ae Re See Wat lf eal ‘ Huh is vA Hoe hey ‘, Ae ANS aoe Nanuet hie mn MN uets NE Abi (ihe TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. VOLUME XVII. LONDON: PRINTED FOR THE SOCIETY : SOLD AT THEIR HOUSE IN HANOVER-SQUARE; AND BY MESSRS. LONGMANS, GREEN, AND CO., PATERNOSTER-ROW. 1903-06. [166 /§ Hitt ants Lae NO A PRINTED BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET. Hat ; Ik JUDE NY, WO CONTENTS. . A Revision of the Amblypodia Group of Butterflies of the Family Lycenide. By GrorGcE T. Beraune-Baker, F.L.S., F.Z.S., &c. (Plates1-V.) . . . pagel Notes on the Transformations of some South-African Lepidoptera. (Continued from the ‘Transactions of the Zoological Society of London,’ vol. xv. pt. 6, April 1901.) By Lieut.-Colonel J. Matcoum Fawcerr. (Plates VI—-VIII.) 165 A Monograph of the Fishes of the Family Loricariide. By C. Tare Reeay, B.A., PZ SU (Plates NXT) owen ie a ee ok) ae oi A Contribution to our Knowledge of the Varieties of the Wall-Lizard (Lacerta muralis) in Western Europe and North Africa. By G. A. Boutencsr, F.R.S., VolPVGS0 (IQ ENGS DOU POD eg) ges 6 Br Gy gy o SOIL . On the Intestinal Tract of Mammals. By P. Cuatmers Mircue tt, JZ.A., D.Sc., Secretary to the Society . 437 Fourth Contribution to the Ichthyology of Lake Tanganyika.—Report on the Collection of Fishes made by Dr. W. A. Cunnington during the Third Tanganyika Expedition, 1904-1905. By G. A. Boutenerr, F.RS., V.P.Z.S. (Bates XOXNOX. ILI: )) iy manera esata re Pc Le 5 en em Iii ILI: OE aw IeAjorsss Copel Ma WOl SO 5 5 ee se 6 oo GO Indexeois Species stQceh i.e" i). rn Se ene a eae ans amen OU) TRANSACTIONS Or THE ZOOLOGICAL SOCIETY OF LONDON. Vou. XVII.—Parr 1. = st] i, alan Instltution ay , & os R > Uj 8 BEP 26 ov AS A x 2 awe Nations WA LONDON: PRINTED FOR THE SOCIETY, SOLD AT THEIR HOUSE IN HANOVER-SQUARE ; AND BY MESSRS, LONGMANS, GREEN, AND CO., PATERNOSTER-ROW. August 1903. Price 30s. Taylor and Francis, Printers, | [Red Lion Court, Fleet Street. TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. VOLUME I. (1883-1835, containing 59 Plates) . VOLUME IL. (1835-1841, containing 71 Plates) . VOLUME III. (1842-1849, containing 63 Plates) . VOLUME IV. (1851-1862, containing 77 Plates) . VOLUME V. (1862-1866, containing 67 Plates) . VOLUME VI. (1866-1869, containing 92 Plates) . VOLUME VII. (1869-1872, containing 73 Plates) . VOLUME VIII. (1872-1874, containing 82 Plates) . VOLUME IX. (1875-1877, containing 99 Plates) . VOLUME X. (1877-1879, containing 95 Plates) . GENERAL INDEX, Vols. I. to X. (1833-1879) VOLUME XI. (1880-1885, contaming 97 Plates) . VOLUME XII. (1886-1890, containing 65 Plates) . VOLUME XIII. (1891-1895, containing 62 Plates) . . Price 7) 72 es) 39 * No copies of these volumes remain in stock. To Fellows. d. ip a 3 Or Se 13 we 6 + Only complete copies of these yolumes are left in stock. To the Public. s. d. 18 O* ts 1 6* Ou (ef 411 0* S) ey Oyp 619 OF 6 0 OF . 1312 oF al S05, 16 2 OF VAIS ere sOr Continued on page 3 of Wrapper. Di aes ACO tO Nos OF PET AOorOGrOxAl sO CTE TY OF LONDON. I. A Revision of the Amblypodia Group of Butterflies of the Family Lycenide. By Guorce T. Bernune-Baker, /L.S., F.Z.S., &e. Received October 22, 1901; read March 4, 1902. [Prares I.-V.] THE large division of the Lycewnide usually known as the Amblypodia group has been in a more or less unsatisfactory state for many years. Hewitson was the first to make any really important contribution to our knowledge of the species; but his descriptions are so meagre that without his types combined with his figures (the latter are not quite satisfactory without the former) they would be useless. About the same time the brothers Felder, in their valuable work the ‘Reise der dsterreichischen Fregatte Novara,’ added more species to the group and adopted the generic name Arhopala given by Boisduval for the largest section thereof, recognizing that they were structurally different from the genus Amblypodia. This name, however, not having been diagnosed, was practically ignored until de Nicéville again adopted it in his ‘ Butterflies of India,’ since then it has been coming more generally into use. I divide the group into six genera, viz.: Surendra, Iraota, Amblypodia, Mahathala, Thaduka, and Arhopala, the last of which alone contains over 200 species and varieties without considering many synonyms. Mr. Moore has tried to split up these into several subgenera; but I cannot see the use of adopting them, for reasons given below. When even a large number of species like this falls naturally into one VOL, XVII.—PART I. No. 1.—August, 1903. B 2 MR. G. T. BETHUNE-BAKER: A REVISION OF THE genus so that even a tyro could recognize them, and when there is no good dividing- line, I cannot see that any object is gained by attempting to split up the genus. Mr. de Nicéville has suggested spelling the word “ Arhopala” with two “7's,” but, according to the strict canon, I prefer to leave it as it was spelt at its first institution. The section dealing with this group in de Nicéville’s work, already referred to, is by far the most valuable treatise we have, but there are so many extra-Indian species and so much material has been received since the publication of that book, the descriptions of which are scattered throughout many periodicals, whilst a good many new ones appear in this monograph, that the present seems an appropriate time to gather up these fragments and present them as a whole for the student. With the exception of one or two species found in China and Japan and in some of the Pacific Islands, the whole group is confined to the Indo-Malayan region, in which I include N. Australia: the insects appear to be confined, to a very large extent, to forest districts—de Nicéville says they seldom or never go down to streams for moisture, neither do flowers attract them ; and the only way to collect them is to beat the bushes and catch them as they fly up; their flight is rapid but always short. In the present paper I have described a dozen new forms, whilst in 1896 I described thirteen new species in the ‘ Proceedings’ of this Society, and I am figuring a considerable number of species that are in an unsatisfactory and doubtful condition. Primaries : 1, la, 1. Discal or cell-spots. 2. Costal spot. 3. Transverse band or fascia. 4. Submarginal row. Secondaries : 0. Costal basal spot. 1. Basal row (of 4). 2. Median row (of 3). 3. Transverse band. 3 a, first spot; 3.6, second spot; 3c, third spot; 3d, fourth spot; 3e, fifth spot; 3/,sixth spot; 3g, angular spot. 4, Submarginal row. Diagrammatic sketch of wing of butter- fly of the genus Arhopala. When I undertook this work I had no idea that it would be so heavy, and I have to thank many entomologists for valuable help. My most hearty thanks are due to Mr. de Nicéville (whose death I deeply regret and in whom India loses a very valuable observer) and to the late Dr. Staudinger—the former having sent me from India the whole of his valuable collection of this group, including all his types, whilst my old AMBLYPODIA GROUP OF THE LYCENID®. 5) friend Dr. Staudinger likewise sent me the most important portion of his collection ; these two alone furnished me with considerably over 2000 specimens, without which my task would have been well nigh impossible, for, with species so closely related as are many of these, it is extremely difficult to differentiate them without the types, especially when they are not described on the same plan. I have therefore in this revision adopted a uniform plan of description for the genus Arhopala, which will be easily understood on reference to the accompanying figure (Text-fig., p. 2). I have further examined nearly all the most important collections in the kingdom, and through the kindness of Herr Georg Semper, who sent me all his Philippine specimens, I have been able to examine this valuable collection, so that my grateful thanks are due to him, as also to many other entomologists both on the Continent and at home; and I would specially thank Mr. Hamilton Druce, who has helped me frequently on various difficult points. The genitalia have had my careful attention, and have often proved a real help in settling the specific value of certain species, especially in the green division of the genus Arhopala. The subgroups, as a rule, follow the main lines of a certain pattern, and in all cases that I am aware of they have a definite specific value. SuRENDRA Moore. Surendra Moore, Proc. Zool. Soc. Lond. 1878, p. 835, et 1879, p. 142; id. Lep. Cey. vol. i. p. 112 (1881) ; de Nicéville, Butt. Ind. vol. iii. p. 218 (1890). Both wings short and broad. Primaries with the costa evenly arched, more strongly in the female, apex acute. Posterior margin excavated below the apex, then strongly convex, especially soin the female. Inner margin waved: costal nervure ending about the apex of the cell; second subcostal nervule with its base midway between the first and upper discoidal ; third subcostal short, rising nearer the apex of the wing than the cell ; discocellular nervules upright, concave, the middle rising from the upper discoidal close to its base and rather shorter than the lower, submedian nervure straight. Secondaries with the costa strongly arched at the base, then almost straight in the male, but slightly arched in the female ; posterior margin evenly rounded to the tail, then deeply excavated and oblique to the lobe, but in the female it is excavated up to the outer tail; abdominal margin excavated above the lobe, then rounded; first sub- costal nervule rising somewhat near the apex of the cell; discocllular nervules with a slight oblique tendency outwards, straight ; the second median nervule rises just at the lower end of the cell; submedian nervure straight, internal nervure straight, then strongly arched. The male has a shortish tail at the end of the first median nervule. The female has likewise, and in addition has another at the end of the second median nervule. Antenne of moderate length; eyes smooth ; palpi long, slender ; legs short. I have been unabie to trace any record of the life-history of any species of this genus, but Mr. Doherty was apparently acquainted with their eggs, for he writes Ba 4 MR. G. T. BETHUNE-BAKER: A REVISION OF THE (J. A. S. B. vol. lviii. pt. 2, p. 409, 1889) that the genus Swrendra Moore and a number of species of Arhopala have hemispherical eggs covered with white globular tubercles and triangular reticulations, six of which radiate from each tubercle. De Nicéville says (Butt. India, vol. iii. p. 220) “that Mr. P. W. Mackinnon has seen the female of 8. guercetorum laying eggs on a species of mimosa in the Dehra Dun.” The genitalia in this genus are more of the usual “ Lycena” shape, the tegumen being less ample. The hooks, however, are short and stout, whilst the clasps are not so reduced as in the genus Arhopala. The penis is long and slender, with the extremity slightly extended (see Plate IV. fig. 1). After examining a very considerable series of this genus (nearly two hundred are now before me, and I must have examined nearly double that number) I find it is quite impossible to retain the species described from India, Ceylon, and the Andaman Isles; they are, without doubt, one and the same variable species, viz. S. guercetorum Moore, and I believe that this name should also give way to 8. vivarna Horsfield, but I have not sufficient material from Java or Singapore to decide, and I therefore treat them as distinct for the present. In a hundred specimens now lying before me I cannot find one dividing-line, the ground-colour of the males ranging from a bright purplish blue, extending over the greater part of the primaries, to an exceedingly dull violet of small extent; the pattern of the underside varies likewise, often tending to disappear, these features obtaining in specimens from the same neighbourhood. Then as to the females, Andaman specimens (as also the Javan 8. vivarna) are entirely dull warm brown; but I have several specimens from the Khasia Hills that are quite indistinguishable from them, whilst from this locality I have also the common form showing the slightly paler discal patch, and one at least which is quite as pale as the average Ceylonese form named by Mr. Moore discalis. The outline of the wings seems equally liable to variation, for I have males from the Khasias with most decided “ hook-tips ” and others with an almost straight outer margin. Turning now to the amisena group, I find, on careful comparison, that Staudinger’s types of palowna (which I have before me) agree precisely with Hewitson’s type of amisena, the tail being but little different, but much exaggerated in Hewitson’s drawing. lorimel of Doherty has no tail, and therefore I retain it as distinct; it is also bright and has a somewhat different underside. Stimula de Nicéville is, without doubt, referable to S. florimel Doherty, as there is no difference except in the larger size of de Nicéville’s forms. SURENDRA QUERCETORUM (Moore). Amblypodia quercetorum Moore, Horsfield and Moore, Cat, Lep. Mus. E.I.C. vol.i. p. 42. n. 63, pl. ia. fig. 7, ¢ (1857); Hewitson, Cat. Lycenide B.M. p. 14 (1862) ; id. Il. Diurn. Lep., Lye. p. 14 (1865). AMBLYPODIA GROUP OF THE LYCAHNID#. 5 Amblypodia vivarna Hewitson (nec Horsfield), Cat. Lyc. B.M. pl. vii. fig. 76, ¢ (1862). Surendra discalis Moore, Proc. Zool. Soc. Lond. 1879, p. 142; id. Lep. Cey. vol. i. p. 112, pl. xliv. figs. 1¢, 1a? (1881). Surendra latimargo Moore, Proc. Zool. Soc. Lond. 1879, p. 142. Surendra quercetorum var. latimargo Wood-Mason and de Nicéville, J. A. S. B. vol. xlix. pt. 2, p. 235 (1880). Surendra biplagiata Butler, Proc. Zool. Soc. Lond. 1883, p. 147. pl. xxiv. fig. 12, g. Amblypodia quercetorum Staudinger, Ex. Schmett. p. 281, pl. xevi. (1888). Surendra quercetorum Moore, Proc. Zool. Soc. Lond. 1882, p. 251 ; Doherty, J. A. S. B. vol. lv. pt. 2, p. 130 (1886). Surendra quercetorum de Nicéville, Butt. Ind. vol. ii. p. 220 (1890). Surendra biplagiata de Nicéville, op. cit. p. 221 Surendra discalis and S. latimargo de Nicéville, op. cit. p. 222 Hab. Sylhet; N. India; Kangra Valley; Bagheswar; Lower Ramganga; Gori and Kali Valleys; Kumdaun; Madras; Ceylon; Andamans; Burma; Assam ; Kashmir. Expanse, ¢ 28-54, 2 36-39 mm. 3. Upperside: both wings violet-blue; primaries with broad black costa, broader posterior margin, and apex black from the termination of the cell; secondaries with the blue extending only in the median area and just into the cell. ‘Tail tipped with white; lobe very much produced. Underside: both wings greyish brown: primaries with an obscure dark dash in the cell and a narrow elongated spot at the close, two other dark dots on the costa, a dark, very irregular, transverse line midway between the end of the cell and the posterior margin from the costa to the submedian nervure, beyond which is a row of small very dark dots, sometimes very obscure : secondaries with a white dot below the costal nervure near the base, beneath which is an obscure, dark, jagged, interrupted line ; from the apex to the centre of the abdominal margin is an obscure band of dark shading, in which is an irregular whitish-edged fascia from the costa (before the apex) to the abdominal margin; a submarginal row of dark spots, sometimes formed into a lunulated line; lobe-spot very dark reddish brown ; anal area sprinkled with greyish scales. @. Upperside: both wings entirely dark brown, with a pale dash beyond the cell in the primaries. ‘Tails edged with white. Underside exactly as in the male, but with all the markings sharper. On examining very carefully a large series of this species, I find it is quite impossible for me to differentiate diplagiata Butler and latimargo Moore from Moore’s original species ; it is so very variable everywhere that it is useless retaining these names. The same remark applies with regard to discalis Moore: this is a mere aberration which is quite inconstant ; in some specimens the ochreous patch is very prominent, in others less so, and I have others from the same locality in which there is no trace of it. 6 MR. G. T. BETHUNE-BAKER: A REVISION OF THE SuRENDRA VIVARNA (Horsfield). (Plate I. fig. 6.) Amblypodia vivarna Horsfield, Cat. Lep. Mus. E.1.C. p.99.n. 31 (1829) ; Hewitson, Cat. Lyczenidee B.M. p. 18. n. 61, pl. vi. figs. 73, 77 g, 75 (1862). Surendra vivarna de Nicéville, Butt. India, vol. iii. p. 221 (1890). Hab. Java; Sumatra. Expanse, ¢ 29-33, 9 31-24 mm. g. Upperside: both wings rather intense deep blue: primaries with the costa broadly, posterior margin more broadly, apex still more broadly black ; secondaries with the blue limited to just over the median area, the rest of the wings black. Tails two, black, the longer one white-tipped. Underside: both wings very dark greyish brown: primaries with an obscure dark dot near the centre of the cell and a small dash closing it, costa with three dark dots; transverse blackish line curved irregularly and scalloped from the costa to the submedian nervure, a submarginal row of blackish dots, margin obscurely dark, internal area quite pale: secondaries with a small white spot about one-third along the costal nervure, below which is an obscure dark short dash in the cell; from two-thirds along the costa to the middle of the abdominal margin is a narrow, indefinite, obscure band of dark shading edged externally and sharply with white, a submarginal row of dark spots forming an almost lunular line; on each side of the long tail is a slight grey scaling. @. Upperside: both wings entirely dark brown. Underside as in the male, but decidedly paler. In this species the outer margin is much produced in the median area, both in the male as well as in the female, but is a somewhat unusual feature in the former, whilst in the female the tip is much less produced outwards, forming less of a hook-tip. SuRENDRA AMISENA (Hewitson). (Plate I. figs.3 ¢,4&5 9 ; PlateIV. figs. 1 & 1a.) Amblypodia amisena Hewitson, Cat. Lycenide B.M. p. 18, pl. vil. figs. 74-78, ¢ (1862) ; Kheil, Rhbop. Ins. Nias, p. 33. n. 123 (1884). Rapala amisena Distant, Rhop. Malay. p. 277. n. 1, pl. xxiii. fig. 18, ¢ (1885). Surendra amisena Doherty, J. A. S. B. vol. lviii. pt. 2, p. 424 (1889). Amblypodia palowna Staudinger, Iris, vol. 11. p. 131, pl. ii. fig. 3 (1889). Surendra amisena de Nicéville, Butt. Ind. vol. in. p. 222 (1890). Surendra palowna Druce, Proc. Zool. Soc. Lond. 1895, p. 588. Hab. Sumatra; Borneo; Palawan; Balabac; Nias; Malacca. Expanse, ¢ 40-43, 2 39-45 mm. 3. Upperside: both wings dull violet to purplish blue: primaries with apical area very broadly dark brown, tapering very narrowly to the anal angle and less narrowly on the costa: secondaries with blue confined within the upper discoidal and submedian nervures; posterior margin narrowly brown. Underside: both wings warm slightly rufous brown. Primaries with the least trace of a most obscure spot AMBLYPODIA GROUP OF THE LYCENID, U in the cell and an equally obscure line closing the cell, often entirely obsolete; transverse band dark, sharply defined, with its centre curved outwards, thus giving it somewhat the shape of a strung bow, the line slightly interrupted at the neryures; a submarginal row of moderately distinct dark dots; internal area pale. Secondaries with a white spot below the costal nervure ; a third along it, below which is an obscure, irregular, short line across the cell; transverse line not far beyond the cell, very irregular, composed of short, white, interrupted marks from the costa (a third from the apex) to near the end of the internal nervure; these white marks have all a dark diffused internal shading; the dark submarginal row is lunular, interrupted at each nervure and edged internally with whitish or palely; the anal angle has more or less metallic scaling, occasionally entirely wanting, often very slight, but sometimes consisting of two large patches of bluish scales with a small spot at the extremity of the fold. The tail of the type specimens is very short, the margin scalloped on each side of it; this is very variable, however. I have before me some without any tails at all which appear from several localities with the tailed form; then I have others in which the tail is reduced to a short tooth-like excrescence. 2. Upperside: both wings glossed all over with slightly lustrous dull violet-mauve, which makes the species immediately recognizable from all others of the genus. I have a second form which is entirely glossed with steel-blue ; both have precisely the same underside as the male just described. This is evidently a case of a dimorphic female. Dr. Staudinger’s types of S. palowna are before me, and they are absolutely the same as the specimen labelled amisena in Hewitson’s collection, each having slight tails. Mr. Druce is correct in saying, in his Bornean paper (P. Z. 8. 1895, p. 588), “that the characters he mentions, of the tails, will be found to be insufficient to separate anusena and palowna.” Hewitson’s figures show the tails very considerably longer than appears in any specimen that I have seen. I hope that my figures of both forms of the female of this species will prevent further confusion. SURENDRA FLORIMEL Doherty. Surendra florimel Doherty, J. A. S. B. vol. lvii. pt. 2, p. 424, pl. xxi. figs. 17 g, 4 9 (1889); de Nicéville, Butt. India, vol. 11. p. 223 (1890). Surendra stimula de Nicéville, Journ. Bomb. Nat. Hist. Soe. vol. ix. pl. O. fig. 23 (1895). Hab. Wagung, Tavoy District; Burma; Sumatra; Java. Expanse, ¢ & 9, 30-39 mm. g¢. Above rich purple-blue from the costal vein to the hind margin, the blue area outwardly angled at the upper median vein, and occupying nearly half of the 8 MR. G. T. BETHUNE-BAKER: A REVISION OF THE fore wing; on the hind wing it extends from just below the lower subcostal to the submedian vein, leaving the upper part of the cell dark; a narrow black marginal line. Below the ground is light fuscous brown, as in amisena, but with the cell and disc of the fore wing much darker and the basal and apical half of the hind wing deep violet-brown. Fore wing with a short oblique dark streak in the middle of the cell, a larger one across its end, and one or two costal streaks; a transverse discal line cf joined lunules (separated in amisena) from the second subcostal to below the lower median, projecting outwardly below the lower radial vein ; apex widely, and outer margin narrowly, pale fuscous. Hind wing with the transverse discal fascia consisting of a broken dull silvery line on a deep brown ground, an obscure, outer, discal, transverse band, pale on the dark apical, and dark on the pale abdominal ground; an obscure metallic patch in the lower median space. 9. Above dull brown, a slightly paler area in the middle of the disc. Below the dark area of the fore wing is confined to the neighbourhood of the median spaces on the disc, that of the hind wing to a band across the wing from the apex to the hind margin, crossing the end of the cell; a distinct whitish spot basally between the costal and the subcostal veins; the inner transverse line united, crossing the dark area subapically; the outer one consisting of pale lunules, bordered, especially outwardly, by a dark band in which there are two dark subapical spots, the second larger. This species has the hind wing strongly angled at the end of the middle median vein, and quite straight thence to the anal angle; there is no trace of tails or lobes. The fore wing is not falcate in either sex. The egg and venation are as in amisena and quercetorum. It is a very distinct species, and the male is very richly coloured. One male and several females taken on the pass near Wagung, avoy District, at 1500 feet altitude. S. stimula de Nicéville, the type of which is before me, is, without doubt, only a large example of this species. Traota Moore. Traota Moore, Lep. Cey. vol. i. p. 101 (1881) ; Distant, Rhop. Malay. p. 253 (1885) ; de Nicéville, Butt. India, vol. ii. p. 218 (1890). Wings less ample, but strong. Primaries with the costa curved at the base, excavated slightly in the middle, then straight; apex rounded; outer margin convex at apex, then concave; inner margin straight. Costal nervure of medium length, ending about the apex of the discoidal cell: first subcostal nervule curved upwards near the base; second subcostal rising midway between the first and the upper discoidal nervule; third subcostal long, from the subcostal nervure nearer to the apex of the wing than to the origin of the upper discoidal; fourth subcostal (not present in the female) very short, rising close to the apex of the wing; the end of the AMBLYPODIA GROUP OF THE LYCANIDA., 9 subcostal nervure terminates just below the apex of the wing in the male, but at the apex in the female; the lower discoidal nervule rises from the upper discoidal not very far from the origin of the latter; the upper and middle discocellular nervules absent ; lower discocellular upright, slightly concave, rising directly after the origin of the discoidal nervule; second median nervule rising just before the lower end of the cell; submedian nervure straight. Secondaries with the costa evenly curved; the outer margin in the male is slightly scalloped and inwardly oblique, in the female rounded and decidedly scalloped; the submedian nervure is produced into a short tail, generally much longer in the female than in the male, sometimes with a second tail at the end of the first median nervule in the male longer than the other—this tail is always present in the female, but is shorter; sometimes there is a third tail at the termination of the second median nervule, which is the shortest of the three. Anal lobe large and well formed. ‘The costal nervure is curved throughout, but very strongly near the base; first subcostal, rising close to the end of the cell, is likewise curved; upper and lower discocellular nervules slightly concave and outwardly oblique, second median nervule rising near the lower end of the cell; submedian nervure slightly recurved, the internal nervure strongly recurved and long. Antenne rather long, over half the length of the costa. yes smooth. Palpi porrect. Thorax stout. Body fairly robust. The male has a small tuft of hairs just before the middle of the inner margin of the primaries, and a very slight glandular space between the bases of the costal and the subcostal nervures. Larva onisciform, smooth; head small; second segment small, then increasing rapidly to the fifth, afterwards decreasing to the anal segment. Pupa smooth, humped on the thorax; abdominal segments very broad at first, but rapidly decreasing to the tail. De Nicéville says he has only taken specimens of this genus on or near the banian tree, on the leaves and small branches of which they alight. He describes the flight as very powerful and rapid, though the insects soon settle after being disturbed. The triangular shape and strong texture of the wings would lead one to expect that all members of the genus were powerful and rapid fliers. The genitalia are specialised, the tegumen being of a very simple hood-like shape (like a sun-bonnet) with long slender hooks attached very simply to its lower extremity; the lateral supports are slender and the clasps well developed. TRAOTA TIMOLEON (Stoll). (Plate IV. figs. 2 & 2a.) Papilio timoleon Stoll, Suppl. Cramer, Pap. Ex. vol. v. pl. xxxii. figs. 4, 4d, 9 (1790). Amblypodia timoleon Horsfield and Moore, Cat. Lep. Mus. H.1.C. vol. 1. p. 44. n. 67, pl. xi. figs. 3 larva, 3a pupa (1857); Marshall and de Nicéville, Butt. Ind. vol. i. pl. ii1., larva and pupa (1882). Traota timoleon Moore, Proc. Zool. Soc. Lond. 1882, p. 249; Staudinger, Exot. Schmett. p. 279, pl. xevi., g (1888). VOL. XVII.—PART I. No, 2.—August, 1903. C 10 MR. G. T. BETHUNE-BAKER: A REVISION OF THE Thecla nila Kollar, Hiigel’s Kaschmir, vol. iv. pt. 2, p. 418. n. 3, pl. iv. figs. 5, 6, § (1844). Traota timoleon de Nicéville, Butt. Ind. vol. iti. p. 215. n. 775, pl. xxvii. figs. 192 ¢,193 2 (1890). Hab. China; Bhutan; Kangra District; Masuri; Dalhousie; Chumba, Dagshai, Faizabab ; Julpigori; Sikkim; Cachar; Pachmarli; Orissa; Ganjam; Rutnagherry ; Khandala; Poona; Nilgiris (found only at low elevations). Expanse, ¢ 48, 9 48-56 mm. g. Upperside: both wings black: primaries with a large patch of brilliant deep lustrous azure below the cell, extending from the base across three-quarters of the wing: secondaries with a larger patch of lustrous deep azure between the upper discoidal and the submedian nervure, and extending but little into the cell, and leaving a rather broad black posterior margin; lobe brick-red, the single fairly long grey-tipped tail adjoining it. Underside: primaries blackish, with the whole of the submedian area grey, and the costa and upper half of the posterior margin broadly red; beyond the centre of the costa is an obscure pale greyish patch ; a silvery-white club-shaped stripe extends rather over halfway along the upper edge of the cell, which is closed by a similar white spot; a straight, interrupted, very obscure, whitish-spotted fascia extends from the costa to the lower median nervule, the lower part of which consists of two lunular white spots, the upper part being very obscure; in the red apical area there is a trace of another short whitish-spotted row. Secondaries clouded grey-black and brown, dusted all over more or less thickly with red irrorations ; from the base of the cell up to just beyond the middle of the costal nervure an irregular white patch extends which is narrowed in the middle, and also has a certain amount of red irroration therein; cell partly closed by a short white dash from the middle of the lower median nervule to the abdominal edge; there is a very irregular whitish line, outside which are three longish red spots touching each other, below which is another white stripe; there is a broadish orange patch in the middle of the posterior margin ; lobe velvety red, with grey fringes. 