terms : '' Antipatha simplicissima, elongata, filiformis, longitudinaliter sexcostata, costis echinatis. Long. 35 mm." Hub. Madeira.

Ben. LEIOPATHXS(Bray), Brook. Corallum much branched : stem and thicker branches polishe spineless, ultimate brauches bearing very small and distant spines. Polyps on all sides of the branches, with 12 mesenteries in the oral cone, 6 below ; tentacles six.

LIIOPATHESGLABEEEIYA (Esper), M.-Edw. Corallum branching on all sides, forming a bush ; stein and main branches thick ; branchlets springing nearly at right angles from opposite sides of the branch subalternately; spines on the ultimate branchlets very short, conical compressed, at right angles to the branch. This species has been found in the Mediterranean, and it is believed to be the only Old World antipatharian that occurs in the West Indies. A fine specimen, 150 centim. (4 ft. 11 in.) and 80 centim. (31 in.) through, was obtained off Seixal, a village on the N.W. coast of Madeira, and has been placed in the Seminario Jluseum, Funchal. It is destitute of its base ; the stem below the first branch is only 12 millim. thick. There are two main branches which run to a greet length and in their lower parts are almost as thick as the stem. These and the secondary branches are strongly and irregularly sinuous, and with the ultimate branchlets form a round bush. The branchlets are very fine and hair-like, and are set with short, broad, conical upright spines at irregular distances apart, not in rows or whorls. The stem and main branches are black, smooth, and shining. Nowhere is there nng fusion of the branches. The branches were thickly covered with polyps of a warm brown colour. The tentacles were subulate in form and much longer than the body. Several other specimens of a smaller size have occurred from time to time. It was remarkable that not a single organism of any kind had established itself parasitically on any part of the large specimen, a great contrast with Aphnipnthes roollastoni when it is brought up to the surface. Hab. Madeira, Mediterranean, W. lndies.

LEIOPATHESEXPANSA, sp. n. (Fig. I., p. 818.) Mwh branched in one plane or in parallel planes to the sixth degree of subdivision. Stem and branches elliptical in section, jet- black, polished, bent into irregular zigzags, the branches being thrown off alternetely on opposite sides. All the branches arise almost at right angles from the parent branches at a distance from each other. The ultimate branchlets are very slender, hair- 8 18 Mk. J. Y. JOHNSON ON THE [June 20 like, tapering, and sharpened off to a point at the extremity. Minute upright conico-subdeltoid spines are irregularly scattered on the ultimate branchlets, the other parts of the cornllum being spineless (fig. 111. 3, p. 823). Polyps pale red, with six tentacles. The remains of the cmnenchyma and polyps are seen on the specimen as a brown pellicle coating the finer branches, and extending as a thin web or film from branch to branch.

Fig. I.

Lewpathes expama, sp. n. Abont 8 nat. mze. From a photograph.

The only specimen of this new species that I have met with waa obtained from a fisherman twenty years ago. It is without a base, and has a height of 405 milliln. (16 in.), with a spread of 380 millim. (15 in.), but its spread when perfect was probably not less than 445 millim. (178 in.). The thickest part of the stem is only 5 millim. in diameter. 1 No fusion of branches is anywhere visible. The elegant flabellate form and delicate habit seem to distinguish this sufficiently from known species of Leiopather. Eab. Madeira. 1899.1 ANTIPATHAEIAN COBAM OF MADBIRA. 819

Gen. ANTIPATHES. Shrub-like, branches not fusing; spines numerous, strong. Polyps large ; tentacles six, radiating, one pair in a line with the oral slit inserted low down, the others at the margin of the peristome.

ANTIPATHEB FURCATA Gray. Antipathes furcatn, ‘Gray, P. Z. S. 1857, p. 201. dntipathes? furcata, Brook, Antipatbaria of the ‘Challenger, ’ p. 104, pl. xi. fig. 2. No specimen of this species has been met with by me, but Gray’s type, obtained by N. Mason at Madeira in 1557, is in the British Museum. Brook believed it to be only a branch of the entire corallum. The habit is different from that of the other bushy Black Corals of Madeira. In order to ’enable collectors to identify any specimens that may occur, an abbreviation of Brook’s description is here given. The specimen is 16 centim. (64 in.) high. The axis is very slender and bears a number of elongate bristle-like branches, which are directed subvertically and reach to about the same height. The branches give off secondary branches at irregular intervals, and the longer ones bear a third series of branchlets, usually on one side only. Nearly all the branchlets are directed upwards and most of them reach the apex of the corallum, and thus it has a corymb-like aspect. The spines are short, triangular, and compressed, with the apex at right angles to the axis. Six longitudinal rows can be seen from one aspect, and the spines in a row are three or four times their own height distant from one another. Polyps? Hub. Madeira (Mason) : British Museum.

Gen. ANTIPATHEILLABrook. Branching in one plane, branches not fusing together ; spines short, upright ; polyps small, with six tentacles, in two series of three each. ANTIPATECNLLABEAULLIS (Gray). (Fig. II., p. 820.) Antipathes gracilis, Gray, Ann. & Mag. Nat. Hist. ser. 3, vol. vi. 1860, p. 311 ; not Antipathlla gracilis, Brook, Antipatharia of ‘ Challenger,’ p. 113. In 1860 Dr. Gray (loc. cit.) gave a short description of a small and delicate antipathnrian from Madeira and assigned to it the specific name gracilis. In 1888 Mr. Brook, when preparing his monograph of the group, was not able to fiud in the British Museum any specimen from Madeira bearing that name ; but he found there a Black Cora1 from the West Indies to which was attached a label with the name in Gray‘s handwriting of Antipathes gracilis. Under these circumstances, Brook in his Report described the WesbIndian specimen under the name of Antipathlla gracilis (Gray). This was a mistake which he would not have committed 820 MR. J. Y. JOHRBON ON THE [June 20 if he had had a specimen of the true A. gracilis from Madeira before him. Indeed he himself felt uncertain whetbr the course he took was right, for he says he was “ at a loss to understand Gray’s description,” and added that it seemed ‘‘ doubtful whether this specimen (that from the West Indies) could be considered to agree with Gray’s definition of the species.”

Fig. 11.

Antipatheh gracilis, about 0“ nat. size.

An examination of specimens obtained at Madeira proves beyond dispute that Gray’s short description of A. yrafiilis ap lies to them; whilst a study of Brook’s description of the West-InJan specimen leads to the conclusion that the latter cannot belong to the same species. I shall therefore proceed to give a fuller account of the Madeiran form under Gray’s name of yracilis; and then, rather than leave the West-Indian specimen without a name, I shall repeat Brook’s description of it and assign to it the name of brooki. In this way I hope that the confusion surrounding the two forms will be cleared away and the nomenclature settled once for all. Antipathella grncilis (Gray).-Corallum black, rery slender, arising from a small round flat base, sparingly and laxly branched in one plane ; branches distant, elongate, straight, never confluent ; ultimate branchlets setiform, tapering to a point, from 20 to 50 millim. long. The corallum is everywhere set with minute triangular upright spines (Fig. 111.4, p. 823), those on the stem forming about seven longitudinal rows. Polyps disposed in a series on one side of the branches, separated by short intervals (fig. III. 5, p. 823). The corallum seldom exceeds 150 dllirn. (6 in.) in height and has a spread of rather more. The thin base has usually a diameter of about 6 millim. and the lower part of the stem is noL more than about 1.5 millim. in diameter. The branches make an angle of 30” or 40’ with the stem. This species seems to live gregariously, as the dredge will sometimes bring up a quantity of it. One specimen was found attached to an old individual of ISSO.] ARTIPATHAEIAN CORALS OP YADDEIEA. 821 Ap7wrtipathes mollaetoni. Others have been discovered seated on the telegm$-cable when hauled up for repairs from a considerable depth. The figure shows one of these, a small but characteristic specimen, only 106 millim. (4i in.) high, with a spread of 110millim. (4+ in.). Hd. Madeira.

[ANTIPAT~LABEOOKI, nom. nov. Aniipathella ? gradlia, Brook, Antipatharia of the ' Challenger,' p. 113, non Gray. The specimen in the British Museum is 56 centim. (22# in.) high, and is related to other flabellate forms included in the genus Antipathella. The base consists of several stems fused together, which give rise to a series of branches not all in the same plrtne, with frequent fusions. Upper portion more spreading, but the larger branches are still strong and are often fused together. In some portions nearly all the branches come off from one side and are placed at irregular intervals. Secondary branches mostly very slender. Medium branches bear branchlets irregularly varying in length from 15 to 100 millim., usually longer on one side than the other. Bmaller branchlets simple and filiform ; the larger ones are again branched irregularly, the ultimate pinnules being very slender and rarely attaining a length of 12 millim. without becoming branched. Spines (Brook, Antip. Chall. pl. xi. f. 8) short and conical with a slender apex, arranged in dextrorse spirals. E'ive rows are visible from one aspect of a pinnule, the members of a row being from two to three lengths apart. (Brook, be. ck) Hub. West Indies (Scrivener).]

Gen. APHANIPATHESBrook. P Corallum paniculate or flabellah ; spines long and slender ; polyps obscure, with short tentacles.

APHANIPATHESWOLLASTONI Brook. Aphanipathes woZZmtoni, Brook, Antipatharia of the ' Chal- lenger,: p. 126. Antzpathes mbpinizata Gray (non Ellis & Sol.), P. Z. S. 1857, p. 293. Colour dark brown. Bushy, shrub-like, branched to the fourth or fifth degree of subdivision. The secondary branches elongate, often reaching to the top of the bush. Ultimate branches very numerous, very slender and varying considerably in length. They and the penultimate branches are thickly set with spines which are atranged in six or seven longitudinal rows as well aa in irregular spirals. The simple, tapering, acute spines rise from a broad thick base and are directed obliquely forwards. The polyps are seated on the upper side of the branchlets in a single row at a distance from each other. Six short, thick, conical 822 ME. J. Y. JOHNSON ON THE [June 20, tentacles, pale brov.vn in colour, surround a puffed-out mouth, the whole covered with vibratile cilia. The polyps are very full of stiiiging-threads (fig. 111. 6, p. 823). “This .is the species which Gray referred to Antipathes sub- pinnatcc E. & s. It differs essentially from that species in the arrangement of the pinnules and in the form of the spines.”- Brook, loc. cit. p. 127. One of my specimens 35 centim. high has a thin base 35 millim. by 25 millim. From this base rise not only the large corallum, but several small ones from 25 to 50 millim. high. This is the commonest of all the species found at Madeira. It is usually attached to loose stones, but in one case within my knowledge a specimen two feet high was growing upon a quaintly shaped metal tankard, the whole exterior of which was completely hidden from view by a crust of bryozoa, worm-cases, &c. Individuals of this species are often made common lodging- houses for the use of a heterogeneous throng of guests. More than 25 different forms, including mollusks (Ostrea cochlear and Avicula tarentina), bryozoa, worm-cases, hydrozoa, Polytremn, and sponges, have been seen crowding on the lower branches of a single specimen ; thus offering a great contrast to specimens of Leiopathes glaberrima, which are always free from parasitical attendants, a difFerence doubtless due to the abundance of spines on the one form and their absence from the other. What is still more curious is that a small stalked cirripede, the Oxttnmpis celntcc of Darwin, is found attached in numbers to the branches of this Aphaniputhes and nowhere else. The antipatharian covers the valves of the cirripede with a thin horny coat beset with minute spines. In Alcide d’orbigny’s list of the zoophytes of the Canary Islands, in Webb and Berthelot’s work, appears the name of Antipathes subpinnata Ellis & Sol. This may have been a specimen of Aph. wollabtoni. Hab. Madeira ; Selvagens or Salvages ; Canary Islands ?

Var. PILOSA, nov. Bushy, branching irregularly ; in general characters resembling the typical species, but a distinct aspect is given to the present form (1) by the ultimate branchlets being stouter with reference to the branch from which they spring ; (2) the angle they make with the branch is more obtuse j (3) they spread in all directions from the branch, whereas in the typical species the ultimate branchlets have a tendency to spread in one plane. The spines have much the same form as in the typical species :that is, they are long, slender, pointed and directed forwards, and are arranged in longitudinal rows; but in the present fodn they differ by being longer in regard to the diameter of the branch on which they are placed, and by being less closely set (fig, III. A). On the lower part of the etem the spines are frequently forked at the top. Thie variety is remarkable in this, that the stem bears numeroue 1809.1 ANTIPATEAEIAN CORALS OF MADXZRA. 823 scattered microscopic hairs, which are seated on thick bases and taper to a fine paint. They sometimes fork near the tip (fig. 111. B). A single specimen has been in my possession for many years. It is without a base, and may possibly be only a branch of the entire corailum. It has a length of about 205 millim. (8 inches) and measures 180 millim. (7 in.) through. Should other specimens occur, a careful examination of them might lead to the conclusion that this form is entitled to rank as a species. Hub. Madeira. Fig. III.

1 2 3

5 6 Arrangement of spines, all X 16. 1. Stichopathes gracilia (p. 815) near apex of stem. 2. Stichqathes setacea ( 816). 3. Lewpathes expansn (p. 817), ultimate branchlet. 4. Antipa- &lla graeilis ( . 819). 5. Same, with two polyps. 6. Aphanipaths wollastoni (p. dl),with polyps. A. Aph. wollastoni var. pilosa (p. 822), branchlet. B. Same, part of stem.

Key to the eigkt Madeiran Species of Antipatharia.

A. SIMPLESTEMS. Stem fy thick as a goose-quill ; spinea simple .. . Stichopathes gracilis (Gray). Madeira. Stem very slender ; spines forked or jagged ...... Stichopathea setacca (Gray). hbdeira. 824 THB SWR~TABYON ADDITIONS m m~MBNAG~RIID. [Nov. 14,

B. BRANCHED. a. Branching in one plane. Short, very slender, sparingly branched ; covered with spines. Antipathella, racilis (Gray). Macleim Spine8 on the ultimate branchlets only ; stom smoo$. Leiopatlaes expama J. Y. J. Madeira. No spines ; stem and branches wrinkled and punctured. 8auagli.a lamarcki (Hairne). Madei a, Mediterranean.

h. Branching on all sides, buahy. Bpiues only on the ultimate branohlets. Leiopathcs glabewima M.-Edw. Ma- deira ; Mediterranean ; w. Indies. Spines sbort, triangiilar, upright, branches arranged in a corymbose manner. Antipathes furcata Gray. Madeira. Spines elongate, directed forwwds... Aphaiiipnthes wollastoni Brook. Ma- deira ; Selvngens. Var. pilosa with hairs on the stem.

November 14, 1899. Dr. A. GBNTHEB,F.R.S., V.P., in the Chair. The Secretary read the following reports on the additions made to the Society’s Menagerie during the months of June, July, August, September, and October, 1899 :- The total number of registered additions to the Society’s Mena- gene during the month of .June was 164, of which 79 were by presentation, 18 by birth, 40 by purchase, 1 was received in exchange and 26 on deposit. The total number of departures during the same period, by death and removals, was 78. Among the additions special attention may be called to a male Sitatunga, or Speke’s Antelope (Tmgelaphus spekii), from the district of Lake Ngami, received from Mr. Oecil J. Rhodes, F.Z.S., in exchange for a female hybrid between Tragelaphus gratus d and T. spelcii 9 (born in the Menagerie on Feb. 12, l896), which was despatched to Mr. Rhodes on April 25th last. The total number of registered additions to the Society’s Mena- gerie during the month of July was 204, of which 77 were by presentation, 19 by birth, 36 by purchase, 1 was received in exchange and 71 on deposit. The total number of departures during the same period, by death and removals,,was 108. Amongst the additions special attention may be called to the fine Ground-Hornbill presented by Dr. Hirst on July 20th, which appears to be a young example of Bucornr abyssinim, and concerning which the following information baa been received from Dr. Hirst :- 1899.1 T€IE BE1oBBTARY OF ADDITIOFB TO THpl MXTNAGm. 828 The Elms, Goldthorn Ha Wolverhampton, July 17, 1899.

9% 6 I have recently returned from West Africa and brought with me a large bird, which I think you may like to have in the Zoo- logical Gardens. The bird, which was caught on the R. Volta, Gold Cvast, W. Africa, stands about 2 feet in height, has an enormous beak, and lives on the banks of the river, wading into the water and catching fish, on which it lives. I have had this bird in my possession some aix or eight months, and it is now quite tame. I atii, Sir, Pours truly, GEO. HIRST, The Secretary, Th Colonial Medical Service. Loologics1 Society, 3 Hanover Bqiiaw, W. The registered additions to the Society’s Menagerie during the month of August xere 102 in number. Of these 37 were acquired by presentation, 40 by purchase, 6 were born in the Gardens, 1 was received by exchange and 18 on deposit. The total number of departures during the same period, by death and reiiiomls, VRS 100.

Clr6vy’s Zebra (female). Amongst these may be specially noticed a pair of Gr6vy’s Zebras (Equus grew;;), deposited by H.M. The Queen on August 14th. 826 THI s~oswlda~ox ADDITIONS TO ‘PHPI YHNAQBRTB. pov. 14, We have now the pleasure of being able to exhibit in the Gardens a pair of the beautiful Zebra of Southern Abyssinia and Somaliland (Epuagrmrii), to the existence of which I have on several former occasions (see P. Z. 8. 1883, p. 721 ; 1800, pp. 412, 461 ; 1893, p. 473 : 1898,‘~.588 ; 1899, p. 713) called attention. The present animals have been sent by the Emperor Menelek as a present to H.M. the Queen, and have been deposited in the Society’s care to be recovered from the effects of their long and arduous journey. I. need hardly descant on the differences between the present animal and the three previously known species of Zebra, both in size and in character of markings, as they are obvious at first sight. These animals were captured somewhere in the south of Shoa, and are Raid to be the only survivors of a considerable number which reached Addis Abbeba. Here they were delivered to Capt. J. L. Harrington, the British Political Agent at the Court of the Emperor Menelek, as a present to H.M. the Queen, and sere brought down under his care to Zeila on the coast of Somaliland, a long march OE six weeks. At Zeila Capt. Harrington handed them over to our Assistant Superintendent, Mr. Arthur Thomson, who had been sent there by the Council, at the request of the Foreign Office, on purpose to take charge of them. Having been conveyed acrous to Aden, where they were kept for about ten days, they were shipped in the P. & 0. 8.8. ‘ Oceana,’ and arrived safely at the Iloyal Albert Dock on August 14th last, and were thence brought to the Society’s Gardens. The female, of which I exhibit a photograph (see p. 825), may now be announced to be in perfect condition : but the male, I regret to say, still shows wounds on the hocks, which, in spite of every care and attention, we have been unable to cure. The total number of registered additions to the Society’s Mena- gerie during the month of September was 126, of which 84 were by presentation, 9 by purchase, and 33 were received on deposit. The total number of departures during the same period, by death and removals, was 90. The total number of registered additions to the Society’s Mena- gerie during the month of October was 1‘70, of which 31 were by presentation, 68 by purchase, 61 were received on deposit, and 10 were born in the Menagerie. The total number of departures during the same period, by death and removals, was 169.

Mr. Sclater stated that in July last he had visited the Zoologicd Gardens of Rotterdam, Amsterdam, and Antwerp, the private collection of Mr. F. E. BIaauw, C.M.Z.S.,pnd the Museums of Brussels and of the Congo Free State at Tervueren near Brussels. Mr. Sclater spoke of several animals of much interest which he had thus seen. Amongst these were a pair of Slow Lemurs (Nyctiecbus tardigradus) at Rotterdam with a young one, which 1899.1 ' MR. ~OL.~!IRON CONTIXTENTALMLIBNILGIIRIEB. 827 was carried by the mother aoross her breast after the manner of other Lemurs ; also a young Pelican (Pelwanus onocrotalua), hatched in the same Gardens and then still in nestling plumage. The Amsterdam Gardens were fortunate in again possessing a living Anser rujkollis, which had been captured about the 10th February, 1899, at Foxhol, near Hoogeram, in the province of Groningen ; also several specimens of the now rare Talapoin Monkey (Cercopi- thecus ta7apoin), and two examples of the beautiful Red Oriole of Formosa (Analcipus arclens), besides a family group of the Pleasant Antelope (Tvagelophus gratus), consisting of an adult pair, two young males, and a newly-born calf. At Antwerp there was, likewise, a small herd of I'ragelaphus yratus, consisting of an adult pair and two young females j also a fine adult male of the Roan Antelope from Senegal (Hippotragus equinus yambinitus) (see ' Book of Antelopes,' iv. p. 15, pl. lxxviii.), and three examples of the true Dama Gazelle (Gazella dama), from Senegal. In the Antwerp Gardens Mr. Sclater had likewise examined a liviug female monkey which appeared to belong to a new species of the genus Cercocebus, remarkable for its prominent crest on the middle of the head and the long hairs on the cheeks. This specimen had been received by the Antwerp Gardens as a present from M. I!'.

Fuchs, the Governor of the Congo Free State, and mas believed to liave been obtained, in March 1899, in the district of Stanley Falls ou the Upper Congo. With the approbation of M. L'hoest, Mr. Sclater proposed to 828 MR. SCLATER’S JOUBR~DYTO THEI (IAPE. [Nov. 14, designate this species Cercoce6w dongincs with the following characters :- CERCOCEBUSCOXQICUS, sp. nov. (Woodcut, p. 827.) Niger, mbth nudiu.sculus, cristd extante Zongd nip4 :genurumpilis productis albis :manibus et pedibzcs cum facie carneis :rnento et pectore albis, ventre nigricante, tibiis albk : brachiis nigris, mzidd albicunte. Long. corp. 2, caucla 3, tot& 5ped. Anql. Hub. Terra Congka. In the new Museum of the Congo Free State at Tervueren, near Brussels, Mr. Sclater had been able to examine the series of specimens of Antelopes of the Congo Territory, which contained examples of the following species :- I. Cephalophus maxwelti (Lower Congo). 2. Cephalophus grinznai (Lower Congo). 3. Cobus penvicei (Matadi). 4. Cervicapru anindinurn (Lower Congo). 5. Hippotrugus epuinw (Lower Congo). 6. Tragelnphus yratzts (Lower Congo). 7. Trageluphus smptus (Cataracts of Lower Congo). Mr. Sclater had been much gratified with the examination of these specimens, as he had never previously seen a collection of -4ntelopes from Congoland. He called special attention to the fact of the Waterbuck being Cobus penricii (see ‘ Book of dntelopes,’ ii. p. 113, pl. xxxv.), as this species was previously known only from the southern part of Angola, also to the fact of the Roan Antelope apparently belonging to the southern form lli1jpotraqus equinus typicus. -- Mr. Sclater gave a short account of his journey to the Cape, from which he had just returned, after an absence of ten weeks. ‘rhe atate of the country caused by the impending war had prevented him from obtnining so many animals as he had hoped to do. But the‘keeprr whom he had taken with him had returned to this country on the 15th inst. with the following animals, which had been mostly obtained from various friends and correspondents :- 1 White-tailed Gnu (Coniiochcetes gnu), d . * 1 Roi Rhehok (Ceruicdpra fulvo-rufuh), 8. 2 Dusty Ichneumons (Heqmstes puluarulfntus). I Cnpe Crowned Crane (Bulearicu replorum). +2 Hchalow’s Tourncous ( Turaczts s( halozui). 4 Cape Turtle Doves (liirtzcr cftpico7a). 1 Vulturine Eagle (Aqziilu verreawi). 1 Tawny Eagle (dqzcila nmioides). * New to the Society’s aollection. t This beautiful pair of Touracous were captured at Npatn, near Mossamedes, Angola, by one of Mr. W. L. Sclater’s correspondents. [They have been well Qqred by Mr. Frohawk in ‘ The Field ’ jvol. xciv. 891, Dea. 2nd, 189!3).] 1899.1 MR. LYDEKKER ON CERVUS DUV?LUCELI. 819 Birr. Sclater also stated that he had given an oddress at a Meeting of the South African Philosophical Society at Capetown on September litli, “ Ou the desirability of establishing a Zoologicrtl Garden in Capetown,” and that the Society had appointed a Com- mittee to consider the subject. Mr. Hclater was in hopes that this inoveiiieut might ultimately lead to good results.

Mr. T,ydekker exhibited the mounted head OF a remarliably fine Swanip-Deer (Cerws dt/uaucdL), sliot by Major C. B. \Vood, of Toiiis IIill, Aldbury, Tririg, on the 6th of Jaiitiary, 1899, ill the

Head of Sw:iinp-Deer (Cuvm dirwwcli) froin Centr,il Priivince;, Iurlio.

Centi*:d Provinccs, India. The specimen a-as noticea1)le oil arcoutit of tlie approximation 01’ the aiiths t.0 those of that variety of the T1i:~iirenyli110~\.11 as C. dtli yltctyc~ros. This was PItlX. ZOOL. SOC.-18!)9, KO. LTV. 54 830 MR. F. VAUQHAH KIRBY ON TEE [Nov. 14, shown hy the circumstance that the main bifurcation took place at a much higher point than usual, and the upper tine of the fork Was folloned by two other tines, thus giving the appearance of the row of small tines in C. elcli plutyceros. Moreover, the very large angle formed by the brow-tine with the beam suggested the continuous curye of the cZdi antler. The specimen clearly demonstrated the propriety of including C. elcli in the same subgeneric group as C. duvctueeli, rather than that of separating the former as Pmolicr.

A communication was read from Seiior Florentino Ameghino, C.M.Z.S., containing further remarks on Xeomylodon listui from Patagnnia. He proposed to id8ntif.y it with the so-called “ Jemich” (or ‘‘ WaterTiger ”) of the Tehuelrhr Indians. This~‘ferociou8beast” had been referred to by Musters, ‘At Home with the Patagonians; ed. 2, pp. 104, 105 (1873). Seii. Ameqhino also considered the So or Succarofh of Lozano (‘Historia de la Conquista del Paraguay,’ vol. i. pp. 285, 286, 1873) to be probably the same rcnimal.

Mr. A. Smith Woodward exhibited, on behalf oflDr. Moreno, the skull and other specimens of Neomyloclon listui (Grypothe&rn) lately discovered in the cave in Southern Patagonia where the original pieces of skin were first obtained, and made remarks on them.

The Secretary exhibited, on behalf of Nr. C. E. Pole Carew, F.Z.S., some malformed horns of the Sambur Deer (Cewus ark- totelis), obtained bg him in the southern province of Ceylon, and read some notes on them sent by Mrs. Cnrew.

The following papers were read :-

1. Field-notes on the Blue Duiker of the Cape Colony (Ceyhalophus monticola). By F. VAUGHANKIRBY, F.Z.S.

[Received May 22, 1899.3 a Although in point of numbers one of the commonest of the Colonial Antelopes, this delicate little creature is of such retiring habits, and its size is so insignificant, that less is known of it than of other Antelopes of the district. In the densely-wooded kloofs and on the hill-sides covered with low scrub and often impenetrable thorn-jungle, which form their home, it is rarely indeed that even the most skilful stalker can.,move with such silence and care as to be neither heard nor seen by them, even though their restless disposition causes them often to move about during those hours of daylight when most other forest-duellers are asleep. Even in the densest bush, the spaces undrrnenth to a height of two feet from lS9O.j BLUE DUIKER OF THE CAPE COLONY. 831 L the ground are comparatively clenr, hence the little Blue Duiker, or Blue-buck, as it is generally called, moves about in what to him is practically open bush, in which objects are iisible at a considerable distance : thus the stooping, struggling form of the stalker worming a passage through an opening two feet square in an unyielding wall of thorns, or striving to free himself from the too firm embrace of a network of “ wacht-een-beetje ” bushes, cannot fail to attract attention long before the little grey watcher, standing motionless in the shadows, has been discovered. At the bush-drives so common in the Colonv, Blue-buck are seldom turned out ; they will keep such dogs as have not learned wisdom of experience tearing round and round a kloof all day, bnt will never venture to break unless by chance a hard-pressed individual takes advantage of some narrow bush-strip at no unwatched, unthought-of point to escape l~yway of it into the next kloof. The Blue-buck may be easily bagged, however, in the early morning by the exercise of a little judgment. The direction of the wind must be studied before all things, then search must be made for the most frequented ‘‘ paths ” or “ runs.” Should one be found which is evidentlv a main path to and from certain feeding- grounds, this can be watched ; but a surer method is to find a spot where they are accustomed to feed ou the surrounding bushes. In such places many converging paths will be seen, in view of which, at a short distance down wind, the watcher mud take hie stand before sunrise, keeping out of sight behind a bush or fallen tree-trunk. Under such circumstances, his patience will not be severely taxed before he is rewarded by a sight of the little grey woocl-elves. In localities where water is handy, the paths to and from it inap be watched. In the heat of summer Blue-buck frequently drink between noon and 2 P.Y., but, as is the case with the Bush-buck, in the extensive arid regions of this Colony, the want of water troubles the Blue-buck not at all ; during the trging drought n hich has raged over the Gamtoos river district for the past fifteen months, countless numbers of both Bush- back and Blue-buck have certainly not tasted water from one week’s eud to another. In little disturbed localities I have seen Blue-buck playing about in pairs on fairly open ground bordering the kloofs aa late as 8 A.M., and towards evening, during the hour before sunset, they ma often be seen standing in or crossing any quiet road which passes t Erough scrub-bush. But they are very quick, and though in the dusk they will stand watching the intruder curiously, yet before the light fails they usually scuttle off very promptly, uttering their sharp, but by no means shrill alarm whistle. They are apt soon to stand again, however, so that if silently followed up, a shot may be obtained. When lyiny up for the clay they usually select spots overgrown with thorn-bush and other vegetation, reaching these from the more open bush, in which they feed, by regularly frequented paths. It will be remarked that while in many respects their hnhits are 54’ 832 ON “3H1 BLUP DUIKYS OF I‘m UAPE UO1joNY. [NOV. 14, aimilar to those of the Red Duiker (C.natalnsis), they entirely differ from them in their rigid avoidance of really open ground ;fnr it is well known that the Red Duilier loves to disport himself on open grassy ridges 200 or 300 pards distant from any bush : I have shot many in such situations. “ Scuttle ” is a word which aptly describes the movements of a Blue-buck when alarmed: unlike C. grimmi and 0.natalensis, they do not bound away, but move at a quick scuttling trot. Blue-buck are almost entirely browsers upon bushes, and it is well known that in order to get st branches which are out of their reach when standing on the ground, they will raise themselves on their hind lege like goats, resting their fore feet against the tree-stem. Perhaps, however, the fact is now made known for the first time (if, indeed, it does not actually amount to the discovery of a hitherto unknown habit of this antelope), that the Blue-buck ran and does climb trees ! Ny brother Mr. E. W. Kirby witnessed this singular feat yesterday morning (21st April), when out stalking, and actually shot one as it stood on a branch, browsing on the leaves around it. He was first attracted to the spot by the low grunting souuds they were making, but, though they were evidently close by, he failed to make one out after carefully scrutinizing the surrounding bush. Advancing cautiously, he soon saw the leayes of a 6‘ boer-boon ” tree shaking violentlg, and for tl moment believed it was caused by either baboons or monkeys : at last, to his surprise, he discovered a Blue-buck moving along a branch of the tree some 12 feet from the ground. -4lthough in pursuit of Bush-buck, this opportunity of shooting a Blue-buck under such peculiar circumstances was not to be lost, so he fired and killed it. At the report of the rifle at least eight other Blue-buck dropped frou the branches, apparently reaching the ground on all four legs at once, and scutt.led off; while, as he stepped forward to secure the dead one, a male, another dropped apparently out of the tree under which he had knelt to fire the shot. That same morning he saw Blue-bucks in four different trees. The boer-boon tree sbove mentioned rose from the ground at an angle of about 50°, but the Blue-buck were not on the main trunk, but amongst the smaller branches. I regret that an accident had prevented me from being with my brother that day, as I had intended ; but he assures me he will be able to point out the spot another day, when he is confident I shall an be eye-witness of this singular climbing feat. Perhaps it will not be out of place if I here make reference to the singular little Antelope which I shot in Nov. 1896, in the Kwaws district, Portuguese East Africa. The skull is still in Mr. Rowland Ward’s hand, (the skin was unfortunately amongst the valuable trophiee purloined in Delagoa Bay by a notorions firm of ‘6 forwarding agents ”), and was, at my request, sent by him to Mr. Sclater for examination. It WR~pronounced to be “ apparently that of a Blue-buck,” but without the skin no definite conclusion

.Farmera’ bean-tree.

P.Z. S. 1899. P1.LV.

P 2 s 1899 P1 LVI

WEST -AFRIC*N ARACHNIDA.

P Z S 1899 P1. LVII

WE ST A? H IC'-+N ARACHNIDA

P.2 S 1899. P1.LVlIl. 1833.1 ON SCORPIONS E~M.FROM T~OPIC~LLWEST ABRICA. 433 could be arrived at. Now at the time this opinion was given my knowledge of the BJlie-buck and its habits was but slight ; but during my residence in the Cape Colony I have studied them very carefully, and 1 am now able to ad\ance further reasons why I am convinced that this Kwawa Antelope was not C. rnonticola. At first I could only point to the entirely different coloration, the Kwawa Rpecimen being a warm yellowish-red, with pure white underpart.4, and a wash of the mousegrey colour, peculiar to C. montiwla, on the frontals and nape of the neck ; but it is now evident to me that the habits of the two are quite dissimilar, the Kwawa Antelope being very restricted in its range, more partial to open clearings, Jess thy, and a less pronounced browser, while its movements are epringy and more resemble those of Nesotrayus liviiigstonianus, for which Antelope I and my native followers at fir& mistook it. However, I hope soon to secure other specimens, and until then it is idle to speculate upon the subject.

2. On the Scorpions, Pedipalps, and Spiders from Tropical West Africa represented in the Collection of the British Museum, By R. I. POCOCK.

[Received May 24, 1899.1 (Plates LV.-LVIII.) With the exception of the Attidm and of some of the more obscure species of other familiee, which I have not attem ted to determine, this paper contains a record of the Arachnida be Ponging to the Orders Scorpiones, Pedipalpi: and Armem, now contained in the British Museum, which have been collected at various tirues in West Africa between Senegambia in the north and the Congo in the south. Senegambia has been fixed as the northern limit,, because it is at that point that the western Ethiopian fauna blends with the western Mediterranean fauna. From the countries lying to the south of the Congo we have very little material ; hence this river has been regarded as the southern limit of the area of which the fauna is discussed in the following pages. By far the richest collection, both as regards numbers of specimens and species, that we have received from this area is the one that has been sent in instalments during the pwt twelve months by Mr. 0. L. Bates from the Benito River in French Congo. This collector, whose name has already been frequently mentioned in the pages of the ' Proceedings ' in connection with various rare mammals that he has procured, has been wonderfully successful in his search after Spiders, having sent home representatives of many new species, and added to the Natioud Collection several others which, although previously known, had never found their way into our cabinets. 884 ME..11. I. rococg OR SCOBPIOXS, PEDIPALPS, [RTov. 14,

Order SCORPIONES.

- Family RUTHIDX. Genus RUTHUSLeach.

BLJTHUS(PR~OXKJRUS) CITEINLJS (Hempr. & Ehrenb.). Androctonus (Prionurus)oitrinus, Hempr. & Ehrenb. Symb. Phys., Scorp. no. 6, pl. ii. fig. 2. Prionurus citrinzcs, Pocock, J. Linn. SOC.,Zool. xxv. p. 306 (1896). Buthus citrinus, Kraep, Das Tierr., Scorpiones, p. 16 (1899). LOC.Senegal (Keys. Coll., and several specimens procured from Heine). Apparently extending right across the Saharan region from Senegal to Dongola and Upper Egypt, where it was first procured.

BUTHUSOCCITANUS (Anior.). Scorpio orcilanus, Amoreux, Journ. Phys. xxv. p. 9, pl. i. figs. 1-3 (1753). (= occitanus and europceus (Linn.) of recent authors.) LOG.Senegal (Heine); Gambia (Sir A. Noloney). Senegambia is the southern limit on the west of Africa of this common Mediterraneau species. BUTHUSHOTPENTOTTA (Fahr.). Scwpio hottentotta, Fnbr. Ent. Spt. ii. p. 435 (1793). LOC.Gambia (Mr. Dalton) ; Sierra Leone (Surg.-Capt. Clements and E. E. Austen) ; Shoiigo ( W. A. Forbes); Niger (Sir R. Nurchi- Son) ; Asaba, 180 miles up Niger (Dr. Crosse). Genus LTCHASC: Koch’. LYCHASASPER (Poc.). leon?etrusasper, Pocock, J. Linn. SOC.,Zool. xniii. p. 445 (1890) ; Kraep. Das Tierr., Scorpiones, p. 49 (1899) jArc7tisonzetrus). Loc. Congo (J. Pinaock and A. Curror, Esq., R.N.);Angola (Dr. Wekuitechl). Genus UROPLECTESPet. UROPLECTESOCCIDEh’TILIS Simon. UropZsctes occiclentalis, Simon, Bull. SOC. Zool. Fr. 1’. 210 (l87G). Ltprcus occidenticlis, POC.P. Z. S. 1890, p. 132, pl. xiv. fig. 4; id. Ann. Nat. Hist. (6) xvii. p. 388 (1896) (Uroplectcs). Y’ityus chincltoxensis, Karsch, %. Baturw. hi. p. 370 (1879). ’ k-dby C. Koch upon two species, L. ?nanclatzcs and L sczktilus, which at the present time are not regarded as congeneric. The firet is the tyl’e of He~nprichand Ehrenberg’sgenus Isosometrzcs. Conseqaentl~L. scufil,usstonds,by elimination, as the type of @&zY. Kraepelin applies this system of elimination to the selection of the typ of the genus Heteromelmcy, but not to Lgchad;. 1899.1 AND SPZDBIW PIGOYTROPICAL WEST AFRICA. 835 LOC.Congo (A. Curror, Esq., XN., and J. Pinnock); Cette- Camma, Gaboon (' Germrd ') ; Angola (DP. Weliuitsch). UROPLECTESANDEEB, 5p. n. Nearly allied to U.occidentalis Simon, but recognizable by having the superior edge of the 4th caudal segment armed posteriorly with a toobh-like bubercle like that which is observable on the preceding segments, and the corresponding edges of the 5th caudal segment ending behind in a rounded bifid lobe with a vertical posterior margin ; the tail, moreover, is parallel-sided, not posteriorly incrassate. Neasurements in mi1limetres.-Total length 58, length of carapace 8, of tail 33, of movable digit 7. Loc. Kassai, on the Loangk River, Upper Congo (Xr.Anclren.).

Genus BABYCURUBKarsch.

BABYCURUSBVJTNERI Krtrscli. Babycurus biitfneri, Karsch, Berl. ent. Z. xxx. p. 78 (1866); Pocock, Ann. Nat. Hist. (6) xvii. p. 429 (1896); Kraep. Das Tierr., Scorpiones, p. 62 (1899). LOC.Benito Xiver (cf. L. Bates), Cette-Camma, Gaboon; mouth of the Loengo River (H. Dugyan).

BABYCURUSKIRKI Poc. Rhoptrurus Irirki, Poc. Proc. Zool. SOC.1890, p. 137. LOC.W. Africa (Kirk.)j Lorn& Togoland (Hiller).

BABYCURUSJOHNSTONI POC. Babycurw biittneri, POC.Proc. Bool. 8oc. 1890, p. 138 (not of Karsch). Babycurus johtwtonii, POC.Ann. Nat. Hist. (8) xvii. pp. 429-430 (1896). LOG.Rio del Rey (U.H. Jo7~nston).

Genus ISOMETRVSIIeinpr. S: Ehrenb.

ISOYETRIJSEUROPXUS (Linn.) I. (= I. rnnculatus De Geer, and almost all modern authors.) LOC.W. Africa (Dr. Kirk) ; Sierra Leone (Jcimes Foxcroft, E. E. Austen) ; Fanti ; Cette-Camma, Gaboon j Cameroons (J.Pitwock) j Angola (Dr. Welwitsch).

The evidence ndduced by Lbnnberg (Ann. Mag. Naf. Hist. (7) i. pp. 86-S7, IS%) in favoiir of his view that the Scorpion described by De Geer as L macu- kztus is f he 5'. eiwqmftsof the 10th ed. of the ' Bystelria ' nppenrs to me to be more cogent than the evidence in favour of the identity of Scorpio mz/irits and Scorpio ciustralls of the latter work. Yet Kraepeliq in his Tierreich' acoepte the two litst, but rejecte S. errropretfi, 536 ME. R. I. POCOCK ON SCORPIONS, PEDIPALPS, [NO\l. 14,

Genus CENTRU~US(Hempr. & Ehrenb.), Pet.

CENTRUEUSMARGARITATUS (Gerv.). Scorpio margnritatw, Gerv. Voy de la Bonite, i. p. 281, pl. i. figs. 13-17 (1841). Centrurus gambiensis, Karsch, MT. Munch. ent. Ver. iii. p. 123 (1879). Loc. Sierra Leone (Surg.-Capt. C'lements); recorded by Karsch from Gambia. Undoubtedly imported from America ; perhaps from Jamaica, where the species is common.

Genus PANDINUSThor.

PANDINUSIMPERATOR (C. Koch). Buthus imperator, C. Koch, Die Arachniclen, ix. p. 2, fig. 695 (1842). Panclinics imperntor, Kraep. Das Tierr., Scorpiones, p. 122 (1809). '(=roeseli, Simon ; nfricnnus, Linn., Thor., POC.,Kraep.) LOC.Slave Coast (Alvan Millson); Gold Coast (Cd~YirF. Frstiq); Fanti (Capt. Marryat) ; Ashanti ( W. B. Adams) 80 miles inland of Axim (H. G. Eames) ; Onitsha on the Niger (Sir J. Marshall) ; Asaba, 180 miles up the Niger (Dr. Crosse); Jebba and 110, on the Upper Niger (Dr. Christy, Capt. Rigby, aud Lieut. Abadie) ; Wegbe, Ho district, Togoland (W. Q. Innes) ; Fernando Po (Mrs. Burton and Cctpt. Birch).

PANDINUSJMPERATOB (C. Koch), subsp. GAMBIENSIS nov. ? Heterometrus imperator (C. Koch), Simon, Rev. Mag. Zoo]. 1872, p. 55 ; Beclier, Ann. SOC. ent. Belg. xxiv. pp. 137-140 (1880). Differing from the more southern form of P. imperntor typicus in having the carapace entirely covered with granules, the tcrga much more closely and thickly gramilar, and the ornamentation of the upper surface of the hand much more distinctly tubercular, the tubercles remaining for the most part distinct and not running together into a network of ridges. Pectinal teeth 16-18. Total length up to about 155 mm. LOC.Gambia (Sir A. Moloney ; type) ; also several specimens from Senegal (Lleine).

PANDINUSDICTATOR POC. Scorpio dictator, POC.Ann. Mag. Nat. Hist. (6) ii. p. 251 (1888) ; Kraep. MT. Mus. Hamb. xi. p. 70 (1894); id. Das Tierr., Scorpiones, p. 123 (1899). LOC.W. Africa (no history) ; Fernando Po (Capt. Biroh ; type) ; mouth of the Loango River (H. L. Duggan). l809.J AND SPIDERS FROM TBOPICAL WEST AFRICA. 831.

Genus OPISTHhCANTHZiS Pet. OPISTHACANTHUSLECOMTEI (Lucas). Ischnurw lecomtei, Lucas, in Thomson’s Arch. Ent. ii. p. 428 (1868). OpistJ~aoanthusduodecini-dentatus, Karsch, Berl. ent. %it, xxx. p. 79 (18a6). [See aleo Pocock, Ann. Mag. Nat. Hist. (6) xi;. p. 318 (1893) ; Kraep. Das Tierr., Scorpiones, p. 149 (L899).] LOG.Benito River (B. L. Bates). This species was not represented in the British Museum until Mr. Bates procured it. OYISTHACANTHIJSAFRICAANUS Simon. OpisthrmnntAus nfricanus, Simon, Bull. SOC.Zool. Fr. i. p. 281 (1876). Opkthacanthus septem-dentutus, Karsch, Z. Naturw. li. p. 372 (1878). [See also Pocock, Ann. Nat. Hist. (6) xii. p. 316 (1893) : Kraep. Das Tierr., Scorpiones, p. 149 (1899).] LOG.Guinea (in Keyseerling’s Coll.) ; Cette-Camms, Gaboon (Gerrnrd); Congo (A. Curror, ESP.,R.N., J. Pimock) ; Stanley Falls, Congo (Gerrrcrd).

Order PEDIPALPI. Genus TITANODAMONPOC. TITANODAMON BASSAMENSIS (Lucas). ? Phnlnngium medium, Herbst, Nat. ungeflugelt. Ins. i. p. 77, pl. iv. fig. 1 (1797). YJ8ynu.s bassamensis, Lucas, Arch. Ent. ii. p. 434 (1855). Phynus kochii and P. granulosug, Butl. Arm. Nat. Hist. (4) xii. pp. 120-122 (1873). l’hrynus suvutieri, Rochebrune, Bull. SOC.Philom. (’7) viii. p. 28 (1884). Titanodamon ba~samensis,POC. Ann. Mag. Eat. Hist. (6) xiv. p. 290 (1894). Damon mediue (Herbst), Kraepelin, Das. Tierr., Scorpiones, p. 238 (1899). hc. Sierra Leone (H. C. Hart, Surg.-Capt. Clements, C. M. Mitford, &c.) ; Dixcove in Ashanti (B. Frend) ; Cape Coast Ctcetle (J. P. Brown); Asaba, 180 miles up Niger (Dr. Cross#); Gold Coast; Wegbe in the Ho district, Togoland ( W. ff. Innes); Jebbn on the Upper Niger (Dr. Christy) ; Cameroons (Cctpt. Burton).

TITANODAMONJOHNSTONI POC. Titanodamon johnstoni, POC.Ann. Mag. Nat. Hist. (6) xiv. p. 291, pl. viii. figs. 2-2 b. 838 ME, B. I. POCOCK ON SL'ORPIOXB, PEIDIPALPS, [R'OV. 14, Danion tnedius johnstoni, Kraep. Dns Tierr., Scorpiones, p. 239 (1899). LOG.FernandoPo( qiyors COW.); Old Calabm (J.V. Cockburn and Misa Kingsley); Cameroons (H. H. 4 J. M. E. Johwton, Miss Kingsky); Rio del Rey (H. H. Johnston); Benito River (ff. L. Bates).

TITANODAYON TIBIALIS (SilnOtl). ? Phrynus tibialis, Simon, Bull. SOC.Zool. Fr. i. pp. 12-1 5 (1876). Titunodainon tiLialis, POC.Ann. Mag. Nat. Hitit. (6) xiv. p. 291 (1894). Dainon ntediua tibiqlis, Kritep. Das Tierr., Scorpiones, p. 239 (1899). LOG.Congo (A. Curror, Esy., EN.) ; West, Africn (Mr. Dulton).

Order ARANEX.

MYGALOMOBPHB. Fnrnily THERAraoaIDx. Subfamily THERAPHOSINX (=Tliei*apliosin.re,sensu stricto, Pocock, P. Z. S. 189i, p. 772.)

Genus SCODEABecker.

SCODRA CALCEAT-4 (Fabr.), Aranea calceata, Fabr. Ent. Byst,. ii. p. 427. Scodm uussweri, Becker, CS. SOC. Ent. Belg. 18'79, p. cxlii ; ibid. 1881, pl. ii. fig. 1. Stroniatopelma alicapillatum, Karsch, Bed. ent. Zeit. 1881, p. 218. Swdra ccilcenta, Pocock, Proc. Zool. SOC. 1897, pp. 755 & it56 (semu stricto). LOG.Accra (G. A. Hiylett); Cameroons (Capt. Burton) ; Ashanti; Afram plains. SCODEAGRISEIPES.POC. Scoclra gr&e+e.s, POC.Proc. ZOOI.Soc. 1897, p. i56, PI. rIiii. figs. 7-7 a. LOG.Sierra Leone (C. Wilson and C. 211. Mifford).

SCODBABRACHTPOUA Poc. ~sodrubrackypoda, POC.IOC. cit. p. i5i, pl. sliii. figs. 8-8 a. hc. Asaba on the Niger (07.. Crosse); Cape Palmas (Akan Millso1r). 1899.2 AND SPIDIRS FUOM TROPICAL WEST APKICS. 830

SCOUUAFUMIGATA, sp. n. 9. Colour. Carapace and upperside of mandible clothed with bright, almost mustard-yellow or greyish-yellow hairs ; upperside of femora and patellre of palpi and anterior legs olive-yellow; tibiae with a basal greyish-white patch and two short median lines ; protarsi and tarsi with median black patch; setie on the limbs greyish ; the upperside of the fringes bordering the leg-segments greyish brown ; upperside of abdomen greyish brown, with symmetrically disposed darker patches and bars ; the entire: underside ?f Zep and plpi, emepting the scopulm, of abdomen, wmz, and sternuna deep sooty-black j circumoral hairs crimson. Thoracic: fovea deep and subcirculnr ; carapace distinctly shorter than patella and tibia of 1st leg, and stout.kr by oue third of the protarsus than the tibia aud protarsus of this limb ; less also thau patella and tibia and than tarsus and protarsus of 4th. Legs 4, 1, 2, 3 in length ; patella and tibia of 4th little greater than of lst, 4th leg about three times as loug as the carapace; segments of legs thickly fringed. Mewurements in millimetres ( Q type).-Total length 44 ; 1engt.h of carapace 21, width 18 ; length of 1st leg 60 (patella + tibia Y3), 2nd leg 58 (patella + tibia 22), 3rd leg 53 (patella + tibia 9), 4th leg 64 (patella + tibia 23.5, protarsus 16). LOG.Benito River (G. L. Bates). The distinguishing characters of the feninles of t,heknoyn species of the genus Scotlra may be re-stated ns follows :- a. Carapace longer t.han patella + tibia of 1st leg and us long as tibia + protarsus of this limb ...... brachypoda POC. b. Uarapnce shorter than patella + tibia of 1st leg and shorter than the protarsus and tarsus of t.his liiub by one tbird of the prot,arsus. a'. 4th leg longer than 1st and about three times a8 long as the caraprtce; patella + tibia of fourth a little longer than those of lse; tibiz of legs Booty blackbslow fiouigata sp. n. 6'. 4th leg shorLcr than lrrt and much less t,hwn three times a8 long aa carapace ; putella + tibia of 4th iuoch less than those of 1st ; tibim not. sooty block below. a*. Lower and inner surface of femora of palp and of Arst pair of legs greyish or yellowish brown ; set= yellowish brown ...... grise@es Poc. bz. Lower side of femora of all the legs and inner surface of femora of palp and 1st and 2nd legs sooty black ; bristles un legs foxy red ...... caleentu. Genus HETEROSCODBA,no'v. This new genus and Scodm, its nearest ally, may be diagnosed as follows :- a. Fourth leg thinner t,han flrst ; the tibia of 4t,h thinner than patella nnd femur, the beighl and width of the tibia about one quarter the length ...... 8rodra Becker. h. Fourth leg thicker and stronger t.han 1st; t.he t,ibia thick, its height a libtle exce,:ding the height of the patella, equal to that, of the femur, and about one half' the length of' the tibia ...... Hcteroscodra, nov. 840 MR. B. I. POCOCK ON SCOBPJONS, PEDIPALl%, [NOV. 14,

HETEROSCODRAMACULATA, sp. n. 9 . Colour. Carapace covered with white hairs round the margin and furnished with narrow-white stripes which radiate from the fovea to this white border, the intermediste area clothed with hairs of an olive-green tint, forming two indefinite longitudinal bands starting with a deep patch of the same tint on each side of the ocular tubercle ; upperside of abdomen olive-green, uiottled with white ; mandibles black, covered with ashy-grey hairs; legs deep red, covered with ashy-grey hairs ; the naked lines on femora, patella, and tibise very distinct ; hairs on femora and tibia whiter than on the rest of the limbs, the distal extremities of which are clothed with foxy-red bristles, with a greenish underclothing ; upperside of tibia with two submedian white lines or spots, on the proximal side of which there are a pair of blackish bands ; a lnrge black spot on the upperside of the tarsus and on the basal half of the protarsus, as in Scodrn j the extremities of the segments with a narrow rim of white hairs ; lower side of trunk and limbs uniform ashy grey, the scopuls bluish green. Carapace longer than wide, the width almost equal to the length from the anterior median eye to the posterior border. Eyes as in Scodra ; the length of carapace a little less than patella, tibia, and tarsus of palp, a little excelling patella snd tibia of 1st leg, less than those of 4th leg, equal to tibia, protarsus, and hslf the tarsus of the 3rd, less than protarsus and tarsus of 4th, greater thau those of 1st; the width a trifle less thau patells and tibia of lst, and just exceeding those of 2nd leg. Legs 4,1,2, 3, without spines, scopulate as in Scodra ; protarsus of 4th longer than tibia of 4th, and as long a8 protarsus and half the tarsus of the 1st. Meusurements in mi1limetrea.-Total length 30 j length of carapace 16, width 14.5; length of 1st leg 42, of 4th 48, patella and tibia of 1st 15, of 4th 18 ; length of tibilt of 4th 105, height 4.5. Lac. W. Africa (F. W. Marshnll). A single female specimen of this interesting new spider was received from this Society, to which it had been presented alive by Mr. F. W. Marshall. This epecimen lived for some months in the Inaqct-house at the 8ociety’s Gardens, and was referred to by Mr. A. Thomson as Scodra caleeatn in his report upon the Insect- house for 1897 (see P. 8.8. 1898, p. 81). Though very closely resembling Scodra culceata in colour, the spider was found, upon closer examination after death, to be the representative of quite a distinct species.

Genus SELENO~YRUSPOC.

SELENOGYBUSUBRULEUS POC. Selenogyrzcs wruleus, POC.Proc. Zool. SOC.1897, p. 768. Loa. bier- Leone (Xiirg.-Cupt. Clevents and E. E. Austen). 1899.J AND SPIDERS PILOX TROPIOAL WEST AFUICA. 841

SDLENOQYEUSAURDUI POC. Selenogyrucr aureus, POC.op. cit. p. 768, pl. xli. figs. 2-2n. LOG.Sierra Leone (no further history). Hapalopiis africanus Simon, from Assinie (Ann. 800. Ent. Fr. 1887, p. 275), perhaps belongs to the genus SeZenog?yrus (see Pocock, P. R. 8. 1897, p. 773,note).

Genus MIASCHISTOPUSPoc. MIASCHISTOPUSRAPIDUS POC. Mia8:hhtopus rapichs, Poc. Proc. zoo]. sot. 1897, p. 770, pl. xli. fig. 5. Loo. W. Africa (Keyserling Coll.).

Subfamily EUMENOPHORINZ, Pocock, Proc. Zool. SOC.1897, pp. 772 & 773. Genus EUMDNOPHORUEPoc. EUMENOPHORUSCLEYENTSI POC. Etmenophorus clementsii, POC.Proc. Zool. Soc. 1897, p. 766. LOC.Sierra Leone (Xurg.-Capt. Clements). Genus PHONEYUSAKarsch. P?honeyzisa, Karsch, Berl. ent. Zeit. 1884, p. 347 (type P. belan- cklua Karseh). Haryccxothmia, Simon, Actes SOC. Linn. Bord. 1889, p. 413 (type, H. antiZope Simon). The species of Spiders here and in my previous paper on the African fanna referred to Phonyusn have been identified aa belonging to that genus on the strength of M. Simon’s statement (Hist. Nat. AraignBes, i. p. 154, 1892) that the type species of Phoneyusa and Hwpaxotheria are congeneric. Nevertheless it is possible that the two fonns will prove generically separable. If Karsch’s description of €‘. belanclana ie reliable, that species differs from all the species which I refer to Phmyusa and which doubtless belong to Harpasotherin, in the strict sense of the word, in the following features :- a. Oarapace about one third longer than broad (34 : 224) ; sternum also about one third longer than broad (154 :104) ; tibia: of legs with inany apical spines, 8-9 on the 1st and 21144 on the 3rd and 4th ...... lelandasa Karsch. b. Length ofcarapace and sternum exceeding the respective widthsb less than one quarter of the length ; tibia: of legs in at most with a pair of apical spines ...... antilope sim. occidenlalis, Lus., luttneri, Karech (P spinal armature). 842 MR. E. 1. POCOCE ON SCOEPIONS, PEDIPALPS, [NOV. 14, But although there is thus a possibility of resuscitating Harpax- theria, it must be remembered that the apparently greater narrow- ness of the cephalothorax in P. belandma may be due to artificial shrinkage, and that too great stress should not be laid upon the tibia1 spine-armature, seeing that in the male of P. greyorii, which is apparently congeneric with P. antilope and P. occiclentalis, the tibia of the 1st leg is armed with 5 spines, the 2nd with 3, and the 3rd and the 4th with 2 each. Again, before Harpcixotheria be rescued from the world of syn- onyms, it will have to be ascertained that it is distinguishable from Karsch’s previously established genus Pe7inobiics, a poiut about which great doubt may be entertained.

PHONEYUSAOCCIDENTALIS (Lucas). Mygale oecid,entnZis, Lucas, Thonison’s Arch. Ent. ii. p. 380 (1858). 9. CoZour. Integument of carapace, mandibles, and legs uniform deep reddish brown, covered with deep olive-green hairs ; the distal segments of the pnlpi and of the first two pairs of legs much redder in the young ; femora, patellp, tibiae, and protarsi with a fringe of yellowish-pink hairs at their divtal extremities j abdomen greenish brown; the long bristles on the legs arid abdomen reddish. Carnpuce considerably longer than patella and tibia of 4th and 2nd legs, a little longer than those of lst, longer than tarsus and protarsus of 4th, much longer than those of lst, a little longer than patella, tibia, and tarsus of palp ; its width a little less than the area between the posterior emarginatiou and the ocular tubercle, much exceeding the 4th protarsus and just exceeding protarsus and tarsus of 1st leg, a little longer than femur of 4th leg. Legs 4,1,2,3 : 4th exceeding 1st by one fourth the length of its tarsuv ; tibis with a pair of distal spines below, protarsus of 1st with 1 median apical spine, of 2nd with 3*spines, of 3rd and 4th with about 9 spines in a transverse row. Measuremenis in mil1intetres.-Total length 62 ; length of carapace 32, width 26 ; length of palp 47, of 1st leg 78, of 2nd leg 69, of 3rd leg 63, of 4th leg 79, patella and tibia of 1st 30, of 4th 27, protarsus of 4th 21 (legs and palpi measured from base of femur). In younger femiiles the legs are much longer as compared with the carapace than in the adult. Loc. Benito River (ff. L. Butes). This species has not been hitherto recognized since it was first established. It certainly differs from P. belnndnna Karsch (Berl. ent. Zeit. 1884, p. 348), from Niam Niam in Central Africa, in having only a pair of spines at the apex of the tibiae, instead of a large number as in 3‘. bekindana ; that is to say, 9 on the 1st tibia, 8 on the 2nd, &c. In P. belundana also the sternum is nearly twice aS long as wide (15 : a), n hereas in P. occide?ztciz~8the length is only a little greater than the width (9 : 7.5). In the characterof the sternum P. occiden- fa& resembles P. bdttneri Karsch (Berl. ent. Zeit. 1886, p. 83), from 1899.1 AND SPIDERS FROM QEOPICAL WEST AFRICA. 843 Sibange Farm, Gaboon, with which it may prove to be identical. Karsch, however, says nothing about the colouring of P. biittrieri, nor about the spine armature of the legs. From P. (Harpa~otheria)nutilope Simon (Act. SOC.L. Bordeaux, xlii. p. 414, 1889), from Toinby in the Congo, P. occidental& also differs, judging by the leg-measnrements that Simon gives. For example, the 4th leg in the type of P. coitilope exceeds the 1st by about 10 mm. (63.2: 53.5), that is to say, by considerably more than the length of the tarsus of either limb, whereas in P. occidenta2is the difference does not amount to more than lialf the tarsus.

PHONEYUSABIDINTATA, sp. n. (Plate LVI. fig. 11.) d. Colour : a uniform dull greyish-brown clothing of hairs on the trunk and limbs, the long bristles on the legs and abdomen reddish grey, the integument beneath the hairs nearly black ; n narrow fringe of pale pinkish hairs at the distal end of the femora, patellae, tibia, and protarsi. Carrcpnce longer than wide, its length slightly exceeding that of patella, tibia, and tarsus of pdp, equal to protarsus and femur of 4th leg, considernbly less than patella and tibia of 4th leg, a little &waterthan patella and tibia of 3rd leg, its width just about equal to femur of 2nd. Legs long and slender (those of 1st pair absent), with two or three spines at the apes of tibise beneath, about 4 on apex of 2nd protarsus, and a row of about 8 on apex ilf 3rd and 4th protarsi; 4th leg exceeding the 2nd by about two thirds of its protarsus, patella and tibia of 2nd about equal to those of 4th, the tarsi about equal : tarsus of Jth, including claw, about equal to patella of 4th ; bristles upon legs and abdomen hooked at their distal ends. Tibin of palp about three times as long as broad ; bulb of palpal organ subcircular and furnished with two spines, the principal spine rather short, stout, not filiform, furnished externally with two keels which pnss on to the bulb, the smaller spine in front of the larger, much smaller than it and curved in the opposite direction. Mensurements in rni&metres.-Total lengeh 4 1: length of carapace 23, width 20.5 ; length of palp 35, of 2nd leg 74, of 3rd 65, of 4th 80, patella and tibm of 4th 27 ; protarsus of 4th 235, tibia of 4th 18, tarsus 9. LOC.Benito River (G. L. Bates). This species differs from all the known species of phoneyusa that are based upon male examples in the possession of a second spiniform process on the bulb of the palpal organ. These species are P. (Harpaxothevia) gracilipes Simon, from the Congo, and P. (Harpaxotheria) ectypa Simon, from Abyssinia (Act. SOC.L. Bordeaux, xlii. pp. 414-415) ; P. gregol-ii Pocock (Proc. Zool. SOC.1897, p. 760, pl. 43. fig. 6), from Masailand, and P. bettoni Pocock (Proc. Zool. SOC.1998, p. 503), from the area between Moinbasa and Uganda. 844 ME. R. I. POCOCK ON SCORPIONS, PEDIPALPS, [NOV. 14, Genus HYSTEBOCRATESSimon.

HYSTEROCRATESQIQAS POC. Hysterocrates gigas POC.loc. cit. p. 762. LOC.Cameroons (J. M. C. Johnston) ; Oil River (U.1% Johnston),

HYSTEROCEATES LATICEPS POC. Hysterocrates laticeps POC.loc. cit. p. 765, pl. xli. figs. 4-4 b. LOC.Old Calabar (MGs Kingsley). HYSTEEOCRATESCRASSIPES Poc. Hysteyocrcites crassipes, Poc. loc. cit. p. 764, pl. xli. fig. 4 c. LOG.Cameroons (U. H. Johnston).

HYSTEEOCEATESHERCULES, sp. n. 9. Colour. Integument black, covered with a thick coating of dark olive-brown hairs, shinipg with greyish silky sheen under reflected light. Carapace a little longer than patella and tibia,of 1st and of 4th leg, longer than tarsus and protarsus of 4th and than patella, tibia, and tarsus of palp ; its width equal to length from tubercle to posterior emargination and to femur and patella of 3rd leg, and to femur and one third of patella of 4th leg; a transverse depression in front of fovea, marking off a transversely oval tubercle. Mundiblea tubercular in front. Legs 4, 1, 2, 3 ; 4th exceeding 1st by nearly half its tarsus ; pntella and tibia of 4th not longer than of 1st ; 4th leg not in any sense thickened, femur more than three times as long as high (25 : 7%), thicker than patella, which is considerably thicker and higher than tibia j tibia slightly concave below and in the basal half above, more thau three times as loug aud wide as high (17.8 : 5.2) ; length of upperside of patella more than take its height or width. Measurements in mil1inietres.-Total length 74; length of carapace 34, width 30 ; length of palp 49, of 1st leg 81, 2nd leg 72, 3rd leg 67, 4th leg 88; patella and tibia of 1st 33, of 4th 32; tarsus and protarsus of 4th 31. LOC.Jebbn, Upper Niger (Lieut. Ahdie). Recognizable from the other known species of the genus by its large size, darker colouring, and by haviug the 4th leg unmodified, not thicker than the lst, and only exceeding it in length by less than half its tarsus. HYSTEROCRATESROBUSTUS, sp. n. Q . Colour olive-greenish brown, with reflexions of greyish-white pubescence. Carapace considerably longer than wide, the width equal to the distance between the ocular tubercle and the median emargintltion ; the length equal to that of patella and tibia of 4th leg, greater than those of 1st and greater than tibia and protarsus of 1st ; the width 1899.1 AND SPIDBRS FROM TROPIUU WEST AFEIUA. 845 slightly exceeding the length of the outer surface of the femur of the 4th leg. Ley 4, 1, 2, 3 ; 4th exceeding the 1st by rather more than the length of its tarsus ; patella and tibia of 4th a little longer than of 1st (24 :22) ; 4th leg strong, but the patella and tibia narrower than the femur ; width and height of patella about equal, but barely equal to half the length measured along the upperside; tibia lightly convex above, its width a little greater than one third of I its length and a little less than its height ; protarsus longer than tibia (15.5 : 13.5) ; femur of 4th leg very robust, its height exceeding one third of its length. Measuremen& in mil1imetres.-Total length 53 ; length of cnra- pace 24, width 20 ; length of 1st leg 55, of 2nd 49, of 3rd 4, of 4th 65 ; length of femur of 4th leg 18.5, height 7 ; length of the tibia 13.5, height 5. LOG.Benito River (G. L. Bates). The subjoined table will show how this species may be distinguished from the rest of the species of Hysterocrates known to me:- a. Width of tibia of posterior leg equal to that of femur ; tarsus of palp more tumid above at base ...... [email protected] POO. h. Width of tibia of posterior leg much less than width of femur ; tarsus riot tumid above at base. ul. Oarapace long ; area between ocular tubercle and posterior median emar ination exceeding the width ... gigus POC. b'. Carapace wider, ita wifth equal to the arm between the tubercle and the osterior emargination. a*. Femur of 4thleg &fender, its height- barely one third of its length. - a3. Hairy clothing of le s reddish brown; 4th leg thickened. the tibia afmost as thick as the femur. luticeps POC. bs. Hairy clothing deep olive-brown ; 4th leg unmodified, the tibia much thinner than the femur ...... hermibs, sp. n. b*. Femur of 4th leg stout,, its height considerably exceeding one third of its length ...... robushcs, sp. n. Family BABYCHHLIDB. Genus CYPHONISIASimon. CYPHONISIAOBESA Simon. CyphonisM obesa, Simon, Act. Boc. L. Bordeaux, xlii. p. 409 (1889). LOG.Benito River (G. L. Bates). Two immature specimens probably referable to this species, which was recorded from the Rio Quiliou (Congo). Family DIPLURID&. GI-enus HETEEOTHIIILEKarsch. Heterothele, Karsch, SB. nat. Fr. Berlin, 1879, p. 64 ; Pocock, Proc. 2001. SOC. 1897, p. 736. PEOcr. ma,Soa.-1899, No. LV. 55 846 MR. E. I. POCOCK ON SCORPIONS, PEDIPALPS, [Nov. 14,

? HETEEOTHELEBABONENBII (Lucae). Mygale yabonensk, Lucas, Arch. Ent. 1858, p. 382. ? Diplurn longipnlpis,Karsch, Zeits. Naturwiss. (3) iv. p. 564. Two €emale specimens of a species referable to Heterothele, obtained by G. L. Bates on the Benito River, are regarded as probably identical with Mygale gabonensis, Lucas. Family CTENIZIDB. Genus ACANTHODON,GuBr. ACANTHODONANGU~TIUEPS, sp. n. 0. Very closely resembling, both in size and spine-armature, hc., A. lacustris, Pocock (Roc. Zool. SOC. 1897, p. 731, pl. xli. fig. 7), from Lake Tanganyika. The two species may be separated as follows :- a. Oars ace broader ae compared with its length (7%: 9), wicfth equal to the distance between the postesor border and the anterior border of the ocular cluster ; the two ocular tubercles higher, the area behind the posterior considerably more depressed owing to the greater elevation of the ce halic prominence; sternum wider as compared wit1 its length (48 : 6) ; lega longer, length of 1st 18.8, or 4th 23, tarsus and rotarsus of 4th 8 ; palp from base of femur 16, its tLee distal segments 9% ...... lamatris POC. b. Uarapacenarrower, its width as compared with its length being 68 : 9, width equal to the length between the posterior border and posterior edge of the ocular tubercle ; ocular tubercle lower, but little elevated, area behind the ocular cluster but little depressed; sternum narrower, very distinctly narrower than lon (4:5) ; legs shorter, length of 1st 17, of 4th 2{ tarsus and protnrsus of 4th 7 ; of palp 15, its three distal Beginenta 9 ...... angzlsticeps, spa n. Slight differences also are observable in the eyes, the posterior medians being a little closer together, more directly behind the medians and a little farther from the laterals. l have very little confidence, however, in the taxonomic value of slight apparent differences in the size and relative positions of these organs. LOC.Benito River (G.L. Bates). A single female example was procured.