2. Upperside: both wings brown, with two tails; the primaries only with a sub- lustrous violet patch. Underside as in the male, but altogether paler. Horsfield and Moore figure the larva as onisciform and smooth, with a very small head; second segment largish ; third and fourth each larger than its predecessor, then decreasing evenly to the tail. Nearly one inch long. Pinkish, shading off into greenish on the sides; the terminal segments entirely green, with three rows of lateral dots. Pupa dark brown, striped with darker brown. Short and thick; head distinct ; abdomen slightly constricted below the thorax, and roundish at the end. De Nicéville (/. c.) writes that both J. timoleon and mecenas present the same apparent dimorphism in the female, and he continues: “ though these two species are apparently so distinct, I hold to the opinion that breeding will probably prove them to AMBLYPODIA GROUP OF THE LYCANIDA. il be but one variable species.” The same opinion is held by Mr. W. Doherty, who writes of J. mecenas, “an extremely variable species wherever I have found it.” I think it by no means improbable that this ‘“ variability is due to seasonal causes, and “that it will hereafter be found that L. timoleon is the rains form and J. mecenas the dry-season form of one species.” The dimorphic female referred to by de Nicéville is very curious, it being occasionally found that some females are as brilliantly and as extensively metallic blue as the males. The prehensores are among the most simple in shape of any that I know; the tegumen is a quite plain hood, with long slender hooks attached to the lower extremities ; the lateral supports long and slight; the clasps fairly developed, ham-shaped, with square extremities whose upper corner is produced. Penis long, with the lower portion ot the tip extended considerably beyond the upper half. Traota M&cENAS (Fabricius). Hesperia mecenas Fabricius, Ent. Syst. vol. iii. pt. 1, p. 271. n. 45 (1793); Donovan, Ins. China, pl. xxxix. fig. 2, d (1798). Deudorix mecenas Hewitson, Ill. Diurn. Lep., Lye. p. 25. n. 27 (1863); Butl. Cat. Fab. Lep. B.M. p. 180. no. 2 (1869). Iraota mecenas Moore, Lep. Cey. vol. i. p. 102, pl. xl. figs. 2 ¢, 2@ 9,26 larva and pupa (1881) ; Doherty, J. A. S. B. vol. lv. pt. 2, p. 126. n. 123 (1886). Deudorix timoleon Hewitson, Il]. Diurn. Lep., Lye. pl. vin. fig. 21, 2 (1863). Traota mecenas de Nicéville, Butt. Ind. vol. iii. p. 216 (1890). Hab. China; Hongkong; North India; Ceylon; Jhulaghat; Hastern border of Kumaun; Tanna, or Nasik District; Egutpura, or the Thul Ghat (October), Bombay ; Bhutan, Sikkim (April, November, and December); Mundi, Maswin; Dinapore ; Bholahat; Calcutta; Orissa; Bangalore. Expanse, ¢ 56-48, 2 42-54 mm. é. Upperside: both wings very dark purplish brown, with blue patches deeper in tone than in ¢imoleon, and perhaps rather more extensive in the primaries. Underside: both wings almost precisely as in ¢imoleon, but not quite so dark, and with no silvery-white patch below the costa of the secondaries ; the outline of this patch is, however, quite distinct as a rule. @. Upperside: both wings brown, with a violet or purplish patch of considerable extent on the primaries, whilst on the secondaries this colour is reduced almost to the disc. ‘This species, like ¢¢moleon, has another occasional form of the female just like the male, and I conclude it was from one of this form that Mr. Moore described the species with the male and female similar in colour. ‘The common torm is as described here. I have one specimen also in which the violet-blue colour of the primaries is reduced almost to the discal area. The earlier authors mixed up this and the previous species considerably, but I believe c2 12 MR. G. T. BETHUNE-BAKER: A REVISION OF THE that de Nicéville (7. c.) has quite correctly elucidated them. Thisspecies may be known from all others by the absence of the silvery-white patch at right angles to the body, just below the costa in the secondaries. It seems to obtain in precisely the same district as témoleon. It would be an interesting life-history to work out, and so prove the correctness of de Nicéville’s theory that it is the dry-season form of that species. Traota RocHANA (Horsfield). Amblypodia rochana Horsfield, Cat. Lep. Mus. E.1.C. p. 108. n. 40 (1829) ; Horsfield and Moore, Cat. Lep. Mus. E.I.C. vol. i. p. 44. n. 68, pl. ia. fig. 10, ¢ (1857). Traota rochana Moore, Journ. Linn. Soc. Lond., Zool. vol. xxi. p. 43 (1886) ; de Nicéville, Butt. Ind. vol. iii. p. 217. n. 777 (1890); Druce, Proc. Zool. Soc. Lond. 1895, p. 587; de Nicéville and Martin, J. A. S. B. vol. lxiv. pt. 2, p. 463. n. 383 (1895). Thecla timoleon Boisduval (nec Stoll), Sp. Gén. vol. i. pl. xxi. fig. 4, 2 (1836). Deudoria timoleon Druce, P. Z. 8. Lond. 1878, p. 352. Traota boswelliana Distant, Rhop. Malay. p. 258. n. 1, pl. xxii. fig. 23, 9 (1885); Staudinger, Exot. Schmett. p. 279, pl. xevi., 9 (1888). Hab. Mergui; Java, Borneo (Kina Balu and Labuan); Penang ; Singapore ; Philippines (Felder); Sumatra. Expanse, ¢ 38-44, @ 43-50 mm. g. Upperside: both wings brownish black, with the blue patches of a pale metallic azure, not the deep blue of the previous two species. Two tails, tipped with white, one at the lobe and one from the end of the lower median nervule; there is also a tooth-like excrescence at the end of the second median nervule. Underside: primaries dark grey with the internal margin quite pale; a silvery-white stripe along the upper discoidal nervure terminating a third before the end of the cell, which is closed by a like- coloured spot, beyond which is a curved row of three silvery-white spots, the lowest being the most outward and forming at the same time the uppermost of a row of three straight, very oblique (inwardly), white spots; the ground beyond the curved row is red up to the submarginal row of white spots, which row is curved outwards in the centre, while beyond this the margin is slightly rufous, and the costal area is also somewhat rufous. Secondaries with the basal half brightly rufous, the extreme basal costa being shortly white; from the base there is a very broad band of creamy silvery touching the costal nervure at and beyond its middle, which sends a creamy stripe to the upper median nervule, thus closing the cell; a toothed whitish stripe runs from the base of the second median nervule to the upper part of the abdominal margin. This rufous area is edged by an irregular, somewhat interrupted, curved stripe of whitish, on the outer margin of which is a row of four or five very deep red, lunular, small spots extending as an interrupted red stripe to the abdominal edge, beyond which is a broad band of neutral grey tapering and narrowing towards the costa and ending in a small red patch ; outside this is a marginal row of spots, the lobe-spot and three beyond it being AMBLYPODIA GROUP OF THE LYCZNIDZ. 13 velvety-black, while above these the other four are red, the central three being the smallest; all are entirely encircled with white, and the spot next that on the lobe is almost entirely obsolete with superimposed submetallic bluish-grey scales. 2. Upperside: both wings of a somewhat dark golden brown, darker at the extremities, with three distinct white-tipped tails, the upper one being short and broad ; lobe-spot black. I have one specimen before me with a trace of violet scaling on both wings; the brown line in this specimen is quite dark all over. Underside: both wings like the male, but all the silvery-white spots and markings are very much enlarged and intensified throughout, and the submedian area and internal margin in the primaries are whitish, whilst the broad tapering band of the secondaries is ochreous rather than grey. In the specimens with violet scaling on the upperside the costal and subcostal margin of the primaries is orange rather than red, and in the other specimens there is a tendency towards this tone of colour. This species has two tails in the male and three in the female. It may easily be distinguished from the preceding species by the distinct black spots in the anal area, the greater extent of white in the secondaries, and by the pale delicate brown subcostal band preceding the black anal spots. Traova LAZARENA (Felder). (Plate I. fig. 2, 2.) Myrina lazarena, Felder, Wien. ent. Monats. vi. p. 293. fraota lazarena, Semper, Schmett. Philip. Ins. Bd. v. p. 195. Hab. Borneo ; Celebes; Philippine Islands (Coll. Senwper). Expanse, ¢ & 2,51 mm. 3. Upperside : both wings brownish black, with a patch of brilliant metallic emerald- green scales, that on the primaries occupying the subdiscal area, but only extending slightly over the submedian nervure, and but little beyond the cell; on the secondaries these scales are more extensive, extending in a line down the median nervure, increasing slightly towards the end of the cell, where it suddenly expands and occupies a good part of the median area; three tails, tipped with white; lobe well developed, with a black eye-spot. Underside: both wings greyish brown: primaries with a white slightly increasing streak occupying nearly the full length of the cell, beyond which is a white spot; transverse band composed of five spots, the first three inclined very obliquely outwards, the first, nearest the costa, being very small, the fourth and fifth inclined very obliquely inwards; a narrow submarginal line of continuous spots; submedian area pale. Secondaries with a short white edging to the base of the costa, followed by a broad stripe of the ground-colour; a broadish white oblique stripe extends from the base to the apex of the costa, interrupted near the apex; in the large area of the darker ground-colour there is a horizontal irregular shortish white stripe ; cell closed by a narrow white line; this dark area is edged yet lower down by 14 MR. G. T. BETHUNE-BAKER: A REVISION OF THE an irregular indistinct whitish line, this is followed by a very broad paler band of the eround-colour tapering towards the apex, edged outside by a fine, irregular, white line, beyond which is a row of marginal spots, all of which are indistinct except those adjacent to the tails, that by the first short tail being quite black, whilst the lobe-spot is intense velvety black; these are edged above, and the space between them is filled in, with a sparse scattering of submetallic green scales. 2. Upperside: both wings entirely warm darkish brown, with three tails, all white- tipped. The costal area of the secondaries is pale. Beneath, the pattern in both wings is like the male, but all the white markings are greatly accentuated and are brilliantly pearly white, and the darker ground-colour is rich reddish brown, This species is nearest rochana Horsfield, but is very much larger, and the male has three distinct tails, a unique feature in the whole group; the metallic colour is distinctly green in all lights, not blue; and in the secondaries below the white marks are distinct stripes and lines, and not broken up as in rochana. I figure a female from Celebes in my own Collection. Traota NILA Distant. Traota nila Distant, Rhop. Malay. p. 462. n. 2, pl. xliv. fig. 24, 9 (1886); Druce, Proc. Zool. Soe. Lond. 1895, p. 588, pl. xxxiil. fig. 1, ¢; de Nicéville and Martin, J. A. S. B. vol. Isiv. pt. 2, p. 463. n. 384 (1895). Hab. Malacca; Borneo; Sumatra; Perak. Expanse, ¢ 39-40, 9 41-42 mm. g. Upperside: both wings blackish with a steely-blue lustre in certain lights, with bright deep blue stripes in the cell, the median interspaces, and the submedian area ; in the secondaries these stripes are confluent, forming a good-sized patch intersected by the blackish nervures, the two tails short tipped with white; lobe small, but well defined ; antenne unusually long and tipped with whitish. Underside: both wings pale ochreous with white spots. Primaries: cell closed by a white spot above and below which are confluent, white marks extending to the costa and lower median nervule, beyond which there is a patch of reddish, edged exteriorly by three white spots, beyond which are three more smaller white spots in the apical area; below the centre of the red patch are two white spots, followed exteriorly by two larger black spots with indefinite reddish margins; submedian area dark grey, below which the ground-colour becomes whitish. Secondaries with three white spots below each other encircled with black rather near the base, beyond which are four more white spots encircled with black, below the lowest of these and nearer the anal angle is an irregular ‘black mark pupilled with white extending to the internal nervure ; at the apex is a good-sized red patch, at the outer edge of which are four small white spots encircled with black; a marginal row of five white dots in the nervule interspaces and an isolated black spot AMBLYPODIA GROUP OF THE LYCANIDA. 16 in the lower submedian interspace somewhat near the outer margin; tails and lobe velvety black, above which are two ill-defined white marks. 2. Upperside: both wings steely lustrous bluish; costa and outer margins of primaries narrowly brown ; costa of secondaries broadly, outer margin narrowly, brown ; two tails, brown tipped with white. Underside exactly like the male. This is a very beautiful and distinct species. Mr. H. H. Druce has figured, in the ‘ Proceedings Zool. Soc. Lond.’ for the year 1895, a male from Borneo with metallic- green interspaces; this is a very beautiful local race. ‘Two males now before me from Sumatra have no trace of the green in any light, but are deep bright blue. _ This species may be recognized at a glance from any others; it is very much smaller in size. AmBiypopia Horsfield. Amblypodia Horsfield, Cat. Lep. Mus. #.1.C. p. 98 (1829) ; Moore, Lep. Cey. vol. i. p. 113; Distant, Rhop. Malay. p. 275 (1885) ; de Nicéville, Butt. Ind. vol. iii. p, 208 (1890) ; Staudinger, Exot. Schmett. (in parte), p. 281 (1888) ; Schatz and Rober, Exot. Schmett. p. 275 (1892). Wings broad. Primaries with base curved, apex pointed and acute, costa arched more strongly at base; posterior margin almost straight or very slightly convex in the male, in the female slightly excavated just below the apex, then decidedly convex ; inner margin produced, slightly excavated in the centre. Costal nervure of moderate length, ending just beyond the apex of the cell; first and second subcostal nervules with their bases close together ; third subcostal nervule rising nearer to the base of the wing than to the origin of the upper discoidal; in the female this ends at the apex of the wing, but in the male only a fourth subcostal is present, which is very long, terminating at the apex of the wing, its origin being much nearer the origin of the third subcostal than the apex of the wing; the end of the subcostal nervure terminates in the posterior margin just below the apex; the upper discoidal nervure rises from the subcostal long before the apex of the cell, whilst the lower discoidal rises from the upper, well beyond its origin, the base of which is deflexed; lower discocellular nervule upright, slightly concave, twice as long as the middle discocellular; second median nervure rising some way before the lower end of the cell; submedian nervure straight at the base, then slightly deflexed towards the apex. Secondaries very consider- ably more ample in the female than in the male; costa and posterior margin strongly curved, the latter excavated before the submedian nervure, which is produced into a short stout tail; a moderate-sized anal lobe, above which the internal or abdominal margin is somewhat excavated, but is afterwards convex. Costal nervure strongly arched, long, ending at the apex of the wing; first subcostal nervule rising some way before the apex of the cell; upper discocellular nervule shorter than the lower, straight and outwardly oblique ; lower discocellular slightly concave, the secona median nervule rises close to the lower end of the cell, submedian nervure almost straignv: internal neryure 16 MR. G. T. BETHUNE-BAKER: A REVISION OF THE short, strongly recurved. Antenne short, about one-third the length of the costa. Eyes smooth. Palpi porrect, longish; thorax stout; body robust; legs short and thick. The arrangement of the discoidal nervules is quite remarkable, and I believe is peculiar to this genus and to Iraota Moore—at least, I do not know of any other “Old World” Lyceenidee which have it. Horsfield figures the larvee of the usual Lycenid shape with small head and second segment much larger, segments getting wider up to the seventh, then decreasing ; the constrictions between the segments are fairly well marked, the sides of the body being slightly bristled. Pupa attenuated in a very unusual manner, with the head round, thorax humped, abdominal segments slender, and tail sharply pointed. Mr. Doherty in describing the egg of A. narada Horsfield says it is “ large, coarse, and overlaid with white, roughly tubercular, and indented with spaces obscurely hexagonal. It greatly resembles that of most of the ‘ Thecline.’” The tegumen and lateral supports of the prehensores are more fully developed in this genus, whilst the clasps though smaller are very unusual in shape. Penis small. AMBLYPODIA NARADA Horsfield. (Plate IV. figs. 3 & 3a.) Amblypodia narada Horsfield, Cat. Lep. Mus. E.I.C. p. 98. n. 80. Thecla narada, Horsfield, op. cit. pl. i. fig. 8 o, pl. iv. figs. 4 larva, 4@ pupa (1829) ; Horsfield and Moore, Cat. Lep. Mus. E.1.C. vol. i. p. 39. n, 51, pl. i. figs. 4 larva, 4.@ pupa (1857); Distant, Rhop. Malay. p. 276. n. 1, pl. xxi. fig. 28, g (1885); de Nicéville, Butt. Ind. vol. iii. p. 210 (1890). Amblypodia taoona Moore, Proc. Zool. Soc. Lond. 1878, p. 835; de Nicéville, Butt. Ind. vol. iii. p. 211. Amblypodia narada Druce, Proc. Zool. Soc. Lond. 1895, p.587; de Nicéville and Martin, J. A. S. B. vol. lxiv. pt. 2, p. 463 (1895). Hab. Mergui; Penang; Malacca, Java, Andamans, Mindanao, Palawan, Mindoro; Borneo; Nias. Expanse, ¢ 51-56, 2 50-58 mm. g. Upperside: both wings bright purplish blue (I cannot call any male Amblypodia that I have seen true ultramarine-blue, as all had a tinge of purple). Primaries with a moderately broad blackish apex, decreasing in width along the posterior margin; costa rather narrowly black, but variable. Secondaries with costa and apex very broadly (but variably) brown, tapering down narrowly to the upper median nervule, whence to the anal angle the margin is quite narrow; internal margin very broadly brown. Underside: both wings rufous brown ; a dark sinuous line passes from the apex of the primaries obliquely through both wings, terminating above the centre of the abdominal margin; from this line to the base the whole of the ground is dark blackish brown with no definite markings therein. In the primaries midway between this line and the posterior margin there is a row of obscure reddish spots darkly edged, those nearest AMBLYPODIA GROUP OF THE LYCENIDA, 17 the apex being only dark dots; the ground-colour is darker at the apex and anal angle. Secondaries with a similar broad row of confluent spots darkly edged, beyond which edging there is a whitish diffused internal and external scaling ; lobe reddish; the lobe is very well developed and adjoins a short stout tail; marginal area dark. 2. Upperside: both wings brown, primaries with a good-sized pale azure patch, secondaries with the azure patch very much reduced in size. Underside: both wings ochreous grey, with the oblique stripe passing through both wings, the ground-colour being rather darker up to the base ; the other markings as in the male, but approaching evanescence. This species is most variable, individuals from every island apparently presenting minor differences which it would be absurd to name. A form found in Nias has the female of the same blue colour as the male, an almost unicolorous dark underside with the oblique stripe as usual, but the lobe has a large velvety-black spot with two or three black marginal spots beyond, all the spots being broadly encircled with glaucous: this form I also have before me from Borneo, but the underside is paler. A form found in several of the Philippine Islands has the male with a pale violet patch on the upperside, while the female is not quite so blue as in the type-species, but the underside is strongly tinged with olive-green or olive-brown and is also almost unicolorous. Specimens of this form occasionally present a trace of the black spots at the anal angle. I have no doubt at all that A. taooana Moore is merely a bright specimen of this species, the blue colour of which varies a good deal. The width of the black border not only varies a good deal in the female but in the males also. The clasps of the prehensores are very peculiar, being small for the size of the insect, but the extremities are strongly serrate and produced upwards into a long straight extension and attached to the laterals by a fine covering of thin chitine; the penis is short and stout. AMBLYPODIA ANDERSONII Moore. Amblypodia andersonii Moore, Journ. A.S. B. vol. li. pt.2, p. 43 (1884) ; id. Journ. Linn. Soc. Lond. Zool. vol. xxi. p. 44, pl. iv. fig. 4, g (1886); de Nicéville, Butt. Ind. vol. iii. p. 211 (1890). Hab. Sampu; Sullivan Island, Mergui (January). Expanse, ¢, 1°75 inches (de Nicéville). “4 smaller than A. faooana Moore. Upperside: both wings of a similar tint of ultramarine-blue, with much narrower marginal border. Underside: both wings much darker coloured, but similarly marked ” (Moore, /.¢.). The unique type of this species is in the Indian Museum at Calcutta. De Nicéville says that the blue colour is rich and beautiful, and is unlike anything else in the genus, except that he has one specimen of the true A. narada Horsfield with which it entirely agrees. The probability is that this therefore is only an aberration of Horsfield’s insect. I keep them separate, as I have been unable to see the type. VOL. XVII.—-PaRT I. No. 3.—August, 1903. D 2 18 MR. G. T. BETHUNE-BAKER: A REVISION OF THE AMBLYPODIA ANITA Hewitson. Amblypodia anita Hewitson, Cat. Lycenide B.M. p. 14. n. 66, pl. vii. figs. 90, 91, g (1862) ; Butler, Proc. Zool. Soc. Lond. 1883, p. 147. n. 8; Elwes, Trans. Ent. Soc. Lond. 1888, p. 403. Amblypodia narada var. erichsonii Wood-Mason and de Nicéville, J. A.S. B. vol. xlix. pt. 2, p. 234. n. 53 (1880) ; et vol. 1. pt. 2, p. 250. n. 74 (1881) ; de Nicéville, Butt. Ind. vol. i. p. 211 (1890). Amblypodia naradoides Moore, Proc. Zool. Soc. Lond. 1879, p. 141; et Lep. Cey. vol. i. p. 114, pl. xliii. figs. 1, 1 a, @ (1881); de Nicéville, Butt. Ind. vol. i. p. 212 (1890). Amblypodia darana Moore, Proc. Zool. Soc. Lond. 1879, p. 141; et Lep. Cey. vol. i. p. 114, pl. xliii. fig. 2, 9 (1881); de Nicéville, Butt. Ind. vol, 11. p. 212 (1890). Amblypodia arracana Grose-Smith, Ann. & Mag. Nat. Hist. ser. 5, vol. xx. p. 268 (1887) ; de Nicéville, Butt. Ind. vol. ii. p. 213 (1890). Amblypodia anita Druce, Proc. Zool. Soc. Lond. 1895, p. 587. Hab. Siam, Madras, Sikkim, Sumatra, Ceylon, Arracan Hills; Mangioli, Sula Island. Expanse, ¢ & 9, 50-57 mm. g. Upperside: both wings dull violet. Primaries with the posterior margin rather broadly blackish, and the costa less broadly so. Secondaries with costa very broadly and posterior margin narrowly blackish, lobe dull red. ‘Tail stout. Underside: both wings rufous brown, with an oblique transverse black line from the apex of the primaries going right through both wings to the middle of the abdominal margin ; from this line both wings are more or less suffused with dark scales—this, however, is very variable indeed, the wings being occasionally almost uniform in colour. ‘The primaries have a very obscure submarginal spotted row, frequently almost obsolete, and the secondaries, as well as the lobe-spot of the secondaries, are black, and there is also an obscure black marginal spot between the second and lower median nervules. 2. Upperside: both wings blackish brown, with the submedian and discal areas dull slightly lustrous violet ; in the secondaries this colour is confined to the discal and median areas. I have before me another form of this sex in which the patches on the wings are brightish purplish blue. Underside of both just like male, but with the transverse lines more marked and a slight grey scaling at the anal angle. I am quite unable to differentiate Mr. Moore’s species and have most reluctantly been compelled to sink his names, as I really cannot find any true points of distinction. When the males are quite indistinguishable I do not consider it wise to make a species on account of the female being of a different form to that usually found: this surely peints to nothing else than a dimorphic female, the original one probably being brown, whilst the purplish one is a more recent development. The colour of {this species being so constant, it appears to be advisable to retain the name as distinct from narada, it being an easily recognisable form. AMBLYPODIA GROUP OF THE LYCENIDE. 19 AMBLYPODIA ANNETTA Staudinger. Amblypodia annetta Staudinger, Exot. Schmett. p. 281, Taf. 96, 3 (1889). Amblypodia annetta var. anna Staudinger, Exot. Schmett. p. 282 (1889). Hab. Batchian; Waigeoe ; Saparua; Amboina. Expanse, ¢ 45-49, 9 53-50 mm. ¢. Upperside: both wings very dark brown, primaries with a discal and submedian patch of submetallic purplish blue; secondaries with the blue confined almost to the discal and median areas. ‘The colour of this patch is given to vary somewhat, as I have a specimen in which it is metallic deep azure and another in which it is metallic blue of lavender tinge. Underside: both wings uniformly rather dark brown, occasionally with a purple tinge; primaries without any mark at all: secondaries with a submarginal lunulated whitish interrupted stripe, preceded internally by a most obscure dark line ; there is also a marginal pale spotted row most obscure at the apex, but quite distinct at and near the anal angle; lobe-spot blackish. 1 have two specimens from Waigeoe, in both of which there is a distinct trace of the transverse line extending through both wings; this is entirely wanting in the type specimens from Batchian. @. Upperside: both wings dark brown, with a much more limited patch (even than the male) of very pale greyish lustrous blue. Underside: both wings pale purplish greyish brown, with a dark oblique stripe through both wings ; the posterior margin of primaries broadly brown (not grey or purplish) ; secondaries with a submarginal tapering spotted row palely edged on both sides, lobe-spot blackish. ‘This description is taken from Waigeoe specimens, as I do not possess a male from Batchian. The var. anna Staudinger is really not worthy of a name, the only difference being that the blue has a slightly lavender tinge, and this, as we have seen, is variable, as also is the pattern of the underside, or more accurately, I might say, absence of pattern; the female on the underside is precisely the same as the type female, but the blue of the upperside has a slight greenish tinge, otherwise there is no difference. The species may be recognised from the others by the very small expanse of blue in the males and by the different shape of the wings. AMBLYPODIA FAISINA Ribbe. Amblypodia faisina Ribbe, Iris, vol. xii. p. 252, Taf. iv. fig. 16, 2 (1899). Hab. Faisi Island (Bismarck Archipelago). This species is very near A. annetta Stgr., having been described from a unique female which is the exact colour above of a female of Staudinger’s species ; but Ribbe’s insect is smaller and can also be recognised by the broad pale band on the underside of both wings. 20 MR. G. T. BETHUNE-BAKER: A REVISION OF THE AMBLYPODIA HAINANA Crowley. Amblypodia hainana, Crowley, Proc. Zool. Soc. Lond. 1900, p. 508. Hab. Hainan. Expanse 56 mm. I have not been able to see this species, but it should be recognised without difficulty by its large size, for it is about 10 millimetres larger than any specimen of the genus that I have seen. Manatuata Moore. Mahathala Moore, Proc. Zool. Soc. Lond. 1878, p. 702; de Nicéville, Butt. India, vol. ii. p. 282 (1890). Apporasa Moore, J. A. S. B. vol. liii. pt. 2, p. 88 (1884) ; de Nicéville, Butt. Ind. vol. iii. p. 224 (1890). Wings broad, short. In the primaries the costal nervure is less than half the length of the costa, terminating well in front of the apex of the discoidal cell; the first and second subcostal and the upper discoidal nervules all have their bases about equidistant ; the third subcostal nervule is short, rising nearer the apex of the wing than of the cell; the middle discocellular nervule is very short, straight, and outwardly oblique ; the lower discocellular is nearly four times as long as the middle, concave, and nearly straight. Second median nervule given off some little distance before the lower apex of the cell; first median bent downwards near the base. In the secondaries the costal nervure is strongly sinuate and bent upwards at the end; the first subcostal nervule arises some way before the end of the cell, and is strongly and somewhat evenly arched. The upper discocellular nervule is straight and very oblique inwardly, whilst the lower is likewise equally oblique inwardly, thus forming a < shape. The second median nervule is given off almost at the juncture of the upper median, but just below it—. e. just in front of the lower end of the cell. The submedian nervure is nearly straight ; the internal nervure short and sinuous. ‘The abdominal margin is decidedly excised, and thus forms a distinct anal lobe. The antenne are very short, about a third the length of the costa. De Nicéville (/. c.) says :—“ Although Apporasa atkinsoni possesses a tail similar to Wahathala ameria Hewitson, and the costa of the hind wing is also somewhat similar, yet it is, I believe, nearer to Thaduka mulicaudata Moore than to that species. It has the costa of the fore wing arched as in 7. multicaudata (con- siderably more so than in J. ameria).”” Doherty (J. A. S. B. vol. lviii. pt. 2, p. 423) says that “the genus Apporasa differs from Mahathala in the less acuminate apex of the hind wing, in the egg (which is covered with triangles and tubercles instead of quadrangles and spines, a difference which in this group is of small importance), and in the more undulate margin of the hind wing”; and he continues :—* But athkinsoni when sitting on a tree-trunk has a marvellous resemblance to a patch of lichen, and AMBLYPODIA GROUP OF THE LYCENID, 21 the irregular outline adds to the effect. Mimicry of this sort is a sign of great flexi- bility of structure, and such genera must be judged by severer canons than others, so perhaps Apporasa had better be sunk in Mahathala.” I entirely agree with Mr. Doherty in this; the neuration is the same, the arch of the costa of the primaries is rather stronger than in J. ameria, but decidedly less strong than in 7. multicaudata; the costa of the secondaries is precisely the same as in Mahathala, and therefore I do not see what use there is in separating the two genera— in fact, I do not see what characters there are to enable us to separate them. The prehensores are highly developed; the tegumen is very ample, with strong straight hooks, with a considerable muscular attachment; the clasps are also unusually ample and peculiar. The penis is short and stout. I have described and figured them under the species IZ. ameria Hew. MAHATHALA ATKINSONI (Hewitson). Amblypodia atkinsoni Hewitson, I]. Diurn. Lep., Lyc. p. 14g. n. 99, pl. iii 4. figs. 48, 49 (1869). Apporasa atkinsoni Moore, J. A.S. B. vol. liii. pt. 2, p. 88 (1884); de Nicéville, Butt. Ind. vol. iu. p. 224. n. 784 (1890). Mahathala (Apporasa) atkinsoni Doherty, J. A. S. B. vol. lviii. pt. 2, p. 423 (1889). Hab. Moulmein; Myitta (Tenasserim Valley), Tilin Maw, and Katha District, N. Burma (Coll. de Nicéville). Expanse 42-45 mm. g. Upperside: primaries very dark blackish brown, tinged with purple; costa lightly spotted, with the discoidal and submedian areas pale lustrous violet-blue (I cannot possibly call it purple); secondaries dark brown, with a similar violet-blue patch as the primaries. Underside: primaries dark brown, posterior margin densely irrorated with grey ; costa irrorated and speckled with a similar colour; a whitish dash across the centre of the cell, which is closed by two broad, short, whitish dashes out- lining a sort of spot, beyond which is a dark spot laterally edged with whitish ; below this several indefinite whitish spots, and the submedian area is also whitish. Secondaries paler brown, mottled and densely irrorated all over with reddish grey or grey tinged with lilac; the cell is closed by a dark mark, and there are one or two spots at the base; a very few metallic bronzy-green scales are scattered here and there over the wing indefinitely. All the markings are most indefinite on both wings, and the few bronzy scales would render the mimicry of lichen (to which Mr. Doherty likens the underside) about perfect. ¢. Exactly like the male, except that the blue of the under surface is paler, bluer. brighter, and more extensive. This is apparently a very rare insect ; it has been described for more than thirty years, and there are but few collections in which specimens appear. It is a beautiful and yery interesting species, and is readily separable from any others of the genus by the 22 MR. G. T. BETHUNE-BAKER: A REVISION OF THE paler blue and broad margins of the male, as also by the curious colour of the under surface. MAHATHALA AMERIA (Hewitson). (Plate IV. figs. 4, 4a.) Amblypodia ameria Hewitson, Cat. Lycenide B.M. p. 14. n. 64, pl. vill. figs. 85, 86, 9 (1862). Mahathala ameria Moore, Proc. Zool. Soc. Lond. 1878, p. 703; Rothney, Ent. Month. Mag. vol. xix. p. 35 (1882). Narathura ameria Distant, Rhop. Malay. p. 268. n. 13, pl. xxi. fig. 30, 2 (1885). Mahathala ameria de Nicéville, Butt. Ind. vol. iii. p. 288. n. 848, pl. xxvii. fig. 200, ¢ (1890). Arhopala (Mahathala) ameria de Nicéville and Martin, J. A. 8. B. vol. Ixiv. pt. 2, p. 470 (1895). Hab. Northern India; Siam; Hainan; China; Barrackpur; Mergui; Perak ; Beerbhum District ; Calcutta; Goalpora; Sibsagar; Chittagong Hill Tracts ; Sumatra (Paul Staudinger) ; Java. Expanse, ¢ & 9, 37-40 mm. 3. Upperside: both wings dull purple; primaries with costa rather narrowly, and posterior margin less narrowly, black ; secondaries with apex and costa broadly, and posterior margin less broadly, blackish. Tail rather long, broad, and _ spatulate, scalloped on each side so as to form a short tooth-like projection. Underside : primaries coffee-brown, with pale submedian area; spots scarcely darker than ground, laterally edged with greyish white, one in the centre of the cell and one more distinct closing it; below these are two divided by the lower median nervule, above them are often (not always) one or two obscure spots; transverse band broken outwardly after the first and after the second spot, and inwardly after the fourth spot, terminating on the lower median nervule beyond this; costa and apex sprinkled with grey irrorations ; marginal row lunular. Secondaries dark brown, with all the ground between the fascize closely irrorated with greyish ; the basal area is very irregularly spotted and blotched with reddish brown; a reddish-brown irregular fascia from apex to near the base of the abdominal margin, beyond which is another broad irregular fascia from the upper discoidal nervule to the middle of the abdominal margin ; marginal row obscure but dark ; lobe well formed. 9. Precisely like the male in every particular, except that the purple area of the upperside is much more restricted. The female of this species is subject to considerable variation both as to the colour of the blue and as to its extent. I have specimens before me with the purple occupying an area only just beyond the outer jimit of the cell, and others which are bright rich blue, with only a broad dark posterior margin and broader apex. L believe that the male also varies in colour, for Mr. de Nicéville says the male is not purple but rich deep blue. The clasping-organs here figured are, however, from a specimen which is quite purple and very dull in tone, and I may say that I have not seen a really blue male. ‘These organs are strongly specialised in this genus. The AMBLYPODIA GROUP OF THE LYCENIDZ. 23 tegumen consists of a very ample and substantial hood, to the lower part of which are affixed, by a strong muscular attachment, the hooks, which are very robust and straightish, with a slight deflection at their tips. The clasps are unusually fully developed, and are excised in the middle so as to form the extremities into two strong points. ‘The penis is short, with the hinder two-thirds very large. MAHATHALA HAINANI, n. sp. (Plate I. fig. 1.) Hab. Isle of Hainan. Expanse 48 mm. 2. Upperside: both wings very dark purplish brown; primaries with a good-sized patch of violet-blue over the discal submedian and two-fifths of the median areas ; secondaries with the violet-blue onlv a little beyond the discal cell. Underside: primaries coffee-brown (more or less dark), with three very short and fine white dashes in the cell; cell closed by a spot laterally edged very finely with whitish, below which is a longish spot edged above with white; basal area dark up to just beyond the cell, whence the submedian area is quite pale; transverse band of moderately equal width, strongly curved from the costa to the upper discoidal nervule, from where it descends almost straight to the lower median nervule, below which is an irregular spot shifted inwards; margin rather broadly dark purplish grey. Secondaries: basal area to well beyond the celi very dark reddish brown, whence to the margin it is of an indescribable reddish grey of a lustrous texture, having almost exactly the same appearance as in an insect that is greasy, the separation between the two areas being quite sharply defined; in the basal area there are traces of a spot across the cell, of one below it, and of a larger one closing the cell, but they are most obscure ; in the lustrous outer area the dark thrice-broken transverse panel is quite distinct from the upper discal to the internal nervure, and is of moderately equal width ; marginal area broadly dark. From J. ameria Hewitson this species may be distinguished by its much larger size and by the blue of the secondaries being so much reduced in its extent, whilst the general appearance beneath is quite different, owing to the uniformity of general tone, though the pattern is very similar; it is, however, much darker, with the whole of the base dark chestnut, and there is a distinct satiny gloss over the whole of the secondaries. I haye two females from the Isle of Hainan. MAHATHALA GONE Druce. Mahathala gone Druce, Proc. Zool. Soc. Lond. 1895, p.593. Hab. Mongolia. Expanse 47 mm. 24 MR. G. T. BETHUNE-BAKER: A REVISION OF THE This species is described from a unique type, and is certainly somewhat different from ameria Hew. It would be very satisfactory if we were able to get other specimens from the same district. THapuka Moore. Thaduka Moore, Proc. Zool. Soc. Lond. 1878, p. 836 ; de Nicéville, Butt. Ind. vol. iti. p. 225 (1890). The neuration is almost the same as in Mahathala, but in the primaries the third subcostal nervule is longer ; whilst in the secondaries the costal nervure is bent down- wards, not upwards, the subcostal nervule is not nearly so strongly arched, and the internal nervure is decidedly longer than in that genus. ‘The two discocellular nervules of the secondaries are each inclined outwards, being thus of exactly the reverse position to that in Mahathala, and the lower discocellular nervule is very decidedly longer than the upper. The costa of the primaries is more strongly arched than in Mahathala. Palpi slender. Antenne short. Body stout. This genus may be known from all others by its species having three tails, two of which are somewhat spatulate. Some species of Jrwota have three tails, but they are long and by no means broad. ‘The genitalia approach closely to the pattern of the genus Arhopala, with the fully-developed tegumen and the small clasps. THADUKA MULTICAUDATA Moore. (Plate IV. figs. 5 & 5a.) Thaduka multicaudata Moore, Proc. Zool. Soc. Lond. 1878, p. 836, pl. li. fig. 7, 2; de Nicéville, Butt. Ind. vol. iii. p. 225. n. 785, pl. xxvii. fig. 196, 9 (1890). Hab. Upper Tenasserim ; Nilgiris (January, August, and September). Expanse, ¢ 38-44, 9 44-48 mm. g. Upperside: both wings purplish black, with the basal, discoidal, and submedian areas very pale silvery blue; abdomen and fold greyisn. Underside: both wings dark vinous brown. Primaries with the posterior half and the submedian space hoary; the central cell-spot is indistinctly outlined by pale bluish, and the spot closing the cell is grey with superimposed bluish scales; the dark catenulated transverse band is distinct and composed of six spots, the upper three oblique outwardly, the second and third being dislocated, the fourth, fifth, and sixth dropping almost perpendicularly from the third, while all are palely edged; submarginal row of <-shaped shading distinct. Secondaries with the three dark, irregular, ill-defined, maculate, transverse bands palely edged generally; these markings are sometimes very indistinct indeed; anal area with a slight suffusion of bronzy-green or bluish scales, which are sometimes scattered sparingly over other parts of the wing. @. Upperside: like the male, but with the blue not quite so pale (not silvery) and bluer. Underside like the male, but not so dark and browner, the markings decidedly more distinct and the metallic-bronzy scaling more abundant. AMBLYPODIA GROUP OF THE LYCENID. 25 I am indebted to the kindness of Sir G. F. Hampson for being able to determine the sexes of this species; be kindly obtained for me a couple of specimens, which proved to be a male and female, whose clasping-organs I mounted. The insect is recognisable at a glance, but I have no doubt that the males have often hitherto been mistaken for females, as the extent of blue is quite the same, and there is very little difference in the colour. In de Nicéville’s collection there is now a good series of this insect, the specimens of which are labelled with dates; some from the Nilgiris are dated January 1; from Atarem Valley (Tenasserim) February ; the majority from the Nilgiris, however, were taken in August and September. The genitalia are rather small. The tegumen is composed of a full hood, the hooks having a strong muscular attachment, straightish as in Mahathala ameria, but with a decided deflexion. The clasps are small, with the extremities projected well outwards, which can be seen better in the figure (Pl. IV. fig. 5) than it can be described. The penis is unusually short and moderately robust. ARHOPALA Boisduval. Arhopala Boisduval, Voy. Astr., Lép. p. 75 (1832). Narathura Moore, Proc. Zool. Soc. Lond. 1878, p. 835; Distant, Rhop. Malay. p. 259 (1885). Nilasera Moore, Lep. Cey. vol. 1. p. 114 (1881). Panchala Moore, Proc. Zool. Soc. Lond. 1882, p. 251; Distant, op. cit. p. 272. Satadra Moore, J. A. 8. B. vol. liu. pt. 1. p. 38 (1884). Acesina Moore, tom. cit. p. 41. Darasana Moore, tom. cit. p. 42. Flos and Iois Doherty, J. A. 8. B. vol. lviii. pt. i. pp. 411, 412 (1889). Arhopala and Acesina de Nicéville, Butt. Ind. vol. i. pp. 226, 280 (1890). Amblypodia (auctorum). Wings broad and ample. Primaries: costa variable, nearly straight, slightly arched or decidedly arched, sometimes somewhat depressed towards the apex; apex acute or subacute ; outer margin occasionally straight or sometimes evenly convex, and some- times concave below the apex, then convex; inner margin generally, not always, slightly hollowed in the centre; costal nervure short but variable, never reaching as far as the end of the cell; first and second subcostal and upper discoidal nervules with their bases nearly equidistant. Third subcostal nervule rising nearer to the apex of the wing than to the apex of the cell, variable in length, but generally short ; middle disco- cellular nervule rising some little way beyond the source of the upper discoidal, very short, straight, outwardly oblique ; lower discocellular nervule variable, sometimes twice as long or up to four times as long as the middle discocellular, straight, generally but not always slightly inwardly oblique. Second median nervule rising some way from the lower end of the cell; lowest median nervule bent downwards and curved. Median VOL. XVII.—ParT I. No, 4.—August, 1903. E 26 MR. G. T. BETHUNE-BAKER: A REVISION OF THE nervure angled upwards to the origin of the latter nervule. Submedian nervure some times straight, often waved. Hind wing generally ample, sometimes with no tail, generally with a tail at the end of the first median nervule, which is sometimes little more than a tooth, but varies considerably, and is not unfrequently long; sometimes there is a second very short tooth-like projection at the end of the second median nervule, and occasionally yet a third very short one at the end of the submedian nervure. Anal angle sometimes round or sharply angled or sometimes distinctly lobed. Inner margin variable, straight, emarginate, or rounded. Costa evenly arched, sometimes flattened, sometimes waved. Costal nervure generally well arched, but sometimes con- siderably straighter. First subcostal nervule well arched, rising generally well before the apex of the cell; upper discocellular nervule shorter than the lower, rather concave, and somewhat oblique outwardly; lower discocellular erect, or somewhat oblique inwardly, or occasionally straight, rarely concave. Second median nervule rising just before the lower end of the cell; submedian nervure waved; internal nervure recurved. Antenne short, generally less than half the length of the costa, with a shght graduated club. Palpi rather long, porrect. Eyes smooth. Body generally decidedly robust. The larva is said to be onisciform, flattened in the anal region, and bristly laterally. Pupa of the usual Lycznid shape. Boisduval created the genus Arhopala and placed in it two new species, one from New Guinea and the other from the Aru Islands, which he named _ respectively A. phryxus and A, meander. Unfortunately he did not diagnose the genus, and until within the last decade the brothers Felder are the only authors who have used it. In 1888 Dr. Staudinger drew attention to it in his ‘ Exotische Schmetterlinge,’ and since then it has been generally accepted in a more or less restricted manner; de Nicéville, however, was the first who used it in its wide and, as I believe, correct sense. Distant, in his ‘ Rhopalocera Malayana,’ does not use it at all. Mr. Moore, however, has split the genus up considerably, creating no less than seven genera out of it, of which I can only retain one, viz. Mahathala. Acesina, which de Nicéville retains with doubt, has somewhat distinct markings below, but they coincide precisely with the general Arhopala pattern, and there is no structural difference whatever. As regards the structural characters relied on by Mr. Moore in the genera that he suggested, there is not one that is constant ; the whole are extremely variable, even in the species which are most nearly allied, and no good can be gained by subdividing a large and well-marked genus like this, merely for the sake of making it less unwieldy, when no definite and good characters can be given. 1 feel, therefore, with all deference to so old and experienced an observer, that I cannot adopt his ideas on this point. I have examined a considerable number of the prehensores of the males, but do not find that they aid much in this respect, and it seems very doubtful if the eggs will be sufficiently well known in our generation to be of any use in the matter. The headquarters of the genus are India AMBLYPODIA GROUP OF THE LYCZNID#. 27 and the Malay region; species occur, however, in China and Japan, also in Australia and some of the small islands of the Pacific, such as New Britain and the Solomon Group. In colour the species are nearly all blue or purple, though a few are of a brilliant lustrous metallic green, whilst one or two are almost brown. In the case of the green species, however, the females are all blue or purple, and were it not for the pattern below coinciding precisely (as it does in every case) they would be, and occasion- ally have been, looked upon as distinct species. Distant remarks upon the shade of the blue varying considerably in the same species, whilst de Nicéville disagrees with the assertion. I think that now the latter author would probably have qualified this had he been alive. I have found decided differences in the colour of the same species, but how that difference has arisen I am not prepared to say. I think it is quite possible that it is a chemical change, for the specimens look quite fresh, so that it is not unlikely that it may have been thus produced. The colours of the whole of the group are aniline—that is to say, they can only be compared to aniline colours,—and I believe the majority of them could not be made up from the colours of the spectrum; they may therefore be more lable to change than otherwise. I have not found the difficulty in recognising Hewitson’s species that some have, and I believe I have been successful in nearly every case; all that I found necessary was time and care. De Nicéville and Martin say (J. A.S. B. vol. Ixiv. pt. ii. p. 470):—*« This genus is one of the most largely represented in Sumatra, which may perhaps be its headquarters, though the Malay Peninsula may possibly possess quite as many species. All are found in forests, but, nevertheless, their more or less metallic-blue, purple, and green colours are not at all conspicuous, and they do little to enliven the sombre depths of the forest, as the restless species of Lampides do. Arhopalas never come to small streams or damp spots on roads to suck up the moisture, or to flowers; they hardly ever fly unless disturbed, and as they always settle with folded wings, of which the undersides present only dull brown, grey, or dull purple colour, little is seen of them. They rest on leaves of shrubs of moderate height, and never fly for any length of time to a distance, feeling themselves much more secure when at rest. There is therefore only one way to see and capture them, and that is to walk through the underwood and disturb them by beating the bushes and low trees, and thus to cause them to fly.” And, again, de Nicéville says (Butt. Ind. /. ¢.): “Many species seem to be common where they occur, others, again, appear to be very rare. Owing to their skulking habits, there is a good deal of difficulty in collecting them.” The genitalia follow one pattern very closely, having the tegumen composed of an ample hood with strong muscular hooks straightish, 7. e. not elbowed, the clasps decidedly small, and the penis of a medium size, Lo (os) MR. G. T. BETHUNE-BAKER: A REVISION OF THE ARHOPALA HERCULES (Hewitson). (Plate IV. figs. 6 & 6 a.) Amblypodia hercules Hewitson, Cat. Lycenide B.M. p. 3. n. 14, pl. vin. figs. 92 & 93, 3 (1862). Arhopala tyrannus Felder, Reise Novara, Lep. vol. i. p. 225. n. 247, tab. xxix. figs. 1 & 2, ¢ (1865). Arhopala hercules Hew., Staudinger, Ex. Schmett. p. 280, Taf. xevi., g (1888). Arhopala tyrannus Felder, Staudinger, Ex. Schmett. p. 280. Arhopala hercules var. herculina Staudinger, Ex. Schmett. p. 280. Arhopala leo Druce, Ann. & Mag. Nat. Hist. ser. 6, vol. xiii. p. 254 (1894). Hab. Makassar; Java; (?) Halmaheira; Waigeoe; N. Guinea, Humboldt Bay ; Gilolo and Ati Ati (N. Guinea). Expanse, ¢ & 2, 62-75 mm. This species is the largest and one of the handsomest of the genus, its wing-expanse frequently reaching 75 mm. It has been recorded from many of the Malay Islands, and extends as far as New Guinea, this being apparently its eastern limit. In the figure of the genitalia, I would draw attention to the straight sword-like hook and the peculiar upward extension of the clasp. Var. LEO Druce. Differs from hercules on the under surface in that the colour is hoary and tinged with very pale green, the transverse fascia on the primaries is more broken, and the same fascia on the secondaries is quite broken up beyond the fracture at the lower discal vein, and does not continuously extend to the internal vein. This was described by Druce from Humboldt Bay, New Guinea, whence I have several specimens; it forms an excellent transition to Staudinger’s form ‘ herculina” from Waigeoe. Var. HERCULINA Stgr. (Plate I. fig. 9.) ¢. Upper surface blue, brighter with less purple in it than hercules. Under surface with a very slight trace of green on the basal area of both wings, which is often entirely absent, the whole surface being pinkish grey; the transverse band in the primaries is more or less broken into a continuous chain of spots, not an unbroken band as in hercules, aud this also applies to the same band in the secondaries, which furthermore is sometimes fractured between the subcostal and the upper discocellular veins, and at the upper and the lower median veins. At the anal angle there is no metallic scaling as is usual in hercules. ?. Brown all over, a shade lighter beyond the discal area of both wings. I have another form brown, with a whitish patch beyond the discoidal cell in the primaries, over which a number of fine pale lavender-coloured scales are very distinctly spread. A third form before me is Staudinger’s blue female, which, as he says in his original description, is almost as blue as the male, the apical and costal areas being broadly brown, the expanse of blue equalling that of leo. AMBLYPODIA GROUP OF THE LYCEHNIDZ. 29 This is a very interesting form indeed, for it is an insect that I should say was in a transitional state, either adapting itself to new conditions, or, finding itself hard pushed for existence, seizing on certain advantages of colour; for far from being of the very constant form of the parent species, individuals of this variety differ inter se on the underside in the pattern very considerably. They are in size rather smaller than hercules. Staudinger described his variety from Waigeoe, but we have since received examples from Halmaheira, from which island I have one without any transverse band on the primaries. Mr. Druce regards the entirely brown females as belonging, possibly, to a distinct species; I have some now before me, but they are without doubt of the same species. It is no more than a case of reversion. Var. TYRANNUS Felder. ¢. Just like hercules Hew., except that the underside is brown with no trace whatever of green, and the pattern is merely of a darker shade. ?. Upper surface entirely dark brown. Under surface like the male. I have one specimen from Ati Ati, on the west coast of New Guinea, which forms a transition from herculina to tyrannus. This species is apparently confined to the Malay Archipelago, where it is evidently dominant, setting up local forms in various islands. I have examined a very large number of specimens and find small divergencies arising in many cases, and have no doubt that the varieties described are local forms of Hewitson’s species. ARHOPALA ATE (Hewitson). Amblypodia ate Hewitson, Ill. Diurn. Lep., Lye. p. 8. n. 32, pl. 1. fig. 4 (1869). Hab. Amboina. Expanse 47-49 mm. ¢. Upperside: both wings purplish blue, very slightly lustrous in certain lights, shaded with a brownish lustre. Primaries with costa and posterior margin narrowly brown, but slightly increasing at the anal angle. Lobe fairly developed; tail brown, white-tipped, and rather long, a tooth-like projection at the end of the second median nervure. Underside brown, slightly lustrous, with darker spots and fascia palely margined. Primaries with a small dot in the cell beyond the middle; cell closed by a rather small elongated spot, below the cell is an obscure darker spot; transverse band almost continuous, slightly-oblique outwardly, rather narrow and tapering rapidly to the lower median nervule where it ends; submarginal row obscure, submedian area paler. Secondaries rather darker than primaries, with three minute basal spots, the upper one the darkest, followed by three more very small spots below each other; cell closed by a very narrow, elongated, curved spot; transverse band continuous, almost straight, moderately broad to the spot between the lower and submedian nervures, Oo 50 MR. G. T, BETHUNE-BAKER: A REVISION OF THE which is shifted well inwards, and is almost confluent with the long spot extending up to the internal nervure; submarginal band broad and distinct to the second median nervule. Lobe-spot black, and a black spot between each of the two following nervure- spaces, each broadly edged above with brilliant, very pale, greenish-blue, metallic scales, over which is greyish irroration. Hewitson says: “ Probably only variety of A. adatha” (I. ¢.). ‘The species has no close affinity with adatha Hew. I have examined his type and have a specimen also from the same locality (ex Mus. Dr. Staudinger) ; it is really more like a miniature hercules, and will be best placed next to that insect. It has apparently neyer been noticed by collectors since it was described, and the only specimens known to me are in Hewitson’s collection and one in Staudinger’s collection, so that it seems to be confined to the little island of Amboina. ARHOPALA PHILANDER Felder. (Plate IV. figs. 7 & 7a.) Arhopala philander Felder, Reise Novara, Lep. vol. 1. p. 226, tab. xxix. fig. 9, 3. Arhopala periander Grose-Smith, Novit. Zool. vol. i. p. 582 (1894) ; id. Rhop. Exot., Lyc. vol. ui. (1898). Hab. Halmaheira, Batchian. Expanse, ¢ 47-00, 2 52 mm. 3. Upperside: primaries and secondaries deep blue, slightly lustrous, with very narrow black costal and posterior margins to the primaries; in the secondaries the costal border is broad, the posterior margin narrow, the abdominal fold entirely dark brown up to the submedian nerve; the lobe is very slightly developed and is black. Underside: both wings uniform brown, internal margin of primaries paler. Disco- cellular spots three, dark brown, encircled with grey, increasing in size, the last one closing the cell, below which and the central spot is another spot palely encircled. The transverse band from the costa to the lower median nervule, increasing in width to just below the second median nervule, is bordered on each side with grey ; touching this band below the third median nervule is another spot placed nearer the base ; submarginal band distinct but ill-defined, with a narrow posterior and a broad interior lighter edge. Secondaries with four small basal darker spots palely encircled below one another, the second placed a little inwards, and the fourth on the outside of the internal nervure, outside these are three larger ones corresponding in position to the three upper basal spots. ‘The cell is closed by a long quadrate spot, below which is another small spot fillimg up the interior end space between the second and lower median nervule. Band continuous, but somewhat fractured at the upper discoidal nervure and also at the lower median, whence it is continued upwards to the internal nervure following the course of the margin. There is an indefinite trace of a submarginal darker shading, between which and the median band is a pale grey AMBLYPODIA GROUP OF THE LYCENIDA. 