AEACHNOMORPHB. Family DINOPIDB. Genus DINOPI~Macleay. DINOPISBUBO Brit. Capello. Dinopis bubo, Brit. Capello, Mem. Ac. Sci. Lisboa, (3) iv. pt. 1 : Arachnideos etc. p. 16, pl. ii. fig. 3. 1899.1 AND SPIDERS FROM TEOPICAG WEST AFRIUA. 847 LOG.Gaboon. Recorded from the River Quilo. DINOPISASPEOTANS, sp. n. 9 . Colour. Carapace yellowish brown, covered with whitish hairs, reddish hairs around the eyes ; mandibles pale yellow, sparsely speckled with black, scantily clothed with yellow hairs ; mouth- parts yellow ; sternum yellow, blackish at the sides, mottled with yellow and white hairs ; palpi yellowish brown, mottled with black : legs brownish, femora infuscate distally, the anterior pairs also infuscate beneath and spotted with black at the base of the spines above, patellse fuscous, tibia? distally infuscate, 3rd and 4th pairs with superior distal spot, protarsi yellowish indistinctly speckled with black ; abdomen clothed laterally and below with whitish hairs, with four white spots and two parallel lines on the area between the epigastric fold and the cribellum ; the upper surface covered with darker hairs, with a lorn crest of hairs passing transversely in front of the prominences and curving backward on the sides. Carapace nearly twice as long as wide, its length equal to that of tibia of 3rd leg, a little more than one third that of the 1st protarsus, less than half (about two fifths) the length of the 1st femur; cephalic area a little wider in front than behind ; superciliary ridges evenly rounded, not prominent or produced into horn or tooth ; the posterior median eyes close together, their radius exceeding the height of the clypeus, anterior medians about two diameters apart. Palpi very siightly longer than carapace. Legs 1, 2,3,4 ; 1Rt twice as long as 3rd, 1st exceeding the 2nd by half its protarsus and its tarsus ;3rd reaching to apex of tibia of 2nd when both are extended ; on the femora of the 1st and 3rd the anterior spines are supported on tubercles which also support small tufts of hair. Abdomen rather more than twice as long as wide, posteriorly pointed and conipressed, widest just in froat of the middle, where it rises into a pair of prominences, from which it narrows anteriorly and posteriorly. Meaezcrements in mil1irnetres.-Total length 19 ; length of carapace 7, width 4; length of abdomen 12, width 5 ; length of 1~t leg 58, of 2nd 45, of 3rd 29, of 4th 28. LOG.Benito Xiver (G. L. Bates) ; a single Q example. This new species may be at once recognized from D. anchieta (Llinopis anchietce, Brit. Cap. Toe. n't. p. 15, pl. ii. figs. 2-2 c), from Rio Quilo, Angola, by the absence of triangular superciliary crests, the greater length of legs, flatter and longer carapace, proximity between posterior median eyes, &c. In D. anchietcz the eyeu are concealed by the crests when viewed from above and are nearly a radius apart; the thoracic portion of the carapace is as wide as long, and although the length of tk trunk is about the same as in D. aspectam, the anterior leg measures only 41 mm. instead of 58. 55' 848 MR. R. I. PO~Kox RCORFIONS, PEDIPALPS, [Nov. 14,

Family AEQIOPIDB. Genus NEPHILALeach. NEPHILAFEMOEALIS (Lucas). Epeira femoralis, Lucas, Thornson’s Arch. Ent. ii. p. 38, pl. xii. fig. 4 (1858). fiephila vittata, Keys. SB. Isis, Dresden, 1863, p. 142, pl. ii. fig. 2. LOC.Sierra Leone (Surg.- Capt. Clemenk) ; Liberia (Keyserling Coll., type of A’. vittata); Gold Coast (T. E. Bowdich); Ashanti (IT.H. Adam) ; Asaba, 150 miles up Niger (Dr. Crosse) : Accra and Wassan (B. A. HGlett) ; Cameroons (Capt. Burton, Miss Kingsley); Old Calabar (Miss Kingsky, H. A. Spencer); Benito River (B. L. Bates) ; Stanley Falls, Congo ; Angola.

NEPHILAWUASI Simon. +eira chrysoguater, Lucas, Thomson’s Arch. Ent. ii. p. 35 (not c7i~*ysogaster,Leach). Nephila lucasi, Simon, Ann. SOC.Ent. France, 1887, p. 270. LOC.Sierra Leone (Sur .-Capt. Clements);Ashanti(Mr. Macarthy); Accra (B. A. Higlett) ; 8ameroons (Capt. Burton, Miss Kingsley) ; Benito River (G. L. Bates); Wathen on the Congo (Miss Macomnick). NEPHILAPILIPES (Lucas). Epeira pilipes, Luma, Thomson’s Arch. Ent. ii. p. 40, pl. xiii. fig. 7. LOC.Fantee, Accra (B. A. HigZett) ; Benito River (B. L. Bates). NEPHILABEAQANTINA Brit. Capello. Nephila bragantinn, Brit. Capello, Mem. Ac. Sci. Lisboa, (3) iv. pt. 1: Descripciio de algunas especiee de . . . . Arachnideos, p. 11, pl. ii. fig. 4. LOC.Braganza, interior of Angola (Brit. Cap.). The British Museum has no examples of this species from W. Africa, but has received specimens of apparently the same form from Kinyamholo, Lake Tanganyika (A. Nutt). N. bragantinn differs from N. keyserlingii Blackw. (= hyrnencea Gerst.) in having the pdpi black, the legs black with the exception of a yellow band at the tip of the tibia and base of protarsus of 1st and 2nd legs, and no dark band in the middle of the sternum.

NEPHILAUEUENTATA (Fabr.). Araneus cruentatus, Fabr. Ent. Syst. ii. p. 427 (1793). Nephila genualis, Gerst. Von der Decken’s Iteisen, iii. 2, p. 502 (1873). Loc. Sierra Leone (D.El. Morgan, Sug.-Capt. Clememts); Ashanti (Mr. Macarthy) ; Onitsha on the Niger (Sir J. Mar8hull) ; Old Calabr (If,A. Spencer); Congo (J.Pinnock); Stanley Falls, Congo. 1899.1 AND smmaFROM TROPIUAL WEIBTAFRICA. 849 The following W. African species of this genus is unknown to me :- NIIPHILA CONSTEICTA Karsch. Ncphila constricta,Karsch, Zeits. gesammt. Naturwiss. lii. p. 834, fig. 4 (1879). Lac. Loango coast. Genus ASIGIOPIIAnd. et Sav.

ARGIOPBPLAVIPALPIS (Lucas). h’peiru jlauipalpis, Lucas, Thomson’s Arch. Ent. ii. p. 49 (1858). Argiope fivipa@s, Brit. Capello, Jorn. Sci. Lisboa, i. p. 83, . pl. ii. fig. 2 (1866). Argiope pechueli, Karsch, Zeits. gesammt. Naturwiss. lii. p. 340 (1879). LOG.Sierra Leone (8ury.-Capt. Clentents) j Old Calabar (Miss Kingsley) ; Cameroons (Capt. Burton and Sir Harr?y Johnston) ; Beuito giver (G.L. Bates). The legs of this species vary in tint: sometimes bhey are noticeably striped black aud yellow, as in the form to which Karsch gave the name pechueli, and sometimes of a very much darker, more uniform hue as in the typical A.JEavipa7pk. The two forms occur at the game locality, and gradations in the coloration of the legs are traceable. I therefore regard A.pec7rueli as a synonym of A.fivipalpk.

ASGIOPIINIGSOVIITATA Thorell. Argiope nigrovittata, Thorell, (Efv. Vet.-Mad. Porhandl. p. 300 (1860). Argiope caudata, Blackwall, Ann. Mag. Nat. Hist. (3) xvi. p. 346 (1865). Argiope zairiensis, Brit. Capello, Jorn. Sci. Lisboa, i. p. 82, pl. ii. fig. 1. drgiope suavissima, Gerstiicker, Von der Decken’s Reisen, iii. 2, p. 495, pl. xviii. fig. 10 (1873). Loo. Congo ; Benguela (J.J. Monteiro). Cfenus ARANEUSLinn. AEANIIKJSEUFIPALPIS (Lucas). Epeira mfiulpis, Lucas, Thomson’s Arch. Ent. ii. p. 422 (1858). Epeira senziunnulnta, Karsch, %its. gesammt. Naturwiss. lii. y. 334 (1879) ( 9 ) j Simon, Bull. SOC.21001. France, ix. p. 14, pl. ii. hgs. 7-8 (1884) ( 6,9 ). ? Epeira penioillipes, Karsch, loc. cit. p. 836 ( 6 ). LOC.Sierra Leone (Xurg.-Capt.Clements) ; Accra (a.A. HGZett); Cameroons (Ccvpt. Burton); Benito River (a. L. Bates). This species also occurs on the eastern side of the continent of Africa. It appears to me that the descriptions given of E. rujipalpis and 850 MR. B. I. POCOCK ON SCORPIONS, PMDIPALPS, [&OV. 14, fi. seinialtnulata were based upon examples belonging to the same species. Karsch's figure of the palp of E. penicillipes resembles that organ in mah examples of E. ru$paZpis sent by Mr. Bates. AEANEUSTHEYS (Walck.). Epeira the!&, Walckenaer, Ins. Apt. ii. p. 53. Epeira moreli, Vinson, Araneides de Madagascar, etc. p. 166, pl. iv. fig. 4 (1863). Epeira eclipsis, Marx, Proc. U. S. Nat. Mus. xvi. p. 590, pl. lxx. figs. 6 a-6 b. This widely distributed tropical species was recorded from the Congo as Epeira eclipsis by Marx. The British Museum has no W. African representatives of it, but has received it in some abundance from Mashonaland (G.A. K. 2liarshaZl).

ARANEUSPACHANUS POC. Araneus pachanus, POC.Ann. Mag. Nat. Hist. (7) vii. p. 447, pl. xiii. fig. 9 (1898). LOC.Benito River (a.L. Bates). Several 9 examples. Previousiy recorded from Karagesi (Emin Pasha) and Ruwenzori (Scott E22iot). This species presents a striking likeness, both in colour, form, and structural details, to the Oriental species that has been described BS A. decens, rump$, vufofeemoralis, &c. But the shape of the vulva seems to separate the African species, the scape being longer and the basal portion much more prominent beneath it.

AnANEus HBMATOUNMMIS, sp. n. (Plate LVI. figs. 8-8 c.) Colour. Carapace either a uniform blackish brown or reddish brown above,.pmsing into black towards the margins ; upperside of abdomen either uniformly blackish brown or ornamented with yellow on the anterior half-the yellow taking the form sometimes of a median field pointed in front and behind, broadest across the shoulder region, breaking up into spots all round its margin and interrupted along the middle line by an irregular black stripe ; sometimes OE a sharply defined median stripe, broadest in front and constricted in the middle and at the posterior end ; sometimes of a transverse recurved etripe behind the large sigilla, the extrkmities of which extend backward as an indistinct yellow stripe on each side circumscribing a median jet-black area which occupies the position of the folium. (In a young specimen the abdomen is testaceous, with jet-black folium and bright yellow median constricted stripe in front of it.) Sides and lower surface of abdomen black. Sternum, labium, maxills, and mandibles deep blackish brown; legs with cox% and femora uniformly blackish brown! patellse, tibia, and protarsi darker or lighter red, with their distal ends black, and frequently a median band ; tarsi black, with narrow red basal band. Carapace moderately elevated ; fovea subcircular, with longitu- 1899.1 AND SPIDMM PBOY TWPICAL WEBT AFRICA. 851 dins1 groove in front of it, somewhat as in Larinia; length o carapace exceeding that of 1st tibia, equal to that of 4th protarsus and tarsus and to patella, tibia, and half the protarsus of 3rd leg; width equal to length of 4th tibia and to 4th protarsus, less than that of 1st or 2nd tibia; median ocular tubercle prominent. Ocular quadrangle much wider in front ; the posterior eyes much smaller than the anterior and nearly a diameter apart j anterior medians about a diameter apart, more than a diameter from the posterior medians and from the edge of the clypeus ; eyes of anterior line when viewed from the front procurved, the upper edge of the laterals scarcely on a level with the centre of the medians ; lateral eyes not quite in contact. Mandibles moderately geniculate at the base ; fang-groove armed with 4 anterior and 3 posterior teeth. Legs armed with many strong spines ; spines black at the base and apex, reddish in the median portion. Abdomen heart-shaped, longer than broad, widely rounded in front, without shoulder-points, narrowly ovate behind, not sur- pasting spinners. Vulva with scape long and bent at right angles, its base as wide as the basal vertical, vestibular portion, which, when viewed from the side, is seen to send out a bitubercular prominence beneath the scape, from which it is separated by a narrow notch. Measurements in mil&netres.-Total length 22 ; length of carapace 9, width 7.2 ;length of abdomen 14, width 12 ;length of 1st leg 34, of 2nd 33, of 3rd 19, of 4th 30 ; patella and tibia of 1st leg 13, of 4th 11 ; tibia of 1st leg 8, of 4th 7. LOG.Benito River (a. L. Bates). Mr. Bates procured many specimens of this handsome species. ARANEUSEIRESIPRONS POC. AranezLs eresifrolzs, POC.Roc. Zool. 6oc. 1898, p. 509, pl. xli. figs. 3-3 b. LOG.Cameroons (H. H. Johnston). Previously recorded from Karagesi (Emin Pasl~~),Likipia (J. W. Gregory), Taru (Sfeuart Betton), and Mombasa (D. J. Wilson). This species is certainly nearly allied to A. strupifer Simon (Ann. SOC.Ent. France, 1885, p. 368), which, according to Simon, occurs both in Senegal and Cape Colony. In the specimens from the Cameroons the abdomen is frequently variegated above with sooty black, a type of coloration not observed in the East-African forms known to me.

ARANTCUSTYLOSUAPUS, sp. n. (Plate Lv. figs. 3-3 b.) Colour. Carapace testaceous ;cephalic region with blackish spots ; legs testaceous, with small black spots ; patellae, tibia, protarsi, and tarsi banded with black ; mandibles testaceous j sternum, labium, and maxillae inf uscate j abdomen olive-yellow, variegated with blackish, marked above with five tmsverse black lines, the anterior of which is the st,rongest and runs from one shoulder-point to the 852 ME. 11. I. POOO~KON SCORPIONS, PEDIPALPS, [Nov. 14, other, with its convexity backwards ; inferior and lateral surfaces darker than the superior, the inferior with a pair of yellowish posteriorly dilated bands running from the stigmata posteriorly towards the spinning-matumillae. Fges of anterior and posterior line recurved when viewed from above, those of anterior line also strongly recurved when viewed from the front, the laterals standing much higher than the medians ; median quadrangle a litt.le wider than long, nearly twice as wide in front as behind, the posterior separated by a very narrow space which barely equals half their radius, the anterior separated by a space which is equal to their diameter. Legs armed with numerous spines arranged iu more or less definite rows ; there being, for example, 6-6 on the lower side of the tibiEe and protarsi of the 1st and 2nd legs j spines black at base, pale distally. Abdomen lo er than wide, widely rounded in front, oval behind ; with distinct lack-tipped shoulder-processes, covered with short white hairs, intermixedY with particoloured bristles. Basal part of vulva very stout when protruded, and consistiug of a right and left, outer sheath, the halves of which do not meet in the middle line. Viewed from below, the two halves of the outer sheath show as a right and left rim surrounding a central pale portion, upon which rests the short but broad scape, the anterior part of which divides the rim of the right side from that of the left. The posterior end of the scape does not project SO far posteriorly as the posterior border of the subjacent portion of the vulva. Measurements in ntil1imetres.-Total length 12 ; length of carapace 5, of abdomen 9, width of latter 7.5. LOC.Benito River (a.L. Bates). In the form of its vulva and other features this species is evidently related to A. medicolu Simon, from Arabia and (according to Pavesi) from Somaliland, to A. mossambicensis, Pavesi, from Mozambique, to A. similis and striata, Bosenberg and Lenz, from Quilimane, and to A. cyrtoscctpus Poc., from the Transvaal.

ARAN~IJSRHINURUS, sp. n. (Plate LVI. figs. 9, 9 a). Colour. Carapace olive-brown, clothed with yellow hairs ; mandibles, palpi, and legs almost the same colour as the carapace ; distal end of femora, tibiae, protarsi, and the tarsi infuscate, especially on the 3rd and 4th legs ; upperside of abdomen chalky yellow, with dark sigilla spots and fine black line between them, also with a fine reticulated ornamentation of lines between the low pigment-spots; the tail and the lower side of abdomen black, with symmetrical bright yellow spots on each side of the spinners. Carapace shorter than tibia 1, about as long as patella and tibia 4 ; cephalic region moderately elevated, tlattish above longitudinally ; ocular quadrangle almost square, scarcely narrowed in front. The eges large and subequal, posterior medians about a 1899.1 AND SPIDEIBBFROM TEOPIUAG W~TAFEIOA. 853 diameter apart, anterior medians a little more than a diameter apart, distance between anterior and posterior medians less than a diameter ; anterior line of eyes straight or nearly so ; anterior medians less than their diameter from the edge of the clypeus. Legs longish, scantily spined, but furnished with long close-set bristles ; the spines setiform ; 1st leg much longer than 4th. Abdomen flat above, heart-shaped, with rounded sntero-lateral angles and anterior border, the posterior apex prolonged into a longish, stout ‘‘ tail,” which about equals the carapace in length ; spinners in the middle of the heart-shaped basal portion of the abdomen. Vulva with its basal vestibular portion not expanded, either laterally or posteriorly, at the base of the scape, which is long, slender, and slightfly curled at the tip. Meusurements in millimetwa.-Total length 9 ; length of carapace 2.5, of abdomen 6.5, of abdomen without tail 4, width of abdomen 35; length of 1st leg 12, of 4th 8. LOG.Benito River (B. L. Bates). A single adult female. This differsfrom all the Tropical African species of Araneus known to nie in having the ocular quadrangle approximately square, with the four eyes subequel, and the extremity of the abdomen produced into a longish caudal process. In both of these features it resembles the Burmese species A. thelurus (Thor.), but may be at once separated from it at least by the much greater length of the scape of the vulva.

Genus CYETOPHOEASim.

CPETOPEORAUITBICOLA (Forsk.). Loc. Benito River (cf. L. Bates). Widely distributed throughout Tropical Africa and Asia.

CYETOPHOEAANQOLBNSIS (Brit. Capello). Epeira angolensie, Brit. Capello, J. Ac. Sci. Lisboa, i. p. 79, pl. ii. fig. 4 (1868). Epeira ohinchoxensis, Karsch, Zeits. gesammt. Naturwiss. lii. p. 333 (1879). Lao. Sierra Leone (Surg.-Capt. Clernents) ; Benito River (B. L. Bates). Recorded by Brito Capello from the Rio Quilo. Karsch’s specimens from Chinchoxo appear to me to be specifically identical with those that Capello described. M. Simon (Hist. Nat. Araignkes, i. p. 775, 1895) adds this species to the synonymy of C. citrioola. But this is undoubtedly an error, C. angolensis, according to my determination, being distinguishable by the posteriorly pointed abdomen and wide head, with the lateral eyes far apart from the medians. The vulva is furnished with a distinct process. Three well-marked colour-varieties of this speoies am met with. 85-4 YE. E. I. POCOUK ON BOOEPIONS, PIODIPALPS, [Nov. 14, In one the upperside of the abdomen is entirely black ; in a second it is also black with a bright yellow transverse stripe crossing it from shoulder to shoulder ; in the third, which seems to be less prevalent than the others, the whole of the upperside behind the anterior shoulders is yellow.

OPHORALARINIOIDES Simon. Cyrtophora larinwidea, Simon, A.m. 8oc. Ent. France, 1894, 1). 156. LOG.Benito River (G. L. Bates). Described from Ogow6.

CYRTOPHORAMARGARITATA, sp. 11. (Plate LV. figs. 4, 4 a.) Colour almost as in C. unicolor Dol., a tolerably uniform yellowish brown; the abdomen darker than the carapace and limbs, legs indistinctly variegated, sternum and lower side of abdomen blackish. Curpuce smooth as in C. citricola; eyes as in the latter species, but the laterals closer together. Legs as in C. citricola, but with tarsi and protarsi shorter ; tarsus and protarsus of lst, for example, being distinctly shorter than patella and tibia of 1st. Abdomen truncate in front, narrowly ovate behind, not lobate laterally, and only weakly bilobate posteriorly, aa in C. anyolends ; studded above with larger and smaller smooth circular bosses, very like those of C. unicolor, but much larger and less numerous. Vulva as in figure (Pl. LV. fig. 4 u). Measurements in millimetrea of type.-Total length 14 ;length of carapace 6.5, of abdomen 9, width of abdomen 75; length of 1st leg 19, its patella and tibia 7, protarsus and tarsus 6. Loc. Benito River (a. L. Bates). At once recognizable from C'. unicdor by the strong curvature of the posterior line of eyes, narrow interval between the lateral eyes, absence of tubercles on the carapace, large size of tubercles on the abdomen, &c. From the rest of the W.-African species known to me it may be at once recognized by the features mentioned in the subjoined table. The four W.-African tipecies of Cyrtophora known to me may be distinguished as follows :- a. Abdomen long and narrow, produced iu front into R longish process overhanging the hse of the car&- pace ...... krinwdea Sim. b. Abdomen truncate in front, broadest at its anterior end. al. Abdomen without diatinct shoulder prominences and no lateral prominences ; its upperside studded with large circular tubercles...... margaritata, sp. n. bl. Abdomen with distinct shoulder prominences and one or more prominences on each side ; upperside studded with smaller tubercles. 4%. Abdomen with one prominence on each aide - behind tho shoulder, posteriorly deeply bifid ; 1809.l ARD SPIDBBS FROM TROPICAL W1ST ATRICA.. 855

eyes of posterior line strongly recurved ; rulve without ecape ...... citricola Forsk. ZJ*. Abdomen with two rominences on each side behind the shoder prominence, apex of abdomen not bifid; poaterior line of eyes much leas strongly recurved; vulva with distinct scape ...... a?cgo&mbB. Cap. Genus ARGPBOBPEIRAEmerton. AI~GYROEPEIRAUNGULATA (Karsch). Xeta ungulata, Karsch, Zeits. ges. Nsturwiss. lii. p. 834 (1879). LOC.Benito River (a.L. Bates). This species, recorded from the Loango coast, also occurs on the eastern side of the African continent. Genus CYOLOSAMenge. CPOLOSA INSULlNA (cost&). LOC.Cameroon Mountains, 4000 ft. (H. H. Joh~tow). Genus AOUSILASSim. ? AOUSILASAFRIOANUS Sim. ?Acusilns afi.icma5, Sim. Hist. Nat. AraignBes, i. p. 785 (1895). LOG.Benito River (a.L. Bates). The description of A. afrkanus is too brief to make sure of the correctness of the identification of the immsture female of a species of this genus which Mr. Bates procured. M. Simon’s example was obtained at Sierra Leone. ? Genus SALASSINASim. SALASSINAFORMOSA (Karsch). Cyclosa fomoea, Karsch, Zeits. ges. Naturwiss. lii. p. 835 (1879). LOC.Cameroon Mountains, 4000 ft. (H. H. Johnston). Recorded by Karsch from the Loango coast. According to M. Simon’s division of the Cycloseae, this species, if rightly determined, as I think is the case, falls apparently nearest to the genus Salassina, having the head short, the median eyes of large size and the quadrangle they form slightly narrowed in front. The tibiae of the legs, however, are not noticesbly incrassate, and there are no angular prominences on the fore part of the abdomen. The lesser recurvature of the eyes md the absence of a median series of spines from the lower side of the protarsus of the 4th leg seem to exclude the species from the allied genus AczCSilas. Genus CXROSTEISThorell. CBEOSTEISAEGOSTIOTUS, sp. n. (Plate LV. figs. 5-6 b.) Cobour. Carapace with thoracic portion reddish above, with s silvery white pstch at the sides ; cephslio portion blackish, with sn 856 MR. B. I. POUOCK ON BCORPIONB, PBDIPAIZS, [NOV. 14, oblong silvery patch of hair on the upperside between the ocular tubercle and the posterior vertical tubercles, and a silvery patch on the upperside of the lateral tubercles ; mandibles deep brown ; sternum bluish black, with central silvery patch; mnxillre and labium black; coxa of legs deep brown with bluish tint; tro- chantera and femora red ; remaining segments nearly black, with bluish lustre especially below; a white band at base of tibia and protarsi below, the tibia1 band broad only on the 1st leg, also a conspicuous white tarsal band on 4th leg; palp with femur red, the other segments blackish with white spots above. Abdomen blackish above and below, yellow at the sides anteriorly, the upperside with a vertically interrupted transverse silver band behind the anterior row of tubercles, and a series of silvery patches and lines forming a longitudinal median band extending over two thirds of the upper surface between the transverse silvery band and the posterior tubercle ; sides of upper surface with narrow transverse silvery stripes ; lower surface with a row of three silvery spots on each side, extending from the epigastric fold to the sides of the spinning-mammillae. Cephalic tubercles long, subspiniform, much longer than ocular tubercle. Length of carapace equal to width of bead, including ocultir tubercles, and as long as protarsus of 1st leg, shorter than protarsus and tarsus of 4th leg by at least half the length of the tarsus. Abdominal processes normal in number, as, for example, in C. mitralis, and all small and tuberculiform ; the bifid projection above the spinners rather prominept. Tibiae and protarsi of legs normally impressed above. Pulvu as in figure (PI. LV. fig. 5 a). Measuremen& in. millimetres.-Total length 14; length of carapace 6, of abdomen 10, width of abdomen 11. Loo. Benito River (B. L. Bates). Easily recognizable from the 8. and E. African species of the genus by the form of the vulva, and by colour, the almost complete absence of the white band at the base of the tibia on the 2nd, 3rd, and 4th legs being exceptional. CBROSTEISALBESCENB, sp. n. (Plate LVII. fig. 16.) Allied to the East-African C. nodulosa Pocock (P. 2.8. 1898, p. 514, pl. xli. fig. 7). Hairy clothing of head mesially white, laterally golden yellow ; upperside of abdomen covered with greyish-white hairs, diversified with black spots on the tubercles, sigilla, and elsewhere, and with narrow transverse black lines which laterally unite, circumscribing transversely elongate pentagonal areas ; lower side of abdomen black ; femora steel-blue ; upperside of patella, tibia, protarsus, and tarsus covered with silvery-white hairs, and varied with pale golden yellow ; extremity of protarsus of 1st and in a lesser degree of 2nd leg slightly infuscate ; 3rd and 4th legs more diversified than 1st and 2nd : legs banded below aa in C. lzOddO8a ; 1899.1 AND SPIDFIRB FROM TROPIUAL WIWT AFRICA. 057 tibia of all with basal white band ; protarsus of 1st and 3rd black, with basal band of 2nd and 4th white with black patch just beyond middle. Carapace with superior and lateral tubercles longer than in nodulosa. Abdomen (in type-specimen) not distended, its anterior portion low, nodular, not elevated. VuZva (as in figure P1. LVII. fig. 16) somewhat resembling that of dulosa,the anterior portion completely divided into a right and left half by a deep median groove j the chamber containing the two fossae smaller, more transversely oblong, with anterior rim less arched. Total length 15.5 mm., width of hend 7.5 mm. LOG.Benito River (G. L. Bat@).

CBROEITRISTURRIGER, sp. n. (Plate LVII. figs. 15,15 a.) Colour. Dorsal surface of carapace, legs, and abdomen a tolerably uniform greyish brown, covered with a coating of yellow and white hairs intermixed ; abdomen mottled with black spots and brownish patches and lines ; legs coloured as in C. nlbescens, but the distal spot on the protarsi reddish brown. Caraiiace with tubercles as in C. albeecens. Abdomen with its anterior portion elevated into a high, broad, subcylindrical prominence, the summit of which is about one third broader than long, with semicircularly rounded anterior tubercular border, and three large tubercles on the posterior border. Vulva as in figure (Pl. LVII. fig. 15 a). Measurements in miZ1imetres.-Total length 16 ; width of head 7; height of abdomen from vulva to summit of promineuce 13. LOC.Benito River (a. L. Bates); also young specimens of probably the same species from Sierra Leone (8urg.-Capt. CZements). Somewhat resembling C. petersi Kartrsch from Inhnmbane (Mon. Ak. Berlin, 1878, p. 324, pl. i. fig. 7), in the elevation of the anterior portion of the abdomen; but in C. petersii the column is narrower, with the summit rounded and not encircled with tubercles. The three species of the genus known from the Benito River may be diagnosed as follows :-

a. Femora of legs bright red ; tibioe of 2nd, 3rd, and 4th le s with scarcely a trace of basal white band; black undersife of abdomen ornaniented with three pairs of silvery spots ; upperside of abdomen velvety blark, furnished with silvery lines and tches ...... argosfictus. b. Femora steerblue ; tibk of 2nd, 3rd, and 4th legs with broad white basal band below ; lower side of abdomen uniformly black, upperside dirty yeuowirrh brown. ai. General colour of dorm1 surfam of body and legs white ; abdomen not elevated in front into a high thick column ...... _...... dbescens. hl. General aspect dirty yellowish brown ; abdomen elevated in front into a broad thick column ...... 2um-k~~. 858 MR. Re I. POCOUK ON SCORPIONS, PBDIPALPLI, mOV. 14, Genus CLADO~LEASim. CLADOMELEA LONGIPES (Cambr.). Cyrtarachne longipes, 0. P. Cambridge, P. Z. 5. 1877, p. 559, pl. lvi. fig. 1. Loc. San Salvador, Congo. Genus BASTERACANTHASund. GASTERACANTHACUETISPINA Gutkin. Basteracantha curvispinu, Gu6rin, Icon. RGg. Anim., Arachn. pl. ii. fig. 8, (1837). Gasteracantha walckenaerii, Lucas, Thomson’s Arch, Ent. ii. p. 425, pl. xii. fig. 7 (7 858). Basteracantha vaccula, Thorell, afv. Vet.-Akad. Forhandl. xvi. p. 301 (1859); id. Eug. Resa, 2001. Arachn. p. 12 (1868). Qasteraenntha retracta, Butler, Trans. Ent. SOC. 1873, p. 157, pl. iv. fig. 12. LOC.Benito River (a. L. Bates) ; Sierra Leone. GASTEEACANTHACONNATA Butl. Qusteracantha cmmta, Butler, Trans. Ent. SOC. 1873, p. 168. LOC.Old Calabar ((hay). Closely allied to the Ceylonese G. geminata Fabr., but with no near affinity to Q. connata 8imon (Hist. Nat. Araigdes, i. p. 847, 1894), which the author makes the type of section 0 of his subdivisions of the genus Basteracantha. GASTERAUANTHAFORMOSA Vins., subsp. NANA Butl. Gagterncantha nuna, Butl. Trans. Ent. SOC. 1873, p. 161, pl. iv. fig. 4. LOC.Congo (type, without further history). The type of this species is young. 3 B. importuna and Q. molesta, O.P. Cambr. (P. Z. 5. IS%),p. 286, pl. xxvi. fig. 1, and pl. xxvii. fig. 13), from W. Africa, are probably the adult forms. GASTEIRACANTHABATESI, sp. n. (Plate LVI. fig. 10.) ? Qasteracantha connuta, Simon, Ann. SOC. Ent. Fr. 1887, p. 266 (nec G. connnta, Butl.). Colour. Carapace black, with a large pale spot on each side of the median eyes ; mandible shining black or red; sternum and coxm fuscous, indistinctly variegated ; legs yellowish brown, indistinctly annulated, tips of tarsi and protarsi black ; upperside of abdomen yellow, with a large anterior median brown spot marked with a thin ;ellow stripe like an inverted T, and on each side a black spot divided by a narrow transverse yellow stripe ; the rest of the ilpper surface variegated brown and black. Cephalic area elevated behind, the middle of the elevation forming a pair of close-set tubercles. 1899.1 AND SPID~SFEOM TEOPIOAL WEST AFRIUA. 859 Legs short ; patella and tibia of 1st less than width of head. Abdomen without inferior tubercle; the anterior lateral spines minute and lying far back behind the middle of the scutum and just in front of the posterior laterals, from the supporting prominence of which, however, they are separated by a rounded notch ; the anterior border lying between these anterior spinules widely and nearly evenly convex, forming an almost completely semicircular arch defined by the normal 10 sigilla, of which the 4 medians form a straight transverse line ; posterior lateral spines erect, very short, but borne upon the summit of a thick cylindrical prominence, the axis of which cuts that of the spines at an obtuse angle; posterior spines resembling the posterio laterals, but their axis in a line with that of the prominence ad directed straight backward. Measurements in mi2limetres.-Total length , of abdomen along middle line 6; width just in front of anterior spinule 9.5; width from tip to tip of postero-lateral spine 8. Loc. Benito River ((7. L. Ba&). This species seems to form a new section of the genus Qastera- mntha.

BASTDEAUANTHA(BTEOUANTHA) ROQDESI 0. P. Cambr. Uasteracantha royersi, 0. P. Cambr. P. Z. 5. 1879, p. 292, fig. 23 (a). Basteracantkt (ATtrocnnthn) sernajlava, Simon, Ann. Soc. Ent. Fr. 1887, p. 268, pl. vi. fig. 2 ( 9 ). LOC.Sierra Leone. Described from Assinie by M. E. Simon, and from Coanza by 0.P. Cambridge. GASTERAUANTHA(ISOXIA) PDNIZOIDDS Simon. Isoxiapenizoides, Simon, Ann. SOC. Ent. Fr. 1887, p. 269, pl. vi. fig. 4. LOC.Benito River (B. L. Rates). Recorded by Simon from Assinie. The foregoing species of Gastwacantha may be tabulated as follows :- a. Lower side of abdomen with large conical t,ubercle; spines large. u1. Anterior and median spines subequal and in contact ...... connata. bl. Anterior and median spines unequal and widely separated. a%. Spines with strong recurvature ...... czwvispilta. ba. Spines not or scarcely recurved ...... formma. b. Lower side of abdomen without tubercle; spines absent or short. us. Spines present. a4. Anterior border of abdomen widely convex; anterior s ines lying far back close to and much smaller than tge median spines ; head tuberculate above ...... but&. b4. Anterior and lateral borders of abdomen cutting at right angles the anterior spines, as large as the medians and widely separated from them ; head not tuberoulate ...... mqy?. bS. Spines absent ; abdomen widely rounded behind ...... pazmda. 860 MR. R. I. POCOCK ON SCORPIOKS, PBDIPAItPS, [NOV. 14,

Genus ARANZCHRAButler. ARAN~ECHRACAMBRIDGBI Butl. Arancethra cambridgei, Butler, Tr. Ent. SOC.1873, p. 175, pl. iv. fig. 8. Arancethra unyari, Karsch, Zeits. gesammt. Naturw. li. p. 322, pl. ix. fig. l(1878). LOC.Fernand Vas River (Du Chaillu; type); Lagos (Capt. Elmes); W. Africa; Fernando Po (Mr. Kalbreyer); Accra (U. A. Hiy!ett) j Loango River (H.L. Duygalz) ; Benito River (cf. L. Babes).

ARANBTHRABWLBRI, sp. n. (Plate L.V. fig. 1.) Colour. Carapace, mandibles, mouth-parts, sternum, and coxae reddish yellow j patella, tibia, and tarsus of palp black ; legs black, the femur of 1st and part of 2nd reddish ; tibiae of lst, 2nd, and 4th pairs with a broad basal yellow ring, the ring incomplete below on the 1st j upper and lower side of abdomen and the tips of the lateral prominences black ; sigilla reddish. Carapace and legs a8 in A. cambridgei. Abdomen of much the same form also as in that species, about twice as wide as long; t,he median portion of the anterior ,border simply emarginate, with a low tuberculiform prominence on each side j lateral portion of carapace furnished with five blunt conical lobes, one on the anterior border just in front of the antero-lateral sigillum, the posterior process lower t.han the other8 ; the entire convex posterior border of the abdomen for a space which exceeds half the length of the abdomen without tubercles; abdominal integument punctured as well as striolate ; sigilla very large, the six on the middle of the back arranged in a circular form, the quadrangle formed by the anterior and posterior pairs only, about twice as long as wide. Menmrements in miZ1imetres.-Length of abdomen 7, width 16. LOC.Benito River (a. L. Bates). Thisinteresting new species, which is dedicated toDr. A. G.Butler, the describer of the genus Arancethra, may be recognized from the typical and hitherto only known species of the genus by the following characters :- (I. Carapace, sternum, legs, &c. a rich dark red or blaok colour ; underside of abdomen entirely black, a black band acroea the fore part of the abdomen above; margin of abdomen furnished with at least 8 strong sharp spinifom processes, the posterior not very far from the posterior middle line ; si illa much smaller. cambridgei Butl. h. Carapace, sternum, corn, &c. reddisi yellow ; tibire of lst, 2nd, and 4th legs with a yellow band ; abdomen flavous below and without anterior black band above; abdomen furnished with only 5 lateral blunt tubercles, the posterior of these being far from the middle of the posterior border ; sigiua very large...... , ...... butleri, sp. n. 1899.1 AND SPIDERS FROM TROPICAL WEST AFRICA. a(i 1 ~ETHRODE~,gen. nov. Allied to Aramthrcc, but differing in the following particulars :- a. Four median eyea elevated on a ronnded tubercle : lateral eyes also on a tubercle ; clypeus equal in height to half the length of the ocular quadrangle, which is longer than broad ; abdomen only oierlapping the posterior third of the carapace up to the median tubercle, its anterior and lateral margins armed with smooth rounded tubercles ; its uppsr surface furnished with about a dozen symmetrically nrranged varions-sized tubercles ...... Bthrodcs. 6. Ocular tubercles low, ocular I uadrangle slightly wider than loug; clypeus verylow, lesa tkan half theoculur quadrangle ; abdoinen overlapping tbe posterior two thirds of the carapace, armed marginally with strong spines, without tubercles above ...... Arunethra. ~ETHEODESMAMMOSA. (Plate LV. fig. 2.) Colour of abdomen a nearly uniform ochre-yellow ; cephalothorax darker. Abdomen. twice broad as its median length j its anterior border siuuous, mesially emarginate, armed with seven tubercles, not including the large tubercle on the antero-lateral angle ; L large tubercle on the postero-lateral angla and two smaller lateral tubercles in front of it ; the posterior border widely convex, with s series of vertical tubercles juet above it and one pointed posteriorly close to the large postero-lateral tubercle ; the sigilla deeply impressed and mostly subcircular ; tubercles arranged as shown in Plate LV. fig. 2. Vulva consisting of a semicircular depression in front and 0 narrow transverse plate above the genital aperture. Mecisurements in millimetres.-length of abdomen 12, width22.5 ; length and width of carapace 7 ; length of 1st leg 18. LOG.Benito River (a. L. Bates). TETRAGNATHALatr. TETRAGNATHACLAVIGERA Simon. Tetragnatha clauigcra, Simon, Ann. Soc. Ent. Fr. 1887, p. 272. LOG.Benito River (G. L. Bates). One female specimen referred to this species, recorded by Simon from Assinie. Family OXYOPIDX. Genus PEUCETIAThorell. PEUCETIALONGIPES, sp. n. (Plate LVII. fig. 17.) Coloour (in alcohol). Carapace pale green, thoracic portion sometimes tinted with brownish, a few black spots marking the position of sete on the head ; side of head with an indistinct or distinct vertical fuscous stripe running to the basal spot on the mandible and coutinued 08 n short stripe on the upper part of that appen- PSOC.ZOOL. S0~.-1899, NO.LVI. 56 862 ME. R. I. POUOUE ON SUOEPIONS, PEDIPALPS, [Nov. 14,

dage, the mandibles otherwise -unstriped ; sternum green ; legs yellow ; all the segments, including cox% and trochanters but not the tarsus, spotted with black at the base of the spines and also at the base of the large hairs on the femora, cox=, and trochanters. Abdomen greenish : a pair of yellow stripes on the ventral surface running from the epigastric fold to the spinners, and a pair of broader stripes of the same colour on the dorsal side, separated by an elongate greenish area, which is marked in front by yellow stripes running obliquely forwards and inwards and meeting in the middle line, and behind by yellow spots or stripes continued like the anterior stripes from the lateral yellow bands. Clypeus scarcely verticF1, the angle it forms with the upper surface of the head slightly obtuse ; carapace as long ILS tarsus of 1st leg, longer than patella, tibia, and tarsus of palp, a little shorter than tibia of 3rd leg, B little more than half the length of the protarsus of the 1st. Abdomen long, broad in front, narrow behind, more than twice as long as broad. Legs very long: the 1st more than six times as long as the carapace : 4th a little less than five times as long, armed with long black spines. Measurements in mi1limetre.k-Total length 19.5 ; length of carapace 6-5, of 1st leg 42, of 2nd 37, of 3rd 27, of 4th 30. LOC.Loango (in the Keyserling Collection). Sereral female specimens. Differs from all the African species known to me in the form of the vulva; further differs from P.puZchra Blckw. and P. foCifera Butl. (?= P. striata Karsch) in having no black bands on the front of the clypeus and mandibles, and from P. luteiceps Simon (Donaldson Smith, ‘ Unknown African Countries,’ p. 391) in having the ocular area less prominent, a single black stripe at the sides of the head and upper end of the mandible, Family LYCOSIDB, Genus OCYALEBud.

OCYALE ATALANI’A SaV. LOC.Accra (G. A. Higbtt) ; Benito River (G. L. But-). Distributed throughout Tropical Africa.

Family PISAUEIDB. Genus TETRAGONOPHTHALMAKarsch.

TETEAUONOPHTHALMAPHYLLA Karsch. ?Dobmedes exilipes, Lucas, Thomson’s Arch. Ent. ii. p. 385 (1858). Tetragonopkthalma phylla, Karach, Zeits. gesammt. Naturwise. li. p. 329, pl. ix. fig. 4 (1878). 1999.1 AND 8PIDER8 PBOY TROPICATI WEST AFRICA. 863 LOC.Sierra Leone (Surg.-Capt. Clements); Benito River (G. L. Bate). The web of this spider has been thus described by Surg.-Capt. Clements :--“Fig. 71 is from a photograph of a web which had a height of between 6 and 7 feet. There were some half-dozen of them built in a partially cleared space in the forest. The bottom part of the aeb is in the form of an inverted widely-spread fuiinel, the top being truncated, leaving a circular aperture of three quarters of an inch in diameter. The spider lived beneath the cone, and gained access to the upper portions of his snare by the hole in its top. Numerous threads arose from the fiinnel and were attached to an overhanging branch more than 6 feet above.” (‘ Science Gossip,’ 1893, p. 116, fig. 71.) It appears to me to be not improbable that this species is based upon adult examples of that which Lucas described as Doloinedrs exilipes. Genus PHALBASimon.

PEALBAFEROX, sp. n. (Plate LV. fig. 6, 6n.) Cokur. Carapace ochre-yellow, sparsely clothed with yellowiah hairs, with a narrow blackish-grey margin, some white hairs at the sides of the head and reddish hairs between the eyes ; legs ochre- yellow, clothed with ashy-grey hairs, which become blarker towards the extremities ; mandibles black, scantly clothed with yellowish- grey hairs ; sternum blackish, clothed, like the maxilla and labium, with black hairs; upper surface and sides of abdomeu clothed with reddish-yellow hairs, its lower surface with greyish-black hairs. Carupace about one-fourth longer than broad, its length about equal to that of tibia of 2nd leg and to protarsus of 1st and 4th legs and to femur, patella, tibia, and half the tarsus of the palp ; its width equal to tibia or protarsus of 3rd leg and almost equal to patella, tibia, and tarsus of palp. Cephalic region high, convexly rounded from before backwards and from side to side; ocular quadrangle much longer than wide, parallel-sided, the eyes aub- equal and ahout a diameter apart, the distance between an anterior and a posterior median of the same side equal to about two diameters, anterior and posterior laterals of either side more than twice aa far apart as anterior and posterior medians ; anterior and posterior laterals situated ou small tubercles ; anterior laterals about their own diameter above the edge of the clypeus ; anterior medians about their diameter above it. Mandibles long ; the fanggroove armed with four teeth along its posterior edge. Lep long and strong, 1, 2, 3, 4 ; the 1st more than four timw as long as the carapace, measured froin base of femur; patella unspined, except for a setiform ~piniileat the extremity of the uppereide; tibiae of 1st and 2nd armed with 4 pairs of spines below and 2 behind and 2 iu front; tibia of 3rd and 4th with 56* 564 MR. R. I. POCOaE ON SCORPIOKS, PEDIPALPS, [NOV. 14, 3 pairs of iuferior spines, 2 behiud, 2 in front, and 1 above; femora serially spined above. Abdomen oval, as wide as high, not twice as long as broad. Vdua as in figure. Jfeasurements in mi1limetres.-Total length 30 ; length of carapace 13, width 10; length of ocular area 3, posterior width 4; length of palyus 16, of 1st leg 52, of 2nd 51, of 3rd 39, of 4th 48 ; patella and tibia of 1st 19.5, of 4th 17. The two known species of this genus, P. ccmescem aud P. vulpiaa Simon (Ann. SOC.Eut. Belg. xlii. p. 10, 1898), from the Congo, are too briefly described to be identifiable. Both are snialler than P. ferox, P. cunescens being 20 am. long and P. vubpina 22 mm. Genus THALASSIUSSimon.

THALASSIUS OUINEENSIS (Lucas). (Plate LvI1. fig. 18.) Olios pineensis, Lucas in Thornson’s Arch. Ent. ii. p. 405, pl. xiii. fig. 6 (1858). 0. Colour. Carapace with narrow black rim, covered above and laterally with a mixture of white and brown hairs, sometimes the white predominating, sometimes the brown ; no distinct white marginal or submarginal band; the clypeus the same tint as the sides of the head ; abdomen a deep rich-broKn or greyish brown above, darker behind than in front, covered with a mixture of whitish and reddish-brown hairs: sometimes the white, sometimes the brown predominating ; ornamented with four pairs of symmetrically- disposed blood-red patches, the posterior patches often indistinct ; no lateral pale band ; lower surface tolerably uniform yellowish brown; legs a tolerably uniform greyish or brownish hue, not banded ; the protarsus sometimes darker at the tip. Carapace just about equal to tibia of 1st and to tibia and protarsus of 4th leg ; longer by about one-fourth of the tarsus than metatarsus of 1st ; its upper surface flat, as high behind as in front. Leys robust, thickly plumo8e, 4, 1 and 2, 3 in length ; patella and tibia of 4th a little less than those of lst, and distinctly less than those of 2nd. Lateral lobes of vulva long, oblique, converging posteriorly and meeting in a short median suture ; the depression between them semioval, widely open in front, aud filled in with a completely chitinous irregular sclerite (see Plate LVII. fig. 18). d. Resembling 2 in colour; legs much longer, but equally strongly plumose. Palp when extended not reaching apex of femur of 1st leg ; its tibia subcylindrical, unarmed ; tarsus piri- form, shorter than patella and tibia taken together. Mensurvnzrnts in nii1limetres.- Q . Total length 24 ; length of carapace 11, of 1st leg 42. 2nd leg 42.5, 3rd leg 39, 4th leg 44. d . Total length 22 j length of carapace 10, of 1st leg 52, 2nd 52.5, 3rd 48, 4th 54. Loc. Benito River (a.L. Batee). This species has apparently not been rediscovered eince Luca 1899.1 AND SPIDERS FROM TROPICAL WEST AFRICA. 865 described it in 1858. Hence its generic position has remained o matter of doubt.

THALASSIUSFORMOSUS, sp. n. (Plate LVII. fig. 19.) 9 . Resembling the preceding in its robust, thickly plumose legs, form of vulva, Bc.; but differing essentially in colour. Bide of carapace covered with a thick coating of yellow hairs, with a narrow brown inferior band above the block border ; upperside mostly covered with deep-brown hairs, which on each side invade arid break the continuity of the lateral yellow posterior area. Clypeus brown, the brown area sharply defined at the sides by the yellow hair clothing the side3 of the head. Upperside of abdomen yellow and rich olive-brown, the latter arranged in distinct patterns, forming a posterior median patch, in front of which there are five transverae stripes, the anterior broader than the posterior ; also small deep brou n spots scattered here and there ou the yellow, and the blood-red patches noticeable on pineensis also present. Upperside of legs yellow, banded with brown; fumora with a broad basal brown baud and a narrower band of the same colour, about one fourth of the distance from the apex ; patella brown, slightly yellow distally j base of tibia narrowly, apex more widely brown ; base and apex of protarsus aud of tarsus brown. Measurements in ~nillimetres.-Total length 21 ; 1eDgth of carapace 10, 1st leg 35.5, 2nd leg 36, 3rd leg 33, 4th leg 38. LOC.Benito River (G. L. Bates). Although not quite adult, the type of the species shows the same form of vulva as in subadult examples of T. pineensis.

THALASSIUSINORNATUS, sp. n. 2. Colour much as in T. guineewsis; carapace and abdomen covered with a uniform mixture of brown and yellowish-grey hairs; abdomen ornamented posteriorlv at the sides with some blood-red patches running into ill-defined stripes ; integument of abdomen olive-yellow, with a median anterior pale narrow lanceolate area ; lega uniformly brown, covered with greyish-white hairs. Curupace a little less than tibia of 1st and 4th and than protarsus of 4th leg, slightly longer than protarsus of 1st. Leys mnch less thickly plumose than in yuineensis; patella and tibia of 4th a little less than those of 1st. Lower side of abdomen covered with short slender hairs, interspersed amongst the normal hairy coating. Lateral lobes of vulva irregularly subquadrate ; the inner edge longitudinally truncate and almost contiguous throughout their length, being merely separated for a short distance anteriorly by a narrow median sclerite ; anterior depression of vulva marked with a pale niembrauous spot on each side, and on the inner side of the spot with a black, thickly chitinous ridge. LOG.Benito River (G.L. Bntes). Iu colour, &c., closely resembling the Somaliland species T. unieo7or Sinion (in Donaldson Smith's ' Through Unknown African 866 U.It. I. POUOCK OX SCORPIONS, PBDIPALPS, [NOT. 14, Countries,' p. 389, 1897). The latter, however, has the lower side of the abdotnen covered with sets, which are longer and more riumerous than- in T. inornutus ; there are no blood-red markings on the abdorneu, and the dqwession of the vulva has no pale lateral spot and black ridge.

THALASSIUSAURATUS, sp. n. (Plate LVII. fig. 20.) Colour. Carapace uniformly covered, except on the clypeus which is brown, with pale hairs, white at the sides and becoming yellowish on the summit j abdomen covered above with golden-yellow hairs and marked in its posterior half with some sniall symmetrical brown spots ; sides of abdomen brownish red, dwker above than below, lower surface bright yellowish brown ; legs uniformly deep chocolate-brown. Curapuce as long m protarsus and one-fourth of the tarsus of the 1st leg, slightly shorter than tibia of 1st and than tibia and protarsus of 4th. Pulvu very like that of 7'. replk, but the anterior depression filled in at the sides and leaving a longitudinally oblong median depression. LOG.Beiiito River (a. L. Bates). A single adult female. Recognizable from the other species by the uniform golden-yellow colour of the upperside of abdomen and carapace, and uniformly deep brown legs.

THALASSIUSLEUCOSTICTUS, sp. n. Colour of carapace brown, with a broad yellow band on each side running from the clypeus almost to the posterior border, its upper edge tolerably even, the lower irregularly jagged on a level with the coxa of the 2nd and 3rd legs, where the space between the stripe and the lateral margin is widest ; a few small white spots on the dorsal surface of the carapace and one on each side above the anterior extremity of the stripe. Abdomen a deep rich velvety brown above, with a broad yellow stripe on each side, the stripe strnight throughout its length, not geniculate, but at its posterior end irregularly jagged above, forming incipient white spots ; anterior part of upper surface of abdomen with a few symmetri- callp arranged yellow spots ; sides of abdomen below the stripe spotted with yellow ; legs and palpi brown, spotted with yellow stripes above, protarsi in addition ringed with darker bands; lower side of legs, sternurn, and mandibles a tolerably uniform fawn-brown j lower side of abdomen a little darker than sternum, with a few white spots at the sides. Carupme shorter than the 1st and 4th tibia and not quite so long as the 4th protarsus, longer than the 1st protarsus; width of carapace just about equal to tibia of 3rd leg. Ocular quadrangle longer than wide, shorter than height of clypeus ; posterior median eyes larger than anterior medians. +gs 4, 2, 1, 3, strongly and normally spined ; protarsi not very noticeably plumose. 1899.1 AND SPIDERS FROY TROPICAL WEIST AFRICA. 86;

Measurements in rnil1imetres.-Total length 14 ; length OE carapace 6.8, width 5.8 ; length of 1st leg 275, of 2nd 28, of 3rd 23.5, of 4th 29. LOG.Benito River (G. L. Bates). ltesembling Y'. spiinosissinaus Karsch, from the same area, in colouring, but recognizable, according to the description of the latter, in having the legs ornamented with white transverse stripes and Bpots. Although the type of this species is not quite adult, the colouring will probably be found to aEord a better criterion of its specific distinction than the form of the vulva.

THALASSITJSLEONENSIS, sp. n. (Plate LVII. fig. 21.) Colour a uniform reddish brown above and below ; carapace and abdomen with a silverg a hite band extending from the sides of the clypeus almost to the spinners, the band on the carapace considerably narrower than the brown margin external to it. Carapace normally high, as high hehind as in front, its width equal to the area between the posterior border and the posterior lateral eye; its length equal to length of tibia or protarsus of 4th leg j slightly exceeding tibia of 1st and 2nd, width slightly exceeding tibia of 3rd. Vulva practically as in T. spenceri P. Cambr., from Cape Colony (see P. Z. S. 1898, pl. iv. fig. 1b), but longer as compared to its width owing to the greater leugth of the lateral lohes. Measurements iii mil1imetres.-Total length 21 j length of carapace 9.5, width 8 ; length of 1st leg 35, of 2nd 35.5, of 3rd 33, of 4th 38 j tibia of 4th 9.5, protarsus of 4th 9.8. LOC.Sierra Leone. The type of this species is the specimen from Sierra Leone referred by Mr. I?. Cambridge to 7'. spenceri, the type of which came from East London in Cape Colony (P. !A. 9. 1898, p. 30). The two specimens appear to me, however, to be specifically distinguishable. In 5". spenceri the carapace is very slightly shorter than tibia 1 (not longer as stated in the synopsis : Zoc. cit. p. 29), and barely exceeds the 1st protarsus, and the width of the carapace is a shade less than tibia 3 j the lobes of the vulva, too, are shorter and smaller.

THALASSIUSBATESI, sp. n. General characters as in T. Zeonenais, but with yellow band on carapace and abdomen broader. Vulva differing in that the hairy lateral lobes do not meet in the middle line, but are separated in front by a median shining hairless sclerite, which is itself divided by a central longitudinal sulcus. Measurements in millimetres.-To~al length 22 ; length of carapace 9.2, width 7.5; length of 1st leg 36, 3rd 36, 4th 39; protarsus of 1st 8, of 4th 9.8. Lac. Renito River (G. L. Bntes). Only the female known. 808 ME. B. I. POOOCK ON SCORPIONS, PEDIPALPEI, [Nov. 14,

TEALASSILTSRBGALIS, sp. n. (Plate LVII. fig. 28.) Colour. Carapace and abdomen with a pair of broad yellow bands extending from dypeus to spinners, covered elsewhere with yellowish rusty-red hairs ; legs strongly banded ; femora of legs yellowish red, those of 2nd pair slightly, of 3rd and 4th pairs distinctly banded with black ; apex of femora, whole of patella, bad seventh and apical third of tibiae, basal and apical fourth of protarsi and tarsi black ; the rest of the legs yellowish red. Carapace a little less than tibia of 1st leg, excelling its protarsus by about one fourth the tarsus, distinctly shorter than both tibia and protarsus of 4th. Lobes of vulva in contact along the posterior half of the inner edge, separated in the anterior half by R median heart-shaped sclerite, the depression in front of the lobes marked with two posteriorly converging black ridges. Measurements in rni1limetres.-Total length 22 ; length of carapace 9, width 8 ; length of 1st leg 36, of 2nd 36, of 3rd 32, of 4th 38 ; protarsus of 1st 8, of 4th 9.8. LOP.Benito River (a. L. Bates). An adult and an immature female. THALASSIUSMSIGKIS, sp. n. (Plate LVII. figs. 23,23 u.) 9. Colour. Carapace brownish red, covered with brownish hairs speckled with white, with a black edge, within the black edge a white rim which broadens on each side of the head, the face uniformly dark, sides of thoracic portion posteriorly blackish j mandibles black with yellowish hairs ; abdomen brownish above, speckled with white hairs and with white spots at the sides ; lower side of abdomen, sternum, and cox= greyish jellow. Legs variegated above ; femora yellow, indistinctlp variegated j patellae yellow, blackish at the base ; tibiae yellow, black at base and apex ; protarsi yellowish,also black at. apex and base ; tarsi yellowish ; the yellowish areas of the legs covered with white hairs, the darker patches with dark hairs. Curapace nearly circular, a little longer than wide, the width equal to the length from the posterior border to the anterior pair of eyes j length slightly less than protarsus of 1st leg, about equal to tibia of 3rd, width slightly shorter than tibia of 3rd ; carapace low, higher in the cephalic region than posteriorly, not strongly elevated behind ; clypeus scarrely exceeding the ocular quadrangle; ocular quadrangle longer than broad, slightly narrowed in front, the posterior median eyes much larger than the anterior median. Legs 4, 2, 1, 3 j the lst, 2nd, and 4th subequal; patella and tibia of 1st and 2nd about equal and a little longer than those of 4th; protarsus of 4th excee ding that of 1st by about one third the length of the tarsus. Abdomen broadest. in its posterior half, gradually narrowed in front, abruptly narrowed behind. Measurements in mil1inzetre.s.-Total length 16 ; length of carapace 7.7, width 7.2; length of 1st leg 34, of 2nd 345, of 3rd 31, of 4th 35. LOC.Benito River (a. L. Bates). 1899.1 AND BPIDEBB XBOM TROPICAL WEST AFEICA. 869 The characters of the foregoing species of TholnsSius may be tabulated as follows:- a. Cam ace flatter and less elevatad behind, cephalic region bigier than thoracic; legs yellow, banded and spotted with black ...... idpis. I. Carapace high and more convex behind, level along upperside; cephalic area not higher than thoracic. a'. A broad yellow lateral stripe extending from side of head to spinners. u'. Lags yellow, strongly banded with black ; patella black.. regdis. 6'. 6'. Legs yellowish brown, at moat spotted above with white. a3. Upperside of legs ornamentedwitb silvery white spots and bands, white spots also on upperside of abdomen above yellow band...... leucostictus. IS. Upperside of legs and abdomen not spotted with white. d. Lateral lobes of vulva in contact, not separated mesially by a shining hairless sclerite ...... Zeoimais. P. Lateral lobas of vulva separated posteriorly in the middle line by a shining hairless sclerite ...... balesi. b'. NO definite longitudinal yellow band extending from head to spinners. as. Legs robust, thickly plumose distally; lateral lobes of vulva elongate, obliquely converging, and meeting posteriorly in a very short suture. d. Legs and upperside of abdomen strongly banded; clypeus darker tban side of head ...... fmsurs. I'. Legs and abdomen not banded; clypeus same colour as sides of head ...... gaineenais. I'. Legs thinner, not thickly plumose ; lateral lobes of vulva subquadrate, meeting or nearly meeting in a long median suture. u'. Upperside of thorax and abdomen nearly unifornily golden yellow ; legs chocolate-brown ...... auratzrs. I'. Upperside of thorax and abdoinen covered with brown and yellow hairs; legs pale, covered with greyish-white hairs ...... inornutas. The following West-African species are unknown to me :- T. spinoeissimus Karsch (Zeits. gesammt. Naturwiss. lii. p. 345, 1879), described as a Ctenus from the Loango coast, will fall apparently under section b' of the above table; but seems to differ from both T.bnteai and T. leonensk in having the brown tield of the upperside of the abdomen laterally spotted with white, perhaps as in T. leucostictus. The description, however, contains no statement to the effect that the legs are ornamental as in T. leucostictue. T. pielus, Simon (Ann. SOC. Ent. Belg. xlii. p. 17, 1898), from OgowB, will according to colour characters fall under section a or as, resembling 2'. insignis in some respects ; but since M. Simon makes no mention of any peculiarity in the form of the carapace, it is not permissible to suppose that his species is identical with T. insignis. Genus DOLOMEDRSLabr.

DOLOMEDESTRALNSFUGA, sp. n. (Plate LVII. fig. 24.) 8. CoZour a tolerably uniform yellowish brown; carapace covered above with olive-brown hairs, with a broad marginal snow-white 870 MB. It. I. POCOCK ON BCORPIONS, PDIPALPS, “OV. 14, band extending from middle of clypeus to side of posterior emargination. Upperside of abdomen covered with short hairs of a rusty-red hue j a snow-white lateral band set off above and below by a narrow darker stripe ; legs and ventral surface not varied. Carapaee nearly circular, its width equal to the length of its upper surface from poaterior median eyes to posterior emargination ; its upper surface posteriorly elevated, a little higher than cephalic region ; length a little less than tibia of 3rd and a little greater than the tarsus of 4th leg, a little excelling half the length of the patella and tibia of 1st and 4th. Eyes of anterior line a little recurved; laterals much smaller than centrals, their upper edges in a straight line ; anterior line noticeably wider on each side than line of posterior medians, the latter much larger than anterior medians j ocular quadrangle less than height of clypeus, much wider behind than in front, its posterior width exceeding its length. Inferior border of mandible armed posteriorly with 4 teeth. Legs long and slender, 4, 1, 2, 3 in length j patella and tibia of 1st and 2nd subequal and a little shorter than of 4th j protarsus of 4th equal to protarsns and one third of tarsus of 1st; patella armed with 3 spines, one on each side and one median apical; tarsi not scopulate, thickly setose below. PaZp extending past middle of tibia of 1st leg ; tibia1 segment longer than patella, distally incrassate, armed externally with a crescmtically upcurved pointed spine, and on the inner side with a short quadrate lightly emarginate buttress ; tarsus about as long as patella and tibia, short in its basal half, subcylindrical distally. Meagurernenta in mil1imetres.-Total length 21 ; length of carapace 18.3, width 9 ; length of 1st leg 56, of 2nd 54, of 3rd 48, of 4th d@. LOC.Benito River (a. L. Batea). An adult male and an immature female. In this description some characters considered to be of generic value have been repeated to substantiate the claim of the species to be ranked in the genus Dolomedea. This genus is, according to Simon, replaced in Tropical Africa by the allied form Tapinothele, which has been recorded from Zanzibar (Hist. Nat. Araignhes, i. pp. 310 t 333, 1898). But the species here named tmnsfup does not appear to differ in any important particulars from the genus Dolomedea as characterized by Simon, and certainly does not fall into the genus Tapinothele.

Family CTENIDB.

Genus CTENUS.

CTENUSBURTONI F. Cambr. Ctenus burtoni, F. Cambr. Proc. Zool. 8oc. 1898, p. 25, pl. iii. figs. 3 a-f. LOC.Cameroons (Capt. Burtoga) j Benito River (G. L. Bates). l899.] AND SPIDERS .FROM TROPICAL WEST ILFBIOA. 871 This species ia apparently nearly idlied to Phoneutriu capulinu Karsch, the colour of the mandibles being apparently the same in the two species and both of them possessing the beard-like fringe of hair on t.he lower side of the legs, upon which Karsch lays SO much stress. Fortunately the question can be without dif8cult.v settled by a comparison between the palpus of C. capulinus and the figures of that of C. burtoni which Mr. Cambridge has published.

CT~NIJSOCCIDEINTALIS F. Cambr. Ctenus om'dentnlis, F. Cambr. Proc. Zool. 8oc. 1898, p. 22. LOC.W. Africa (without further history).

CTENUSXIRGSLEYI F. Cambr. Ctenus kingsleyi, F. Cambr. Proc. Zool. SOC.1898, p. 21, pl. iii. fin.v 6. LOG.Cameroons (Miss Kinysley).

CTENUSSCOPULATUS, sp. n. (Plate LVII. fig. 25.) Colour. Carapace deep mahogany, clothed with short reddish- brown hairs, hairs on the face and upper half of mandible deep carmine ; mandibles shining black : legs same colour as carapace above; femora redder below, especially the base of the 1st and 2nd in front ; scopulao greyish black. Abdomen indistinctly variegated black and red above, the sides especially in front clothed with bright reddish hairs interspersed with longer hairs of a paler tint j the lower surface behind the epigastric fold entirely covered with a broad velvety-black field, narrowed and oval behind and sharply marked off at the sides by t4e red hairs of the lateral surface and in front by the reddish-yellow epigastric region; steruum and coxao blackish brown. Carapace longitudinally horizontal above, a little shorter than palp measured from base of femur, longer than patella and tibia or than tarsus and protarsus of 3rd leg, a little shorter than 4th protarsus, just about equal to protarsus and tarsus of 2nd j its width just about equal to tibia of 2nd and of 4th; ocular quadrangle longer than wide, narrowed in front, exceeding height of clypeus, which nearly equals three diameters of anterior median eyes. Lgs 4, 1, 2, 3; tibiae of 1st and 2nd with 5 pairs of inferior Rpines, not spined above, of 3rd and 4th with 3 pairs and 2 anterior and 2 posterior spines and 3 superior spines ; patellse of 1st and 2nd unspined, of 3rd and 4th with 1 anterior and 1 posterior spine; tibiae of 1st and 2nd scopulate below, of 3rd scopulate below at its distal end ; protarsus of 4th scopulate except in the basal fourth of its length; protnrsi of 1st and 2nd without inferior apical spine. Vulva cousisting oE a reddish, heart-shaped, convex, median sclerite, with a small shining black dentiforin process on each side of its posterior extremity. 872 MB. E. I. POCOCK ON SCOEPIOKE, PEDIPALPB, [Noo. 14, &eaeurements in mil1imetres.-Total length 31 ; length of carapace 15, width 12 ; length of palp 17, of 1st leg 50, of 2nd leg 47, of 3rd leg 38, of 4th leg 61 ; patella and tibia of 1st leg 26, of 4th 17, protarsus of 1st 12, of 4th 16. Loc. Benito River (0. L. Bates). Rivalling the preceding species, C. Xringsleyi and C. occiclentalis,in size, and resembling them in t,he absence of the inferior apical protarsal spine of the 1st and 2nd leg; but at once recognizable from both by its coloration, especially of the lower surface of the abdomen. It further resembles C'. kitigsleyi in having the carapace shorter than the 4th protarsus and very much in the form and size of the vulva, but differs from it and approaches C. occicZenta2is in habing the carapace longer than the patella and tibia of the 3rd leg. The black patch on the lower side of the abdomen in C. swpulatus calls to mind the somewhat similar colouring found in Phoneutria melanogastra of Bosenberg and Lenz (JB. Hamburg. Mus. xii. p. 12, pl. i. fig. 14), from East Africa; but the two species are certainly quite distinct, seeing that P. melarwgastra is less than half the size of C. scopulatus and has a median pale stripe 011 the carapace. And lastly from Simon's species C. erythrocldis from Landana (Bull. SOC. Zool. France, i. p. 222), which also has red hairs on the base of the mandible, C. scopulatus may be recognized by the absence of pale bands on the carapace and the presence of the black field on the lower side of the abdomen. Ctenue auricularis and C. capulinus of Karsch, from Chinchoxo, also seem to differ in colour from C. scopu2atus, as well as from C. occitlentalis and C. kinysleyi (see Zeits. gesammt. Nnturw. lii. pp. 347-348, 1879).

CTENUSRIVULATUS, sp. n. (Plate LVII. figs. 26, 26 a.) 0. Colour. Carapace covered with blackish or olive-brown hairs, with a median pale golden stripe, and a broad irregular interrupted Rubmarginal band extending from the sides of the clypeus to the posterior border ; upperside of abdomen ornamented with a median golden-yellow or greyish band, narrow in front and iudented at the sides with four pairs of large black spots ; sides of abdomen blackish, mottled with yellow or grey spots ; on the lower side the pale spots are arranged in definite longitudinal posteriorly converging lines : sternum, coxae, and lower side of legs uniformly deep brown ; upperside of legs varied, especially on the femora, with golden-yellow bands, the rest of the segments tinted with yellow ; palpi uniformly dull brown ; mandibles black in front, without bright coloured base. Carapace slightly elevated above behind the fovea, as long as patella+tibiaof 3rd leg, aj tibia of lst, and a little longer than protarsus of 1st and than tibia of 4th, about four fifths of the protarsus of the 4th ; ocular quadrangle more narrciwed in front than in C. scopulrrtua, its eyes relatively larger, the airterior separated by a space which about equals the radius. Legs 4, 1, 2, 3; patellattibia of 4th a little less than of 1st 1899.1 AND SI’IUERY FIXOM TItOPICbL WEST AFRICA. ‘873

and a little more t,han of 2nd ; protarsi of 1st and 2nd legs without iiiferior apical spine ; tibia of 1st lightly scopolate. Vulva consisting of a median red sclerite, shaped like the ace of spades, as broad as long, its anterior border with a median for- uwdly directed prominence and convex sides ; the lateral tooth less prominent than in C. scqiu.lntus. 6.Similar in colour to $? ,legs much longer (cf. measurements). Legs not fringed as in C. burtoni aud C. capdinus; tibia1 spur of palpns long, viewed from above diverging externally, then turned forwards at apex ; seen from the outside aspect, the upper edge is nearly straight, the lower concave, with the apex stronglg bifurcate, the upper branch of the fork slightly hooked apically, the lower truncate with acute inferior angle ; tarsus with only a small tuberculiforrn promiuence at base on outer side. Mecisurements iii mi11imtres.- 2 . Total length 26 ; lenglh of carapace 12, of 1st leg 44, 2nd 49, 3rd 34, 4th 47.5. 8. l’ot,al length 20; length of carapace 11, of 1st leg 57, 2nd 50, 3rd 41, 4th 57. Lnc. Benito River (G. I;. Bates). CTEYUS(LEPTOCTENLJS) AGILIOR, sp. n. (Plate LV. fig. 7.) 6.Colour. Carapace with pale median band constricted just behind the head, 1at)erally infuscate, pnler above the border, area around eyes black ; mandibles pale, infuscate externally ; palpi and legs pale, obscurely handed with greyish black ; upperside of ahdomm clothed with olive-yellow hairs, and diversified with red and black marliing ; underside flavous. Cwoprtce with a deep depression betiiiid the head ; head and posterior portion elevated ; clypeus narrower than anterior median eyes. /Icy? wry long antl slender : tarsnl and protarsal scopulae very scxnty ; tibia of I st, arnit:tl with 4 anterior and 4 posterior spines and with (ipairs of inferior spines ; protarsus of 1st with 3 pairs ol‘ inferior spines, arid with 1 anterior and 3 posterior spines ; pat,c,lla: spined in front and behind. I’ctlp with tibia longer than pat-ella, and furnished externally wii h a Iaige bitid prominence, of which the upper branch is longer antl sharper tlian the lower ; base of t,arsus above furnished with n conical process ; tip OF tarsus prolonged and subcylindrical ; t,arsus not- so long as patella and t,ibia taken together (for struc- turd details of tarsus sec P1. LV. fig. 7). Mectsuremeds i,~,,riZZii,1Pt,.Ps.-Total lcngtli 10 ; length of campace 5, of palpus 9, of 1st leg 33, 2nd 2!j, 3rd 24? 4th 37; tibia of 1st 10, of 4111 9. LOC.Henito ltiver (G. L. Bates). A single adult 6 example. ill. Simon has recently described two species of Cteitus (/+to- ctmir~)from West Africa, namely C. (L.) I!/cosiitus and C. (L.) cccvclcnf~s, both from t,lie Jjio Pungo (Ann. SOC.Ent. Prance, 1897, pp. 4!)3-4). Both are based upon females, and are not comparable with the wale described above. From C. (L.) modestus (id. Zoc. 874 MR. It. I. POCOCK ON SCORPIORS, PmDIPbLP8, [Nov. 14, dt. p. 492) from Zanzibar, of which the male is known, C. (L.) agilior certainly differs in having the tibia of the palp longer than the patella and both together longer than the tarsus, and jn the form of the tibia1 apophysis, which is described as acute and erect. Family HETEROPODIDB. Genus SELENOPSLatr. SmmNoPs ANNULATUS Simon. Selenope annulatq Simon, Bull. SOC.Zool. Fr. i. 1879, p. 15. LOC.Benito River (a. L. Bates). Becorded from bpchoxo. The following species from the Congo Region may be identical with the preceding :- S. eectov, Rarsch, Zeita. ges. Xaturw. 1879, p. 342. S. biichneri, id. Bed. ent. Zeits. xxv. p. 94 (1881). S. brownii, Marx, P. U. 8. Nat. Mus. xvi. p. 589 (1893). Genus TOEANIASimon. TOBANIAOUCIDBNTALIS (Simon). Isopeda occidental&, Simon, Ann. SOC. Ent. Fmce, 1887, p. 264. Loc. Benito River (G. L. Bates), Accra (G. A. Hiylett) ; Afraln Plain, inland of Ashmti. Recorded by Simon from Assinie. TORANIAVABIATA, sp. n. (Plate LVIII. figs. 30-30d.) 9. Colour. Carapace castaneous, clothed with ashy-grey hairs, wiLh o thickened white stripe along the posterior slope, the white emphasized by a conspicuous bicrescentic black stripe in front and a pair of narrower black stripes behind ; ocular area black ;clypeus clothed with white hairs ; mandibles nearly black, clothed with greyish hairs above; maxilla and labium black; pnlpi clothed with grey hairs, mottled with black, tarsi fuscous; legs grey, mottled with black above ; femora with three largish black spots above, ashy black below, those of the posterior pairs paler than the others ; patellae blaok beneath ; tibiae white at base and apex below, black in the middle ; scopulae blackish grey ; sternum testaceous ; cox= also testaceous, those of first three legs and of the 4th slightly black in front ; abdomen mottled ashy grey above, with a strong transverse sinuous black stripe in the posterior fourth of its length and irregular blackish patcbes at the sides and smaller spots in the middle ; lower Bide of abdomen yellowish grey, with four narrow blackish longitudinal lines. Carapace as broad as long, not quite so flat as in T. ocoidentalis, a8 long as tibia of 1st leg and as protarsus and half the tarsus of the 4th leg, a little longer than patella, tibia, and tarsus of palp. Etjea .of posterior line recurved; distqpce between medians less than 1899.1 AND SPIDBES FROM TROPICAL WEST AFRICA. 875 between medians and laterals ; eyes of anterior line with superior edges on a level, the laterals nearly twice the diameter of the medians, at all events much larger ; ocular quadrangle longer then wide, parallel-sided, the anterior eyes considerably larger than the posterior and closer together. Mundibles weakly geniculate basally ; inferior edge with 4 teeth behind, 2 in front. Legs 2, 1, 3 and 4 ; 2nd excelling 1st by length of its tarsus ; spine-armature as in T. occidentalis, but with anterior spine on all the patellae and a posterior on all the patellae but that of 4th leg. Vulva as in occidentalis, consisting of an anterior central depression followed by a pair of subcircular lobes separated in the middle line by a median groove (Plate LVIII. fig. 3Od). 8. Resembling 9 ,but less vividly coloured and with longer legs; carapace equal to length of tibia of 3rd leg : palpal organ as in fignre (Plate LVIII. figs. 3Ob-30~);the tibia with a pair of strong subequal external spurs. Meaeurements in millimetree.- 9. Total length 18 ; length of carapace 8, of 2nd leg 36, 4th leg 28. 8. Totallength 16.5 ; length of carapace 8, of 2nd leg 42, of 4th leg 33. LOG.Benito River (U.L. Batee). This species and the foregoing may be at once distinguished as follows :- a. Larger, up to 30 mm. or over; anterior median eyes scarcely smaller than anterior laterals ; sternum black ; cox= of 1st and 2nd legs deep blackish brown; femora witlf exception of the lst, which haa a black basal spot, yellowish red beneath ...... occidentalis Sim. b. Smaller, barely 20 mm. in length; anterior niedian eyes much smaller than anterior laterals ; sternum testa- ceooa; cox= of 1st and 2nd legs infuscate only in front ; femora of legs, especially of 1st and 2nd pairs, why black beneath ...... uuriata, sp. n. Genus RNMMIUSSimon.