31 shading ; above the anal angle the lobe-spot is black, edged above with a green iridescence, which is extended up to the second median nervule, and underneath which black spots show through the green scaling. ‘The margin has a very fine grey border at the anal area. Tail brown, tipped with white. 2. Upperside: blue rather brighter than in male, extended almost all over both wings; primaries have the costa blackish beyond the costal nervure; the apical area is broadly black, narrowing to quite a moderate border both towards the costal vein and the anal angle; in the secondaries the blackish costal margin is broad, the posterior margin is rather narrow, increasing a little at the anal angle; the whole of the abdominal fold up to the submedian nervure is brown. This is another species that appears to have been unobserved and unrecorded since its description by the Bros. Felder. When examining the Felder Collection in the Tring Museum I found the unique type, and recognised it as close to some specimens in Staudinger’s and my own collections, and I find they are identical. I have before me one female (coll. Staudinger) from Halmaheira, which I make the type female, as it corresponds with Felder’s type; I have also five males and one female from Batchian. From meander Boisd. it may be recognized by the uniform brown colour of the under surface, and by the transverse band of the primaries being narrower at the costa than at its termination. I cannot really separate periander Grose-Smith from this species. The genitalia have a most ample tegumen; as usual in this genus, the clasps are distinct, somewhat fan-shaped, but more extended on one side. ‘The penis has a somewhat boat-shaped orifice, as will be seen in figure 7 a. ARHOPALA HYLANDER Grose-Smith. Arhopala hylander Grose-Smith, Novit. Zool. vol. i. p. 582 (1894); id. Rhop. Exot., Lye. no. vil. pl. xv. figs. 6, 7, 8 (1898). Hab. Biak. Expanse 65-70 mm. This species is a close ally of philander Feld., scarcely differing from it, the one distinction being that in Felder’s species the transverse band near the posterior margin of the primaries is straight, whereas in the present species it is broken at the upper median nervule. ARHOPALA EuRISUS H. H. Druce. Arhopala eurisus Druce, Proc. Zool. Soc. Lond. 1891, p. 370, pl. xxxi. figs. 11 9,12 ¢. Hab. Aola; Guadalcanar Island; Florida Island; (Solomon Islands) New Britain. Expanse, ¢ 51, 2 48-51 mm. (Se) bo MR. G. T. BETHUNE-BAKER: A REVISION OF THE I have a female before me from New Britain, in which the blue is less silvery, but deeper and richer in tone, and the underside is also rather darker. This species is near philander Feld., but is smaller and is paler in colour above; whilst beneath it is more uniform in tone, and the transverse bands are more irregular. The female is also much paler and silvery in its blue. Var. TINDALI Ribbe, Iris, Bd. xii. p. 251 (1899). ‘This variety appears to be merely a darker form of Druce’s species. ARHOPALA MEANDER Boisduval. (Plate IV. figs. 8 & 8a.) Arhopala meander Boisduval, Voy. ‘ Astrolabe,’ Ent. p. 76 (1832). Amblypodia meander Hewitson, Cat. Lycenide B.M. p. 4, pl. il. figs. 4, 5, & 6 (1862) ; id. Tl. Diurn. Lep., Lye. p. 4. n. 6 (1869). Arhopala meander Boisduval, Miskin, Annals Queensland Mus. no. 1, p. 70 (1891). Arhopala meander Boisd., Bethune-Baker & Druce, Proc. Zool. Soc. Lond. 1896, p. 658. Arhopala adkerbal Grose-Smith, Rhop. Exot. vol. i. p. 10, pl. xxv. fig. 3 (1902). Arhopala appianus Grose-Smith, Rhop. Exot. vol. i. p. 11, pl. xxv. fig. 4 (1902). Hab. Aru; New Guinea; Waigeoe; Halmaheira; New Britain; New Zealand (Mioko); Queensland and South-east Borneo; Woodlark and Fergusson Islands ; Duke of York Island. Expanse, ¢ & 2, 47-59 mm. g. Upperside: both wings violaceous blue; primaries with narrow black costal margin and rather broader posterior margin ; secondaries with broad costal and narrow posterior margin; abdominal feld brown. ‘Tail black, tipped with white. Underside: both wings pinkish lilac-brown, all the spots cimmamon-brown, margined with pinkish erey. Primaries: internal area up to about the second median nervule paler and not tinged with pink; three discal spots increasing in size, and the usual two below the cell; transverse band continuous, inclined outwards as it nears the anal angle; this band generally reaches the submedian nervure, but the last two spots are shifted much inwards; an indistinct submarginal shading. Secondaries with four small basal spots below one another, the bottom one shifted inwards beyond the internal nervure, beyond which are three larger ones corresponding in position to the upper three; cell closed by a long quadrate spot, below which and touching is another small spot closing the end of the space between the second and lower median nervures; transverse chain- band not absolutely broken, but consisting of two spots above the upper discal nervure, and two spots shifted outwards below it, one between the upper and second median nervure shifted slightly inwards, the next spot is outwards, and that between the submedian nervure and the lower median nervule is moved inwards and forms an acute angle on its interior edge, thus joining the long spot reaching the internal nervure ; submarginal shading darker, with a faint pale outward border; small spot at anal AMBLYPODIA GROUP OF THE LYCENID. 30 angle (there is no lobe) black, surmounted with a green iridescence, which is extended slightly into the marginal area between the submedian nervure and the second median nervule. 2. Upperside: both wings as in male, but with a broad costal apical and posterior marginal border to the primaries ; in the secondaries the costal margin only is broader than in the male. Underside precisely as in male. This is probably the most variable species of the genus; the blue varies through violaceous to purplish up to lustrous cobalt, and the underside presents some wonderful divergencies. I have before me about 60 specimens, and have examined a considerable number more, and find that the ground-colour underneath runs from the typical pinkish through all shades up to deep rich purplish red and blackish red; then, again, a not uncommon form is rich reddish closely irrorated with bronzy-green scales, giving the insect a very beautiful colour. The markings are also subject to very great exaggera- tion, the spots being enlarged all over the wing, so that in the hind wing nearly all the spots and bands touch one another. The richest example before me is one from South- east Borneo, kindly lent me by Herr Ribbe. The species is evidently common all over the Malay Archipelago. I have specimens from New Guinea, Halmaheira, Waigeoce, New Britain, Mioko, Aru, and Queensland. Boisduval’s description of this species would suit almost any of the group. It is necessary therefore to take the first figure, which is Hewitson’s (B. M. Cat. p. 4, pl. ii. figs. 4, 5, & 6); he knew Boisduval’s insect, and the figures 4 and 5 are good, so that the insect can be recognised at once therefrom. Mr. Grose-Smith has described two species, A. adherbal and A. appianus, both of which must, I fear, be sunk under meander Boisd. As already mentioned, I have had a very large number of specimens from the Malay Islands and elsewhere, showing a most remarkable range of variation, specimens varying from the same locality, some of which quite correspond with Mr. Grose-Smith’s figures, and I cannot but regard them as aberrations. The genitalia have subovate clasps, and the penis is rather slender. ARHOPALA CONSTANCE de Nicéville. Arhopala constancee de Nicéville, Journ. A. 8. B. vol. lxiii, pt. 2, p. 34. n. 80 (1894). Hab. South Andaman Isles. Expanse, 2, 48 mm. Mr. de Nicéville described this species from a unique example taken at Port Blair by the late Mr. R. Wimberley. It may be merely an aberration of meander or one of its allies. ARHOPALA AMYTIS Hewitson. (Plate IV. figs. 9 & 9a.) Arhopala amytis Hewitson, Cat. Lycenide B.M. p. 4. n. 16, pl. i. figs. 7,8, & 9, 3 (1862); id Ill. Diurn. Lep., Lye. p. 4. n. 4 (1869). you. xvi.— PART 1. No. 5.—August, 1903. FE a4 MR. G. T. BETHUNE-BAKER: A REVISION OF THE Amblypodia ribbet Rober, Iris, vol. i. p. 70, pl. v. fig. 5, ¢ (1886). Amblypodia cryonthe Miskin, Proc. Linn. Soc. N.S.W. (2) v. p. 43 (1890). Arhopala amytis Hew., Miskin, Annals Queensland Mus. no. 1, p. 70 (1891). Arhopala cryonthe Miskin, 1. c. Hab. Queensland; Aru; Waigeoe; Woodlark Island; Fergusson Island; Tokal Ikey ; and New Guinea. Expanse, ¢ 47-54, 2 48-58 mm. ¢. Upperside: both wings brilliant morpho-blue, shading into purplish towards the costal and posterior margins of the primaries and towards the posterior margin of the secondaries. Costa of primaries narrowly black ; posterior margin broader, increasing in width at the apex. Costa of secondaries broadly blackish ; posterior margin narrow, abdominal fold entirely brownish ; tail black, tipped with white; lobe scarcely deve- loped. Underside: both wings reddish brown, with all the spots of a darker hue with pale margins. Primaries: inner marginal area paler brown, apical area washed with very pale pinkish lavender; three discal spots increasing in size, the upper portions of each side of which are margined with pale silvery bluish, the usual two spots below the discoidal cell; transverse band continuous, inclined outwards, terminating on the lower median nervule, below which, but unconnected with it, are one or two semi-coalescent spots shifted well inwards ; submarginal row of darker spots very indefinite; the upper and apical areas are dusted over with grey. Secondaries with a broad subcostal streak of pinkish lavender, with the four basal spots followed by the three large ones as usual; cell closed by a large quadrate spot, below which is a small spot filling the terminal space between second and lower median nervules. Transverse band unbroken, extending round up to the internal nervure ; the outer edge of this band is very jagged and irregular. Submarginal shading continuous, between which and the band is a broad space of grey. Margin rather broadly and very darkly bordered; anal spot small, black, surmounted by slight green iridescence, which extends slightly over the marginal area between it and the second median nervule. ?. Upperside: both wings brilliant, slightly greenish morpho-blue ; primaries, very broad apical and outer marginal areas and a narrow costal margin black. Secondaries with costa and apical margins very broadly blackish; posterior margin not so broad, but increasing in width at the anal angle; abdominal fold brownish. Underside as male. I have another form of the female which has the blue darker, almost like the male, but a shade paler perhaps. The genitalia have the clasps peculiar, with excised extremities; the penis has a globular orifice. AMBLYPODIA GROUP OF THE LYCANID#. (Sis) on Var. RIBBEI Rober. 3¢. Upperside rich purplish or violet-blue; primaries with very narrow black costal and posterior margins ; secondaries with broad blackish costa and very narrow outer margin; tail moderately long, stout, black, tipped with white; lobe slightly developed. Underside exactly as in amytis Hew. ¢. Upperside paler than the male and more violet in tone; primaries with very broad costa, and much broader posterior margin and apex; secondaries with very broad margins all round. Underside exactly as in amytis Hew. This insect is evidently a local form of Hewitson’s species amytis, but it differs in the deep tone of the blue of the upperside, and the very narrow margins of the males. I have specimens of it before me from Aru (the types), Tokal Key, one male from Waigeoe, and one male from Ati Ati Onin (New Guinea). Miskin has also recorded two specimens of it from Cape York under the name eryonthe, but, as Rober described it two years previously, Miskin’s name falls. Amytis is somewhat variable as to the tone of the blue, and decidedly variable as to the pale subcostal streak on the underside of the secondaries, which is subject to obsolescence. I have several specimens in which it is practically obsolete, and many where there is only a slight trace of it, and these forms occur irrespective of locality. Through the kindness of Mr. H. Druce, I have been able to examine some specimens from Fergusson and Woodlark Islands, all of which have no trace of the pale sub- costal streak ; one female from Fergusson Island is unusually large and robust. ARHOPALA XONE (Hewitson). (Plate ILI. fig. 22, ¢.) Amblypodia exone Hewitson, Il]. Diurn. Lep., Lye. p. 5. n. 15, pl. iil. figs. 20 g, 242 (1869). Hab. Waigeoe; New Guinea; Halmaheira; Aru; Kiriwini Island. Expanse, ¢ 44-51, 9 43-49 mm. ¢. Upperside: both wings brilliant silvery morpho-blue, purplish at the costa and apex of the primaries, which have a very narrow costa, rather broader posterior border of black ; the secondaries have a broadish black costa, narrow black posterior margin, and dark grey abdominal fold; the tail is long, stout, and white-tipped, and above it at the end of the second median nervule is a tooth-like projection. Underside: both wings very dark; the primaries have three discal spots traced out by silvery-bluish edges, almost obsolete as to the internal edge of the third spot, which has a pale outer edging; otherwise the ground is very broadly quite pale, with more or less of a lilac suffusion, the paleness extending throughout the submedian area. Below the second and third discal spots are two other obscure spots divided by the lower median nervule. The transverse band in the middle of the pale area is continuous, slightly curved, broadish, and generally slightly tapering narrower to the lower median nervure, where it terminates; it occasionally has a distinct bend or angle. Submarginal row present as usual. Secondaries: the basal three-quarters of the wing very dark indeed and F 2 36 MR. G. T. BETHUNE-BAKER: A REVISION OF THE markless, followed by a pale broad internal edging to the usual submarginal row, this band of pale edging having a lilac tinge at the anal area; a black spot at the lobe and between the second and lower median nervules, between and over which is a scattering (more or less plentiful) of greenish metallic scales. 2. Upperside: both wings blue, slightly paler and less lustrous than in male. Primaries with costa broadly, posterior margin very broadly, and apex yet more broadly black ; secondaries with costa and upper part of the posterior margin very broadly black indeed, but the margin of the anal area much more narrowly black. At first sight this unmistakable species scarcely looks in its right place here; its size and shape, however, bring it at once into this group, and a careful examination of the under-surface pattern soon locates it; the markings of the primaries are exceed- ingly close to micale Blanch., whilst in a female in my collection the transverse band of the secondaries is quite traceable, and in other specimens before me isolated spots are discernible, so that I have no hesitation in fixing it here. ARHOPALA SOPHROSYNE (Grose-Smith). Amblypodia sophrosyne Grose-Smith, Eut. Mo. Mag. xxv. p. 300 (1889). Arhopala sophrosyne Grose-Smith, Druce, Proc. Zool. Soc. Lond. 1891, p. 370, pl. xxxi. fig. 9, g. Hab, Guadaleanar Island (Solomon Group). Expanse 49-53 mm. This species is described by Mr. Grose-Smith with tails, but figured by Mr. Druce without tails, his specimen having the tails broken off. It is a tailed species and very distinct. ARHOPALA MICALE Blanchard. Arhopala micale Blanchard, Boisd., Hombron & Jacquinot, Voy. Pole Sud, Lép. pl. iii. figs. 11, 12. Amblypodia micale Boisd. MS., Hewitson, Cat. Lycenidee B.M. p. 7 (non adatha Hew. pl. iv. 3, figs. 29, 30, 31) (1862); id. Ill, Diurn. Lep., Lyczenidz, p. 8 (1863). Amblypodia superba Rober, Tris, vol. i. pt. 4, p. 199, Taf. ix. fig. 3 (1887). Arhopala micale Blanch. et superba Rober, Staudinger, Ex. Schmett. p. 280, Taf. 96, g (1888). Hab. (Amboina?); New Guinea; Batchian; Halmaheira; Gilolo. Expanse, ¢ 52-61, 9 45-59 mm. g. Upperside : both wings lustrous ultramarine-blue, shading into purplish in certain lights on the costal and posterior margins in primaries; costa in secondaries broadly blackish, posterior margin narrow; abdominal fold entirely fuscous up to submedian nervure ; tail tipped with white. Underside: both wings darkish brown, all the markings dark brown, with paler margins; primaries with three discoidal spots increasing in size, partly margined with pale blue, this colour being confined to the inner edge of the large spot closing the cell, the two usual spots just beneath the cell; the transverse band, edged with whitish on the upper part of the inner edge, almost AMBLYPODIA GROUP OF THE LYCANIDA. oF touches the third cell-spot and generally tapers down considerably to the lower median vein, where it ends; on the other side of this vein is another spot liable to obsoletion, shifted well inwards; outside this band is a pale grey shading, forming the inner edge to the almost obsolete submarginal row of spots. Secondaries with four basal spots, the second and fourth shifted inwards, beyond which are three larger ones; cell closed by a long quadrate spot, below which is a small one occupying the termination of the space between the second and lower median nervule. Transverse band continuous, extending round to the internal nervure; this band has exceedingly uneven and jagged margins, but it is unbroken, though fractured to a certain extent at the lower median nervule; outside this fascia, and following its exact course, is a line of unbroken crescentic pale grey marks followed by a pale brown submarginal area. Anal spot black, surmounted by a bluish or greenish scaling, which extends up to the second median nervule and is margined by a whitish line; there is also a small black spot at the margin of the bluish scaling between the second and lower median nervule. Tail stout, blackish, tipped with white. ‘here is often a pale subcostal stripe more or less developed. @. Upperside: both wings much paler than the male; primaries with costa of medium width, posterior dark margin very broad, apical area broader still ; secondaries with broad margins all round. Underside exactly as the male. The form usually found in New Guinea, whence I have several specimens, is deep purplish, almost the colour of A. hercules Hew., but not lustrous, with the underside paler and more uniform. I have, however, one typical female. In this species the underside pattern is somewhat liable to obsoletion, more particularly in the secondaries; this is caused by a darkening of the ground-colour and a slight lightening of the markings. The insect should be easily recognised on account of the brilliancy of the blue, whilst the under-surface pattern is also well marked in tone of colour as well as pattern. Every author that has hitherto mentioned this species has added a little to the confusion of its identification. The original figure is bad beyond comparison, and Blanchard’s description is much too vague to serve for identification. We then come to Hewitson, who first confused it with his own adatha, but afterwards discovered his error (Il. Diurn. Lep., Lyc. /. ¢.) and raised both to specific rank quite correctly ; unfortunately, however, he did not separate the two species in his cabinet, where they remain up to this moment mixed together. He knew Boisduval’s specimens, and there is no difficulty whatever in separating the two species. Staudinger next refers to it (/.c.), and says that he regards amytis Hew. as a synonym of this insect; but he is wrong in this, the two species being quite distinct: superba Rober is, however, synonymous; I have the type of this species before me, and it is like typical micale, except that the underside of the wings is paler and the whitish is more developed and the paler subcostal stripe of the secondaries strongly marked. This form was described from Batchian, but I have it also from Halmaheira, Gilolo, 38 MR. G. T. BETHUNE-BAKER: A REVISION OF THE and Ternate, with these characteristics more or less developed, and therefore the name should fall, as it is quite impossible to draw a line between this and typical mécale, and all occur indiscriminately in the same districts. Staudinger’s reference of amytis to this species leads me to say that micale can be at once separated from amytis by the rounder apex, narrow black outer margin, the fuller secondaries, the intense blue of the upper surface, and the strong brown tint of the under surface. It appears to be confined to the Malay Islands. I have seen no specimens from Queensland, where it may be replaced by Hewitson’s species amgtis. The figures given by Hewitson (Cat. Lyceenide, /. ¢.) do not belong to this insect, but to his adatha. ARHOPALA KIRIWINII, n. sp. (Plate I. fig. 10, ¢.) fab. 'Trobriand and Fergusson Islands, Expanse, ¢ & 9, 54-58 mm. 3. Upperside: both wings dark purplish blue; primaries with linear costal and outer margins; secondaries with a broad costa and linear outer margin, fold ample, greyish. ‘Tail fairly stout, white-tipped. Underside: both wings warm brown washed with purplish, with darker spots palely edged. Primaries with three increasing cell- spots, the first one very small, below the third a largish one in the lower median angle ; transverse band composed of six confluent spots, each increasing in size up to the fifth spot, which latter is very large, the sixth spot tapers slightly towards the lower median nervule, the first four spots are outwardly oblique, the fifth and sixth have their inner margins shifted well inwards, and below these is another smaller spot shifted inwards ; submedian area paler; submarginal row fairly distinct, marginal row dark and almost coalescing with the submarginal in its upper part. Secondaries with a small spot at the extreme base of the costa and four small basal spots, the third shifted well out- wards, followed by three large spots below each other, the lowest one being elongated somewhat outwardly ; cell closed by a large subquadrate spot, below which is another in the submedian angle connecting it with the elongated spot just mentioned; trans- verse band composed of cight large spots, the first two extremely large, touching the spot closing the cell, the third shifted well outwards, fourth slightly out, fifth with its inner margin shifted far inwards, touching the bottom edge of the spot closing the cell, sixth spot shifted slightly outwards and larger, seventh angular spot very large and confluent with the eighth very large rather long spot; submarginal row indefinite ; marginal row narrow, just below the apex is a patch of dark brown clouding; a black spot at the anal angle edged above with bright metallic-blue scales, which are also scattered more or less along the margin up to and just beyond the tail. 2. Upperside: both wings brighter blue and paler thanin the male. Primaries with a broad blackish costa, increasing in width beyond the apex of the cell; outer margin very broad; secondaries with a very broad blackish costa and a rather broad AMBLYPODIA GROUP OF THE LYCAENID. 39 outer margin, this margin is decidedly less broad than usual. Underside just like the male except that the ground-colour is paler, so that the markings stand out in greater contrast. I have before me three males and one female of this insect from Mr. Druce’s collection, and [ have recently received other specimens from the same locality: it will come before centaurus Fabricius, but may easily be separated from that species by the difference in the blue and by the remarkable transverse bands in both wings on the underside ; the general pattern and appearance are also different, and there should be no difficulty in recognising the species. ARHOPALA CENTAURUS (Fabricius). (Plate IV. figs. 10 & 10 a.) Papilio centaurus Fabricius, Syst. Ent. p. 520. n. 829 (1775) ; id. Sp. Ins. vol. ii. p. 117. n. 523 (1781) ; id. Mant. Ins. vol. 1. p. 68. n. 646 (1787). Hesperia centaurus Fabricius, Ent. Syst. vol. i. pt. 1, p. 275. n. 63 (1798). Polyommatus helus Godart, Enc. Méth. vol. ix, p. 652. n. 119 (1828). Polyommatus centaurus Godart, Enc. Méth. vol. ix. p. 658. n. 138 (1828). Amblypodia centaurus Horsfield, Cat. Lep. Mus. E.I.C. p. 102. n. 33 (1829); Hewitson, Cat. Lycenide B.M. p. 3. un. 15, pl. i. figs. 10,13 ¢ (? figs. 12 ¢, 11 2) (1862); Moore, Proc. Zool. Soc. Lond. 1865, p. 775; Butler, Cat. Fab. Lep. B.M. p. 179. n. 1 (1869) ; Druce, Proc. Zool. Soc. Lond. 1874, p. 107 ; Butler, Trans. Linn. Soc. Lond., ser. 2, Zool. vol. i. p. 548. n. 1 (1877). Amblypodia pseudo-centaurus Doubleday, List Lep. B.M. vol. ii. p. 24 (1847). Narathura centaurus Distant, Rhop. Malay. p. 261. n. 1, pl. xxi. figs. 4 3,5 9 (1885). Nilasera centaurus Moore, Journ. Linn. Soc. Lond., Zoology, vol. xxi. p. 44 (1886). Arhopala centaurus, var. coruscans Wood-Mason and de Nicéville, Journ. A. 8S. B. vol. xlix. pt. 2, p. 234. n. 54 (1880). Amblypodia nakula Felder, Wien. ent. Monatsch. vol. iv. p. 395. n. 4 (1860) ; Druce, Proc. Zool. Soc. Lond. 1878, p. 853; Kheil, Rhop. Nias, p. 33. n. 121 (1884). Arhopala nakula Felder, Reise Novara, Lep. vol. ii. p. 222. n. 224, pl. xxix. fig. 14 (1865) ; Moore, Proc. Zool. Soc. Lond. 1877, p. 590, et 1878, p. 835; Wood-Mason and de Nicéville, Journ. A. S. B. vol. 1. pt. 2, p. 251. n. 77 (1881). Nilasera pirama Moore, Lep. Cey. vol. i. p. 116, pl. xliii. figs. 3,30 ¢,3a ?, 3c larve and pura (1881), Nilasera pirithous Moore, Proc. Zool. Soc. Lond. 1883, p. 531. Arhopala centaurus de Nicéville, Butt. Ind. vol. iii. p. 234. n. 786 (1890). Narathura centaurus Fab., Hagen, Berl. ent. Zeit. xxviii. p. 152 (1892). Arhopala centaurus Druce, Proc. Zool. Soe. Lond. 1895, p- 588, et 1896, p. 671. Arhopala centaurus Fabricius, de Nicéville and Martin, J. A. S. B. vol. Ixiv. pt. 11. p. 464 (1895). Hab. India, K. Himalayas; Assam; Bhutan; Burma; Malay Peninsula; Siam ; South India; Ceylon; Andaman Isles ; Sumatra; Java; Borneo; Sumba and Sumbawa. Expanse, ¢ & 2, 58-64 mm, This species and its various forms are now so well known that I will only make 40} MR. G. T. BETHUNE-BAKER: A REVISION OF THE reference to those forms in which I differ somewhat from the conclusions arrived at by de Nicéville regarding them. Coruscans (Wood-Mason and de Nicéville) and pirama Moore.—These two forms are the same; I can trace no difference whatever in the colour of either males or females from Ceylon and the Andaman Isles. 3. Upperside: both wings brilliant ultramarine-blue, sometimes paler at the base, but not always, with a very narrow black costa and outer margin. 