REMMIUSTULTUOSUB Simon. Remmius uultuoszur, Simon, Ann. SOC. Ent. .France, 1896, p. 484. Loc. Benito River (a.L. Bates). ? REMMIGETULPINUS Simon. Remmiua vulpinus, Simon, Ann. SOC.Ent. France, 1896, p. 485. LOG.Benito River (U.L. Batee). A single immature female doubtfully referred to this species.

Genus 8PAEABSUB Latr.

SPARASSUSBENITENSIS, sp. n. (Plate LVI. &s. 12, 12a, & Plate LVIII. fig. 27.) 9. Colour of carapace a uniform ochre-yellow, clothed with 876 WE. B. I. POCOCK ON BCORPIONLI, PBDTPALPB, [NOV. 14,

yellow hairs : mandibles, maxillae, and labium mahogany-brown ; sternum and legs much the same coloiir as the carapace; femora redder below, indistinctly spotted above : protarsi covered above with a clothing of greyish hairs, the rest of the segments scantily hairy; abdomen covered with brightish yellow hairs, marked above and at the sides with short darker stripes and with a darker median longitudinal stripe emphasized by three pairs of black spots ; dark blackish brown over the spinners. Carapme a little longer than tibia of 4th leg, equal to the length of its protarsus, a little longer than patella, tibia, and tarsus of palp, and a little shorter than rotarsus and tarsus of 3rd leg ; its width about equal to tibia o! 4th leg, and to patella, tibia, and tarsus of palp. @ea of posterior line subequally spaced, space between the medians almost equal to twice their diameter ; ocular quadrangle parallel-sided and about as long as wide ; the anterior median eyes a little larger tban the posterior median, about a diameter apart, less than a diameter from the anterior laterals, which they equal in size and with which they are in the same straight line ; clypeus less than diameter of anterjpr median eye. - Legs long and slender ; tibiae of legs armed with 11 spines, the odd spine being in the dorsal middle line ; patellae armed with an anterior and a posterior spine ; protarsi of 1st and 2nd scopulate to the base, of the 2nd less closely than of the 1st ; basal portion of 3rd scarcely scopulate ; 4th protarsus only scantily scopulate in its distal half. Vulva as in figs. 12, 12a, Plate LVI. Measurements in milZimetres.-Total length 14; length of carapace 6, width 5.5 j length of 1st leg 28, of 2nd 29.5, of 3rd 20, of 4th 235 ; patella and tibia of 1st 10.5, of 4th 8. 8. Colour. Carapace yellow, clothed with ferruginous hnirs and mottled with white bands and spots ; legs mottled with white blotches and bands ; upperside of abdomen variegated with white and olive-green hairs. Carapace distinctly longer than broad, the length a little less than tibia of 3rd and than half patella+ tibia of 1st leg. Eyes of posterior line practically stmight ; anterior medians about half their diameter apart and B little nearer to the laterals, which are distinctly smaller. Legs moderately long; 2nd surpassing 1st b.7 barely half its tarflus; 3rd slightly surpassing tibia of 2nd and middle of protarsus of 4th; 2nd a little less than six times, 4th a little less than five times, as long as carapace ; tibiae with a pair of apl’cal inferior spines j patellae with anterior and posterior spine. PaZp with tibia considerably longer than patella, armed with 5 long spines; the apophysis short at the base, its distal portion bent downwards nqrly at riglt angles and forming a slender spiniform proeess; tarsus ovalq about as long as patella and tibia. 1899.3 AND SPIDERS FROM TBOPICA& WBT APRICA. 877 bfeastcrenbent8in miUinzetres.-!l'otal length 13 ;length of carapace 6, width 48 ; leugth of 1st leg 33, of 2nd 34, of 3rd 22.3, of 4th 28; patella and tibia of 1st 12.5, tibia of 3rd 6. GC.Benito River (G. L. Bates). SPARASSUSBATESI, sp. n. (Plate LVIII. fig. 29.) $? . Colour. Carapace, sternum, palpi, and lags flavous, less darkly coloured distally, with blackish-grey scopulae on the protarsi and tarsi ; tarsus of palp also with blackish-grey scopulae; mandibles black ; abdomen golden yellow, intermixed with brown, with transverse darker stripes passing outwards from the middle line ; sides of abdomen golden yellow, indistiuctly mottled ; lower side of abdomen yellow. Carapace high, about as long m broad ; cephalic region strongly convex from side to side and from behind forwards ; width of head almost equal to its length measured from the anterior end of the fovea ; eyes more widely spaced than iu 8. benitetLsis ; the ocular quadrangle distinctly narrowed in front. MunrEibles with 4 teeth behind and 2 in front. Legs m in the preceding species, but without patellar spines aud with only two pairs of spines on lower side of tibia. Vulva as in fig. 29, Plate LVIII. Measurements in mil1imetres.-Total length 14; length of campace 6, width 5.9 ; width of head 4 ; length of 1st leg 26, of 2nd 28, of 3rd 19.5, of 4th 23. Lac. Benito River (G. L. Butes).

SPARASSUSTRIFURCATUS, sp. n. (Plate LVIII. fig. 28.) d . Colouv. Integument pale olive-yellow, clothed with heirs of a golden-yellow hue ; mandibles pale ; tarsus of palp infuscate. Carapace high, convexly rounded, a little longer than broad, its length nenrlg equal to tibia of 3rd leg and to half the patella and tibia of the 1st. Eye8 of posterior line very slightly recurved, subequal, subequally spaced, the medians a little less than two diameters apart; eyes of anterior line straight by their centres, medians larger than laterals, half a diameter apart and a little more than that from the laterals. Leg8 2, 1, 4, 3; 2nd surpwsing 1st by a little more than its tarsus, 3rd barely surpassing tibia of 2nd and not reaching tip of protansus of 4th ; 2nd leg only a little more than five times BB long as the carapace, the 4th rather morethan four times; patells of lst, 2nd, and 4th with an anterior spine ; no apical spines on under side of tibiae. Palp with tibia a little longer than patella, both without spines ; tibia1 apophysis consisting of 3 strong teeth, the upper the thickest, obliquely truncate apically, and ending in a sharp point, the inferior shorter und thinner than the rest ; tarsus ebngate oval, about as long as patella + tibia taken together. Meusurements in mi1Zirnetres.-Total length 13; length Q€ carapace YROC. ZOOL. 8OC.-1899, NOi LvlI. 57 878 XI&. R. I. POC‘OCK ON SCORPIONS, PEDLPALPS, [NOV. 1-19

6-5, width 58; length of 1st leg 31.5, of 2nd 35, of 3rd 24, of 4th 28; patella + tibia of 1st leg 12, tibia of 3rd 6. LOG. Cameroons.

SPARASSUSRUFILATUS, sp. n. (Plate LVI. figs. 23, 13a.) d . Colour much as in S. trifurcatus, except that the legs are more orange-red, though the bases of the femora remain pale ; the upperside of the abdomen studded with dark reddish-brown spots and the lower side behind the epigastric fold uniformly pinkish brown. Carapace vbry slightly longer than wide, less than tibia of 3rd and than half the patella + tibia of 1st leg. Eyes of posterior line slightlp procurved, subequal and subequally apaced; eyes of anterior line closer together, equally Spaced, distance between medians about half their radius. Legs very long, 2nd surpassing 1st by its tarsus and one third of its Frotarsus; 2nd leg about seven times as long as carapace. Tibia without apical inferior spines ; patellae unspined. Pa$ with tibia cylindrical, much longer than patella, and armed internally with a single long spine : patella unspined ; tibia1 apophysis consisting of a forwardly directed spur, the apex of which is pointed and hooked ; tarsus very large, its external edge concave, internal convex ;* palpal organ constrticted as in fig. 13, Plate LVI. Heasurements an miZliinetres.-Total length 14; lengthof carapace 6.5, width 6 j length of 1st leg 40, of 2nd 45.5, of 3rd 29, of 4th 33 ; patella + tibia of 1st 14.8, tibia of 3rd 7.8. LOG.Cameroons.

The females of the foregoing species may be recognized i~s follows :- a. Head narrower,its width ouly equal to thelength between the anterior end of’ the fo\ea and the posterior line of eyes ; oyes closer together, the ocdar quadrangle scarce1 narrowed in front; mandiblesreddish brown ; tibia wid 3 pairs of inferior spinea ; patellse at least on 2nd and 3rd legs #pi.@ in front and behind ...... bemtcmia, sp. n. b. Head bdder, its width almost equal to the entire length measured from the apex of the fovea ; eyes more widely spacpd, the ocular quadrangle distinctly narrower in front; mandibles blark; tibim with two pairs of spines beneath, patelloe unspinrd ...... butesi, sp. n. The males of the species here described may be determined as follows :- a, Tibial process of ~alptrifnrcate (tibise of legs without inferior apical spines ; 2nd Ie a good deal loneor thon let, and only a little niore 8an five tiwes aa long as carapace; patella and tibia of palp without spines) ... Ir$mutcls, up. n. b. Tibia1 process of pnlp simple, undivided. d. Tibial rocess stout at bale, its distal hnlf slender and bent iwnwards almost at right angles ; tibis of legs with two pairs of inferior spines; 2nd leg only a little longer thsn lst, and less than six times as long a8 carapace ; tibia of palp strongly spined ...... benitemis, q~.n 1899.3 AND SPIDEIM PROX TROPICAL WEST API~ICA. 8i9

b'. Tibia1 process of palp stout, straightisb, hooked at apex; tibioe of lege with 3 pairs ofspines; 2nd leg considerably longer than 1st and about seven times as long as carapace; tibim of palp with only 1 spine ... rujlatzis, sp. n. Spnrssua (OZios) uZluuudi Simon (Ann. SOC. Ent. Fr. 1887, p. 264), from Assinie, resembles S. batesi in certain characters, for example in having the carapace as wide as long and the mandibles black ; but according to Simon the eyes of the anterior line are procurved in S. alluaudi, whereas they are straight in 8. bat& aud the upperside of the abdomen is ornamented with a longitudinal, lanceolate fuscous band.

PALYSTODES,geu. nov. Carapace abont one-third longer than wide, rather low, moderately convex, flat above longitudiually ; the cephalic area not raised. Eyes of posterior line nearly equidistant, very slightly recurved, the medians larger than the laterals, which are sessile ; those of anterior line also equidistant, their lower edges on a level, the medians much smaller than the laterals, their diameter less than the radius of the latter j cl! peus low, less than the diameter of auterior medians ; quadrangle of median eyes much longer than wide, wider behind than in front, the eyes subequal. Alandibles armed below with 3 teeth in front and 3 behind. Legs 1, 2, 4, 8, completely laterigrade, very long and slender. Recognizable from Pdystes, to which it, is most nearly related, by having the carapace narrower and its upper surface flat from the eyes to the posterior end of the thoracic fovea. In Palystes the carapace is strongly convex longitudinally. PALYBTODESl'LuMosus, sp. n. (Plate LVIII. figs. 31, 31a.) d. Colour. Carapace castaneous, covered with a coating of brownish hairs, mottled with darker and lighter patches; mandibles deep mahogany, clothed with mottled greyish-yellow hairs ; labium and maxillae black ; sternum black, striped with red bands, which cross it transverselg on a level with the coxa ; coxae covered with rich yellowhh-red hairs, black in frolit ; legs blackish, femora spotted with white below, white spotted with black above, with tufts of brownish hairs in front ; patellse white below ; tibiae with two white stripes below, one median and one apical ; upperside of tarsi and protarsi brownish, mottled with tufts of hair; scopula of tarsi and protami rusty red: the colouring of the 4th leg less distinctly marked than that of the others ; apex of femora, tibia, and protarsi, especially of 4t.h leg, with short tufts of hair, soinewhat as in some species of Pandercetes ; palpi black, mottled above, the hairs niore uniformly flavous below ; tarsus covered with ferruginous hairs, fuscous apically below. Abdomen mottled with greyish- yellow above, with tufts of brown hairs, a short anterior mediirn black stripe, and on each side of the middle line a large soini- circular black stripe, the anterior and posterior ends of which 57" 880 YE. El. I. POCOCK OX SCORPIONS, PEDIPALPS, [NOV. 14, are almost in contact in the middle line, the former just in front of the middle of the upper surface, the latter a little distance in front of the spinners; lower aide of abdomen covered with a broad undivided field of deep rich, nearly blood-red hairs, a narrow arched transverse black stripe behind the epigastric fold ; the epigastric region blackish, relieved, especially posteriorly, with golden-yellow hairs; sides of abdomen above the red field greyish, iiiottled with red anteriorly. Carapace equal to length of 4th tibia and 3rd femur, width equal to about half the length of protarsus and tarsus of 2nd leg. Legs-spines of 1st and 2nd femora, 3, 2, 3 ; of 3rd, 2, 2, 3 ; oE 4th, 3, 2,2 ; 1st and 2nd patellse, 1,l; of 3rd and 4th, 1 anterior ; tibia, 2,2, 2 below, 2 in front, 1 above, and 2 behind ; protarsi, 2, 2 below, 2 behind, 3 in front; tibise of 1st and 2nd sinuous; femora of 1st and 4th equal ; 1st leg nearly five times as long as carapace, 3rd leg about three and half times as long, 4th about four times as long. Abdomen more than one third longer than broad, voluminous, broadest just behind the middle, then narrowed abruptly to its termination. Vulva consisting of a transversely oval pit, followed by a pair oE skeletal pieces separated by a narrow median sclerite. Mecrsurements in mi1limetres.-Total length 26 ; length of carapace 11, width 8; length of 1st leg 56, of 2nd 52, of 3rd 39, of 4th 43 (all measured from base of femur) ; patella and tibia of 1st 21, of 4th 16 ; protarsus of 1Rt 14, of 4th 12, LOC.Benito River (a. L. Bates).

Family THOXISIDA.

(bus PHRYN.ARACIINEThorell.

PHRYNARACHNBBUGOSA (Latreille). !l'homim~ruyosus, Latreille, Now. Dict. d'Hist. nat. xxxiv. p. 62. Thornisus foka, Vinson, Arandides Madagascar, tc. p. 69, pl. xiv. fig. 4 (1863). Loc. Benito River (G'. L. Bates). The specimen of Phrynarachnt! obtained by Mr. Rates seem3 identical with an example in the British Nluseum froin Mada- gascar.

PHBYNARACHNEMARMORATA, sp. n. (Plate LVI. fig. 14.) 2. CoZaour. Carapace with a large jet-black patch on each side of the thoracic portion j the margins yellowish brown and the middle liue clearer yellow ; Lead infuscate, the ocular tubercles ainber-yellow, a yellowish-white band passing from the median eyes to the lateral angles of the clypeus ; upper portion of mandible yellowish white, lower portion infuscate ; basal half of palpi yellowish white, tibia and tarsus iafuscate; legs of 1st and 2nd 1899.1 AND lPIDEBS FROM TEOPIOAI, WElT AFltlCA. 881 pairs with coxae, trochanters and femora, patellae, and basal half of tibiae yellow, clouded or marbled with black ; distal half of tibiae and protarsi black, slightly variegated with yellow, tarsi yellow ; 3rd and 4th legs black, with tarsus and basal half of femur yellow, and the rest of the segmeiits varied with yellow marks; sternum yellow, clouded with black ; upperside of abdomen black in its anterior half, with a median yellow stripe, the posterior half reddish brown in the middle, jet-black on the posterior angles ; B large yellowish-green triangular patch spreading in from the sides in front of the posterior augles ; sides of abdomen whitish yellow ; ventral surface yellowish brown, varied with black marlrings. Carapace a little excelling tibia of 1sb lrg in length ; flat above, not elevated posteriorly as in rugosa, the tubercles high and conical ; eyes of anterior line a little more recurved ; median ocular quadrangle inore narrowed anteriorly than in that; species. Legs longer than in rugosa ; the tibia of 1st and 2nd pairs more strongly bowed ; tubercles on femora larger j spines on protarsi more numerous and longer, those on upperside of the segment forming a distinct long pectinated series, which distally largely overlaps the base of the tarsus. Abdomen widest along its posterior border, which is truncate ; width across the middle less than the length ; width across posterior angles exceeding the length ; tubercles large and conical, not low as in rugosa. VuLa consisting of a median auteriorly narrowed piriforln lobe, lodged in a depression bounded by a semicircular border. Measurements in mi1limetres.-Total length 9 ; length of carapace 42, width 3.8 ; length of abdomen 5, width posteriorly 6. LOG. Benito River (C?. L. Bates). Easily recognizable from P. rugosa by the variegated colouring, stronger tuberculation, longer anterior legs, Lc.

Genus THOMISUBLatr.

THOMIltJl TBIPUNUl'ATUS LUCaS. Thomisus trkunctatus, Lucas, in Thornson's Arch. Ent. ii. p. 400, pl. xii. fig. 3 (1863). LOC.Sierra Leone (Sui*g.-Cupt.Clements), and subadult specimens probably referable to this species from the Benito Ever (a. L. Bates). Genus PLATYTHOMISUBDol. PLATYTHOMISUl NIGBICEPS, Sp. n. (Phk?LVI1.T.. fig. 32.) 9. Colozcr. Carapace coal-black, with narrow yellow margin at the sides and posteriorly j mandibles, maxillae, labium, and sternum black; palpi yellow, infuscate quite at tip of tarsus; coxm and trochantere of all the legs yellow j femora of 3rd and 4th ptrirs also yellow except for a black apical ring ; femora of 1st and 2nd pairs deeply infuwate, yellow quite at base ; remaining segments of all 882 ME. R. I. POCOOK ON SCORPIONS, PNDIPALPS, [Nor. 14, the legs jet-black. Abdomeu variegated j upperside yellow, with three pairs of large black patches, the anterior pair united by a narrow bridge, the third pair the smallest and triangular j these are followed by a transversely crescentic black stripe, the concavity of which is posterior; behind this are two median black spots, one behind the other ; on each side of the ribdoinen is a broad black stripe, which unites with its fellow of the opposite side above the pedicel in front and stops short some lirde distance above the spinners behind ; spinners black, with a dorsully incomplete black ring arouud thein ; the ring is connected in the middle line below with a pair of broad black stripes, which unite posteriorly, are separated by a narrow median yellow stripe, and extend from the epigastric fold to tho posterior third ol the lower side of the abdonien ; region of vulva arid lung-books brown, darker mesially. Carapace as wide as long, strongly convex ; cephalic area sloped downwards and forts ards j its total length about equal to tibia of 1st leg and to patella and tibia of 4th. Eyes of anterior aud postcrior lines distinctly recurved when viewed from above; distance between anterior and posterior medians equal to that between anterior and posterior laterals; distalice between posterior medians greater than that between posterior median and lateral on each side ; distance between anterior medians less than between median and lateral on each side; ocular quadrangle much wider thau long, much narrowed in front, the distance between the posterior medians at least one third greater than that between the anterior medians. Legs long ; 1st leg about four times as long as carapace j tarsi of 1st and 2nd nith thick apical scopula; protarsi and tarsi of 3rd and 4th also scopulate j tibia and protarsi of 1st and 2nd uot visibly spined beneath. Abdomen voluminous, not quite twit-e as long as wide, broadest just behind the middle, rocinded in front, narrowed and somewhat pointed behind. Xeasurements in miZ2intetres.-Total length 17; length of carapace 5.8, of 1st leg 22.5, 2nd leg 21, 3rd leg 12,4th leg 145j length of abdomen 12, width S.3. Lor. Benito River (G. L. Bates). A sihgle 0 example.

PLATYTHOMISUSINSIGNIS, sp. n. (Plate LVIII. fig. 33.) 9. Colour. Carapace orange-yellow, with a narrow black rim, incomplete behind but complete along the clypeus ; ocular area involved in a broadiah transverse black stripe ; mandibles yellow, broadly black externally and apically, maxills and labium yellow, the latter infuscate marginally, the former internally at the apex j sternum yellow, with auteriorly and posteriorly incomplete median black liue j pdpi yellow, tip of tarsus blackish j legs coloured as in nigi-iceps, but with more black at the distal half of the femora. Abdomen mostly yellow, the upperside with three pairs of black patches, the posterior pair small and followed by a small median black spot ; anterior surface of abdomen with a broad transverse l899.} AND SPIDERS FROM TBOPICAL WEST AFRIOA. 883 black stripe, which on the sides of the abdomen breaks up into a fan-like arrangement of narrow black longitudinal stripes extending as far as the spinners ; spinners black ; the lower surface of abdomen testaceou8 yellow. Structurally allied to P. niggceps, but the ocular quadrangle wider as compared with its length : legs distinctly shorter, the 1st less than four times as long as the carapace; length of carapace much greater than that of tibia of 1st leg and distinctly greater than patella and tibia of 4th ; tibiae and protarsi of 1st and 2nd legs distinctly spined below; abdointn about one third longer than wide, a long oval, not noticeably wider behind the middle. Mensurements in ntillimetres.-Total length 21; length of carapace 7, of abdomen 14.5, width of abdomen 9.5; length of 1st leg 25, of L?d 25, of 3rd 16, of 4th 17. Loc. Benito River (G. L. Bates).

The previously described african species of this genus are :- P. heraldiws, Karsch, Zeits. ges. Naturwiss. li. p. 315, pl. viii. fig. 4 (1878); Simon, Hist. Nat. AraignBes, i. p. 1016, fig. 1077 (1895). From Zanzibar and Lake Tanganyika. [I have seen no specimens of this species, but it certainly differs from all those enumerated in the table below in having the upperside of the abdomen black with jellom spots, instead of yellow with black spots.] P. sex-maailatus, Simon, in Donaldson Smith's ' Through Un- known African Countries,' p. 388 (1897). Prom Somdiland. P. pant?ierinus, Poc., Ann. t Mag. Nat. Hist. (7) ii. p. 445, pl. xiii. fig. 7 (1898). From Nyasalaud. The four Tropical African species of this handsome geniis, represented in the British Museum by the typical examples, may be distinguished as follows :-

a. Carapace, mandibles, labium, maxillse, and sternum jet-black; lower side of abdomen with a pair of broad black narrowly separated bands, which unite posteriorly ...... nigriceps, sp. n. I. Carnpace principally yellow or red, margined or spotted w:th black, and black in the ocular region; mandibles, maxillse, labium, and sternum principally or wholly yellow ; abdomen without black bands on the middle of its ventral surface. ul. Carapace red or yellow, with a black margin and transverse black frontal stiipe; mandiblee externally black ; black abdominal spots large. I?. Femora of 1st and andlegs deep blackish green ; the black margin of' carapace extending right round beneath the clypeus but not joining the ocular stripe : sides of abdomen ornamented with narrow branching; stripes arranged longitudinally ; no black rings round the spinnera below ...... imigwis, sp. n. 1'. Femora of 1st and 2nd legs clear ycllow; black margiii on carapace not extending on to face, 884 ON SCORPIONSm. FBOM TROPICAL WEST AFRICA. [Nov. 14,

stopping short on a level with the 1st leg; frontal stri involving lateral angles of campace and 8% of abdomen occupied by a continuous blackstripqextending to tbe s inner6 ; spinnera surrounded by ring of black gelow .. . se.zvmmdatwa 8im. b' Oampaoe yellow,not margined with black,spotted with black ; mandibles not black externally ; black abdominal spots emall ...... pafltkinctd Poc.

EXPLANATTON OF THE PLATES. PLATELV. Pig. I. Arawthra butkn' (p. 860). x 4. 2. Bthrob mzmmosu (p. 861). x 3 (nearly). 3. Araw tyloacapus (p. 851). x 2. 3a. ,, ,, vulva from below. 3b. ,, ,, vulva from behind. 4. Cyrtopbra margaritata (p. 854). x 2. 4a. ,, ,, vulva. 6. Cmostris argoslictw, (p. 856). x 2. 5 a. ,, ,, vulva. 5 b. front leg from below. 8. Phalza fero3.i' %3), vulva. 6 a. Lgram of arran ement of eyes. 7. Cdk (&toctmw)ogilior (p. &3), tibial Rpurs of pulp (spurs shaded).

PLATELVI. Fig. 8. Araiieus hatociiemis (p. 850). x 14. 8a- 9, var. 81. 1, var. 8c. ,, vulva, lateral view. 9. Araneus rhinurrus (p. 852). x 2. 9 a. vulva, lateral view. 10. Ga&racant& bat& (p. 865). x 2. 11. Phrmeyscsa baentata (p. 843), palpal organ. 12 & 12a. 8parassua benitenlria (p. 8T5), two forms of vulva. 13. 8parambs rujilatus (p878), palpal organ froin below. 13a. ,, ,, tibial spur of palp. 14. Phyiiarachne maimata (11. 880). x 3.

PLATELVII. Fig. 15. C'wostris turriger (p. @7), side view. x 14. 15a. ,, vulva. 16. ,, db&ens (p. 8%), vulva. 17. Pmcetia longipes (p. 861), vulva. 18. 17ralasmka uineensis (p. 864) vulva. 19. ,, !mom @.\865),luteml view of carapace. 20. ,, auratrcs (p. 866), vulva. 21. ,, Zeownsis (p. 867), vulva. 22. ,) regdia (p. as), vulva. 2% ,, ins&Jnia (p. 868),vulva. 23 a. lateral view of carrrpaca. 24. Dol&eda tra&fya (p. 869), tibid spur of d. 25. Ctenua scopuh (p. 871), vulva. 25. ,, rivubtw (p. 8'72) tibial spur of d. 26a. ,, ,, central portion of palpal organ.

P.ZS.1899 P1 LX,

SSRowardel PJSrmthth M7ntni-i Fr LII II t TADPOLES 1 MICROHYLA OIWATA 2 MICROHVLA(7 SF'? 3 BUFO PENANCEh'S1:i 1899.1 BATEAORIANS OF THE MALAP PBNINSULA AND BIABL 885 PLATELVlIL Fig. 27. r9parasmu, hit& (p. 876), tibial spur of 8. 28. ,, tqtlmafus (p. 877), tibial spur of d. 29. ,, bateei (p. 877). vulva 30. Tmaniu vuriatu (p. 874), tibial rpurs (upper or right, lower or left). 30a. ,, ,, lower tibial s ur from below. 306. ,, ,, palpal organ from below. 30c. ,, . ,, membrane proceee of -.palpal organ. . 30 a. vulra. - 31. Pa&todeS &mwwa (p.-- 879),- 9.. x If. 31 a. vulva. 82. Plat&omisus :i@q~ (p. Wl), Q. x 14. 33. ,, idgnis (p. 882), 9. x 4.

3. Notes on a Second Collection of Batrachians made in the Malay Peninsula and Sam, from November 1896 to September 1898, with a List of the Species recorded from those Countries. By STANLEYSYYTH FLOWER, 5th Fusiliers, F.Z.S. [Received May 29,1699.1 (Plates LlX. & LX.) To the List of 34 Batrachians from the Malay Peninsula published in the Society’s ‘ Proceedings,’ 1896 (pp. 897-914), eight species can now be added, vie. :-&nu kuhlii Schleg., R. macro- dactyls (Gtiinth.), R. larutensis Blgr., Aficrohyla ornata (D. & B.), M. leucostigma Blgr., Bufo divergem Peters, and two large species of Rhacophorus ; making a total of 42 species. In the British Museum Catalogues of Batrachians 11 species hare been recorded from Siam. Nine species can now be added, viz. :--Rnna macrodnctyla (Qiinth.), R. nigrovittata (Blyt$h),Cab- phrynuspleuroatigma Tschudi, Microhylu ornata (D. & B.), M.inornata Blgr., M. pukhra (Hallow.), H. aiatina (Boie), M. berdmorii (Blyth), and Bufo macrotis Blgr. ; making a total, of 20 speciea. This list must represent, however, but a small proportion of the forms which will eventually be found to inhabit this part of the world. In this paper n little-known frog, Rana plicatdla, first made known by Stoliczka, is redescribed, and the tadpolee of the following species are described, I believe, for the firat time:- Ram macrodon, R. tigrina, Hicrohyla ornata, Bufo pwngemk. Besides these there is a very remarkable t,dpole from Penang, the adult form of which, so fr aa I have been able to discover, is undescribed; it is hoped this notice of it may attract other collectors to observe it and to find out to what species it belongs. I have to acknowledge my sense of obligation to Mr. G. A. Boulenger, F.R.S., for his invaluable advice and very kind help in answering many questions for me in letters during the last three years. 886 MR. STANLEY 9. FLOWER ON TIlE [Nov. 14,

Order ECAUDATA.

Family I~ANID.~.

1. O~~c~ossusLIMA (Qravenh.). 0.rygbssus limn, Blgr. cat. Batr. Sal. p. 5. I‘ &to’ limpong ” of the Malays of Kedah. Localities. In the P. Z. S. 1896, p. 897,l wrote :--l‘This species is said to occur in the Malay Peninsula, but I have not been able to find it recorded south of Tenasserirn, though it occurs again in Java ” ; but since then, in June 1898, I found it numerous near Alor Star and at Jenan, in the State of Kedah, Malay Peninsula. M. Mouhot obtained specimens in Siam and Cambodia. I obtained specimens at the following places in Siam:-one Bangkok (July), many Sapatoom (August), niany Ayuthia (December), one a few miles north of Ayuthia (February), many Pakpreo (June), and one Bawtong Kabin (March). Habits. This is a thoroughly aquatic frog, to be found in small ponds. Colour (in life). Above olive-brown (Ihbin specimen dark, Pakpreo ones light), lighter and greener on limbs, with or without a pale yellow vertebral line ; beneath yellow, more or less handsomely marked with dark olive-brown, including a narrow line along middle of abdomen; back of thigh conspicuously marked with two yellow lines or a broad black one. Iris : narrow gold line surrounding the black diamond-shaped pupil, remainder brown. A particularly handsome individual from Bangkok was coloured in life :-Above olive-brown mottled with dark brown, a broad black-edged vertebral line grass-green anteriorly, shading to olive- green posteriorly; beneath pale yellow, a pair of very distinct black lines from the chin to the breast (the tubercles on these lines form small yellow spotR), a black line beneath each arm, three longitudinal irregular dark brown lines on each side of the body, an L-shaped black line on each side of the base of the thighs, a very broad black line along the hinder side of the thigh, above it being two narrow parallel ones, the interspaces forming pale yellow lines. The eyes of 0. limn in life are very prominent and look upwards and outwards. Size. The largest specimens measures snout to vent 33 mm. Dietribution. Lower Bengal, Burma, Southern China, Siarn, Cambodia, Cochin China, Malay Peninsula, Java.

2. Ox~a~ossusLBTIS Giinther. O.m~gl08SUShis, Blgr. Cat. Batr. Sal. p. 6 ; Blgr. P. 55. S. 1597, p. 228 (tadpole). In the Museum at Taiping there are specimens from swamps 1899.1 BA’PBACHIANS OF THE MALAY PBNINSULA AND SIAM. 887 near Taiping, Perak; and Ur. Hauitsch (Rep. Libr. & Mus. 1898) records the species from swamps near Iph, Pen& (March). fitribution. Burma, Malay Peninsula, Sumatra, Sipora (Menh- wei Islands), Borneo, Flows, Celebes, Philippines.

3. Ox~o~~ssnsMAETEXSII (Peters). PhynogZossue martenuii, Peters, Mon. Berl. Ac. 1867, p. 19. O.~ygZossusmarteilsii, Blgr. Cat. Batr. Sal. p. 6. Localities. I obtained specintens at the following places in Siarn :-one Bawtong Kabin (March), one Chantnboon (January), many in the Dong Phya Fai (November), elevation 900 feet, and a few near Ayuthia (February and June). Habits. This species does not seem so thoroughly aquatic as 0. lima; out of seventeen individuals I caught at the above places only two were actually in ponds, though all were near water. Cbbur(in life). Sometimes a narrvw yellow vertebral line is present. Behind the eyes a broad transverse patch of red-brown and orange is couspicuous. lris yellow, closely and finely speckled with brown. The diamond-shaped pupil is dark led in colour. Size. The largest specimen measurea snout to vent 27 mm. fitributwn. Sinm.

4. HANACYANOPHLPCTIS Ychneider. Rana leecherzaultii, Cantor, p. 138. Rana cyunophZyctis, Blgr. Cat. Batr. Sal. p. 17; Anderson, P. Z. S. 1895, p. 660, pl. xxxvii. fig. 2 (tadpole). This frog does not seem to hove been observed in the Malg Peninsula since Cantor’s time. Habits. South Arabia, Baluchistan, Cashmere, Himalayas (up to 6000 ft.), India, Ceylon, Malay Peninsula.

5. RANAEUHLII Schleg. Rana kuklii, Blgr. Cat. Batr. Ssl. p. 20; P. Z. 8. 1809, p. 166. hxzlities. This frog, not previously recorded from the Malay Peninsula, waa found by Dr. Hanitsch in Perak in 1897. In April 1898 I obtained four s imens in the same locality, Maxwell’s Hill (Larut BTs, Perag.. ), at an elevation of 3500 feet. Colour (in life). Above olive-green, with irregular, indistinct small blmk spots; a distinct black transverse line between the eyes, the skin immediately in front and behind this being yellowish olive. Limbs yellowish olive, extensively s otted with dark olive- brown, which spots have a tendency to Porm cross-brs. Chin and sides of head, neck, body, and limbs chrome-yellow. Lips extensively mottled with dark olive-brown. Lower surfaces bufl. Tympanic fold black. Web of hind feet pale yellowish olive. Iris : a narrow rim of gold rouud black pupil, remainder bronze with a black cross (much as in Rana mmh). ,Si:c. The largest measured : snout to vent 90 mm. ; width of 888 MR. B'l'ANLEY 8. FLOWBR ON THE [Nov. 14, head 38 mm. ; srm 46 mm. ; leg 115 mm. The largest 2 measured : snout to vent 70 mm. ; width of head 67 mm. Dktribution. Southern China, Burma, Malay Peninsula, Sipora (Mentawei Islands), Java, Borneo, Celebes.

6. RANALATICBPS Blgr. Rana laticeps, Blgr. Cat. Batr. Sal. p. 20, PI. i. fig. 1 ; S. Flower, P. Z. 8. 1896, p. 897. Dr. Haoitmh (Rep. Raffles Libr. & Mus. 1898) records obtaining Ram Zatieeps from Gunong Kledang, Ipoh, Perak, in March 1898, at an elevation of 2000 feet. Distribution. Khassya, Bengal, Malay Peninsula. 7. RANAMACBODON Kuhl. (Plate LIX. figs. 1, 1a) Rana macrodon, Blgr. Cat. Batr. Sal. p. 24. In the P. 8.5. 1896, p. 898, I pointed out that there appear to be two forms of this species in the Malay Peninsula ; since then I have seen a large number of these frogs, all of which were referable to one or other of the varieties, but I found large specimens of the Penang variety may approach the 8ingapore variety in colour (having the upper parts reddish or brownish yellow), and in the distance of the nostrils apart being less than the interorbital space. The stomachs of these big frogs generally contain a good deal of foreign matter, bits of leaves, small twigs, snnd, fine granite, gravel, and angular pebbles as much as 10.5 mm. in diameter. Their food consists of snails, crabs (Thelphusidae), caterpillars, beetles, crickets, ants, &c., and twice I have fouud the remains of scorpions (Homurus sp. incert.) in their stomachs. Sometimes there are a number of parasitic worms between the kidneys and the back. The fang-like bony prominences in the lower jaw are sometimes very sharp and nearly 5 mm. in length. huditiee. Penang Hills, 1900 to 2200 feet (November 1896, March and April 1898) ; Laruti Hills, Perak, 3200 to 3400 feet (April 1898): in both of these localities the Penang variety is numerous about water (mountain streams and pools) and grows to a large size. I also got two very typical specimens of this variety near the foot of Gunong Phi, Johore (September 1597), from which State this species had not previously been recorded ; and it is interesting to thus find the Penang variety at a low elevation (about 200 feet) and so near the island of Singapore. So far I have not come across the Singapore variety on the mainland or at Penang. Cohur (in life). One specimen from Penang was unusually coloured, the upper parts being rich bronze-red and the linibs handsomely marked with yellow and dark brown ; below (as usual) the chin wafi white and the remainder pale orange. Iris (noted from many specimens) : a narrow ring of bright gold round the 1899.1 BATRAOHIANS OF THE MATAY PININSULA AND SIAX. 853 black pupil, remainder pale goldon finely speckled with black, and a broad horizontal line a,nd a narrow perpendicular line forming a black cross on the eye. Size. 8. Snout to vent 120 mm. : width of head 48 mm. ; leg 183 mm. 9. Snout to vent 88 mm. : widkh of head 30 mm. These measurements are from Perak specimens, Penang ones attain an equal size. Singapore specimens reach snout to vent 165 mm. : width of head 76 nim. (P. Z. Y. 1896, p. 901). Distribution. Upper Burma, Tenasserim, Mday Peninsula, Sipora (Mentawei Islands), Java, Lombok, Flores, Natunas, Borneo, Philippines.

Tadpole. In March 1898 I found tadpoles of this species iu the clear, swift hill-streams of Penang, about 1900 feet above sea-level ; they frequented the edges of the stiller pools. They are typical Rana tadpoles in habits and appearance ; the mouth, I find on comparing my drawing with the figure by Boulenger (P. Z. S. 1891, pl. xlv. fig. 1a), seems similar to that of R. escuknta.

Deswiptiow of the Tadpole. Leiigth of body about once and a half its width, about half the length of the tail. Nostrils about halfway between the eyes and the end of the snout. Eyes on the upper surface of the body, nearer the end of the snout than the spiraculum, the distance between the eyes once and a half to twice as great as that between the nostrils, and slightly greater than the width of the mouth. Ypiraculum on the left side, pointing backwards and upwards,nearer the anus than the end of the snout, visible from above and from below. Anus opening on the right side, close to the lower edge of the subcaudal crest. Tail about four times as long as deep, acutely pointed ; upper crest convex, slightly deeper than lower, not extending to base of tail ; the depth of the muscular portion of the tail, at its base, about & of the greatest total depth. Mouth. Beak broadly edged with black. Sides and lower edge of the lip bordered with papillae ; upper lip with a long series of teeth, followed on each side by a short series of very fine teeth. Lower lip with three series of tedth : 1st very short, weak and uninterrupted ; 2nd also uninterrupted ; 3rd longer, narrowly interrupted and stronger than the two outer rows. Colour (in life). Above light reddish brown, of exactly the same colour as the sand on the bed of the streams they inhabit, mottled in places with darker brown ; a dark brown line through the eye, sometimes other dark lines radiating from the eye. Crests and posterior portion of muscular portion of tail colourless, with $90 MR. ST.LNLEP 8. FLOWER ON THE [XOV. 14, irregular dark brown vertical bars. Belly buff and transparent, showing the coil of the intestines. Iris golden, with black cross marks. Size. Largest specimen measured in total length 34.25 mm. ; length of body 11.25 mm. ; width of body 7-75 mm. ; length of tail 23 mm. ; depth of tail 5.5 mm. The recently transformed young measure from 9.5 to 11 mm. from snout to vent. The smallest frog which bad the yellow vertebral line developed measured 13 mm. from snout to vent.

8. RANAPLIOATELLA Stol. 1Zann ylimtelln, Stol. J. A. 8. B. 1873, p. 116, pl. xi. fig. 1 ; Blgr. Cat. Batr. Sal. p. 26. Rana plicatella waa described by Ytoliczka froin a single specimen (apparently), which had been at least two gears in RpiPits; its locality was either Penang or Province Wellesley : unfortunately he does not mention either the sex or size of the type specimen. A frog caught by me in Penang I refer to this species, as it entirely agrees with his description and figure, except in the following points :-lst, a larger head; 3nd, the presence of a knob- shaped prominence on the occiput ; 3rd, the upper surface of the body has about 12 longitudinal folds instead of 8 ; 4th, difFerences o€ colour, mostly attributable to Stoliczka’s heiiig a spirit-specimen, except that mine has 5 transverse dark bands 011 the femur instead of 6, the “ horseshoe-shaped yellow mark, open below, round the anus” is absent and a pale vertebral line is present. The 3rd and 4th points may well be individual variations, while the 1st and 2nd may be secondary sexual characteristics, my specimen being an adult male. This frog resembles Rana Zuticep Blgr., known from India and M~WCB,and Ram dorice Blgr., from Burma, in many respects; but differs from the former in, lst, the distinct and larger tympanum ; 2nd, the less broadly webbed toes : and from both in, lst, the distinct longitudinal glandular folds on the back ; 2nd, the prominence on the occiput. Description. Vomerine teeth on two straight ridges, commencing on a line with hinder edge of choanae and running obliquely back from them, converging behind so as to meet, if prolonged, in an angle rather greater than a right angle. Lower jaw with two fang-like bony prominences in front. Head very large ; snout bluntly pointed ; canthus rostralis distinct, rounded ; loreal region slightly concave ; occiput swollen at the sides, interorbital region convex, the swelling produced posteriorly and ending in a knob- shaped prominence 1.5 mm. in height ; in life this feature is very prominent and at once attracts the attention. Nostrils lateral, somewhat directed upwards, rather nearer the end of the snout than the eye, their distance apart is slightly greater than the interorbital space. The breadth across the gape is much greater than the distance from angle of mouth to end of enout. Eye prominent. The interorbital epace is much greater than the width of the upper 1839.1 B1TllACHIdNS OF THE MALAY PEXlXBlJLA AND SIAM. 691 eyelid. Tympanum distinct, slightly larger than the eye. A prominent tympanic fold. Fiugers moderate, 1st slightly longer than 2nd ; toes about three-quarter webbed, the web reaching on the 4th toe “ to scarcely beyond the basle of the third ultimate joint,” on the otber toes to the base of the tips, it is deeply emarginate. Tips of fingers and toes dilated into amdl but distinct disks. Tarsal fold very Blight. Suharticular tubercles of toes well developed. Inner metatarsal tubercle elongate, pTomiuent : practically uo outer tubercle, a mere white spot (this can be distinguished iu Stoliczka’s figure, dorsal aspect, on the right foot). The hind limb being carried forward along the body, the tibio-tarsal articulation reaches to the nostril. Head smooth above, hinder half of the eyelid tuberculated : skin of back with long longitudinal folds, with numerous smdl tubercles between thetn j loreal region, chin, angles of jaws, parts of forearm, and sides of body scattered with small rounded tubercles ; anal region densely studded with small tubercles ; back of tibiae scattered with small tubercles, which are pointed and prominent in life; back of thighs, feet and lower surfaces of throat, body and limbs smooth. Testicles white, 2.3 mm. in length. Size. Length, snout to vent 39 mm.; arm 24; leg 68 ; femur 17.5; tibia 21.5 ; foot 29 ; width of head 19 ; diameter of eye 4.23 ; diameter of tympanum 4.5 ; length of inner metatarsal tubercle 3. Colour (in life). Above bronze (changing from bright yellowish red to dull brown shades), with a bright yellow narrow vertebral line, a black chevron (pointiog backwards) between the ejw ; the tympanic fold and some of the longitudinal folds are edged with black. Limbs with dark brown cross-bars. Belly very bright sulphur-yellow ; underneath of chin, t.hroat, and limbs pellowish bronze. (In spirit the bronze colour of the upper surfaces turned to olive-brown.) Eye. Pupil black (dark red in some lights), horizontal diamond- shaped. Iris : a very iinrrow ring of red round pupil, remsinder golden, closely speckled with dark bronze, a narrow black vertical liiie across iris and also a broad ill-defined black horizontal one. One specimen, d,found in the jungle on Penang Ell, at an elevation of about 2400 feet, 19th March, 1898. It took wonderful long hops. Distribution. Malay Peninsula (Penang). 9. RANATIGRINA Daud. (Plate LIX. iigs. 2, 2(1.) 22ana tigrim, Blgr. Cat. Batr. Sal. p. 26; 5. Flower, P. Z. S. 1896, p. 901. Siamese. “ Kop ” : term also applied to other species of frogs. J found this species common on the Bangpakong river (from Tahkumen to Kabin) in March and April 1897, where it ww to be found among the sedges by the river’s bank ; itH croak (a loud “opp, opp, opp,’’ repeated 8 or 9 times), which sometimes was 898 Ylt. STAXLEY 9. FLOWER ON THE “ov. 14, heard in the middle of the dav, betraying its presence. During the dry season I have not seen this frog at Bangkok, but from the middle of May to August they are very plentiful in suitable ponds and are to be bought in the market, as R. tiyrino is eaten by the Siamese ; I have heard Malays speak of the Siamese by the name (* Frog-eaters,” in consequence of this. Small Siamese boys may be seen fishing for these frogs with a short rod and a very fine line with a small hook with a white bait : this is danced and dangled over the surface of the water near the edge of a pond 5t9 if it were a fly hovering about : when a frog springs at this the angler has to ‘(strike ” very quickly, to catch the hook in the frog’s mouth to draw it out of the water. I also found this species at Ayutbia in June. Breeding-eeccson. A pair were observed in copuld in Bangkok in May, the male was considerably smaller than the female j during the latter half of July females were to he found with the ovaries distended with spawn, ready for expulsion. But well-grown tadpoles were found as early a3 the first week in June, in Aynthia, and on the 28th July in Bangkok, when young frogs were leaving the water. Thus the breeding-season must extend over some months. Colour (in life). Description dmwn LIP from adult specimens of both sexes caught in Bangkok in May :- Above olive-brown or pale olive-green, with very dark greenish- brown or black spots on the head, sides, and limbs ; the back either spotless or with large dark spots; on the legs the s otu may become broad dark transverse burs : no vertebral line. BeP ow, body and limbs immaculate silvery white, between which and the green upper surface there is an irregular spaee of lemon-yellow with greenish-brgwu spots. The lips, chin, and uuderneath of head are lemon-yellow, with more or less distinct dark spots; sometimes there is a median dark line from the synipbysial angle to the breast. The vocal sacs of the male are dark grey, shaded with pink. Iris dark brown, with narrow golden ring round the black pupil. Half-grown specimens are often prettier coloured than the adults, being rich green above and pure white below, with many Mack markings. Size. An adult $? from Bangkok measured snout to vent 142mm.; another individual (sex not recorded) 153 mm. Distribution. Nepal, Sikhim, India, Ceylon, Burma, China, Formosa, Siam, Malay Peninsula, Java, Borneo, Celebes, Philip- pines, Lombok, Ombaai, Sumba. Description of the lhdpole. Drawn up from specimens obtained in June 1897 at Ayuthia, Sam. Length of body once and a half its width, rather more than half length of tail. Nostrils a little nearer to the eyes than to the end of the snoub. Eyes on the upper surface of the body, nearer the 1899.l BATRACHIANS OF THE MALAY PNNINMJLA .4ND SIAM. 893 end of the snout than the spiraculum, the distance between the eyes twice the distance between the nostrils, and about equal to the width of the mouth. Spiraculum on the left side, directed backwards and upwards, a little nearer the anus than the end of the snout, visible from above and from below. Anus opening on the right side. Tail from 3$ to 4 times as long as deep; acutely pointed; upper crest convex, a little deeper than the lower, not extending on to the back ; depth of the muscular portion at its brave abont half the greatest total depth. Mozcth. The large powerful beak is entirely black j the upper mandible terminates in front in a long sharp tooth-like prominence ; the lower mandible is bicuspid, each ‘‘ tooth ” being long and sharp. The lips are bordered with very short fleshy papillae: inside the upper lip are five series of fine, black teeth; the 1st series is uninterrupted, the second slightly interrupted by the individual teeth being (‘grouped with intervals ’’ about the centre of the line ; the remaining series are broadly interriipted, the 5th being very short and difficult to distinguish : the lower lip has also five series of teeth; the 1st is short and uninterrupted, the second long and uninterrupted, the remainder broadly interrupted and very short. Colour (in life). Above yellowish brown, mottled with darker brown, a very distinct dark brown crescent-shaped mark above each nostril. Below white, purplish grey about the chin and throat. Tail yellow, mottled with brown, an horizontal dark line along the median line of the basal third of the muscular portion. Iris golden. Size. Total length 54 mm.; length of body 18.5 ; width of body 12.5 ; length of tail 33.5 ; depth of tail 9. Habits. Those of ordinary Rnnn tadpoles.

10. RANALIMNOCHAEIS Boie. Ram gracilis, Blgr. Cat. Batr. Sal. p. 28. Localities 4 Hnbits. This species is very numerous in Bangkok, where I have observed it in the monthR of Jan., Feb., Mar., June, July, Bug., Oct., Nov., and Dec. Small specimens abound in the evening, hopping about the grass, they are very active. In the hot weather, in spite of the burnt up condition of the ground and grass, these frogs still appear at night, and, when one tries to catch them, take refuge down the sun-cracks in the parched earth. Specimens over 50 mm. snout to vent are of comparative rare occurrence. I have also found this species common along the Bangpakong river, at Chantaboon, Paknam Menam, in fields beyond Sapatoom, Ayuthia, Pakpreo, Dong Phya Fai (up to 900 feet elevation), in Siam ; at Taiping, Kuala Kangsar, and Chumar in Perak, and at Alor Star and Jenan in Kedah. In a former paper (P. 8.S. 1896, p. 902) I wrote of this species PROC.ZOOL. SOO.-1899, NO.LVIII. 58 894 ME. BTMLBY 8. FLOWm ON “HB mov. 14,

at Singapore :-cr It does not attempt to escape by jumping into the water.. . .but even if touched squats down close. . . .so is easily caught.” The individuals I met at Chumar had this habit, but elsewhere, at Taiping, Bangkok, Bc., I found them very agile and difficult to snare. In captivity they feed readily, eating insects in the same manner as R. temporniia does; winged termites they devour in large numbers and ail1 also manage grasshoppers of comparatively large rize ; when suddenly seized in the hand or when caught by a snake, they utter shrill piercing shrieks of alarm. C’oZoozc~ (in life). Bangkok specimens not nnfrequently have n very distinct grass-green vertebral line, others none at all j in the Dong Phya Fai the commonest colour-variety was one I have not seen elsewhere, there being transverse bands of bright grass-green across the back, but specimens with yellow vertebral lines were also to be seen. The sexes are coloured alike, except that the male may have a broad black M-shaped mark on the throat. The following description applies to Taiping specimens :- Above olive-brown, irregularly mottled with darker ; vertebral line either absent, or 11 ell defined, narrow and yellow, or irregular, broad and orange-coloured. Limbs extensively marked w ith dark brown ; hinder portion of thigh yellow, marbled with dark brown. Underneath of head, neck, and body pure white, but lower surface of limbs yellow. Both upper and lower lips white, with large distinct dark brown blotches ; the point of the snout white, with a well-defined very dark brown blotch on each side. Size. A pair caught in eopulh in Bangkok, 24th July, 1898, measured :- 6.Snout to vent 46 mm. Q . Snout to vent 52 mm. Distribution. Sikbim, India, Ceylon, Burma, China, Hongkong. Hainan, Formosa, Japan, Yiam, Malay Peninsula, Java, Lombok, and Borneo.

11. RANAHASCHEANA (Stol.). Polyped

12. RANAERYTHEBA (Schleg.). Rana erythrcea, Blgr. Cat. Batr. Sal. p. 65. “Katak pisang” of the Malays according to Mi.. L. \Yray. Personally 1 think ‘(Kata’ pisang ’’ (or Banana Frog) is the proper Malay name for a Bhacophoms, but no doubt any frog with large digital disks would he considered a tree-frog and called by tlie same name. Localities 4 Habits. I found this pretty frog fairly numerous on the Bangpakong river, from Tahkamen to Kabin, duriug March and April 1897, in places where there was long grass on the banks ; also in ditches at Sapatoom, near Bangkok, in Jan. 1898, where its green aud yellow colours exactly matched the leaves of the water lettuce (Pistin stratiotes’) among which it was to be found. In Taiping, Perak, in April 1898 it was very numerous in ponds and ditches j two individuals were caught in Penang, April 1898, and I also observed it near 91or Star, Kedah, in June 1898. It is a true water-frog, and appears to occur only at low elevations. Colour. The Bangpakong frogs were more ornamental than Singapore specimens (vide P. 8. S. 1896, p. 203, pl. xlv. fig. 2), although the scheme of coloration was the same. They were as follows in life :- Above brilliant green ; along each side a line of similar green, separated from the green back by a broad light yellow line (from above the eye to the side of the vent), bordered above and below by broad lines of intense black ; below the lateral green line is another line of pale golden yellow (from the angle of the mouth along the side, gradually narrowing and disappearing before it reaches the inset of the hind legs), bordered above by a broad line of intense black aud below by an irregular line of dark green with iridescent golden shades. A black line on each side of the head from snout to eye ; upper lip very pale golden yellow ; tympanum dark rufous brown, with a bright green spot in the centre. Lower parts white, with pale iridescent golden shades. Limbs as in Singapore specimens, above yellowish brown speckled with dark brown, below immaculate buff. Iris: very narrow golden ring round pupil, remainder golden-bronze finely speckled with black. Size. The large Siamese specimen measured, snout to vent 74 mu. Distribution. Burma, Siam, Malay Peninsula, Sumatra, Borneo, Celebes, Philippines. 13. RANAWAUBODAOFYLA (Giinther). Eanu naacrodactykr, Blgr. Cat. Batr. Sal. p. 54. Loodities. Of this species, which had hitherto been recorded 1 Kindly identified for me by Mr, H. N. Ridley, F.L.S. 58’ 896 'MR. STARLEY 8. FLOmON vov. 14, from Burma and Southern China, I obtained many specimens near Jenw and Alor Star, Kedah, in June 1898, and near Sapatoom, Bangkok, Siam, in A-ugust 1898. Habits. In each case they were living among swampy paddy- fields ; they sat in the grass on the ridges between the submerged fields, and as one walked along took great leaps away into the water, which, however, they did not seem to like, as they nearly always at once suam back to the bank. They are remarkably nimble, active frogs, and specimens I had in captivity used to climb up a vertical surface of glass like a true tree-frog. Colour (in life). Above rich dark brown, spotted with black, and in some individuals mottled with dull yellow and vivid green, with five very distinct longitudinal lines, which are white with golden shades (the centre line is from the snout to vent, the next pair from the upper eyelids to the sides of the vent, and the outer ones from the tympanum to the inset of hind leg). Below, head and body white, limbs yellowish. Upper Rurface of limbs reddish yellow, handsomely marked with dark brown, three lines along the back of the thigh being most conspicuous. his golden. Size. Snout to vent 38 mm. ; hind leg 63 mm. Distribution. Burma, South China, Siam, Malay Peninsula. 14. RANANIGROVITTATA (Blyth). Rana nigrovittata, Blgr. Ann. Mus. Genova, (2) xiii. 1893, p. 334, pl. viii. fig. 3. This species, not hitherto recorded from Siam, I found very numerous along the banks oE the river at Muok &ek in the Dong Phya Fai, elevation 900 feet, in November 1897 ; both during the heat of the day, in the afternoon, and in the evening they kept up a continuous, loud and rather musical croaking, something like that of Rhacophorccs leucomystua, only the note is louder and uttered much more often. They were very active. Size. The largest specimen caught measured from snout to vent 58 mm. Distribution. Burma, Siam.

15. RANALABIALIS Blgr. Rana labialis, Blgr. Ann. & Mag. Nat. Hist. (5) xix. 1887, p. 345, pl. x. 6g. 1; S. Plower, P. Z. 5. 1896, p. 903, pl. xlv. fig. 3 (tad pole). In Sept. 1898 in Singapore I saw several specimens of this frog both in the Botanical Gardens and in the jungle on Bukit Timah ; it was in each case observed sitting on the leaves of plants or in bushes, so evidently is not a true water-frog like R. erythrma. Rana labialis can change its colour rapidly from green to brown. Distribution. Malay Peninsula and Mentawei Islands. 16. RANALUCTUOSA (Peters). Rann luotuosa, Blgr. Cat. Batr. Sal. p. 68. Localities. Common on Penang Hill, elevation 2000 to 2200 feet (Nov. 1896 and March 1898), rand obtained by Mr. A. L. Butler 1899.1 BAFRAOI~~ANSOP vrn ~A~AYPSAWSULA AND SUM. 897 in the Larut Hills, Perak, elevation 4000 feet, in March and April 1898. Colour. In the P. Z. 5. 1896, p. 904, I described the life coloration of this species, but haling since then examined more specimens I think the following account to be more complete :- Top of head and back rich dark chocolate-brown (in very small frogs of this species the back is a very bright red, more vermilion than chocolate), bordered on each side from the nose to above the vent by a very distinct line, usually all white but sometimes white only on the head, merging on the body to yellow and then orange. Sides of head, neck, and body are very dark brown or black : somp- times a very distinct line of lemon-yellow spots from behind the nostril to angle of mouth, smaller anteriorly and getting larger posteriorly; sides of the body with a few white spots iii nu irregular line from angle of mouth to thigh, or else evteiisivelg spotted with small white or yellow spots. Tympanum dark reddish brown, sometimes nearly black. Limbs very dark brown or bluish black, with marblings iisually bluish white or very pale grey in colour but varying from white to orange ; the hands, feet, and toes may have as dark a ground-colour as the limbs and as distinct light marblings, or the black may turn to brown and the marblings be less conspicuous. Lower surfaceB, chin, and throat dark brown, sometimes near17 black and immaculate (March), or dirty buff like the abdomen (Nov.); body dirty huff or dark brown mottled with yellowish buff j limbs brown, sometimes spotted with white. Iris pole golden bronze, extensively marked with very dark brown. Size. An adult 9 from Penang measured:-snout to vent 51 mm.; arm 30 mm. ; leg 82 mm. Diekibtctioa. Malay Peninsula, Borneo. Tadpoles. Both in November 1896 and in March 1898 I found many tadpoles of this species in small ponds on Penaug Hill, but none with legs developed (though in March 1896 there were tadpoles with legs and also recently transformed young frogs about, in the same locality); they agreed with those described and figured P. 2;. 5. 1896, p. 904, pl. xlvi., except that some had about 8 lony papillre along edge of lower lip.

17. RANAGLANDUMSA Blgr. Rana ~/luntlulo.sic,Blgr. Cat. Batr. 8a1. p. 73, pl. vii. Known from Perak (specimen in Museum at Taiping), Malacca ( Hervey), and Singapore (Riclley). Distribution. Malay Peninsula, Borneo, Palawan.

1s. RANAESCULENTA L. Bnna esculenta, Blgr. Cat. Batr. 8al. p. 38. The British Museum contains specimens of the variety chirceiayis Osbeck, from Bangkok, Siam. Distiibutiow. Central and Southern Europe, Northern Africa, Western Asia, Cora, Japan, China, Siam. 898 MR. BTAIiLEIY 8. FLQWMR ON THE [Nov. 14,

19. HANALARUTEN~IS Blgr. Rana larutensis, Blgr. Ann. & Mag. N. H. (7) iii. 1899, p. 273, pl. xi. fig. 1. In April 1898 these frogs were numerous sitting on the rocks in a swift rocky mountain-stream in the Larut Hills, Perak ; their colour harmonized wonderfully well with their surroundings. They were difficult to catch, being exceedingly active jumpers and climbers. C'olour (in life). Above pale yellowish green, the head and body very extensively blotched with black, the limbs with black transverse bars. Below, head and body pure white, limbs pale green and grey, underneath of hands and feet very dark, and web between toes black. Ins : very narrow gold ring round pupil, remainder dark olive-brown. Size. Snout to vent 60 mm. Distribution. Malay Peninsula.

20. ? RRAUOPHORU~REUTICUS (Peters). hliucwphorus ?hecticus, Blgr. Cat. Batr. Sal. p. 78. In the Museum at Taiping there is a large tree-frog from Iciuta, Perak, which is perhaps of this species. Uistribution. Samar Island in the Philippines, and possibly Malay Peninsula.

21. BHACOPI~ORUSLEUCOYYBTAX (Gravh.) (Plate LIX. figs. 3, 3a.) Rhacophorus nuzcukctus, part., Blgr. Cat, Batr. Sal. p. 83. " Kata' pisang " of the Malays. ['' Pisang "=banana.] Coewlities. Since writing of this frog in the P. Z. S.I8FKi, p. 905, I have observed it in the following localities: Sia~i-Bangkol~ (Jnne and July), Kabin (March), and obtained one pouug specimen from Chantaboon : Malay Peninsula-Singapore (March, April, May, September, and October); Larut Hills, Perak, 3400 ft. (April); Penang Hills, 2200 ft. (March, April, and November) : Alor Star, Kedah (May and June); and there are specimens from Kuala Lumpor, Selangor, in the Museum at that place. Hrtbits. This species apparently breeds at various times of the year, specimens are frequently to be seen in the evenings in c0pu.U on the edges of the rain-water butts of houses, both in March and April (Singapore and Penang Hills) and in October (Singapore), and probably in other months also. The spawn %oats on the surface of the water enclosed in an envelope resembling white foam. I have noticed tadpoles in the following months: January, February, March, April (Singapore); May (Kedah) j June and July (Bangkok) ; November (Penang) ; December (Singapore) ; and newly transformed young were just leaving the water in September 1898 in Singapore. C'olour. Beuides the six varieties of colour mentioned P. Z. 8. 1896, p. 906, I have observed in one cam on Penang Hill another 1809.1 BAPBAOEIILNB OF mm MALAY PIDNINSULA AND AM. 890 very notiiceable one :-Upper parts pale cream-colour, a black line along each side starting from behind the eyes, some black spots on posterior part of back, and brown cross-bands on limbs. Newly transformed young in Bangkok (June) were coloured as follows :-Above pale olive-brown, with on each side a black line from the snout passing through the eye to the inset of the hind leg ; a black spot on each supraorbital region prolonged backwards to the neck as a black line, these lines converge but do not meet ; ou the shoulders are two diverging black lines ; the remainder of the back is spotted with black ; the hind limbs are transver,sely banded with dark brown ; the lower surfaces are white. Size. The largest of a series of adults examined on Peiiang Hill in March 1898 measured :- d . Snout to vent 51 mm. : hind leg 77; testicles, length G mm. 2 . Snout to vent 73 mm. ; hilid leg 107. Largest eggs in ovaries 2 mm. in diameter. Distribution. Sikhim, Aesam, Burma, Southern China, Hoiigkoilg, Formosa, Cambodia, Cochinchina, Siam, Malay Peninsula, .Javn, Yumba, Borneo, Celebes, Philippines. Tadpoles. Specimens from Bangkok agreed with the dsscriptioii (P. Z. 8. 1896, p. 906, pl. xliv. fig. 2), except for some differences in the mouth and in colour, which were as follows :- Mouth. Beak broadlg edged witb black, lower mandible finely serrated along the cutting-edge ; sides and lower edge of the lip bordered with small, ehort, round papillae, except in the centre of the lower lip, where there is a space devoid of papille. Upper lip withJive series of fine teeth, the uppermost uninterrupted, the second narrowly iuterrupted, the remainder broadly so, the fifth series is very short ; lower’lip with three long series of teeth, third narrowly interrupted, the other two uninterrupted. Colour (in life). Above yellowish brown, mottled darker and lighter; below white, purplish grey on the throat; muscular portion of tail very pale brown, crests colourless and transparent j a very noticeable light yellow spot on the point of the snout ; a dark brown line from corner of mouth to eye. Iris yellow. A specimen from the Waterfall Gardens, Penaug, agreed with these in having the 3rd series of teeth in the lower lip narrowly interrupted in the middle, but differed from them and from thone originally described from Singapore in having at each of the upper corners of the mouth four loug fine papillae.

22. XHAUOPHORIJSNIGROPALMATUS (Blgr.). Rhacop?~omsnigropakncitus, Blgr. Ann. C Msg. N. H. (6) xvi. p. 170 (1895). In the Museum at Taiping is a single specimen of a large frog caught in Upper Perak by Mr. L. Wray (jun.); by his kind per- mission I was allowed to examine it. Description. Vomerine teeth in a straight line between the front edges of the choanae. Snout rounded. Centhus rostralis distinct. Loreal region decidedly concave. Nostril nearer the tip of the snout; than the eye. Interorbital space broader than tbe upper eyelid. Tympanum half the width of the eye. Fingers and toes 900 MR. STANLEY 8. FrBWEk ON T- [Nov. 14, webbed to the disks, which are as large as the tympanum. Sub- articular tubercles distinct ; a small inner metatarsal tubercle. The hind limb being carried forward along the body, the tibio-tarsal articulation reaches nearly to the nostril. Skin smooth, granulated very markedly on the belly and under the thighs. Fold round tympanum not noticeable. A large flap of skin behind the forearm. Cutaneous fringe on both 1st and 4th finger, and 1st and 5th toe. A transverse flap of skin above the vent, and a flap of skin on the tibio-tarsal articulation. Length, snout to vent, 98 mm. Mr. Wray sent me the following note about this specimen, which has been identified by Mr. Boulenger. “ Flying Frog collected in Piali Valley, Upper Perak : above, it is a lovely bright green, with hands and feet tinted yellow and a white patch on each thigh. Below, it is pink dotted over with yellow. The sides of the body are chrome-yellow, and its webbed hands and feet are yellow and black.” 111 answer to further enquiries Mr. Wray tells me he did not see it fly, but caught it sitting on a tree. The name “Flying Frog” seems to be taken from A. R. Wallace (‘The Malay Archi- pelago,’ edition 1894, fig. on p. 30). Distvibution. Borneo and Malay Peninsula.

23. BHACOPHORIJSLEPROSUS (Schlegel). lthacophorus lep1-08~6,Blgr. P. Z. S. 1890, p. 284. This species has been obtained by Mr. Wray in the Larut Hills, and the Batang Padang Mountains, Perak (J. 8. B. R. A. S. 1890, no. 21, p. 144). Distrzbution. Malay Peninsula, Sumatra.

24. IXALUSPICPUS Peters. Ixdus pictus, Blgr. Cat. Batr. Sal. p. 99; S. Plower, P. 2. S. 1896, p. 908. Distribution. Malay Peninsula, Borneo. 25. IXALUSAWER Blgr. Ixulus mper, Blgr. P.Z. 8. 1886, p. 416, pl. xxxix. fig. 1. There are specimens, obtained by Mr. Wray in Perak, in the British Museum and in the local Museum at Taiping. Signor Fea obtained two specinieus at Thao, Karin Hills (Blgr. Ann. AMUS. Civ. Genova, series 2, vol. xiii. [xxxiii.] 1898, p. 37); and it is also recorded from hills between Burma and Siam (W. L. Sclater, P. 8.9. 1892, p. 347). Distribution. Malay Peninsula, Burma.

Family ENBYBTOMATIDB.

26. CAL~PHRYNUSPLEUROSTIBYA Tschudi. Culophryiam pbzo.ostyma, Blgr. Cat. Batr. Sal. p. 158; S. Blower, P. Z. 8. 1896, p. 908. 1899.l BATRACHIANB OF ‘PHp1 MALAY PBNlN8TJLA AND STAY. 901 Localities. Of this interesting frog, which does not seem to have been previously recorded from Siam, I obtained a specimen at Bawtong, Kabin, in March 1897. I also got one near the foot of Gunong Pulai, Johore, in September 1897; and Mr. Ridley obtained a specimen in Selangor, in July 1897 ; these are the two first reported occurrences of this species on the mainland of the Peninsula. Habits. Nothing is known definitely of the habits of this frog, but it is supposed to be the author of a remarkable strident call heard in certain Malay jungles, which may be written LL waalk, waalk.” There is a big field of interesting work in determining the animals whose voices are heard both by day and night in ungle-clad districts j neither the English pioneers nor the Malays k now for certain what animals make some of the most noticeable jungle calls : as an instance I may mention thht in June 1898 in the woods round Jenan, Kedah, a cry of 6L koop ” was to be heard, even at high midday, which we imagined to be made by some batrachian, but diligent search on the part of several local Malays, my Siamese “boy,” and myself failed to discover anything in the spots whence the sound seemed to have proceeded. Colour (in life). Kabin specimen (March).-Above blight yellowish bronze j a broad dark brown line from snout to inset of hind leg, passing through eye. At inset of hind leg is a conspicuous black spot surrounded by a white ring, the greater part of the thighs is bright vermilion. Johore specimen (September).-Above rusty red-brown ; a narrow black lifie (bordered above with yellow) from snout to inset of hind leg, passing through eye. Just above where this black line terminates on either side of the back is a conspicuous round black spot. The lower surfaces are a paler red than the upper, and the lower aspect of the hind leg is marbled with dark brown on a buff ground. Size. Kabin specimen, snout to vent 36 mm. Distribution. Burma, South China, Siam, Malay Peninsula, Natunas, Borneo. 27. MICEOHYLAOBNATA (Dum. & Bibron). (Plate LX. figs. 1, la, 1b.) Microhylu ornatu, Blgr. Cat. Batr. Sal. p. 165. Localities. This little frog was obtained by M. Mouhot in Cambodia, but it does not seem to have been previously recorded from either Siam or the Malay Peninsula. I have found it in Bangkok in the months of Jan., Feb., March, April, May, June, July, August, and December j at Paknam Menam in August ; in the Dong Phya Fai, at an elevation of about 900 feet, in November ; at Ayuthia in February ; at &bin in March ; at Chantaboon in January; and in the Royal Siamese Museum etores were some specimens labelled (‘Bangpain, Oct. ’93.” In March 1898 I obtained a single frog in the Waterfall Gardens, Penang, which is referred to this species; and in June 1898 found it numerous 902 MR. dANfEY 8.l?LOWEl& ON T~0 [Nov. 14, near Alor Star in Kedah, and also near Jenan in the same State caught a particularly handsome little frog which is referred provisionally to M. ornata, but which differs considerably in appearance from Bangkok individuals. Habib. This very actrive,elegant frog is to be found hiding during the day under stones, logs, &c. in the crevices of the mud in dried-up pools and among dead leaves. Once I found two in an ants' nest, situated in the ground under some brickwork. It comes out at dusk and seems Lo remain abroad all night. At night in December and January these frogs may be heard croaking in Bangkok ; considering their small size they produce an aston- ishing volume of sound, the noise seems to me indescribable on paper. Colour (in life). Usual Bangkok specimens.-Above reddish olive, with a large dark brown mark on the back beginning between the eyes, then narrowing and then widening as it extends to the hind part of the body ; a broad darker brown line along the side of the head and body ; limbs with irregular dark brown cross-bars ; a dark horseshoe mark round the vent. Lower surfaces white, extensively spotted witch brown ou the throat and chest. Iris : golden ring round pupil, remainder golden speckled with bronze. The Penang specimen was above yellow marked with rich dark brown, and below pale immaculate buff. Size. The largest Siamese specimens noted measured-snout to vent, d 22 mm., 9 23 mm. Penang specimen (sex not recorded), snout to vent 24 mm. Distrihzctiori, Kashmir, India, Ceylon, Burma, Southern China, Cambodia, Siam, Malay Peninsula.

Tadpoles. At the end of December 1896 I fouiid tadpoles of this species in a small pond in Bangkok; they were numerous all through -January and February 1897, aud the young frogs were leaving the water at the beginuing of March. On revisiting the saine poud tbe following winter (1897-983 I €ailed to see a single specimen. DeecriptiotL. Length OE body once and three quarters its width, a little more than half the length of the tail. Nostrils placed close together on the upper surface of the head, nearer the end of the snout than tbe eye. Eyes on the sides of the body, visible from above and from below, their distance apart is about five times the distauce between the nostrils, and also much greater than the width of the mouth. On the back between the eyes are a pair of shields, oval in outline, placed side by side ; they are not conspicuous in the living tadpole, but in specimens shrunk in spirit they become so. Spirrtculum median, on lower surface of body, opening into a transparent sheath of skin, in front of the anus. Anus median, opening in the lower edge of the subcaudal crest. Tail about four times as long as deep, ending in a very finely produced point ; upper crest not extending on to the back ; 1899.1 BATRAUHIANS OF THE MALAY PEN~SUL~AND SIAM. 903 lower crest deeper than the upper. The mouth has neither hard beak, labial teeth, or papillse, but consists of a simple upper lip and a contractile lower one. Colour. In life these tadpoles are transparent and almost colourless. The eyes and viscera are therefore very noticeable. But there are a number of minute yellowish-brown spots, particularly on the back, where they form a somewhat diamond-shaped figure, and on the muscular portion of the tail. Size. Length of body 7 mm. ; width of body 4 mm.; length of tail 13 mm.; depth of tail 3 mm. The recently transformed young measure about 7.5 mm. from snout to vent. A remarkable feature of these tadpoles is that the hind feet are for a time completely webbed, the web is very fine and colourless; when the young frogs leave the water this web disappears. N.B.-At various times I have found iu Penang some very remarkable tadpoles (Plate LX. figs. 2, 2a-2c), the affinities of which could not be conjectured, till, having made out the tadpole of Microhykz ornata, there seems no doubt they belong to some Engystomatoid Batrachian. Although the species they develop into is still unknown, I think it desirable to describe the tadpole in this paper. The first tadpoles with the spiraciilum thus placed, and with this simple mouth, that I came acrosa, I caught in Singapore early in 1896. At the time I imagined them to belong to Callula pulchra (and still do so), but was unable to prove it ; on later occasions I have found these “ transparent tadpoles” in Bangkok (where I was able to observe them grow into M. ornata), in the Dong Phya Fai, in Kedah, and on Penang Hill, all of which I haw no doubt are of the genus Microhyla. These ‘‘ transparent tadpoles ” (of which I have observed four or five different species), besides differing in structure, differ entirely in habits from the tadpoles of the Banidre and Bufonidse. Instead of passing a great deal of their time on the bottom, they are usually just under the surface of trhe water, continually opening and shutting their mouths ; they are very delicate, and difficult to transport alive in a bottle even for a few miles.

Description of a Transparent Tudpoles,” Pemy (Nov. and Dec. 1896). Fm.The length of the body is about 14 its width ; the length of the tail is from 14 to 1s the length of the body. Nostrils. Distance of nostrils apart 1.4 mm.; distance from nostril to end of snout 25 mm. ; distance from nostril to eye 4 mm. The nostrils are placed close together on the upper surface of the bead, and are nearer the end of the snout than the eye. The distance to the eye is from 14 to 1: the distance to the end of the snout. The distance between the nostrils is about Q the distance betv een the eye#. 904 ME. ISTANLBIP 8. FLOwDIl ON THE [Nov. 14, Eyee. Distance of eyes apart 7.5 mm. The eyes me on the eide of the head (looking out horizontally), and are visible from above and from below. Syiranclum. Distance of opening of spiraculum from the eye 10.5 mm. The spiraculum is median, on the lower surface of the body, opening into a transparent sheath of skin directed backwards and downwards j the opening being in a vertical line drawn behitad the body. The width of the opening of the spiraculum is from 2 to 3 mm. Anus. Median, opening into a dark-coloured sheath of skin directed backwards and downwards, behind and nearly parallel to the sheath of the spiraculum, but longer than it. The opening is B longitudinal cleft. These two tubes are very prominent in the live tadpole. Tail. The tail is of remarkable appearance, owing to its coloration, the pigment in the crests ending abruptly, and also being continued further along the outside of the crest than along the part nearest the muscular portion ; consequently the tadpole when seen alive in the water has apparently a trifurcated tail, a very long centre point and short upper and lower one. The end of the tail is prolonged to a very fine point or filament, which in life is almost continuously being vibrated rapidly from side to side, the end frequently curving round so as to be almost parallel with the rest of the tail. The upper crest is convex and does not extend on to the back. The lower crest is deeper than the upper, and forms a double curve, the deeper being that behind the opening of the anus. Mouth. The mouth is 3.5 to 4 mm. in width. The mouth is situated at the extremity of the head, and not on the lower surface; the lower lip projects beyond the upper : this is particularly so in young specimens, where the mouth appears to be on the upper surface of the head. These young tadpoles frequent the surface of the water, and their mouths are constantly expanding and contracting (which words seem to imply the motion better than ‘‘ opening and shutting ”). There are no papilla round the lips, labial teeth or horny beak, but the mouth consists of simple upper and lower lips, the latter with a very deep notch in the centre which expands when the mouth is open. CoZour (in life). These tadpoles are very transparent and the amount of colouring varies in individuals. The upper surfaces and sides of the hinder parts of the body are generally yellowish brown, mottled along the vertebral line with very dark brown. The lower surfaces and sides of the head are colourless or a pale dirty buff. The muscular portion of the tail is pale yellowish brown, getting darker towards the point. The crests are buff mottled with yellowish brown, along both edges are irregular dark brown marks; the brown mottling gets more continuous and darker towards the point of the tail till it ends abruptly, 8s described above. Iris bright yellow. 1899.1 BATRAOHIANS OF THB MALAY PBNIXSUU AND SIAM. 906 Size. A good specimen with hind legs developed was in total length 37.5 mm.; length of body 13.5 mm.; width of body 10 mm. j length of tail 24 mm. ; depth of tail 11 mm. The toes are fully webbed and tips well dilated.

28. MIOBOHYLAINOBNATA Blgr. Microhyln inornata, Blgr. P. I;. S. 1890, p. 37. This little frog was described from specimens obtained by Prof. Moesch near Deli, Western Sumatra (P. Z. S. 1890, p. 37); subsequently it was found at Palon, Yegu, by Signor Fea (Blgr. Ann. Mus. Civ. Genova, 1893, p. 39), and near Bangkok, Siam, by the late Dr. E. Haase (Boettger, 2001. Anzeiger, 1893, no.433, p. 430). I have not met this species alive myself, but I received a number of specimens said to have been caught at Chantaboon, Sism. Distr&twn. Bum, %am, Sumatra.

29. MIOBOHYLALBIUOOSTIBMA Blgr. Mimohyk leucostigma, Blgr. Ann. t Mag. Nat. Hist. (7) vol. iii. 1899, p. 275, pl. xii. fig. 1. I obtained three specimens in the Larut Hills, Perak, at an elevation of 3500 feet in April 1898 ; two of these (the types) sre now in the British Museum, and the third in the Raffles Museum at Singapore. A pair were caught in cop&: as the mode of embrace does not seem to have been recorded for any Asiatic Engystomatoid Batrachian, it is interesting to note the embrace was axillary, and the fingers of the male did not meet on the breast of the femnle. Colour (in life). Above intense iridescent black, with very small scattered bluish-white spots, which get larger towards the sides and in the anal region. Below very rich dark brown (blacker on the throat, redder on the belly and thighs), nearly covered with large very distinctly defined spots of intense yellow. Upper surface of limbs reddish brown, with black cross-bars, and thickly studded with very small white spots. Upper surface of hands and feet brown, with bright yellow spots. Lower surface of hands and feet reddish brown. Iris very dark brown, very minutely spotted with pslo gold. Pupil circular. The sexes do not appear to differ in coloration. Size. 8. Snout to vent 25 mrn. ; arm 14 ; leg 39. 2. Snout to vent 27 mni. j arm 155 ; leg 405. Distribution. Malay Peninsula (Perak). Mr. Boulenger informs me the same frog bas been discovered in Borneo, on the Baram river, by Mr. C. Hose.

30. MICROHYLAPULCHBA (Hallow.). .&~Oh~kc yuZohru, Blgr. Cat. Batr. Sal. p. 165. M. Mouhot obtained this species in Cambodia, and I found 8 single specimen under a stone in the Dong Phya Fai, Siam, in 906 ME. STANLEY 8. FIIOWB~EON TH.B [Nov. 14,

November 1897. It is a singolarly handsomely marked frog, and well merits its specific name ofpulchra. Distribution. Siam, Cambodia, China, Hongkong.

31. MICROHYLAACHATINA (Boie). ;IficrohyZa uchatinn, Blgr. Cat. Batr. Sal. p. 166. Localities. This pretty little frog was known to inhabit the Malay Peninsula from d and 9 specimens sent to the British Museum from Malacca by Mr. D. 3’. A. Hervey ; it does not seem to have been previously recorded from %am. I have obtained specimens on Penang Hill, at from 2000 to 2500 feet elevation, in Nov. 1896 and April 1898 ; in Taiping, Pemk, in May 1898 ; in Bangkok in July 1898 ; in the Dong Phya Fai, about 900 feet elevation, in Nov. 1897 ; and I have received specimens from Chantoboon. Habits. A very active frog; at times taking very sudden, long hops like a “grasshopper ” insect, at others using its dilated digital disks in climbing like a true tree-frog. Colour (in life). Upper parts vary from very pale light bronze- brown to rich bronze-red, speckled in irregular longitudinal lines, with very small dark brown spots, a very pale yellow vertebral line, and a conspicuom dark brown or black pattern on the back. The sides are rich dark brown or black. Lower parts purplish buff. Iris golden. Size. Snout to vent 20 mm. Distribution. Tenasserim, Sinm, Malay Peninsula, Sumatra, Java, Moluccas.

32. MICROHYLABERDMORII (Blyth). Microhyln berdmorii, Blgr. Cat. Batr. Sal. p. 166. Localities. This species, although known from Burma, Malacca (Davison), and Cambodia, does not seem to have been recorded from Siam. In Nov. 1897 I found it numerous near Hinlap, in the Dong Phya Pai, Shm, elevation about 700 feet. Habits. Nocturnal, frequenting the neigltbourhood of water, an extraordinary good jumper (even for a frog). Size. Snout to vent 43 mm. Distribution. Burma, Siam, Cambodia, Malay Peninsula.

33. CALLULAPULCHEA (Bray). Hyltedactylus bivittatus, Cantor, p. 143. Calluln pulchrn, Blgr. Cat. Batr. Sal. p. 170 (hand &c. fig.); 5. Flower, P. 8.S. 1896, p. 908. “ Eung-ahng ” of the Slamese. ‘‘ Bull Frog ” of the English in Singapore and Siam. LOWk&%. This species apparently does not occur in Penang ; but is now common in Singapore, having been (from all mounts) imporked into that island from Siarn. The only instances of its occurrenoe on the mainland of the Peninsula (that I know of) are 1899.1 BATICAOJXIANS OF TEE MALAY PEIPIB~JLAAND SIAM. 907 the single male obtained from a field near Malacca by Cantor, and specimens in the Museum at Kuala Lumpor, from that place, where it is said to be common. In Shm I have observed it in Bangkok (Jan., May, June, July, Aug., Nov., and Dec.), at Paknam Menam (Bug.), at Tahkamen on the Bangpakong river (April), and I have also received specimens from Chantaboon. Habit-s. From having kept many specimens in captivity for months at a time, and also observed them frequently in their native haunts, I think CnUu7a pu2chi.a is the clejerest batrachian I have come across : they are good swimmers, can hop \!ell on laud and also cliaib fairly, though slo\vly; ours in captivity in the evening often go up the glass hide of their case, but they manage better in a corner than on a plain vertical glass wall. During the raius, when every evening swarms of insects flew into the house attracted by the light and were a great annoyance at dinner-time, we were in the habit of putting a Callula or two on the dinner- table : they seemed to understand what they were there for, and instead of jumping off the table or being alarmed by us or the servants, caught and ate the 0yiug insects, one after another, as they alighted on the cloth. Termites, ants, moths, small beetles, crickets, and grasshoppers they devour eagerly, but the larger crickets aud grasshoppers they cannot manage to hold to got them into their small mouths ; they seem more clever in catching their desired prey than either Rana or Bufo, and also show curious discrimination in not attempting to seize the wiuged bugs, which often come into the house at the same time as the swarms of ants, termites, Qc. During the rainy season in Bangkok almost every evening, after a wet day, the whole air is full of the booming of these frogs-“eung- ahng, eung-shng, eung-ahng,” now rising, now falling, and the sound continues all night. In some of the roads where there is low land and much water on each side, and Callula swarms, you can hardly hear yourself speak for the noise, but at the distance of a quarter or half a mile the sound is not unpleasant and is like that of a great weir or waterfall. In Singapore possibly they croak on suitable evenings all the year round : personally I have noted them doing so in the months of March, April, Ma]-, June, July, Sep- tember, October, and December. In captivity they continue to make their characteristic sound ; also apparently they can make a quite different noise : on more than one occasion we were disturbed at night in Bangkok by shrill screams apparenhly of a person in great fear and pain ; the noise seemed to come from the room where the Callula were kept, but on procuriug a light and going there, I found them sitting quietly in their vivarium as if nothing had occurred, so it cannot be proved that they were the authors of these really alarming cries’.

1 Our knowledge of the strange cries that animals make at times must still be very meagre. Various noises occurred from time to time in the old ruinous palace I lived in at Bangkok that I did not succeed in tracing : the natives (as usual) attributed them to the supernatural, but I have no doubt they were so8 ME. armmy 8. PLOW~BON ma [Nov. 14, 1 bve been told the Laos eat Ccillula pulchra, but the Siamese in Bangkok do not, though they esteem Rana tigrina as food. Distribution. India, Ceylon, Burma, South China, Siam, Cambodia, Malay Peninsula, Celebes.

34. PHEYNELLAPULCHBA Blgr. Phrynella pulchra, Blgr. A. M. N. H. (5) xix. 1887, p. 346, pl. x. fig. 2. Distribution. Malay Peninsula (Malacca), Snmrttra, Mentawei Islands.

35. PRRYNELLAPOLLICARIS Blgr. Phrynella pulchra, Giinth. A. M. N. H. (5) xx. 1887, p. 313, PI. xvi. fig. B; L. Wrap, J. 8. B. R. A. S. 1890, no. 21, p. 141. Phynellapollicaria, Blgr. P. Z. 9. 1890, p. 3’7. Diatm’bution. Mday Peninsula (Perak).

Family BUFONIDIE.

36. NEGTOPHRYNEGUENTHERI Blgr. Nectophryne guentheri, Blgr. Cat. Batr. Sal. p. 280, pl. xviii. fig. 3 ; 8. Plower, P. Z. 5. 1896, p. 910. Mr. Ridley obtained another specimen on Bukit Timab, Sing]&- pore, in March 1898. Distribution. Malay Peninsula (Singapore), Mentawei Islands, Natuna Islands, Borneo.

37. Bmo PENANGENSIS (Stol.). (Plate LX. figs. 3, 3a.) Ansonia penanyen&, Stol. J. A. S. B. 1870, p. 152, pl. ix. fig. 4. Bufopenangensis, Blgr. Cat. Batr. Sal. p. 287. Lowlities. I have found this species in the hills of Penang, elevation 2000 feet, in March 1898, and in the Larut Bills, Perak, elevation 3000 feet, in April 1898. Dr. Hanitsch (Bep. Baffles Libr. & Mus. 1898, p. 5) records specimens from Gunnng, Kledang, Perak, elevation 2100 feet, caught in March 1898. Habite. My Penang specimens I caught after dark hopping on the ground on paths through the bill-jungle ; the Lariit specimens I found by daylight crouching on the nearly vertical face of some rocks on the side of a rushing mountain-stream : they were easily caught in the hands. The iris in life is golden.

made by animals, probably reptiles or batrachians, that we generally oonsider to be mute. I have not seen it recorded that the Lizard Uromastix czgyptius has a voioe, but specimens now living in my house here olbn make a low noiee, a sort of guttural aackling, audible 3 or 4 yards OK Testudo margbata at times utters a plaintive cry, very like a shee bleating ; and Testzcdo rudiata has a low querulous bark; probably many otEer instance8 could be given.-S. S, F., Ghiwh, Egypt : 13-699, 1899.) BATEAUHIANS OF rn MALAY PININSULA AND SIAY 909 Size. The largest specimen from Penang measured, snout to vent, 37 mm. Distribution. Malay Peninsula, Borneo. Tadpole. Tadpoles and newly transformed young toads were found in abundance in two streams on Penang Hill, at elevations of about 1800 feet, during March 1898. Habits. These tadpoles live in the swift-flowing hill-streams, and are to be found where the torrent is rushing fastest, fixed to the face of the granite boulders which obstruct the stream; a favourite place of theirs was a perpendicular wall of rock which the water fell over in a small cascade j they hold on so fast with their mouths that they cannot easily be pulled off, but have to be plucked away from the rock between one's finger and thumb. They move upstream and about the face of the rock by means of their mouths; when placed in a glass bowl they never laid on the bottom (as most tadpoles do), and seldom swam about but fixed themselves to the glass sides. In captivity they died in a few hours, the still water probably not suiting them. Descriytioi, of the Tadpole (in the 3rd period). Forin. Length of body from rather more than once and a half to rat,her less than once and two thirds its width, nearly half the length of the tail. Nostrils much nearer the eyes than the end of the snout, about a quarter the distance. Eyes on the sides of the head, looking outwards and upwards, not at all prominent in life ; the distance between the eyes is rather more than once and a half ns great as that between the nostrils, and little more than half aa great as the width of the mouth. A strongly marked lachrymal canal from in front of the eye to the nostril. Mpiraculum on the left side, directed backwards and upwards, rather nearer the eye than the anus, not at all prominent in life. Anus median. Tail six times as long aa deep, acutely pointed; upper crest only on posterior two-thirds of tail, lower crest whole length of tail, but only the posterior two-thirds are pigmented ; crests of equal depth "or herslightly deeper. Mouth. The large mouth forms an organ for adhesion and locomotion. Beak white ; lower jaw edged with black, upper with a conspicuous black diagonal mark on each side. The lips form the rim of a sucking-disk, when not fully expanded they take a crenular form (in spirit-specimens this is very marked) ; the upper lip, which has its edge turned in and terminating in the 1st row of upper labml teeth, is smooth and free from papilla; the enlarged muscular lower lip is thickly studded with very small short rounded papilla. There are two uninterrupted series of upper labial teeth of equal length, the 2nd being slightly stronger than the 1st : three uninterrupted series of lower labial teeth of equal length but shorter than the upper series ; the 3rd is the strongest and the 1st the weakest. PEW. %OJi BOO.-1899, NO. LE. 69 9io MI. S~AIVLBY8. Pbtmk 0% TBB mov. 14, Cotou~(in life). Upper surface sepia-brown, mottled darker and lighter j the disk round the head is translucent, colour ellow very finely speckled with sepia-brown. Lower surface ye17 owish buff, the intestines showing the transparent skin as a dark pu lish patch. Muscular portion of tail sepia-brown mottled with yexw ; the crests are transparent, finely speckled with sepia-brown towards their edges, which are dark brown. The legs as soon as they appear have the bright colour and distinct markings of those of the young toad j when the fore limbs appear the back begins to take the markings of that of the young toad. Iris, a narrow ring of reddish yellow. Colour (in life) of newly transformed young.-Upper surface of head and body yellowish brown, extensively marked with black ; sides of head and body spotted with orange and yellow. Limbs red, with dark brown cross-bars. Below purplish grey, with numerous very small whitish-yellow spots. Size. Taclpoles (3rd period) : total length 34 mm. ; length of body 12.5 j width of body 7 ; width of mouth 6 ; length of tail 21-5 j depth of tail 3.5 j depth of mouth 5. Newly transformed young : snout to vent 13 to 14 mm.

38. BUFOYELANOSTIOTUS Schneid. Byfo mctanostictus, Cantor, p. 142 j Blgr. Cat. Batr. Sal. p. 306 ; Blgr.Faun.Brit. Ind., Rept. p. 505(fig. p.506). ; 8. Flower, P. Z. S. 1896, p. 911, pl. xliv. fig. 3 (tadpole). 6‘ KBkong,” ‘‘ KBtak pfirii,” of the Malays of the Peninsula, according to Cantor. ‘‘ King-kop ’’ of the Siamese. Localities. This is the common toad of the Malay Peninsula and Hiam, to be found in abundance at all seasons of the year j I have observed it in the following localities :-Penang (from seaJeve1 to the summit of Western Hill, 2725 feet) ; Alor Star, Aiiak Bukit and Kulim, Kedah j Taiping, Perak ; Johore Bahru ; Siiigapore ; Paknam Menam, Bangkok, Ayuthia, Pachim, Kabin, and Chanta- boon in Siam ; and at Kosichang, an island in the Gulf of Siam, where exceptionally large individuals were seen. Specimens from the same localities vary considerably in roughness, some are nearly covered with strong spinous warts. HaLits. This species resembles Bufo vulgaris in habits andmanner of feeding,and does well in captivity, readily eating beetles, termites, ants, crickets, grasshoppers, &c., but refusing millipedes. As a rule it frequents cultivated places, or the neighbourhood of paths and clearings, only once have I found a specimen in virgin jungle. At certain seasons the males make a good deal of noise croaking, both when wild or when kept in a vivarium ; while croaking the single vocal sac under the chin is distended into a globular form. I have heard them croaking in February (at Ayuthia), in March (at Pachn), in July (at Bangkok), and in November (in the Penang Hills). The Siamese are much afraid of toads j a man I employed 1899.1 BATBACHIANB OF TRPl !&Atby PBNINSTh.4 AND BtA& 911 in collecting would handle frogs, snakes, lizards, &c., and pick up scorpions, large spiders, &c., but could never be persuaded to even touch a toad. C’olouv. In the breeding-season the inales aamne a very haud- some appearancet the throat becoming bright chrome-yellow, and the sides of the head and cheet yellow, spotted with black. Size. The largest Bangkok specimen measured 116 mm. from snout to vent. Disttibution. India, Ceylon, Burma, Yikhirn Himalayas (up to 10,000 feet), Southern China, Hongkong, Cambodia, Siam, Malay Peninsula, Java, Borneo, Philippines. Tadpoles. I have observed the tadpole of B. mdanostictus in April and October in Singapore; in March in Penang (mouth exactly as described in P. Z. S. 1896, p. 912 ; but colour of caudal crests almost none, transparent) ; in July and August in Bangkok; and at Batu Gajah, Perak, in December I caught a tadpole exactly like t’hose of this species, except that the 2nd series of teeth in the upper jaw was uninterrupted.

39. BUEOMACROTIS Blgr. Bufo inaci-Otis, Blgr. Ann. Mus. Civic. Genova, (2) v. 1887, 1). 422, pl. iv. fig. 3 ; Blgr. Faun. Brit. Ind., Rept. p. 502. Localities. This toad, previously known froin Upper Burma and Pegu, has not been recorded from Siam before. I found it fairly iiumerous at Bawtong Kabin in March 1891. This species and B. melanostictus frequented the same spots. Size. Snout to vent 46 mm. Distribution. Burma, Siam.

40. BUFOPABVUS Blgr. Bufoparvus Blgr. A. M. N. H. (5) xix. 1887, p. 346, pl. x. In November 1896 I found a single specimen under some dead leaves in jiingle by the Waterfall Gardens, Penang. It was a very active toad, hopping like a frog ; in captivity it refused to feed. The eyes are large and prominent, and the tympanum very distinct. Colour (in life). Above rich dark reddish brown; limbs dark brown mottled with light red. Below yellowish buff and purplish grey, speckled on the chin, throat, chest, nud lower legs with dark brown. Iris golden and vandyke-brown, minutely speckled with black. Distribution. Pegu, Malay Peninsula, Sumatra.

41. Bmo QUADRIPORCATUS Blgr. Bufo pundriporcatus, Blgr. A. M. N. H. (5) xix. 1887, p. 347, pl. x. fig. 4; Giinth. A. M. N. H. (5) xx. 1887, p. 314, pl. xvi. fig. c. 59. 912 MR. STANLEY 8. PLOWBR ON THEl [Nov. 14, Distribution. Malay Peninsula (Perak and Malacca), Sumatra, Borneo. K.B.-Mr. L. Wray (Perak Museum Notes, vol. ii. part ii. p. 135) credits me with having presented the Museum at Taiping with a toad of this species from Penang ; the toad in question, however, was a specimen of B. asper.

42. BUFODIVERGENS Peters. Mr. A. L. Butler informs me he has obtained this species in Balangor during the latter half of 1898 ; it had not previously been recorded from the Malay Peninsula. Distribution. Malay Peninsula.

43. BUFOABP~ Gravenh. Bufo asper, Blgr. Cat. Batr. Sal. p. 313; 8. Flower, P. Z. 8. 1896. D. 912. This is the grandest batrachian known to inhabit the Malay Peninsula, and interesting in many ways : its size, rugged coat, and prominent yellow eyes at once attract attention ; its muscular strength is unusually great for an animal of this class ; the strong scent of musk it gives out when excited or alarmed is remarkable ; its habit of pretending to be dead is very curious j and, lastly, its motive for frequenting only particular caves and waterfalls would be most interesting to work out. For this reason I give each place and date on which I have met the species :- Great Waterfall, Botanical Gdns., Penang : 24.3.95 ; four. 19 7) 9, 30.3.96 ; two ( d d ). 3) 3, 9, 23.4.26 ; three ( Q .) Y9 33 19 8.3.98 ; two. 97 9, Y, 14.6.98 ; one. On 1st Jan., 1896, J searched this place without seeing a single toad. Penaug Hill, elevation about 2000 feet ; 26th November, 1896, one individual. Penang Hill, elevation about 2000 feet; April 1898, three individuals. Batu Caves, Selangor, 28th and 30th June, 1898. In the dark part of these caves there were numbers of B. asper; on the 28th I saw about twenty individuals, some were several hundred yards underground (perhaps half a mile) and in places where no daylight could ever penetrate. All the toads seen in the caves were well-grown specimens, apparently adult ; their rugged backs and colour exactly match many of the rocks in the caves. The excrement of these toads contained wings of small beetles and cockroaches. 1889.1 BATEAUHIANS OF IHBI MALAY P~NSULIIAND SYAM. 913 Colour (in life). General colour varies from yellowish brown, irregularly darker on upper surface, to rich reddish brown and blackish. Iris: a narrow golden line round pupil, remainder golden, very finely speckled and vermiculated with very dark bronze. Size. Largest Penang specimen I have measured waa, snout to vent 170 mm. (69 inches). Distribution. Tenasserim, Mergui, -Malay Peninsula, Java, Borneo.

N.B.-BuFo OAT~EATUBGiinther. Bujb pleatus, Blgr. Cat. Batr. Sal. p. 314. The type specimen, a female, was obtalued by M. Mouhot in Cambodia, so the species may be eventually found in Siam.

Family PELOBATIDX.

44. LEPTOBEAUHIUHHASSELTI Tschudi. Leptahrcic7~iuin7iasseltii, Blgr. Cat. Batr. Sal. p. 441 ; S. Flower, P. %. 8. 1896, p. 913 ; Hanitsch, Rep. Raffles Libr. & Mus. 1897, p. 8. Distribution. Burma, Malay Peninsula, Java.

45. ~~EOALOPHEYSNABUT.4 (Schleg.). .MegaZop7~ysinontana var., Cantor, p. 140 ; part., Giinth. Rept. Brit. Ind. p. 413. Meyabphrys nasutrc, Blgr. Cat. Batr. Sal. p. 443; 8. Flower, P. Z. 8. 1896, p. 913. “Katah bertandu” of the Malays of Perak, according to L. Wray, jun. Cantor obtained two specimens on the Pentland Hills, Penang, at an elevation of about 1800 feet. When in Penang during March and April 1898, I obtained one adult specimen from a valley ; and two adults, two small specimens, and many recently transformed young from the hills at elevations of from 1800 to 2000 feet. This species is also found in Perak (specimens in British and local Museums) ; in Selangor (one specimen in lo(:a] Museum, caught about 15 miles from Kuala Lumpor, 1898); Malacca (Raffles Museum) ; Johore (Rdes Mus.) ; and on Bukit Timah, Singapore (Raffles Mus.). Colour. Cantor gives a good description of this species, but 6‘ above, pale greyish-brown ; ” in four specimens which I observed alive for several days I found they were capable of altering their colour to a great extent-olive-brown, red-bronze, yellowish bronze or chocolate, but in every case the colours and sbdes resembled those ~eeuin dead leaves ; the rich dark-brown 914 MR. STANLEY 8. FLOWER ON THB [Nov. 14, markings on the sides vary very much in intensity from time to time. Size. d from Penang: snorit to vent 80 mm. ; width of head at angle of jaw 36 mmT . 9 from Yenang : snout to vent 90 mm. ; width of head at angle oE jaw 46 mm. D&tribrttion. Malay Peninsula, Sumatra, Borneo.

N.B.-MFIGIALOPHRYSMONTANA Kuhl. MegaZophrys mmtann, Blgr. Cat. Batr. Sd. p. 442. A frog in the Museum at Taiping, said to have been caught in Perak, apparently belongs to this species. ZXqtribzctwn. Java, Sumatra, Dinagat Island j Malay Peninsula (possibly).

46. MEGIALOPHRYSLONQIPFIS Blgr. MegaZophrys long+es, Blgr. P. 8.H. 1885, p. 850, pl. lv. j Uiinth. A. M. N. H. (6) xx. 1887, p. 316. This species was first discovered by Mr. Wray; in April 1898 Mr. Keilich again obtained it in the Larut Hills in Perak, at, an elevation of 4500 feet. Distribution. Malay Peninsula (Perak).

Order CAUDATA.

Family SALAMANDRIDAL

47. AMBLYSTOMAPPRBIMILE (Gray). Amblystoma per&de, Blgr. Cat. Batr. Grad. Lc. p. 47. Two specimens were collected by M. Mouhot in Siam, (‘probably at a considerable altitude.” Distribution. Siam.

Order APODA. Family CBOILIIDX. 48. ICHTHYOPHISQLuTmosua (Linn.). Qicrium glutinosum, Giinth. Rept. Brit. Ind. p. 441 (1864). Ichthyophis glutinosus, Blgr. Cat. Batr. Grad. &c. p. 89, pl. iv. 6g. 2 (1882); Blgr. Faun. Brit. Ind., Rept. p. 515, fig. p. 516 (1890). 6‘ Ngu pling ” or leech-snake ” of the Siamese. 1899.1 BATBACUNS OF TRE~&LAY PE~NI~CBULAAND SUM. 916 I have come across the following specimens :-

Number of Approximate length Locality. oiroular folds. in millimetm. (!) Bangkok ...... 366 560 (11) I, (Feb. 1898) ...... 820 304 (eij; ,, ...... >%4 2!M (1v) ,I ...... 312 282 (q)* I9 ...... 305 180 136 (m)* 99 ...... 308 (vii) A uthia (June 1887) ...... __ - (viii) Clnntnboon ...... 332 304 ’ (ixj * Penang, 1800 ft. 297 m (April 1898) 287 iMarch1898) 298 196 1::: 188 (xiv)* ,, f April 1898). 294 174 The counted number of circular folds must always be more or less approximate, as they are not all complete rings, some bifurcating in places ; but t’his table shows how very much they vary in number in individuals regardless of the length. Mr. Wray hm shown me a specimen obtained bey him in the Larut Hills, Perak, at between 3000 and 4000 feet elevation, and Mr. A. (3. B. Van Sommeren had in his collection one from the Penang Hills from about 2200 feet. Dr. Hanitsch (Rep. Baffl. Libr. 86 Mus. 1898, p. 5) records specimens from Ipoh, Kinta district of Perak, caught November 1897, and from Giinong Panti, Johore, caught June 1898. Habite. In life the tentacles are constantly being rotruded and retracted, and the throat is in constant motion, like a Prog’s. These creatures are gentle and make no attempt to bite ; although their usunl movements are very slow and deliberate, when they want to they can wriggle away with surprising speed. They do not feel at all slimy when handled. These remarks apply equally well to I. momchrouu. Colour (in life). Very dark rich purple, each circular fold showing as a narrow paler ring. Along each side of the body a bright lemon-yellow line, very distinct nnd sharply defined at the edges, but varying very much in width in specimens of about equal length. The eyes, though so small, are bright; they are black, with a very narrow pale brown ring round them. The tentacles are white. Distribution. Mountains of Ceylon, Malabar, Eastern Himalayas, Khdisi Hills, Burma, Siam, mal lay Peninsula, Sumatra, Mentswei Islands, Borneo, Java.

* In the ten speoimens marked with sn asterisk the tentacle is considerably nearer the eye than the nostril ; of its position in the remaining specimens I 5nd I have made no note. 916 BAPBAUHIANS OF THI MALAY PENINSULA LV SUM. [Nov. 14,

49. IOHTHYOPEISMONOCHEOUS (Blkr.). Ichthyophis monochrous, Blgr. Cat. Batr. Grad. Btc. p. 91, pl. iv. fig.- 1 (1882). In April 1808 on different days I obtained two specimens from under a stack of firewood near ‘‘ Maxwell’s Bungalow,” in the Larut Hills, Perak, elevation 3380 feet. 1st. Number of circular folds about 313 ; length 208 mm. 2nd. Number of circular folds about 309; length 167 mm. As in I. glutinoaus, some of the circular folds either bifurcate or converge into each other ; therefore the number, in counting the same individual at different parts of its circumference, varies. Colour (in life). Uniform purplish black. Tentacles white. Anal region and tip of tail pale pinkish. The eye appears as an inconspicuous black speck (but turns whitish in spirits). Distribution. India (Sikhim, Western Ghsuts, Surnt, Malabar), Malay Peninsula, Borneo, Java. ADDaNDA.-Mr. A. L. Butler has recently obtained in the Malay Peninsula examples of two species not included in this list-Rann jerboa from the Batu Caves, -Selangor, and Nectea atbasper from Sungei Buloh : he also obtained in Kuala Sumpor a male Runn macrodon, measuring ‘‘ exactly nine inches ” (228; mm.) from snout, to vent.

EXPLLLNATION OF THE PLATES.

PLATELIX. Fig. 1. Rana macrodon (p. %8). Tadpole. x 2. la. ,, ,, mouth. x 15. 2. Rana t

PLATELX.

Fig. 1, 1 a! 1 h. Microhykz oniata (p. 901). Tadpoles. x 4. 2. Mzcrohyla (i’), sp. ? (p. 903). Tadpole, side view and iipper view. x 3. 2a. ,, ,I ,, front view of mouth, shut. x 6. 2b. ,, . ,9 ,, front view of mouth, open. x 6. 2c. 11 ,, side view of mouth, shut. x 3. , 3. Bufo penailgensis (p. 908). Tadpole, iipper and lower views. x 2, 3a. ,, ,, mouth. x 5.

P.Z .S-.1899.Pl.LX.

J.Srmt dd etlith. MxmternBroe imp. MALE CHILIAN GUEMAL. (Mazamabisulca .) 917

4. Specific Characters of the Chilian Guemal. By R. LYDEKKER. [Receired August 30, 1899.1 (Plate LXI.) When describing that subgeneric group of American Deer commonly known as Guemals in ‘ Deer of All Lands,’ I had no mounted specimen of the Chilian species $0 compare with the one of the Peruvian Guemel in the British Museum; the latter having suffered considerably from fading. Consequently, I was compelled to rely on the descriptions of others ; and now that the Museum (thanks to Dr. H. P. Moreno) has acquired a beautiful male of the Chilian Guemal, I find that there are several iuaccuracies in my description. In the first place, the Chilian Guemal is a considerably larger animalthan thePeruvian species, theshoulder-height in the mounted specimens of these species in the Museum being respectively 39fr and 334 inches. Secondly, it is much more uniformly coloured than its northern relative, the greater portion of the under-parts, limbs, and buttocks being of the same tint as the back, instead of very much lighter. The faded condition oE the ikhseum specimen of the Peruvian species does not admit of the original tint of the hair being precisely determined ; but it was evidently speckled after the manner of the Chilian form. In the latter, the general colour of the head and upper-parts is hright greyish-yellow speckled with black. A broad black band runs up the middle of the face from the muzzle to terminate in a fork between the eyes; the sides of the muzzle being brown and the extremity of the chin whitish. The upper surface of the tail is coloured like the back, while the under surface is white ; there is no trace of the brown patch on the rump and the brown upper surface of the root of the tail characteristic of the Peruvian species. The under-parts and limbs, with the exception of the inguinal region, the front and upper part of the inner surface of the thighs, and a streak on the postero-internal surface of the fore legs (which are greyish white), are also coloured like the back ; thus presenting a very striking difference from the Peruvian animal, in which they are very much lighter. The tarsal tuft, too, instead of being dark umber-brown on a whitish ground, is likewise of the same speckled hue as the upper-parts. In regard to’the antlers, they are distinguished from those of the Chilian species by the forking taking place at a considerably greater distance above the burr, so that between the latter and the upper surface of the fork there is an interval of nearly two inches instead of less than an inch. The antlers of the specimen figured in the drawing (Plate LXI.), which came from Patagonis, are comparstively thin and smooth. In a head from Ultima Eaperanza, Patagonis, recently acquired by the 918 MR. It. LYDEKKEX ON THE CHILIAN GIUEMAL. POV. 14, Museum, the antlers are, however, much stouter and more rugose, perhaps indicating an older animal. Moreover, in the same specimen (represented in the figure, p,, 018) the general tone of the hair is greyer and less rufous, while the black inark on the face

Head of an adult male Chilian Guemal (Mumma biawlcn). irr narrower and less deep in colour. Now this specimen is stated to have been killed in June, that is to say in the middle of winter. And it would accordingly seem that the Quemals, like so many 1899.1 ON A SHARK-~H)THED DOLPHIN FROM PATAQONIA. 919 Deer, exhibit a reddish phase in summer, and a more greyish (blue) tint in winter. The Chilian Guemal was originally named from specimens obtained in the Andes of the country from which it takes its popular title ; probably on the east side of the main range. I can find no reason for separating the Patagonian animal, even racially. It has sometimes occurred to me that the Peruvian and Chilian Quemals might be nothing more than local forms of one widely- spread species ; but the important points of difference indicated above leave little doubt as to the propriety of regarding them in t.he light of separate species.

5. On the Skull of a Shark-toothed Dolphin from Patagoilia. By R. LYDEKKER. [Received September 7, 1899.1 In 1893 I described and figured ’ an imperfect skull of a Shark- toothed Dolphin from a Tertiary deposit at Chubut, Patagonia, which was clearly generically dbtinct from Squdodon, and seemed to me to require a new name. I accordingly suggested the title of Prosqualodon australis. From Squalodon this Dolphin evidently differed in the smaller number of teeth, and apparently in the shorter and more laterally curved lower jaw. Moreover, I. came to the conclusion that the nasals, instead of formiog mere nodules of bone lying in depressions of the frontals, were of triangular form, and to a certain small extent roofed over the base of the nose- cavity. Unfortunately, the extremity of the rostrum was so broken as to preclude the possibility of estimating the total length of the skull. This deficiency is supplied by a skull from the same deposit recently acquired by the British Museum, to which my attention has been directed by Mr. C. W. Andrews. This specimen has B general resemblance to the skulls of the short-beaked Dolphins of the present day, such as P?Locmaa, Granipus, Globiceps, &c. In size it apparently comes very close to the skull of I’seudowci crasdens, but is relatively shorter, and therefore more like that of Cogia brevicqs, so far as proportion is concerned. It agrees in all respects with the type skull ; and there is one detached tooth remaining, which is of the same Squalodont type as those of the latter. With the exception of a certain amount of damage to the region of the blow-hole, the new skull, in spits of numerous fractures, is comparatively but little imperfect on the upper surface. On the under surface the pterygoids, which afford such characteristic features in differentiating the skulls of the existing Dolphins, are wanting. Of the lower jaw only the greater portion of the right mandibular ramus is preserved.

1 An. Mus. LaPlata,-Pal. Arg. vol. ii. art. 3, p. 8, pl. iv. (1893). 920 YR. R. LYDEEKEB ON A [Nov. 14, The extreme shortness of the skull (only a very small portion of the tip of the rostrum being miseing) indicates the wide difference of the genus from Squnlodow; and we t,hus have evidence that the Squalodontidce, like the Delphiniclce, were represented by a long-beaked and a short-beaked group. The question will therefore arise whether the two groups of the last-named family may not be inde endently derived from the two corresponding groups of the Squayodonts. Pig. 1.

Upper surface of a skull of Prosqaalodon australis, from the Tertiary deposits of Ohubut, Patagonia. W.., frontal ; m., maxilla ; Pmx., prernaxilla ; Ero.. supra-occipital ; Pa., parietal. 1 a. Naiial region of specimen in the La Plah Museum. Be this as it may, the general characters of the fossil skull are essentially those of modern Dolphins, asymmetry being but slightly developed, while the proximal extremities of the prernaxillae and maxilla overlie aud conceal the frontals to the same extent. The prernaxillae, although their vomerine borders are perhaps slightly imperfect distally, seem, however, to have roofed over the mesethmoid channel to a less degree than in existing Dolphins, thus 1899.3 SHARK-TOOTEBD DOLPEIB FROM PATAWNfA. 921 leaving in the dry skull the greater extent of the vomer exposed, very much as in Oogia. In life the exposed mesethmo-vomerine channel was doubtless occu ied by a large mesethmoid cartilage. It is much to be regrette $ that the narial region of the British Museum specimen is imperfect, the nasal bones being wanting. I have therefore had reproduced (fig. 1 a, p. 920) this region from my original plate, from which it will ,be seen that the nasal bones, instead of being reduced to irregular nodules lyiiig in depressions of the frontals, form a slight penthouse to the upper end of the blow-hole. In my original paper I stated that the niolariform teeth were double-rooted, like those of Squabdon ; but, a detached specimen (fig. 2a, p. 921) shows that the two fangs have coalesced, although separated by a deep groove. And it appears that the same feature,

Fig. 2.

Lateral aspect of the specimen of Provpzlodon azlstralia re resented in 5g. 1. 3 a. A molar tooth associated with the sku{. judging from the sockets, obtains in all the teeth of this type. Home of the hinder molars mere, however, siugle-fanged; and the whole number of taeth did not apparently exceed ten or eleven pairs in each jaw, against the fifteen pairs of Xqunlodon. Whether the anterior teeth were of the slender inciaiform type of the similarly situated teeth oE SquaZodon cannot be ascertained. The CUE~Son the molars are leas developed than in the latter. The new specimen accentuates the distinction of P~*oequalodon from the last-named genus ; and whereas in the structure of the nasals the South-American genus is the more generalized of the two, in the characters of the teeth it is the more specialiied. It is worth mention that the retention of roofiug nasals in Pro- sqzutlodolz and in a second Patagonbn genus, for which I have suggested the name Argyrodelphis, removes any difflculty, so far 822 MB. B. LSDEIKKER ON THI DEINTALFORMULA OF [Nov. 14 as this part of the skull is concerned, in deriving the Whalebone from $he Toothed Whales. But whether such is the true phylo- geny may be left an open question; and I may add that, for several reasons, I do not propose on this occasion to discuss the geological age of the deposits from which Prosqualodon was obtained.

6. The Dental Formula of the Marsupial and Placental Carnivora. By R. LYDEKKER. (Plate LXII.) [Received October 21, 1899.1

Since the views expressed in the Study of Mammals ' 1 with regard to the dental succession in the Mammalia generally, and the homology of the individual teeth- of the cheek-series of the Marsupials with those of the Placentals, are out of harmony with the results of recent investigations, 1,think the tiine is ripe for a statement that I, as the surviving author of that work, no longer hold them. And I do this the more readily because it appears to me that some emendations in regard to the names employed for certain of the teeth of the cheek-series are urgently required. I may commence by the statement that I fully accept the view that the milk-teeth plue the so-called true molars constitute the first, or original series, and that the premolars form the second series ; this being precisely the opposite of the view taken in the work referred to 2. Apart from other considerations, 1 regard the fact that the last tooth of the milk-molar series (as well as sometimes the tooth in advance of it) is always similar in structure to the true molars as a very strong argument in favour of this view. And I likewise accept the view that the whole of the teeth of moderii Marsupials, with the exception of the single replacing pair in each jaw, belong to the first series. This being so, I come, without further preliminaries, to the consideration of the special subject of the present communication ; that is to say, the serial homology of the individual cheek-teeth in the Marsupial and Placental Carnivora, and the dental formula that will best express this homology. It will simplify matters to confine our attention in the main to the teeth of the lower jaw, as what holds good for these will be likewise applicable in the case of those of the upper jaw. To go no further back than the publication of his ' Odonto- graphy,' we find Sir R. Owen in that work' giving the lower dental formula of Canis, which may be regarded as typical for the

' Elower and Lydekker, 1891. * I do not propow to take into oonsideration the evidenae in favour of the occasional mence of an aborted eucaessional wriea to the true molars. Page &. P.Z S 1899 P1 LXII

J Smt del ethth MmternBros y"p LOWER TEETH OF PLACENTAL &MARSUPIAL CAFtNIVORA

$809.] TmYAEBUPIhL AND PIIAU~AZUABNIVOBA. 923

Placental Carnivora, as i. 3, C. 1, p. 4, m. 3 ; while he gives that of the Marsupial Thylacinus’ as i. 3, c. 1,.p. 3, m. 4. Nothing is said as to any replacement in the dental series of the latter genus, or in Marsupials generally; the division of the cheek-series into premolars and molars having been apparently made solely from the form and characters of trheteeth themselves. But, it is important to recognize that the premolars and molars mere regarded as being numericallay just the reverse of one another in the Dog and the Thylacine ; and that this view has been accepted by almost all subsequent writers till quite recently. In 1867 Sir William Plower a carried matters one stage further by proving that, when any replacement at all occurred, only one pair of teeth in each jaw was changed in the modern Marsupials ; this pair bein the third of the cheek-series of seven. It was further arguef that this replacing pair of teeth corresponded to the fourth cheek-tooth of the Dog, thus indicating that one premolar tooth (the first) was wanting in the Marsupial cheek-series, and hence suggesting that the full series in that group was originally i. 3, c. 1, p. 4, m. 4. It is, however, noteworthy that the three premolars of the Thylacine were still called p. 1, p. 2, and p. 3 ; and that the same notation was retained in the article ‘‘ Mammalia” by the same writer in the 9th edition of the ‘Encyclopsdia Britannica.’ By the date of the issue of the third volume of‘ his ‘ Anatomy of Vertebrates ’ (1868), Owen had likewise recognized the fact that only a single pair of teeth were replaced in each jaw of the Marsupials ; this, he said, “giving the extent of the theoretical deciduous series.” From this it may be inferred that he did not accept the homology of the replacing tooth of the Marsupials with p. 4 of the Placental series. But in a later part of the second volume (pp. 378 & 379) occurs the following very remarkable statement, which, although not altogether an exact solution of the problem, makes a very near approach to it :-“ The observed phenomena of the development and change of the teeth led to the generalisation that the Mar- supial differed from the Placental Diphyodont Mammals in hsving four true molars, i.e. m. 4 instead of m. 3; and also that they differed in having only three premolars, i. e. p. 3 instead ofp. 4; the typical number of the grinding series, 7, being the same ; and it waa convenieut for comparison to symbolise them accordingly. Since, however, there is reason to conclude that m. 1 in the Pla- cental Diphyodonts is R continuation of the deciduous series of molars, which might be symbolised as dm. 5, and only becomes a permanent molar because there is no premolar developed above it, so we may regard the tooth marked rn. 1 [that is to say, the fourth of the cheek-series] in Thylacinus as being an antecedent tooth of the deciduous series, rendered permanent by a like reaaon, the suppression of p. 4. In other words, that m. 1 in Thykcinus is the homologue of dm. 4 [the last milk-molar] of Xus [or Ounis], IbicE. p. 371. Phil. Trans. 1867, p. 631. * PSge285. 924 MR. E. LYDEKKEE ON TEE DIWTAL FOEMUI,A OF mov. 1.4, and that the true homologue of pm. 4 is not developed in the Marsupialia.” In this passage, then, the great anatomist recognizes, firstly, that the first true molar of Placentals belongs to the first series of teeth ; and, secondly, that the fourth cheek-tooth of the Marsupials is a persistent last (fourth) milk-molar. And it is merely in order to obtain general recognition for these two important facts that the present paper is chietly written. The next amendment in the dental homology of the Marsupial and Placental Carnivora was made by Professor A. Gaudry I, in 1878, who, shck by the resemblance between the teeth of the Creodont and Marsupial Carnivora, applied the same formula to both, thus making the lower dentition of Thylacinus i. 3, c. 1, p. 4, m. 3, or the same as that of Hymodon and Cank. He then pointed out that although the Creodonts differed from ThyZacinw and its allies by a complete dental replacement, yet the former likewise differed from modern land Carnivora by the circumstances that all their three true lower molars were of a carnassial type, and that they closely resembled the corresponding lower teeth of the Thylacine. No attempt was, however, made to show why the latter animal, in common with its kindred, should have four teeth of this same carnassial type. In 1887 appeared a paper by M-r. 0. Thomas 2, in which the replacing tooth of the Marsupials was definitely regarded as representingp. 4 of the Placental series, and was accordingly termed the fourth premolar ; the second tooth of that series being regarded as missing in the modern Marsupials. In this communication the author suggested the use of the term ‘‘ milk-premolars,” in lieu of milk-molars. Mr. Thomas’s nomenclature of the Marsupial series mas adopted in the ‘ Study of Mammals.’ It was some years after the appearance of the paper last referred to that the researches of Messrs. Kiikenthal and Rose afforded grounds for regarding all the teeth in advance of the replacing premolar of niodern Marsupials as milk-teeth, and the identification of the true molar series as corresponding serially with the milk set rather than with t)he premolars. To these discoveries I need not refer further than to say that a useful summary of them is given by Professor Osborn in the ‘ American Naturalist ’ for 18933. I accordingly pass on to two papers by Seiior Florentino Ameghino, in the course of which the remains of certain Marsupial- like Mammals from the Tertiaries of Patagonia are described and figured under the group-name of Bparassodonta.” In the Erst of these communications the animals in question are said to be referable neither to t,he Carnivora Vera, the Creodontia, or the

‘ Lea Enchainetnente, etc.-Mammifk-es Tortiairw,’ pp. 13-19. Phil. Trans. 1887, p. 447. Many of the views propounded here were modified in a paper published in the Ann. Mag. Nat. Hist. mr. 6, vol. ix. p. 308 (1892). Vol. xxvii. p . 493-508. Bol. Ac. CoAoba, vol. xiii. pp. 259-462 (1894). 1899.1 TIiE MARSUPIAL AND PLACEJTAL CARNIVORA. 92.3 Dnsyuridce, although stated to present resemblances to each of these groups. Apparently in all cases the palate is devoid of the un- ossified vacuities characteristic of existing Marsupials. In mauy instances the upper incisors exceed the number occurring in modern Placentals, one of the genera (ProtJiyylacinus) having the same incibive formula as in Thylacinus, namely :. The cheek-teeth (as shown in figs. 3, 5, and 6 of Plate LXII.) are also of a Rlarmpial type, the total ~iuniberbeing seven, of which the last four are molariform. And in his first communication Seiior Ameghino divides them according to the formula usually accepted for the Marsupialia; that is to say, into three premolars (p. 2, p. 3, and p.4) and fonr molars. He also goes on to observe that while the milk-dentition is more reduced than in the Carnivora, it is less so than in the Dasyuridm. The genus in which the reduction is carried to the greatest extent * is the one named Borhycem (Plate LXII. fig. 3), in which only the canine and the fourth rheelr- tooth have vertical successors. On the other hand, in the other genera (e.g. Prothylacinus, fig. 5, and Aotphiproviverra, fig: 6), both the second and third cheek-teeth, in addition to the canines, are thus replaced. In regard to the incisors there is no evidence. In the drawing (Plate LXII.) I have had the lower jaws of the three genera mentioned figured alongside of those of the Creodont genera Hycenodora (fig. 1)and Pteroclon (fig. 2) above, :md of the Mar- supial Thylacinus (fig. 4) below. And an inspection of these will show that, whereas the jnwa shown in figs. 1 and 2 have but three molariform teeth, all the others have four. The general resein- blance is, however, so striking between the whole series, that it is almost impossible to conceive that the seven cheek-teeth are not serially homologous with one another in the six genera. And this idea has been developed in Seiior Ameghino’a second paper, published in the Society’s ‘Proceedings ’ for the present year’. Thus on page 556 he writes that he assigns to the teeth behind the canines the progressive numbers 1to 7a,since they are perfectly homologous in the Placentals and Marsupials, the only difference being that some teeth may belong to the first series in cprtain genera (e. g. the fourth in Marsupials) and to the second in others (e. g. the fourth in Placentals). This view is in fact the one advanced by Owen, when he said that the fourth cheek-tooth of the Thylacine was a milk-molar rendered permanent by the suppression of its vertical successor. And looking at the number of forms described by Seiior Ameghino which serve in some degree to connect the Creodontia with the Dngyuridce, it appears to me, as already indicated, impossible to avoid accepting the above interpretation. The fourth cheek-tooth in the Prothylacinidce (Sparassodonta) indisputably belongs to the

Supra, pp. .555-671 ; I am not prepared to admit the Cretaceous age of Borne of the apeciuieiis described therein. This nomenclature had been long since proposed by Dr. H. Winge, Vidensk. Med. ~jobenhavn,1888, p. 66. PROC.ZOOL. Soc.-l899, No. LX. 60 928 MR. B. LYDEICKEE ON THI DENTAL FORMULA OF [Nov. 14, first series, as it does in Thylacinus ; and we have now to ask, is there any evidence that this tooth eber had a successor in allied forms ? The only instance with which I am acquainted where this question could possibly be answered in the affirmative is that of the Purbeck genus Trico)iodon (Triacanthodon),in which, as shown by Mr. Thomas ’, there are at least seven cheek-teeth, of which the fourth has a vertical successor. And it appears to me highly probable that we have in this genus an ancestral type of Marsupial in which all the first four cheek-teeth were replaced, as in the Creodonts. From this we pass to Prothylacinus anddmphiprowivarra of the Patagonian Tertiaries, in which (if Seiior Ameghino’s observations are trustworthy) only the canine and the second and third cheek-teeth are replaced j to Borhyrpna, in which replacement is restricted to the canine and third cheek-tooth; then to Didelyhya, in which only the third cheek-tooth has a successor and that at a fairly advanced stage of life ; and finally to Thylacinus, in which the same tooth is replaced in utero. Accepting, then, the foregoing interpretation, namely that the even lower cbeek-teeth respectively met with in Canis, Hytenodon, Prothylacinus, and Thykcinus are serially homologous one with the other, I come to the msin object of my paper, that is to say, to the formula we must adopt in order to indicate this. When I first considered the subject, I thought it would be necessary to adopt the plan proposed b.y Seiior Ameghino, and to term the teeth respectively 1 to 7. If this view were adopted, it would, however, be necessary to use the term “cheek-teeth” in place of ‘‘ molars,” as the latter has a special restricted signification. Were we starting de novo,I think this would be the better course ; but it is exceedingly inconvenient to interfere with the accepted use of familiar terms, and Mr. Thomas has suggested to me a way out of the difficulty which involves very little change. If we agree to call the RrRt four cheek-teeth of all the animals under coneideration ‘‘ premolars,” as coming in advance of the cL molars,” which never have successors, then we may designate t,hose that belong to the first series as (‘milk-premolars,” atid those of the second series as “ permaneut premolars,” with the respective symbols of mp. and pp. The adult dental formula of Hymnodoo,~will then staud us follows, viz. :- i.l.i.2.i.3 c.1 pp.l.pp.2.pp.3.pp.4 m.l.?n.2.nZ.3 2. 1 .i.2.1.s-c.3-2)p. 1 .ppP.2.pp.3.pp.4-1n. 1. In. 2.m. 6’

That of Borhyma will be :- _~_____? - c. 1 mp. 1 .pp. 2,pp. 3. mp. 4 111. I . w. 2.ni. 3 i. I . i. 2. i. 3 CTi-Cp. 1 .pp. 9 .pp. 3. mp. 4-m. 1 . nz. 2. m.3’

Phil. Trans. 1887, pl. xxvii. 5g. 10. 1899.1 THE YARIUPIAL AND PLACENTAL CAltNIVORA. 927

Prothylacinus, on the other hand, will have the formula :-- mi. 1 . mi. 2. mi. 3. mi. 4 mc. 1 mp. 1 . pp. 2. pp. 3. mp. 4-m. 1 . m. 2. m. 3 mi. I.mi. 2.mi. 3 me. I mp. I .pp. 2 .pp. 3. mp. 4 m. 1.m. 2. m. Y’

Finally, in Thyltcciwus we shall have :- mi.1.mi.2.mi.3.mi.4 mc.1 mp.1.mp.a.pp.3.mp.4 m.l.m.2.m.3 mi. 1. mi. 2. mi. s -GX-cmp. 2 .pp.3. mp .4-nr. 1 .m. 2. m. 3‘

In ordinary practice, however, when tJhe number, rather than the siiccessional homology, is the point to be elucidated, we may follow a modification of the practice now employed. 143 Hpnodon will remain as before, viz. i. i,c. 1, p.a, m. 5; and Prothylacinus and Th!ylacinus will be indicated by i. i, c i, nap. & 43 11. i, m. 3’ Possibly an emendation may be necessary in regard to the detailed formula of HycPnoilon, for as the first cheek-tooth (as in almost all other Placentals) is not replaced, it may really be a persistent milk-premolar instend of a permanent premolar. Indeed the condition occurring in Bhinoceros suggests that such is probably the case. In conclusion, I may depart so far from the subject indicated by the title of this paper as to express my opinion that the Prothy- kicinitlm (for I see no reason for regarding the (‘Sparassodonta ” as representing tiiore than a single family) are undoubtedly Mar- supials, and that they are not very far removed from the Dnsyuru&, of which they mav represent the ancestral type. They also appear to be related to the Creodontia, which are themselves in all probability the ancestors of both the modern Carnivora and Insectivora. The Creotlonts, on this view, have retained a tooth-change which is lost in the modern Marsupials ; and both groups may be derived from Mesozoic ancestors like Trimnodon and Amphitheriuni, in which, as appears to be indicated in the first-named of these, there must have been a complete tooth-change. Evidence of such an- cestry is afforded Iiy the retention in Mtlrmecohiuu of the numerous true molars distinctive of some of the Mesozoic genera ; while, 1d8 an abnormality, four trne molars may occur in other modern Marsupials, such as Didelp7zys. If these Mesozoic mammnls be rightly regarded as the common ancestors of both Creodonts and Dasyurids, it is more than doubtful if they (*an any longer be classed as LL Marsupials,” sensu stricto, for, in additiou to possessing B complete tooth-change, it is, in the light of recent researches, quite possible, if iudeed not probable, that they may have also been placentiferous. I may add that the nomenclature proposed for the teeth of the Placental Carnivora will also be applicable to those of the other Piaceutal orders. 60% 928 ON THE WOOD-CAT OF ABBENTINA. [Nov. 14,

EXPLANATION OF PLATE LXII. Outer side of left ramus of lower jaws of Placental and Marsupial Carnivorn. Fig. 1. Hymdon 1eptorh.pchua (nfter Gaudry), ~~92.5. 2. Pterodom dasyuroides (after Gaudry), p. 920. 3. Borhyema fera (after Ameghino), p. 935. 4. Thylacinus cynocephadw, p. 925. 5. Prothylacinus patagonicua (after Ameghino), p. 9%. 6. Ainphiproviaerra manzanhna (after Ameghino), p. 925.

7. Field-notes on the Wood-Cat of Argentina (Felisgeofroyi). By ERNESTGIBSON, F.Z.S. [Received August 9, 1899.1 During the last tweuty-five years I have had many opportunities of observing the habits of the “.Gat0 Months ” (Fe& yeofroyi) in this district, where it is not uncommon, frequenting the woods and gntss-coverts. Too wild to approach poultry-yards (notowithstanding A~ra’sstatement), it preys upon small rodents (Cauizaustralis and Ctenomys brusiliensis) and birds ; and I greatly doubt the accusa- tions made as to its attacking young lambs. That it can give a good account of itself with dogs is quite true; and it has been known to fly at man, or even a horseman, when brought to bay. I have seen it taken at night in one of the large and powerful traps employed for the Vizcacha (Luyostomus t?

November 28, 1899. Dr. HNNEYWOODWARD, F.R.S., V.P., in the Chair. Nr. Oldfield Thomas exhibited the skull of a Baboon raceritlg obtained at Aden by Messrs. Percival and Doason. It appeared to represent a new species allied to Papio hamadryas, but distinguished by its small size, the row of upper cheek-teeth being only 41.5 mm. in length. This species was proposed to be named Papio a~ahiciis.