2. Like the male in tone, but with the bases of the wings beautiful silvery greenish metallic blue extending over nearly half the surface, with costa and outer margin in the primaries very broadly black, increasing towards the apex ; in the secondaries the costa is very broadly black and the outer margin by no means broad, but increasing in width towards the anal angle. De Nicéville (d. ¢.) says, referring to coruscans: “'The male may be known from A. pirama on the upperside by its more brilliant and lighter blue coloration, and in the female the brilliant basal blue portion is lighter and also better defined from the darker deep blue portion beyond.” ‘This is perfectly correct as regards de Nicéville’s specimens, which, through his great kindness, I have before me; but it is not so at all with those I have had from my friend Dr. Staudinger and others, the males being precisely the same in every respect, and I have only one female to which it would apply—all the others from Ceylon are absolutely the same in each particular as those from the Andaman Isles. Pseudocentaurus Doubleday (amazona Pagenstecher).—This I take to be the Java form of centaurus, of which I have a number of specimens before me: the blue of some, not all, is rather brighter and bluer than ordinary centawrus Fabricius ; it is, however, a curious fact that my only specimen, out of a great number, which is exactly typical with the type specimen of centawrus in the Banksian collection is one from Java, and is not of the dull purple of the common form, but rather brighter and blver. The blue of these Javan specimens is not nearly so bright as coruscans, but it is of exactly the same shade as my Bornean specimens of centaurus Fab. and also as some specimens I have from Singapore and the west coast of Sumatra; the under surface is somewhat more variegated, but this is so very variable throughout all the forms that no name can be based on it; pseudocentawrus and amazona must therefore be sunk under centaurus Fab. The species being a very common one is also very variable, and it appears to me to serve no useful purpose to retain names for such a very slight variety as Doubleday’s, It is one of the most widely distributed species of the genus, but unfortunately it is scarcely possible to state correctly the distribution of the type-form. Pirithous seems 19 occur mainly in Sikkim and Assam and on the eastern side of India. Besides these districts and Bhutan, I have it from Upper and Lower Burma; the type-form occurs with it in both these latter localities. Coruscans Wood-Mason and de Nicéville, as far AMBLYPODIA GROUP OF THE LYCENIDA. 4] as I can trace, appears to be confined to Ceylon and the South Andaman Isles, though a specimen I have from Nias comes very close to it. The genitalia of all these are indistinguishable from each other, and they also are exceedingly close to A. micale Blanch.; the tegumen of centawrus is not so large and the clasps less excised at the extremities. ARHOPALA EUPOLIS (Miskin). (Plate I. fig. 11, ¢.) Amblypodia eupolis Miskin, Proc. Linn. Soc. N.S.W. (2) v. p. 42 (1890). Arhopala eupolis Miskin, Annals Queensland Mus. no. 1, p. 70 (1891). Hab. Queensland, Tokal Key Island. Expanse, ¢ & 2, 52-58 mm. 3. Upperside: both wings dull violet; primaries with the costa and outer margin very narrowly brown ; secondaries with the costa widely and the outer margin very narrowly brown; abdominal fold grey. ‘Tail longish, very broad, black, tipped with white; nervures on each side slightly elongated, so as to form very short tooth-like projections. Lobe moderately well developed. Underside: both wings dull, very pale brownish grey, uniform in tint, with darker spots palely encircled. Primaries with three increasing cell-spots, the third quadrate; below the second and third are two large spots divided by the lower median nervule; transverse band composed of six confluent spots, forming an almost straight fascia of equal width placed somewhat obliquely outwards; below the band is a seventh spot shifted inwards; marginal and submarginal bands very obscure. Secondaries with four basal spots, the lowest shifted inwards, followed by three smallish spots below each other; cell closed by a narrow subquadrate spot, below which is a small one; transverse band highly irregular and peculiar, composed of two large spots below each other which touch the narrow sub- quadrate spot, below this and shifted outwards is a broad irregular fascia occupying the place of spots 3-6 inclusive, the seventh angular spot is shifted inwards and is almost confluent with the eighth long spot; marginal and submarginal fascia very obscure ; anal lobe blackish, edged above, and the space between it and the second median nervule filled in, with bluish metallic scales. 2. Upperside: both wings bright pale shining blue, darker towards the edges; primaries with the costa broadly blackish, increasing towards the apex, where there is a large blackish patch and the outer margin is very broadly blackish ; secondaries with the costa very broadly dark brown and the outer margin narrowly so, but increasing somewhat towards the anal angle. Underside as in the male. This species is evidently the Australian form of centawrus Fab. The colour above is duller and paler and the dark outer margin is of a width intermediate between that of that species and of pirithous Moore. On the underside the colour is very much paler and is altogether more uniform in tone, the pattern is smaller in the primaries, there is no metallic edge to the cell-spots, and in the secondaries the lobe is more VOL. XVI.—PART 1. No. 6.—August, 1903. G 42 MR. G. T. BETHUNE-BAKER: A REVISION OF THE developed. The female closely resembles coruscans W.-M. & de N. in the coloration of the upperside. This is an interesting species, forming as it does a good “ transit ” to the species amatria de Nicéville. The hue of the underside varies somewhat, as in centaurus Fab., being occasionally washed with violet or grey, but it is always quite pale in tone. The species has been known for years in the larger collections, having been placed either with amytis Hew. or centaurus Fab., but Mr. Miskin discriminated it in 1890. Until this year it has only been recorded from Queensland, but I have just received from Dr. Staudinger a perfectly typical female from Tokal Key. ARHOPALA AMATRIX de Nicéville. Arhopala amatrix de Nicéville, Journ. Bombay Nat. Hist. Soc. vol. vi. p. 370. no. 19, pl. G. figs. 23 ¢@, 24 2 (1891). Hab. Burma. Expanse, ¢ & 9, 52-58 mm. This species is a close ally of amantes Hew., but can be immediately distinguished by the broad black borders of the male, while the blue is not nearly so deep and brilliant. So far as is yet known, it is confined to Burma, but probably replaces Hewitson’s species there ; de Nicéville seems also to intimate this in his original description, for he says “ True amantes apparently do not occur in Burma.” ARHOPALA AMANTES (Hewitson). (Plate IV. figs. 11 & 11a.) Amblypodia amantes Hewitson, Cat. Lycenide B.M. p. 4. n. 17, pl. ii. figs. 2 & 8, ¢ 9 (1862). Arhopala amantes Wood-Mason and de Nicéville, Journ. A. S. B. vol. 1. pt. 2, p. 250, n. 76 (1881). Nilasera amantes Moore, Lep. Cey. vol. i. p. 115, pl. xliv. figs. 24,26 3,2 9, 2c larva and pupa (1881). Nilasera apella Swinhoe, Proc. Zool. Soc. Lond. 1886, p. 429. n. 64, pl. xl. fig. 4, 2. Hab. India; Ceylon; 8. Andaman Isles; Makassar; Sumba; Timor; Babber Isles (Burma). Expanse, ¢ & 9, 57-62 mm. De Nicéville (Butt. Ind. vol. ii. p. 240) says it is widely spread throughout India, occurring “all along the foot of the Himalayas, in the plains of Bengal, the Central Provinces, and in Bombay, thence southwards to Travancore and Ceylon. Its flight is strong and rapid, and it conceals itself with wings always folded under leaves (Hutchinson).” It is readily recognisable from ordinary centaurus Fab. by the brilliancy of the blue and by the brighter females, but it is not so easily separated from coruscans W.-M. & de Nicéville. The blue is, however, deeper and more brilliant and the underside AMBLYPODIA GROUP OF THE LYCZNIDA. 45 is not subject to such violent contrasts as in that insect, the tone being much more uniform and the pattern more compact. Hewitson pointed out that the fourth spot of the transverse band of the primaries was projected outwards; this, however, is not always so. I have specimens from Sumba, Timor, and the Babber Islands, and in each of these the band is unbroken, though for all my specimens from the mainland and from Ceylon his remark is correct; the anal lobe, moreover, is always very much more developed and is deep velvety black beneath. I have seen no specimen from the Andaman Isles. De Nicéville says (/. c.) that he has seen one which is smaller and in which the area of the blue coloration of the upperside is different (does this mean the specimen was a female ?), and all the markings below are smaller, more compact, and prominent. This specimen is not among those Mr. de Nicéville kindly lent me, so I can only record his note. [Since writing the foregoing note on the Sumba, Timor, and Babber specimens, I have very carefully re-examined them, and am inclined to regard them as a transit from this species to araxes Felder ; the upperside colour is almost the colour of amantes, but the underside pattern is that of Felder’s insect. The range of true amantes will therefore be restricted to the mainlands and the Andaman Isles, the Makassar specimens T believe to be the transition form. In the genitalia the clasps assume a subtriangular form, whilst the base of the penis is angular instead of being almost straight. | ARHOPALA ARAXES Felder. Arhopala araxes Felder, Reise Novara, Lep. p. 224, pl. xxix. figs. 3&4 9,5 ? (1865). Amblypodia arazes Hewitson, Ill. Diurn. Lep., Lyczenide, p. 14a. Hab. Celebes, Mangioli, Sula Isles; Banka Isle. Expanse, ¢ 53, 2 50 mm. ¢. Upperside: both wings brilliant morpho-blue, shading into purplish near the margins, in the costal, apical, and posterior marginal areas, costa itself finely black, posterior margin less narrow. Secondaries: costa broadly blackish, posterior margin narrowly black, abdominal fold grey. Underside: both wings pale brown, with darkish spots with pale margins. Primaries with three discal spots increasing in size, all margined with bluish or silvery white except on their costal edges; below the spot closing the cell isa triangular spot, touching the inner edge of which is a large spot in the paler submedian area followed by another near the base; transverse fascia ending on or below the lower median nervule and consisting of six, seven, or eight confluent spots regularly below each other with a slight outward tendency, the seventh and eighth sometimes detached; submarginal row indistinct; the ground between these two latter fasciz is greyish. Secondaries more uniform in colour with four basal spots, the second shifted slightly, and the fourth considerably inwards, the latter beyond but touching the internal nervure ; outside these are two large spots below each other; cell G2 44 MR. G. T. BETHUNE-BAKER: A REVISION OF THE closed by a quadrate spot, below which is a small triangular one; transverse fascia consisting of the following spots—two large upper almost confluent ones, two others below shifted outwards, fifth spot small, shifted inwards a little, sixth very slightly outwards, the angular seventh spot well inwards and almost confluent with a long eighth spot which touches the internal nervure; submarginal row very indistinct ; anal spot black, over which are scattered metallic-blue scales, as also over the marginal area up to the second median nervule, and these have an internal edging of dark brown bordered with white and following the course of the transverse fascia up to the internal nervure. The second median nervure is elongated a little, so as to form a short tail beyond the white-tipped longish one. ©. Upperside: both wings rich dark brown ; the primaries with a bright pale blue patch on the lower half of the wing, extending two-thirds along the inner margin, while in the secondaries this blue patch extends almost to the posterior margin below the lower discoidal nervure. Underside as in male. This species is very close to amantes, but may be distinguished from it at once by the very brilliant morpho-blue of the male, Hewitson’s insect being very much darker ; whilst in the female of Felder’s species the blue is very much less in extent, is not lustrous, and is paler. Felder also points out another distinguishing character, viz. that the transverse band of the primaries is straight and not broken as in amantes Hew. Araxes is also decidedly the larger insect of the two. I have specimens of it before me from 8. Celebes, N. Celebes (Minahassa), Banka Isle, and Mangioli, Sula Isles, but I cannot trace it elsewhere, so that its distribution appears to be very confined and local. ARHOPALA AGLAIS Felder. Arhopala aglais Felder, Reise Novara, Lep. vol. iii. p. 223, pl. xsix. fig. 11, go (1865). Amblypodia aglais Hewitson, Ill, Diurn. Lep., Lyc. p. 14. (1869). Hab. Philippine Islands. Expanse, ¢ & 2, 56-64 mm. 3. Upperside: both wings deep blue with a slight mixture of purple with narrow black borders, the costa of the secondaries is rather broadly black. Abdominal fold up to the submedian nervure brown. ‘The submedian nervure and the second median nervule is each elongated a little, so as to form a very short tail on each side the usual tail. Underside: both wings brown, paler on the primaries below the lower median nervule, all the spots very slightly darker brown and margined with whitish. Primaries with three largish discal spots increasing in size; below the spot closing the cell is another small one, and below that a third larger spot occupying the terminal AMBLYPODIA GROUP OF THE LYCENIDA. 45 portion of the second median and lower median spaces, and below the lower median nervule, beneath the second discal spot, is another spot, larger and more distinct than usual; the slightly bent and uneven transverse band ends abruptly on the lower median nervule; submarginal row fairly distinct, with the inner margin shaded with violet in the upper portion, which shading extends to the outer margin of the transverse band. Secondaries with four basal spots, the second and fourth shifted inwards and outside these three large ones ; cell closed by a large subquadrate spot, below which is another touching the lowest of the three. Transverse band irregular, extending to the internal nervure and consisting of two large squarish spots touching the spot closing the cell, four smaller ones shifted outwards; the next spot forming the angle is shifted inwards and touches the spot extending up to the internal nervure ; submarginal row obsolete, but its position is indicated by the internal pale crescentic border; a large black velvety spot occupies the lobe and marginal space between the second and lower median nervules sprinkled with brilliant metallic-green scales, which extend into the marginal area between these two spots. @. Upperside: dark brown, with a large patch of deep blue (not a tinge of purple in it) extending over a considerable part of the primaries, but restricted in the secondaries almost to the area between the upper discoidal and submedian nervures. Underside as in male. In the colour of the upperside this species is near micale Blanch., but is darker ; it is, however, better placed here on account of the underside pattern, which is very dark and marked, the pattern being very prominently outlined with whitish and with no bluish metallic outline to the cell-spots of the primaries, so that there will be no difficulty in separating it from its near allies. Felder described it from Luzon and it appears to be confined to the Philippine Islands. In the genitalia the clasps are more ovate and the penis unusually short. ARHOPALA ACETES (Hewitson). (Plate III. fig. 18, ¢.) Amblypodia acetes Hewitson, Cat. Lycenide B.M. p. 5, pl. iii. figs. 14, 15, 9 (1862); id. Tl. Diurn. Lep., Lye. p. 4 (1869). Hab. Makassar, Celebes. Expanse, ¢ 64-67, 2 62-64 mm. This species can at once be recognised from ag/lais Feld. by its larger size, duller blue with merely linear black margins, and by its underside being much more uniform in tone and more spotted in pattern. It is one of the largest of the group and appears to be confined exclusively to Celebes and is a very distinct species. 46 MR. G. T. BETHUNE-BAKER: A REVISION OF THE ARHOPALA CAMDEO (Moore). Amblypodia camdeo Moore, Horsfield and Moore, Cat. Lep. Mus. E.I.C. vol. i. p. 41. n. 58, pl. ia. fig. 6, ¢ (1857); Hewitson, Cat. Lycenide B.M. p. 5. n. 28, pl. iil. figs. 25, 26, ¢ (1862). Arhopala camdeo Staudinger, Ex. Schmett. p. 280, pl. xevi., ¢ (1888); de Nicéville, Butt. Ind. vol. iii. p. 256 (1890). Hab. Sikkim tarai; Jalpaiguri; Assam; Chittagong Hills. Expanse, ¢ 58-64, 9 60-66 mm. This species can be recognised from any of its neighbours by the very pale lilac of its upperside, and by the pale grey of its underside with decided black spots encircled distinctly with whitish. It is a beautiful and very distinct species and is not very closely related to any, but is probably best placed here; de Nicéville says (J. ¢.), “Occurs in Sikkim tarai (not in the hills) in July and August; it is also found at Jalpaiguri and eastwards throughout Assam, and the late Mr. H. M. Parish obtained it in July in the Chittagong Hill tracts. It appears to be confined to the lower hills and the country at their feet.” ARHOPALA PaDUS Felder. (Plate I. figs. 12 ¢, 13 9; Plate IV. figs. 12 & 12a.) Arhopala padus Felder, Reise Novara, Lep. vol. ii. p. 230. Amblypodia viola Rober, Iris, vol.i. p. 199, pl. ix. fig. 4 (1887). Amblypodia padus Hewitson, Ill. Diurn. Lep., Lye. p. 146. n. 79 (1865). Hab, Halmaheira; Batchian; Banka; Mangioli, Sula Island. Expanse, ¢ 50-53, 2 51 mm. 3. Upperside: both wings pale lilac-blue, with fine black costal and posterior margins; in the secondaries the dark costa is rather broad. Abdominal feld pale greyish. Tail black, tipped with white. Underside: both wings darkish brown, spots precisely the same colour as ground, but broadly margined with whitish. In the primaries the space below the lower median nervule is pale greyish. Primaries with three discal spots increasing in size; above the second are two very small confluent spots and one small one above the very irregular and large spot closing the cell, while touching it beneath is a large one; the spot below the discoidal cell is large and distinct. The transverse band consists of four distinct spots—the first on the costa very close to the terminal discal spot, the other three shifted outwards and outwardly inclined, below which are two large spots shifted much inwards but with an outward inclination; immediately below these is a very small detached dot; the sub- marginal row of somewhat quadrangular spots distinct, extreme margin whitish. Secondaries with four largish basal spots below each other, the fourth shifted inwards, followed by three longer ones; cell closed by a large spot, below which and touching it is another shifted a little inwards; the transverse chain-like fascia is composed of four pairs of spots, the upper of the costal two touching the upper one of the three larger AMBLYPODIA GROUP OF THE LYCHNIDZ. 47 spots mentioned before ; the second two are shifted outwards, the next two inwards, and the final two further inwards again and quite detached from them; the last spot of these, touching the internal nervure and joining the angular one, is very long and extends almost up to the lowest basal spot; submarginal lunular row distinct. Lobe- spot black; spot between the lower and second median nervules black, over which, as also over the marginal space between them, are bluish or greenish metallic scales. Extreme margin finely white. @. Primaries: upperside blackish, with a large patch of very pale iridescent blue extending a little beyond the discoidal cell to the inner margin and occupying about two-thirds of the median area of the wing, in which area isa largish white patch showing through the iridescence. Secondaries: upperside greyish brown, with a pale slightly iridescent blue central area from the submedian to just over the discoidal nervure. Underside precisely as in the male. These descriptions are taken from a male and female in my collection from Halma- heira; the male agrees quite with Felder’s type. I have male specimens from Batchian which agree entirely with the type from Mangioli: however, the males differ slightly in having a broader suffused black posterior margin to the primaries; in some specimens this border is rather narrow, but in others it becomes very broad from the lower median nervule up to the apex. The two female specimens I have before me are entirely brown all over, with a few superimposed pale blue scales on the basal area, in one specimen only just sufficient to be seen, in the other more or less over the cell, and more thickly beneath the cell in the primaries only. Herr Rober described his species viola from Bangka, which is evidently only a slight variety of Felder’s species ; the colour is quite similar, but the apex and outer margin of the lower median nervule are broadly suffused with blackish. The insect is evidently variable both as to the width of the border and the tone of brown beneath: I have specimens from Mangioli in which the border is moderately narrow, and others in which it is almost as wide as in the Bangka specimens of Réber; whilst beneath in some specimens the ground-colour is rather pale with the pattern slightly outlined with whitish, and in others it is very dark with very marked whitish outlines. In Mangioli specimens the female is entirely brown with a slight scattering of pale blue scales over the basal area, and occasionally even these are almost obsolete. This species has apparently not been recognised by anyone since Felder described it : all the specimens before me have been sent me as ertdanus Feld. (a species about two- thirds its size), though the Felders in their work (/. ¢.) speak of it as reaching the size of camdeo Hew., and on the previous page they speak of eridanus as a third smaller than camdeo. There is now, however, no doubt as to the correct identity of each, for I have examined the original types very carefully. The distribution is narrow; I have only seen specimens from Halmaheira, Batchian, Bangka, and Mangioli (Sula Islands), and 4§ MR. G. T. BETHUNE-BAKER: A REVISION OF THE can find no records of it at all. With a distinct species like this we may expect to find the genitalia somewhat specialised; as will be seen from fig. 12, Plate IV., the clasps are peculiar, and the tegumen is not so ample, whilst the penis is small. ARHOPALA ELFETA (Hewitson). Amblypodia elfeta Hewitson, [ll]. Diurn. Lep., Lye. p. 143, pl. iiia. fig. 40, ¢ (1865). Hab. Sulu (Celebes). Expanse, 2, 54 mm. ?. Upperside: both wings brown, with a slight scattering of very pale silvery-blue scales over the basal area, not extending beyond the cell. Tail present but broken. Underside dull brown, with all the spots but slightly darker than the ground and distinctly encircled with pale greyish or whitish. Primaries with three large cell-spots, the third very large and irregular, extending down to the lower median nervule, with a very small spot above it on the costa. Transverse band consisting of six spots—first spot small, second, third, and fourth evenly and obliquely inclined outwards and shifted well out from the first small spot ; fifth spot shifted well inwards, almost touching the third large cell-spot; sixth spot below the fifth but with an outward inclination ; below this is another very small spot; submarginal row fairly distinct, marginal row obscure ; subdiscal area dark, submedian area pale. Secondaries with four large basal spots, the fourth shifted inwards; above the uppermost is another on the extreme basal edge of the costa; these are followed by three large spots below each other, the lowest being prolonged outwards. Cell closed by an irregular spot, below which is a small one in the lower median angle. ‘Transverse band composed of eight distinct spots—the second shifted outwards, the third and fourth shifted outwards, fifth inwards, sixth outwards, seventh angular spot well inwards and touching the eighth which is unusually long and straight ; submarginal row distinct and crescentic, marginal row fairly distinct; three blackish spots at the anal angle, the middle one almost covered with submetallic pale bluish scales. This species is a very close ally of azata de Nicéville, of which at present only the male is known; the spots are larger, especially that closing the cell, but the pattern is similar, and I expect when more material comes to hand that de Nicéville’s name will probably fall as a synonym of this. ARHOPALA AzaTa de Nicéville. Arhopala azata de Nicéville, Journ. A. 8. B. vol. lxiv. pt. i. p. 466. n. 404 (1895) ; id. Journ. Bomb, Nat. Hist. Soc. vol. x. p. 172. n. 21, pl. T. figs. 32 ¢, 33 ¢ (1896). Hab. Malay Peninsula (Perak) ; Sumatra. Expanse, ¢, 45 mm. This species, coming between padus Felder and eridanus Felder, may be known from the former by its much smaller size, the colour of the upperside is very much AMBLYPODIA GROUP OF THE LYCANIDA. 49 darker and duller, the underside is close in both colour and pattern. From eridanus it can be separated in the same way as regards its upperside, whilst beneath it differs much more, Felder’s species having a great deal more white, whilst this insect is more uniform and sombre in general tone. ARHOPALA ERIDANUS (Felder). (Plate I. fig. 14, ¢; Plate IV. figs. 13 & 132.) Amblypodia eridanus Felder, Sitzungsber. Akad. Wiss. Wien, xl. p. 453 (1860). Arhopala eridanus Felder, Reise Nov., Lep. vol. ii. p. 229, pl. xxix. figs. 16 & 17. Amblypodia polita Rober, Iris, vol. i. p. 199, pl. ix. fig. 14 (1887). Amblypodia eridanus var. dilutior Staudinger, Iris, vol. ii. p. 123 (1889). Amblypodia eridanus Feld., Hewitson, Il. Diurn. Lep., Lycenide, p. 5 (1863). Hab. Amboina, Palawan, Cagayan (a small island off the coast of N. Borneo), and Balabak. Expanse, ¢ & 2, 46-50 mm. 3. Upperside: both wings pale lustrous violaceous blue, with very narrow black costal and posterior margins in primaries; in secondaries the costa is broadly blackish, posterior margin very narrowly black, abdominal fold to submedian nervure whitish grey. Underside: both wings brown, spots scarcely darker than ground, broadly encircled with white. Primaries with three discal spots increasing in size, that closing the cell very large and irregular, at the internal upper extremity of which and over the middle spot are one or two small white marks, indicating the trace of other small spots; on the costa directly over the large spot is a very small one and beneath the cell-spot is another largish one ; below the middle discal spot is the usual one in the submedian area which is large and distinct ; the transverse chain-like fascia is composed of one small spot on the costa, beyond which and quite detached are three chain-spots outwardly inclined, followed below by two more shifted inwards and detached from the three above; submarginal lunular row fairly distinct; extreme margin finely white. Secondaries with a half-spot on the extreme internal edge of costa, the white margin of which extends into a white short line along the costa; four longish basal spots below each other, the fourth being shifted inwards, followed by three others beyond; cell closed by a longish reniform spot, touching which below is another long spot; the chain-like fascia is composed of two spots generally touching each other, but not always, below which are four chain-spots, the third one shifted well inwards; the angular spot, detached and shifted inwards, touches the very long spot which almost reaches the fourth basal spot; submarginal lunular row distinct, a largish black spot occupies the lobe and the marginal space, above which and over the space intervening are scattered bluish-metallic scales. Extreme margin finely white. The second median nervule and the submedian nervure are slightly elongated, so as to form the trace of a tail on each side of the usual tail. VOL. XVII.—PART. I. No. 7.