Mr. W. Saville-Kent, F.L.S., F.Z.S., stated that he had devoted considerable attention siuce the meeting of the last session to the subject of trichromatic' or three-colour photography as applied to the correct colour-registration of Zoological and Botanical mbjects. With the aid of the lantern he submitted a series of examples upon which he had successfully experimented. These included various species of tropical butterflies, orchids, fishes, lizards, and birds. Among the slides displayed, that of a peacock's feather, in which the characteristic tints were reprohced with marked fidelity, was p%rticularly referred to as B successful demonstration of the capabilities of the system. Gold and Silver Carp, Cuckoo Wwses (Labrus mi.&us), and other marine species, taken by Mr. Saville-Kent at the Plymouth Zoological Station, yielded appropriate illustrations of the process as applied to the colour-registration of the more brilliant but notably evanescent hues of fishes. In the bird-section, especial prominence was given to the correct colour-portrayal of the gaily plumaged Australian finches Poephila yoitlili and I-'. nairnbilis. These wera 930 OX TBIOEWXATIO PEOTOOBAPHY. [Nov. 28, kepresented as lantern-transparencies prepared from photographs of artistically preserved specimens, and also from the replica of a water-colour drawing of a group of these birds executed by Mr. J. G. Keulematls from living examples. Mr. Baville-Kent explained that all these photographs had been taken by him with the Sauger-Shepherd colour-screens, of which he exhibited a set, in conjunction with the Cadett (‘Lightning Spectrum-plate.” These screens represented the three primary spectrum colours, red, green, and blue-violet, as enunciated by the late Prof. Clerk-Maxwell, and a separate negative of the subject had to be taken through each respective screen. The transparent positives prepared from these negatives were stained with tints complementary to those through which they were severally taken. That was-the positive resuking from the red-screen negat,ive WRH stained blue or minus red; that from the green screen, red or minus green ; and that produced from the blue-violet, screen, yellow or minus blue. Due care being exercised in obtaining tlie right, tint-gradation, and the three stained positives being then superimposed in precise register, au optically perfect presentment or counterfeit of the original subject wa.8 mechanically produced. The special method of developing and staining the positives exhibited was, as in the case of the producbion of suitable colour- ecreens, associat,ed with the name of Mr. Bauger Shepherd, with whom Mr. Saville-Kent had been working in collaboration. It was recommended, for the acquirement of perfect registration, that all three of the respective negatives should be taken on a single plate in conjunction with a specially constructed multiple back, of which a sample was exhibited. This, however, was not absolutely necessary. It was competent, in fact, for anyone possessing an ordinary camera to secure correct colour-replicas of desirable objects, using only in conjunction with his instrument the Cadet,t spectruul-plates and colour-screens referred to. As an illustration of this fiwt,, Mr. Srtville-Kent explained that the Peacock‘s feather, and several other subjects exhibited that evening, had been taken hy hiin wit,h a large-siz4 Xodak camera, across the lens of which he had simply slung, with the aid of elastic bands, consecutive sections of ltis innltiple-back screen. The negatives taken for tlie production of these lantern-trans- pareucies were also available for three-colour printing or process work. In conjunction more especially with such a perfected machine as tbe newly introduced Orloff Colour-priuting Press, there was evidently a wide field thrown open for the cheaper reproduction, by printing methods, of Zoological and Botanical subjects in their correct natural colours. For lantern demonstmtion, at any rate, the Sanger-Shepherd process as illustrated by him that evening would, Mr. Saville-Kent ant,icipated, strongly recommend itself to adoption by t,lte many naturnlists who had hitherto employed their cameras for the delineation in mono- chrome only of the subjects of their studies. In the instances, m,ore especially, of the brilliant but fleeting tints of reptiles, fishes, 1899.1 AN EXPEDITION !TO THE GAYBIA. 83 1

and marine subjects, as also in those of the tegumentary tissues and appzndages of birds, this method of colour-registration would be of invaluable aid to the working artist, who having possession of a correctly prepared transparency might utilize it at his leisure for the elaboration of a finished painting. The successful application of the same process to the duplication, as lantern-trausparencies, of coloured figures of zoological subjects had been demonstrated by the example of the pictures of the Goiildiau Pinches submittrd to the meeting. In a like nianner, colourecl illustratioiis from other more rare and costly zoological works could be correctly reproduced. - The following papers were read :-

1. General Account of an Expeditiou to the Gambia Colony and Protectorate in 1898-99. By J. S. UUDGETT, F.Z.S. [Received November 38, 1899.1 I propose to give a short general account of an expedition recently made by me, under instructions from the Couucil of this Society, to the river Gambia. This expedition hiid for its object the general study of the vertebrate I’aiina of the Ganibia, and especially the investigation of the habits of Protopierus aud Polyp ter us. The river Gambia lies between the 13th and 14th parallels of North latitude. It flows due west through country which, lying about 100 miles to the north of the equatorial forest-re,’mion, IS nowhere densely wooded but mosrl y covered with a somewhat eparse vegetation consisting largely of leguniinous trees interspersed with gigailtic baobabs (Adansonia diyitatn), the African mahogany (Kciyu sene!/alemis), figs and sycamores. Exterisive open plnins, which in the rainp season become flooded, border this river along the greater part of its course, while at a very variable distance from the river-bank low hills of dark red conglomerate rise, often abruptly, and occasionally in steep cliffs, to form level plateaux, which in the upper river may be 200 feet high. The river-bank itself is cluthed throughout the year with a rich luxuriant vegetation extending usually about 100 yards from the rater’s edge. Though here the trees and creepers remain green the pear round, yet away from the river the trees lose their foliage in the dry season as completely almost as our own trees in winter. From the mouth of this river to the country just below Nianimaru, the river is shut in by an almost impenetrable wall of mangroves, sometimes 30 feet in height. Above this point the river, though 532 MR. J. 5. BUDGETT ON HIS "OV. 23, tidal, is perfectly fresh. The tides in the dry seasou make themselves felt for over 200 miles up the river, in fact to the end of navigable water, where there is about a foot rise. The dry season extends from November to May. Torndoes usually begin in June, while during July, August, and September there is a total rainfall of about 50 inches. During these months, though the tides make themselves felt, yet there is no change in the direction of the flow, while in August there runs a steady current of about 3 or 4 miles an hour. In passing up the river the first place of iiiterest is the old Fort James, which as formerly the port of export of the Gambia for the black-ivory trade. It is now being slowly washed away. About 20 miles further up, the Viiitang creek joins the Gambia, andat the junction of the two streams is the village of Vintang : it is seldom that a purely native village is seen at the water's edge, 3s they are usually on higher ground a mile or so from the rivei.. If there are any tall trees in these villages, they are sure to be the nesting-places of Pelicans and Marabou-birds, which in the neighbourhood of the villages are strictly preserved. The vast flocks of these birds aud also of the Balearic Crane are a great feature in the lower river, where there is little else to be seen but continual walls of mangroves, though now and again the monotony is broken by the passage of ti native canoe or some trading cutter; but further up the vfiriety of the vegetation is much greater. Of particular interest to myself were patches of a Pandanus growing in the swampy grouud at the river-side. The native name of this was Fang jani, which means " It burns itself." It certainly looked as though it deserved this name, for wherever it was seen a portion of every patch was charred with fire, and it was not easy to imagine how this could have been set alight by an external agency. The great trading station on the Upper Gambia is M'Carthy's Island. To this place the trading cutters bring their cargoes of ground-nuts, the fruit of the plant Arnchw hypogcen, to be d~ippedto Europe by the Ocean steamers which make tlieir way up to this islaud. On M'Carthy's Island there are two trading establishnients or 6 factories ' as they are termed, and the remains of an ancient mili- tary settlement, consisting of Government House, Officers' quarters, and Barracks, formerly occupied by a detachment of the West India Regiment, which was withdrawn about 1870. The Government House alone of these buildings has been kept in repair; and here I established myself in company with Mr. Wainewright, the Commissioner of the district, who, though usually travelling about the district, yet spends B considerable portion of his time here as Governor of the island. I stayed on M'Carthg's Island about one third of my time. To the Governor of the Coloi~y,Sir Robert Llewellyn, I am indebted for allowing me the free use of the Colonial steamer, 'Mnnsah Kilah,' and also for much hospitality. TQ Mr. Wainewright, the Travelling Cominissiouer in the M'Carthy's 1899.1 EXPEDITION TO TIiE OAMBIA. 933 Island district, I am greatly indebted for allowing me the use oE a portion at‘ the Government House at M‘Carthy’s Island, end also for the use of his huts in the main towns of his district. Very soon after my arrival at my headquarters, I made a tow through the district with the Commissioner to get some idea of the kind of countray that surrounded me. We started from Nianimaru, which was subsequently made my second headquarters, and where I spent even more time than at M‘Carthy’s. The chief interest in this tour lay in the people themselves, the country we travelled through not being of great interest from the point of view of its scenery. Travelling was not difficult, as porters were plentiful, and were employed from one village to the next at the rate of 3il. a man, if the distance was not more than 5 miles. At the important towns IL court was held, and a stay was made of two days. The courts were held in the open, the chief, the head-man of the town, and the people all sitting round the Commissioner’s chair. There was plenty of time for shooting and no need to carry much in the way of provisions. The bag usually consisted of Bush-fowl and the Barbary Quail, Pterocles ydricincta, Guinea-hen, Qi’dicne- naus, various Spur-winged Plovers, especially Lobivanellus senegalus and Hoplopterus spino8us, also Doves and Pigeons as marly as were required. The finest of these, as game, was the Green Pigeou (Treron culva), which is never seen to approach the ground, being especially fond of the fruit of the fig-tree. The commonest birds around us, which were not shot for the pot, were nunibers of four specie? of Coraciaa, a Centropus knoun ns the “ foolish bird ” from its fearless habits and its call, which resembles a soft laugh, Reveral species of BuceroticEte, generally seen flying clumsily from tree to tree in small flocks; while overhead hovered large flocks of Bee-eaters (Merop nubious), swallow-like in flight and song. Other common birds everywhere seen in large flocks were the Metallic Starlings (Lamprocoliwr nuratus and L*oaudatus) ; Wood- Hoopoes (hisor senegalensis) seen in smaller flocks ; while the commonest solitary birds were the Long-tailed Shrike (Corvinalltr corvina) and a species of Drongo (Dicrurus assirnilis). The bushes of course swarmed with Ploceida? and Nectnriniida. It being the beginniiig of the dry season, the grass was everywhere yet high, and it was out of the question to do any mammul- shooting ; the only mammals visible were Climbing Squirrels and Monkeys. Burrows of Orycteropwr were seen, though the animal does not appear to be very common in this region. The towns visited during this tour were mostly far from the river and were taken in the order Nianimeru, Sukuta, Kaihai, Demfai, Tabanani, &mi, Koreantab, and back to M‘Carthy’s Island. Near Kaihai there were news of a Giraffe having been seen, but they appear to be extremely rare in these parts. I heard indirectly that there were two in captivity at KIisies on the Senegal river. 934 XR. J. 8. BODGE'lT ON 1119 [Nov. 28, &wing returned to M'Carthy'a Island on December 5, I devoted myself again to fishing and catching Polypterus. I found that all the specimens of Polypterns Zapradii had already returned to the river from the swamps, where they come up to spawn in the wet season. However, large numbers of the youug of Polypterus senqclalus could stiU be caught by damming up the swamp-oudets. This is a favourite way of fishing with the natives. They make dams across the creeks at short intervals, and then leave them in connection with each other for some days. Then damming up t.he connections, tbey bale out the water from the lowest compartment, collect the fishes, and proceed to the next compartment. Very much more difficult is it to catch t.he Yolypteri in the river. Nets which were very successful with other river fishes, failed utterly with Polypterus. The seine-net and trammel were given up, and the native cast-net was used with better success. The 1-eSlilt5 of weeks of patient work were not encouraging however, and I gradually realized that the the to catch Pol?jpter.us was during the rainy season, when it had betaken itself' to the flooded lands. However, during these fishing days at Nianimaru, many interesting fishes were caught, and most of the common small Passerine birds were skinned. Moreover, this firhing was not without its dangers and excitement, as a look-out had ever to be kept for Hippopotami wbich swarmed iu all the creeka. Moreover, frequently in the morning, when the trammel-net was examined, a Crocodile (Croco- dilus ccttCfphTaCtU8)or a Sawficlh (Prktisperotteti) had to be slain. Several specimens of the latter were thus caught up as far as M'Carthy's Island, some of them measuring 9 feet in length. Fly-fishing was tried wit,hout success. The line and hook were used more by the natives than myself. The trammel was found to be the best kind of net to use for the Nomyride, which were seldom caught in other ways. The Morniyrids apparently keep to the bottom of the river, and were seldom taken in the seine near shore. s It, was noticed that a very large proportion of the fishes caught in this river were brilliantly coloured red in the ventral posterior portion of the body. Of fishes I believe 40 species were obtained, including 2 Selachians, Protopterns annectcm, Polypterus lapradii and P. senegalus, 8 species of Biluroids and i Morinyrids, and 18 others belonging to various groups. Most of the fish were tried as food, but there was only one that was really good eating : this was, I believe, a grey mullet and was taken far up the river. Often the creeks in which the cast-net was thrown were very narrow, and the canoe did silently amongst the most luxuriant vegetation abounding with Bee-eaters and Flycatchers. Altogether representatives of 108 species of birds u ere shot, measured, and described ; but skins were made only of the smaller birds, of which examples of 52 species were obtained, belonging to 23 families. .. Wit,h Dr. Gadow's assistance, most of these have been identified. Of the Uim..idre, in addition to the gregarious Irrim already 18'39.1 EIPEDITIOX TO TB1 GAMBIA. 936 mentioned, several Npecirnens of Sooptslue utern'niur were seen and a skin of a male preserved. On April 4 I took my two fishermen, my cook, and cmnm up to a small village in the Kunchow creek called Alimsks, and t,herehad some huts built. At this place again the trammel, the seine, and the cast-net were worked with hope of obtaining numbers of Polyptorus. As a rule in the afternoon I went out to shoot, and fouud it a fair place for game. During the fort,night thus spent at Alimaka, only six PoZypteri were caught. There were aught also in this creek several specimens of Qymnarchwr nilotincs and sonie fine specimens of a fresh water Turtle, Cydanorbis set~egalensis. Lions were heard here frequently, and Leopards were seen, but at neither did I get 3 chance of a shot. On April 20 t.wo English gentlemen and a Frenchman arrived at M'Carthy's Islaud, on their way to some supposed gold-mines about 300 miles to the easb of M'Carthy's Island. I accompanied them s short distance beyond the east,ward Britkh frontier to the town of Netebulu j t,he river is not navigable beyond that poiut. Netebulu is an important native t,own, where a powerful chief named Yandian had his castle and harem. Here we stayed several days as the guest.s of the chief, and then I parted from tbe gold expedibion, and made my way back overland to M'Carthy's Island, staying on the way a ~eekat, Koina. About 50 miles above MTarthy's Island the river-banks become high and precipitous, the country around being composed of high plateaux intercepted by valleys. Frequently, however, the edges of the plateaux retreat from the river-bank a mile or 80, surrounding wide phius, where one could be fairly certain of finding gcime. Along the steep cliffs of t'he river-bank, vast numbers of Dogfaccd Baboons (C;ynocephalus babuin) might be seen wending their uay. Sometimes the clifh extended YO far along the river-side that the Antelope were forced to come down to drink at certain places, and here the ground would be covered with their spoor. April and Alay are the best months for big-game shooting. At Koina, large herds of Tankong (Dnmaliscus kori-igum) were seen almost every day. Several were shot aud a complete skin was made, which, however, suffered severelp from the attacks of dogs and insects before it reached Englmd. These herds were composed of' males, females, and young of every age. The largest males seemed to lead the herd, though fine msles mingled with the females and young as they daily made their way back in long processiou from the river-banks to the higher lands. Large herds were also seen of Hippotmgus equinvs, the Roan Antelope, or Dakoio as the natives call it, but this species was not so plentiful as the Tankong in these parts. A herd of' Elands (Oreas derbianus) are believed to have been seen in the distance, aud I was presented with a skull taken by Mr. Wainewright from a Carcass floating down the river. 936 AN EXPHDITIOH TO THE GAMBIA. [R'ov. 28, In the open plains, where cluuips of tall dead grasR were shaded by a few trees, one might generally count on starting a Kanko- tong (Cobus kob), some Gazelles, or a Harnessed Antelope. The smaller solitary Antelopes were usually found in pairs. Enquiriee were instituted everywhere as to the existence in this region of a Zebra, but I could hear nothing of it. The horns either collected by me or from the natives included those of 9 specieu :-Bubcilis major, Damdiscus korriguwc, Cobus ttnctuo.sus, Cobus bb, Cervicnpra reduiwcr, Hippotragti.9 epuinus, Tragelaphus scriptus, a second species of l'mgelaphts not yet deterniined, and Oreas tlerhinttus. Buffaloes were said to he common on Deer Island, but they were not seen by me, though horns of two forms were obtained from natives. On the way back, a cutter was taken from Fatotenda to M'Carthy's Island, and after a few days spent at the Government House attending to my collections, and my living fishes and reptiles, I paid a final visit to Nianimaru. During this time, being the latter part of May, the rainy season began and the swampy places became filled with water. The Frogs began to spawn, and treveral Neries of stages in developinent of the different forms were preserved. Here I first obtained free swimming Protopterus with ripe ovaries : examples of 8 Frogs, 3 Chelonians, 5 Lacertilia, and 9 Ophidia, including a Typhlops, were also collected about this time. Returning to M'Carthy's Island, it was found that a number of €'ol!ypterus lrtpradii which hnd been kept in a pool connected with the river in the hope of getting them to spawn had been set free by the rising river. However, durinq the latter part of June and July a large number of Polypterus of both species were obtained, the females of which were crowded with ripe eggs. Artificial fertilization was tried with these, without success. Many were kept in confinement, mid some, of which a pair are now exhibited, were successfully brought alive to England. About the 10th of July, in the same swamp where these fishes were obtained, several nests of eggs were found. These eggs coin- cided in measurement exactly with the ovarian eggs of Polypterim. The young larvse possessed cement-organs on the front of the head bo characteristic of Ganoid larvse; and other characters led me to assume that they were the young of Polypteiws. None were reared beyond the larval state, and their identity could not well be established. However, having stnyed on the Gambia three months longer than I had intended, and having a number of healthy Polypteri full of spawn, I decided to return hoine. Just a day or so before leaving M'Carthyk Island I obtained eggs of Protopterm. These were watched through the early stages of segmentation, but the young could not be reared. On July 25 I left N'Carthy's Islrrnd and returned to England. Several Polyptei-i and Protopteri, 12 young Crocotlilus cat+ 1809.1 MR. L. A. BOItRdDhrLE OX THE PAGURIXE LAND-CRABS. 937 phructus, a P\ithon, 3 Qclunorbis senegalensis, 2 Hinged Tortoises, some Chameleons, aud a, Serval Cat were brought home alive. Since iny return to England, I have definitely decided that the eggs and larvae obtained are not those of Polypterus. I have, however, I believe, learned enough about the habits of Polypterm to encourage me to make a second attempt next year to obtain the developmental stages. In conclusion, I wish to thank the Society for lending me influence and support, without which the little that has been done by this expedition could not have been accomplished.

2. A Note on the Hatching-stage of the Pagurine Laid- crabs. By L. A. BORRADAILE,M.A., F.Z.S., Lecturer in Natural Sciences of Selwyn College, Cambridge. [Received October 12, 1899.1 The life-history of the Land-crabs of the family Ccenobitidre is one to which considerable interest attaches, and of which, at present, nothing appears to be known. The family comprises the genera Birgus and Camobita, the robber- or coconut-crab and the land hermit-crabs, ail of which have given up a sea life for one on land. It need hardly be remarked that changes in habitat, particularly from sea to land or fresh water, have frequently necessitated the suppression of larval stages in the life-history. Among Crusta- ceans the instances of the cry-fiahes, the ditch-prawn (Pahmonetm uarians), the freshwater crabs (Poturnon), and at least one species of land-crab (Geccrrcinus) come at ouce to mind. The pm- sibility was thus suggested that the land-pagurines might also have lost the whole or a part of their larval life, and leave the egg in something like the adult condition. On the other hand, it bad to be borne in mind that some species of land-crabs and all the strand-crabs (Ocypoda, &c.) retain the habit of setting free xoea-larvs in the sea, where they pass through their earlier stages. It was probably with these considerations in his mind that vou Willemoes-Suhm, when, in October 1874, the ‘ Challenger ’ arrived at Zamboanga in the Philippine Islands, wished to investigate the development of the robber-crab from the egg. Unfor- tunately the time of year rendered it impossible for him to do this, but he was told by an LL intelligent native ” that the young were born resembling the parent. This statement has since been accepted in a tentative manner by certain text-books, in spite of the fact that the small size of the eggs (and, indeed, of the female genital opening) made it improbable that the development was a direct one, depending on food-yolk. As for the statement of the 935 YE. L. A. BORRADAILE OX TUE PAGCRINE LAND-CRABS. [XOV. 28, native, it is only what was to be expected. A ‘‘ native,” being un- familiar with the idea of a metamorphosis, will always give the same answer to any question on the subject, namely that the young are born exactly like the adult-but very small. All doubt on this point has, however, now been removed by the discovery by Dr. A. Willey of a female Birgus on the rocks at the brink of the sea at Lifu in the Loyalty Islands. The abdomen of this specimen was covered with hatching zoam which were being washed off into the water. The time of the year was the month of January l. With regard to the genus Canobita, I have myself recently taken specimens of two epecies (C. rugosus and C. perlatus) in Ceylon and the island of Minikoi in the months of May and June, bearing zoaas. The animals were taken on the stretch of wet sand just above the waves. The catches of the tow-net in the island of Minikoi not having yet been examined for specimens of the zoaa of Canobita, the possibility is not coinpletely excluded that the larva may undergo the whole or a part of their development within the shell of the mother, which is always wet with salt water. The larva, however, did not, on a cursory examination, give any indication to justify such an assumption, and when placed in sea- water lived for a short time and showed powers of swimmiug in B lively manner. An attempt to rear them nnfortunately failed, but this was only to be expected in view of the known difficulty of the operation. It is perhaps worth noticing that the above two species of Coenobita are those whose habits keep them nearest to the sea. If an abbreviated development is to be found in the genus, it woiild more probably occur in forms such as C. spinosus which live at considerable distance from, or at least a considerable height above, the sea. Prom the observations just recorded, it is clear that the early stages of the development of the C’wnotrititlm present no ver-y remarkable features. It may be presumed that later stages follow the ordinary course. The only points of interest remaining for investigation are the assumption of the adult form by Bi?yue and the transition from sea to land, which we may hope to have described by some future traveller in the Pacific Ocean.

The prnsent writer is under great obli ation to Dr. Willey for Landing over to him a number of these zow, which w& be described and figured in Part V. of Dr. Willey’s ‘‘ Zoologird Resulte” now heing published by the Cambridge University Press. It io intended also to poblisb an account of the larvte of Cmobita.

PZ.S 1899 PL.LXlll

10.

I 2 3 @! 4 4: 4:

P.Z.S.1899. PL.LX1V.

i 2 3 4

2 3

2 2 3 3 4 4

BaledDanizlsson L!d imp.

EFFERENT BRANCHIAL VESSELS OF TELEOSTEAN FISHES.

P.Z.S.1599. PL.LXV

I 2 3 4

I e 3

EFFERENT BRANCHIAL VESSELS OF TELE0STEA.N FISHES. 1899.3 ON THE BLOOD-VESSELS OF TBLEOSTEAN FISHES. 939

3. On the Relations of the Eff2rent Branchial Blood-vessels to the Circulus Cephalicus ” in Teleostean Fishes. By W. G. RIDEWOOD,D.Sc., F.L.S., Lecturer on Biology at St. Mary’s Hospital Nedical School. [Received Auguat 1, 1899.1 (Plates LXII1.-LXV.)

INTBODUCTION. When engaged in dissecting a Herring some years ago, I was struck by the fact that the four efferent branchial vessels all reached a median vessel which was continuous behind with the dorsal aorta, and that the circulus cephalicus was situated anteriorly to the first pair of vessels. This condition was 80 totally differvent from what I knew to be the arrangement of the vessels in the Cod, that I examined the same parts in a third form, the Salmon, and here found a condition intermediate between the two preceding. The results appeared to warrant a further inquiry, and the present investigation was undertaken with a view to ascertaining what are the commonest, and what the extreme modifications of the etferent branchial arteries to be met with among Teleosteau fishes. The greater part of the work was done during the summer vaca- tion of 1892 at the Marine Laboratory of St. Aiidrews, Scotland ; but, owing to the difficulty of making the series of fishes thoroughly representative, the investigation has been protracted over a period of seven years. Late, however, as it may now appear, I take the opportunity of expressing my warmest thanks to Prof. W. C. McIntosh, M.D., F.R.S., for his kindness in placing at my disposal for six weeks during the summer of 1892 all the facilities that are afforded by the 8t. Andrew Laboratory for the procuring and in- jecting of the indigenous fishes. I have also to thank Prof. G. B. Howes, LL.D., F.B.S., of the Royal College of Science, London, for specimens of Pema, Trachinus, Lophius, Fistulavia, Motella, Ammodytes, Silurus, Exocmtus, &ox, Albula, MeguIops, Chiroeentrue, and Hippocawgms, and Mr. Q. A. Boulenger, F.R.S., of the Katural History Museum, for specimens of Cowinn, Equula, Goobius, Sphyrcenu, Hemichroniis, Clurias, Saccobrunchus, Malapterurus, C’nhhthye, Scopelus, Cobitis, Illnrczruenius, Balistes, Tetroclon, and Orthagorisczts. The series of forms examined includes 61 species belonging to 57 genera. 940 DR. Vi. 0. RIDEWOOD ON THE "ov. as,

LI@TOF 8l'EOIES EXAMINED. Ophidiidw. ACANTHOPTERYOII. Ammodytes lenceolatus. Ammodytes tobianus. Percidse. Pleuroucctidse. Perm flnviatilis. Hippoglossus vulgaris. hbrax lupus. Pleuronectas flesua. Mullidm. Mullus barbatw. Sciaenidse. PHFSOSTOMI. Corvina nip. Siluridm. Canmgidse. Ularias magiir. Equula edentula. Saccobranchus fossilin. Cyttidse. Silarns glanis. Zeus fnber. Liocassis longirostrin. Scombridse. Malapterurus electricus. Scomber acorn brus. Cnllichtbyn littoralis. Trachinids. Saopelidse. Trachinns draco. Sco elua boops. Pediculuti. CgpriniL. Lophiiis pismtorius. Oyprinus carpio. Cottidse. Hypopbthalmichtliysnobilis. Cottan acorpio. Cobitis twnin. Trigla garnardus. Scoinhresocidm. Trigla cuculus. Exoraetus bracliysoma. Discoboli. Esocidre. Ogcloptarus lumpus. Esox luciue. ff obiidie. hlormpridm. Oobius giuris. Marcusetiius plagiostorna. Blenniids. Salnionidse. Anarrhichas lupus. Salmo sdar. Blenniuv pholis. Salmo truth. Cantronotun gunellus. Osinerus eperlanua. Zoarcm viviparus. Coregonus oxyrhynchtib. Sphyrsenidae. Clupeidse. Sphyrtena vulgaris. Enpraulia encrnsicholus. Mugilidse. Clupeu harengus. Mugil rapito. Clupa sprattus. Glastrosteidse. Albula conorhynchus. Cfastrosteus spinachia. Megalopa cyprinoicles. Fistulariidse. Chirocentridie. Fistularia tabaccaria. Ohirowntrus dor8b. Cfobieeocidte. Murmnidie. Lepadognster gouanii. Anguilla vulgnris.

PHARYNGOGNATHI. LOPHOBRANCH€&. Labridse. Syngnathidle. Labrus maculatue. Sgnpnathua acus. Chroniidea. Hippocam p~sgii ttuln tus Re~ichromiafasciatua. PLECTOGXATHI. ANAC ANTHINT. Scleroderini. Gadid%. Bu lia teq twilent u s. Gadus segtefinw. Gym n oclontea. Molva vulgaris. Tetrodon palembangensis. Notella tricirrhnb. Orthagoriscue trunmtiis. ia99.1 BLOOD-TESSELS OF TELEOSTEAN FISHES. 941 LITERATURE. On searching through the literature of the subject, one cannot fail to be struck by the fact that the arterial system of Teleostean fishes has been very greatly negIected, and the disposition of the efferent branchial vessels particularly so. Meckel and Hyrtl appear to have been the only anatomists to undertake anything like a systematic study of these latter. Meckel in 1831 (14. p. 192) pointed out that the mesial ends of the third and fourth efferent branchial vessels are usually close together, and unite with one another before joining the aorta; and he furnished a few observations on the efferent branchial system of Giidus, Trigln, Perca, Pleuronectes, Lophius, and ilfitrcrxa. Seven years later Hyrtl (7) gave a table showing the proportions of the circulus cephalicus in fourteen species of Teleosteans, supplemented by a considerable amount of information concerning the vessels associated with the circulns, and good figures of the efferent branchial system of Perca, Gudus, and Y‘incci. Staniiius in 1849 (23. pl. v.) published some fairly reliable figures of the efferent vessels of Cyclopterus, Gadus, Salnto, and Sconaber, but the blood-vessels were only introduced into his figures to act as landmarks for the recognition of the sympathetic nerves, and must not be treated too critically. In his ‘Handbuch’ of 1854, however (24. p. 242), he made reference to the fact that the circulus cephalicus is wide in Gatlus and Lota, where all the efferent branchid vessels open into it, whereas it is narrow in Scomber and Sdmo, in which genera the last two open directly into the aorta. The only other information on the subject is that conveyed by the figures of the Carp by Duverney (6. pl. ix. figs. 17 and 18), tho Perch by Laurillard in Covier’s ‘ Histoire Nat. des Poissons ’ (5. pl. vii. fig. l),the Cod by Miiller (16. pl. iii. fig. 13), the Trout by Vogt (1. pl. L. fig. 2), the Pike by Maurer (13. pl. xi. fig. l), the Cod by T. J. Parker (20. p. 117), and the contributions by Hyrtl on Beterotis (9), Gymnarc7~us(11)’ Chnnos (12), and other genera. GENERALPART. As may be gathered from the title, the observations recorded in this paper concern the efferent branchial vessels and the vessels formed by their confluence. The cceliaco-mesenteric and subclavian arteries usually arise in relation with the hinder part of the circulus cephalicus, or with that part of the aorta which receives the third and fourth efferent branchial vessels. The positions of these arteries are indicated in the figures, and occasional references are made to them in the text; but the investigation does not profess to deal exhaustively with these vessels, nor with the hyoiclean, anterior carotid and posterior carotid arteries, which also are associated with the circulus cephalicus. The dotted liues in the figures si8uify that owing to the failure of the injection-mass to pass, or owlng to the siuall size of the fish PEOC. ZOOL. SOC.-lb%), KO. LXI. 61 942 DR. W. G. BIDEWOOD ON THE [Nov. 28, and the consequent dif6culty of dissection, or to the bad state of preservation of the parts, it was not possible to trace out the vessels so delineated, but that, from analogg with other forms, there is reason to believe that the vessels occupy the poditions indicated. There appears to be no reason to suppose that the circulus cephalicus is ever incomplete in front. The lumen of the transverse commissure between the roots of the anterior carotid arteries may possibly be closed in some cases, but the failure of the injection-mass to pass into the commissure is not necessarily a proof of the fact, since the pressure duriug the process of injection is equal on the two sides of the circulus. The transverse vessel is usually of small size, and its traversing the parasphenoid bone makes it difficult to dissect out with any degree of neat,ness. In the selection of characters by which to classify the various types of arterial disposition, it has been assumed t.hat the condition found in Clupea (Pl. LXIII. fig. 2)and Eyruulis (fig. l), where the circulus cephalicus is small, and does not involve the second, third, and fourth efferent branchial vessels, is the most simple and primitive, and that the connection of all four efferent branchial arteries with the circulus, such as occurs in Cfadus (PI. LXV. fig. 34), is the most specialized. This assumption is based partly upon the fact that the Gadoids are highly specialized in numerous other respects, whereas the Clupeoids-are generally recognized as among the lowest of the Teleostean series; partly upon the fact that in Amin, an admittedly primitive Ganoid with Clupeoid affinit.ies, the last three efferent branchial vessels are unconnected with the circulus cephalicus j partly also upon the researches of Ayers (4) upon Elasmobranch fishes, which go to prove that the right aird left sides of the circulus cephalicus are not the primitive paired aortae such as occur in Amph%o.Tusand in embryos of the true Vertebrata, but that the true dorsal aorta may persist as a median vestigial vessel traversing the circulus cephalicus, in the same nianner as, according to Miiller (l0), it does in the Cyclostomi. Pursuing this line of argument, we may legitimately conclude that where, as in the Salmon (fig. 7), Mackerel (fig. 6), and Carp (fig. 13), the circulus cephalicus, receiving the first and second efferent branchial vessels, is separated from the point of entry of the third and fourth by a length of the median aort.a, the condition is more primitive t’han that in which the third and fourth vessels open at the posterior extremity of the circulus cephalicus, as in the Base (P1. LXIV. fig. 17). And further, the separation of the third and fourth vessels in t.he Anchovy (fig. 1) by a portion of the aorta indicates a more lowly condition than that seen in the Herring (fig. 2), where the two vessels open close together. There are thus two lines upon which we may regard specialization as proceeding : firstly, by the circul us cephalicus If such exists. The circulus appears to be suggested in Auis’s figure (3.pl. xxxvi.), but, itsexistence is denied by Ramsay Wright (25. p. 495). The arrangement in Lepidostcus is somewhat similar to that of’ Amia. See Hprtl, 8. p. TYD, and Mdler, 17. pl. v. fig. 6. 1899.1 BLOOD-VBSSBLS OF FELEOSTBAN FISHES. 943 involving the second, and later the third and fourth efferent branchial vessels ; and secondly, by the progressive suppression in length of the median aorta, bringing about an approximation of the dorsal or proximal ends of the last three efferent branchial vessels on each side. Owing to the fact that the one variety of specialization may occur quite independently of the other, or in conjunction with it, it becomes very difficult in some cases to compare the ultimate degree of specialization attained, since there is no evidence to show whether the inclusion of the efferent vessels into the circulus or the approximation of the efferent vessels by the suppression of the aorta is the more important. In the Salinon (fig. 7), for inetance, the second efferent vessel opens into the circulus cephalicus-an indication of specialization ; but a length of aorta persists between the circulus cephnlicus and the third efferent vessel-a primitive character. In Baliates (fig. 5) the second vessel is free from the circulus cephalicus, and yet there is obvious specialization in the complete suppression of the aorta in the branchial region, resulting in the second, third, and fourth vessels opening close together, immediately behind the circulus. Who shall say whether, in the dispogition of the efferent branchial vessels, the Salmon or the File-fish is the more primitive’? Having recourse to the other anatomical feature8 of these two forms, one would conclude that, the Salmon being in general structure the more primitive, the abbreviation of the aorta is as a mode of specialization more irnportant than the backward extension of the circulus to include the second efferent branchial vessels. The conclusion is further justified by the fact that Albula (fig. ll), which is undoubtedly allied to Megalops (fig. 4) and Chirocentrus, differs from these genera in this lat,ter respect. This hypothesis, however, opens up the further question as to how far a backward extension of the circulus cephalicus is due to the longitudinal splitting of a part of the median aorta. Has, for instance, the condition found in Megalops and Chiroeeritrtle,in which the circulus cephalicus extends back to the second branchial vessels, been brought about by the longitudinal division of a median vessel such as exists in Clwpea (fig. 2) and hhgrawlis (fig. 1) between the first and second efferent branchial vessels? The suggestion has much to recommend it; more especially as the suppression of the median aorta cannot have operated here, or the two anterior carotids would be arising close together at the bottom of the fork of the first efferent branchials. Another line of specialization, independent of the two former, can be traced in the confluence ot the third and fourth efferent vessels. Having assumed that the separation of two consecutive efferent branchial vessels by a portion of the median aorta is a primitive feature, it follows that the separate entry into the aorta. of the third and fourth vessels in hhpmlis (fig. 1)is au iudication of less specialization than the debouching of the two vessels together, as in Clupeci (fig. 2) ; and further, that, this latter Cl * 944 DR. W. G. RIDEWOOD ON THE [Nov. 28, condition is more primitive than that found in Veplops (fig. 4), in which the third efferent vessel unites with the fourth on each side to form a short common trunk which carries the double charge of blood to the aorta. If this line of argumentation be extended to the cases in which the third and fourth vessels open into the circulus cephalicus instead of into the aorta, we are led to the conclusion that Blennius (fig. S.?), having the Y -shaped system, is inore specialized than Gadus (fig. 34), where the third and fourth vessels are disposed in the form of a V, and that this latter is more specialized than Synpathzcs (fig. 33), where the two vessels find separtite outIets into the circulus. Whether much importance, however, can be attached to this last feature is open to considerable question, for the V-shaped system obtains in Srrlmo (fig. 7) and the Y-uhaped oue in Osmerus (fig. lo), and a similar relation exists between the Chinese Carp ~f~~o~hthnlmic7~tli~~and our native Carp Cyprinus (fig. 13), while among the Siluroidu, a premmably natural arsemblage of forms, there are gradations from CiiLlid~tl~ys(fig. 19) and Liocassis (tig. 18) with the V-shaped system, through Claricrs (fig. 20), and Silurus (fig. 33), to Saccobranchus (fig. 31), with a typical Y-shaped arrangement. The disposition of the eft'erent branchid vessels is independent, of the shape of the head, except in so far as the slope of the vessels and the shape of the circulus is concerned. There is no connection, that is to say, between the shape of the head and the degree of suppression of the median aorta, or the entrg of the efferent branchial vessels into the circulus rather than into the aorta. In long-headed forms like Amnaodytes (fig. 3), SphyrcjPna, Fi8tzclaria (fig. 30), AllcJualla (fig. 16), and Syngncithzis (fig. 33) the circulus cep1)nlicus is elongated in an antero-posterior direction ; while in Cottiis (fig. 27), Lophius (fig. as), and others, with a broad, flat head, the gills are widely separated, and the rircnlus csphslicus is proportionately broad, the c~i~inontrunks formed by the fusion of the third and fourth efferent vessels being also lengthened. The differences in the arrangement of the efferent branchial vessels relatively to the circiilus cephalicus aid the aorta are not correlated with any differences in the position and extent of development of the epipharyngeal dentition. At the commencement of the iiiquiry, the suggestion occurred to my mind that the development of a large and elaborate dental appzratus might, by some process of natural selection, have resulted in the blood-vessels taking up a position of safety, out of the line of direct pressure between the epiphnryugeal bones and the vertebral centra or the base of the skull. A minute examiiiation of the individual cases shows, however, that the vessels do not experience any displacement under the circumstances, but obtain sufficient protection by running in grooves or arches in the epipliaryngeal bones ; in fact, id9 often as not, the epipharyngeal teeth lie immediately below certain of the efferent branchial blood-vessels. As an instance of two forms 1899.) BLOOD-VESSELS OF TELEOSTEAN FISHES. 9 45 with very similar vascular arrangement, but with widely different epipharyngeal dentition, may be mentioned the Salmon and the Mackerel. In the Salmon there are a pair of small patches of teeth borne by the fotirth epibranchial bones, lying inimediately ventral to the aortic extreinit,ies of the third efferent branchinl vessels; but in the Maclierel there are a pair of great den- tigerous pads underlyiiig the mesial eiids of the second, third, aud fourth efferent bmichial vessels, as well as a considerable part of the circulus cepha1ic.u~. In the Carp the dorsal aorta obtains the necessary protection by actually traversing the great horn-covered bony projection of the basioccipital, against which the lower pharyngeal teeth bite. In the Wrasse the vascular arrangement conforms to a very average trype, being apparently quite unaffected by the large and elaborate pharyngeal mill developed in the vicinity of the posterior half of the circulus cephalicus. In &oc.mtzts, again, the epiphaqngeal pad is of relatively enormous size, underlying t,he upper extremities of all four efferent branchial vessels, tlie anterior ends of the aorta and cceliaco-inesenteric artery, aiicl the whole of the circulus cephalicus except the extreme anterior part ; yet, there is nothing very remarlrable in the disposition of these ressels which might be accounted for by their relation to the epipharyngeul apparat,ns. Only in Gohius, Cottw, aud AloZvn, of the forms examined, are tlie epiphargngeal dental pads situated ent,irely within the circulus cephalicus. The evidence afforded by forins devoid of epipharyngeal teet.h is probably inconclusive, on account of the great possibility of the edeutulous conclitioii haviiig been arrived at independently iii diit'ereiit groups of fishes. In the Sprat and Pipe-fish the types of \~asculararrarigeinerit are widrly divergent. The Sprat closely reseiubles the Herring (fig. 2) ; the yessels of Syngnatlius are shown in tigure 33. In order to discuss intelligibly the different forms of vascular arrangement met with, some forin of clamification, however artificial, is essential ; and after careful consideration I have found it convenient t,o adopt the following scheme, based on the relations of' the effereut branchial vessels to the circulus cephalicus and the dorsal aorta. In consideratmion of the unsatisfactory nature of the present classification of Teleostean fishes, B purely artificial scheme, founded upon the single character which forms t,he basis of the present communication, is likely to prove of more permanent utility for purposes of subsequent reference, than one which relies upon a classification \I-hicb sooner or later may prove to be an unnat,ural grouping. The great majority of the forms studied will be seen to come under the headings B and C, while the rarer and more extreme modifications occupy terminal posit,ions in the classificatory scheme. The division of the groups B aud C into the subgroups 6 and c, differing only in the t,ransverse or obliqce position of t,he confluent, third and fourtli efferent branchiu vessels, appears arbitrary, bnt in prnctice there are very few tornis which fai to 946 DR. W. G. RlDIWOOD ON TEE [Nov. 28, fall definitely into the one or the other subgroiip. Without this subdivision, such remarkably different forms as Anarr?&hm (fig. 23) and Z"ri&x (fig. 28) would come under the same heading.

CLASSIFICATIONOF THE GENERAEXAMINID. Group A. The first efferent branchial vessel opens into the circulus cephalicus, but the second does not. Subgroup a. The third and fourth vessels open into the median aorta sepnrately.-E/l//rcl167is (fig. 1). Subgroup b. The third and fourth vessels open into the median aorta together.-Clupen (fig. 2), Ammo(l!~tes (fig. 3). Subgroup c. The third and fourth vessels on each side unite to form a common trunk, which reaches the aorta some distance behind the second vessel.- Chirocentrus, Megalops (fig. 4). Subgroup d. The third and fourth vesssls on each side unite to form a common trunk, which reaches the median aortn immediately behind the second vessel.- Bnlistes (fig. 5). Group B. The first and second efferent brnnchial vessels open into the circulus cephalicus, and the third and fourth iuto the median aorta at some distance behi;id the circulus. Subgroup a. The third and fourth vessels open into the aorta together.-Scomber (fig. S), H~~popl~tl~nlnzich~?~ys, Xalmo (fig. T), Coreyonus. Subgroup b. The third and fourth -vessels unite before opening into the aorta, the common stem being transverse to the length of the body.-Mu$ (fig. S), Ilippoglossus (fig. 9), Pleuronectes, Osmerus (fig. lo), Alb& (fig. 11). Subgroup c. The third and fourth vessels unite, the common stem sloping backwards towards the aorta.- Mu7aptemrw (fig. 12), Scopelus, Cyprinus (fig. 13), Cobitis (fig. 14), Esm (fig. 15), Mcmusenius, Anguilla (fig. IS). Group C. The first and second efferent branchial vessels open into the circnlus cephalicus, and the third and fourth into the aorta immediately behind the circulus. Subgroup a. The third and fourth vessels open into the aorta together.-Labraa (fig. 17), Lwcassis (fig. 18), Ccillichthys (fig. 19), Clnrias (fig. 20). Subgroup b. The third and fourth vessels unite before opening into the aorta, the common stein being transverse to the length of the body.-Percu (fig. 21), 1899.1 BLOOD-VESSELS OF TNLEOSTEYIN FIRHES. 947 Mullus, Corvinn, Equuln, Zeus (fig. 22), Trachinus, Anurs*?iiahas(fig. 23), Zoarces, Centronotus, Sphyrane, Labrus (fig. 24), Utmichromis, Exomtus (fig. 25). Subgroup c. The third and fourth vessels unite, the common stem sloping backwards towards the aorta.- Lophius (fig. 26), Cottus (fig. Z),Trigla (fig. 28), Cyclo- pterus (fig. 2M), Gobius, Fktularin (fig. 30), Lepaclo- gnster, Snec-obranc?ius (fig. 31), SiZurus (fig. 32). Group D. The four efferent branchial vessels open into the circulus cephalicus. Subgroup a. The third and fourth vessels open separately. -Synpccthw (-fig. 33), L%fotebki. Subgroup b. The third and fourth vessels open together.- Gastrosteus, Gadus (fig. 34), Molvci, Hippocampus. Subgroup c. The third and fourth vessels unite before joining the circulus cephalicus.--Bleniiius (fig. 35), Orthagoriseus (fig. 36), Tetroclon (fig. 3’7).

SPECIAL PART. In the case of species not figured, comparisons will be drawn with those figured forms which, in the arrangement of the efferent brnnchial vessels, they most nearly resemble, irrespective of the degree of affinity which upon other grounds may be considered to exist between the forms.

FURTHERREMARKS ON GROUPA. In Engrnulis (fig. I), there is a eonsidernble difference in the size of the four efferent branchal vessels; the second is the broadest, and the first the narrowest. The dorsal aorta behind the fourth branchial vessel is wide and very thin-walled. This is also the case in CZupea. Although the scheme of classification which it bas been found convenient to adopt brings Ammoclytes (tig, 3) under the same subgroup as Clupea (fig. 2), there are several features which go to show that the association is an uiinatural one. The circulus cephzlicus in the former genus is much longer than broad, and extends bark nearly to the point of entry of the second pair of vessels into the aorta; whereas in Clupea the posterior part of the circulus is transverse to the axis of the body, and forms with the first part of the aorta a T rather than a Y. The cceliaco-niesenteric artery arises in Ammodytes close behind the fourth efferent vessele, but much further back in Clupea. The origin of the subclavian arteries is slightly more posterior in Clupea than in Anzmodytts. There are no differences between Ammoclytes lanceolatua and A. tohianus, nor between flupea hwengus and C. sprattus. Chirocentrus closely resembles Mega10178 (fig. 4). In bot,h genera the median aortic stem found in Engraulis and Clupea 948 DR. W. 0. RIDEWOOD ON THl [Nou. 28, between the first and second efferent branchial iressels is wanting, owing to the backward extension of the circulus to the point of entry of the second branchial vessels into the aorta. The same feature is to be observed iri Baliates (fig. Fj), where the aortic stem between the second vessel and t,he comnioii trunk of the third and fourth is also suppressed. The cceliaco-mesenteric artery of Balistes arises, not from the aorh, but from the third and fourth branchial vessels of the right side, immediately after their anastomosis. If we disregard the modification of the efferent branchial circu- lation brought about in G?/ntiini~c7iitsby the increased functional importance of tlie swim-bladder, the efferent branchial system of this genus can be seen, from thtl desoription and figure published by Hyrtl (11. p. 11, and 111. 4. fig. 4), to conform with the type which characterizes subgroup A c.

FURTHER REMARKSON GROUPB. The efferent branchial system of Seombw (fig. 6) bears a cloPe resemblance to that of Snli)?o(fig. ’i),but t,hr cceliaco-mesenteric artery arises from the aorta immediately behind the fourth efferent branvhial vessels and the subclavian arteries some distance fiirther back, whereas in Salmo the positions of the cceliaco-niesenteric and subclavian arteries are reversed. In the figure of Scomber given by Staunius (23. pl. v. fig. 4) the circulus cephalicus appears much too large, and the iihrval between the circulus and the entry of the third and fourth branchial vessels into the aorta too short. In Salmo truttn, as also in Osmerus and Coregonus, the dorsal norta is wide and thin-walled ; but this is not the case in A”a2~no sulnr. The distance between the posterior angle of the cireulns and the point of entray of the third efferent branchial vessel is proportionately longer in C‘oregoi~~isthan in Salmo sdcw, and proportionately shorter in Sdmo tmtta. A very reliable figure of the efferent branchial system of S. fruttn has been given by Vogt (1. pl. L. fig. 2). The figure by Stannius of S. snlnr (23. pl. v. fig. 3) is incomplete, but is correct so far as it’ goes. In Hypo- p7~thalmic7~thysboth the cceliaco-mesenteric arid the subclavian arteries arise some distance behind the fourt,h efferent vessel. The ctirculus in this genus is broader than long. In H+po.qlossus (fig. 9), Pleuronectes, Mu!$ (fig. 8), Esom (fig. 15), and Mawusenitis the distance between the posterior angle of the circulus cephalicus and the point of entry of the common trunk of the tjhird and fourth branchial vessels is so short that these forms epproach somewhat closely those included in subgroups CI, and Cc. In Pleuronectes the circultis cephalicus is longer than brond, whereas in Hippo!Ilossus it is broader than long ; the anterior carotids are closer together, and the transverse commissure between them is shorter than in Ilippoqlossus. Tlie common trunk formed by the union of the tliircl and fourth branrhial vessels of the right side is much shorter than that 1899.1 BLOOD-VE88EL8 OF TELEOSTEAN FIBHBS. 949 on the left, and the origin of the cceliaco-mesenteric artery is nearer the middle line than in Hippoglossus. In Mugil (fig. 8) the circulus cephalicus is considerably longer than broad, and the cceliaco-mesenteric artery arises from the median aorta behind the entry of the third and fourth branchid vessels, and not as in the two preceding genera. In Osmerus (fig. 10) and Albula (fig. ll), also, the coeliaco-inesenteric artery arises from the median aorta, but the subclavian arteries take their origin immediately behind the point of entry of the common trunk of the third aud fourth branchial vessels into the aorta, and not behind the coeliaco-mesenteric artery as in Mugil. In Scopelus the circulus cephalicus is small and nearly circiilar in shape, and the portion of median aorta intervening between its posterior angle and the mesial ends of the common trunks of the last two branchial vessels is unusually long. In Malapterurus (fig. 12) and Esox (fig. 15) the coeliaco-mesenteric artery arises from tlie aorta immediately ventral to the mesial ends of the common trunks of the third and fourth vessels; in Ctp-inics (fig. 13), Cohitis (fig. 14), and Marcusenius it arises more posteriorly, and in Anguilla (fig. 16) considerably farther back. Except in this latter respect, the arrangement of the vessels in Marncsenius very closely resembles that of Esox. Judging by the excellent figure of Tinca published by Hprtl (7. pl. iv.), this genus exactly resembles Cy2~iv~usin the dispo- Bition of the vessels of the efferent branchial system. The efferent vessels of Cyprinus \\ere figured by Duverney (6. pl. ix. figs. 17 & 18) nearly 140 years ago; and although the figures are incomplete, the essential features are correctly represented. In Cypri- nus, and according to Hyrtl in T'inca also, there are two pairs of subclavian arteries. The anterior pair, arising in front of the common trunks of the last two branchial vessels, supply the upper part of the pectoral arch, the posterior pair the lower part of the arch and the pectoral fins. A somewhat similar arrangement obtains in Esox. The mode of origin of the anterior pair has been remarked by Miiller, and quoted by Stannius (22. p. 103, footnote 3) and Owen (18.p. 270, and 19. p. 4891, and these vessels are bhown, although not named, in Maurer's figure (13. pl. xi. fig. 1). The circulus is large in Malapterurris and Anguilla; and in the latter genus a median vessel, occupying the position of the anterior continuation of the primitive median aorta described in Selachian fishes by Ayers (4), may be traced forward from the posterior angle of the circulus. It soon forks, and is ultimately lost in the mucous membrane of the roof of the pharynx. In Anguilla, also, the origin of the posterior carotid artery is much farther removed from the entry of the first branchial vessel into the circulus than is usual. Meckel has stated (14. p. 193) that in the marine Eel, iVurcenophis helena (Murcena helena), the anterior lateral trunk formed by the union of the first aiid second branchial vessels is three times as long as the posterior one formed by the union of the third and fourth. 950 DR. W. Q. RIDEWOOD ON THE] [Nov. 28, The efferent branchial vessels of Chanos have been described and figured by Hyrtl (12. pl. i. fig. 1). The first and second vessels open into the circulus cephalicus, the third some distance farther back into the median aorta, while the vessels from the fourth gills, after uniting with those from the epibranchial organs, unite with one another and open into the aorta at a point as far behind the opening of the third branchial vessel as the latter is hehind the circulus cephalicus. The genus thus falls into group B, but, owing to the exceptional disposition of the fourth pair of vessels, it cannot be included in any of the three subgroups recognized. FURTHERREMARKS ON GROUPC. The circulus cephalicus is narrow in front in Labrax (fig. 17), but it is broad in the Siluroids Liocassis (fig. le), Callichthys (fig. 19), and Clcirins (fig. 20). In Lahrrra the cceliaco-niesenteric artery arises from the aorta immediately behind the fourth efferent branchial. It has the same relations in Clarias, but is somewhat more posterior in Callichthys, and considerably so in Liocassis. The subclavian arteries arise behind the cceliaco-mesenteric artery in Lahrcix, Clarias, and Callichthys, but close behind the fourth branchial vessel in Liocccssis. Although CEnrias is introduced into the subgroup C a, it really occupies an intermediate position between C a nnd C b, since there is a very short common trunk on each side between the last two branchial vessels and the aorta. The genus Heterotis I have not been able to examine, but it is tolerably certain, from the description given by Hyrtl(9. p. 87), that it should be included in the subgroup Ca. In Mullus and Corvina the disposition of the cceliaco-mesenteric and siibclavian arteries is as in Perm (fig. 21), but the circulus cephalicus is broader in front. The circulus has the form of a regular heptagon in Afullus, while in Corvina it is pear-shaped, the broad end being anterior. In Equula it is oval in shape and longer than broad j but otherwise the relations of the parts are as in Perca. With regard to Peren itself, Hatchett Jackson, in his edition of Rolleston’s ‘Forms of Animal Life ’ (21. p. SS), states that the “ cceliaco-mesenteric artery * * * springs from the right epibranchial artery before it fuses with its fellow.” This does not accord with my own observntions. Hyrtl in his figure of Lucio- perca (7. pl. i. fig. 1) shows the common trunks of the third and fourth branchial vessels entering the aorta at some distance from the posterior angle of the circulus. The transverse commissure, also, between the anterior carotids (“ die vorderen oder kleinen Kopfarterien ”) is situated farther forward than in Perccc. In Zeus (fig. 22) the circulus is broader, and the cceliaco-mesenteric artery arises, not from the aorta, but from the common trunk of the last two branchial vessels of the right side. Zoarces resembles Anarrhichas (fig. 23) in the oval shape of the circulus and in the narrow anterior prolongation of the latter, but tlie 1899.1 BLOOD-VESSELS OF TELEOSTEAN FISHES, 951 common trunk formed by the union of the third and fourth branchial vessels is shorter. In Centronotus and Trachinus the anterior prolongation of the circulus is wanting, but otherwise the vessels o€ the pharyngeal roof are disposed as in Aria+ rhichns. The circulus of Spl~yranccis twice as long as broad, and the posterior angle is very acute ;the subclavian arteries arise farther forward than in Anarrhichns, and have more the relations of those of LaBrus (fig. 24). Hemkhromis differs from fiibrus in having a much broader circulus cephalicus, and in the more posterior origin OF the subclavian arteries. The circulus of Emccztus (fig. 25) is oval and longer than broad ; the cceliaco-mesenteric artery arises from the circulus cephalicus immediately to the right of the aorta. Owing to the absence of the fourth gill and its efferent vessel iri Lophim (fig. 26), the right of this form t,o rank under subgroup Cc rather than Ca is somewhat conjectural. The matter is, however, of no great importance. The circulus is very wide, and the cceliaco-mesenteric artery, which is considerably thicker than the aorta, branches soon after its origin. Concerning Lophius Mecltel has written (14.p. 192): “Lophius yiscatorius hat, shtt der gewiihulichen drei bis vier, nur zwei sehr lange Kiemenblutadern. Von der vorclersten Kieme entsteht ein einfacher Stamm, der zweite wird durch die Vereinigung der xweiten nnd dritten Kie- menblutader gebildet, die ungeffihr eben so lang als der geniein- schaftliche Stamm getrennt verlaufen.” My own observations are thus not in accord with those of Meckel. Most ichthyologists admit, with Muller (17. p. 47), that the three gills present in Lophius are the anterior three of the four gills more normally present ; and t,he coupling of the first,and second efferent branchial vessels, the third remaining solitary, is by analogy with allied forms a far more intelligible arrangement than that described in the above-quoted passage from Meckel’s text-book. In Cottus (fig. 27), although the circulus is so wide in front, the anterior carot,icIs lie very close together, and the transverse commissure, which takes 11curious bend forward, is therefore short. The cceliaco-mesenteric artery of Trigla (6g. 25) is a double vessel arising from the common trunk of t,he last two branchial vessels of the right side. There are no differences between the efferent branchial systems of Trigln cuculue and T. gurn~crclw. The anterior part of the circiilus cephalicus of Cyclopterus (fig. 29) is very curioualy shaped, and the transverse vessel may possibly be wanting. I. have been unable to find it in the three specimens dissected. In the figure of Cyclopterua given by Stannius (23. pl. v. fig. 1) t>he subclavian arteries are drawn too wide ; and they are incorrectly described on page 156 as branchial veins. Lepdogaster does not differ materially from Cyclopterus, except that the anterior part of the circulus cephalicus conforms more to the normal type. Gohius differs from Cottus (fig. 27) in the wider separation of the anterior carotids, the greater breadth of the circulus cephalicus, the more posterior entry of the second 952 DR. W. 0. BIDEWOOD ON THE [Nov. 28, branchial vessels into the circulus, and the more posterior origin of the subclavian arteries. In Fistuluria (fig. 30) the slope oE the common trunks of the third and fourth efferent branchial vessels is so slight that the genus might with equal propriety be classed under subgroup C b, and the fact that these common trunks do not enter exactly at the posterior angle of the circulus cephalicus makes it dificult to uphold its claim to come into group C at all. The circulus is long and abruptly terminated in front. The aorta, after giving origin to the coeliaco-mesenteric artery, is unsymmetrical, and runsto the left side of the vertebral centra. In Silztrus (fig. 32) the posterior angle of the circulus cephalicns is not exactly coincident with the mesial ends of the common trunks of the efferent branchial vessels 3 and 4, as it is in Saccobrnnchus (fig. 31); and thus the form redly occupies an intermediate position between the subgroups C c arid B c, in which latter subgroup the Siluroid genus Malapterurus (fig. 12) has already been placed. With regard to Saccobrancl~us, it has already heen pointed out by Hyrtl(l0. p. 306) that the first branchial vessel unites with the second (which is another way of stating that they both open into the circulus), the third with the fourth, arid that the efferent vessels of the lung-sac open into the fourth branchial vessel.

FURTHERREMARKS ON GROUPD. In Syynathus (fig. 33) the dorsal aorta is not median but runs slightlg to the left side of the vertebral centra. The caeliaco- meeenteric artery arises at the posterior angle of the circulus cephalicus. From the same place arises a single vessel forking posteriorly into the two subclavian arteries. Hippocampus closely resembles Synynathus, but since the openings of the third and fourth efferent vessels are closer together, the genus comes under the second subgroup, D b. The circulus cephalicus, also, is less elongated than in Syngnathus. In both genera, however, the front of the circulus is broad and its posterior angle very acute. The dorsal aorta of Hippocampus is median, and the subclavian arteries arise from its sides directly, and not through the intervention of a common root. In ffastrosteusthe circulus cephalicus is iiot more than twice as long as broad. The subclavian arteries arise D short distance behind its posterior angle, and the caeliaco-mesenteric a considerable distance behind. Molvn differs from ffadus(fig. 34) mainly in the fact that the subclavian arteries arise froin the circulus cephalicus farther from the median line, and consequently more remote from the dorsal aorta. The cmliaco-mesenteric forks close to its origin in both genera, as it also does in Hotella. ilfotallri very closely resembles Molwci, and it is only the slight separation of the mesial extremities of the last two branchial vessels which causes the genera to be placed in separate subgroups. GVidus callarias (G. morrhun), figured by Miiller (16. pl. iii. fig. 13) and by Stannius (23. pl. v. fig. 2), does not appear to differ materially from Oadzcs ceglej’nus, 1899.1 BLOOD-VESSELS OF TELEOSTBAN FISHES. 953 except that the subclavian arteries arise farther from the median line, just as they do in Allolva. In Lota, as described and fi4iired by Hyrtl (7. pl. i. fig. d), there is a tendency for the last two efferent brmchial vessels to unite before opening into the circulus. The fact that the cceliaco-mesenteric artery arises from the circulus is shown in Hyrtl's figure, and is quoted by Stannius (22. p. 108, footnote 3) and by Owen (19. p. 490). In Blennius (fig. 35) the circulus cephalicus is narrow in front, the cceliaco-mesenteric artery arises from the aorta immediately behind the circulus, and the subclavian arteries are just behind this. In Ortl~agooriscw(fig. 36) the aorta is unsymmecrical, running to the right side of the vertebra ; the cceliaco-mesenteric artery is formed by the union of a pair of vessels arising from the common trunks formed on each side by the confluence of the third and fourth efferent braiichial vessels. The mesial ends of all the efferent vessels are closely approximated. The gills in Orthugoi*iscics are remarkably prolonged in a backward direction. The efferent branchial vessel of each emerges from near the middle of the full extent of the gill, aud not, as is more usual, from the upper end. It is formed by the union of one vessel coming from the lower part or gill proper, with another from the dorso-posterior prolongation, in a manner already made clear by Alessandrini (2) and Milne- Edwards (15. p. 335, footnote 2). Judging from Alessandrini's description, the peculiar mode of formation of the caeliaco-mesenteric artery in 0. truncatus does not occur in the species examined by him (0.nzola). In Tetrotlon (fig. 3i)there are only three gills on each side, and three efferent branchial vessels, the fourth of the normal series being absent. The reason for putting the genus in the subgroup D e, characterized by the union of the third and fourth efferent vessels into a coninion trunk, is to be found in the relation of the cccliaco-mesenteric artery to the third efferent vessel of the right side. The association is such that, were the fonrth vessel present, it vould not reach the circulus betweeii the third vessel and the cceliaco-mesenteric. Both the right and left subclavian arteriee arise from the right side of the circulus, and the latter is nearly circular iu shape.

C'ONCLUSIONS. It will be setw from the foregoing observations that very considerable diversity in the arrangement of the efferent branchial blood-vessels is to be met with in the Telro.;tean fishes. The typo of vascular arrangement is constant for different species of the same genus, and does not vary to any considerable extent in different genera of the same family. If, as in the Siluroid fishes, some widely divergent types ire included in the same family, there are to be fouucl intermediate types which act as connecting-links between these extremes. With regard, however, to families which, in the at present accepted taxonomy of the group, are brought into close relationship, 954 DR. W. 0. RIDEWOOD ON ‘PEE [Nov. 28, the characters of the efferent bronchial system fail to afford any convincing evidence as to the correctness or the relerse of such association. Thus, while the Gadoids have one sharply marked type of arterial arrangement, the Clupeoids another, and the Salmonoids a third, in the disposition of the efferent vessels the Lophobranchii resemble the first, the Ophidiida the second, and the Scombrida the last,-a most unnatural coupling of the families. It does not appear, therefore, that the facts revealed by the present inquiry can he applied with any probability of success to the interpretation of the afinities of the family and larger groups of Teleostei.

LIWOF REFERENCES. 1. AUASSIZ,L., et Vow, C.-“Anat. des Salmones,” MQm.de la Yoc. des Sci. Nat. de Neuchbtel, iii. 1845 (1846), pp. 198, fourteen plates. 2. ALESSANDRINI,A.-“ De pisciuni apparatu respir., turn speci- atim Orthagorisco,” Nov. Comment. Acad. Sci. Bononiensis, iii. 1839, pp. 259-381, four plates. 3. ALLIS, E. P.-‘‘The Crauial Muscles atid Cranial and First Spinal Nerves in Amin calvu,” Journal of Morphology, xii. Boston, 1897, pp. 487-808, nineteen plates. 4. AYERS,H.--“ The Morphology of the Carotids, based on a Study of the Blood-vessels of C‘hlun~ytloseliichusayu~nezm,” Bull. Mus. Comp. Zool. Harvard Coll., xvii., Cambridge, Mtiis., 1889, pp. 191-223, one plate. 5. CUVIER et VaLENCIENNES.-IiiStOirt: hTatUrel1f3deS POiSSOllS. Paris, 1828. 6. DuvERNEY.-oeuvres Anatomiques, ii. Paris, 1761. 7. HYRTL,J.--“ Beob. aus dem Gebiethe dar kergl. Gefasslelire,” filed. Jahrb. der Oesterr. Staates, xxiv. (n. P. xv.), Wien, 1838, pp. 69-91 and pp. 232-248, tour plates. 8. HYRTL,J.-“ Uber das Arterien-System des hpidosteu8,f’ 8itzungsber. dkad. Wiss. Wien, viii., 1852, pp. 234 241. 9. IIYR‘FL, J.--“ Beitrag zur Auat. von fleterotis e?Lrelzberyii,” Denkschr. Akad. Wiss. Wien, viii., 1854, pp. 73-88, three plates. 10. HTRTL,J.--“ Zur Anat. von Saccobranchus sinyio,” Sitzungs- ber. Akad. Wiss. Wien, xi., 1854, pp. 302-307, one plate. 11. HYUTL,J.--“ Anat. Mitth. iiber Morniyrus uncl Gymnarchus,” Denkschr. Akad. Wiss. Wien, xii., 1856, pp. 1-22, six plates. 12. HYRTL,J.-Ober besoridere Eigenthiimlichkeiten der Kiemm und des Skeletes von Lutodeira chi(9~08,‘’Denkschr. Akad. Wiss. Wien, xxi., 1863, pp. 1-10, one plate. 13. MAURER,F.-“ Ein Beitrag zur Kenntnis der Pseudobranchien der Knochenfische,” Morph. Jahrb., ix., Leipzig, 1884, pp. 229-252, two plates. 14. MECEEL,J. F.-System der vergl. Anat,, v. H~nlle,1831. 1899.1 BLOOD-VESSELS OF 'TELEOSTEAN FI8HHS. 955 15. MILNE-EDWARDS,H.-Lepons sur la Phys. et 1'Anat. comp., iii. Paris, 1858. 16. MOLLBIZ,Jox.--" Vergl. Anat. der Myxinoiden, Dritte Port- setaung, "Abhandl. Alrad. Wiss. Berlin, 1839 (1841). 17. MULLEB,Jorr.-ober den Bau und die Qrenzsn der Oanoiden, und iiber das natiirliche S.ystem der Fische. Berlin, 1846. 18. OWEN,R. -Lectures on Comp Anat., i., Fishes. London, 1846. 19. OWEN,R.--dnntorny of Vertebrates, i. London, 1866. 20. PASKER,T. J.-Zootomp. London, 1884. 21. RoLLnsroN, G.-Forms of Animal Life: 2nd Edn. by W. Hlttchett Jackson. Oxford, 1888. 22. STANNIIJS,H.-Lehrb. der vergl. Anat., ii., Wirbelth. Berlin, 1846. 23. STANNIIJS,H.-Das periph. Nervensystern der Pische. Rostock, 1849. 24. STANNIUS,K.-Handbuch der Anat. der Wirbelthiere, i., Fische. Berlic, 1854. 25. WBIGHT,1E. R.-" On the Hgoniandibular Clefts and Pseudo- braachs of Lcpidosteus and Ainia,.' Journ. Anat. nnd Phys. xix., London, 1885, pp. 476-499, one plate.

EXPLANATION OF THE PLATES. Efferent Branchial Blood-vessels of the first, second, third, and fourth Gills of Teleostean Fishes, as Been after st,ripping off t.he mucous membraiie from the root' of' the pharynx. The dorsal aorta, the caeliaco-mesenteric art,ery, and the dxlaviaa artery are respectively indicated in all the figures by the letters, a, c, and 8. PLATELXIII. Fig. 1. Engruslis e~~crm'cholus,p. 947. 2. Chipea hareups, p, 947. 3. Ammodytcs laiueolattw, p. 947. 4. Meploy cypinoides, p. 947. 5. Ba isteJ acubentus, p. 948. 6. Scomber scombrws, 7. salrrio snlar, p. 948 948* 8. .Wig3 capuito, p. 948. 9. Hippoglowus vulgaris, 948. 10. Ommw qerhnus, p. Jig. 11. AlCulu conorhynchus, p. 949. 12. Mabpterumw electricus, p. 949. 13. Cgprinu carpw, p. 949.

PLATELXIV. Fig. 14. Cohitis tcenia, p. 949. 15. Esox lucitw, p. 949. 16. AiLguilln vulgaris, p. 949. 17. Labrax lupzis, p. 950. 18. Liocusis longirosfris, p. 930. 19. Callichthys littoralis, p. 950. W. Chius nzuyur, p. 95U. 956 ME. ff. A. BOULENffnB ON REPTILES, [Nov.28, Fig. 21. PercafEuviatilis, p. 950. 22. Zeus faber. p. 950. 23. Anarrhichas lupua, p. 950. 24. Labrzls maculatus, p. 951. 25. &occptus brachysoma, p. 951.

PLATELXV. Fig. 26. Laphiiis piscatorius, 951. 27. Coltus Scorpio, p. 95r: 28. Tr@a ci~c14dzu.p. 951. 29. Cyclopterwr lumpur, p. 951. 30. FGtularin tdaccaria, p. 952. 31. Naccobramhusfossilis, p. 952. 32. Sdurus qlanis, p. 952. 33. rFyn,qn&hus acus, p. 952. 34. Gadus cegvems, p. 952. 35. Blmnius pholis, p. 953. 36. Ortltaqoriscus tncncatus, p. 953. 37. Tetrodon palembangewsia, p. 953.

4. On the Rephles, Batrachians, and Fishes collected by the late Mr. John Whitehead in the Interior of Hainan. By G. A. BWLEN~ER,F.R.S. . [Received October 31, 1899.1 (Plates LXV1.-LXIX.) During his short stay in Hainan, where he died on June 2 of the present year, Mr. John Whitehead had succeeded in collecting a small number of cold-blooded vertebrates in the Five-fiuger Monntains, in the interior of the island. The fact that so many of the few species represented in the collection are new, tends to show how rich a hameet these unexplored mountains would have yielded but for the fatal climate which has deprived the zoological' world of one of its most enthusiastic and successful members.

REPTILES.

1. DRACOWHITEHEADI, sp. n. (Plate LXVI. fig. 1.) Head small ; Pnout considerably longer than the diameter of the orbit ; nostril lateral, directed outwards ; tympanum scaly. Upper head-scales unequal, strongly keeled j 8 or 9 upper labials. Male's giilar appendage very large, once and a lialf as long as the head. A rudimentary nuchal crest. Dorsal scales B little larger than ventraIs, irregular, obtusely keeled ; on each side of the back a series of enlarged, keeled dorsal scales. The fore limb stretched forward reaches the tip of the snout, the hind limb between the elbow, md the axilla. Reddish brown abova, with dark transverse P Z 5.1893 P1.LXVI.

I DRACO WHITEmADI 2 ACANTHOSAURA HAINANENSIS P I, S 1899 Pl.LxvlI

FJ Snut. d.1 et hM mrLternnxo6 1 RANR GRAMINEA 2 STAUROIS HAINANLNSIS 3 RHAC OPHORUS OXYCEF’HPLLIJS P 2 S 1899 P1 LXV7

J Green del et bth C OFLEOPERCA IVHITEHEADI

1809.1 MR. L. A. BOItRdDhrLE OX THE PAGURIXE LAND-CRABS. 937 phructus, a P\ithon, 3 Qclunorbis senegalensis, 2 Hinged Tortoises, some Chameleons, aud a, Serval Cat were brought home alive. Since iny return to England, I have definitely decided that the eggs and larvae obtained are not those of Polypterus. I have, however, I believe, learned enough about the habits of Polypterm to encourage me to make a second attempt next year to obtain the developmental stages. In conclusion, I wish to thank the Society for lending me influence and support, without which the little that has been done by this expedition could not have been accomplished.

BATRACHIAN S.

1. RANAGSAMINEA, sp. n. (Plate LXVII. fig. 1.) Vomerine teeth in two short oblique series between the choaiise, nearer to each other than to the latter. Head depressed, as long a'i broad ; snout rounded, scarcely projectiug, as long as the diameter of the orbit ; canthus rostralis well-marked ; loreal region concave ; nostril nearer the end of the snout than the eye : interorbital space as broad as the upper eyelid ; tympanum very distinct, three fourt'hs the diameter of the eye. Fingers and toes rather slender, with small but well-developed disks: first finger not extending beyond second ; toes nearly entirely webbed ; a single, feebly prominent, oval, inner metatarsal tubercle. The tibio-tarsal articulation reaches bepond the tip of the snout: tibia as long as the distance from end of snout to sacrum. Skin smooth ; it moderately broad, feebly prominent glandular lateral fold ; another, fold from below the eye to the shoulder, followed by a strong glandule. Bright green above, brownish on t,he sides of the head and body, below t.he canthus rostralis and the dorso-lateral fold, and on the limbs ; upper lip white ; limbs with regular dark cross- bars ; hinder side of thighs marbled dark brown and yellow : lower parts white. Male with two external vocal sacs, in front of the arms ; no humeral gland. From snout to vent 48 millim. Allied to R. erythrcm Schleg. Distinguished by the shorter snout, the longer hind limbs, the external vocd s&cg, and the coloration. Also allied to R.jerhoa Gthr. and R.whiteh.endi Blgr., in which the digital disks are larger and the hind limb Ionger still. Two male specimens.

2. RANAANDERSOXI Blgr. This species, first discovered in the Hotha Valley, Punnaii, by Dr. J. Anderson, has since been found in the Kalihyen Hills, Upper Burma, by Signor L. Fea, and at Kuatun, N.W. Fokien, by Mr. J. D. La Touche.

3. STAUROISRAINANENBIS, sp. n. (Plate LXVII. fig. 2.) Head as long as broad or slightly broader than long ; snout short, truncate, projecting : cauthus rostralis strong ; loreal region nearly vertical, concave ; nostril midway between the eye ~iid the end of the snout; interorbital space as broad as the upper eyelid ; tympauum distinct., one third or two fifths the diameter' of the eye. Fingers slender, first, longer than second, with very large disk8 ; toes webbed to the disks, which are a littde smaller thaq those of the fingers ; subnrt.fc.ular tubercles feebly prominent ; a very indistinct iiiner nietntarsal tubercle. The t,ibio- tarsal articulation reaches the t'ip of the snout; or a little beyond. Skin smooth above in the adult, warty in the young ; lo\\ er part 1899.1 UATRACHIANS, AND FISHES FROM HAINAN. 959 smooth. Olive above, spotted with black, or blackish with pale olive markings ; limbs with dark cross-bars ; hinder side of thighs with a black reticulation. From snout to lent 58 millim. Larva aith a large pectoral adhesive disk (see P. 8. S. 1893, p. 526). Beak formed of two pieces, an upper and a lower, feebly denticulate, not ribbed ; lower lip not fringed ; the horny teeth form 3 uninterrupted and 2 paired series on the upper lip, 2 uninterrupted and 1 narrowly interrupted series on the lower lip, 3 03 an arrangement that may be expressed by the formulaL el. 13 Closely allied to Xtcturois nrifntor Gthr. Distinguished by the shorter head. The tadpole, on the other haud, stands nearest to that of Rnnci lntopnlmcitu Blgr. Three specimens : a female, a young, and an advanced tadpole.

4. RHACOPHORUSLEUCOXYRTAX Gravh.

5. RHACOPIIORUSOXYCEPIIALUS, sp. 11. (Plate LXVII. fig. 3.) Vomerine teeth in two obliqiie series betReen the choanse, the inner front edge of which they nearly touch. Head as long 8s broad ; snout pointed, as long as or a little longer than the diameter of the orbit ; canthus rostralis distinct; loreal region concave; nostril a little nearer the tip of the snout than the eye; interorbital space a little narrower than the upper eyelid ; tympanum distinct, half the diameter of the eye. Fingers with a distinct rudiment; of web ; toes entirely webbed ; disks of fingers nearly as large as the tympanum, of toes a little smaller ; a very small inner metatarsal tubercle. The tibio-tarsal articulation reaches beyond the tip of the snout. Skin smooth or with small warts ahove; belly granular. Greyish or brown above, spotted or marbled with darker; a dark transverse band or triangular marking, base forwards, between the eyes ; limbs with dark cross-bars ; groin and back of thighs marbled black and yellow ; lower parts white. Male with an internal vocal sac. From snout to vent 57 niillim. Four specimens.

6. BUFOYELANOSTICTUS Hchn.

FISHES. COREOPEWX. Coreoperca, Herzenstein, Ann. Mus. Zool. St. Pdtersb. 1896, p. 11. Body compressed ; scales small, cycloid, concentrically striated, Laterd line complete ; tubes straight, occupying the greater length of the scale. Mouth large, protractile ; mpxillary exposed, with fjP 960 MR. 6. A. BOULENGER ON REPTILES, [Nov. 28, supplemental bone; villiform teeth in jaws and on vomer and palatines ; no canines ; tongue smooth : head partly naked : praeopercle serrated, ,with a few antrorse spines on the lower border ; opercle with two spines. Gill-membranes separate; seven brnnchioAtegals j pseudobranchiae present. Dorsal tins confluent, XIV-XV 11-14, the spinous portion much longer than the soft ; a,iial short, III 7-11. ; caudal rounded. Pectoral symmetrical, rounded, rays 16. Ventrals below the pectorals, close together, with a etrong spine and five branched rays, the last of which is connected with the belly by II membrane. The type species of this genus of Serranitkce, allied to Siniperau, is from the interior of North Corea'. It is highly interesting to add a secood species from the interior of Hainan.

1. COREOPFIRCAWHITEHEADI, sp. n. (Plate LXV.LIT.) Depth of body equal to length of head, 3 times in totd length. Snout 14 diameter of eye, which equals interorbital width, 4 length of head ; lower jaw projecting beyond the snout; maxillary extending a little beyond vertical of posterior border of eye, the width of its distal extremity a little less than diameter of eye ; praeorbital entire ; cheeks and opercles scaly, rest of head naked ; prsopercle finely serrated, without enlarged spines at the angle ; opercular spines strong, Dorsal XV 14, originating above base of I pectoral j spinous portmiontwice as long as the soft j spines st,rong. ehort, increasing in length to the sixth, which equals 2 length of head ; longest soft rays nearly 6 length of bead. Pectoral $ length of head. Anal 111 11 ; second spine longest, a little shorter than longest dorsal spines. Caudal rounded, subtruncate. Sq. 80 i; ; 1.1. 63. Brown, with dark marbling8 and whitish dots; a dark streak from below t,he eye to the angle of the prsopercle and another from the eye to a large, black, whit,e-edged ocellar spot Haring had the privilege of examining the type specimens of Cowoperca herti, Herzenst. 1. c., preserved in the St. Petersburg Museum, I add a description of them for comparison with C. whiteheudi :- Greatest depth at origin of dorsal fin, 3 to 3; timed in total length, length of head 23 to 3 times. Snout as long as diameter of eye, 4 length of head, and t,wice width of interorbital region ; lower jaw not projeoting ; maxillary extending to below posterior third of eye, the width of its distal extremity about half diameter of eye ; przorbilal ent.ire ; cheeks and opercles scaly, rest of head naked ; pwopercle with two stroug bifid spines at the angle and two or three antrorse spines on the lower border ; opercular spinesstrong. Dorsal XIB 11- 12; originating above base of pectoral, epinous portion twice as lovg t19 the soft ; spines stron increasing in length to the fifth, which equals length of head, but, is consi&rably shorter than 'the longest soft rap. Pectoral p length of head. Anal I11 7, spines very strong, third a little longer than first., second longest and a little longer than longest dorsal spine. Caudal rounded. Sq. 76-82 :$ ; 1. 1. 51-56. Brown ; a dark streak from below the eye to the angle of tbe rzopercle. a black spot, edger1 with white anteriorly, between the opercuyar spines f body with some &irk brown spots intermixed with whitish dots ; a regular series of dark sp0t.s along the base of the dorml. Total length 85 niilliin. Pung Tung, Corea. 1899.1 BATRACHIANS, AND FISHES FROM HAINAN. 961

between the opercular spines j some light dots on the soft dorsal and anal and on the membrane between the ventral rays. Total length 155 millim. A single specimen.

2. I)ISCOC+~-ATHTJSIMBEBBIS Vincig. A species described from the Karen Hills, Burma.

3. GYMNOSTOMUSLEETUBUS, sp. n. (Plate LSIX. fig. 1.) Depth of body 4 times in total length, length of head 5 times. Head 1$ as long as broad ; snout broad, rourded ; width of mouth nearly half length of head ; lower jaw with a sharp, horny edge ; diameter of eye equal to length of snout, 3$ times in length of head, 18 in interorbital width. 1)orsal 111 8, midway between end of snout and base of caudal ; first branched ray longest, a little shorter than the head, last ray longer than those preceding it, 8 length of head. Pectoral as long as head. Ventrals below middle of dorsal. Anal I11 6, as deep as dorsal. Caudal deeply bifurcate, 14 length of head. Caudal peduncle thrice as long as deep. Sq. 49 ; 4 scales betneen the lateral line and the ventral. Olive above, silvery beneath ; an ill-defined dark lateral streak. Total length 165 millim. A single spacimen.

4. BAEILIUSIIAINANENSIS, SP. n. (Plate LXIX. fig- 2.) Depth of body equal to length of head, 43 times in total length. Head twice as long us broad ; snout pointed, not projecting beyond the mouth, as long as diameter of eye, which is 33 times in length of head and equals interorbital width; mouth extending hardly to below anterior border of eye ; suborbitah entirely covering the cheek. Dorsal 11 7, originating just behind yentral and situated at equal distaiice from the eye and the root of the caudal; first branched ray $ length of head. Pectoral a little shorter than head, not reaching ventral. Anal 11 14. Caudal deeply bifurcate, as long as head. Caudal peduncle nearly thrice as l011g ns deep. Sq. 46 3". Silvery, darker on the back; scales abore the laFeral line black at the base. Total length 130 millim. A single speL'']men.

5. OPSABIICHTHYSPLATYPUS Schleg. A species known from Japan and Formosa.

EXPLANATION OF THE PLATES.

PLDE LXVI. Fig. 1. Draco whiteheadi, p. 966, with side-view of head. 2. Acalzthosaura binanensis, p. 967. 968 DR. A. G.%UTLER ox BUTTERFLIES [Nov. 28, PLATELXVII. Fig. 1. Raw graminea, p. 958. la. ,, ,, Side view of head. 2. Stauroia hainanensis, p. 958. 2a. ,, ,, larva, lower view of body. 3. Rhicophorus oxgcephalw, p. 959. PLATELXVIIT. Coreoperca whiteheadi, p. 960. PLATELXIX. Fig. 1. Gynmwstomwr leptzlrus, p. 961, t nat. size. 2. Burilius hainanensis, p. 961.

5. On a Collection of Butterflies made by Mr. Richard Crawshay in British East Africa. By ARTHUR0. BUTLER,Ph.D., F.L.S., F.Z.S., &c., Senior Assistant- Keeper, Zoological Department, British Museum. [Received September 20, 189O.J (Plate LXX.) During the past summer I received from Mr. Crawshay a box of Lepidoptera and a letter dated February 8th, 1899, addressed from Xeugia, as follows :- ‘‘ A few lines to let you know that I have lately returned from a journey into Maraiiga, the S. and S.W. slopes of Mt. Kenya; and that I was able to take some Butterflies, which, I think, will please you. “From this-I mean the mention of mighty Kenya and its 18,600 feet-you must not infer that these insects have been collected at any great altitude. Maraiiga is not so high as other parts of Kikuyu to the westward,-for instance the neighbourhood of Fort Smith, which is 6400 feet. As a matter of fact it does not average, I suppose, more than 5600 feet ; rising to the N. and N.E. gradually into the mighty belt of forest surrounding the mountain for many thousands of feet, and falling away to the W. and S.W. OF the Tana River, which, where we crossed it, is 3850-3900 feet. ‘‘ In all the thousands of miles I have travelled in Africa, I hye never seen a more lovely and more possible country than Marafiga ; nor more splendid specimens of its peoples than are the Wakikuyu- though they are at present suspicious of and hostile to everyone from the outer world. The Wakamba are the most veritable worms in comparison with them. However, you want to know something more of the surprise8 which I hope are in store for you in the shape of the Butterflies. “The most promising of these are Skippers and Blues-one a very large and powerful Blue with almost black wings on the inside, which show a Purple-Emperor-like glow, though with a P z s 1899. P1 LXX

FW Frohawk del et Lth Wcat,Nemchrorno BUTTERFLIES FROM BRIT. E. AFRICA.

1899.1 FaQM BRITISH EAST AFRICA: 963

tinge of port-wine colour about it I. This Blue (of which I think I took a pa.ir) and another very active and cunning insect-as you will see from the note appended on bhe paper envelope-were taken on a stinking crocodile’s head, than which nothing smells more foully ; otherwise-I mean without this-I should probably never have seen the former, nor taken either of them. “ A hippopotaniue skull hr gone in putrefaction also proved attractive, and did me several good turns in enabling me to get on equal terms with other insects: on this I took bhe only Charaxes I saw, I think, on the whole journey, though one I know well. Of Bloi.hs I also took a nice lot.” On his return Mr. Crawshay had a narrow escape from a wounded bull Rhinoceros, but by a fortunate shot whilst lying in the grass he managed to hit it in the heart, and so escaped with his life. The collection of Moths will be worked out by Sir George Hampson. The Butterflies are represented by 127 specimens referable to 69 species, of which 3 are new to science ; the most interesting to me is n new species of Chloroselns allied to C. tamaniba of Walker from Suakin, but apparently distinct, and clearly proving that Prof. Aurivillius was in error in referring the Somali insect C. esmercildn to that species. C. tamaniba is considerably larger than C. esmerdda, has the posterior half of the primaries, including the base, blue; and the anal orange spot of the secondaries is \vell-defined as in the present species.

NYM PE A L ID >N. 1. MONOTRICHTISSAEITZA Hewibs. Muthalubi River, Ndya, 7th January, 1899. ‘6 They swarm 011 the banks of the streitlll ” (3.C.). 2. NEOC~NYRADUPLEX Butl. Slopes of Nthatha Hill, Kitwi, 4700 feet, 31st December, 1898.

3. NBOC(E?ITRAGREGORII Butl. Muthambi River, Ndya, 7th Janwiry, 1899. One much worn example.

4. CHARAXESVARANES Cram. 8, Mutharnbi River, Ndya, 13th January, 1899. ‘6 Takeu on the crocodile’s head (hippopot.amus’ skull?) ; the only one of this species which I have seen during my week here ” (R.C.). 5. PRECISCLOANTHA Cram. 8, Plains N. of the Tana River, Kikuyu, 4500 ft?et, 5th January ; 9,Muthambi River, Ndya, 12th Jaiiuary, 1899.

1 This proves to be a well-known apecies of CIrmiu, which occurs SO in South Africa.-A. 0. B. 964 DB. A. G. BUTLER ON BUTTBRFLIES [Nov. 28, The female, though in excellent condition, was caught by Bvalamkambi, one of Mr. Crawshay’s native servants.

6. PRECISELGIVA Hewits. d , Q , Muthambi River, Ndya, 7th January, 1899.

7. PRECISNATALICA Feld. 9 , Muthambi River, Ndpa, 13th January, 18‘39. ‘‘ Fairly plentiful : but this is, I think, the only perfect specimen I have seen” (R.C.). 8. HYPOLIMNASMISIPPUS, var. INARIA Cram. Q , Muthambi River, 4500 feet, Rdya, 11th January, 1899.

9. CEENIS BOISDUVALII Wllgl.. 6 d, Muthambi River, 4500 feet, Ndya, 12th January, 1899. 6L An insect which I should probably never have seen, and certainly not taken, had it not been for the putrefyiug crocodile’s head, on which it descended as it were from the clouds. It is worthy of note that this insect shows a lovely tinge of purple or blue on the inside of the wings if these ore viewed at an angl? in a good light.” (22. C.) 10. CRPINISHOWENSIS Staud. d d, Muthambi River, 12th January, 1899. “ The most restless, active, and difficult insect to take I have ever come across ; for six days have 1 been watching and following the movements of some half-dozen in the glade of fig-trees in which I am encamped, trying all I could to take one, but without success. They flip about amongst the trees and occasionally perch on the trunk and branches, but always out of reach. By great good luck this morning two have been tempteJ to come down and feed on a putrefying crocodile’s head, and I was thus enabled to take them.” (R.C.) One specimen (a male) is said to have contained a whitish-green ovum, but this must have been an error of observation, as the insect has strongly pronounced male claspers.

11. HAMANUMIDADZDALUS Fabr.

“ Plentiful enough on undulating plains N. of Tana River, 4200 feet, 14th January, 1899.” A single dry-season male of this abundant species.

12. NEPTISAQATHA Cram. Q , d, Muthambi River, 4500 feet, Ndya, Kikuyu, 6th & 7th January, 1899.

13. NPIPTIDOPSIBOPHIONB, var. VELLEDA Mab. Muthambi River, 4500 feet, Ndp, 6th & 13th January, 1899. Mr. Ckawshay appears to have seen three examples of this 1899.1 FROM BBlTIClH BAST dFBIUA. 965 species on the 6th January, of which he caught two then, and the third a week later.

14. BPBLIAILIYTHIA Drury. Wet phase.-Machakos to Neugia, 16th December, 1898. Intermediate and dry phases.-Neugia, 31st January, 1899. 6' Quite the commonest butterfly hereabouts these days ; indeed I have never anywhere else seen this insect in such numbers" (R.C.). 15. ACRBACABIRA Hopff. Clue to exact locality and date lost. An example leading to the variety A. apecida.

16. ACBBALYCIA Fabr. (vars. sganzini and daira). Muthambi River, Ndye, 7th & 11th January, 1899.

17. ACRBAONERATA Trim. d 9 in copula, Neugia, Kitwi, 11th February, 1899. The sexes are remarkably alike, the wing-borders of the female slightly heavier and the body spotted.

18. ACRBAACRITA, var. PUDORINA Staud. d 8, Neugia, Kitwi, 16th February, 1899.

L Y c BNID B.

19. LACHNOCNEMABIBULUS Fabr. Ndya, Kikuyu, 4500 feet, 6th January, 1899.

20. AXIOCERSESHARPAX Fabr. 9, Kitwi, 18th January, 1899. '' Grass-green ova '' (R. C.).

21. TERIOMIMAPALLIDA Trim. E. of Athi River, about 4300 feet, 18th December, 1898. Mr. Crawshay regarded this as a worn example of T.hildegarda, to which it certainly is very closely related, but the small and inconspicuous spots on the under surface of the secondaries give it a somewhat different aspect ; it would not surprise me to find that it was only another variety of that variable species. Mr. Crawshay says he took it for a moth until it was in the killing- bottle. 22. CATOCHRYSOPSPECULIARIB, var., Rogenh. (Plate LXX. fig. 1.) 9 9, Neugia, Kitwi, 24th 6 29th December, 1898. Of the first example Mr. Crawshay writes--"A hardly won capture. Caught in my Terai hat when out shooting, and kept under this on the ground, while I covered my head from the 966 DR. A. 0. BUTLER ON BUTTERFLIES [Nov. 28, blazing sun until my net reached me, fully three quarters of an hour.” The two specimens are of much interest, as they clear up one of the greatest muddles which has been made over any species of LycmaidR :-In 1891 Rogenhofer described a butterfly which was quite unknown to English entomologists under the name of 66 Ch-ysophmus peculiavis ” ; naturally nobody expected a Cuto- chrysgs to be called a Chvysophams. In 18‘32 Ur. Holland described the same species in the ‘ Entomologist, uiider the name of Lyccenaperpul~hvu;in 1893 I described a inale from Wasiii atid a female from the Victoria hTyanza as Cnstrtlius 7~ypoleucus; aud in 1894 Mr. Trimeu described (7. peculiaris agaiu under the name of Lycma exclusa. Ur. Hollaud subseqilently pointed out that C. hypoleucus and L. exclusa were synonynious with his L. perpulchra. In 1898 Mr. Trirnrn received some large evamplen from Mashunaland which he rightly stated to be identical with our Nyanza female ; but instead ot adopting my name for this form, he called it Lyccenn giyuntea, stating that I had confounded the female with that sex of L. peryulchrtr (entirely overlookiiig the faat that the Byanza female \i as described by me as the type of that sex of Castalizis hypoleucirs). At the commencement of the present year, when working out the species of Chrysophnnus, I recognized C.peedinria Rogenh. as the oldest nxme for the present species, aud entered it in my paper on Mr. Crawshay’s last collection. Professor Aurivillius also recognized Hogenhofer’s species in his ‘ Rhopalocera Bthiopica,’ where, however, he retained Mr. Trimen’s name for the larger form, ignoring the fact that iny female unquestionably takes priority. The two examples now sent home by Mr. Cranshay are quite intermediate in character between the large and small forms, the colouring of the upper surface agreeing most iiearly with C. peculiaris $? , the expanse of wing being nevertheless equal to that of my Victoria Nyanza female. Like the latter and a worn and faded female from Zoinba (uhich Trimen refers to as the Pu’yasa female), they have from oiie to two extrn black spots in the discal series on the under surface of tbe primaries; the black discocellular spot is smaller than in the little Xomba female, but varies in size in the two examples. It is absolutely impossible to sny that these examples belong to one form dher than to the other, and I do not doubt that they repreeent an intermediate phaBe between C. lL~j~oleucus=gigcinten thp wet phtise, and C‘. peculiaris the dry phase. of one and the same hpeciee.

23. AZANUSNATALENSIS Trim. d,Tana River, 3800 feet, 16th January, 1899. ‘‘ The only specimen seen” (R.C.).

24. AZANUSZENA Moore (Lycmau mciculen.ga, Trim.). , Kibwi, 4000 feet, 30th December, 1898. 1899.3 FROM BRITISH EABT AFIIICA. 967

2.5. EVERESKEDONGA Orose-Smith. (Plate LXX. fig. 4.) 8 8,Plains N. of the Tam River, Kikuyu, 5th January, 1899. Mr. Crawshay sent us the female of this extremely pretty species in his last collection (see P. Z. S. 1899, pl. xiv. figs. 3, 3 a).

26. TARUCUSPLINTUS Fabr. 2, Athi Valley, 16th December, 1898 j 8, Ndya, 4500 feet, Kikuyu, 16th January, 1899. Of the female Mr. Crawsbay observes-“Common, and all evidently newly emerged.”

27. ZIZEEAKNYSNA Trim. 9, Slopes of Nthatha Hill, 4700 feet, Kitwi, 31st December, 1898; d, 9, Muthambi River, 4500 feet, Ndya, 8th & 10th January, 1899.

28. ZIZERAGAIKA Trim. 8 8,Tam River, 3800 feet, 2nd January ; Neugia, Kitwi, 7th Fzbruary, 1899. Of the second example Mr. Crawshay observes-“ The smallest butterfly I have ever seen.” It is a starved specimen.

29. LYCXNEBTHESAMARAH LefAb. 9, Slopes of Nthatha Hill, Kitwi, 4700 feet, 31st December, 1898 ; 8, Plains N. of the Tana River, 5th Jaunary, Tam Hiver, 3800 feet, 16th January, 1899.

30. CACYREUSLINGEUS Cram. 8 8,Clue to exact loc~lityaud date lost. 31. PHLYARIAVIRGO Butl. 9,Muthambi River, 4500 feet, Kdya, 10 January, 1899. “The only specimen I have seen iu these parts, and like, if not identical with, another which I used to take in the mountaius of Nyilra” (R.C.). According to Prof. Aurivillius this is the female of the West Coast P. heritsin, but some of the details of marking make me hesitate to accept this dictum until I have seen East-African males. In my opinion the border of the secondaries is too narrow and the markings on the under surface of these wings too feeble for 2’. heritsic1 : I may prove to be wrong, but 1 very strongly object to putting species together by guess.

32. CHLOEOSELASAZUEEA, sp. n. (Plate LXX. figi. 2, 3.) Intermediate between C. tcimnniba from Suakin aud C. esrneralcltc from Somali-land; evidently nearest to the former. It is considerably larger than C. esmeraldn, the shot-colouring on the upperside of the wings being glistening deep sky-blue instead of emerald shaded with blue: in the primaries this colouring is restricted to the outer half of the internal area and the iuternal margin nearly 968 DR. A. Q. BUTLER ON BUTTEBFLIES [Nov. 28, to base ; the remainder of these wings is smoky brown, slightly cupreou~,with a large ill-defined but distiuctly darker patch over the end of the cell : secondaries cupreous brown, brilliantly glossed with sky-blue between the second subcostal branch and submedian vein ; a conspicuous orange anal spot between two short tails : anal lobe small, silver-spotted and fringed with black ; body blackish ; frons silverp whitish, collar with pale edges ; antennzP annulated with white, the club externally edged with tawny: under aurface pale fleshy buff, with the discoidal area of the primaries and the anal patch of secondaries orange-ochreous; the internal area of the primaries blackish grey tou ards the base, white slightly opalescent be.yond, the marking on the wings much as in C. esmeralda, silver with black margins; body below white. Expanse of wings 26 millim. The female is rich copper-brown with white fringes, and an orange spot on the secondaries between the tails, as iu the male ; the under surface similar to that of the male : size uniform. d, Slopes of RTthathrt Hill, Kitwi, 4700 feet, 31st December, 1898; 9, Plains N. of the Tana River, 4500 feet, Kikuyu, 5th January, 1899. C. iamaniba, according to Walker, expands one inch and one line (or 28 millim.), and is therefore the largest species in the genus ; the primaries are described as being shot with blue at the base and on the hind half, and the secondaries as blue, with a narrow brown border ; ‘I an orange spot adjoining the tail, into which it extends, bordered on the outer side with glittering chaly- beous.” These characters do not at all correspond with those of C. esmerakda (which Prof. Aurivillius has unaccountably placed as a synonym of it), and differ considerably from those of the present species. Many years ago (1870) I saw the type, but I cannot pretend to remember exactly what it was like. I am, however, quite certain that Walker’s measurement is rather under than overstated : the specimen is, of course, incorrectly described as a female. 0. esmeralda not only differs from C. pseudozeritis in having only one tail to the secondaries, but in the absence of the brown clouding 011 the under surface of these wings. It is mere guesswork to suppose that a small insect like this from Natal is at all likely to be identical with one from Somali-land, or even that the latter is likely to be the same as one from the western shore of the Xed Sea. It is quite possible that there are many species of Chloroselas scattered over Africa, and that, ten years hence, the species hitherto confounded will be generally regarded as amongst the best marked of all. Structurally, the species may at present be separated by the tails of the secondaries as follows :- Species with two tails. Small-C. pseudozeritis Trimen. Large-C. azurea Butler. Species with one tail. Small-C. esmeralda Butler. Large- C. tamaniba Walker. 1899.1 FROM BRITISH EAST AFRICA. 969 But in many respects the two sinall species and thp two large species are more nearly related to each other respectively in other characters-the bluer shot and the internal position of the blue colonring, as well ae the more conspicuous anal orange spot of the secondaries, being characteristic of the larger species, the greener shot and the basal position of the greenish colouring, as well as the less conspicuous anal orange-tinted spot of the secondaries, of the smaller species. 33. MYRINA EICEDULA Trim. 8,Neugia, Kitwi, 2Snd December, 1898. Taken sitting on an outstanding branch of a small tree about noon in the hottest possible sun ” (R.C.). 34. VIRACHOLAANTALUS Trim. 8,Neugia, Kitmi, 22nd January, 1899.

35. VIRACHOLADEXONA ? Smith. E. of Athi River, some 4300 feet, Kitwi, 18th December, 1898. “ I have no recollectio!i of having taken this insect before. It was moat accommodating too in awaiting about a quarter of mi hour. I was stalkiug a Hartebeeste and my net was behind.” (R.C.) This appears to correspond with Mr. Grose-Smith’s descriptiou, but it is a soniewhat faded example.

PAPILI o KID i~, 36. MYLOTHRISAGATHINA Cram. 8 8,Muthambi River, 4500 feet, Xdya, 7th & 10th January, 189:). 37. NYCHITONAMIDUSA, var. ALCESTA Cram. 8,Tana River, 3850 feet, 4th January, 1899. 38. COLIASELECTRA, var. EDUSA Fabr. Ndya, Sth January, 1899. “The first specimen of c. edU8n I have seen since leaving Mrsssai ”, (R.C.). 39. TERIASSENEGALENSIS, var. BISINUATA Butl. 8,Tana River, 4th January, 2899.

40. TERACOLUSCALAIS Cram. 8,Kitwi, 19th January, 1899. “ An average specimen of this species as regards size” (R.C.). 41. TEEACOLUSERIS l

42. TERACOLUSINCRETUS Butl. Intermediate phcise.- 2 , Athi Valley, 4000 feet, 16th December ; 8 2 , Athi escarpment, eastern side, Kitwi, 18th December ; 9 , Msokani, 20th December, 1898 ; d , Tana River, 3800 feet, 16th January, 1899. Dry phase.- 8, 2 9, Tana River, 3809 feet, 16th January, 1899. Mr. Crawshay says of the male-‘‘ Very plentiful just here (Athi Valley) : this and the following taken in numbers when playing together, with one stroke of the net.” Of the female he says- &‘ An insect I have never before taken, and which, until settled, I imagined to be another, rather common just here-plain sulphur with an orange tip.” In thip conjecture Mr. Cruwbhay was quite correct. 43. TERACOLUSXANTHUS, var. METAGONE Holl. 8, Tana River, 3800 feet, 16th January ; 2, Neiigia, Kitwi, 30th January, 1899. 8. “ Fairly conlmon, frequents the more open country, dry and desert-like, and covered with thorny scrub ” (R.C.). This is distinctly a dry-season phase, and differs from IT. xunthus var. comptus in the entire absence of the internal grey streak 011 the primaries.

44. TERACOLUSANTEVIPPE Boisd. d , var. subvenosus, Kitwj, 18th January, 1899.

4.5. TERACOLUSGAVISA Wallgr. 9,Kitwi, 19th January, 1899.

46. TERACOLUSCALLIDIA Grose-hit h. d d , Intermediate and dry p?mst!s.--Tana River, 3800 feet, 16th January, 1859. ‘‘ Fairly plentiful ; but, if once missed, by no means an easy insect to take” (R.C.). The typical orange form of this species is by no means too well represented in the Museum series. 1899.1 FROM BRITISH EAST AFRICA.

47. TERACOLUSCATACHRYSOPS Butl. 8,Msokaui, Kitwi, 20th December, 1898. “Fairly common, but not an easy insect to take : a fast flier and a desperate donbler ” (R.C.). The single example sent belongs to the wet phase, which differs from the dry phase in the better defined, browner, and much darker bands on the under surface. This character, apart from other differences, amply serves to provs its entire distinctness from T. inutans of Southern Africa, whilst at the same time it; indicates some relationship to T‘. protomedia.

48. BELENOISSEVERINA, Cram. Tana River, 4th January, 1899. Rather a curious male, having a wet aspect on the npper surt‘acr, but the under surface characteristic of a late intermediate phnse, the black veins being feebly indicated.

49. BELENOISMESENTIJA Cram. 2, Nthatha Hill, Kitwi, 4700 feet, 3lst December, 1698 : d , Neugia, 30th January, 1899.

50. BELENOISWEYTWOODI Wllgr. Tana River, 4th January, 1899.

51. PINACOPTERYXASTARTE, sp. u. (Plate LXX. figs. 6, 7,) The wet phase of this species, of which we received a male from Fwambo in 1897, is not very unlike that sex of P. falkerlateinii, hut has the costal margin of the primaries shorter and the outer border more interrupted, represented only by spots at the extremities of the first and second median branches ; the costal margin of the secondaries below is deep orange. Expanse of wings 60 millim. The intermediate phase, of which (as well as of the male dry phase) we received sexes in 1889 from Tanganpika, differs in the reduction of the width of the outer border in the male and the paler under surface. The female is bright orange, redder at the base; the primaries with a marginal series of rather large greyish spots becoming black externally, and a similar smaller spot on the disk beyond the middle of the second median interspsce ; the secondarirs with smaller black marginal spots at the extremities of the nervures : below, all the spots are small and black, and there are five tiny squaniose spots across tbe disk of secondaries parallel to the outer margin : body blackish above, whitish below. Expanse of wings, d 63 millim., 2 61 millim. The dry phase (uow sent b-y Ah. Crawsbay) is smaller; the marginal blackish border of the male primaries is reduced to more or less connected spots, the number of black spots on the secondaries is reduced, and on the under surface those of the primaries are absent, whilst the orange costal border of the secondaries is 9 72 DB. A. 6. BUTLER ON BUTTEBPJJES "OV. 28, reduced to less than half its width. The female differs in the same way, but retains its vivid orange colouring. Expanse of wings, d 56-60 millim., $2 54 millim. d d, $2, Tana River, 3800 feet, 4th & 16th January, 1899. Of the male Mr. Crswshay writes-" Plentiful on a bush with a red flower, where I could have taken any number in season ; but nowhere else have I seen this insect." This is probably the Eastern representative of P. pigea. [An allied species of Pinn- coyteyx commou in the same country, but hitherto identified with P. orbona, is described below I.]

52. PINACOPIERYXSPILLERI Staud. d d, Tana River, 3800 feet, 4th & 16th January, 1899. In Staudinger's fignre the under surface of the secondaries is represented as unspotted; this is the case with the single example taken on the 4th January ; all the others have a sene8 of grey spots across the disk; the tint of these wings below varies from sulphur- to butter-yellow. From Mr. Crawshay'a note it appears that this was taken in company with P. astnrte on the same red-flowered bush on the river's bank whilst he waited for his men to find a crossing to the other side.

53. PINACOPTERYSUERDA arose-Smith & Kirby. Intermediate phase.- 0 (shattered), Muthainbi River, 4500 feet, Ndya, 10th January ; d (perfect), Tana River, 16th January, 1899. Dry phase.- d (shattered), Muthanibi River, 6th January ; $2 (perfect), 11th January, 1899. Of the first female Mr. Crawshay writes-"Oblong ova, of a greenish-yellow colour " ; of the second one-'' A perEect specimen at last I The first I saw of this species was on the wet mud of the Tana River, where-when waiting for my net-it was devoured by a dragonfly ; the second and third which I took are both rags. Greenish-white spike-shaped eggs." Professor Aurivillius questions the distinctness of this species from P. simnna, and suggests that the lntter may be a seasonal

1 PINACOPTQRYXVIDUA,sp. n. (Plate LW. figs. 8, 9.) Allied to P. orbma, which it represents in Eastern and Northern Africa : it has well-defined eeasonal phases, the wet base bein most like P. orbona ; the male, however, has a narrower marginal torder to &e primaries ; the apex of these wings and the secondaries creamy on the under surface, instead of white : the female shows far less grey basal shading, not filling the discoidal cell ; the dim1 spota are wanting from the secondarie#, and the orange froin the under surface of the primaries (which in characteristic of P. Znrirnn, the female of P. orbona in my opinion, not of Belenois thysa). Expanse of wings, 8 45- 46 millim., 9 46-50 millim. d ,d, White Nib, F6da (Emin). and British East Africa (Gregory); 9, Wasin, and British East Africa (Gregory). Formerly I referred this species to P. ortygna (of. P. 2. S. 1888, p. 76), me Aurivillius, Rhop. Bthiop. p. 411. It is certainly not P. gerda, which is more nearly related to P. sEmana. 1899.1 FROM BRITISH HAST AFRICA. 973 form of P. charinn. I am quite sure that all collectors of South- African species will dissent from his last suggestion, because the wet, intermediate, and dry phases of P. charina are well known, and (apart from size) can readily be distinguished from P. sinoana by the uniform character of their upper surface at all seasons, and by the absence of the black veins on the upper surface and the black discal spot on the under surface of the male primaries. We do not possess the wet phase of P. gerdcr, in which the outer border attains to a width of 4 millimetres; but even the intermediate phase has a wider border than the wet phase of P. simcina ; and in all the specimens now received the grey at the base of the primaries ie rather more diffused and the black veins are obliterated almost to the outer margin, bringing the species nearer to P. chccrinn excepting for the black discal spot of the male and the absence of the dense speckling which characterizes the dry phase of the Natal species ; the primaries in P. gerdri are somewhat shorter, and therefore less acutely triangular, than those of P. simuna. Before leaving Pinacopteryx it is perhaps as well to point out that “.himvenatus” proves to be s female of a species undoubtedly referable to this genus, and apparently most nearly related to P. liliana ; it certainly has nothing to do with Belenois.

54. HERP~ENIAMELANABOE, var. ITERATA Butl. 9 , Kitwi, 18th January, 1899. 6‘ Dark yellow ova ’) (R.C.). 55. PAPILIOANTHEUS, var. UTUBA Hsmps. Tana River, 3800 feet, 3rd January, 1899. b‘ This, a poor specimen rather, is one of only two seen here in as many days ” (R. C.). 56. PAPILIONIREUS Linn. Q , Undulating plains N. of Tana River, 4200 feet, 14th January, 1899. 66 By no means plentiful; the second or third only I have seen” (R. C.). HESPE R I I D B.

57. SARANGESAELIMINATA Holl. E. of Athi R., some 4900feet alt., Kitwi, 18th December, 1898. 6‘ Undulating uplands, timbered with thorny scrub and very dry. This is, I think, an insect very familiar to me in British Central Africa.” Mr. Crawshay obtained examples of this species at Michako’s, in British East Africa, but not in Nyasa-land; he is probably thinking of S. synestalrnemis. 58. EEETISDJBLm Wugl’. Clue to exact locality and date lost. paOa. ZOOL S0~.-1899, NO.LSIII. 63 974 DR. A. 0.BUTLER ON BUTPERFLIE8 [Nov. 28,

59. hmmsPARADISEA Butl. Tana River, 3800 feet, 16th January, 1899. ‘‘ Taken on a flowering shrub in one of the hottest places and under one of the hottest suns I have ever experienced ” (R. (7.).

60. PYRGUSOOLOTES Trim., var. 8 d, Kangonde, 4500 feet, Kitwi, 31st December, 1898. “Taken when playing together, with a single stroke of the net ” (R.C.). These examples are blacker above and browner below than Angolan specimens, but they are spot for spot alike in other respects.

61. GouuELMA Trim. Muthambi River, 4500 feet, Ndya, 12th January, 1899. ‘‘ An insect new to me ; taken in the early morning on the flags growing on the river’s bank” (R.C.). Mr. Crawshay obtained this insect previously at Kikuyu and Ngongo in July and August, hut it is impossible for anybody to remember every obscure little thing that has passed through his hands j it surprises me that Mr. Crawshay remembers so many.

62. AOLEROSMACKENII Trim. 8 d, Muthambi River, 10th & 16th Janunry; 0, 12th January, 1899. 8. (‘Quite a new insect to me.” ? ‘L No doubt the other of the pair which I saw of which I took the one five or six days ago in precisely the same spot-in a cool glade in my camp close to the stream. I have meantime each day visited the same place in the hope of securing the other insect, but without success until now.” 9 . “ Taken in the early morning on the banks of the stream about twenty yards from the spot where I took the pair” (h?.C.). Either the white spots on the primaries are sometimes almost obliterated on the female or the example taken on the 16th January is a male; it, however, has the white border of the secondaries almost entirely restricted to the fringe; the body having been squeezed out I cannot be sure that claspers are present. In my opinion it is a female (as Mr. Crawshay evidently believed), but an aberrant one.

63. ANDRONYMUSPHILANDER Hopff. Muthambi River, 11th January, 1899

64. PADRAONAZENO Trim. Muthambi River, 3rd January, 1899. 2. (6 Containing two rather large yellow spherical ova. The only specimen of this species which I have seen ” (R.C.).

65. QEQIIXES LEITPEESTEIDTI wllgr. d , Muthambi Riiver, 7th January, 1899. 1899.1 FROM BEITI8H EAST AFRIUA. 975

66. BAOEI~AuRrfmC!cUs Butl. cf , Tana River, 3800 feet, 16th January, 1899. " Taken in my two hands and then manipulated so as to enable me to transfix it with a thorn, after which I was able to administer the necessary pinch " (R.C.). 67. BAORISMARANQA, sp. n. (Plate LXX. fg. 6.) Not nearly related to anything known to me. Wings deep smoky brown; the male with three unequal streaks of bronze- brown only visible in certain lights, divided by the median rein with its first and second branches; fringe towards anal angle of secondaries and anal tuft paler : wings below slightly more silky than above ; primaries with an oblique series OE three whitish dotr beyond the cell; secondaries with a central patch slightly paler than the ground-colour. Expanse of wings 32 millim. The female chiefly differs in having five transparent spots in the primaries-three in an oblique series beyond the cell (answering to those on the under surface of the male), and two placed obliquelp near the base of the median interspaces ; fringe towards anal angle of secondaries and anal tuft whitish. Expanse of wings 34 u~illim. Muthambi River, 4500 feet, Ndya, 7th & 10th January, 1899. Of the male Mr. Crawshay writes, " A new skipper to me;'' and of the female, " Greenish-white spherical ova." This species, though unlike any other African species that I have seen, either in collections or plates, and unrecognizable in any published description that I have met with, is perhaps more nearly related to B. alberti than to any other known species. 68. RHOPALOQAMPTAANCHIsES Gerst. Q , Bondoni Plains, 5100 feet, 8th December, 1898. In the present collection Mr. Cramha-y unfortunately did not follow his usual plan of putting the exact locality on every envelope, but in some instances referred to a number ; thus :- See no. 166, 125, or 146, ~.laceand date. Not being at all prepared for this, I took no special note of these collector's numbers until I came to one of the notes, by which time no. 166 had been inco orated and I had lost all clue to it ; the others I happily foun?: this will account for the absence of exact localities and dates for several species, they are all referred to no. 166 (cf. Acrcta cabira, Caqreus Enyeus, and Eretis djdcelre). One other species was obtained by Mr. Crawshy-a possibly new Ppthima,-but in such a shattered condition that it is neither fit for description nor for the collection ; it is also referred to the unliicky number. EXPLANATION OF PLATB LXX. Fig. 1. Catochryaope pecudiaris, Q (intermediate phase), p. 965. 2, 3. Chkwowh azureu, d Q, p. 967. 4. Evercs kedonga, 8, p. 967. 6. Baaria maranga, 8,p. 976. 6,7. Pinaccrptetyx darte, cf Q , p. 971. 8, 9. ,, vidw, $2, p. 972. 63 @ 976 DE. A. 0. BUTLER ON BUTPEEFLIES [Nov. 28,

6. On a small Collection of Butterflies from the Nandi District, TJganda Protectorate, Eastern side of Lake Victoria; made by Captain Hobart, of the Grenadier Guards. By ARTHURG. BUTLER,Ph.D., F.L.S., F.Z.S., &c. [Received November 1, 1899.1 The collection of which this is au account is small but interesting, the whole of the specimens having been captured by Capt. Hobart on the march. Among them is a Cyrnothoe which appears to be quite new, and of considerable interest from its affinity to Western forms; also Planemn poggei d, and Acrma leucop-apha, an extremely beautiful variation of the Western Acrcea admatha. NYMPHA L I D B.

1. AMAURISAT~BIMACULATA Butl. The white spots on the primaries are reduced in size in all the specimens, but we have similar examples in the Museum collection. 2. CHARAXESCANDIOPE Godt. One male example.

3. CYMOTEOEHOBARTI, sp. n. The male above vermilion suffused with carmine, the costal and outer margins narrowly black j primaries with a small apical patch and one subapical spot, sometimes continued indistinctly as a submarginal series ; secondaries with well-defined submarginal black spots commencing with a sagittate spot at apex and terminating in an obtusely biangulated linear marking above anal angle; abdominal border pale brown; body bronzy brown, the palpi and under surface of antenna1 club tawny. The under surface is of a sandy-brownish hue, with fleshy and weak olivaceous change- able tints ; the general pattern is that of C. useldcc, the markings on the basal half being sharply outlined in black; the nearly straight dividing line beyond the middle is dark rich brown in the type, but weakly defined in a second smaller male ; the series of A-shaped markings beyond the dividing stripe are alternately pale pearly pink and olivaceous greyish, the outer series of the latter tint uniting into a continuous wavy submarginal line dotted with blackish between the nerrures j the femora are whitish. Expanse of wings 42 to 58 millim. The female nearly resembles that sex of C. adeZa, excepting that the basal area is internally suffused with olivaceous greyish and externally with pale sandy yellowish, the discal series of sagittate markings is weaker in the primaries and very much more so in the secondaries ; on the under surface the general appearance is 1899.1 FaOY NAND1 DISTRIOT. 977 even closer to that of C. adela Q , but the central stripe is straighter and very dark, with the irregular series of spots which bound it internally white, the outer border (excepting at its extremities) sandy brown, and the discal markings very indistinct. Expanse of wings 64 dim. Two males and one female were obtained.

4. PRECISAURORINA Butler. A female (very much shattered) of ail example in which the tawny band crossing the wings is nearly half as wide agsin as usual. If this should be proved to be n constant difference in Uganda specimens, it would be necessary to regard them as representing a distinct local ram.

5. PRECISCEBRENE Trimen. Two males.

6. PRWISBOOPIB Trimen. One shattered male. 7. ~EIUISGREGORII Butler. A pair in poor condition.

8. CYREISTISEILEQANS Boisd. A much shattered specimen. 9. EURYTELAHYAEBA Fabr. Two examples. It is difacult to decide whether these Eastern examples should be assigned to typical E. hyarbn or yap. angustata ; they seem to vary as regards the width of the white band. 10. VANESSULAMILCA Hewits. An example with an unusually broad tawny band, remarkably resembling the female of Precis nurorincc in the same collection excepting in size. 11. PLANPMALATIBASCIATA 3. M. Sharpe. Two males. This ia new to us. 12. ACRXAADMATHA, var'. LEUCOGRAPHA Ribbe. Two males. This is an extremely pretty local race of the species, which Capt. Hobart assures me was quite common.

P APILI ONI D B.

13. NYCHITONAXYDUSA, var. IMMACULATA Auriv. One male. 978 MR. J. Y. JOHNSON ON THE [Nov. 28,

14. TBRIASBBNEBALENSIS, wr. BISINUATA Butler. One female.

15. BELENOISNEBENTINA Crainer. A pair.

16. GLUTOPHRISSASABA, yar. cos;TRAcrA Butler. One male.

17. LEUCERONIATHALASSINA Boisd. One male.

7. Note on the Habit and Mode of Growth of the Corals belonging to the Genus Pleurocoralliurn. By JAMES YATEJOHNSON, C.M:Z.S.

[Received Noveiiiber 2, 1899.1 In my comniunication to the Zoological Society on the Coralliide of Madeira (P. %. S. 1899, p. 57) nothing was said as to the cause or meaning of the peculiar habit and mode of growth of the four known species of Pleuromrdlium. Whilst the species of Corullium, such as the red cord of the Moditerranean, branch in all directions and put forth their polype-cells on all sides of the branches, the i’leui*owrnllin ramify more or less in one plane and their polype- cells are confined to one face of the brancheg. When Dr. Gray first alluded to the matter (P. Z. 8.1867, p. 125) he said, “ I have no doubt that it (the coral under description) grows out horizontally from the rocks, and that they (the polype-cells) arise from the upper surface of the branches.” This would appear to have been only a conjecture; but in his Catalogue of Litliophytes or Stony Corals in the British Museum (p. 24) he went further, and described them as “ growing horizontally from the sides of rocke,” witbout citing my authority for the statement. The facts about to be mentioned seem to throw doubt on the correctness of Dr. Gray’s view, and to suggest another conclusion. In describing the only known specimen of Pleurocorallium nt - Clereiise (Zoc. dt.), I stated that some zoophytes of rare occurrence at Madeira were growing parasitically upon it. Two of these were branched specimens belonging to the genera Xuberea and Steiulla, and they were seated on different parts of their host at a distance from each other. The point to which I wish to draw attention is, that both these Alcyonarians had groan in the plane of the host. If that had extended horizontal1y, they too had extended hori- zonhlly. But can we suppose it possible that they would take in 1899.1 GBNVS PLI~~OOOEALLIU~~, 979 growing any direction other than a vertical one, whatever might be the position of the coral on which the embryos had settled? May we not hold it as certain that they had grown upright, that is towards the surface of the sea ? If so, the Pleurocorallium must have taken the same direction. Another piece of evidence having the same bearing is afforded by some specimens of a simple Madreporarian coral (Desmophyllum). All four examples were attached by their bases to the front of the coral, but two of them had twisted themselves round and had pushed their calyces between its branches to the other, that is the posterior side. It is not probable that tho Dcsmophylla would have acted so if the supporting coral had possessed a horizontal position, because their calyces would then have been directed downward to the bed of the sea. Again, it is observable that in a specimen of another species of .Phrooorallium in my possession where a small branch by some accident had been broken off the living coral, it bad fallen upon the spreading base, to which it had in course of time been made to adhere by an extension of the growing coenenchyma of the base. It is not easy to understand how the fractured branch could have lodged upon the base in the manner it has done unless the coral had been upright. In view of these facts, I do not see how we can adopt Dr. Gray’s hypothesis, however plausible it may appear at first sight, or come to any other conclusion than that these coral8 assume in their growth a vertical, not B horizontal position. But if this is so, what is the meaning of the fact that the polype-cells are confined to one face of the branches ? If we suppose that the habitat of the corals is in that part of the sea’s bed where a constant current is flowing, it is clear that it would be more beneficial to the colonies if all their polypes were turned towards the direction from which their food comes, than if half of the polppes were turned in the opposite direction. A colony will obviously obtain the largest possible supply of nutriment when all its members face the current that carries it, we may say, into their mouths. To summarize what has been said : in order to account for the fan-like mode of growth of the l’leurocorallia and the unilaterality of the polype-cells, Dr. Gray maintained that the corals grow in a horiiontal poeition. As such a position is not easily reconciled with the facts above stated, the suggestion is now put forward that the corals grow upright in the path of a submarine current; with all their polypes opposed to the on-coming stream. 980 ON THE MOULT OF THP~KING PEINGUIN. [Nov. 28,

8. Further Notes on the Moult of the King Penguin (Apteno- dytes pennanti) living in the Society’s Gardens. By W. E. DE WINTON,F.Z.S. [Received November 24,1899.1 Some observations on the moult of the King Penguin were offered by me in November 1898 (see P. Z. 8. 1898, p. 900). I nm now able to supplement those notes by further important particulars, the same bird, still alive and well in the Society’s Gardene, having again gone successfully through the moult. The dates given for last year were again closely adhered to, and the succession of changes were made pretty much in the same order ; hut I am able now to explain fully the very. strange appearance of the old feathers, which were then likened to withered leaves. The coloured portion of the sheath of the bill was shed as before. An examination of the freshly moulted feathers showed that the bases were unlike those of the feathers of any other bird 80 far as I am aware. A

Feathers of Aptenodytes pesmati. A. Part of uew feather of flipper, with old feather still attaohed. B. Moulted breasbfeather, showing flexible sheath attached to its base.

The quill does not end sharply with a contracted base, but is shed with a flexible sheath attached to it (fig. B). This sheath, which is a continuation of the outer coat of the quill, is, in the body-feathers, as long as the naked part of the shaft. On handling the bird it was found that its enlarged puffed-out appearance at the beginning of the moult aro8e from the old plumage being actually raised, and now adhering only to the new feathers that were growing into the

P.z.s.1839. P1.LXXi

J Smrt del ethth kterr.3rcs Imp SKIN OF SMITHEMAN'S KOB (Cubus sm1thenianl 1899.1 ON THIO SKIN OF AN ATIIPIOLOPBFBOM ZAK~MWEII~U. 981

bases of the old ow,the thin sheaths attached to the bases of the feathers being occupied by the points of the new feathers. All the feathers so raised had lost most of their ori@nal colour, or lost it entirely, the yellow feathers of the neck havlng bleached white, and the slate-blue feathers of the back and flippera being dull brown or drab except at the extreme tips. The greater part of the old plumage is removed by the bids bill as soon as the new feathers upon which it is raised are sufflciently developed to form a covering, but many feathem upon the back and flippers may be left for a longer period (fig. A) ; theee scattered feathers, adhering to the now nearly fully developed new plumage, produce the appearance of small crinkled leaves, which puzzled me so much last year. The small body-feathers have a large downy aftershaft ; the quill-feathers have a naked shaft as long as the plumed portion. The feathers of this bird are so unlike those of any other bird, in the entire absence of a raised midrib as well as in their umbilical portion, and the nature of the moult seems to open up such , interesting questions, that I have placed all the materials possible in the hands of my frieud Mr. W. P. Pycraft, who I hope will shortly publish the result of investigat,ions which he is about to make upon these specimens.

9. Description of the Skin of an apparently new Kob Antelope from the Neighbourhood of Lake Mweru, with Note on a Skull and Horns of an Antelope of the same Genus. By R. LYDEKKER. [Received November ZO, 1899.1 (Plate LXXI.) I am indebted to Mr. Rowlrtnd Ward, F.Z.S., for the opportunity of exhibiting to the Society this evening the skin of the very handsome and apparently new species of Kob Antelope forming the subject of the drawing (Plate LXXI.). The specimen, which consists of a flat skin, wanting the head, feet, and the greater portion of the tail, was obtained by Mr. F. Smitheman, F.Z.S., in the neighbourhood of Lake Mweru, situated to the south-west of the lower end of Lake Tanganyika. It arrived in England during last summer, and there were hopes that the head would follow ; but if the latter was ever despatched at all, it has evidently miscarried. The skin has the appearance of belonging to an adult animal, and there is every probability that it pertained to a male. On the under surface there are rudimentary mamma, which indicate that the skin must be that either of a male or of a young female. If it were that of a young female, it would indicate that the adult animal was of very large size; but I think it may be pretty confidently assigned to an adult male. t!From the general characters of the pelage and its coloration, and 982 ME. It. LXDPKPEB ON "ECB SKIN [Nov. 28, especially from the long shaggy hair on the nape of the neck and the absence of a mane, the specimen may be confidently assigned to the genus Cobus. Additional evidence in favour of this reference is afforded by the circumstance that the long hair on the middle liue of the back is reversed from a point some distance in advance of the loins to the withers, exactly as in the Piiku (C. vardoni). In size the auimal to which the skin pertained must evidently have been considerably larger than the species last mentioned, and may have been more nearly comparable in this respect with C. mugu of the Bahr-el-Ghazal. In colour, the nape, the sides of the buttocks, and thighs are bright chestnut-tawny ; the middle line of the back, the hinder portion of the shoulders, and the hind-quarters are the same chestnubtawny mingled with blackish brown ; the fore part of the shoulder, a line on the under surface of the neck, the flanks, and the front surface of the fore-limbs and of the lower part of the hind limbs are of a deep glossy blackish brown, the under-parts being dirty white. The portion of the skin of the neck remaining, which seems to have been cut off a considerable distance below the head, is suggestive of a comparatively long-necked animal. And if this be a correct inference, it would be natural to expect that the horns were of a comparatively long and slender type. Now, in its dark colour the skin is more like that of Cobus maria, of the swamps of the White Nile, which is a species with comparatively long, slender, and doubly curved horns ; and it is to that animal, rather than to any other member of the genus Cobzcs, so far as the materials ermit of forming an opinion, that I am inclined to consider t e form represented by the skin before us most nearly related. AltogetherB apart from the distance between the White Nile and Lake Mweru, the skin under consideration is broadly distinguished from the male of Oobw maria by the absence of tho white patch on the withers and the white line down the back of the neck. As regards the female of the latter species, there is a dehite statement in the 'Book of Antelopes,' vol. ii. p. 122, that it is similar in all respects to the buck, except for the lack of horns; but in the plate a female is figured without the white saddle and neck-line '. As there is no other Antelope of which the skin is known that presents any close resemblance to the specimen under consideration, and since it is certainly distinct from the male of C. muria (being itself probably a male specimen), I take leave to regard it as representing a new species, for which I propose the name of Cobus smit7wnani ; the skin represented in the figure (Plate LXXI.) being of course the type. The Rpecies may be provisionally defined rn a large-sized Kob, differing from every other species of the genus exce t C. mariu, and distinguished from the male of the latter by t E e absence of the white line down the back of the neck and the patch of the same colour on the withers, in which region the present species is chestnut. From a ~pecimmthat has recently oome under my notioe, the white neck- line does not seem to be eonstant even in the bucks of C. muria. 1899.1 OF AN ANTI%OPM FROM LAK.M MWEEU. am I may now mention another piece of evidence in favour of the relationship of this Cobtts smithernalti to C. maria. A correspondent of Mr. Rowland Ward writes that in the neighbourhood of Lake Mweru he has seen an Antelope very like a Situtunga, with similarly elongated hoofs, but with horns more like those of a Lechwi, although longer and slighter. Such a description would admirably fit the present species, if, as I suggest, it be more nearly allied to C. maria than to the Puku. Supposing it to have elongated hoofs, I should not regard such difference as, at the most, of more than subgeneric importance.

Skull arid horns of Loder’s Puku (Cohuardoni loden’).

Before concluding this communication, I may call attention to the skull and horns of a Puku-like Antelope in the collection of Sir E. G.Loder, which I first thought; might belong to the same species as the above skin. This skull and horns, of which the locality is unknown, are shown in the accompanying drawing. The 984 ON mm MEINOF AN A~~PEIF~OMLAKE ~RU.[Nov. 28, specimen is essentially of the Puku type, but broadly distinguished by the circumstance that while the skull itself is slightly shorter than that of an average-sized Puku skull in the British Mumum, the horns are very much longer and stouter. In the Puku skull the length from the fronto-parietal suture to tfhe tip of the nasals, measured in a straight line, is 8.5 inches, while in Sir E. Loder’s specimen the corresponding dimension is but 8 inches. In the present specimen the horns hare more ridges (17) and relatively shorter tips than any Puku horns 1: have seen ; they measure 20.3 inches along the front curve, 8.0 inches in basal circumference, and 8.1 inches between the tips. Now the. only horns assigned to the Puku with which I am acquainted that have anything like thew dimensions are a pair obtained by Mr. Smitheman from the Luswesi Valley, in the neighbourhood of Lake Bangweolo, which lies S.S.E. of Lake Mweru; these horns measuring 204 inches along the curve, €34 in basal girth, and 124 from tip to tip ’. The wide interval between the tips I consider of no importance, but in other respects these horns agree very closely as regards measurements with Sir E. Loder’s specimen. And they differ from the next specimen of Puku horns (IS& in.) in Mr. Rowland Ward’s list by the much greater girth, the basal circumference of the latter being 68 inches. Accordingly, so far aa horn-measurements alone are concerned, there would seem a probability that Mr. Smitheman’s Lake Bangweolo skull may be specifically identical with Sir E. Loder’s specimen. And if this be so, there arises the question whether both are not referable to C. mithemani. But if the evidence of the correspondent quoted above as to the LechwLlike character of the horns of the Antelope presumed to be identical with the latter be reliable, this can hardly be the case. It must also be remembered that Lake Bangweolo is a considerable distance from Lake Mweru, so that each district (in spite of the fact that the true Puku and Lechwi extend from the Chobi-Zambesi Valley to Lake Mweru) may have its own piirticular species or race of Kob. The matter is one of great difficulty, and I may be accused of rashness in what I propose to do, which is to consider, for the present, Sir E. Loder’s specimen as typifying a large-horned race of Puku to be known aa cobus uardoni loderi, until it can be either proved to be the same s 0. mithemani or entitled to rank aa a species by itself. W dether Mr. Smitheman’s large Puku head from Lake Bangweolo belongs to the same form may be left an open question.

See Rowland Ward, ‘ Records of Big Uame,’ p. 189 (1899). * Loc. cs‘t. 1899.1 ME, SCLATEllt ON OVIBOS MOSCILITUS. 985

December 19, 1899. Dr. HENRYWOODWARD, LL.D., F.R.S., V.P., in the Chair. The Secretary read the following report on the additions to the Society’s Menagerie during the month of November 1899 :- The registered additions to the Society’s Menagerie diiring the month of November 1899 were 123 in number. Of these 35 were acquired by presentation, 8 by purchase, and 1in exchange, 78 were received on deposit, and 1 was born in the Gardena. The total number of departures during the same period, by death and removals, was 110. Amongst the additions attention may be specially directed to the two Snake-Wishes (PoZypterus mnegalus) from the River Gambia, obtained by Mr. J. 5. Budgett, F.Z.S., during his recent ex edition to the Gambia, and presented by him on Nov. 22nd. TK ese are believed to be the first examples of this remarkable fish ever brought alive to Europe.

On behalf of Mr. G. 9. Mackenzie, P.Z.S., s photograph ww exhibited of two remarkably large tusks of the African Elephant (Elephas africanus) recently sold at Zanzibar, and stated to have been obtained in the district of Kilimanjaro. They each measured, on the outside curve, 10 feet 4 inches in length, and weighed respectively 235 lbs. and 225 Ibs.

Mr. Sclater exhibited the hind portion of the skin of a Giraffe, which had been shot on the east bank of the Great Loangwa River, Northern Rhodesia, in latitude 13O South, and read the following extract from a lett,er on the subject addressed to him by Mr. Alfred Sharpe, dated Zomba, June 14th, 1899 :- 6‘ As you know, there have been from time to time reports of Giraffes existing north of the Zambezi on the Loangwa, but no one has been able actually to verify this until now. This skin was sent to Mpeseni’s while I waa there, the beast having been shot by a prospector. He stated that they were not plentiful at all, and were restricted in area, but that he had seen a herd of 35. The skin was sent down to Capt. Chichester, and was not complete, 89 it consisted of the hind-quarters only, but possibly this will be sufficient for determination as to whether it belongs to a different variety from the 8. African Giraffe.” Mr. W. E. de Winton, F.Z.S., who had examined this specimen, was of opinion that it was decidedly referable to the Southern form (cfircrfa capensis).

Mr. Sclater stated that during a recent visit to Woburn he had had the pleasure of inspecting, under the guidance of the President, two young male Musk-oxen (Oviboe moachatue) which had been 986 YR. w. B. DB WINTON ON DBNDEOMYS LOYATI. [Dec. 19, lately added to the collection. According to information kindly supplied to Mr. Sclater by (Xrosserer Anton Nilss, of Tromso, these animals had been captured on Clavering Island, near Cape Mary, East Greenland (about 74' N. Int.) on the 16th August, 1899.

Young male Musk-ox, livine at Woburn. (From a photograph taken by H.a. The Duchess of Bedford.)

Mr. Sclater exhibited photographs of these animals taken by the Duchess of Bedford, and stated that he believed these specimens to be the first examples of this remarkable mammnl that had renched Europe alive. - Mr. W. E. de Winton exhibited a remarkable Mouse of the genus Dendromys, obtained by Lord Lovat at Managatha in Southern Abyssinia, for which he pruposed the name Dendrmys lavati. This new species was of about the same size as D. typicus, but was striped to almost the same extent as the Barbary Mouse (Atvkunthis barbarus). Perhaps the markings would be more easily realized if likened to those of the Chipmunks (Tuamias): the broad black dorsal stripe was divided by a narrow grizzled 1809.1 ON THEI SKULLS OF SOYEl MALAGASY LBhfURS. 987 central line ; on either side of the black stripes were pale fawn stripes ; outside these sgain were black stripes. The general body- colour was soft greyish brown. The fur was very soft, like that of Malacothrix, so that the general effect of this colouring was patticularly pleasing. The tail was barely so long as the head and body, and was thickly covered with short hairs.

Mi. R. E. Holding exhibited, on behalf of Mr. William Pierpoint, a series of the horns of the Siberian Roebuck (Capreolus pygargw) brought from the Gulf of the Obi, Siberia, and pointed out some remarkable vnriations in the form and size usually characteristic of the horns of this species. Mr. Holding also exhibited a pair of the horns of the dltai Deer (Cervus eustephanua) from the same district, which were mainly interesting on account of the absence of the third tine on both horns--a somewhat unusual case, mi the third tine in this group of Deer is the most persistent, the “bez” tine. being usually arrested. A pair of horns, probably of the same species, also showing the third tine absent, had been shown by Mr. H. J. Elwes at a recent, meeting of the Linnean Society and figured iu the Journal of that Society for 1899 (Zool. vol. xxvii. p. 32).

Dr. Forsyth Major, F.Z.S., exhibited several skulls of fmtal Malagasy Lemurs, partly collected by himself and partly lent to him by the Hon. Walter Rothschild, M.P., and Prof. Charles Stewart, F.R.S., and made the following remarks :- All the Malagasy Lemurs, Chivomye included, exhibit a remarkable peculiarity of their tympanic bulla, the annulus tympanicus taking no part whatever in its conformation. This condition is unique amongst the Mammals, if we except the Insectivorous form Tupah (Winge), to which I am able to add the nsarly related genus Ptilacercus. To decide the question whether this is a primitive condition in Malagasy Lemurs, we have in the first place to investigate how the bulla is developed. In the youngest stage available to me for examination, the fcetus of a Chiromys, there is no trace of an osseous bulla; the completely ossified annulus lies almost horizontally underneath the periotic. In a second stage (Lepidalemacr) ossification begins to be developed from the lower sharp margin of the periotic, which adjoins the annulus. In a third stage (LepicEolemur) this outgrowth appears increased, and has a shell-like shape, with the concavity turned outward; the annulus is gradually being uplifted by it. In a fourth stsg0 (Lemur rmbriventer) the shell-like ossification is still more increased, and begins to cover the median part of the annulus ; and this state of things is still more increased in the fifth (Lepidolemur) and sixth stage (Avahis laniger), with the result that first the median part, and eventually the remainder of the annulus becomes invisiblewhen viewed from below, being shut by theperiotic 9 88 ON SUBFOSSIL MAMMALS FROM MADABASUAB. pec. 19, In the adult (as will be seen by the skull of an adult Lemur mcbri- ventm which I exhibit) the annulus is represented by a bony ring- the size is scarcely larger than in the youngest stages-which hangs freely in the tympauic cavity, being coalesced with the squamosum only in one part, viz. anteriorly to the stylo-mastoid foramen. Ontogeny thus teaches us that the annulus of the adult is not a secondarily detached part of the bulla. In the second place, I have to state, in connection with the above, the important fact, that in the Tertiary Aclapis the annulus tympanicus is a free ring, independent of the bulla, absolutely as in the Malagasy Lemurs. Besides, in the large development of the bulla and in the conformation of the whole of the bnsicranium (in the shape and position of foramina &c.), Adupis closely resembles the Malagasy Lemurs. So that, far from agreeing with Osborn and Wortman, who place Adnpis among the “primitive Anthropoidea,” I now see no reason for separating it as a family from Malagasy Lemurs. In the Oriental and Ethiopian Lemurs both the annulus and an outgrowth from tho petrosum enter into the composition of the bulla. In a young Nycticebus (which I exhibit) it is to be seen that the median part of the bulla is,as in Malagasy Lemurs, formed from an appendage of the periotic, which becomes co-ossified with the annulus ; in the specimen exhibited the suture between them is distinctly visible. The annulus, in its turn, no longer plays the passive partthat it does inMalagasy Lemurs, but grows out laterally, so as to form the lateral part of the tympanic cavity, which, however, never reaches the dimensions it has in Malagasy Lemurs. I have not, for the present, sufficient material to follow the process of development in detail in other Malagasy Lemurs. In the skull of a half-grown Q;zlugo, it may 6e seen that the composition of the bulla is essentially the same as in Nycticebua. From the close agreement in cranial characters between the last-named and Loris and Perodicticus, it may be safely argued that in the development of their bulla they also agree with Nycticebus. The same holds good with regard to !Pursiua, as shown by a young skull of Tarsius spectrum now exhibited. -- Dr. Porsyth Major also asked leave to exhibit specimens of two subfossil Mammals from Madwascar, which would be fully described later on ; but he preferred not to delay their exhibition, as very soon it would probably be no longer in his power to exhibit them. Dr. Forsyth Major made the following remarks :- This almost complete skull, togetther with a mandibular ramus, represents a new species of Nesopithecw, which may be called Nesopithecus australis, sp. nov. It is distinguished from N. roberti by its smaller size, by the less steep facial profile, by the position of the lachrymal foramen situated on the margin of the orbit, and not inside as in I?. roberti, and by the slightly outward direction of the orbits. This beautifully preserved ripecimen shows that the genus Blobilemur founded by me an the posterior, and Nesopithecwr roberti 1899.1 MR. w. L. ECLATER ON TnB 'FAUNA OF SOUTH AFBICA.' 9d9 founded on the anterior parb of a crauium, are one and the same and it further shows that the skiill of Xesopithecus is provided with several features characteristic of the Malagasy Lemurs, amongst them being the character of the bulla before described, in which it completely agrees with them. The bull= are very spacious, and the outer opening of the matus nuditorius being very large, the free tympanic ring can be seen through it without difficulty. Other features of this remarkable skull are the following :- The orbits are open behind. The number and notation of the teeth are the same as in N. robe& In the number of the upper series the latter agrees with American monkeys, but at the same time with the Lemurinm. In the !ower series the number of premolars is as in the latter ;the number of incisiform teeth is two, as in the lndrisinae. On the other hand, several features presented by N. roberti, and in a minor degree by the present species, are decidedly those of the Anthropoidea, and scarcely a single oue of the characters considered to distingiiish the Lemuroidea from the Anthropoidea holds good in the case before us. In both thespecies oE Nesopithecus the upper incisors are not separated in the median line; in their shape they decidedly resemble the incisors of the Cercopithecidae, the lower iricisiforui teeth being inserted nearly vertically. The true molars, as previously stilted (cf. Geol. Mag. 1896, p. 435), present the pattern of the Cercopithecidae. The facial profile is steep in both species ; very steep in N. roherti, in which the orbits are directed straight forward and the lachrymal foramen is situated inside the orbit. These resemblances to the Monkeys are not -limited to the skull, but extencl to almost every oue of the bones of the skeleton, most of which are at haud. As the question at present stands, we have then to inquire whether Nesopithecus is the most highly evolved of the Lemiiroidea or the lowest of the Anthropoidea-that is to say, are its Simian characters independently acquired or not? I do not iutend to enter into this difficult question this evening, as an attempt to answer it can only be made after afuller description and discussion of all the characters. . The beautifully preserved upper molar teeth of Mqaladapis now exhibited form part of a complete set OF teeth recently received from Madagascar. and agree in all particulars almost exactly with the correspondinhg feet of Neycila&piS ntad~igascuriensb,but they are at least one-third larger, and thus indicate a huge Lemurid, the skull of which must have had the approximate length of 330 mm. I propose to call this new species MegaEadnpis insignis.

Mr. W. L. Sclater, F.Z.S., Director of the South kfricau Museum, Cape Town, explained the nature of a series of volumes, contributed by various authors, which he proposed to issue under the title of the ' Fauna of 8ouBh Africa.' The first volume, which he hoped would be ready in a few days' time, would deal with the first portion of tJhe PasseFine birds ; it had been prepared by the PROD. ZOOL. 800.-1899, NO. L?(lV. (i 4 990 MESSES. B. C. A. WINDLE AND F. G. PARSONS ON [Dec. I!) late Dr. Arthur Cowell Stark, whose tragic death at Ladaysmithon Nov. 18th last had been recently reported. Dr. Stark, who had spent many years in different parts of South Africa, and bad made its avifauna his special study, bad, it was believed, near1.y completed a second volume containing the remaining families of the Pnsseres, the MS. of which Mr. Sclater had good reason to hope would be recovered. Mr. Sclater bad himself completed an account of the Mammals, which was already in the printers’ hands. Other volumes ou the Reptiles, Batrachians, Fishes, and some of the groups of Inverte- brates would follow. The area embraced in the ‘Fauna of South Africrs’ would be that portion of the continent which 1a.y south of the Zambesi and Cun6n6 Rivers, and would contain the English Colonies of the Cape and Natal, Southern Bhodesia, the two Dutch Republics, and the adjoining German and Portuguese territories. Passing on specially to the Mammals of this area, Mr. Sclater pointed out that, so far as his present information went, about 236 species had been hitherto recorded within these limits, but tbtlt there could be no doubt that, especially among the smaller forms, a great many more species remained to be added to the list by future investigators. Mr. Sclater concluded his remarks by speaking itbout some of the older travellers and collectors to whom we are mostly indebted for our earliest knowledge of South African zoology.

The following papers were read :-

1. On the Myology of the Edontata. By BEKTRAMC. A. WINDLE,D.Sc., M.D., M.A., F.R.S., Professor of Anatomy iu Mason University College, Birmingham, and F. G. PARSONS,F.R.C.S., Lecturer on Human and Comparative Anatomy at St. Thomas’s Hospital, late Hunterian Professor in the Royal College of Surgeons, England. [Received November 9, 1899.1

Part II.-MUSCLES OP THE HINDLIMB; AND SUMNARYOF CONCLUSIONSRESPECTING THE MUSCULATUREOF THE ORDER. The first part of this paper, dealing with the musculature of the trunk, head and neck, and anterior limb, was read before tbis Society on March 7th, 1899 (cf. Proc. 55001. SOC.1899, p. 314). For convenience of reference we again append the list of animals on the dissection of which our generalizations are founded. The Arabic numerals before the name of each auiind refer to the mention made of it in the text, whilst the Roman numerals following each name relate to the bibliography at the end of the paper. Where no such numeral is affixed it may be understood that we 1899.] TLLE MllOLOGY OP THE IDBNThTA. 991 are responsible for the dissectioii ourselves. We have to thauk Mr. Burne, of the .Royal College of Surgeons' Museum, Londou, for kindly permitting us to use the notes of his dissection of a specimen oi Chlamydophorus. List of Animals. Family BBADYPODIDIE. 1. Braclypus tridactylus.

2. ,9 99 (Humphry, 1V.) 3. Y, YY (Macalister, XlV.)

4. 9, 99 (Meckel, XI.) 5. 19 9, (Mackintosh, XVI.) 6. I1 5, (Cuvier & Laurillard, XVIL.) 7. 99 99 8. Cholctps clidactylus.

9. 9, 9, (Huinphry, IV.) 10. 9, 9, (Mlnckintosh, XIII.) l'ainily MYBNECOPIIAGIDB. I 1. iWyrmacapkaga jubata I. 12. 93 9) (Couvreur & Bertaillon, 11.) 13. 99 1, (Macalieter, I.) 14. Tamnnclun tetradactylu.