—August, 1903. H 50 MR. G. T. BETHUNE*BAKER: A REVISION OF THE ?. Upperside: both wings dark brownish, the primaries with a large patch of very pale iridescent blue from the discoidal to the submedian nervure and slightly on to the internal margin; in the posterior area a patch of white shows through the blue iridescence; there is a large dark spot closing the cell, which also shows through the iridescence and off which the blue scales are liable to get rubbed. Secondaries with the pale blue iridescence not extending quite so far towards the posterior margin. Underside as in male, but with paler ground and darker spots (in contrast) and the white encircling broader. In the primaries the spots over the middle discal spot and up to that closing the cell are quite distinct though irregular in shape, and the submedian area 1s whitish. In the secondaries there is a half basal spot on the extreme internal edge of the costa. The type of this species, a single female, was described by Felder from Amboina. I have therefore described a male from that island and also a female agreeing almost exactly with the type; the colour of the underside is not, however, constant in examples from the same locality—some from Amboina being quite pale brown, others dark brown ; the pattern, however, appears not to be subject to variation. In the female the pale upperside patch occupies a greater or less area in different specimens quite irrespective of locality. Dr. Staudinger has described a variety from Palawan as var. dilutior, of which I have the type male and female before me; the male is precisely the same as Amboina specimens of eridanus Feld., and the female differs only in the extent of the pale upperside, which is slightly greater than in the type—this, however, varies in different specimens. Herr Rober has also described his species polita from Ceram (Iris, /. ¢.), the type of which is also before me, and it is absolutely the same as a specimen from Amboina, so that both these names must sink. I have before me, through the kindness of Mr. Cator, a series of eight specimens, four males and four females, from Cagayan, which present no differences from evidanus Feld., except that the tone of colour of the male is distinctly bluer than usual. Dr. Staudinger calls his var. didutior a small local form of eridanus Feld.; this remark no doubt arises from the fact that he had not differentiated this species from the large padus of Felder, the two former being similar in size as well as colour and pattern. The genitalia are very similar to those of A. padus Feld., but the hooks attached to the lower part of the tegumen are recurved, as shown in Plate IV. fig. 13. ARHOPALA CAROLINA Holland. Arhopala carolina Holland, Novit. Zool. vol. vii. p. 78 (1900). Hab. Buru Island. Expanse, ¢ & 2, about 40 mm. I have not seen this species, which is described from specimens in Dr. Holland's collection. I admit it with some doubt, believing it will turn out to be merely a slight AMBLYPODIA GROUP OF THE LYCANIDA. dL variety of ertdanus Feld. Dr. Holland says it is like polita Rober (a synonym of eridanus), but the spots below, instead of being round, are elongated transversely. In many species of this genus I have noticed that the spots are occasionally liable to this curious malformation. ARHOPALA OPALINA (Moore). Nilasera opalina Moore, Proc. Zool. Soc. Lond. 1883, p. 581, pl. xlix. fig. 1, 3. Arhopala opalina de Nicéville, Butt. Ind. vol. iii. p. 257 (1890). Hab. Khasia Hills. Expanse, ¢, 38 mm. This species can be recognised at once from its neighbours by its small size and the quite distinct pattern of the under surface. It was described from a unique male specimen in the collection of Mr. Moore, which specimen is unique to the present time. ARHOPALA VIOLA Semp. (non VIOLA Rober). (Plate I. fig. 7, ¢; Plate III. figs. 29 & 30, 2.) Arhopala viola Semp. Schmetterl. Philip. i. p. 196. Hab. Mindanao. Expanse, ¢ 36, 2 36-37 mm. 3. Upperside: both wings violaceous blue, slightly lustrous ; primaries with a narrow black costa and very broad apical area and posterior margin, the latter becoming narrower at the anal angle; secondaries with costa broadish black, and posterior margin narrow ; abdominal fold grey; from the submedian nervure up to the upper discal nervule is a marginal row of black lunules decreasing in size. Underside: both wings darkish brown, spots rather darker than ground, encircled, or partially so, with whitish. Primaries with two basal spots; the third closing the cell is not margined at its costal or anal edge with white, but occupies a large irregular area from the second subcostal to the lower median nervule ; the transverse fascia consists of four confluent spots outwardly inclined, below which are three more confluent spots shifted inwards and reaching almost to the inner margin; submarginal lunular row distinct, extreme margin finely edged with white; there is a small light patch near the anal angle on the inner margin. Secondaries with half a spot on the extreme upper basal edge, the white border of which extends shortly along the costa; beyond and below this are the four basal spots, the lowest: shifted inwards on to the internal nervure; outside these are three largish spots, the top one of which is not margined with white on its costal edge; cell shifted inwards. Chain-like fascia very irregular, the second spot shifted outwards, the third still further outwards, the fourth outwards somewhat again, fifth well inwards, sixth slightly outwards, the seventh spot at angle inwards, touching the long final spot, which nearly reaches the lowest basalone. This fascia is not absolutely H 2 52 Mk. G. T. BETHUNE-BAKER: A REVISION OF THE fractured anywhere, though so irregular ; submarginal lunular row distinct. A black spot occupies the lobe and the marginal space between the submedian and lower median nervures and between the latter and the second median nervule; the first and last are slightly edged above with bluish-metallic scales, and the middle spot has its entire central area filled with these scales, so that it appears as two little black spots at each extremity. Extreme margin finely white. Tail black, tipped with white. 2. Upperside: both wings pale azure-blue; primaries with very broad costal and posterior margins. Secondaries with very broad costal and posterior margins, the latter narrowing at anal angle; abdominal fold grey; anal angle with two black spots, lobe with a small black spot. Underside as in male, but spots more broadly encircled with white, and in the primaries the submedian area is all whitish, with the usual spot below the middle discal spot. This area in the male is almost uniform in colour with the ground. I have before me a second female in which the blue is deep slightly lustrous ultramarine, and the markings of the underside are not so confluent, and the ground darker brown. A male and two females from Mindanao. Both these females are figured on Plate III. figs. 29 & 30. E Maus. Staudinger. This species is much smaller than eridanus Feld., and may be- known by this fact and also because the apex and outer margin are broadly blackish, whilst in the female the pale central area is not white but blue and is very restricted. The under surface is darker in both sexes, and with the white outlines of pattern narrower. ARHOPALA ANNULATA (Felder). (Plate IV. figs. 16 & 16 a.) Amblypodia annulata Felder, Lep. Amboin., Sitz. Akad. Wiss. Wien, xl. p. 452 (1860) ; Hewitson, Cat. Lycenide B.M. p. 9 (1862) ; id. Ill. Diurn. Lep., Lycenide, p. 5, pl. ii. fig. 21 (1865). Amblypodia tristris Rober, Iris, vol. i. p. 200, Taf. Ix. fig. 9 (1887). Amblypodia erebina Staudinger, Iris, vol. i. p. 123, Taf. 1. fig. 14 (1889). Hab. Amboina; Banka Island; Palawan and Saparua and Mindoro. Expanse, ¢ 32-38, ? 35 mm. 3. Upperside: both wings brown, with a few blue scales scattered over the basal half of the wings; in one specimen before me these blue scales occupy the whole of the basal, central, and submedian area, and in the secondaries the same area, but to a more limited extent. Tail brown, tipped with white, not long. Underside: both wings brown, with slightly darker spots encircled with whitish. Primaries with three discal spots increasing in size ; below the spot closing the cell is another smaller one, and in the submedian area is another beneath the second discal spot; the transverse chain- like band is composed of seven distinct spots, all touching each other, except that in the submedian area, which varies much in size. The first spot on the costa is very small, and is followed by three shifted outwards and outwardly inclined; the fifth is AMBLYPODIA GROUP OF THE LYCANIDA. 53 shifted well inwards; the sixth outwards again, immediately below which is the seventh detached spot. Submarginal lunular row distinct, beyond which is a marginal row of dark dashes interrupted at each nervure; extreme margin pale; subdiscoidal area slightly paler than ground. Secondaries: a half-spot at the extreme basal edge of the costa; four basal spots below each other, the lowest shifted inwards, followed by three large spots; cell closed by a large, somewhat irregular spot, beneath which is a small triangular one shifted inwards; chain-like band composed of eight spots, the two costal ones inclined outwards, the next two shifted well outwards, fifth spot shifted decidedly inwards, sixth outwards, the angular seventh spot slightly inwards, whence the long eighth spot runs upwards to the lowest basal one, which it sometimes touches ; submarginal lunular row distinct; marginal row less distinct ; extreme margin finely paler; a dark spot occupies anal angle, the marginal submedian space, and the marginal space between the lower median and the second median nervule, over which are slightly scattered, but more thickly in the central space, bluish-metallic scales. 9. Upperside: both wings as in male, but with a large central and basal and sub- median area of superimposed pale bluish scales; this area is more restricted in the secondaries. In one specimen before me this area is whitish, tinged with lavender. A dark spot shows through, closing each of the cells. Underside as in male, but spots darker and more broadly encircled with white. Dr. Staudinger’s species erebina, from Saparua, agrees precisely with annulata in every particular, and so does Réber’s tristis from Banka—in fact, Hewitson’s drawing might have been taken from Dr. Staudinger’s type, they are so close. This species may be easily recognised from viola Semp. by its brown colour and smaller size; the shape of the wings is also different, being narrower than in Semper’s insect. The female is variable; I have one in which the blue is quite bright (almost ultramarine), others almost brown, and one whitish. The superimposed blue scales are evidently rubbed off easily in both sexes, and are often very sparse indeed. In the genitalia the same recurved (at the base) hooks appear as in eridanus, with subtriangular clasps and small penis. ARHOPALA TEPHLIS (Hewitson). (Plate I. figs. 15 & 16,3; Plate IV. figs. 14 & 14 a.) Amblypodia tephlis Hewitson, Ill. Diurn. Lep., Lycenide, p. 14 ¢, figs. 57 & 58, ¢ (1865) Amblypodia bicolora Rober, Iris, vol. i. p. 71, Taf. v. fig. 7 (1886). Hab. Gilolo ; Celebes ; -Halmaheira. Expanse, ¢ & 2, about 47 mm. 3. Upperside: both wings dull violaceous blue, with a not very narrow black costa and a broadish black posterior margin to the primaries, and fringes tipped with snowy white for the anal half. Secondaries with broad brown costa, and apex increasing in width at the anal angle ; abdominal fold grey; extreme margin in anal area bright blue ; 54 MR. G. T. BETHUNE-BAKER: A REVISION OF THE tail black, tipped with white, fringes tipped with snowy white; the submedian and lower median nervures are slightly elongated, so as to form two more very short tails, the upper one being the shorter. Underside: both wings brown, with spots marked out by broad whitish margins, but in the primaries the spots are not encircled, though in the secondaries most are completely encircled. Primaries with three discal spots, the large irregular one closing the cell quite open above and below; submedian area whitish ; below the middle discal spot is another spot in the submedian space; trans- verse fascia marked out by three confluent spots, which are quite open above and below—these have an outward tendency ; below are two more distinct confluent spots, shifted so much inwards that their outer margins touch the inner margins of the previous three—these two have a very broad internal white margin ; immediately below them is a double crescentic dark spot; submarginal line distinct, between which and the upper three spots of the transverse fascia is a dusting of white scales. Secondaries with a half-spot at the upper basal extremity of the costa; four basal spots below each other, the second and fourth shifted inwards, the latter being outside the internal nervure; beyond these are three large spots, the upper one of which is situated in a large whitish patch; cell closed by a large irregular spot, below which is a smaller one shifted a little inwards. The transverse chain-like fascia is composed of two very large upper spots, open above and below; the third, shifted right out on to the outer margin of the second, is very small indeed and entirely encircled with whitish ; the fourth, equally small and encircled, is again shifted on to the outer margin of the previous one; fifth smaller, still shifted inwards, below which is the sixth; the angular seventh spot much larger and shifted slightly inwards, and confluent with the eighth, which runs upwards to the internal nervure; following the course of this fascia is a dark submarginal line with an interior pale edging, and the space between the two is more or less suffused with whitish. Lobe-spot black, between which and the second median nervule the marginal area is covered with bright blue metallic scales, a very few being also sprinkled above and below the lobe-spot. ?. Upperside: both wings blackish brown, with a small brightish blue patch occupying the basal, submedian, and discoidal areas. Underside precisely as in male. This is a common species in South Celebes, and is very distinct from all others, its nearest ally being bazaloides Hew. This insect is subject to great variation, the form found in Celebes and in Halma- heira, named by Rober dicolora, being of a brighter blue and less violaceous ; it has very narrow black borders, the underside of which has very much less white and the spots are only margined very finely with white. I have, however, from 8. Celebes specimens exactly like the type-form also. Herr Réber had evidently overlooked Hewitson’s inadequate description and also his figures; the two insects are identical. The genitalia are compressed, with the clasps larger and more rounded ; the penis, with a trumpet-shaped orifice, is small. AMBLYPODIA GROUP OF THE LYCANIDA. 55 ARHOPALA BAZALOIDES (Hewitson). (Plate I. fig. 17; Plate IV. figs. 15 & 15a.) Amblypodia bazaloides Hewitzon, Ill. Diurn. Lep., Lycenidz, Suppl. p. 21, pl. vii. figs. 62 & 63. Arhopala bazalus de Nicéville, Butt. Ind. vol. iii. p. 249 (1890). Hab. Palawan; Mindanao; Zebu; Assam ; Travancore and Canara. Expanse, ¢ 37, 2 40 mm. 3. Upperside : both wings deep purplish, with broad costal and posterior brown margins. In the secondaries these borders are decidedly broader than in the primaries; abdominal fold brown. ‘Tail brown, white-tipped, the nervures on each side of which are elongated and form two very short tails. Underside: both wings brown, with the spots margined with whitish. Primaries with three discal spots increasing in size, that closing the cell being open above and below; beneath it is a short white dash; the transverse band is just discernible by the interior and exterior interrupted white margins; the upper three spots shifted outwards, the next two inwards, so that their inner edge touches the white dash just mentioned; below this is a small dark submarginal row, fairly distinct ; below the second discal spot there is a trace of another in the pale submedian area; apex with a few scattered whitish scales. Secondaries suffused more or less almost all over with whitish scales, with half a spot at the extreme basal edge of the costa, the usual four basal followed by the three larger spots, the upper ones of these two series being confluent ; cell closed by a large spot, beneath which is a small triangular one. ‘Transverse band indefinite, consisting of four pairs of spots: the upper pair being dark and large; second pair small, indistinct, and projected outwards; third small pair shifted inwards, almost confluent with the seventh angular spot, which is also confluent with the eighth long spot, which latter is broken up into two, that touching the angular spot being small and round. ‘There is a distinct submarginal dark waved line ; anal spot black ; a black spot also between the sub- median and lower median nervure, and between the latter and the second median nervule, —these two being covered with bluish-metallic scales, which also edge the anal spotabove. @. Upperside: both wings brighter purplish than in the male, with very broad, dark brown apical and posterior margins, and broad costa in primaries. In the secondaries the purplish is confined to nearly the area of the cell. I have before me a male and two females of this Philippine form. ‘There are decided differences between them and dicolora; it is only about three-quarters its size, the white margins of the spots below in the primaries are reduced so much as to become in some cases almost olsolete—the colour is quite different above and below; whilst in the secondaries the colour is also different, and the chain-band spots decidedly larger, though the insect is smaller. In de Nicéyille’s collection are quite typical specimens from the Khasia Hills (Assam), Travancore, Canara; but he mistook Hewitson’s species bazalus (his teesta) for this insect. The genitalia are similar to those of tephlis, but the clasps differ, as 56 MR. G. T. BETHUNE-BAKER: A REVISION OF THE will be seen on reference to the figures (Plate IV. figs. 15 & 15a), and the hooks have a decided upward curve at their extremities. ARHOPALA THEBA (Hewitson). (Plate IV. figs. 17 & 17a.) Amblypodia theba Hewitson, Ill. Diurn. Lep., Lycenide, p. 6, pl. v. figs. 29 & 30 (1865). Hab. Mindoro; Mindanao. Expanse, ¢ 41, 9 42 mm. ¢. Upperside: both wings brilliant pale morpho-blue, primaries shading into rich deep purple on the costa from the end of the discoidal cell to the apex and down the posterior margin to about the lower median nervule; costa narrowly black, posterior margin rather broadly black. Secondaries with a broadish dark costa; abdominal fold grey ; posterior margin very finely black, and dotted with black between each nervure ; tail black, tipped with white. Underside: both wings very dark brown, with the spots no darker than the ground, but marked out by broad white margins. Primaries with two discal spots not margined on the costal edge; the third spot, closing the cell, very large and irregular, is open above and below, and extends from the second subcostal on to the submedian nervule from the lower edge of the cell; the inner white margin is projected suddenly outwards so as to almost touch the outer margin on the lower median nervule ; the transverse band is irregularly marked out by the white margins, which rapidly taper from just below the lower discoidal nervure until they meet on the lower median; margin spotted with white between each nervule, with brown central dashes; submedian space whitish; the usual spot in the subdiscal area is present just below the second discal spot; there is a short white dash on the costal nervure up to the first discal spot. Secondaries with a half-spot slightly extended outwards on the extreme basal margin of the costa; four basal spots below each other, the lowest shifted inwards beyond the internal nervure, almost touching the margin itself ; beyond these are three larger spots, the upper one not margined on the costal edge ; cell closed by a large subquadrate spot, open above and below, beneath which the usual spot can be traced by its internal and external margins; the transverse band marked out only by the lateral margins of the usual spots, but the margins are very irregular, thus the inner margin of what should be the second spot is projected well inwards, the margins of the third slightly outwards, and the fourth more outwards, the fifth well inwards, the sixth outwards, the angular seventh inwards and it is quite confluent with the eighth long spot, running up to the internal nervure ; outside this fascia there is a suffusion of white scales ; marginal row distinct the internal margin of which is very finely white; extreme margin white; anal spot black; spot between the second and lower median nervule black, the space between being filled with ereenish-metallic scales, which are also scattered more or less over the spots. @. Upperside: both wings brown, with a large pale blue patch over the basal and AMBLYPODIA GROUP OF THE LYCANIDA. 57 central areas, extending right into the inner margin in the primaries; this patch in the secondaries is larger in extent, extending further towards the posterior margin. Underside precisely as in male. This very lovely species forms a natural connecting-link, by its underside pattern, between bazaloides Hew. and the most beautiful (though not the largest in size) subgroup of the genus. From bazaloides Hew. it is immediately separable by its gorgeous colouring above, whilst below the pattern, though similar in character, is equally distinct in the very marked contrast of almost pure white and very dark brown. I have specimens from Mindanao as well as from Mindoro, so we may expect it to occur in other islands of the Philippine group. The genitalia are very distinct ; the tegumen is not a hood, but is open at the top, and the laterals are small, the side supports being ample; the hooks are double-jointed, with the terminal arm sharply bent at the elbow; the clasps are curved and rather small, whilst the penis is long, slender, horn-shaped, with a trumpet-shaped orifice. ARHOPALA ARGENTEA Staudinger. Arhopala argentea Staudinger, Exot. Schmett. p. 281, pl. xevi. (1888). Arhopala clarissa Grose-Smith, Novit. Zool. vol. iv. p. 366 (1897) ; id. Rhop. Exot., Lyczenide, pl. xiv. figs. 4 & 5 (1898). Hab. Minahassa and South Celebes. Expanse, ¢ 45, 9 46 mm. I have, through my friend Dr. Staudinger’s kindness, his two types of this most unusual and lovely insect. The best description of the colour of the male is to call it mother-of-pearl, for it quite shows the greenish and mauve translucent tints of that material. It is at once separable from all other species by this colour. Mr. Grose- Smith has since described it from South Celebes, so that it occurs both in the north and south of that island. ARHOPALA SANGIRA Bethune-Baker. (Plate I. fig. 20, ¢ .) Arhopala sangira Bethune-Baker, Ent. Mo. Mag. vol. xxxii. p. 202 (1897). Hab. Sangir. Expanse, ¢, 44 mm. This unique and beautiful little insect has only been recorded from the one island. It should be recognised at a glance whenever found. ARHOPALA ARONYA (Hewitson). Amblypodia aronya Hewitson, Ill. Diurn. Lep., Lycenide, p. 14: e, pl. iii 6. figs. 45 & 46 (1865). Hab. Mindanao. VoL. XviI.—Part 1. No. 8.—August, 1903. I 58 MR. G. T. BETHUNE-BAKER: A REVISION OF THE Expanse, ¢ & 2, 46-48 mm. Hewitson’s figures are very fair. I have a male from the same locality as his type, in which the colour is rather deep purplish blue, much deeper than usual. ARHOPALA ANTHORE (Hewitson). (Plate IV. figs. 18 & 18 a.) Amblypodia anthore Hewitson, B. M. Cat. Lye. pl. iii. figs. 21 & 22 (1862); id. Il. Diurn. Lep., Lye. p. 6 (1865). Arhopala anthore Hewitson, Staudinger, Exot. Schmett. p. 280, pl. xevi. (1888). Hab. Batchian, Halmaheira, and Ternate. Expanse, ¢ 45-48, 2 about 44 mm. From the preceding species anthore may be recognised by its larger size, brilliant blue, and by its underside being spotted rather than lined, as in Felder’s species. From helius Cram. it may be known in that its underside is uniform dark brown, the spots being merely marked out by the white edging, and no darker than the ground, whilst in Cramer’s insect the spots are very decidedly darker than the ground-colour. The colour above is almost the same, but anthore Hew. is usually more suffused with purple in the primaries than is heliws Cram. In distribution the species is very local though apparently plentiful where it occurs. I can trace it only from Batchian, Halmaheira, and Ternate. The genitalia have the tegumen hooded, extending right down to the extremities of the side supports; the hooks are straightish, the clasps suboval, and the penis small and slender. ARHOPALA HELIUS (Cramer). (Plate I. fig. 21 g, & var. ANTHELIUS Stgr., fig. 22 3.) Papilio helius Cramer, Pap Exot. pl. cci. figs. F, G (1782). Polyommatus helius Lucas, Lep. Exot. pl. xliv. fig. 4. Arhopala phryzus Boisd. Voy. de Astrolabe, Ent. p. 75. Amblypodia phryxus Westwood, Doubl. & Hewitson, Gen. Diurn. Lep. p. 478. Amblypodia helius Westwood, Doubl. & Hewitson, Gen. Diurn. Lep. p. 478. Amblypodia helius var. anthelius Staudinger, Exot. Schmett. p. 281 (1888). Amblypodia tamyrus Felder, Reise Novara, Lep. vol. ii. p. 230, pl. xxviii. fig. 30 (1865). Amblypodia helius Ribbe, Iris, vol. ii. p. 256 (1889). Amblypodia sophax Mathew, Trans. Ent. Soc. Lond. 1887, p. 47. Arhopala heltianthes Grose-Smith, Rhop. Exot. vol. iii, p. 11, pl. xxv. figs. 5, 6, 7 (1902). Hab. Amboina; Aru; Tokal Key; New Guinea; Ceram; Waigeoe; Solomon Islands ; Fergusson Island. Expanse, ¢ , 41-44 mm. 3. Upperside: both wings brilliant morpho-blue. Primaries shading into deep rich slightly purplish blue over the cell and in the apical area, and slightly on the upper half of the posterior margin; costa dark brown from costal nervure to margin, beyond extremity of nervure very narrowly brown; posterior margin less narrowly AMBLYPODIA GROUP OF THE LYCANIDA. 59 blackish. Secondaries with costa broadly grey; narrow brown posterior margin ; abdominal fold pale grey. Underside: both wings pale brown, with a purplish shade in many parts; spots dark and velvety, broadly margined with whitish. Primaries with a pale submedian area ; three discal spots increasing in size; above the outer edge of the second are two very small spots over each other; two or three are over the third spot; below this is another, and a trace of one in the pale submedian area below the second spot; transverse fascia consists of six confluent spots, the second shifted inwards, the third well outwards, the lower three below each other and shifted slightly inwards ; there is sometimes just a trace of a small detached spot below the submedian nervure ; submarginal row distinct. Secondaries with a half-spot at the extreme basal edge of the costa ; four basal small spots, second and fourth shifted inwards, the fourth beyond the internal nervure, the first spot being a double one; beyond these are three large spots, the upper one being shifted right outwards, its inner margin touching the outer margin of that below it, and the lowest spot runs outwards into the seventh spot of the transverse fascia; cell closed by a large suboval spot; transverse fascia consists of nine spots, the second slightly shifted outwards, the confluent third and fourth right out, their inner margin touching the outer margin of the second, the fifth and sixth each slightly inwards, the usual angular seventh spot is divided into two, the upper of the two running into the lowest of the three across the central discal area, and the lower of the two—or the eighth, as I call it—confluent with the long spot which extends up almost to the lowest basal spot; submarginal band distinct, the white margins of which are rather suffused ; lobe-spot black ; spot between the second and lower median nervures black, the marginal space between which is broadly coloured with greenish-blue metallic scales which extend considerably over each spot; extreme margin whitish ; tail stout, rather short, brown, tipped with white. 2. Upperside: both wings pale silvery azure-blue. Primaries with costa narrowly, apex very broadly, and posterior margin broad at first, then tapering to quite narrow, black. Secondaries with costa broadly dark grey, and posterior margin narrowly black. Underside as in male, but paler. This species, which is one of the two on which Boisduval founded his genus Arhopala, may be known from anthore Hew. by the pale submarginal suffusion and by the fact that the markings are decidedly darker than the ground and are much more spotted in appearance. Var. anthelius Staudinger is a very pretty smaller form described from Aru and Waigeoe, but the Aru specimens are without doubt ordinary helius; the Waigeoe specimens, however, are readily separable by the smaller size, especially noticeable in the females, and by the dark purple suffusion of the undersides, which is particularly prominent in the secondaries. Mr. Grose-Smith’s helianthes, of which I have specimens from Milne Bay, New Guinea, is referable to Staudinger’s variety; it corresponds precisely with it. 12 60 MR. G. T. BETHUNE-BAKER: A REVISION OF THE ARHOPALA MINNETTA (Butler). Amblypodia minnetta Butler, Ann. & Mag. Nat. Hist. ser. 5, vol. x. p. 152 (1882). Hab. Duke of York Island; New Britain ; New Guinea. Expanse, ¢, 41 mm. This species is probably a local race of Arhopala helius, but is separable easily by the quite pale under surface and by the spots being somewhat smaller. ARHOPALA ALBOPUNCTATA (Hewitson). Amblypodia albopunctata Hewitson, Ill. Diurn. Lep., Lye. p. 14c. n. 89, pl. iii d. figs. 43 & 44, g (1869). Narathura albopunctata Moore, Journ. Linn. Soc. London, Zoology, vol. xxi. p. 44 (1886). Arhopala albopunctata de Nicéville, Butt. Ind. vol. ii. p. 261. n. 821, frontispiece fig. 126 (1900). Hab. Burma. Expanse, ¢ 37-44, 2 33-42 mm. The figure of the female given on the frontispiece by de Nicéville, in his ‘ Butterflies of India,’ vol. iii., is not like Hewitson’s female specimens ; the underside markings are almost entirely disintegrated and practically resolve themselves into white spots and dots, though it is just possible to trace the sequence of the pattern, and there is scarcely any trace of the oval black spots at the anal angle of the upperside mentioned by him. This species can be immediately separated from dycwnaria Feld. by the metallic morpho- blue of its upper surface, which in the latter insects is somewhat shining pale violet, and the strong contrast of the dark ground and white spots and dots of the underside is very different to lycenaria. ARHOPALA LYCHNARIA (Felder). Amblypodia lycenaria Felder, Wien. ent. Monatsch. vol. iv. p. 396. n. 8 (1860) ; Hewitson, Cat. Lyc. B.M. p. 12. n. 54 (1862) ; Druce, Proc. Zool. Soc. Lond. 18738, p. 354. n. 12. Arhopala lycenaria Felder, Reise Novara, Lep. vol. ii. p. 232. n. 258, pl. xxix. fig. 138, ¢ (1865). Narathura lycenaria Distant, Rhop. Malay. p. 269. n. 15, woodcut n. 79, g (1885). Amblypodia olinda Druce, Proc. Zool. Soc. Lond. 1873, p. 354, pl. xxxiil. fig. 5, 2. Amblypodia buxtoni Hewitson, Il. Diurn. Lep., Suppl. p. 22. n. 104, pl. viii. suppl. figs. 68 & 69, ? (1878). Narathura buxtoni Distant, Rhop. Malay. p. 464. n. 20, pl. xliv. fig. 18, 9 (1886). Arhopala lycenaria de Nicéville, Butt. India, vol. iii. p. 262. Arhopala buxtoni, de Nicéville, loc. cit. Arhopala olinda Druce, P. Z. S. 1895, p. 591. Hab. Malacca (interior); Penang ; Singapore; Borneo; Sumatra. Expanse, ¢ 34-38, 2 36-39 mm. A, lycenaria is a species that has been rare hitherto; I do not think, however, that olinda Druce can be separated from it. Felder’s original type is small and weak-looking ; but I have other specimens before me, undoubted lycenaria, that are AMBLYPODIA GROUP OF THE LYCANID#. 