15. 9, 9, (w?p, 111.) 16. 9, 9, (Cuvier & Laurillard, XVII.) 17. Cyclothurus clidactylus. (Humphry, IV.) 18. 9, 33 (Macalister, I.) 19. 99 $3 (Meckel, V.) 20. 71 9, (Galton, Vl.) 21. 99 99 (Cuvier & Laurillard, XVLI..) Family DASYPODIDIE. 22. Dasyplis viuosus. 23. ,, seacinctus. (Balton, X.) 24. ,, 9, (Cuvier & Laurillard, XVH.) 25. Tatueia pba. (Macalister, VII.) 26. ,, sp. inc. (Meckel, XI.) 27. Chlumyclophorus truncntus. (Macalister, VII.) 28. 99 ,, (Hyrtl, XII.) 28 a. Yf ,, (Burne, MB. notes.) Family MANIDB. 29. Mnnis maorura. 30. ,, '? 31. ,, sp. IUC. 32. ,, aurita. (Humphry, IV.) 38. ,, tricuspis. (Macalister, 1.) 34. ,, jawanica. (Macaliskr, VlI.) R.C.S. Nag. Collection. 64* 992 YESSRS. B. 0. A. \VINULE AND E. U. PAILSONB ON [hC. 19, Family ORYCTEROPODIDB. 35. Orycteropus capemis. (Galton, VIII.) 36. 9, 13 (Humphry, IX.) 37. 9, ,9 (Cuvier & Laurillard, XVII.) Ectogluteus ( ffluteus maxirnus) and Cctiido-fmoTalis (Ayitntoi. cctu&e).-In the Bradypodidce the ectogluteus rises by fascia from the sacral and caudal spines ; according to some authors from the crest of the ilium also. As its origiu is fascial, there is clearly cou- Riderable scope for diversity of opinion and description on this point. The insertion is into the shaft of the femur frotn just below the great trochanter to the middle or rather lower. No separate caudo-femoralis has been described in these animals, except in Cuvier & Lrturillard’s specimen (G), in which it was very small. It seems quite probable that;, in the other specimens of which descriptions exist, it is fused with the ectogluteus. This hypothesis is supported to a certain extent b.y the fact that in our specinieu of ChoZceyus (8) the muscle had a double iusertion, the anterior fibres passing to just below the great trochanter, thus obtaining the normal insertion of the ectogluteus in mammals, whilst the posterior fibres reached the middle and lower thirds of the femur. Among the Myrmecophagidce the ectogluteus and caudo-femoralici can usually be separated from one another, the former passing to the upper part of the femur and to the fascia lata, the latter to the lower part of the bone. This was certainly the case in Myrmeco- phnga (11) and Tumcindua (14,16) and, to B less extent, in Cyclo- thurzcs (21), though most of the dissectors of this animal do not seem to have recognized the caudo-fernoralis as a separate muscle. In the Dcisypodidm and Xanicke both the ectogluteus and caudo- femoralis are present, though usunlly closely united. In Chlan~y- clophorus (28 a) the latter rose sepnrately from the spharoma end of the lower spharonia support. In the Orycteropodiclce the ectogluteus is inserted into just below the middle of the femur, and the caudo-fernoralis into the lower end (37). Tensor fnscim jhnoris and ilio-tibieilis (Sartorius).--In the Brdypodidce the ilio-tibialis is a well-marked muscle rising front the crest of the ilium and passing to the inn-r side of the head of the tibia. This applies to Brurlypus (1, 4, 5, 6) and Choltqms (8,Q). The rnuscular fibres rising external to this and in the same plane, instead of going to the fascia lata, accompany the ectogluteus to the outer surface of the shaft of the femur, but we are of opinion that they represent the tensor fascire femoris. In the Myrmecophagidm the ilio-tibialis is distinct and has the human attachments in &Iyn~eeop7zap(11) and Cyclothurus (17,21). In TcimnncZua (14) and Cyclothurus (20) it rises from the tendon of the psoas magnus or parvus. The tensor fascire femoris is iti- separable from the ectogluteus. In the Dasypodidm both muscles are present, but whereas the ilio-tibialis is a delicate muscle in Uasypus (22, 23, 24), it is thick and fleshy in Chlamytlophortur 1899.1 maMYOLOGY OF THE EDFNTATA. 993

( 27,28). In the Manide (29,32,33) the ilio-tibialis, tensor fascire femoris, and ectogluteus form a continuous sheet as in the majority of mammals, but the ilio-tibialis is remarkable for the frequency with which it is wholly or partly inserted into the inner side of the patella. In the Orycteropodidce, Galton (35) dewribes the ilio- tibiaiis as arising from the ilio-pectinpal tubercle, and Humphry (36) as coming from the last rib just external to the psoas. Jlesoyluteus, h’ntog~uteus, aid Pyrifornzis.-The Edentata are remarkable for the imperfect differentiation of the meso- and entogluteus and the pyriformis, which three muscles form a large fleshy mass. Such being the case, it is not surprising that the literature of the subject preseuts us with very varying accounts of their condition. The fleshy mass above mentioned has the usual origin from the outer surface of the ilium and from the fascia which also gives origin to the ectogluteus. In Bradypus (1) and Dasypus (22) we separated a small entogluteus with some difficulty; it was inserted as usual into the front of the great trochanter. In 0r.ycteropus (85, 37) the entogluteus was present, but Humphry says that in his specimen (36) it was scarcely distinguishable from the mesogluteus. The pyriformis is generally discernible in Brudypur (1, 4,5,6) and (J‘lwloepus(8, lo), rising, as usual, from the inside of the pelvis. In I’aanzandua (14), Dasypus (22, 24), Chlamydophorus (28 rc), and Orycteropus (35, 36,37) it was also clearly made out. In all the other forms of which we have records it is described as absent or iuseparable from the mesogluteus. Qlutemprofundus and ventralis.--8 gluteus ventralis (9. quartus) was found in Tnmandrca coming from the whole ventral border of the ilium. This is probably the third gluteal described by Rapp (15). It is also mentioned iu Tatusia (25) and Chol~p.~(lo), but it does not seem to be clearly differentiated with any frequency in the order. The gluteus profundus (g. quintus) has, so far as we have been able to ascertain, never been seen in Edentates. Obturator internus tcncl Bemal1i.-Guvier states in his ‘ Lesons,’ “that, in animals which have the ischiurn anlrylosed to the sacrum, a muscle coming from the external face of the ischium takes the place of the obturator internus and gemelli.” This, in our opinion, is equivalent to saying that the intrapelvic portion of the obtiirator internus is absent. We found this arrangement existing throughout the Edentata, with the exception of the Orycteropodidce, in which the typical mammalian arrangement occurs, both gemelli and the obturator internus being recorded as present (35,36). In one case, however, Oalton (35) states that the anterior gemellus was double, one being attached as usual, the other coming from the posterior half of the sacral edge of the great sacro-sciatic foramen ; it seems possible that this so-called second gemellus may really have been a pyriformis. Obturator externus.-In the Bradypodicle this muscle was present and possessed the usual attachments in Bradypus (1,2,4) and Chokepus (8); but in a second specimen of the latter (10) it is described as double, the upper part coming from the horizontal 994 MESS=. B. c. A. wmnm AND F. G. PARSOKS ON [Dec. 19, ramus of the pubes near the acctabulum, the lower from the obturator niembraiie and horizontal ramus of the pubes. In the NywnecophugidtE the muscle is present in Tamanrlua (14,15) and Cyclothunrs (17,20). In the Dasypoditlce it is present. in Da-~ypus (22,23) and Il’utusicc (25). and was fouiid in Chliimydophorus (28n), t,hough not in the other t,wo specimens (27,28). In the Namidce it tends to fuse with the quadratus femoris (29,32,33), but in oiie case (34) it was quite distinct. In the Orycteropoclitlcc! (35,36) it is present and normal. Qucidrcctus fenioris.-This muscle is present with the usual nttachinents in Brtrtlypus (1, 4,5, S), though Humphry failed to iiiid it in his specimen (2). It is also present in Cholceps (8, 10). In Tnmtcndua (14) among the Myrrne~ophagidczit is present, but in Cyclothzcrrrs (17, 20) it is described as wanting. In the Dasiypodidm it is a strong, distinct, rounded mass in Dasypus (2%,23,24),which rises from the ischial ramus under cover of the adductor mass, and is inserted into tjhe posterior aspect of the lesser trochant’er. In 3’utu.sia (25) it was absent, while in Chlaiuydophoms (27, 28) it was present and triangular in shape. In the ilfadc~the frequency with which it becomes fused with the obturator externus has already been noticed. In the Orycteropodidrr it was absent in both Humphrp’s and Galton’s specimens (35, 36). 2’ectineus.-In the Braclypotl&?, Bradypus is remarkable for the extensive insertion which this muscle possesses, as it is attached to the whole length of the shaft of the femur (1,3). In one specimen (4) it consists of superficial and deep layers, and in that figured by Cuvier & Laurillard (6) it is divided longitudinally. In Ch0hpu.s (8,9, 10) it rises from the pectbeal tubercle, and is inserted into the upper half or somewhat less of the femur. In the 21fyrmecophayidce,the muscle is single in Myrmecophap (1 1, 12) and CycZothurus (17, 19,21). In the former animal it arose froin the brim of the pelvis opposite the ilio-pectineal eminence, and wi1s inserted into the upper two-thirds of the shaft of the femur. It was entirely supplied by t,he anterior crural nerve. In Tamandun (14) it was double, and the part which rose superficially was inserted by a small tendon just above the middle of the femur. The deeper portion was inserted above the last and in the same line with it, reaching aa high as the lesser trochanter. In another specimen (15) Happ describes the muscle as very thick. In the Dasypodiclce the muscle seems usually to be single and small. In the -Mcmidce it is also small but distinct, and is inserted just below the lesser trochanter. In one specimen of Orycteropm (35) Galton found the pectineus double; one portion was strap-shaped and passed from the ilio-pectineal eminence to the linea aspera, the other part from the same origin extended to the posterior inter- trochanteric line. In Cuvier and Laurillard’s specimen it was also double (37), whilst in the animal dissected by Humphry it was single (36). Adductor .fernoris mass.-In Bradyps (1) the adductor longus was distinct and, rising from the ilio-pect,ineal line, extended to the 1699.1 THE MYOLOGY OF THE EDENTATA. 995 femur just above the internal condyle. The rest of the mass came from the sub-pubic arch and was inserted into the middle third of the femur. The other two specimens of which we have records (2, 5) were dissected at a time when the presemimembranosus was as yet unrecognized as a separate muscle, and we suspect that the description of this muscle is included in that of the adductors. It is worlh noting that the pectineus and adductors rise as near the middle line of bhe body as they can, and wrap well round the back of the femur so as to act powerfully as external rotators. In Chologzrs (a), wher? the presemimembranosus is kept separate, the adductors closely agree wit>hthose of Bradypux. Amoug the Myrrnecophagidce, fl!jmrecq.phuguhas the mass divided into tmo planes, the more superficial of which rises from the hori- zorital ramus and anterior half of the symphysis of the pubes and is inserted into the lower half of the femur, This part is pierced by the branch of the ohturator nerve to the posterior adductor cruris (gracilis), and possibly, therefore, corresponds to the human adductores longus et brevis. The more posterior part of the maw has the same origin as the last, but lies deep to it and obtains origin from rather more of the symphysis ; it is inserted into the lower two-thirds of the femur, and probably corresponds to the human adcluctor magnus. The whole of t,he adductor mass is supplied by the obt,urator nerve. In Tamccndwr (14) two layers can also be made out, both inserted below the middle of the femur. Rapp (15) says of this animal that the adductors cannot be divided into three layers. In CycZothurus, Humphry (17), Meckel (19), and Galton (PO) were all able to distinguish three layers, but possibly one of these was the presemilnernbranosus. The Dasypodidce (Busypus 22, 23, Tutusia 25, and Clilanydopliorus 27,28, %a), all present an indivisible adductor mass. In the Mun,itZu! we were able to make out tliree distinct parts, viz., (u) to the outer border of the femur above the condyles under cover of the outer head of the gastrocnemiw (which reaches unusually high up), this is probably adductor longus ; (b) to the middle third st' the femur (adductor brevis): nnd (c) the most posterior (adductor magnus) to the lower part of the femur. In other specimens (32,33, 34) the dissectors all agree that three layers can be seen. In tile Orycteropditlce (:35,36) three layers are also described. A&uctor ci-uris (Grucilis).-The Brudypodidct; seem remarkable for the constant presence of a double adductor cruris, but the insertion differs from that which one is accust.omed to msociate with the double condition of that muscle amongst the Mammalia. In our specimeu of Brcidypus (1) the anterior adductor cruris rose from the inner part of Poupart's ligament, and was inserted int.0 the upper part of the inner silrface of the tibia. The posterior rose from the symphysis pubis, and, after reaching the knee, passed in front of the shaft of the tibia to be incierted into the fascia on the outer side of the leg below, and continuous with the insertion of the flexor cruris lateralis (biceps). Both parts were supplied by the obturator nerve. In Cuvier and Laurillard's 9!J6 MESSES. B. c. A. WINDLE AND F. G. PAESOXS ON [Dec. 19, specimen (6), the posterior muscle, instead of passing round the front of the leg as it did in our case, wrapped round the back oE the thigh, and was inserted just below the insertion of the flexor cruris lateralis. Humphry (2) also found the muscle double, but Mackintosh does not seem to have noticed this coudition in his specimen. In C?~ohpuswe have records of three specimens (8, 9, lo), all of which agree in calling the muscle double. Among the Mymtecop7~ayidot.,J4y1-m~cophap ( 11) has n double muscle, the anterior part of which rises from the ramus of the pubes internal to the ilio-pectineal eminence, and is inserted into the upper two-thirds of the cnemial crest of the tibia. The posterior part rises from the syinphysis and descending ramus of the pubes, and is inserted below the last, iuto the loner part of the cnemial crest and shaft of the tibia as low :IS the middle of the bone. In Y’amnndua (14) the miiscle rises from the sub-pubic arch, and in section would appear V-shaped, with the apex of the V directed mesially and its concavity including the adductor mass. Though there is no actual line of fission along it, yet the apex of this V clearly is equivalent to the line of separation into anterior and posterior portions in JIyrnzecophuga. The wide and strong insertion of this muscle is into more thau half of the inner side of the tibia. Rapp describes this muscle as very broad in his specimen (15). In C~yclothurus(17, 19, 20) the inuscle is single and broad. Among the Dasypod~dce the adductor cruris may be either single or double, the latter condition obtaining as far as its insertion into the fascia of the leg frcin the knee to the ankle in our specimen (22). In another specimen (24) it was single. Galton (X.) describes it as a thin muscle in Daqpus (23), but Macalister (VII.) says that it is broad in Y’utiisiu (25). In Ciilamydophorus (2i, 28 a) it is single and thin. In the Manidcz the muscle may tho be single or double. In two cases (32, 34) it fell under the former categorg, and in other two (29, 33) under the latter. In all cases the muscle is of specially large size in this family. In the Orycteropodicke (38, 36, 37) the adductor cruris is single and broad. ~emimembrunoaus.-In nll the Edentates the semimeuibranosus is a very constant muscle rieing from the tuber ischii and part of the ramus, and obtaining insertion into the upper part of the internal surface of the tibia, deep to the long internal lateral ligament. The tibia1 insertion is especially extensive in the Da8~ypodidce,and it is remarkable that in these animals the long internal lateral ligament is attached nearly as low as the middle of the tibia. Presernimembranosus.-By many observers this has not 89 yet been recognized as a separate muscle, some including it in the adductor mass, others describing it with the semimembranosus. Haughton calls it the adductor primus ; Macalister, the adductor magnus condylnidee. In our specimens of Brudypus (1) and Cho7oepue (8) it was a perfectly distinct muscle, rising from the tuber ischii, and being inserted into the femur just above the 1899.1 THE YPOLOGY OF TEE EDENTATA. 09; internal condyle and the inner head of the gastrocnemius. iWyimecophaga (11, 12) and Tamaiidua (14, IS), amongst the Nyrmecophagidce, have also perfectly distinct representatives of this muscle; and in Cyc7othurus (17, 19,20) it is also evidently present. Qalton (20) speaks of it as a second head of the semitendinosus. In the Dasypodidce, Manidce, and 0rycteropodid.a the muscle is also evidently present, but seems to be rather more closely united with the semimembranosus. Semitenclin0sus.-The chief point of intereRt about this inuscle is the varying presence of ischial, caudal, or both heads. In the Bradypodidce, Bradypws (1, 2, 4, 5, 6) and Cholaps (8, 9, 10) possess ouly thg origin from the tuberosity of the ischiurn, the insertion being as usual into the upper part of the internal surface OE the tibia. In the Xyrmecophccgidce, our specimen of Nyrmecophaga (11) was especinllv interesting *in that in it the caudal and ischial origins remained separate right down to their insertions, and so formed two distinct muscles. The first of theee (se~uitendinosus anterior) rose from the tuber ischii, and was inserted into the internal tuberosity of the tibia just below the insertion of the semimembranosus. The second (semitendinosus posterior) rose from the anterior caudal vertebrse, continuing the origin of the caudo-feuoralis, and was inserted into the tibia juet below the last. Both these iuuscles were supplied by the great sciatic nerve. In Taniandun (14, 15) only the caudal head was preseut. In Cyclothurus (17, 20, 21) the caudal head was alone present, but in mother specimen (19) the ischial head was the oiily one found. Among the Dasypodiclce, Dasypus (22, 23) and Chlumydophorus (27, 28u) had ouly the ischial head, but in Tatusia (25)both ischial aud caudal heads were present. A strong prolongation, extending to the heel, from the insertion of the muscle was noticed in Dnsypus (22). In the Mrcnidce (29, 32, 33, 34) the caudal head is always present, and in two instances (33, 34) an ischial head was also found, though it was small in the firsbmentioned specimen. In the Orycteropodidce (35, 36, 3i) the ischial head alone was found, and the mme prolongation to the heel already alluded to in Dasypus was noticed. It will thus be seen that the ischial head is alone present in the Bradypodidce and Orycteropodidce ; it is also alwayu present in the Dasypoclida, though occasionally accompanied by a caudal head. In the Nanidm the caudal head is always found, and the ischial, if present, is sub- sidiary ; while in the Myrrnecophagida either ischial or caudal, or both heads may appear. The presence of a tendinous intersection in the muscular belly of the semitendinosus has evidently been sought for by many of the dissectors of these animals, but so far we have only found its presence recorded in one specimen of Chlamydophorzcs(27) described by Macalister. Flexor cmrk luteraliv (Biceps fernoris).-Among the Bmdy- podidce, Bradypus (I, 2, 4, 5) and Cholcepzcs (S, 9, 10) have the muscle rising from the tuberosity of the ischium, but not from the caudal vertebrze. The insertion is usually into the faacia of the 998 MESSES. B. c. A. WINDLE AKD P. a. PARSOW ON [Dec. 19, leg, by which its fibres may be traced to the fibula and tibia in their upper ends. In addition to this long and constant niammalian part of the muscle, a femoral origin was found in all the specimens of which we have records. This femoral or short head always seem to have a very extensive origin from tbe shaft of the femur, and in our specimen of Cholops it was om of the largest muscles in the hind limb. Its insertion is usually lower than t,hat of the long head, and may be connected with the terido Achillis ; moreover it is not always fused even at its insertion with the long head. Among the Jiyrmecophcigicle the long head, in our specimen of H!pnecoyhago (ll), rose from the ischiuni and was inserted into the fascia over the upper pirt of the fibula. The short head was not so large as in the Brutl!/podidce, and inut,ead of rising from the shaft of the femur, it took origin from the insertion of the caudo-femoralis, it then crossed deeply to the long head, forming an X with it, to be inserted into the gastrocnemius at the point where the tendo Achillis commenced to exist’. In our specimen of ITama)iduii (~14)there was no femoral head, neither do Rapp nor C‘uvier arid Laurillard mention one in theirs (15, 16), though Macalister states that it is to be found in this animal. In ~yc~othurus(17, 18, l9,20,21) all the authorities are agreed as to the presence of the femoral head ; aud Meckel (19) says the,t it is inserted into the outer malleolus, an assertion with which Cuvier and Laurillard agree. In the Dnsypodida we find no indication of a femoral head in Das?ypm (22, 23, 24), ITutusiu (25), or Chluny- ‘dop?iorus (27, 28, 28 u). In the Moniclce a femoral head was found (29, 32, 33), while in the Orycteropodiclce no femoral head is recorded by any of the observers from whom we quote. Tenuissinaus (Bicipiti accessorius).--For reasons which will appear later, we wish to coritrast the presence or absence ol‘ this muscle with the condition of the long head of t,he last deacribed muscle. First, however, we call attention to the fact that in most mammals the tenuissimus arises from the anterior sacral vertebrae under cover of the ecto-gluteus or caudo-t’emoralis and passes down, as a narrow ribbon-like muscle, to be inserted with the lowest fibres of the flexor cruris lateralis, i. e.. with those fibres which most nearly attain the ankle. As these fibres are often inserted into the tendo AchilliR or gastrocnemius, it is clear that this will not be an uncommon insertion for the tenuissimus. In the Bradypodidce we noticed that the femoral head of the flexor cruris lateralis is always present. In uo specimen of Brotlypus (1, 9, 4) or ChoZepus (8, 9, 10) is the presence of a tenuissimus mentioned, while in our specimens (1, 8) we specially looked for it and can definitely state that it was absent. Amongst the Myrmecophagidce, ,Wpnecopluz/a (11, 1%)has a short head for the flexor cruris lateralis, which, instead of rising From the feinur, come6 from the surface of the caudo-femoralis. This animal tias no tenuiasimna. Our TumancZua (14) had a typical uammalian

1 MU.Oouvreur and Bertaillon describe an identical arrangement in their apecimen (12). 1899.1 THE XYOLOGP OF THE EDENTAYA. 999 teniiissimus, but no short head of flexor cruris lutentlis, and the same condition existed in Cuvier and Laurillard's specimen (16). All the specimens of Cyclothurua hnd femoral heads for flexor cruris lateralis, but in none of them is a tenuissimus recorded. In the Dasypodida? a short head of flexor cruris laberalis is never found, buh in Dasypus (22, 23) and Tatusia (25) the tenuissimus wss present, though not in Chlamydophorus (2'7, 28, 28 a). The Bunide have a femoral head to the flexor cruris lateralis. In two cases (29, 33) there was no tenuissimus, in the others its presence is not mentioned. The Oryeteropoclida?have no femoral head to the flexor cruris lateralis, but Galton (35) evidently found a tenuissimus in his specimen, though he calls it the second part of the semimembranosus. The fact that these two muscles, the short or femoral head of the flexor cruris 1at.eralis and the tenuissimus, never in our records and dissections have been found to co-exist, made us suspect that the two muscles uright be identical. This suspicion was strengthened when we noticed how similar the insertion of the two muscles was, the origin alone differing. The condition met with in our specimen of Myrmeeo- phaya (ll),as well as iu that of MM. Couvreur aud Bertaillon ( I2), makes us thiuk that the caudo-femoralis acts as R '' muscle- slide," down which the origin of the tenuissimus slips until it reaches the shaft of the femur, when it becomes a short head of t,he flexor cruris lateralis. Dues this explain the morphology of the femoral head of the human biceps femoris'? We are not prepared to commit ourselves to a definite statement of opinion until we have examined into the matter more fully, as we propose to do at a subsequent time, but so far as the evidence before us goes we are inclined to answer Ihe question in the affirmative. Quadriceps extensor orurk- In all our dissections of Edentates we found the rectus rising bg one broad head from the dorsal and cephalic margins of the acetabulum, aud fusing more or less with the capsule of the hip-joint. Most of the other writers who give any details of the origin of this muscle speak of it as eiugle-headed. Humphry, however, states that in Orycteropzco (36) its origin is as in Man; though Gnkon, in his description of the same animal (35), says that the rectus rises from the superior and posterior margins of the acetabulum, not mentioning the fact that there are two distinct heads. Our experience of myology makes us think that in the Edentata, as in most mammals, there is one broad head, and that in Man this head becomes differentiated into two as an adaptat,ion to the erect position. In other wordR, we believe that in the Edentates both the acetabular and iliac heads of the rectus are present, but that they are fused or not yet differentiated. Of the deeper parts of the quadriceps there is little to say beyond the fact that the quadricipitis lateralis (vastua externus) is always very large in proportion to the mesialis (vastus internus). In Chlamydophorus (27) Macalister found an extra head to the lateralis rising from t,he ilium beneath the origin of the super- ficialis (rectus femoris). This was not seen in the other two 1000 MESSRS. B. c. A. WINDLE AND F. G. PALI~SONSON [Dec.19, specimens (28,28 a). It is elwags difficult to separate the mesialis from the profundus (crureus). Tibialis anticus.-Among the Bradypodidce it is rather difficult to determine the line of demarcation between this muscle and the extensor hallucis, more especially as in both kinds of Sloths the hallux is aborted. In four specimens of Brccdypus (1, 4, 5, 6) the muscle rose from the anterior surfaces of the tibia and fibula, and was inserted into the rudimentary first metatarsal boue. In Chohpus (8, 9, 10) the origin wat? the same, but the tendon, instead of being inserted into the metahrsnl, ~indsround the ankle to the plantar surface of the foot, and is inserted into the long flexor tendons. In Humphry’s specimen (9) and in our OWTI (8) it divided into threeEdips which joined other three slips from the flexor longus digitorum (tibialit?? fibularis ?), while in Mackintosh’s animal (10) it onlg joined the flexor of the middle toe. Among the Myrniecophagidp, the muscle is single in Mymtecophagn (11) and rises from the upper $ of the tibia to be inserted into the entocuneiform and slightly into the base of the first metatarsal. In Tamccndua (14) there were tibia1 and fibular origins, and the insertion was into a sesamoid bone on the inner side of the nnvicular. In Cyclothurus (17,19,20,21) there were tibial and fibular origins, and the insertion in all cases was into the entocuneiform. Among the Dasypodidce, Dasypus (22, 23) has tibial and fibular origiua, and an insertion into the entocuneiform, but in Tcitusia (25) and Chlamydophorw (27) only the tibinl origin was found. In this family the muscle is particularly large. In the Manirlce (29,32, 33) there are tibial and fibular origins, and the insertion is iuto the entocuneiform and first metatarsal. In the Oryateropodidce (35, 36, 37) it rises from the upper half of the tibia and from the fibula. In Humphry’s and Galton’s specinlens (35, 36) its tendon divides, and is inserted into the first metatarsal and entocuneiform. Extensor prcpriua halluck--In the Brcic~ypodirLzwe found no separate representative of this muscle in our specimen (1). Humphry (2) and Meckel (4), however, found a small muscle rising from the lower end of the fibula and passing to the rudimentary first metatarsal, a condition also figured by Cuvier and Laurillard (6). In cholapus (8, 9, 10) a similar conditioii was observed. Among the Myrmecophailidce, Myrnzecophaga (11) and Tanzandua (14, 15) have the muscle rising from the lower end of the fibula and inserted into (11, 12, 14) the terminal phalanx of the hallux, but in Rapp’s specimen (15) it also went to the second toe. In Cyclothurus the extensor proprius hallucis is noticed by both Humphry and Galton; but the loug extensor muscles of the foot nere evidently imperfectly differentiated, for Humphrg (IT.)found the muscle joining the tendon of the tibialis anticus, whilst in Galton’s specimen (20) it united nit11 that of the extensor longus digitorum. In fhe Daypodida, Dasypue (22, 23), Tatusia (25), and Chlamydophorzrs (27, 28 a), the muscle always rises from the lower part of the fibula and is 1899.1 THE MYOLOGY OF TlIE EUEPYTATA. 1001 inserted into the hallux. In the Mriniclce (29, 82) the same description applies, but in one case (32) a small slip was given to the second toe as well as to the first. In the Orycteroyodida (35, 36) the muscle rises rather higher up from the fibula, and in both cases had a slip of communication with the extensor longus digitor um. Extensor longus digitorurn.-In the Brudypodich, Bradypus (1, 2, 4) and C/ioZu?pus (8, 9) have the usual origin froni just above the external condyle of the femur ; but in Meckel's specimen of Brnd,ypus (4) tibia1 and fibular origins were also met with. In Bradypus the insertion is never into the toes. In our specimen (1) the tendon divided into two slips, which were inserted into the bases of the innermost and outermost of the three developed metatarsal bones ; it will be observed that the outermost of these is the same muscle 8s the humnu peroneus tertiua. Cuvier and Laurillard's figure agrees very closely with our specimen, though they call this part of the muscle the peroueus brevis (6). In Humphry's specimen (2) the whole muscle wa9 apparently inserted into the outermost metatarsal, while in Meckel's (4) it went to the innermost. In Cholcepus the insertion may be into the dorsum OF all three toes (8, lo), or only into the second arid third (9). Among the Myrmewphayidct., the origin is condylar in iMyrmeeopAagci (1 1) and l'anmtdua (14), but iii C'yelothurus it rises from the tibia only (17, 19), or the tibia aud the tibuh (20). The insertion is usually into the four outer toes, though in Cuvier aud Laurillard's figure (21) it appears as if slips only went to the third and fourth toes, a.nd in Couvreur and Bertaillon's specimen of Myrmecophaga into the three outer toes. In the Dasypoclich a femoral origin is never found, the muscle rising from the upper part of the fibula only, and being inserted into the four outer toes. This description applies to Dmypus (SY), Tatueia (25), and Chlamydophrus (27, %a). In one specimen of Dcwypus, however, (22) a slip was sent to the hallux as well as the other four toes; otherwise this specimen agreed with the rest. In the Manidm the chief origin id from the tibia and fibula (29, 32, 33, 34), but in some specimens (32, 34) a feeble femoral origin is found. The insertion is into the outer three (29,83) or foiir (32) toes. In the Orycteropodidm (35, 36) there is the normal femoral origin, and the insertion into the four outer toes (35, 36, 37). We have already drawn attention to the connection with the extensor proprius ballucis. F.&nsor brevis digitorun8.-ln the Bt-icclypodida this muscle usually rises from the lower end of the fibula (1,4,6), but sometimes from the tarsal bones (presumably cdcaneum) (2). It generally sends slips to all three toes (1,2), but in one case (4) it only sent a tendon to the inner toe. In Cholcepus (8,9,10) the muscle always rises from the tarsus and is inserted into the long tendons of all the toes. In the Mymnecophagidce the muscle always rises from the tarsus. In Mpnecophqa (11) and Tamandm (14,15) it sends slips to all five toes, but in Cyclothurw (17,19,20, 21) only to the four 1002 MESSRS. B. c. A. WIXDLE AND E. G. PARSOXS ON [Dec. 19, outer. In the Dmyporlirkr the origin is always tarsal (calcaneal with occasional additional slips). In Dasypus (22,24) the inaertion was into the second, third, and fourth digits, though in another Rpecimen (23) into the first, second, and third. In Z'a'atusia (25) its insertion was into the second, third, and fourth toes, while in Ch1amydopJmw.s it was inserted into the four outer toes. In the Manidre the fibular origin sometimes occurs as in Bradypus, at least this was the case in OUP specimen (29), though in Hurnphry's (32) it came from the tarsus. The insertion may be into the four outer toes (32, and apparently 34), or into all of them (29). In the Oycteropodiclre (35,36) the inuscle rises from the calcaneum, and is inserted in one (37) into the three inner toes, in another (36) into the three middle toes. Peroneus lonps.-In the Braclypodiclm this muscle was found in three specimeus of Bradypus (2, 4, 5) rising from the condyle of the femur and upper part of the fibula, its insertion in all cases being into the base of the outermost metatarsal bone. In our specimen of Brodypus (1) we failed to find any peroneus longus at all, and it is absent in Cuvier and Laurillard's plate (6). In Cho7mpus (8, 9, 10) the muscle only rises from the upper part of the fibula aid is inserted into the base of the outermost metatarsal bone ; so that in the family of the Bradypcdidre we think we are able definitely to state that the perooeus longus tendou never runs across the sole of the foot. Among the iVlyrmecrvphugidre, the liluecle rises from the tibia and fibnla in illyrmecophaga (ll), from the fibula aud semilunar cartilage in I'arnadua (14), and from the fibula only in G'yclothurwr (17). In no member of this family has a femoral head been found. The tendon always runs across the sole and is inserted into the innermost metatarsal bone or bones, into the entocuneiform, or (20) into the naviciilar. Among the Dnypodide the muscle has pntellar and fibular origins in Dasypus (22)' and Tatwria (2.7), but only fibular in Chlamytlo- phorus (27, 28). In another speciiner (28 a) of this animal there was an additional origin from the patella. In all these animals the tendon passes across the sole of the foot. In the Mznidre (29,32, 33) the muscle rises from the fibula and passes across the sole of the foot. In the Orycteropodidce there iR no definite femoral origin, but the peroneus longus rises from the external lateral ligament and semilunar cartilage as well as froin the tipper part of the fibula. The tendon passes across the sole to the first meta- larsal bone. In those animals, sitch as the Brrdypodidce and Dtqpodidre, which have a femoral origin for the peroneus longus, there is a distinct external lateral ligament in addition. This fact is of considerable importance in arriving at a conclusion with regard to the morphology of that ligament. I'eronewr be&.-In the Bradypodidce this muscle was present in three specimens (2, 4, B), but was absent in the fourth (1). It rose from the lower part of the fibula and was inserted into the

This wan aka Been in another dissection at the Royal Oollege of Surgeons. 1899.1 THE MYOLOGY OF THE EDENTATA. 1003 base of the outermost metatarsal. In Cholapus, Humpbry (9) found the muscle as iu Braflypus, but in another specimen (10) as well as in our own (8) it was not seen. In the hQraecophagidce the muscle is present in Myrmewphaga (11, 12), Tamundzsa (14, l6), and Cyclothurus (19). In Dasypus (22 and another) among the Dasypodidce the muscle rose from the outer side of the fibula and from the external condyle. It is also present in CJdamydo- phorus (27, 28, 48 a). In the Manib (29, 32,33) and Oryctmo- potfib(35, 36) the muscle is present and normal. Peronei tertius, quarti et yuivtti diyitt.-It is difficult, in reviewing the literature, to feel perfectly certain as to the identity of these muscles. Macalister (VII.) say8 “there is no trace whatever of a true peroneus tertius in any of the species examined, Chla mydophorus, Tatusia, Dmypus, Bradypus, Cholapus, Orycteropus, Pholidotus, C?yclothurvs,or Taniandua. The m uscles described a3 such by authors are, in reality, peronei quinti.” We have little doubt that Macalister is right in inany cases, but we have Rome doubt as to whether the statement applies accurately to all, since we have seen and nlready described a typical peroneus tertius in Bmrlypus. The pointrs w hirh. in our opinion, enable a right decision to be arrived at in the case oE a doubtful peroneus tortius or p. quinti digiti are (a) its nerve-supply ; (a) its relation to the ankle, whether anterior or posterior; (e) its insertion into the extensor tendon or into the metatarsal bone of a digit. Unfortunately many writers fail to give derails on some of these points or on all of them, and for this reason any generalizations which we may venture to offer must be taken with a reservation. In the Bradyporlida a well-marked peroneus tertius, inserted into the base of _the metatarsal bone, passing in front of the ankle, and supplied by the anterior crural nerve, was found. Meckel (XI.) and Mackintosh (XVI.) mention a peroneus quinti, which the latter savs is inserted into the tuberosity of the outer metatarsal. In Cholcepus (9, 10) a peroneus tertius is described, Hnmphry (9) stating that it came from the front of the fibula and was inserted into the bases of the tvro outer metatarsal bones. The relation to the ankle is not mentioned, but frm what we learn as to its origin and insertion we are inclined to agree as to the correctness of this denomination. It should be borne in mind that both Cholczpus and Bradypus have only three toes on the hind foot; so that a peroneus quinti is not, so far as we have learnt the lessons of mammalian myology, a muscle with which one would expect to meet. In the Nyrmmophayidce the outer or fifth toe is always developed, and we find that both Myrmewphaga (11) and Tamartdua (14)have a peroneus quinti digiti, but we have not been able to satisfy ourselves as to the absence or presence of this muscle in C7yclothurus. In the Dcisypodidce the peroneus quinti is present in Dmypus (22 and another) and ChlamyclopAoriis (27, 28 a). In the Manida (29, 32, 33) we find no account of a peroneus quarti or quinti. In the Orycteropoclidm (36, 36) all four peronei were present, viz., longus, brevis, quarti et quinti digitorum. 1004 MXSSRS. B. c. A. WINULE ANU 8. o. PARSONS ON [Dec. 19, G'astrooiemius.-This muscle, amongst the Edentata, has, with certain exceptions, the typical mammalian arrangement. In the Brudypodih rhe two heads do not unite until they reach the calcaneum, but they are not twisted in such a way that the inner becomes superficial and then external (cf. Journal of Anat. & Phys. vol. xxviii. p. 414). This is true of Bradypua (I, 2) and Cholapw (8, 9, 10). In the Manicla (29, 32, 33) the external bead is very large and rises a long way up the abaft of the femur-a condition far exceeding anything which we have hitherto observed in any other mammal. It is interesting to notice that nearly all observers have recorded the absence of fabellae except in the Orycteropoclidre. So1eu.s.-Among the Brudypodidce the soleiis often rises quite low down on the fibula in Bradypus, in which animal it arose in one case (5) from the middle, and in another (1) from the lower third of the bone. It is inserted into the calcaneum without joining the tendo Achillis (1, 2). In Cholcepus the chief insertion is also into the calcaneum in front of the tendo Achillis, but Humphry noticed that some of its fibres were continuous with those of the accessoriris. Among the illyrmecophugiah its origin was chiefly from the fascia over the deep flexor muscles in iL%y-niecophri.yu (11) '. In the last-mentioned animal, in l'umanduu (14), and in C~yclothurtis(19.20) it is inserted as in the Rradypodidre. In the Dasypoclidce it seems usually to join the outer head of the gastrocnemius, but our information is not very clear upon this point. In the MunidfF?(29, 32) and Orycteropodiclre (35, 36) its insertion is as in the Braclypodida and Myrmecophayidce. It will thus be seen that the Edentata as an order are characterized by the separate insertion of the soleus and the absence or incompleteness of the tendo Achillis. Plantaiis.-This muscle is liable to a good deal of variation in the Edentata, and is likely to be Confused, on the one hand, with the femoral head of the flexor cruris lateralis (biceps), and, on the other, with the flexor tibialis and fibularis. In the Bmdypodidoe the muscle was absent in one specimen (l), but in three others (2, 4,5) it was present as a very large muscle which rose from above the external condyle of the femur, and was inserted into the long flexor tendons in the sole of the foot. It is described by some writers as an extra head of the long flexors of the toes. In C'holcepus (8, 9, 10) the muscle is absent, but the condition in this form will be again referred to under the head of the tibialis posticus. Among the Byraeeophctgicla?it is present in Myrnoecophuga (11) and has the usual mammalian insertion into the plantar fascia. In Cyclothz~rus(17, 19, 20) it is also present, and is inserted into the elongated ossicle on the tibia1 side of the foot. This insertion is intere5ting when compared with that which is found iu the band of €'eclete.s (cf. Proc. Zool. SOC.1898, p. 867), in which the palmaris longus, the serial homologue of the plantaris, is inserted into the

In (12) it had the generalized mamm~lisnorigin from the back of the head of the fibula 1899.1 TEE MYOLOGY OF TEE EDENTATA. 1006 radial ossicle, or so-called pre-pollex. In Tarnundua (14, 15) the muscle is absent. Among the Dasypodid~zthe plantaris rites frnm the ridge above the external condyle in Dasypus (22, 23, and is continued into the sole of the foot, where it flattens out and sends slips to three or four of the digits, which slips are perforated by the tendons of the flexor longus digitorurn. In Y’atusia (25) there is no separate plantaris. In Chlumydophorus (27, 28, 28 a) the tendon passes into the sole of the foot, where it divides iuto four slips. In the Manidm (29, 32, 33, 34) there is no separate plantaris, it is probably fused with the very large external head of the gastrocnemius. In the Orycteropoclirlm (35, 36, 37) the generalized mammalian arrangement is found ; the plantaris passes under the tuherovity of the calcaneum, and forins a fibrous flexor brevis digitorum for the four outer digits. Flexor breuiv cZigitomciri.-In the BwdypocZidce (1, 2, 5, 9, lo), Myrmecophayidce (11,14, 15, 17), and Mmiicke (29,36), the muscle rises from the posterior part of the lower surface of the calcaneum, and has no conuection with the plantaris when that muscle is present. The insertion is into the three middle or four outer digits. The tendons are usually inserted into those of the flexor longus, instead of being perforhted by the latter. The information, however, as to the manner of ending of these tendons is very scanty. In the Dasypodidce (22, 23, 27, 28) and Oycteropodich (35, 36, 37) the flexor brevis is a continuation of the plantaris, as in most generalized mammals. Flexores tibiulis et jbu1aris.-In the Bradypodide these two tnuscles are difficult to distinguish: in any case they coalesce before reaching the ankle, and then divide into three tendons, which pass to the second, third, and fourth toes respectively. This applies to Brudypus (1, 2, 5) and ChoZ~pus(8, 9, 10). In some records of Bmdypus (2,5) a femoral head is also recorded, but a consideration of the conditions has decided us to regard this as a plantaris. Among the Myrmecophrigidce the two muscles are practically inseparable ; they form a single flat tendon, 1% hich in Myrmecophap (11) and Tmandua (15) has a sesamoid body where it passes into the sole of the foot. In Tamunduu (14, 15) tendons pake to all five toes, but in Cyclothums (17, 19, 20) there is no slip for the hallux. In the Unsypodidce (22, 23, 25) the tibia1 and fibular heads unite in the lower part of the leg and are inserted into a very large sesamoid bone in the sole ot the foot, which is held in place by a fibrous band from the caleaneum, the equivalent of the nccessorius. From the front of the fiesamoid bone five tendons pass to the terminal phalanges of the fire digits. In Chlumyclophorus (27) the sessmoid hone nas replaced by 8 cartilaginous nodule. In the Manidce (22, 2H, 25) the two muscles are much more distinct, the flexor fibulnris forming a very large tendon into the inner side of which the small tendon of the flexor tibidis is inserted. There are tendon8 for the four outer toes, but none for the hallux. In the Oyderopodidce (35, 36) the flexores tibialis et fibularis fuse in the leg and from the tenrlnn PEOC.ZOOS. SOC.-l899, SO.LXV. 65 1006 MESBRS. B. c. A. WIRDLE AND F. G. PARSONS OR [Dec. 19, slips are given off to all five toes. The tendon is joined in the sole by a slip from the tibialis posticus accessorius. Popliteu8.-In the Bradypodidce (1, 2, 4, 8, 9, lo), Myrmeco- phayidce (12, 14, 17, 19, 20), and Manidm (29, 32, 33, 34) this muscle is large, occupying the upper half of the tibia ; its origin is from the outer condyle, and a sesamoid cartilage or bone is developed in its tendon. In the Orycleropodiclce (35, 36) its insertion is singularly extensive, but no sesamoid is mentioned as having been observed. In Dasypus (22,23) two tendons of origin were noticed, the anterior and larger coming from the outer side of the coudyle and the external semiluuar cartilage, the posterior and smaller from the posterior part of the coudyle. In Tutusic1 (25) and Chlamydophorw (27, 28, %a) this double origin was not seen. In no member of the Dasypodidn: was a sesamoid cartilage observed. The Edentata are characterized, as an order, by the large size of their popliteus. Tibialis posticus. --In the Bratlypodide this muscle is single and small; it usually rises from the middle or lower part of the shaft of the tibia, and is inserted into the eutocuneiform bone. In the other fainilies (i+fyrmecophagidce,Das.ypoclidce, ManicEte, and Or?ycteropodicZce) the muscle is usually double, the more external being inserted into the navicular or sometimes the entocuneiform, whilst the other is often larger and passes to the tibia1 olisicle on the inner side of the foot. This additional tibialis posticus is called by Galton tibialis posticus secundus, and by Hyrtl tibialis posticus accessorius. Awssorius.--In the Bradypodidm this muscle is always well developed. Humphry states that in Cholupa (9) and Bradypug (2) it was continuous with the tendou of the soleus. We have carefully dissected both these auimals, but found no conuection whatever between the two muscles, nor have other observera described it. In the Myrrnecophngidce (11, 14, 15, 17, 20) and Manic& (29, 33, 33) it rises as usual from the lower surface of the calcaneum, and is inserted into the conjoined deep flexors in the sole. In Myrrnecophugu (11) it is especially large. In (12) it gives off a special slip to the hallux tendon. ln the Dasypodida (22, 23, 25, 27) there is no muscular accessoriue, but its place is taken by the fibrous band which binds the great sesamoid boue of the sole of the foot to the under surface of the calcaueom. In the Orycteropodidce (35, 36) the muscle is also replaced by tendinous bands from the calcaneum to the outer side of the long flexor tendons. Lumbrioales.-In the Bradypodidce we found no lumbricales in our specimens of Bradypa (1) and Chohpw (8), and their presence is not mentioned by other observers. In Tanaa12duo (14) and Cyclothurus (17), among the Mymnecophayide, they were not seen, but Galton in another specimen of Cyclotlww found three. In Myrmecophaga (12) there were four. The Daeypodida (22, 25,27) always have at least four lumbricales, and in one specimen of Dasypus (23) Galton describes seven. In the Manicla (29, 32, 34) 1899.1 THIP XYOLOGP OB TIIE EDENTATA. 1007 three lumbricales are usually found, while in the 0ryctet.opoclicla (35, 36) there are four. IWwculi breves pec1is.-The abductores hnllmis et minimi digiti are of course absent in the Bradypodida, but in all the other families they are present. In the Orycteropodidm, however, the arbductor hallucis is replaced by fibrous tissue. The flexor brevis hallucis is present in all Edentates except the Braclypotlidce. In the OrycteropodicZm (35, 36) its absence is noted, but Humphry (36) states that it is replaced by fibrous tissue. It is present in all other Edentata, including Cyclothurus. The abductor minimi digiti is present in the iUyrrnecophagidm, Dasypoditlcz, Manida, and Oycteropodidce; it usually rises from the base of the fifth metatarsal bone, instead of from the calcaneum as in most mammals. The superficial layer of deep muscles, i. e. those which lie superficial to the deep branch of the external plantar nerve and are usually called adductors, are wanting in the Bradypodidre and Manitlm. In the following animals adductores hallucis et minimi digiti were found :-Cyclot?izwus (20), Dusypus (22), Tatusin (25), and Oyc- teropus (35, 36). Tarnnndua (14) in addition had an adductor indicis. The interossei are usually present as paired flexores breves, but in the Dasypodida and .Manidre a differentiation into dorsal and plantar groups is noticed ; their exact arrangement, however, differs in different specimens of the same animal.

Abdominal Mzcscles. kJerratug dorsalis (S. posticus).-The two portions of this muscle, thoracis and lumbalis, seem to be but little developed in the Eden- tata, and this statement particularly applies to the former of the two. Amongst the Braclypodidm the muscle is represented only by a fibrous sheet in Bradypus (l), whilst in Cholqm (lo) the thoracic portion is absent, but the lumbar, attached to the lower ribs, is present. Amongst the Myrmecophagidce n fibrous representative of the muscle was fouud in Tamandua (14); it is not figured, however, in Cuvier and Laurillnrd's plate of the same species (16). Dasypus amongst the Dasypodicla does not seem usually to possess any representative of this muscle, for none was found in (22) ; Gaiton (23) does not mention one, nor is any figured by Cuvier and Laurillard (24). Indeed we coulil have asserted that there is none but for the fact that Macalker (VI.) seems to have found a representative of the muscle in one case. In l'rctusia (25) a very feeble thoracis was found, and a lumbalis attached to the lower four ribs. Chhmyclophorus (2,7, 28) had a similarly arranged lumbalis but no thoracis. In Mnnis (89) a fibrous representative was alone found. In Oi*yctcropus (37) Cuvier and Laurillard figure a continuous sheet extending from the third to the tenth dorsal vertebra which appears to be an unusually well-developed thoracis. Batus ventraZis.-This muscle in the majority of forms reaches as high a point of attachment as the first rib, though sonietimes 65" 1008 YESBBB. B. C. A. WINDLE AND F. 0. PARSONS OR [Dec. 19, it falls short of it. Amongst the Bradypodih this is particularly the awe. In Bradpus (4, 5) it reached as high as the fifth rib, and in the latter had four tendinous inscriptions ; in another (6) it was attached to the ribs from the fifth to the eighth, and in a fourth (7) from the tbird to the seventh. In this last two inscriptions were noted. Amongst the iWyrmerophagidce, Taniaiidua (14) has the attachment extending up to the firsb rib, and three indistinct intersections were seen. In Cyclothurus (19) the attachment was to the eight upper ribs with three tendinous intersections, whilst in another case (YO) only the second to the sixth ribs gave origin to the mosele. Amongst the Dasypodicle the first rib was the point of attachment in Duypus (22), ll'atusia (26), and Chlamydophorus (27, 28). Iu the laat-named form one tendinous intersection was alone noticed. In Manis (29) the attachment was from the first to the fifth rib, and in Oryctevopus (35) the highest rib was also attained. We are not inclined to place much reliance on the number of intersections, as in small mammals these are often very indistinct. Obliqui abdominis wternus et intcrnus.-The former of these muscles was in Brcidypus (4) attached to the lower ribs, in (7) to the ribs from the 6th to the 14th, and in (6) to the last six. The condition of the internal oblique is only mentioned in the second of these cases, where it is said to have gone to the last rib. Tamandua (14) amongst the Myrmecop?iugidce has the first muscle arising from the fourth rib backward and the second from the last five ribs. Amongst the Dasypodich, in Dasyps (22, 23) the external oblique risea from the ribs from the third to the seventh, whilst in another specimen (24) it is stated to have arisen as high up as the first. In Chlamydophorus (27) it rose from the lower six ribs. In Maniu (29) the external attained the first rib and the internal was attached to the last six. In Orycteropus (35) the same high attachment of the external oblique is recorded. P~otcsma:pzus.-This muscle seems to be always present, though frequently very closely associated with the iliacus, a condition which we think accounts for the fact that in one or two cases it bas beer1 reported as absent. It is large and more or less connected with the iliacus in the BratlypocEidce, @adypus (4, 5) and CJ~ohpus(10). Amongst the illymaecophngide its presence is noted iu Myrmecophap (12), Tamandua (14), and C"@thurus (17). In none of these cases is any spacid association with the iliacus mentioned. Amongst the Dmypotli&e it arose from the sides of all the lumbar vertebra in Daypus (22, 23). In Tutu& (25) it is described :LS inseparable from the ilincus, a condition phich we think must also have obtained in G'J~lrtnayilophorus(28), where it is stated not to have been found. In another specimen of the same (27) it is said to have been very small, though separate from the iliacus at its origin, where it lay as a thin strip along the side of the ilium. In the account of a third specimen (28a) no mention is made ot this muscle, though the presence of the psom parvus is alluded to. It is therefore probable that here also it was inseparable from the 189U.I TIIE MTOLOGP OF THE EDBINTATI. 1009 iliacus. In Mania (29, 31, 32) the muscle arose from the trans- Terse processes of the three lower lumbar vertebra. It is also described as present in Orycteropus (35, 36). Psoasparvus.-This muscle is generally present, and when present always inserted into the ilio-pectiuesl tubercle as usual. In 1Ymdypu.q (1, 2, 4, 7) it seems always to be a feeble muscle, and iisually to come only from the first lumbar vertebra. In Chohpus (10) it is described as present. Amongst the M?yrmecophagidct? it i4 noted as having been present in My)*mecophaya (12), Tamandua (14), and Cycl0thul.w (17), and in the second of these it arose from the laid dorsal and first two lumbar vertebrae. Amongst tho Drrsypodicke it arose iu Daypue (2!2,23) from the last dorsal and first two lumbar vertebrae; it was present, though mall, in Tatusin (65),and is also noted as present in Chlainyclopkorus (27,28 n). In the former of these it is described as a strong muscle. In Manis it is always present and stroug, generally arising from five lumbar vertebra (31,32,33,34). In Orwterouus (36) it aroye from the bodies of the lumbar vertebrae and ;so dkhtli frbm the last rib, aud its presence is also noted in (35).\I Iliacus.-As has been mentioned above, this muscle is often inore or less fused with the psoas magnus. It also not infrequently obtains a much larger insertion into the femur than is the case in human myology. This is the case in Cyclothurus (li),Chlamy- dophorus (28 a), the proximal third of the femur, Manis (32), more than half the femur, Oyycteropus (35, 36), half the femur. In another animal of the same species figured by Cuvier and Lauril- lord (37)the ilincus was divided into two bundles, an external and au internal.

iUyologicnl C7iaractet&tics of the uario us Families of Edentata. BRAUYPODIDZ. 1. The dorsal part of the panniculus is feebly marked, and there is no sterno-facialis or sphincter colli. 2. The sterno-maxillaris is absent. 3. The sterno-glossus is absent. . 4. The rectus tboracis lateralis is present. 5. The splenius colli is present in Bradypus. 6. The rhomboid has no occipital origin in Bradypus, though there is one in Choloepw. 7. The subclavius is large. 8. The clavicular deltoid foruis a cephalo-humeralis in B*adypus, not in Cholmpus. 9. Bradypus has the middle part on!y oE the coraco-brachialis, Choloppus the short and long portious. 10. Rradypus has humeral and glenoid heads to the flexor longue cubiti (biceps), Cholmpus the glenoid head only. 11. The extensor cubiti (triceps) has only one scapiilar head. 12. The flexor carpi radialis does not reach the metacarpus. 1010 YEWRS. B. C. A. WINDLE AND F. ct. P?LESOh’8 ON [Dec. 19, 13. The palmaris longus is distinct. 14. The flexor sublimis digitorum is absent in Brudypur, feeble in Cholqw. 15. The flexor profundus digitorum has no palmar sesamoid. 16. The pronator quadratus is very small. 17. The supinator longus is usually double. 18. The extensor minimi digiti becomes an extensor brevis digitorurn and often rises from the carpns or metacarpus. 19. The extensor ossis metacarpi pollicis i8 inserted into the trapezium. 20. The supinator brevis is inserted iuto the upper third of the radius. Sometimes it is bilaminar in Cholaps. 21. The cauclo-fenioralis is fused with ectogluteus. 22. Pectinatis is sometimes double in Braclyps. 23. The adductor cruris (gracilij) is double. 24. The semiteridiuosus has only an ischial head. 25. The flexor cruris lateralis (biceps) always has the short femoral head. 26. The tenuissimus is absent. 27. The extensor longus digitorum has a condylar origin and is inserted into the metacarpal bones. 28. The extensor brevis digitorurn has a fibular origin in Bradypus, tarsal in C~iolwp,~a.s. 29. The peroneus longus never runs across the sole, sometimes it has a femoral origin, sometimes the muscle is absent altogether. 30. The peroneus brevis is sometimes absent. 31. The gastrocnemius has no fabella. 32. The plantaris is often absent. 3.The flexor brevis digitorum rises from the calcnneum. 34. The tibialis poeticus is single. 35. The acceesorius pedis is preseut.

MYIZXECOPHAQIDX. 1. The sterno-maxiIlnris is present. 2. The sterno-glossus is present. 3. The rectus thoracis lateralis is present. 4. The splenius colli is absent or very slightly developed. 5. The rhomboid has no occipital origin. 6. The subclavius is absent. i. When there is a corsco-brachialis it is only the lung head which is present. 8. The flexor brevis cabiti (hrachialis anticus) does not rise from the neck of the humerus. 9. The extensor cubiti (triceps) has more than one scapular head. 10. The flexor carpi radialis reaches the metacarpus. 11. The palmaris longus is usually indistinct or absent. 12. The flexor sublimis digitorum has only one tendon. 1899.1 THE XYOLOBY OF THB BDENT-4TA. loll 13. The flexor profundus digitorurn bas an extra head from the extentlor cubiti. There is no palmsr sesamoid in it. 14. The pronator quadratus extends the whole length of the forearm. 15. The slipinator longus is usually double. 16. The supinator brevis 4 inserted into the lower part of the radius. 17. The caudo-femoralis is usually distinct. 18. The ndductor cruris (gracilis) is sometimes double. 19. The semitendinosus may be double, the ischial and caudal heads remaining separate, or either one may alone be preaent. 20. The ,femord head of the flexor cruris lateralis (biceps). -_ is usually present. 21. The tenuissimus is present when the last-mentioned head is absent. 22. The extensor longus digitorum is usually femoral in origin, but sometimes tibial. 23. The extensor brevis digitorum always has a tarsal origin. 24. The pzroneus longus has no femoral head and its tendon crosses the sole. 25. The gastrocnemius has no fabellre. 26. The flexor brevis digitorum rises from the calcaneum. 27. The tibialis posticus is double. 25. The accessorius is present. DASYPODIDB. 1. The panniculus is highly specialized for moving the carapace. There is no definite sterno-facialis (sphincter colli). 2. The depressor mandibulae (digastric) is absent or small. 3. The sterno-maxillaria is present. 4. The rectus thoracis lateralis is present and deep to the rectus ventralis. 5. The splenius colli is absent. 6. The rhomboideus has a large occipital origin. 7. The subclavius is large. 8. The clavicular deltoid does not unite with the trapezius to form a cephalo-humeral muscle. 9. The coraco-brachialis longus is always present. 10. The flexor longus cubiti (biceps) may have glenoid and coracoid heads (Daqpm), or only glenoid (Tatusia and Chlamydophorus). 11. The flexor brevis cubiti (brachialis anticus) rises from the neck of the humerus. 12. The extensor cnbiti (triceps) has more than one scapular head. 13. The flexor carpi radialis sometimes fails to reach the metacarpus. 14. The palmaiis longus is indistinct or absent. 15. The flexor sublimis digitorum has only one or two tendons. 1012 YESSRS. B. o. A. WINDLE AND F. G. PARBOBSON [Dec. 19, 10. The flexor profundm digitorum has a large palmar sesamoid developed in it. 17. The lumbricales as a rule are absent. 18. The pronator quadratus is absent. 19. The supinator longus is always absent. 20. The extensor ossis metacarpi pollicis is inserted into 1 he first metacarpal bone. 21. The eupinator brevis is sninll or absent. 22. The caudo-lemoralis is separable from the ectogluteus with difficulty. 23. The semimeinbranosus has a very extensive insertion into the tibia. 24. The femoral head of the flexor cmris lateralis is absent. 2.5. The tenuiesiinus is usually present. 26. The extensor longus digitorum never has a femoral origin. 27. The extensor brevis digitoruin nlways has a tarsal origin. 28. The peroneus longus sometimes ha8 a fenioral origin and its tendon alwnys rims across the sole. 29. The gastrocnemius has no fabellae. 30. The plantaris is present. 31. The flexor brevis digitorum is the continuation of the plantaris into the sole. 32. The flexores tibialis et fibularis have a large sesamoid bone in the sole. 33. The tibinlis posticus is double. 34. The accessorius pedis is represented only by a fibrous band.

MANIDB, 1. The panniculus is arranged as in a generalized mammal. 2. Sterno-facialis and sphincter colli are absent. 3. The sterno-maxillaris is absent. 4. The sterno-glossus is present. 5. The rectus thoracis lateralis is present and deep to the rectus ventralis. 6. The splenius colli is absent. 7. The clavicular deltoid forms a ceyhalo-humeralis with tlie trapeziu 8. 8. The coraco-brachialis is absent. 9. The flexor longus cubiti has only a glenoid head. 10. The flexor brevis cubiti rises from the neck of the humerus. 11. The extensor cubiti has more than one scapular head. 12. The flexor carpi radialiv reaches the metacarpus. 13. The palmaris longus is usually distinct. 14. The flexor sublimis digitorum has only one or two tendons. 15. The flexor profundus digitorum has a feeble palmar sesamoid. 16. The pronator quadratus is absent. 17. The supinator lougus is present or absent. 18. The extensor ossis metacarpi pollicis sometimes reaches no farther than the trapezium. 1899.1 FHB MYOLOOY OF fHB IDEINTATA. 101s 19. The supinator brevis is inserted into nearly the whole of the radius. 20. The caudo-femoralia is separable from the ectogluteus with difficulty. 21. The adductor cruris is single or double. 22. The semitendinosus always has a caudal head, though the ischial may be present or absent. 23. The femoral head of the flexor cruris lateralis is present. 24. The tenuissimus is absent. 25. The femoral origin of the extensor longus digitoruni is feeble or absent. 26. The extensor brevis digitoruni sometimes rises from the fibula. 27. The peroneus longus has no femoral origin and its tendon passes across the sole. 28. The gastrocnemius has no fabellae. 29. The plantaris is absent.

30.I The flexor brevie digitorum rises from the calcaneum. 31,, The tibialis posticus is double. 32. The accessorius pedis is present and fleshy.

OBYOTEROPODIDB. 1. The sterno-facialis is well-marked and extends back (caudal- wards) superficially to the pectorals to form a stern& as in Erhaaceuu and Bathyergus. 2. The sterno-mazillnris is absent. 8. The rectus thoracis Lteralis is absent. 4. The splenius colli is absent. 5. The rhoinboideus has an occipital origin. 6. The subclavius is large. c 1. The clayicular deltoid is a separate muscle and is inserted into t.he radius. a. The coraco-brachialis longus alone is present. 9. The flexor longus cubiti has only a glenoid bead. 10. The flexor brevis cubiti rises from the neck of the humerus. 11. The extensor cubiti has more than one scapular head. 12. The flexor carpi radialis reaches the metacarpus. 13. The palmari8 longus is indistinct or absent. 14. The flesor sublimis digitorum has four complete tendons. 15. The flexor profundus digitorum has no palmar sessmoid. 16. The pronator quadratus extends over the whole length of the forearm. 17. The supinator longus is present. 18. The extensor ossis metacarpi-~ pollicis is inserted into the trapezium. 19. The supinator brevis is inserted into the upper half of the radius. 20. The cauda-femoralis is fairly distinct. 21. The obturator internus has an intra-pelvic portion. I014 MESSRS. B. c. A. WINDLE AND F. (t. PARSONS ON [Dec. 19, 22. 'l?ht~pectineus is often double. 23. The adductor cruris is single. 24. The Remitendinosus has only an ischid head. 25. The femoral head of the flexor cruris lateralis is absent. 26. The tenuissimus is present. 17. The extensor longus digitorum has a femoral origin. 28. The extensor brevis digitorum is always tarsal in origin. 29. The peroneus longus has no definite femoral origiu, its tendon passes across the sole. 30. The gmtrocnemius has fabellae. 31. The plantaris is present. 32. The flexor brevis digitorum is continuous with the plantaris in the sole. 33. The tibialis posticus is double. 34. The accessorius pedis is fibrous. With the object of rendering comparison more easy we have arranged some of the more important muscles in a tabular form (p. 1015). We have now to consider what lessons may be learnt concerning the relations and systematic position of the animals included in the order of the Edentata from the study of the muscles. Flower (Proc. Zool. SOC.1882, p. 358), in a paper on the mutual affinities of the animals composing this order, says that &' the two Old-World forms Xunidce and Oyoteropodidce are so essentially distinct from all the American families, that it may even be considered doubtful whether they are derived from the same primary branch of mammals, or whether they may not be offsets from some other branch, the remaining members of which have been lost to knowledge." In using the muscles in the endeavour to deal with this problem, the first consideration necessary is to ascertain whether there are any departures from the generalized arrangement of mammalian muscles which itre common to all the families of this so-called order, for, if such exist,, they are not likely to be adapta- tions to similar conditions of life in animals far removed in relationship. For instance, if similar wanderings from the generalized mammalian arrangement of muscles can be found in the Pangolin and the Sloth, these wanderings are more likely to be the result of kinship than of an adaptive modification to meet similar conditions of life, for few animals more dissimilar in their habits could be imagined than these two. Everyone who has worked at Edentste myology will at once think of two curious muscular modifications which are not usually found elsewhere amongst the Mammalia, namely, the rectus thoracis lateralis and the fenioral head of the flexor criiris lateralis, or biceps. Both these musclea are present in the two families, although, RO far as we know, the rectus thoracis lateralis is never found as a distinct muscle outside the order with which we are now concerned, whilst the short head of the flexor cruris lateralis is only to be seen in the Edentates, Platyrrhine Monkeys, and Anthropoids, There are other peculiarities Mymnecqhagida. Daspodida. Masida. Orycteropopodidm.

----- I ----- 1 Sterno.maxillaris ...... Absent. Present. Present. Absent. Absent. Reetos thoracis latefalis. Presdht. Present. Present.. Preaent. Absent. F1. biev. cubiti (Brach. ant.) Seldom. Never. Always. Always. Always. from neckof humerus. No. of scapular bead3 of One. More than one. More than one. More than one. More than one. ext. long cubiti (Triceps). Pronator quadratas .. . . . Very small. Whole of forearm. Abseilt . Absent. Whole of forearm. Supinator longus...... Uuunliy double. Usually double. Absent. Preaent or nbeent. Present. Ext. brev. dig. from carpus. From carpus. Occasionally from Not froin carpus. Not from carpus. Not from carpus. carpus. Adductor cruris ._...... Double. Sometimes double. Single or double. Single or double. Single. Femoral. head of Flex. Always present. Usually present. Absent. Present. Absent. eruriR lateralis (Biceps). Tenuiasimus...... Absent. Usunlly absent. UaualIy present. Absent. Present. Condylar origin of Ext. Present. Usually present. Abseil t. Feeble or absent. Present. long. dig. F'ibutarorigin of Ext. brev. Present in B~udypzls. Absent. Absent. Sometimes present. Absent. dig. Fabllse in gaetrocuemius. Absent. Absent. Bbseiit. Absent. Present. Flex.brev. digitorurn...... _ Prom ralcis. From calcie. Continuous withplan- From calcis. Continuouswith plan- taria. taris. Tibialii pstious ...... Single. Double. Double. Double. Double. Fleshy. Fibrous. Fleshy. Fibrous. - - -- 1016 MESSRS. B. c. A. WINDLE AND F. 0. PARSONS ON [Dec. 19, common to the two animals, such as the total absence of fabellae from the gastrocnemius, the occasional presence of a fibular origin for the extensor brevis digitorum pedis, and the absence of the sterno-facialis (sphincter colli) part of the paniiiculus, which are not so striking as the former two, but which, taken together, are enough to make us think that there is a closer kinship between the Sloths and the Pangolins than they are generally supposed to possess. It would be easy to pick out points of similarity between the Sloths, Ant-eaters, and Armadillos by reason of which they differ from the generalized mammalian type, and which clearly point to their near relationship with one another ; it would also be easy to iudicate by means of its muscles that, although Manis cannot be a verfdistsnt relation of the Bratlypodidp, it is more closely allied to the Myrmecophagidce and Dasypodidce. When we come to cousider the Oryctm-opodiclcP,however, we are more struck with the generalized mammalian arrangemeiit of its muscles than by any special edentate characteristics : the three points on which we laid so much stress in claiming a place for the Pangolins in the Edentate order are wanting in the Aard-varlr. There is no rectus thoracis lateralis, no femoral head to the flexor crurie lateralis, and it has fabellie in its gastrocnemius just like any other mammal. In addition to this the sterno-facialis, which in all other Edeutates is suppressed, ia very strongly loarhed and covers a pnrt of the pectorals as in Erinaceus among the Insectivora adBiithyergus among Bodentia. There are, however, a few points in which the Aard-vark differs from most mammals and resembles the Edentata. One of these is the presence of more than one scapular head for the extensor cubiti (triceps), and another is the double tibialis posticus. We have never yet seen either of these arrange- ments in any other mammals but the Edentates; and we cannot help regarding this animal as a link between the Edentates and the more generalized stock from which that order has diverged. We have read with much iuterest a paper by Dr. Elliot Smith (Trans. Linn. SOC.,2nd ser. Zool. vol. vii. pt. 7, p. 387) in which he says that ‘(if the brain of Orycteropus were given to an anatomist acquainted with all the other variations of the mammalian type of brain, there is probably only one feature which would lead him to hesitate in describing it as an exceediiigly eimple Ungulate brain.” Changing the word muscles for that of brain, this is practically our own view. There are only one ortwo points which nould cause us to hesitate in describing Orycteropus as a generalized type of mammal, but theae one or two are certainlp in an edentate direction. We further read (ib. p. 390) that Miinis has certain cerebral features which point to a relationship with the American Edentate group ; a statement which strongly confirms the view which we have already expressed. Taking all these facts into consideration, we think that the systematists do well to retain the order of Edentata, although the name is certainly a misguiding appellation. We also think that it is not wive to lay too much stress on the articulations of the 1899.1 TR0 YYOLOGP OF THE EDESTITA. 1017 vertebrs of the American forms, and to press these into a separate order of Xenarthra to the exclusion of the Manidce. Tlie Oryctero- potlidce, too, present Borne feeble claipl to be taken into the order, for, gmeralized though they are, their m usculur peculiarities eeeni to point, so far as we at present know, more towards the Edeutatat than to any other group of mammals.

BIBLIOGRAPHY. I. MACALISTIR.--" Report on the Anatomy of the Insecti- vorous Edentates," Trans. B. Irish Academy, xxv. p. 491. II. COUVREURET BATArLLoN.--Annales SOC. Linn. LyOn, n. s. xxxviii. 1891, p. 83. 111. BaPP.--Bnat. Unterauchungen ub. die Enentaten. Tu- bingan, 1852. TV. HUMPHILY.-"On the Myology of the Limbs of the Untlu, the Ai, the two-toed Anteater, and the Pangolin," Journ. Anat. & Phys. iv. p. 17. V. &[ EC'KEL.--" Anat. des EM eizehigen Ameiuenfresser," Meckel's Archiv, v. 1819, p. 1. VI. GAULY)N.--" The Myology of C5jclothurus tlickictylus," Ann. C Mag. Nat. Hist. (4) iv. 1869, p. 244. VII. ~~~ACALISTER.-~'AMonograph on the Aniitoiny of Chlavtry- dophortts trutrcutus, &c.," Tribns. R. Irish Academy, xxv. p. 219. VIII. GALTON.---"The l\lyology of the Upper aid Lower Ex- tremities of Orycteropus cupensis," Trans. Liun. SOC. xxvi. p. 567. IS. HIJMPHRY.-'~On the Myology of Oryctteuopuu capensis," Journ. Anat. & Phys. ii. p. 290. X. GALTON.--"The Muscles of the Pore and Hind Limbs in Dcisypus sexcinctus," Trans. Linn. Socx. xxvi. I). 523. XI. &fEUIiEL.-TrBit& gdn&rald'bnatomie comparke, vol. vi. XII. HYrtrL.-~eulisc.hr. d. k.-k. Akad. d. Wissensch. in Wien, Bd. ix. XITI. N,LCKINTOSH.-~'On the Muscular Anatomy of ChoZ~~ptcs di&u%ylus," Pror. K. lrish Academy, ser. ii. vul. ii. p. 66. XIV. ~~IAUALIS'I'I~L--"On the Myologg of Brrdypus tridnc- tylus," Ann. &, &lag. Kat. liiqt. (4) iv. 1869, p. 51. XV. OWBN.--" 011the Anatomy of tlie Great Anteater (Myr- mecophayu jubata, Linn.)," Trans. Zool. 8oc. iv. pp. 117, 179. XVI. &$ACKINTOYTI.--" On the l$yology of the Genus BrcLdy- pus," Proc. R. Irish Academy, ser. ii. vol. i. p. 517. XVII. C'UY~ERet LAUMLLARIL-~Planches de Myologie,' 1840. 1018 MIL. W. P. PYCBAFT ON THE [Dec. 19,

i. Introdocbry Remarks, p. 101R. viii. The Pectoral Limb, p. 1037. ii. The Skull of tho Adult, p. 1019. ix. The Pelvic Limb, p. 1037. iii. The Skull of the Nestling, p. 1@28. x. Summary, p. 1041. . iv. The Vertebral Column, p. 1035. xi. Key to the Osteology of the v. The Ribs, p. 1034. Pygopodes, p. 1042. vi. The Sternum and Pectoral Girdle, xii. List of Works referred to or p. 1035. consulted, p. 1044. P.Z. S 1899.PL.LXXl1

OSTEOLOCW OF THE PYGOPODES

1899.1 OST&Ou)BY OF TJXB PPOOPODES. lolo a given bone belongs, when only thrvt portiun of the akelebn, as frequently happens, comes up for determination.

ii. THI SEULLOF TBE ADULT. The skull of the Pygopodes resembles on the one hand that of the Alcib and the Penguins, and on the other the Bails. Itm resemblance to the two former rests chiefly upon the structure of the palate, which in all is schizognathous, and of the deep mpra- orbital grooves, when these are present. It can at once be distinguished from the Alcids by the holorhinal nares, which in the Alcids are schizorhinal, and from the Impennes by the rod- shaped pterygoids, the inflated basitemporal platform, the laminate maxillo-palatine processes, and the great width and shallowness of the temporal fossa, when present. Its resemblance to the Rails is confined to tbe smaller Grebes, and in these it is very striking. The skull of the Grebe, however, is always to be distinguished from that of the Bail by the conspicuous development 'of a cerebellar prominence, similar to that of the larger Urebes, the Divers, Penguins, and Auks. At the base of this prominence is a well-marked deepening of the posterior region of the temporal fossa which is never found in the Rails, where the fossa is only barely indicated by s very shallow depression. The Occipitcil Rqion.-The occipital condyle is more or less reniform in the Grebes and hemispherical in the Divers, though even here the flattened upper surface is slightly hollowed. The form and development of the paroccipital processes resemhle those of the Penguins; they pass upwards into the lsmbdoidal crest and forwards into the squamod prominence. In the smaller Grebes these processes are but feebly developed, being represented only by small and somewhat infloted bosses laterad of the base of the foramen magnum. In the Grebes, caudad of the inner end of the lip or interior free border of the process is a deep groove which is not present in the Divers. The supraoccipital is not pierced by lateral fontanelles, but there is a small median foramen above the foramen magnum in the Divers ; this is wantiug in the Grebes, and the cerebellar dome-formed by this bone-is marked by a more or less well-defined median vertical ridge, or low keel, forming s supraoccipital crest, differing in this respect from both Penguins, Petrels, and Auks. This crest joins the median sagittal crest, dividing the temporal fosse, at the lambdoidal ridge. The squsmoso-parietal wings, in the Divers, rise in the form of sharp lateral ridges for the whole height of the skull as in many Penguins, terminating in the middle line in a more or less diagonal expansion, which passes forwards into the median sagittal crest. The free edges of these wings give a sharply defined crescentic outline to the skull when seen from behind. These wings, as in the Penguins, occupy the position of the lambdoidal ridge. 1020 MR. W. P. PYCRAFT ON THE [Dec. 19, In the Grebes, the squamoso-parietal wings do not attain to the height of the skull. In the larger species they cease abruptly at about halfway, a thin raised ridge running upaards from this point to the sagittal crest representing the lambdoidal ridge. In the smaller Grebes the squamoso-parietal wings are feebly developed. As in the larger species, the lambdoidal ridge is represented by a thin raised line, terminating at the sagittal crest. The Roof of the C'raniunz.-The parietal region in the Divers and larger Grebes is impressed by wide but shallow temporal fosss, divided, in the fully adult bird, by a narrow sagittal creit (Plate LXXIT. fig. 1). It is interesting to note that in a nearly full-grown skull of Colymbus septcntriodis in the Museum Col- lection this crest (Plate LXXII. fig. 2) is represented by a broicd plate of bow, whilst in an immature C. !$addis, apparently a little younger than that of C. septentriondis, inasmuch as Borne sutures are yet distinct, the sagittal crest ia as sharply defined as in the adult. In the larger Grebes the sagittal crest is sharply defined, a8 are the temporal fosss. In the smaller species the anterior buundar.y- line of the temporal fossa is barely visible, posteriorly the fossa is moderately deep. The form of the fossa differs from that of the lar,ner species and Divers, in that it is relatively shorter from betore backwards, the cerebellar region of the skull only slightly projecting backwards beyond the cerebral. In the former the backward extension of the cerebellar prominence is very marked, more so thnn in the Penguins and Petrels. The mpmorbital grooves of the frontals in the Divers are separated by a median knife-like edge ; externally, they are bounded b.y a broad supraorbitel ledge the free edge of which is flattened, as in many Penguins. Anteriorly, the supraorbital ledge fuses with the posterior-dorsal limb of the lachrymal on either side. The anterior inner border of the ledge, immediately behind the lachrymal, is pierced by a large foramen for the paRsage of the lachrymal duct. In the Grebes, the supraorbital region of the frontals is marked by a wide shallow median furrow; supraorbital grooves can hardly be said to exist, being represented only by a faint excavation along the free edge of the frontal. The supraorbital ledge, in the Divers, posteriorly combines with the alisphenoid, to form a prominent postorbital process. In the Grebes, this process is only very feebly developed ; moreover, this region of the skull differs markedly in the two forms, in that, in the Divers the postorbital process is continued forwards as the supraorbital ledge, whilst in the Grebes the region in front of this process is mlrrkzd by a shallow depression for the insertion of the muscles of the jaws. Bchntophoms seems to be the only Grebe to which the above remarks do not apply ; in this genus the periphery af the fossa lodging the muscle is produced outwards into a broad shelf-like emarginate postorbital process. 1899.1 08TEOLOGP OF ‘PHI PYGOPODES. 1021

Zhe Base of the Skull.-The basitemporal plate of the parasphenoid is 1narksd1.y inflated, and has a bevelled anterior border, the free edge of which, in the Divers, is overhung by a downgrowth from the alisphenoidal wings of the parasphenoid. This, ill the Grebes, by its fusion with the free edge of the basitemporal plate, forms R pair of closed tubes opening on either side of the skull, behind the quadrate, and below the squamosal prominence into the aperture which serves also aa the mouth of the tympanic cavity. The inferior and posterior of these two runs transversely across the skull, and forms the Eustachian tubes of the right and left sides of the head. The connection with the choanm is by means of a single median aperture immediately under the rostruni. The anterior rum forward as n pneumatic cavity into the body of the parasphenoid to terminate beneath the level of the foramen opticum. In the Divers, the form of trhis aperture is tubular, recalling that of the Penguins, the anterior wall of the tube being continued outwards behind the equamosal, but in the Grebrs the anterior wall is deficient. Mammillary processes are but feebly developed ; in the Divers the paroccipibal notch is wide and shallow, it can scarcely be said to exist in the Grebes. There is a precondylar fossa, or rather groove, in the larger species of both families. The parasphenoid rostrum, in both Grebes and Divers, is somewhat inflated at the base, owing to the presence of the pneumatic cavity already described. The Laternl Aspect of the Cranium (Plate LXXII. figs. 3-6).- The tympanic cavity has n eharply defined aperture in ths Divers (Colymbi), by reason of the considerable lateral development of the alisphenoidal wing of the parasphenoid. Within its mouth call be seen, distad, two large apertures, lying immediately behind the alisphenoidal wing just referred to : the upper is the pneumatic aperture of the parasphenoid, the lower is the Eustachian aperture ; caudad, and separated by a broad column of bone, lie the fenestra ovalis and the fenestra rotunda. The temporalis reoess’, so well developed in the Steganopodes and Petrels, and to a lesser degree in the Penguins, is here represented only in the Divers, by a moderately deep fossa ; in the Grebes it is wanting. The posterior pneumatic cavity opening downwards, behind and above the fenestru ovcilis, so well developed in the skull of the Tubinares, is wauting in both Grebes and Divers.

1 In my recent paper on the Osteology of the Penguins the temporalis recess was described as “leading eventually, in the dried skull, into the cranial cavity.” This is quite a mistake. The correct interpretation of this is as follows :-In many skulls, e. g. PuJimts, above the trigeminal there lies a second foramen for t,he sinus transversus of‘ the vena cephalica posterior-at times this is corifliicrit with that, for the trigeminal, e. y. Divers-and both these lie immediately outside, below and mesiad of the mouth of the recess in question. In the Penguin the foramen suppleineiitary to the trigeniinal lies immediately within the mouth of the recess, piercing its inner wall; owing to imperfect ossiRcation, the mouth of the foramen in the dried skin extends upwards nearly the whole length of the recess. hO0. ZOOL. SOU.-1899, NO. LXVI. 66 1032 MR. W. P. PTCRAFT ON THE [Dec. 19, The tympanic cavity in the Grebes is not so sharply defined as in the Divers, owing to the slighter development of the alisphenoidal wing of bhe parasphenoid ; in other respects, however, it closely resembles that of the Divers. The squamosal prominence is only concerned with the suspension of the quadrste and does not, in addition-except very imperfectly, and only in the Divers--form the roof of a temporalis recess, as in the Tubinares atid the Impennes. The paroccipital processes are well developed in the Colymbi, forming, as in the Impennes, 8 backward continuation of the squamosal prominence ; in the Grebes (Podicipides) these processes are feebly developed, being represented only by an abruptly truncated lamina of bone. me tempora2 fossa-In the Colymbi (Divers), and in the larger Podicipides (Grebes), the temporal fossse are very wide and shallow and sharply defined. They are separated in t,he mid-dorsal line only by a sharp median sagittal crest. Within the confines of the fossa, on each side, are defined the limits between the cerebral and cerebellar regions of the skull, the squamoso-parietal winp running transversely across the cerebellar dome, as in the Petrels, and not, as in the Penguins, traversing the boundary line between these two regions. In the smaller Grebes, e. g. Tachybaptes, the temporal fossa is only faintly defiued, moreover it is almost entirely con6ned to the cerebral, and extends scarcely at all on to the cerebellar dome. This is due, not to any essential difference in the form or position of the fossa, but to the extremely slight development of the last-mentioned dome. The trigeminal foruwn, in the Divers, lies immediately outside the inner wall of the mouth of the fossa representing the temporalis recess. There is no foramen for the sinus trausversus branch of the vcna cephalim posteriov above this as in the Penguins and Petrels. In the Grebes the trigeniinal and the venous foramen immediately above are confluent ; it pierces the Mall of the alisphenoid immediately in front of the articular surface for the head of the quadrate and below the anterior border of the squauosd prominence. It lies relatively fiirther forwards thau in the Divers, inasmuch as, in the Grebes, this foramen and the anterior border of the squamosal prominence lie only a short distance behind a vertical line passing through the postorbital process ; in the Divers both these points lie far behind this line. The orbits, in the Divers, are overarched by the supraorbital ledges, the postorbital process bounding them posteriorly and the lachrymal anteriorly. The perforated interorbital septum forms a mesial partition-wall. In the Grebes the supraorbital ledges and postorbital processes are wanting, and the lachrymal is very small, so that the orbit in tliis group is not nearly so well defined. An osseous interorbital septum is wanting; in the Divers a large extent of the middle region is supplied by membrane, so that in the dried skull the septum is largely fenestrated. 1899.) OSTEOLOGY OF TJD PYBOPOUES. 1023

The ortitosphenoid in the Divers is completely ossi6ed j in the Grebes this is largely represented by membrane, so that the anterior wall of the brain-case in the dried skull benrs a more or less considerable fenestra. In a skull of Bhmophorus kindly lent me by Mr. Beddard this fenestra is very small, the orbitosphenoid is ossified dorsad so as to close in and form tubes for the olfactory nerves as they leave the brain. In the Divers the interorbital septum forms a vertical bar in front of the optic foramen, this is wanting in the Grebes. The ethmoidal region.-The mesethmoid is indistinguishably fused with the presphenoid ( =interorbital septum) behind, and the parasphenoidal rostrum below j it expands dorsally ns usual into 8 pair of lateral aliethmoidal plates under the nasal and frontal bones, the free edges of which curve slightly downwards under the outer border of the frontal and along the inner border of the lachrymal. Its posterodorsal border is continued backwards to terminate in B sharp, spinous, crista-galli forming a median partition between the olfactory nerves. The antorbital plate in the Divers is represented by a thin ridge of bone running from the mesethmoid outward and forward to the lachrymal. In the Orebes this is represented by a narrow band-shaped scroll of bone from the lower and hiuder border upn ards to the nasal, immediately to the inner side of the dorsal end of the lachrymal. A comparison may profitably be made here between the 1110s- ethmoid of the Pygopodes and that of the Iinpennes and Tubinares. In the two first mentioned groups the mesethmoid is relatively smaller than in the last, and only very sliglitly pneumatic. In the Divers and Grebes its anterior border curves gently forwards, carrying with it a pair of lateral wing-like ridges, the whole eventually terminating in a sharply truncated border running transversely across the skull immediately under the free end of the nasal processes of the premaxilla. Its posterior border is deeply hollowed by the interorbital feneetra. Its dorsal border, anteriorlg, expands into a pair of lateral diethmoidal plates, tapering from before backwards j posteriorly it runs backwards, in the Divers in the form o€ R deep, and in the arebes in the form of a very narrow knife-like ridge, the free end of which terminates as a pointed " cristegalli "- within the olfactory fossn. In the Impennes the form and relatiqns of the mesethmoid closely resemble those of the Pygopodes. In the Tubinares, the mesethmoid differs from the forms de: scribed on account of the fact that its upper and lower regions are brought into sharp contrast by reason of the great pneumaticity of the lower region, which causes the upper non-pneumatic half, with its gently arched aliethmoidal wings, to assume a cavern-like form, which passes backwards in a tubular manner into the olfactory fossa. Moreover, the crista-galli takes the form of a median pillar dividing two large tubular apertures for the olfactory crura ; whilst in the other fqring the crista-galli is 66* 1024 ME. W. P. WCILAPT ON THE [Dec. 19, reduced to a spine-like process dividing two greatly reduced passages for the olfactory nerves, the crura lying caudad of the crista, and not passing on either side. The a2iethrnwid.s are the only ectoethmoidal ossifications in either the Impennes, Tubinares, or Pygopodes. They constitute the antorbital plates. In the Colymbi, when present, they resemble in form those of the Tubineres-plates of bone jutting out from the mesethmoid to the lachrymd, sloping obliquely forwards and downwards. They appear, however, in the Colymbi to be but rarely ossified, and never so well developed as in the Tubinares. In the Podicipides their true nature is well seen. Here, they a pear as scroll-like bars of bone running continuously backwards, lownwards, and inwards from the expanded dorso- lateral plates of the mesethmoid itself, to pass eventually into the middle of its posterior border. The o7,factory chamber is of comparatively small size. The lachrymal in the Poclicipides is free, sniall in size, roughly semilunar in shape-with the convex border forwards-and apparently a disappearing structure. It articulates by its superior limb with the outer border of the nasal bone. Though conspicuous from a lateral view, it is scarcely if at all visible from the dorsal aspect of the skull. In the Colymbi, the lachrymal is roughly of the same shape ay in the Podicipides j but it differs therefrom, markedly, in several points. Its superior limb, as in the Grebes, articulates with the nasal; but it is free only in the young bird, later it becomes indistiriguishably fused with that bone ; moreover, it sends backwards a long spur to fuse with the supra-orbital ledge, and to enclose with its aid a passage for the lachrymal duct, as in many Alcidae. Its lower limb, at its free end, is more or less markedly sigmoidally curved. It is, on the whole, a larger and stronger bone than in the Grebes. The Cranial Cavity.-The metnncephcibic fossa of the Pygopodes is relatively both longer and shallower t,hau in the Impennes and Tubinares, but is deeper in the Grebes than iu the Divers. The vayus jornmen occupies relatively the same position as in the two last mentioned groups ; the two condgloid foramina, similarly, lie mesio-caudad of this. The internal auditory meatus lies immediately under the mouth of the floccular fossa and in front of the vagus foramen. The abducent foramen pierces the anterior border of the foe%&,passing on either side of the pituitary fossa, and emerging in the Grebes within the rim of the ventral border of the optic foramen, and in the Divers caudad of the foramina for the oculomotor nerve and internal ophthalmic artery. The cerebellar fosacc agrees with the Tubinares, and differs from the Impennes, in its greater relative size. It lacks, however, the transverse system of gro-oves and ridges representing the cerebellar sulci and gyri so marked a feature in this region of the skull in the Tubinares. 1899.1 OSTEOLO(3Y OF THB PYGOPODES. 1025 The mesencephalic fossa agrees with that of the Tubinares, in that the groove for the sinus transversus branch of the vena cephalica posterior (pp. 1021-2) appears as a deep tunnel excavated out of the inner wall of the skull. It runs upwards, backwards, and downwards, following the curve of the anterior semicircular canal, finally piercing the wall of the supra-occipital tunnelwise, leaving the skull by a small aperture on either side of the foramen niagnuni, but much lower clown than in the Impennes or Tubinares. The Pygopodes differ from the two last inentioned groups in that this cephalic vein enters the skull through the’trigeminal foramen, arid not by a separate aperture. The floor of the fossa bears a deep groove for the orbito-nasal nerve. The pituitary fossa differs conspicuously both from that of the Impennes and of t,he Tubinares by its peculiar shallowness. In the Colymbi it form a moderately deep pit sloping gently backwards; but in the Podicipides it is represented only by a very slight oblong depression, bounded on either side by a strong ridge forming a tunnel for the abducent nerve. This pituitary ridge, from the point where the nerve enters, is continued upwards and outwards, to terminate at the groove for the cerebral vein, already described. This second ridge forms the posterior boundary line of the mesmcephalic fossa. There is a well-developed dorsum sellce and prepituitary ridge in the Colymbi, the latter flattened to form an optic platform. The pre-optic, as usual, passes on either side into the tentorial ridge. In the Yodicipides there is no dorsum sellrp, the internal carotid apertures opening directly 011 to the floor of the fossa. The optic foramen in the Colymbi is bounded in front by a vertical plate of bone, being that part of the interorbital septum bounding the interorbital fenestra posteriorly. This plate is wanting in the Podicipides. The cerebral fossce in the Divers and Grebes closely resemble one another, they both agree in that this region of the brain-cavity is greatly depressed dorso-ventrally, as in the smaller Tubinares. The tentorial ridge is sharply defined, particularly so in the Divers. The bony falx, which dips dowm between the pallial fissures of the brain, is not 80 strongly marked as in the Tubinares. There is only the faintest indication of the bony ridge marking the position of the Sylvian furrow such as is found in the skull of Diomecka, and this is entirely dorsal, and not lateral as in Diomedea. Again, the tentorial ridge of the Pygopodes and emaller Tubinares lies horizontally to the long axis of the skull ; whilst in Dwmedea, for instance, it is almost vertical in position. The olfactory fossce, like those of the Impennes, are extremely small, and pass insensibly into the cerebral fossze behind ; wherein they stand strongly contrasted with the large tubular chambers of the Tubinares. They are divided in the middle line, forwards, by a small knife-like crista-galli. In the Tubinares, it will be remembered, this crista-galli is columnar. 1026 YB. W. P. PYCRAFT ON TEB [Dec. 19, The Premaxilla, Nasal, and Lachymal. The nnso-premaxillary region of the upper jaw bears B strong superficial resemblance to the Alcide on the one hand, and certain genera of Penguins on the other, e. 9. that of Begaeudyptee. It may readily be ilistingiiished from the former by the form oE the nostrils, which are schizorhinal in the Alcitle ; from the latter, apart from the rest of the skull, it would be difficult to distinguish it. The nnsal, in the adult, both in the Colynibi and Pygopodes is completely fused with the premaxilla and frontals. In the Divers it is, furthermore, fused with the outer border of the lachrymal. It is not deeply cleft caudad, tlie form of the external narial apertures being holorhinal ;they are also, by the way, in the dried skin pervious, there being no nnsnl septum. The hchrymal is a fairly large bone in the Colymbi, with a peculiar notch in the posterior border of its free end; in the Podicipides, as elsewhere remarked (pp. 1024, 1031), it shows eigns of degeneration. The Maxillo-jupl Arc?&. The mrc&lTa, in the adult of both Grebes and Divers, is indistinguishably fused with the premaxilla. The mnxillo-palatine processes are Charadriiforrn in type, closely resembling those of the -4lcide, being leaf-shaped, and in the Colymbi more or less fenestrated. The atatrum is very shallow. They are widely separated in the middle hue, and tlie palate is therefore schizo- griathous. The form of these processes rather closely resembles that of the smaller Tubinares, e. g. Pelngodroma, Pi*ocPllaria, but differs markedly from that of the larger forrim, such as Pu$nus for iiistance. In these, it will be remembered, the maxillo-palatine processes do not extend backwards into the Lchrymo-nasal fossa, ond are hollowed out to form a spacious antrum of Hiyhnaore. The anterior end of the maxillo-jugal arch does not trend upwards to meet the lachrym~l,as in many Procellariidae, but the lower limb of the lachrymal in the Colymbi is very long, so much SO as nearly to touch the bar, and thus to completely shut off the triangular lachrgmo-nasal €omftom the orbit. The greater part of this arch is made up by the quadrato-jugal bar.

The Yomer, Palatihe, and ~teryyoirl. The vomer in the Colymbi appears to be free throughout life ; in really adult or advanced Podicipides it fuse8 with the palatines. In the Colymbi it somewhat closely resembles that of the Diomedeidae, appearing knife-shaped from below and provided uith a pair of lateral wings along its dorsal border. It differs from this type, however, in that it lacks the conspicuous dorso- ventral curve. Its dorsal aspect is trough-like, and posteriorly lodges the anterior end of the parasphenoidal rostrum. In the Podicipides it is knife-like, and lacking the lateral wings, 1899.1 OSTEOLOOY OF TEE PPGOPOPES. 1027 when seen from below. From the dorsal aspect, the trough-like region is found to be restricted to the posterior end, immediately underlying the anterior end of the rostrum. Thepalatine (Plate LXXII. fig. 7), in its general form, and in its relations with the vomer and maxillo-palatine processes, somewhat closely resembles that of the Alcidae. It differs conspicuously from that of the larger Tubinares iu its greater relative length, and in the feebler development of the paired median and lateral keels formed by the inner and outer borders of the palatine caudad of the maxillo- palatine processes. Its general contour, from the ventral aspect, may be described as rod-shaped, iu front of the maxillo-palatines, passing behind these into a shallow trough-like expansion. Seen dorsally, the inner free edge of the expanded portion, cephalad, rises dorwlly into a scroll-shapeci plate to articulate mesially with the vomer. Seen laterally, this scroll-like dorso-lateral plate does not fit closely up to the posterior border of an almost vertical maxillo-palatine process as in the Tubinares, but leaves a large space between. The pterygoid is rod-shaped, and more or less triangular in section. That of the Colymbi sends outwards, from the outer border of its extreme proximal articular end, a small plate to abut ngainst the base of the orbital process of the quadrate. Moreover, it still further differs from that of the Podicipides, (1) in that it possesses a strong inward curve, causing the pterygoids to embrace the parasphenoidal rostriim, as in many Procellaria, from which it differs, however, in that no glenoid articular surface is dereloped for this purpose j and (2) in that it seuds forward and upward, from the external ventral border of its extreme distal end, a delicate claw-like process, to embrace the articular end of the palatine- or more correctly of the anchglosed Lemipterygoid. This point was first noticed by Coues (3). The quadrate more nearly resembles that of the Tubinares and Impennes than that of the Alcidae. It differs from all in the very elongated and rod-like form of its orbital process. Its otic and squamosal heads are sharply divided. The glenoid mandibular surface nearly resembles that of the larger Tubinares. Its external condyle is hollowed from before backwards ; its internal condyle may be divided into relatively large anterior and posterior facets, meeting one another in the mid-ventral line. Immediately dorsad of the posterior facet of the internal condyle is a sharply defined pterapophysial facet for the articulation of the pterygoid. It is non-pneumatic both in the Grebes and Divers, in which point it agrees with both Impenzles and Alcidae, and differs from the Tubinares, in which it is pneurnrbtic

The Mandible. Is slender, elongated, and slightly recurved at the tip. It can be more or less readily distinguished from that of other groups--such us some Ardeidae, which it somewhat closely resembles-by the 102% MR. W. P. PYCRAFT ON TEE [Dec. 19, form of the coronoid. This is long, narrow, arid pointed. Its anterior half remains distinct throughout life, the posterior region fuses with the anguhre. The dentary suture also remains distinct. In the Divers the supra-angular is piel-ced by a large foramen, which remains permanently open : furt,hermore the jaw may be distinguished from that of the Urebes by a deep notch situated immediately behind the outer border of the glenoid surface for the external condyle of the quadrate ; in this last it agrees with many Alcida. The internal angular process is feebly developed. The mandible of t,he Podicipides (Grebes) lacks the notch just describd, and the foramen piercing the supra-angular is much reduced in size, aud sbiit off from within by the base of the roronoid ; the internal angular process is moderat,ely well developed. The Ifyoid. The ityoid of the Pygopodes differs markedly from that of the Alcida, both in the form of the basihyal and basibranchial ossifications, neit,her do they bear any close resemblance either to those of the Tubinares or of the Iiapennes. In each of the three last-mentioned groups the first and second hasibranchials are anchylosed, the latter being continued backwards in the form of R median pointed bony st,yle between the ceratobranchials to form t.he “urobyal.” In the Pygopodes the first and second basi- branchials appear to remain distinct throughout life. In the Pypopodes the first bssibranchial-the main body of the bone-takes the form of an oblong plate hollowed dorsally, and with a slight median keel ventrally. The basihyal is partly ossified, the ceratobranchials are relatively rery long, the epi- brnnchials short and slender. In the Colyinbi the first basihrauchial takes the form of a flattened oval, and the anterior region of the ventral median keel iR strongly developed. T)ie second (“ urohyal”) is apparently but imperfectly ossified, only the very centre of the rod-shaped style being bony. In the Museum skeletons there is no basihyal ossiticntion, but this may have been lost in maceration.

iii. THESKULL OF THE NESTLING. For the nestling which forms the subject of the following notes I have to thank the Hon. Walter Rothschild, who kindly provided it specially for this purpose.

a. Cartilage-bows. The basioccipitaZ seen externally is linguiform j it forms the central portion of the occipital condyle posteriorly, and anteriorly is underfloored by the basitemporal plate of the parasphenoid. It joins the exoccipital by harmony suture. Internally it is hounded laterally by the exoccipital and anteriorly by the bnsi- sphenoid. 1899.1 OSTBIOLOGY OF TEN PYQOPODEB. 1029 The R~:oc~pifaZ.-Exterually it is bounded supero-internally by the supraoccipital, and supero-externally by that portion of the prootic cartilage which lodges the floccular fossa. Its internal border is <-shaped. The posterior limb is free and bounds the fornmen magnum, the anterior runs along the outer border of the basioccipital. Its external border has fused with the prootic cartilage, the boundary between the two being indicated by a notch superiorly. The vagus foramen pierces it near its centre. Internally, it is largely concealed by the opisthotic, only its inner half being visible. The vagns foramen appears on this side at the bme of the opisthotic, and indicates how much of the exoccipital is concealed by this bone. The eupraoccipital is cleft in the middle line superiorly for more than half its length. Its rentri-lateral border is fused with the epiotic, but the distinction between the two bones can yet be made out. Its internal does not differ much from its external form. The prootic appears externally in the form of an oblong mass of cartilage separating the squamosal from the anchylosed opisthotic and exoccipital bones. Internally its size is seen to be considerable. Supero-dorsally it is notched to form the inferior border of the floccular fossa. Its outer half lies immediately in front of the squamosal, which bone it almost entirely shuts out from the inner surface of the cranium, only a small semicircular strip of about $ mm. being risible, and forming the floor of a groove separating the prootic from the parietal and alispheuoid bones. Its antero- ventral horder is linguiform and imbedded in a mass of cartilage. Its inner lateral border is in pnrt (superiorly) fused with the opisthotic and in part (inferiorly) imbedded in cartilage in common with the linguiform auterior end. The meatus internus is very deep. The epiotic is only ossified at its junction with the supraoccipital ; the rest of the posterior vertical semicircular canal is yet cartilaginous, and forms the superior boundary of the floccular fossa. The opiathotic has almost completely fused with the prootic, but traces of the original suture still remain. It fuses posteriorly with the exoccipital. The vagus foramen passes between its nutero-internal border and the exoccipital in front. There is no trace of the opisthotic visible externaliy. The baeisphmoid is not visible externally, being underfloored by the parasphenoidal rostrum and its basitemporal plate, .with which furthermore it has now completely fused. Below the pituitary fossa, and that portion of the bauisphenoid immediately behind it, is a large air-sinus, and this forms the only indication of the division between the para- and basisphenoidal regions. The pituitary fossa forms a moderately deep pit. The dorsum sellae is yet membranous. In the adult the pituitary fossa seems to have become almost obliterated (p. 1025). Immediately behind the pituitary fossa the basisphenoid is marked by a deep < -shaped notch, dividing a median portion from a pair of lateral wings. The membrane stretched between the two wings forms the dorsum 1030 U.W. P. PYCRAFT OY TRI [Dec. 19, sellm. This arrangement suggests an ossification of the basisphenoid from three centres-a median longitudinal, running forward to lodge the pituitary fossa, and fwo lateral wings. Thirr is a point for further investigation on fresh material. The basisphenoid is bounded laterally by the alisphenoid, postero-laterally by the prootic, and posteriorly by the basioccipital, and anteriorly by the presphenoid. The aliqhenoid has partly fused with the orbital process of the frontal, but is otherwise at present free. Externally it is roughly circular in form. The posterior convex border follows the outline OE the concave anterior border of the squanio~al,but the two borders do not as yet even touch. Neither has the ossification of its ventral border extended dow-nwardsas far as tbe parasphenoid. The orbit* and prqhenoida have not as yet begun to ossify. They are represented only in the dried skeleton by a thin transparent sheet of tissue-the remainu of the original cartilage. The mesethmoid is almost completely ossified and subcrescentic in form, its convex border forwards. It extends vertically from the parasphenoid below to the nasal above. The qwadrate has not yet assumed its fully adult form, the distal end of the orbital process being only cartilaginous and relatively shorter than in the adult. The columella is represented only by its base, which is ossified ; the stapedial rays are not distinguishable in the dried skull. The artimlure can still be distinguished aa a separate element. b. Membrane-bones. The parietaZ externally is oblong in form, with its borders nearly straight and at right angles one with another. Immediately above the antero-ventral angle of its anterior border it is overlapped by a tongue-shaped process from the frontal. Its inferior border rests upon the dorsal border of the squamosal, than which it is a trifle broader. Its posterior or hinder border ventrd abuts against the epiotic, which at this point is cartilaginous, and dorsad forms a harmony suture with the supraoccipital. Internally the angle formed by its hinder and ventral borders is cut off from the inner surface of the skull by a portion of the pro- and epiotic bones. The frontal has its hinder border nearly straight. Just above its postero-ventral angle it sends backwards a slight linguiform pmes to overlap the parietal. Its outer, ventrd border caudud extends dow-nwarde to within a short distance of the level of the Rquamosal, oephalad it sends downwards a small orbital process to overlap the alisphenoid. The equmosciZ Been externally is almost quadrate in form, but widest along its dorsal border, which, like the remaining three, is slightly hollowed. It articulates by a close harmony suture with the parietal. In the hollow of its hinder border there is a small os8eous nodule representing a portion of the opisthotic. The squatnosal is almost entirely excluded from the cranial 1899.1 OSTEQLOGY OF THB fyQOPODElS. 1031 cavity, being covered by the prootic. It appearr, however, as a small semicircular tract of bone curving round the outer lateral border of the prootic, and bounded above by the frontal and in front by the alisphenoid. The nasal extends backwards to a point corresponding in the adult with the level of the free posterior border of the anterior portion of the fenestrate interorbit,al septum. The external processes extend forwards to the anterior extremity of the external nasal fossa. The form of the nasal cleft is fiolorhinal. The posterior border of the naeal is produced backwards into a point. In the Grebe this point is separated from its fellow of the opposite side by a median forward extension of the frontal ; in the Diver the two points meet in the middle line. The mesial edges of the right and left sides are separated one from another, for a cousiclerable distance behind the level of the posterior narial aperture, by the n:rsal proesses of the premaxilla. In the adult the extreme posterior limit of these processes is indicated by a more or less well-marked transverse groove, corresponding with the “ nasal hinge ” in forms in which tbis is present. The lachrymal in the Qrebes is subcrescentic in form, and appareutlg in process of degeneration. In the Divers it is still t) moderately large bone. It articulates entirely with the nasal, from which it projects laterally, for a considerable distance on either side of the skull, in the form of a subcrescentic backwardly directed spur. The inner border of this spur is hollowed and forms the anterior limit of the supraorbital groove. The extreme posterior end of the spur anchyloses with the supraorbital ledge, which anteriorly between itself and the inner border of the lachrymal is deeply hollowed ; thus a large supraorbital fenestra is left for the passage of the duct of the naml gland. A similar supraorbital fenestra is found also in many Charadriiform birds, e.y. Alcidre. The inferior limb of the lachrymal in the Divers extends downwards to within a short distance of the quadrato-jugal bar. Its free end is deeply notched, the lower projecting buckwards for a considerable dietance beyond the level of the upper limb of the notch. Thepl’maxilZu in the Pygopodes is produced forwards into a pomt. The median cleft dividing the nus1 promsses does not extend so far forwards as in the Impennes. They rest upon the internal processes of the nasals. The maxilla extends backwards rodwise to form the inferior border of the anterior half of the quadrato-jugal bar, underlying the jugd, and forwards as a long triangular splint below the maxillary process of the premaxilla. It is bounded on its inner side by the palatine. The maxillo-palatine process takes the form of a concavo- convex lamella, which in the Colgmbi extends further backwards into the lachrymo-nasal fossa than in the Podicipides. There is never more than a vestige of the antrum of Highmore. The jugal takes the usual elongated splint-like form. It over- lies the maxilla anteriorly and the quadrab-jugal posteriorly. 1033 MR. W. P. PYCRAFT ON “HE [Dec. 19, The quadratojugal extends forwards beyond the middle of the quadrsto-jugal bar, passing to the inner side both of the jugal and maxilla. It articulates posteriorly with the quadrate, fitting into a deep cup-shaped cavity. The vcmer in the Colymbi articulates with the hemipterygoid posteriorly, and with the palatine by means of its dorsal border. In the Podicipides, in our Museum skeleton, the relations of the vomer to the hemipterygoid are not easily made out, owing to the fact that the hemipterygoid has not yet split off from the pterygoid (p. 1026). The palatine is of great length, extending forwards as a long slender rod to within a short distance of the tip of the jaw. Posteriorly it is more orless grooved along its ventral aspect ; from its inner dorsal border there arises caudad D leaf-like plate of bone turning inwards towards the middle line, the free border of which articulates with the superior border of the posterior end of the vomer. The pteygoid, in so far as its general form is concerned, has been already described. We are concerned here only with the segmentation of its anterior end to form the Bmipteygoid.-This can best be studied in the Colymbi. Here it bears a very close resemblance to that of the Impennes. In the skulls of two Divers in the Museum Collection the segmentation between the pterygoid and hemipterygoid is not only complete, but a perfect joint has formed between the two. The heniipterygoid itself hm not yet fused with the palatine, but articulates with it by suture. Its form is that of a short triradiate spike extending forwards above the proximal end of the palatine, which underfloors it, to overlap the extreme posterior end of the vomer, which, as previously remarked, articulates for the most part with the palatine. Iu the Grebe, in the youngest skulls, segmentation has not yet taken place ; but at the point where this is about to happen there is an indication of a fracture, having jagged edges similar to that figured and described recently in the Impennee, only that in this case the separation is less distinct. In a nearly adult Grebe the form and relations of the hemipterygoid agree exactly with those of the Diver just described. In the Tubinares, it will be remembered, the form of the hemipterygoid differed from that just described. The denta y does not appear to undergo any appreciable change of form between nestling and adult periods. The upleninl is precisely similar in form, both in Grebes and Divers. It resembles a flattened cone, the base contributing to form the ventral border of the jaw. The cormid is at first rod-shaped, then turns abruptly upwards and expands into a flattened trowel-shaped blade, which remains more or less distinct throughout life. The angulare is only just distinguishable as an independent bone. The supra-anqulare can be distinguished as a separate bone only in the youngest of the Grebes in the Museum Collection. 1899.1 OSTEOLOQY OE THN PPQOPODES. 1033

iv. THE VERTEBRALCOLUMN. The vertebra seem to resemble those of the Steganopodes more nearly than of any other group, and, amongst the Steganopodes, they most nearly approach those of Phalnorocorax. They are quite different from those of the Impennes or Tubinares. They can, however, be at once distinguished from those of Phalacrocorm, in that the thoracic vertebra are heteroccelous j but they differ also in other respects. The odoittoid Zigamejat of the atlas is not ossified. The neural arches of the anterior cervicals are not, like those of the Irnpennes and Tubinares, deeply notched posterior1.y. In the Colymbithey are sharply truncated and very broad, in the Podicipides they are, as in Phalacrocorm, marke I bg a slight notch j this notch, however, is cot out of the coalesced buses of it pair of hgperapophyses and lies behind the postzygapophyses ; ordinarily such a notch is formed by cutting away the neural arch itself so as to leave the postzygapophyses as a pair of articular surfaces, each at the termination of a A-shaped fork. The hyperapophyses of these vertebm in the Divers take the form of stout pillars, grooved at the top. In the Grebes the pillars become mere tubercles placed close together and deeply grooved superiorly. In this they resemble the vertebrm in the same region of Phrdmrocorax. The neural arches of the posterior cervicltls do not present any very noticeable features. In the Colymbi the 5th to the 10th vertebra bear catapophyses, which, rapidly converging, give place to hypapophyses. These run backwards to the extreme end of the centrum in the form of a strong median keel. The vertebrae 14and 11-13 bear cntapophyees. In the Podicipides the cervical catapophyses from the 3rd to 13th vertebra form deep tubular grooves for the carotids, recalling those of Plotus and l'halacrocorax. The cervicals 1-3 and 16-23 bear well-developed hypapophyses. The thoracic vertebra-and the last cervical-in the Divers are all free, save the last, which is anchylosed with the synsacrum. 1 to 5 bear median hypapophyses, with broadly exparided free ends, a8 in Alcidse and some Impennes, e. g. Pygoscelis. In the Grebes the laet ~mvicaland the thoracic8 1-4 are anchylosed to form one mass ; the 5th thoracic is free, but the 6th and 7th are fused with the synsacrum. The synsscrum of the Pygopodos is remarkable for the extraordinary lateral compression which it has undergone, accompanied by an almost complete suppressiou of the di- and parapophysial elements. Pleurosteal elements appear to be wanting. Prom the evidence obtainable from the synsacral region of a nestling Grebe we may perhaps be justified in holding that the synsacrum of tbe adult includes some 15 to 17 vertebm. Of these the 1st is thoracic, the next 4 are lumbar, then follow 3 lumbo-swral, 2 sacral, and 5 or 6 caudal. The 3rd and 4th lumbar bear small nipple-like parapophysial processes at the base of the neuron,behind t hese follow, as just stated, 3 lumbo-sacraland2 sncial. 1034 MIL W. P. PPCBIFT ON THE [Dec. 19, These last bear only a roughened diapophysial surface on the neuroid. There is no indication of a pleurapophysial element (sacral rib). If these two vertebra are really sacral then they lie more caudad than usual, being behind the acetabulum and directly opposite the middle of the ilio-ischiadic foramen. The characteristic lumbar enlargement lies between the 2nd lumbar in front and the 1st sacral bebiod. The free caudal vertebre vary from 6-7 in number, including the pygostyle. They are very feebly developed in the Grebes. Intercentra occur below the caudal vertebrae both in Colymbi and Podicipides, but are reduced to mere veetiges in the former. The lateral compression of the synsncrum is less marked in the nestling than in the adult ; and the high neural crest of the adult preaoetabular region is wanting in the nestling. v. THERIBS. The anterior anchylosed cervical ribs in the Pygopodes, in their form and position, recall those of P?halucrocorna. In the adult they are completely fused above with a downgrowth from the ventral surface of the anterior qgapophysis and below with the anterior and ventral borders of the catapophyses so as to form D bony canal for the vertebral artery. In the Grebes they are found only from the 2nd to the 9th vertebm, and are comparatively feeble, though long ; those of the 2nd vertebrae are mere vestiges. In the Divers (Colymbi) they start from the 3rd vertebra, but terminate, as a pair of vestigial proceesee on the 10th or 11th ; they differ markedlay from those in tbe Grebes by their great length and thickness, extending back- u ards so as to embrace the catapophyses of the vertebra next behind, when the neck is straigbtened out. The posterior free cervical ribs in the Podicipides are two in number (see next page). The penultimate, borne by the 21st vertebra, is long, styloid, and without an unciuate; that of the 22nd vertebra is longer, extending down to the level of the top of the sternal rib immediately behind it. It bears B large uncinate, but no sternal segment. In the Colymbi there is only one free cervicaI, apparently corresponding to the antepenultimate rib of the Grebe. The thoracic ribs in the Podicipides are 7 in number, the last two being overlapped by the ilium. 1-5, like the last cervical, bear large uncinates ; these are absent on the 6ith and 7th. There are 8 pairs of sternal ribs, the 8th being bound by membrane to the posterior border of the 7th. Thns there is evidence of the loss of at least one pair of thoracic ribs. It should be remarked, by tbe way, that the 7th pair of sternal ribs do not articulate with the sternum. The thoracic ribs in the Colymbi number 8 pairs, all but the i~stof whicb articulate with the sternum. The lnst 3 pairs are overlapped by the preacetabular ilium, The 8th pair are mere vestiges. *1899.] OSTEOLOQY OF THB PYGOPODBB. 1035 It is probable that the long styloid free rib of the lmt cervical or cervico-thoracic vertebra was originally larger and connected with the sternum by means of a sternal rib ; in other words, this represents a thoracic vertebra which has been transferred to the cervical series by the 108s of the sternal segments and its ribs. Thus, what i8 now the first was earlier the second thoracic vertebra snd rib. In the Podicipides this transfereuce of vertebra from the thoracic to the cervicd series iR still more marked, inasmuch w what now fornis the first thoracic vertebra and rib in the Diver is in the Grebe the la& cervical. This seems the most satisfactory way of exphining the presence of the long free ribs in both Diver and Grebe, and wherever else they occur. The transference of the %nd pair in the Grebe seems to have been comparatively recent, inwmuch as the unciniite is still retained. If this interpretation be correct, mid it is one which was, I believe, originally put forward by the late Prof. T. J. Parker, then one more thoracic segment is represeated in the Grebe than in the Diver, inasmuch as what now answer8 to the 1st thoracic of the Grebe really represents the 3rd, and what now answers to the 6th-the last vertebra now connected with the sternum by a sternal rib- represents the 8th thoracic vertebra. The 6th and 7th vertebrae have already fused with the synsaerum. The 7th (tr9th) ceases to be conuected with the sternurn, and the rib of the 8th (=10th) vertebra is represented only by its sternal segment. In the Diver thwe is only evidenee for 9 thoracic vertebrs, the 9th now venturing but a minute stglet partly fused with the preacetabdar ilium, and projecting from its ventral border as a small spine. The presence of these free ribs is exceedingly interesting, they form one of the many links in the chain of evidence, hinted at by Mr. Beddard (l),which goell to show that a ehortening of the sternum has taken place. The ribs and uncinatie in both Grebes and Divers are relatively broad and flat and of moderate length. In all these particulam they differ markedly from the Alcidle, to which the Pygopodes bear a superficial resemblance. In this last group the sternal and thoracic ribs are of great length. Especially is this the case with the hindmost ribs, which nre of enormous length, extendiug backwards so a8 to pruject beyond the level of the free end of the pubes. vi. THE STERNUMAND PECTORALGIRDLN. The sternum of the Colymbi is very long and bears a superficial resemblatce to that of some Alcids. It can be immediately distinguished therefrom amongst other things bv the shallower rltrina, the feebly developed spina externa, and the large linguiform metasternuin, which projects considerably beyond the postorior lateral processes. The sternum of the Podicipides differs very markedly from that of the Colymbi. In the first place, it is conspicuously shorter. In 1036 YR. W. P. PYCRAET ON THE [Dee. 19, Podicip jluvitctilis, for instance, the width across the posterior lateral processes may equal the whole lrngth of the corpus sterni ; in other words, the sternum may be as broad as long. In the Divers the width across the widest part is about one-third the total length of the sternum. There is no spina externa nor interns; instead, this region of the sternum is deeply hollowed. The lower lip of the coracoid groove is very large, making the groove exceedingly deep. In the Diver this lip is not greatly developed. The metasternum is deeply notched and not produced backwards into a linguiform plate as in the Divers. The anterior lateral processes are larger and project forward. In the Diver they are sbarply truncated, the free anterior border sloping distinctly backwards. The coracoid is short and straight, both in Grebes and Divers. In the former, the epicoracoid is marked by a wide articular surface running transversely across its ventral aspect. The procoracoid process is absent. In the latter the broad articular surface is absent on the ventral aspect and there is a sniall procoracoid process. In both there is a well-marked processw lateralis. There is no supracoracoid foremen, as in the Alcidse; the posterior free border of the epicoracoid is almost knife-like and not, as in the Alcidm, squarely truncate. The scapula, as compared with that of the Alcidae, is relatively short, and has but a very narrow transverse articular surface, instead of a very wide one as in Alcidae. In the Podicipides there is a well-marked acromion process projecting downwards from the shaft beyond the level of the coracoid articular surface. The clavicle is not provided with an external lateral facet fw articulat,ion with the coracoid, as in many Steganopodes and AIcidae. There is a small hypocleideum. The right and left limbs of the clavicle are very broad and laterally compressed in the Divers. In the Grebes the upper free end of each limb is pointed, and runs along the antero-internal border of the scapula.

vii. THI PNLVICGIRDLE. The form of the pelvic girdle in the Pygopodes is unique amongst living birds. Its most characterist.ic feature is the extraordinary elongation and lateral compression which has taken place. Although the epnsacrum has been involved in this compression, it is not, at first sight, so marked as in the innominate bones. The preacetabular ilium is small and narrow, and widely separated from its fellow of the opposite side, but is not otherwise very remarkable. The postacetabular ilium, however, takes the form of a broad, flat, almost or quite vertical lamina. This in the Grebe meets its fellow of the opposite side, in the Diver is separated by the knife-like ridge formed by the neural spines of the anchylosed synsacral vertebrae. The ilio-ischiadic foralneu is moderately large; the obturator foramen in the Colymbi remains permanently iu connection with the fissure of that name, in the Podicipides the 1899.1 OSTElOLOQP OF TEN PYGOPODElS. 1037 foramen is shut off from the fissure by a bar of bone. The pubis is long aucl rod-shaped throughout in the Podicipides, but becomes spatulate at its free eud in the Colymbi. There is no pectineal process. The preacetabolar ilium, the ischinm, and pubis become more or less completely ossified at a much earlier date than the postacetabular ilium. This last is as yet for the most part still cartilaginous. The separate elements ot’ the innominate bones are still very distinct.

liii. ‘]‘HE PEcmaaL LIMB, The wings of the Grebe and Diver bear a very close similarity, and perhaps more nearly resemble those of Phalacrocornx than of any other group. The wing OE the Diver can be readily distinguished from that of the Grebe by the great relative length of the metacarpals. As Shufeldt (18) has pointed out, the Divers in this particular probably stand alone. Phe delto-pectoral crest is larger in the Diver, and the fossa for the brachialis internus is deeper. There is no ectepicondylar process nor subtrochauteric pneumatic fossa. The clelto-pectoral crest in the Divers is separated from the crista inferior by a deep gorge-the plununa iiztertuberculare j this in the Grebe is represeuted only by a shallow depression. The coraco- humeral groove takes the form of a deep pit veutrad and distad of the caput humeri. The foreccrna iu the Grebe i5 nearly as long as the arm, considerably less so in the Divers. In the manus the great length of the metacarpals in the Colymbi has already been commented on; the 1st phalanx of digit II. in the Divers is relatively shorter and broader than in the Grebes ; the same applies to the remaining phalanges. The carpus does not seem to call for any special remark. For further details concerning the fore limb, see Key (p. 1041).

ix. THEPELVIC LIMB. The pelvic limbs of the Grebe and Diver bear an exceedingly close resemblance one t,o aiiother, but differ in almost every particular from those of any other group. The femur is very short and thick, with a strong dorsal curve. Its proximal and distal extremities are greatly elongited transversely. The head lies rather below the level of the antetrochanter, and bears a deep fossa for the ligaiiientum teres. The fibular condyle is of great size, and lies considerably below the level of the tibinl. The tihio-tnrszw is remarkable for tha enormous development of the onemial crests, which form a large pyramidal process projectiug vertically upwards beyond t’he fenioral articular surface. This PROC. zooL. soc.-18~, NO. ~~~1711. 67 1038 MR. W. P. PPCRAFT OR TIIB [Dec. 19, process in the Diver (fig. 1) may esceed the femur in lengkh. In the Grebes (fig. 2) it is not more thou half as long. The ecto-

Fig. 1. Fig. 2.

Outer aspect of the pelvic limb of Col.pynlbw. sepfent&uaZis (fig. 1) and Podicipa crktntns (fig. a), adult. ec, cnemial crest; p., patella; J,femur; jh., fibula; t., tarsus.

and entocnemial crests bear about equal shares in the formation of this cnemid process. 18!19.] OSTEOL00Y OF TRE PYUOPODES. 1039 The Tubinares and Alcids both develop large cnetnial processes, 1% hich, as in the Pygopocles, project vertically beyond the femoral articular surface. But these never attain the size of those of the Pygqocles, and differ, moreover, in form. In the Alcida the ento- and ectocneminl crests bear about equal shares in the formation of the process, hut the former starts suddenly from the shaft just below the head of the tibia, in the P-ygopodes it arises near the middle of the shaft and more or less gradually increases in size, and in the Tubinares it arises as in Alcida, but immediately expands into R more or less flabelliform plate.

Pig. 3.

Outer aspeut of tbe pelvic liwb of Podicip cristatzls, neatling.

p.t., proximal tnrsnl ma88 ; other letters rn in fig. 1 & 2.

About the exact homology of the great cnemiol crest of the Pygopodes there seems to be some doubt, even now. According to Bhnfeldt (18) it is to be regarded as representing the olecranon of the ulna, and both are to he treated “as mere extensions of the shaft of the bones” to which they belong. The cnemial crest, or “rotular process,’’ is stated by hini to have a separate centre of ossification, separate from that of the tibia1 epiphysis. The patella, which has been held by 67” 1040 YR. W. P. PYCElAFT ON THE [Dee. 19, Vicq-d’Axyr (20), Owen (13), and others to be the homologue of the olecranon, is considered by Yhufeldt as a sesamoid only. Prof. D’Arcy- Thompson (19) sees, apparently, like Selenka (15) and Flourens, in the cneinial crest of the Grebes and Divers nothing more than ‘‘ the upper extremhy of the tibia.’’ In a preparation in our Museum Collection, of the pelvic limb of a nestling Grebe, the cnemial crest forms a part of the t,ibial epiphysis in which a ceutre of ossification is just making its sppearmce (fig. 3, p. 103‘3). From this it would appear that the process in question is really only a greatly elongated epiphysis. The jibula in the Colymbi extends downwards to the level of the superior border of the extensor bridge ; it terminates in the Podicipides near the distal 4 of the tibio-tarsus. Thepatella in the Grebe is a very large, laterally compressed pyramidal bone, the apex projecting above the level of the cnemial process, whilst its inner surface is more or less closely applied to the outer border of this process. Its base forms a longitudinally elongated glenoicl surface for articulation with the femur. In all the skeletons of Colynibi in the Museum Collection, unfortunateiy, the patella is missing. According to Shufeldt (18) and others it is, however, represented by a small flake-like bone. Prof. D’Arcy Thompson (19) holcls that the patella proper of the Divers has fused with the cnemial procws, and that the small patelliform plate is to be regarded as a sesamoid, and not 11s the homologue of the free patella found in Podicipes, Hespwornis, aud other forms. This is a point which could probably be settled by an examination of nestlings or embryos. Amongst the Alcidae, e. g. LTtG u&, the patella is more or less quadrate in form, and articulates by the lower half of its anterior surface wit,h the apex of the cnemial process, this being ver.y much less developed than that of the Grebe and Diver. Thus, the upper half of this border serves as a further exteusion dnrsacl of the process it’self. From t’his it will be remarked that the position, size, and form of the patella, and t,lie development of the cnemial process, in the A1cid:e is distinctly different from that of the Colymbi, as also, it will be remembered, is the form of the pelvis. In all these particulars it will be noticed that, though there is a tendency in the Alcids to modification along the same lines,- to a convergence of characters, due to siruilar methods of progression, resulting in a similar upright carriage when on land,-tht? Alcids are less specialized than the Colynibids, which possibly had its origin in that of the stock of the Cretaceous Xesperor&. The pelvic girdle and limb, in coniuion wit11 the rest of t.he skeleton, of this bird, are, as is well-known, nlrnost indistinguishable from those of the modern Colymbi. Indeed, when we eliminate the presence of teeth, the Ratite sternum, vestigial wing, and cotaplete ilio-ischiadic fissure, the only points of dit€erence appear to he such as serve to distinguish species one from another. 1899.1 OBTMOLOQP OF TEE PYBOPOUES. 1041 The form of the patella and cnemial process of the Alcids more nearly resembles that of the Impennes, as does, to a lesser extent, the pelvis. The form of the tarso-metatarsus in the Grebes and Divers is very similar. That of the Grebe may be distinguished from the Diver by the larger size of the intercotylar tubercle and the great depth of the inner glenoid surface for the inner tibio-tarsal condyle. In both groups the tarao-metatarsus is much compressed laterally, and the ectotrochlea is mucL reduced. The hypotarsus is simple. Other characters will be found in the Key (p. 1044). The phalanges of the toes are much flattened dorso-ventrally, the ungunl phnlunx especially so. The 4th digit is longer than the 3rd. x. SUMMARY. The present paper affords good evidence in favour of the views of Beddnrd, Purbringer, Gadow, and others who hold that the Grebes and Divers are closely related, but refuse to associate them with the Auks and Gulls as was done by Huxley and others. The Pygopodes (=the Colymbi of Beddard) seem to be nearlg related to the Tubinares, the Impennes, aud the Steganopodes : but, as Mr. Beddard remarks, “any comparisons bristle with ditiiculties.” That Heqierornis rightly belongs to this sub-order there cau no longer be any doubt, after Prof. D’Arcy Thompson’s admirable memoir; there is one point which has apparently escaped the notice of this writer, however, \\it11 regard to the pelvic girdle of Hesperorniu. This differs from that of both Grebe and Diver, in that the pre- and postacetabular ilium form one great, vertical and laterally compressed blade of very considerable depth. In the Grebe and Diver the preacetabular ilium takes the form of a narrow blade, twisted so as to lie in an obliquely horizontal position. Furthernrore, ZTesperornis seems to be peculiar in that the innominate bones meet throughout in the mid-dorsal line, above the neural crest of the synsacrum; in this particular, however, it approaches the Grebes, where the postacetabular ilium behaves in this \jay j similarly it agrees aith the Grebes in the shortening of the sternum and the large size of the patella. These last two points, however, must be regarded as coincidences rather than indications of affinity; that is to say, Hesperornis must not on account of these points be regarded as more closely allied to the Grebes than to tbe Divers. Indeed, its sternum differsmaterially from that of both these families in that it was keelless, whilst the patella differs from that of the Grebe in being pierced by a foramen for the ambiens. But these and other points will be found exhaustively discussed in the memoirs of Marsh and Dhy Thompson. Mr. Beddard regards the Grebes and Divers a8 representing two 1042 MR. W. P. PYCRhPT OR THE [Dec. 19, families ; Dr. Gadow would regard these as of subordinal value. Which of the two views will become ~lt~imatelyadopted remains to be seen. Probably the first is a sufficiently wide aeparation. Finally,-and it hdalmost esraped mention,-the skeleton of the Pygopodes is non-pneumntic.

xi. KEYTO TEE OSTEOLOGY‘OB TEE PYGOPODES. A. SKULL. (Plate LXXII.) Holorhinal and schizognathous ; nares pervious ; vomer cleft posteriorly ; basiptery oid processes ahent ; lachrymal small, feebly developed, not extending downwads to joic the quadrato-jiigal bar; quadrate with an elongate orbital process ; maxillo-palat,ine processes in the form of horizontal lamina, never extending as far backwards as the scroll-like antero-internal border of the palntine ; basiteinporal plate of the parasphenoid with an iutlated anterior border converting the Eustachian grooves into tubes, with a median aperture below the parasphenoidal rostrum ; temporal fossre more or less well developed. Deritary suture of mandible tending to dieappear in the adult. Angulare truncated. A. Supraorbital grooves very deep, with a rvell-cleveloped ledge ; temporal foasre wide, separated one from anotlrer superiorly by a median Bagittal ridge ; lachrymal more or lesj completely fused with the nasal ; vomer grooved and laterally expanded dorsally ; with a deep median, veiitral keel, and with a strongly marked ventral keel in front ofthe parasphenoidal rostrum ; Eustachian grooves never completely closed ; large postorbital and paroccipital processes ...... COLYMBIDB. (Only one genud-Colymbu.) B. Snpraorbital grooves feebly developed or absent ; lachrymal free, not roiecting posteriorly from t.he sides of the supraorbital margin ; vomer kkh-shaped ; Eustachian grooves completely closed ; poatmbital imd paroccipital processes obsolete ...... PODICII*EDID.%.

Key to the ffeneraof the Fanlily Podicipedids’. A. Without a broad biiid, overhanging postorbital procese. GROUPa. (Type P. oristatus.) With a wide and distinct temporal fossa. and strongly msrlied cerebral prominence ; postorbital region of the frontal marked by a deep scar, for the ternporalis muscle, the soprior border of which has 1% rough edge; upper jaw longer tliun cranium.

I do not feel justified in attempting to form “ Keys” to the species. either for the skull or any other part of’ the skeleton, of‘ the forms couiprisiwg the two sub-families dealt with in this paper. Inasu~uchas of the Colyinbidz I Lave 3nlg two species, C. glacialis aud C. septentrionalis, and these are easily recognizable by the difference in size alone. In the Podicipedidz I have only 8 out of a powible 19 species of the genus Podicipes ; only one skeleton of Echntophorus, and no bones whatever of Podylimbzw. From what I can gather from our inaterial, the difference between the three geuera recognized in the British Museum Catalogue vol. xxvi. is very slight, and that between the species com rising these genera is even less. The genus Podicipcs seems to divide itself into two groups-one of the type of 1’. Jisuiutilis, and one of the type of P. crzitutus. The differences upon which such separation rests concern the skull only, and de end mainly upon size; the smaller species having a relatively shorter an$ wider skull, and ill-de5ned temporal foam. 1s99.1 OSl'EOLOGY OF THE PYGOPODEEI. 1043 UnonP b. (Type P. $u&utdis.) Temporal foam ill-defined; cerebral prominence mall; fossa for insertion of temporalis muscle not deeply excavated on postorbital region of frontal8 ; with a small postorbital process ; upper jaw not longer than cranium. Podicipea. B. With a broad bifid, overlianging postorbital process ...... Rchmophmzis.

B. VERTEBRE. Emargination of neural arches of cervical vertebra caudad, slight, never extending forward beyond the level of the postzygapophyses ; anterior cervical ribs of reat length ; all cervicala save atlas and axis with a bony vertebrarterial canal cfosed externally by the cervical ribs ; Rynsacrum elongated, and laterally compressed, 80 a8 to be almost styliform. A. Neural spines of anterior cervicals in form of long median ridge ; hjpera- popbyses of anterior cervical8 widely separated ; neural arch of 2nd, 3rd, and 4th with a rounded free posterior border. All thoracic vertebroe free, the 1st-3rd with long kshaped h papophyses; 6 free caudals, excluding the pygostyle ; pygostyle broad an$ strong ...... COLYYB~DB. B. Neural s ines of anterior cervicals almost obsolete ; catapophyses of anterior cervicaE meeting in the middle line ; last cervico-thoracic and thoracic 1-4 anchylosed; free end of hppapophyses scarcely, if at all, laterally expanded ; 7 free caudals, excluding pygostyle; pygostyle feeble. PODICIPEDIDZ. It is almost impossible to distinguish, in the synsacrum, lumbar, lumbo- sacral, and sacral vertebrte. The lumbar region has apparently undergone a great shortening. There are no vertebra which can be distinguished-in the dried skeleton-as sacral. The postsynsacral region-that lying behind the acetabuluxn-is of great length. The vertebral formula must therefore stand as follows:-

Syn. sc. ---h-- Y Cv. 13. Cv. th. 1. -Tb.6+2. L. Lb. sc. SC.?~. Cd. 10+7=44. 8 -17 Co Zy mbide. Syn. sc. __-- A--- Y Cv. 90. Cv. th. 2. Th.6+1. L. 4. Lb.sc.3. sO.3. Cd.6+7=51. v c 13 Podicapedida.

C. STERNUMAXD PECTORALGIRDLE.

Corpus stemi with a pair of notcheR posteriorly ; coracoid grooves dbep ; metastemum either notched posteriorly, or broadly linguifom, and projecting far beyond posterior lateral procesEes ; anterior lateral processee not projecting forwards beyond tho level of the coracoid grooves ; no supracoracoicl foramen ; clayicle not articulating bF a glenoid snrl'ace with acromion of scapula. A. Greatest width of sternum acroEs anterior lateral processes rather lass tha n 4 the length ; coracoid groove moderately deep ; spina externa sessile and bifid ; keel very ehallow ; metaeternum linguiform, pro'ecting far beyond level of posterior lateral processes ; precoracoid very smah ; posterior lateral processes of corucoid large and spinose ...... COLYYBID~. 1014 MB. W. P. PYCRAFl’ ON TEE [Uec. 19,

B. Greatest width of sternum across posterior lateral processes more than +- sometimes equal to-its whole length ; metasternuin deeply notched, not projecting so far back as ends of posterior lateral processes; coracoid groove very deep ; no spiw ezteriza or iirteriza ; no wecoracoid ; posterior lateral processes of coracoid short and blunt ; scapu!a with long acromion. PODICIPEDIDAC. D. PELVICGIRDLE. Greatly elongated and compressed laterally behind the acetabuluni ; closely resembles that of the Cretaceous Hcspcroriris, and differs entirely from that of any other living Carinate bird, the postncetabulnr ilium being represented by a nearly vertical plate of bone ; synsacrum almost styliform ; pleurostea caudarl of acetabulum wanting ; parapophysos vestigiul. A. Free end of pubes spatulate ...... COI.YIBIDR. B. Free end of pubesnot spatulate ...... PODICIPEDII:.U.

E. PECTORALLIM~. The bones of the wing are relatively long ; the humorus has a moderately welldeveloped delto-pectoral crest, the corilco-humerd groove takes the form of a deep pit; there is a shallow fossa for tho brachinlis infermrs; the fosu suhtrochairtericus is blind. The forearm is uearlg or quit,e RS long a8 the arm. Mc. 111. is long, Blender, and runs parallel with Mc. 11. A. Carpo-metacarpus much elongated ; Mc. I. TerX long, t ns long as Mc. 11. ; manus as long as forearm ...... COLYAIBIDR. B. Oarpo-metacarpus not greatly elongated ; Mc. I. very short ; manus shorter than forearm...... Yoorcriwornm.

F. PELVICLIMB. (Figs. 1-3, pp. 1038-9.) Femur very short; tibio-tarsus with an enormous cnemial crest; tarso- metatarsus laterally compressed ; outer toe longest ; ungual phalanges mucli flattened. A. Cnemial crest aslong as or longer than femur; fibula extending downwards to the tarsus, or very nearly so; the ecto-trochlear foratlien of tarso- metatarsus tubular ; hypotarsus simple ...... COLYNBIDB. B. Cnemial crest shorter than femiir; fibula brminating near the lower + of the leg ; hypotarsus simple ; ectotrochltur jbrumen in the form of 8 gr0OVe...... PODICII’EDIDb~

xii. LISTOF WORKSREFEPBED TO OR CONSULTED. 1. BBDDARD,F. E.-Structure aud Classification of Birds, 1898. 2. BRANDT,J. B’.-Beitriige zur Kenntn. der Naturgesch. der Vogel. MQm. Acad. Imp. des Sciences St. PQtersbourg, a&. vi. vol. iii. 1839-40. 3. COUEB,E.-The Osteology of the Colymbus torquatus; with Notes on its Myology. Mem. Bost. Soc. Nnt. Hist. yol. i. part ii. 1867, pp. 131-172. 1899.1 OSTEOLOGY OP THE PYGOPODES. 1046

4. COIJES,E.-’l’he Crania of Colpbus torpuatus and C‘. aclnntsi compared. Proc. Acad. Nat. Sci. Philad. vols. xvi.-xvii. 1864- 65, pp. 21-22. 5. F~RB~IN~ER,M.-Untersuch. zur Morphol. und Systemat. der Vogel. 11. Allgem. Theil, 1888. 6. GADOW,€1.-Bronn‘s Thier-Reich, Ed. vi. Vogel, 1891. Auatoui. Theil. 7. GADOW,I-I.-Ibid. Syst. Theil, 1893. 8. GABROD,A. H.-On certain Muscles of the Thigh of BirdB. P. Z. S. 18i3 (Part I.)’ 1874 (Part, 11.). 9. HUXLEY,T. li.--Uu the Classification of Birds. P. Z. S. 1867. 10. LYDEKKER,It.-Cat. E’oss. Birds Brit. Mus. 1891. 11. MECPEL,J. F.-Syst. der vergleich. Anat ,Theil 2, Abtlieil. ii., 1825, p. 129. 12. MILNE-EDWARDS,A.--ltechrches pour sen ir h 1’HLtoire des Oiseaux Fossilrs de la Prance, rol. i. pp. 278-300 (1867-68). 13. OWEN,R.-Todd’s Cyclopedia of Anatomy, vol. i, pp. 286-7. 14. PPCBAFT,W. €’.-Contributions to the Osteology of Birds (Tubinares). P. Z. S. 1899, p. 381. 15. HELENKA, E.-Bronn’s Thier-Reich, Bd. vi. Vogel, 1S91. Anatom. Thcil, p. 83. 16 SHUFLLDT,B. W.-Concerning the Taxonomy of the Rorth Anierican Pygopodes, based upon their Osteology. Jouru. Anat. & Phgsiol. vol. xxvi. 1892, pp. 199-203. 17. S~XUFEL~T,K. W.-Contributions to the Comparative Oste- ology of Arctic and Sub-Arctic Water-Birds, Part VI. Journ. Anat. & Phpsiol. rol. xxir. 1890, pp. 169-187. 18. SHVFBADT,B. W.-Concerning Rome of the Forms assumed by the Patella in Birds. Proc. U.S. Nat. Mue. vol. vii. 1884, pp. 324331. 19. THOMPSON,D’ARCY.-OI~ the Systematic Position of Ilesper- o~izi8. Studies Mus. Zool. Dundee, 1890, vol. i. art. s. p. 97. 20. VIcQ-DyAzYR.-M6moire sur lee Rapports qui se trouveiit eutre les usages et la structure des quatre extrt(mit8s dans 1’Hoinme et dans les QuRdrupAdes. Hist. de l’acad. Roy. Hci. 1774, p. 261.

EXPLANATION OF PLATE LXXII. als.=alisphenoid. j‘J:=floccular fossa. 6t. =basitemporal latform. .jr. =f t-ontal. Ir.oc. = bnsioccipita! hpt.= hetnipterygoid. 6.8. =basisphenoid. ZJ: =lachrymal foramen. b par para sphenoidal rostrum. me. =meatus internue. c.=occipital condyle. m~.=tuesethmoid. ep.=cerebellar promiiience. n. =nasal.

c.~.= coronal ridge. ~ Il.pmx.=nasal process of premaxilla cl.=dentary. q.0. =opisthotic. ep.0. =epiotic. p.=porietal. ex.0. =exoccipital. pa.=palatine. 1046 ON THE OSTEOLOGY OF THE PYGOPODEB. [Dec. 19.

pro.=prootic. s0.2. = supraorbital ledge. pt.=pterygoid. sy. = squamossl. g.=quadrute. q.p.= squalnosal prominence. S.U. =supra-angular. sqpw. =squumoao-parietal ring. a.c.=sagittal crest. t.f.=temporal fossa. S.O. =supraoccipital. h.f =trigeminal foramen. sag. =supraorbital groove. w.=vomer. Pig. 1. Dorsal aspect of the skull of Podicipes c&talss (p. 1020). to show the well-developed sagittal crest, temporal fossle, coronal ridge, squanioso- parietal rings, the freu lachrymal, and the feeble supraorbital groovea. Fig. 2. Doreal aspect of the skull of Colymbus septentrionalis (p. 1020), to contrast with fig. 1, with the great deveiopniect of the supraorbital grooves and ridges and the lachrymal fontanelle. Fig. 3. Lateral aspect of fig. 2 (p. 1021), showing the conspicuous cerebellar prominence, tem oral fossa, and supraorbital ledge. Pig. 4. Lateral aspect of tEe skull of a nestling Podieipcs crist'utus, outer view (p. 1030), to show the unclosed sutures. Fig. 5. Lateral aspect of the skull of a nestling Podicipes cristatus, inner iiew, to show the unclosed sutures. Fig. 6. Ventral view of skull of an adult Cul@~ks septe?ttrio?ialzs(p. lo%), tu show the schizognathoua palale. Fig. 7. Lateral aspect of a portion of the pterygoid and palatine of CcZphs gluciri2is (p. 1032), to show tlie hemipterygoid.