61 fully as large as ordinary olinda and likewise almost identical in colour, whilst the general pattern of the underside is quite inseparable. They appear to obtain in exactly the same localities, and I think they must be synonymous. ARHOPALA ALITAUS (Hewitson). (Plate IV. figs. 19 & 19a.) Amblypodia aliteus Hewitson, Cat. Lyc. B.M. p. 7. n. 29, pl. v. figs. 45 & 46, 3 (1862) ; id. Ml. Diurn. Lep., Lye. p. 8 (1865). Amblypodia viviana Rober, Iris, vol. i. p. 200, Taf. ix. figs. 13 & 14 (1887). Hab. Makassar ; South Celebes; Banka; Mindoro. Expanse, ¢ 35, 9 32 mm. Rober has described specimens from Banka as viviana which, without doubt, belong to this species; both male and female are exactly like Hewitson’s insect in every particular, except that the posterior border of the latter is slightly wider. The form found in Celebes is much brighter blue, though dark, and not violaceous, whilst in the female the blue is brighter still and confined to a small spot just below and at the end of the cell of the primaries, and the underside is a little more strongly marked, but exactly the same pattern. Dr. Staudinger has sent me a specimen from Mindoro, and there is also one in Semper’s Philippine collection. It is at once recognisable from its allies and is fairly well figured by Hewitson. The genitalia are less robust, the hooks are spatulate, the clasps subtriangular, with a moderately robust penis. ARHOPALA MIRABELLA Doherty. Arhopala mirabella Doherty, J. A. S. B. 1889, vol. lvii. pt. 2, p. 420. n. 13. Hab. Mergui; Myitta. I have not seen this species; from the description it appears to be an exceedingly close ally of alitwus Hewitson, so near that, without seeing the specimens, it is scarcely possible to differentiate them ; but with species that are so close, as in this genus they are, an opinion can scarcely be formed in the absence of a good figure or the specimen itself. ARHOPALA MYRTHA (Staudinger). (Plate I. fig. 23, ¢.) Amblypodia myrtha Staudinger, Iris, vol. ii. p. 127 (1889). Hab. Palawan. Expanse, ¢, 39 mm. 3. Upperside: both wings deep ultramarine-blue, quite lustrous as it catches the light, but in some lights very deeply subdued. Primaries with a very broad blackish posterior margin; fore part of costa narrowly brown up to the costal nervure. Secondaries with broad blackish margins all round; abdominal fold greyish; tail black, tipped with white. Underside: both wings warm cinnamon-brown, spots and fascia scarcely darker than ground, palely encircled. Primaries: three discal spots 62 MR. G. T. BETHUNE-BAKER: A REVISION OF THE imcreasing in size, below which are two spots separated by the lower median nervule, and over which are two indistinct small spots, with a third spot on the costa, that over the second spot beneath the costal nervure, between these, but further from the costa, is another indistinct dot; transverse chain-band composed of seven spots, first four detached from the three lower ones, the row of four is inclined outwards and the second and fourth spots are shifted outwards, the fifth and sixth are confluent, shifted well inwards, but with an outward inclination, the seventh spot smaller and indistinct; submarginal row fairly distinct; extreme posterior margin finely dark with a fine paler interior edge; submedian space paler. Secondaries with four basal darker spots, the third shifted outwards, followed by three larger ones below each other; cell closed by a large spot margined laterally, below which is a small triangular spot; transverse fascia with the first and second spot confluent, the second the larger, with its inner edge shifted well inwards and touching the spot closing the cell near the centre of its upper edge; third spot shifted well out- wards, fourth spot smaller, shifted outwards again, fifth inwards, its inner edge often almost touching the spot closing the cell, sixth slightly outwards, the seventh angular spot confluent or touching the eighth and shifted right inwards, its inner angle being very acute and its inner upper margin almost touching the small triangular spot; sub- marginal row distinct ; lobe-spot black; a black spot between the submedian and lower median and also between the latter and the second median nervule, which are almost obliterated by metallic-blue scales, which also edge the lobe-spot above and below ; extreme margin finely brown with a fine interior edging of white. This species may at once be recognised by the peculiar lustrous shading of the upper surface and by its unusually broad black outer margins, and also by the very irregular transverse band of the primaries beneath. I have before me three male specimens, all from Palawan. ARHOPALA MINDANENSIS, n. sp. (Plate I. fig. 24, ¢ .) Hab. Mindanao; Mergui; Singapore. Expanse, ¢ 32-36, 9 45 mm. ¢. Upperside: both wings deep sublustrous purplish violet, shading into lustrous brown in certain lights. Primaries: costa narrowly black, posterior margin with a broad black border. Secondaries with a very broad blackish costa and a broad posterior margin; abdominal fold pale brownish. Tail black, tipped with white. Underside: both wings greyish brown, with darker spots encircled with white. Primaries with three discal spots, the second very large, quite as large as the third ; below each of the latter two is another spot separated by the lower median nervule, that below the third having its white margin confluent with the inner margin of the third spot ; between these two discal spots and the costa are three or four very small indistinct spots; transverse fascia composed of seven spots, the first on the costa small, the AMBLYPODIA GROUP OF THE LYCANIDZ. 63 second larger, shifted outwards, third slightly inwards, fourth well outwards, these all having an outward inclination, fifth and sixth confluent and shifted inwards, seventh below the lower median nervule inwards again and irregular in shape ; submarginal row almost evanescent at apex but more distinct towards the anal angle. Submedian area whitish. Secondaries with four basal spots larger than usual. Second and fourth shifted inwards, third outwards, the upper three almost touching each other, followed by three larger ones, the lower two touching ; cell closed by a large irregular spot, below which is a small triangular one; transverse fascia with the second spot larger than the first, its inner margin being shifted inwards (though I have one specimen with the outer margin shifted outwards, but in this the spots are all compressed together, being a small specimen), third spot shifted right outwards, fourth outwards again, fifth inwards, sixth outwards and tapering in shape, seventh angular spot right inwards, detached, and confluent with the eighth long spot ; submarginal row distinct; extreme margin finely dark, edged by an interior fine pale line; anal spot blackish, and a black spot between the second and lower median nervures, the space between which and also over each being filled with metallic greenish-blue scales. ?. Upperside: both wings purplish brown. Primaries with a small purple patch occupying the discal and submedian area, but not below the submedian nervure. Secondaries with a purple patch very obscure and confined to the cell. I have one specimen before me without any purple on the secondaries. Underside precisely as in the male. This species may be known from myrtha Staudinger by its deep purplish-violet colour and by the border not being half as wide. On the underside also it differs by its pale colour and by the distinct white edging to the spots, and also by a slight whitish suffusion in the secondaries; the transverse band is, moreover, less fractured and the spots larger and slightly different in position. The tone of blue varies somewhat, one specimen being decidedly bluer than the type. A female from Mergui in de Nicéville’s collection is quite bright azure, and its underside pattern much more strongly marked. ARHOPALA MYRTALE (Staudinger). (Plate I. fig. 25, ¢.) Amblypodia myrtale Stgr. Iris, vol. 11. p. 126, pl. i. fig. 17 (non 16) (1889). Arhopala myrtale Druce, P. Z. 8. Lond. 1895, p. 592. Hab. Palawan; Borneo (Sandakan and Labuan). Expanse, ¢, 42-46 mm. The female of this species is not at present known. In his description Dr. Staudinger refers to some specimens doubtfully as the females, but he is so doubtful that he suggests the name myrtha should be given them should they prove distinct. There is no doubt that they are distinct, and, moreover, they are not females, but males having broad borders. This and the paler blue will at once separate them from myrtale; the 64 MR. G. T. BETHUNE-BAKER: A REVISION OF THE latter is also a larger insect with a darker underside and a very much less broken-up transverse band. ‘The same characteristics on the underside will separate it also from mindanensis B.-B., but the blue of my species is very decidedly darker and purpler and with broader borders than in myrtale. These three species are somewhat closely allied, but are, I believe, quite distinct and are easily separable. The legend to the plate in the ‘ Ivis’ (/. ¢.) is incorrect: fig. 17 represents myrtale and fig. 16 agesilaus. ARHOPALA AIDA de Nicéville. Arhopala aida de Nicéville, Journ. Bomb. Nat. Hist. Soe. vol. iv. p. 168. n. 7, pl. A. fig. 1, 3; id. Butt. India, vol. iii. p. 247. n. 802 (1890); Doherty, Journ. A.S. B. vol. lvin. pt 2, p. 422. n. 21; Druce, P. Z. 8. Lond. 1896, p. 666. Hab. Burma; Labuan (Druce). Expanse, ¢ , 44 mm. The female of this species is at present unknown. De Nicéville describes (J. ¢.) a female from Mergui, but is evidently doubtful as to its identity and notes certain differences. I have carefully examined his specimens and believe them to be a very strongly marked form of mindanensis, of which I have five specimens from Mindanao ; the female, however, is paler and brighter than my females, and the male is exactly the same above. Aida differs from myrtale Stgr. in its darker and more violet-blue and its broad outer margins; the undersides are, however, very much alike. ARHOPALA DOHERTYI, n. sp. (Plate I. fig. 8, ¢; Plate IV. figs. 20 & 20a.) Hab. Celebes. Expanse, ¢ & 2, 38-42 mm. g. Upperside: primaries very dark deep blue, slightly violaceous, somewhat irid- escent, with very dark and broad purplish-brown posterior margins, the blue has a very uniform and defined posterior margin following the course of the wing-margin; costa narrowly dark brown. Secondaries dark brown, with a small deep blue patch occupying the central area of the wing ; abdominal fold greyish brown. Underside: both wings brown, with darker spots lightly encircled. Primaries with three increasing discoidal spots ; beneath the third is rather a large triangular spot; another small spot just beneath the inner end of the lower median nervule. Transverse chain-like band consisting of seven spots, of which the upper four have the inner edges uniformly curved outwards, but the outer edges of the two lower spots are projected outwards ; the fifth and sixth spots are shifted well inwards and are followed by a smaller spot . below the lower median nervule, usually, but not always, shifted slightly outwards ; submarginal row distinct ; extreme margin pale ; submedian area greyish. Secondaries with a small spot at the extreme basal edge of the costa, followed by the four basal spots below each other, the lowest shifted inwards beyond the internal nervure, three AMBLYPODIA GROUP OF THE LYCANIDA. 65 larger spots across the central discal area; cell closed by a subquadrate spot, below which is a smaller one connecting it with the third central discal spot; chain-like band composed of eight spots, the two distinct upper spots outwardly inclined, third and fourth spots confluent, shifted well outwards, fifth spot small, shifted inwards, sixth spot outwards, the angular seventh confluent with the long eighth spot which touches the internal nervure; submarginal row distinct, extending very narrowly over the metallic spots up to the internal nervure, the light edging of which is more pronounced in the anal area; the black spots at the anal angle are almost entirely covered with metallic bluish-green scales up to the second median nervule; tail brown, tipped with white ; submedian nervure very slightly prolonged, so as to form a very short tooth-like anal projection ; extreme margin pale. 9. Upperside: primaries rich purplish black, with a patch of purplish blue occupying the cell, the submedian area, and about half of the posterior area below the upper median nervule. Secondaries dark brown, with a small patch of purplish blue occupying about the discoidal area; tail darker, tipped with whitish ; fringes grey. Underside as in male. I have described the blue of the male as iridescent ; this is not absolutely correct, but is the best term I can use, the one wing in the light looking very much brighter than the other, and in one particular light the wing looks lustrous brownish. It is evidently a common insect in Celebes, judging from the number Mr. Doherty sent over here. This is a very distinct species and should be recognised at a glance on account of the peculiar blue of the male and its exceedingly broad outer margins, while on the underside of the primaries the upper four spots of the transverse band are peculiar, the first spot quite small, second rather larger, third and fourth very much larger; the whole of the underside is also very distinctly marked. The genitalia have the usual hooded tegumen and straightish hooks, the clasps are ham-shaped with squarish extremities. The penis is rather small for the size of the insect and is straight. ARHOPALA LABUANA Bethune-Baker. Arhopala labuana Bethune-Baker, Proc. Zool. Soc. Lond. 1896, p. 667, pl. xxx. figs. 12 9,13 ?. Hab. Labuan; Mindanao. Expanse, ¢ 49, 2? 44 mm. From dohertyi this species may be recognised by the brighter blue, very much narrower borders, much shorter tail, which is little more than a tooth-like projection, whilst below in the primaries the transverse band is not dislocated, but in the secondaries it is strongly dislocated. It is much nearer arzenius Felder, but is very much larger than that species; the distinctions are given in the description of that insect. VOL. XVII.—Part 1. No. 9,—August, 1903. K 66 MR. G. T. BETHUNE-BAKER: A REVISION OF THE ARHOPALA ARZENIUS Felder. Arhopala arzenius Felder, Reise Novara, Lep. vol. ii. p. 236, tab. xxix. fig. 18 (1865). Hab. Luzon; Batchian. Expanse, ¢ 40, 2 41 mm. 3. Upperside: both wings deep purplish blue, bright in some lights but exceedingly dull in others. Primaries with costa narrowly, outer margin widely, and apex very widely black; secondaries with costa broadly, and outer margin less broadly, black. Tail short, not much more than a tooth. Underside: both wings ochreous brown, with dark brown markings with pale outlines. Primaries with three small but increasing cell-spots, below each of which is another ; transverse band of nearly uniform width, rather wide, nearly straight, but with a very slight outward inclination, composed of six spots, the first of which is so small and indistinct as to be nearly obsolete, the sixth spot ends on the lower median nervule ; submarginal row distinct, composed of six spots terminating on the submedian nervure; submedian area very pale. Secondaries with four basal spots inwardly oblique, followed by four larger spots also somewhat inwardly oblique; cell closed by a subquadrate spot, below which is a small one connecting it with the third spot in the second series of four; transverse band composed of seven spots, the second of which touches the quadrate spot—hind spot shifted right outwards on to the outer edge of the second, fourth slightly outwards, fifth smaller oval spot inwards, sixth outwards, seventh angular spot well inwards; submarginal row distinct; a few metallic-blue scales near the anal angle. @. Upperside: both wings paler violet-blue; primaries with the costa broadly, outer margins more broadly, apex yet more broadly dark fuscous; cell closed by a blackish dash; secondaries with the costa very broadly, and the outer margins less broadly, fuscous. Underside precisely as in the male, but the metallic-blue scales near the anal angle, very few in the male, are absent in the female. From labuana this species may be recognised by its much smaller size and by the broad borders of the primaries, while on the underside the transverse band of the primaries is of uniform width, absolutely unbroken and nearly straight, and there is scarcely a trace of the metallic-blue scaling at the anal angle of the secondaries. ARHOPALA VIHARA (Felder). (Plate II. fig. 1, ¢.) Amblypodia vihara Felder, Wien. ent. Monatsch. vol. iv. p. 395. n. 5 (1860); Kheil, Rhop. der Insel Nias, p. 33. n. 122 (1884). Arhopala vihara Felder, Reise Novara, Lep. vol. ii. p. 228. n. 253, pl. xxix. fig. 7 (1865). Narathura vihara Distant, Rhop. Malay. p. 270. n..16, woodcut n. 80 (1885). Arhopala vihara de Nicéville, Butt. India, vol. iii. p. 241 (1890); Bethune-Baker and Druce, P. Z. 8. Lond. 1896, p. 662. Hab. Malacca; Nias Island, Borneo (Labuan and Tameang-Lajang), Sumatra, and Upper ‘Tenasserim ! Expanse, ¢, 44 mm. AMBLYPODIA GROUP OF THE LYCAHNIDA, 67 3. Upperside: both wings deep blue; primaries with a narrow black costa and less narrow black posterior margin; secondaries with a very broad brown apical area, tapering off towards the base of costa; posterior margin rather narrowly black; abdominal fold greyish brown. Underside: both wings golden brown, with dark spots palely margined. Primaries with three discal spots, beneath each of which is another, two below the median nervure, the third in the angle of the lower median nervule ; transverse band composed of a curved row of four confluent spots, followed by two larger spots shifted inwards and another much smaller one below the lower median nervule; submarginal row somewhat indistinct; submedian area paler. Secondaries with four basal spots, second and fourth shifted inwards, the latter beyond the internal nervure, followed by three larger spots; cell closed by a subquadrate spot, below which is a small triangular one; chain-like band composed of two distinct upper spots, the third and fourth confluent shifted right outwards, fifth inwards, sixth slightly outwards, the angular seventh spot inwards and confluent with the long eighth; submarginal row fairly distinct, the pale edging of which is much more pronounced at the anal angle above the metallic spots; a black velvety spot occupies the anal angle, and a small dark one the margin between the second and lower median nervule, between which the space is filled with metallic-blue scales, which also edge the spots above ; tail brown, tipped with white. I have not seen a female of this species ; it isa near ally of A. adorea de Niceville, but may be recognised by its darker blue, rather wide black borders, whilst below it is decidedly darker in colour, and the transverse band of the primaries is quite different in shape, which also applies to this band in the secondaries; the difference is, however, so difficult to convey in words that I must refer my readers to the figure; there is no costal spot in the primaries in front of the transverse band. In de Nicéville’s collection I find several males from Sumatra and I have two before me from Borneo. ARHOPALA ADOREA de Nicéville. Arhopala adorea de Nicéville, Butt. Ind. vol. iti. p. 238. n. 789, pl. frontispiece fig. 189, 3 (1890). Arrhopala adorea de Nicéville and Martin, Butt. Sumatra, Journ. A. S. B. vol. Ixiv. pt. 2, p. 465. n. 391 (1895). Hab. Sikkim; Khasia Hills; Tenasserim Valley; Burma; Singapore; Sumatra; Malacca. Expanse, ¢ & 2, 50-56 mm. There appear to be two forms of this species, one distinctly violet in shade, the other almost blue, and they have their respective females: beneath they are the same, but the colour above is readily separable; the blue form has rather broader outer margins than he purpler one, and I believe has often been confused with “ vihari Feld.,” with which it K 2 68 MR. G. T. BETHUNE-BAKER: A REVISION OF THE is very closely allied. It may be separated best by the underside pattern of the primaries, the spots being decidedly less than in Felder’s insect, which latter has also a different transverse band beginning with quite a small spot on the costa, and the fifth and sixth spot being much larger and shifted further inwards. The genitalia are similar to those of dohertyi, but larger, with subtriangular clasps, and more slender penis which has a long opening at the orifice. ARHOPALA PHENOPS Felder. (Plate I. fig. 27, 3.) Arhopala phenops Felder, Reise Novara, Lep. vol. 1. p. 227. n. 251 (1865). Hab. Luzon and Philippine Islands. Expanse, ¢ & 2,36-39 mm. This species is easily separable from adorea de Nicéville, for it is very much smaller in size and the transverse band of the underside is not dislocated, and it is altogether more uniform in tone of colour. From sandakani mihi it may be recognised by its much duller and darker upper surface and by the broader margins ; beneath the colour is darker, the pattern is slightly different and is also more uniform than is my species. The figure will show at once the various points of difference. The genitalia are similar to those of dohertyi in pattern, but with smaller clasps rounded at the lower extremity but sharply angled at the upper extremity. ARHOPALA SANDAKANI Bethune-Baker. Arhopala sandakani Bethune-Baker & Druce, Proc. Zool. Soc. Lond. 1896, p. 671, pl. xxxi. fig. 2, 3. Hab. Sandakan; Java; Silam; Pulo Bai. Expanse, ¢ 35-45, 9 42 mm. This species is easily separated from the following by its smaller size, narrower margins, brighter and bluer colour, and by the darker tone of the undersides. Since describing this species, Mr. Cator has taken many in Pulo Bai and one in Silam, Borneo. The geni- talia are very similar to those of phanops, but the hooks are slightly curved upwards. ARHOPALA QUERCOIDES (Rober). (Plate I. fig. 26, ¢.) Amblypodia quercoides Rober, Iris, vol. 1. p. 72, Taf. v. fig. 9 (1886). Hab. Celebes. Expanse, ¢ & 2, 37-42 mm. This is a common insect in Celebes: in the male the violaceous colour has in certain lights a brown shimmer over it; it may be readily recognised from its near allies by its very much greyer under surface and very distinctly marked pattern and the unusually even transverse band. ‘The genitalia have oval extremities to their clasps, but are otherwise similar to those of phwnops. AMBLYPODIA GROUP OF THE LYCHNID. 69 ARHOPALA DRUCEI Bethune-Baker. Arhopala druce: Bethune-Baker & Druce, P. Z. S. Lond. 1896, p. 661, pl. xxx. figs. 1 ¢, 2 2. Hab. Kina Balu, Borneo. Expanse, ¢ 55, 2 50 mm. This isa very marked species and should be recognisable immediately ; I have only seen a male and two females, all from Kina Balu, as mentioned in my original description (J. c.). ARHOPALA APHA de Nicéville. Arhopala apha de Nicéville, Journ. Bombay Nat. Hist. Soe. vol. ix. p. 287. n. 21, pl. O. fig. 31, 3 (1895). Hab. Tenasserim; Burma; Lombok. Expanse, ¢ 48-58, 9 48 mm. This species is a very ciose ally of adatha Hew., but its under surface is of a pinker and richer brown, whilst the spots are very much darker than in Hewitson’s species. The upperside of apha is of a very much brighter and more shining blue and its outer borders are rather broader: there should be no difficulty in separating these species. I have recently received a pair of apha from Herr Fruhstorfer, taken in the Isle of Lombok. ARHOPALA ADATHA (Hewitson). Amblypodia adatha Hewitson, Cat. Lyc. B.M. pl. iv. figs. 29, 30, & 31, g (1862); id. Ill. Diurn. Lep., Lycenide, p. 8. n. 28 (1865) ; Butler, Trans. Linn. Soc. Lond., ser. 2, Zoology, vol. i. p. 548. n. 4 (1877). Narathura adatha Distant, Rhop. Malay. p. 265 (figures incorrect) (1885). Arhopala adatha de Nicéville, Butt. Ind. vol. iii. p. 247. Hab. Amboina; Malacca; Singapore; Celebes; Philippine Islands (Mindanao and Mindoro). Expanse, ¢ & 2, 48-54 mm. 3. Upperside: both wings dark, somewhat purplish blue, primaries with costa narrowly and posterior margin less narrowly black, secondaries with costa very broadly and posterior margin rather narrowly blackish ; lobe spot black, edged above with whitish ; tail blackish, white-tipped. Underside: both wings rufous brown, with dark spots palely encircled. Primaries with three increasing discal spots ; above the third is a small spot on the costa and below it (the third) is another in the lower median angle, the subdiscal area is darker ; submedian area quite pale up to the submarginal row; transverse band oblique, composed of six confluent spots, the first on the costa being the smallest, the second, third, and fourth are continuous, fifth very slightly shifted inwards, 70 MR. G. T. BETHUNE-BAKER: A REVISION OF THE sixth ending on the lower median nervule shifted slightly outwards, below this is a detached or occasionally almost detached spot ; submarginal row obscure, but quite perceptible: marginal row almost obsolete. Secondaries with four basal spots, the fourth beyond the internal nervure, followed by three larger spots below each other ; cell closed by a large subquadrate spot, below which in the lower median angle is another very narrow spot; two large spots below each other from the costal nervure to the subquadrate spot, the lower one of which also touches the transverse band, which begins on the upper discoidal nervule and is composed of two curved confluent spots edged laterally, followed below by the third spot shifted inwards, fourth shifted outwards, fifth angular, shifted right inwards and completely confluent with the sixth long spot, so that it appears as an angulated fascia with one side abnormally long; submarginal row obscure; marginal sublunulated row likewise obscure ; lobe-spot black, followed by two other black marginal spots, which are almost entirely obliterated by bronzy-green metallic scales, which also edge the lobe-spot above, the metallic scales being margined above with very pale greyish. ?. Upperside: both wings dark purplish brown, with a very large patch of rather brighter purplish blue occupying the cell, the submedian and three-fifths of the median areas of the primaries, and in the secondaries just over the cell and about two-fifths of the median area; the blue of the primaries is invaded at the end of the cell by a wedge of ground-colour. Underside precisely as in male. There has been considerable confusion with this species, which was figured by Hewitson as adatha in his Cat. Lycenidee B.M. pl. iv. figs. 29-31, but in the text was sunk under micale Boisduval, and in his Ill]. Diurn. Lep., Lyc. p. 8, was raised to specific rank again, but without any remark. His type labelled adatha from Amboina is quite evidently that shown as figure 31, which is a very fairly accurate figure. Had it been left thus the matter was perfectly simple, but unfortunately Mr. Distant refers to it in his work Rhop, Malay. p. 265, tab. xxii. figs. 1 ¢, 2 9, thus:—“ The figures here given are those of a male and female specimen captured in Malacca by Captain Pinwill and now contained in the British Museum; these figures do not altogether correspond to those given by Mr. Hewitson as representative of his species. I have, however, carefully compared them with the type specimens of Hewitson, and find that they are correct and faithful.” In Hewitson’s collection there is one specimen labelled adatha: this Mr. Distant, I doubt not, has called the type specimen; I have done it likewise, more especially as Hewitson’s figure 31 is quite evidently drawn from it; but how Mr. Distant can consider his figures are “correct and faithful” to it, I cannot under- stand—they have no resemblance whatever, either in form, colour, or markings, to Hewitson’s specimen labelled adatha; they evidently depict an insect belonging to quite a different section of this genus. But are figures 29 and 30 the same species as 31%t—de Nicéville considers them probably distinct (vide Butt. India, 7. ¢.). I feel however, that I really cannot separate them. All the Amboina specimens have the AMBLYPODIA GROUP OF THE LYCANIDA. 71 transverse band of the primaries unbroken, and one specimen from Minahassa (Celebes) has this band likewise unbroken, another has it slightly fractured at the fifth and sixth spot, whilst several Philippine specimens have similar slight fractures, but one from Mindoro has the band distinctly fractured and is exactly represented by figure 29 from Singapore ; otherwise both in size and colour they are just the same and also as to the width of the margin. Under these circumstances I do not see any advantage in splitting the forms up, as, until we have very much more material there can be no certainty of being correct in so doing. The identity of adatha Hew. is, however, quite settled, and I feel myself that the other is merely an aberration, perhaps a gradual setting up into a local form of the same species. There is nothing unusual about the genitalia. ARHOPALA MALAYICA, n. sp. (Plate IL. fig. 2, ¢.) Hab. Philippine Islands ; Java. Expanse, ¢ & 2,50-58 mm. g¢. Upperside: both wings deep purplish blue, primaries with the costa rather narrowly brown and outer margin rather wider and blackish, secondaries with the costa broadly and the outer margin not very narrowly blackish. Tail long, black, tipped with white, the nervule on each side of it being elongated so as to make a slight tooth; lobe well developed, with a black spot edged slightly above with metallic blue. Underside: both wings ochreous brown, with dark spots palely edged. Primaries with three small but increasing cell-spots; below the third is a large one in the lower median angle; subdiscal area dark with a very faint paler spot in the middle; transverse band narrow, composed of six confluent spots, with a very slight outwardly oblique inclination, the band being nearly straight, except that the fourth spot is curved very slightly outwards and the fifth spot has its inner edge shifted slightly inwards below the sixth spot; in the pale submedian area is another spot (divided into two in the type male, but not so in the female) shifted inwards ; submarginal row distinct but indefinite. Secondaries with small spot at basal extremity of costa; four quite small basal spots, the third shifted well outwards, followed by three larger spots below each other; cell closed by a largish spot, below which is another in the lower median angle; shifted well out from the spot closing the cell are two spots which usually form the upper two spots of the transverse band between the costal and upper discoidal nervures, the two spots are isolated and con- fluent, the second being the larger ; the transverse band beginning on the upper discoidal nervule and shifted right out is composed of six spots, the first four almost confluent, but with the second having in the type, but not always so, a slight outward shifting and the third a very slight inward shifting, the fifth angular spot confluent with the long sixth spot is shifted right inwards; submarginal row broad and fairly distinct; lobe- 72 MR. G. T. BETHUNE-BAKER: A REVISION OF THE spot large, velvety black, and a deep black spot beyond the tail, well covered, as is also the space between the two spots, with bright pale bluish-metallic scales, which likewise edge the lobe-spot above. 2. Upperside: both wings violet-blue; primaries with very broad black costa and outer margin, and yet broader apical area; secondaries with very broad margins all round. Underside like that of the male, except that the ground-colour is rather paler. One specimen from Mindoro has the costa and outer margin of the primaries decidedly narrower than in the type. This species comes between adatha Hew. and silhetensis Hew. ; on the upperside it has wider borders than the former but rather narrower than the latter. In general pattern of the underside it is more like adatha, but has the spots nearly half the size and very much darker, thus forming a decided contrast with the ground-colours, and the transverse band of the secondaries is quite different, this being caused by the isolation of the upper two spots. The spots are also decidedly smaller than in si/hetensis ; there is also no spot on the costa of the primaries, and the transverse bands in both wings are not dislocated as they are in that species. In the genitalia the clasps have conical extremities. ARHOPALA SILHETENSIS (Hewitson). Amblypodia silhetensis Hewitson, Cat. Lycenide B.M. p. 7. n. 31, pl. iv. figs. 27, 28, ¢ (1862). Arhopala silhetensis de Nicéville, Butt. Ind. vol. 11. p. 237. n. 788 (1890). Arhopala arama de Nicéville, Journ. Bombay Nat. Hist. Soc. vol. ix. p. 285. n. 20 (1895). Hab. Sylhet; Sikkim Dooars, Sikkim; Burma (? Cochin China). Expanse, ¢ & 2, 49-54 mm. ¢. Upperside: both wings deep purplish blue; primaries with the costa narrowly and the posterior margin broader black ; secondaries with the costa very broadly and the posterior margin less broadly, but increasing at the anal angle, black ; abdominal fold very broadly brown; lobe well developed ; tail blackish, tipped with white. Underside : both wings dull brown, with darker spots bordered with blackish with an outer encircling of white. Primaries with three increasing cell-spots; above the third near the costa is a small one, and below it (the third) another; subdiscal area dark, with a paler spot in the middle; submedian area paler; transverse band composed of seven spots, the first four placed obliquely outwards and slightly curved, the second and fourth being the largest, fifth and sixth spots shifted well inwards, but with an outward inclination, the seventh elongated spot is also shifted inwards, submarginal row distinct. Secondaries with a spot at the basal costal extremity, and the usual four basal spots, the fourth shifted inwards, followed by three larger ones below each other ; cell closed by a subquadrate spot, below which is a small one, and above which are two confluent spots, the lower larger than that next the costa, these usually forming the AMBLYPODIA GROUP OF THE LYCZNIDZ, 73 first two spots of the transverse band, which band begins on the upper discoidal nervule, composed of six spots, the second shifted outwards, the third well inwards, fourth outwards, fifth angular spot right inwards, detached and touching the sixth long spot which extends up to the internal nervure; lobe largely developed ; lobe-spot black finely edged above with metallic-blue scales, which extend slightly into the nervule interspace beyond, above which is a broad dash of greyish scales; submarginal row obscure. 2. Paler, and the blue more violet in shade ; in the primaries the costa and posterior margins are very broadly, and the apex yet more broadly, purplish black ; the secondaries have exceedingly broad margins all round. Underside as in the male. In the Hewitson collection is a female, labelled Cochin China, in which the blue is much brighter and paler than usual. A. silhetensis can be separated from malayica mihi by the pattern beneath; the spots are all larger, the transverse band of the primaries is distinctly fractured (whilst in my species it is not), and the transverse band of the secondaries is different, the third, fourth, fifth, and sixth spots being twice fractured, but being in malayica almost confluent ; there is also a distinct black spot in the lower median nervule space in my species, absent in Hewitson’s, and his species has only a trace of a very sparse metallic scaling, whereas mine has an abundance of bright blue metallic scales. From nicévillet it is easily separable because of the violet colour of that speciess, and by the pale greyish underside with the blackish spots standing out in strong contrast to my species. This little group of four species allied to adatha Hew. has been difficult to arrange : adatha Hew., malayica mihi, silhetensis Hew., and nicévillei mihi are all closely allied yet quite distinct in the appearance of the underside; of malayica I have nine specimens, and of nicévillec five, each constant in pattern: apha de Nicéville is close also, but I think there is no doubt that they are distinct from each other. A. arama of de Nicéville falls as a synonym of this species. My friend Mr. de Nicé- ville had never seen this species, and from Hewitson’s description he mistook the insect. I have carefully compared his type of arama with Hewitson’s insect, and find that they agree absolutely. Under the circumstances I have thought it well to describe these fully, more especially as this last species is very rare and Hewitson’s descripfion is very inadequate. ARHOPALA NICEVILLE, n. sp. (Plate II. fig. 5, 3.) Hab. Burma and Sikkim. Expanse, ¢ 56, 9 48-56 mm. g. Upperside: both wings violet-blue, slightly lustrous, in certain lights dull violet ; primaries with the costa and outer margins narrowly dark fuscous: secondaries with the costa and abdominal margins broadly and the outer margin very narrowly dark fuscous; abdominal fold grey; lobe highly developed, with a blackish spot; VoL. xvil.—part 1. No. 10.—August, 19038. ii 74 MR. G. T. BETHUNE-BAKER: A REVISION OF THER tail long, dark fuscous, tipped with white, nervures on each side of this being slightly elongated into a tooth-like projection. Underside: both wings greyish ochreous brown, with very dark brown, almost blackish spots encircled with whitish. Primaries with three increasing cell-spots; above the third a small one near the costa, and below it (the third) a large one in the lower median angle ; a very large subdiscal spot, with a pale centre in it; submedian area pale grey. ‘Transverse band composed of seven spots, highly dislocated between the fourth and fifth; the second spot is large and shifted outwards as to its outer margin only, third spot smaller with its outer margin consequently inwards, fourth spot shifted somewhat outwards altogether, fifth spot shifted right inwards, with its outer margin on the inner margin of the fourth spot, sixth long spot shifted slightly outwards, seventh spot shifted inwards, but with an outward inclination ; submarginal row distinct, but indefinite. Secondaries with a spot at the basal costal extremity ; four basal spots, the lowest shifted inwards, followed by three larger spots below each other; cell closed by a large quadrate spot, below which is another connecting it with the lowest of the three afore-mentioned ; above this cell- spot, touching it, are two large spots above each other, extending up to the costa ; these usually form the first two spots of the transverse band, but in this species they form an inner fascia with the cell-spot, the small one below it, and the third spot of the series of three. ‘Transverse band beginning on the discoidai nervure, composed of four spots and a large crook-shaped mark; the second spot is shifted outwards, third slightly inwards, fourth slightly outwards, the crook-shaped mark, composed of the angular spot and long spot combined in one, shifted right inwards and extending nearly up to the lowest of the basal spots. Submarginal row indefinite ; lobe-spot deep black, with a slight black marginal spot in each of the next two nervure interspaces, over which latter are superimposed greenish-metallic scales, which also edge the lobe-spot above; between these and the transverse band there is a slight scattering of greenish-grey submetallic scales on a darker background. @. Upperside: both wings a little bluer than the male; primaries with broad costa, broader outer margin, and very broad black apical area; cell closed by a black dash: secondaries with very broad black margins all round. Underside: both wings precisely as in the male. This species, of which I have five specimens before me, is nearest st/hetensis Hew., but may be readily distinguished by the paler violet-blue and the very narrow black borders to the primaries; beneath it is much paler and greyer, whilst the blackish spots, all standing out most prominently, make it recognisable at a glance. ARHOPALA NOBILIS (Felder). (Plate II. figs. 4 ¢,5 2.) Amblypodia nobilis Felder, Sitzb. Akad. Wiss. Wien, xl. p. 453 (1860). Arhopala nobilis Felder, Reise Nov., Lep. vol. ii. p. 226. n. 249, tab. xxix. fig. 6 (1865). AMBLYPODIA GROUP OF THE LYCEHNIDA. 10 Amblypodia nobilis Hewitson, Cat. Lye. B.M. p. 7. n. 33 (1862); id. Ill. Diurn. Lep., Lye. p. 8 (1865). Amblypodia alce Ribbe, Iris, vol. ii. p. 256 (1889). Arhopala alcestes Grose-Smith, Rhop. Exot. vol. iii. p. 9, pl. xxv. fig. 1. Hab. Amboina; Ceram. Expanse, ¢ 47, 2 49 mm. ¢. Upperside: both wings shining ultramarine-blue; primaries with a very narrow black costa and narrow posterior margin. Secondaries with broad blackish costa, narrow posterior margin, blackish-grey abdominal fold; anal angle with a large black spot; tail black, white-tipped, a short black tail also at the end of the second median nervule; lobe highly developed, black, edged above with brilliant cerulean. Under- side: both wings warm brown, with dark velvety spots palely edged. Primaries with three large and increasing spots, the second not edged on its costal side, and the third edged laterally only ; anterior to the second and over the third are two small spots on the costa; beneath the second spot in the paler submedian area is a darker shading forming a very indefinite spot: transverse band composed of six spots, a small costal one confluent with the much larger second spot; third and fourth larger still, shifted well outwardly, confiuent, with no outward pale edging, but diffused right into the submarginal row, the upper of these two spots forms an interior obtuse angle with the lower one; fifth spot shifted right inwards, the sixth also slightly inwards and terminating on the lower median nervule; submarginal row distinct. Secondaries with a small spot at the costal base; a large, irregular, basal spot, followed by two small roundish ones, the lower one being shifted beyond the internal nervure ; these are followed by three longish spots below each other; cell closed by a subquadrate spot, below which is a small, very dark spot in the lower median angle. Two large confluent spots are just below the costa, which should form part of the transverse band, but are quite disconnected from it; the upper one touches the middle spot, and the lower one touches the spot closing the cell. Transverse band begins on the upper discoidal nervure with two confluent quadrate spots, followed by the third shifted well inwards, the fourth outwards, fifth angular spot slightly inwards and confluent with the sixth long spot; submarginal row almost obsolete. Lobe-spot large, velvety black, and more or less black in the marginal spaces between the three succeeding nervules, over which is a considerable suffusion of greenish-metallic scales, which are edged above by a black stripe up to the internal nervure ; over this again are more metallic-greenish scales mixed with grey ones. Most of the spots on the secondaries have pale centres. 2. Upperside: both wings bright azure-blue, with broad purplish-black costal ani posterior margins to the primaries, and broad margins all round in the secondaries. Underside as in male. This species is very close to alce Hew., but the blue of the upperside is much brighter and of a different tone, and the underside markings are much less and are not L2 76 MR. G. T. BETHUNE-BAKER: A REVISION OF THE at all confluent as in Hewitson’s insect. The shape of the wings is also different, the apex of the primaries being sharper, the posterior margin straighter, whilst the secondaries are more excavated towards the tails. Alce also has the black margins less narrow than in this species. Herr Ribbe recorded alce from Ceram (Iris, J. ¢.), but I have the specimens before me, and they are nobilis. Alcestes Grose-Smith must sink as a synonym of Felder’s species; the underside pattern is quite similar, the only difference being the whitish patch by the subcostal nervule of the secondaries, and as there is sometimes a lightening of the scales there it can only be a pale aberration. ARHOPALA ALCE (Hewitson). (Plate V. figs. 1 & 1a.) Amblypodia alce Hewitson, Cat. Lyc. B.M. p. 5. n. 21, pl. iii. 9 figs. 18, 19, g 20 (1862) ; id. Ill. Diurn. Lep., Lye. p. 4, fig. 9 (1865). Arhopala athara Grose-Smith, Rhop. Exot. vol. iii. pp. 9, 10, pl. xxv. fig. 2. Hab, Aru; Halmaheira. Expanse, ¢ 50-51, 2 60 mm. This is probably a local form of nobilis Feld., and both appear to be rare species. It may be separated by its larger size, brighter blue colour, and below by the extremely large size of all the spots, which are much confluent, and at the base of the secondaries are strongly so, and they thus form a very irregular, dark, basal blotch. I cannot find any distinction between A. athara Grose-Smith and alce Hew. ; the position of the spots mentioned by Mr. Grose-Smith is by no means constant in Hewitson’s species, of which I have before me a fair series, so that it would not be possible to separate them on this character only. ‘The clasps of the genitalia are peculiar, becoming attenuated at the extremities; the hooks have a downward curve. ARHOPALA ANUNDA Hewitson. (Plate III. figs. 19-21; Plate V. figs. 2 & 2a.) Amblypodia anunda Hewitson, Til. Diurn. Lep., Lye. p. 14a. n. 73, pl. iia. fig. 32. Narathura anthelus Distant, Rhop. Malay. pl. xxiu. fig. 4, 2, and in parte p. 263. Amblypodia anthelus var. saturatior Staudinger, Iris, vol. 1. p. 122. Hab. Borneo; Philippines (Mindoro, Mindanao, Palawan) ; Sumatra and Java. Expanse, ¢ 45-52, 2 46-51 mm. g. Upperside: both wings brilliant shining violet-blue; the primaries with black linear costa and posterior margin; secondaries with broadish brown costa, linear black posterior margin, and very pale grey abdominal fold. ‘Tail black, white-tipped. Underside: both wings pale greyish brown, with dark spots broadly edged with whitish. Primaries with three large increasing spots, touching the second of which are two or three somewhat confluent spots up to the costal nervure; over the third is a largish spot on the costa, and anterior to it is another just touching the transverse band; beneath the second and third are two spots separated by the lower median os f bof AMBLYPODIA GROUP OF THE LYCANIDA. nervule, that in the grey pale submedian area being continued up to the base; transverse catenulated band composed of five spots from the third subcostal to the lower median nervule, the second of which is shifted inwards, the third well outwards, the fourth and fifth confluent well inwards, but with an outward inclination; sub- marginal row obscure ; around the uppermost spot of the transverse band, the interior costal spot touching it, is a considerable greyish-white suffusion. Secondaries with a spot on the costal base, followed by two more touching each other ; below the third are three smaller basal spots, the lowest shifted on to the internal margin, then come three larger spots below each other ; cell closed by a largish irregular spot, below which is a smaller triangular one; transverse band composed of eight spots, the upper two being round and separated from the others by a pale suffusion, the lower of these two is the smaller and shifted outwards; the fourth spot is shifted outwards, the fifth inwards, sixth outwards, seventh angular spot touching the eighth long one and shifted inwards ; submarginal row obscure; a black spot at anal angle and between the second and lower median nervules, the space between and over which is filled with bronzy-greenish metallic scales, edged above by a dark suffusion. In the secondaries the spots near the costa are all very dark rich brown, the others quite pale; there is also more or less of a lilac suffusion in the basal half of both wings. 2. Upperside: both wings violet, with very broad brown costal and posterior margins to primaries and less broad borders to the secondaries. Underside as in male. These descriptions appear to be of the commonest form of this species. Hewitson described his male (from Borneo) as brilliant morpho-blue, but his labelled type is violet-blue. I have specimens similar to this from Palawan, Mindoro, and Mindanao, and in all these the ground-colour beneath is darker, but the markings are the same. I have one female which might well answer to this male, as it is brilliant azure with very broad blackish borders. I have also three females in very good condition which are entirely brown, with no trace of blue whatever above. I am inclined to think that the species is dimorphic, as I have both forms from various islands of the Philippines (both forms from the same island), and I have also the violet-blue form from South-east and North-west Borneo now lying before me. This shows that both the morpho-blue and the violet-blue forms co-exist ; it would therefore be most interesting to investigate the life-history of the species. ‘The underside in both forms is precisely the same, except that the more brilliant of the two is darker on the underside. It is fortunately a well-marked species, so that there is no difficulty in recognising it. ‘The figure given by Distant (Rhop. Malay. /.¢.) as A. anthelus is without doubt anunda, and his text evidently partially refers to this species, he having, I imagine, not differentiated the two species. Dr. Staudinger has also described (wide Iris, /.¢.) var. satwratior as a variety of anthelus ; it is, however, a very bright and pale blue form of this species. The genitalia are somewhat similar to alce Hew., though the clasps differ, and the hooks are straight, whilst the penis is long, bent, and large at the orifice. =1 wo MR. G. T. BETHUNE-BAKER: A REVISION OF THE ARHOPALA ANTHARITA Grose-Smith. Arhopala antharita Grose-Smith, Novit. Zool. vol. i. p. 583 (1894) ; id. Rhop. Exot. part 44, Lye. no. 8, pl. xv. figs. 9 & 10, 2 (1898). Hab. New Guinea? Expanse, 2, 78 mm. I admit this species with doubt. It is described from a single specimen, and has, as Mr. Grose-Smith says, a very strong resemblance to anunda Hew., which species is decidedly variable, even in its females, some of which are entirely brown on the upper- side, as this species is. There being only one specimen extant, it is impossible to say more than this at present. ARHOPALA ANTHELUS (Doubleday and Hewitson). Amblypodia anthelus Doubleday and Hewitson, Gen. Diurn. Lep. vol. i. p. 478. n. 5, pl. xxiv. fig. 6, f (1852); Hewitson, Cat. Lyczenide B.M. p. 5. n. 22, pl. i. figs. 23 & 24, 2 (1862). Satadra anthelus Moore, Journ. A. 8. B. vol. ii. pt. 2, p. 88 (1884). Narathura anthelus Distant, Rhop. Malay. p. 263. n. 3. Arhopala anthelus de Nicéville, Butt. Ind. vol. iii. p. 259. n. 818 (1890). Hab. Moulmein; Upper Tenasserim and ? Mergui. Expanse, ¢ 65-66, 9 60-62 mm. ¢. Upperside: both wings brilliant, lustrous, rather deep morpho-blue, deepening slightly on the outer edges; primaries with a rather narrow brown costa and wider black outer margin; secondaries with broad black costa, narrow outer margin, and a broadish patch of blackish at the anal angle. Lobe only moderately developed. Tail stout, long, black, tipped with white, the nervule on each side of which is slightly elongated so as to form a very short tooth. Underside: both wings very uniform, soft, rather pale ochreous brown, with spots, some very dark, and some with only an inner linear edging of dark brown, edged rather palely. Primaries with a heavy dark double spot at base of cell; a much larger treble confluent spot in the middle, the two upper spots having paler central irrorations ; cell closed by a very large distinct spot, above and below which is another one, the upper one having the greyish central scaling ; all these spots are almost black; subdiscal area very dark brown; transverse band composed of six spots, the upper four palish with a double edging, the inner dark, the outer slightly paler than ground, the fifth and sixth almost black; the second spot is shifted well outwards, third very slightly inwards, fourth (a darker spot than the previous ones) outwards but with an inward inclination, fifth spot large, shifted inwards, sixth very large, with an outward inclination ; a trace of a submarginal very dark row. Secondaries with a rather dark narrow costal stripe; a blackish spot at the basal extremity of the costa, with another blackish spot beside it; four basal spots, the AMBLYPODIA GROUP OF THE LYCHNIDA. 79 upper one largish, touching the second costal spot, the second and third very small and dark, the fourth rather larger, paler, and shifted inwards, followed by three spots, the upper one near the costa blackish, the second one in centre of cell, small, dark, with paler central scaling and shifted well inwards, the third larger, paler, with a good deal of paler scaling and shifted slightly outwards; cell closed by a large pale spot with a double edging, the inner one dark and outer one rather paler than the ground, below which is a smaller similar spot in the lower median angle; transverse band composed of eight spots, the first large, roundish, and black, with a little central paler scaling, second round, very small, dark, with paler centre, shifted outwards and quite isolated, third quadrate, shifted outwards, fourth right out but not detached, fifth well inwards, sixth well outwards, seventh angular, touching the long eighth spot, shifted well inwards ; all these last six spots are pale, with the double edging of dark and light; a trace of the submarginal row; lobe-spot blackish, small, and another small blackish spot on the other side of the tail, slightly edged above with greenish-bronze metallic scales, which also occupy sparingly the intermediate space. 2. Upperside: both wings lustrous silvery blue; primaries with a broad purplish- brown costa and outer margin increasing rapidly towards the apex, and with a large black cell-spot beyond the end of the upper part of the cell; secondaries with a broad brown costa, narrow outer margin increasing very rapidly from the second median nervule to the anal angle, where it is very broad. Underside like that of the male. There is generally in both sexes a light clouding, slightly mauvish, towards the end of the costa in the primaries, and below the subcostal nervule in the secondaries. This species is a very close ally of anuwnda Hew., and though the blue of the latter sometimes approaches it in shade (the type, however, being quite purplish), it is never so brilliant and morpho-like, and the black costa and outer margins are much narrower, being almost linear usually. The patterns beneath are almost the same, but the two species look quite distinct by the lack in anthelus of the white edging which is so very conspicuous a feature in anunda; the edging in the former is little paler than the ground-colour, the spots are also smaller in the secondaries, and the transverse band more irregular. Anwnda is subject to considerable variation, both as to the tone of blue and as to the underside, which sometimes is very dark indeed, and the pattern is occasionally so enlarged as to become confluent over a large area of the wing, Anthelus, on the other hand, appears to be exceedingly constant in coloration; it is a very local species, but I imagine not excessively rare in its localities. ‘The shape of the wings is likewise different, being somewhat excavated as to the outer margin of both primaries and secondaries, whilst in anwnda they are very full and rounded, and the latter are decidedly more ample. 80 MR. G. T. BETHUNE-BAKER: A REVISION OF THE ARHOPALA SUBFASCIATA (Moore). Nilasera subfasciata Moore, Proc. Zool. Soc. Lond. 1883, p. 532, pl. xlix. fig. 2. Arhopala subfasciata Doherty, Journ, A. 8. B. vol. lviii. pt. 2, p. 418. n. 5 (1889); de Nicéville, Butt. Ind. vel. 11. p. 260. n. 820 (1890). Hab. Tavoy ; Tenasserim Valley; Burma. Expanse, ¢ 44, 9 48-49 mm. This is a very beautiful species and should be distinguishable at a glance. It is, however, very close to ¢awensis mihi, but is a much less robust insect, has a much larger blue area, and is very decidedly paler beneath ; my species is more uniform in tone, and there is only a trace of the very distinct white subcostal stripe so marked in Moore’s species. ARHOPALA IJAUENSIS Bethune-Baker. (Plate I. fig. 28, ¢.) Arhopala ijanensis (vecte ijauensis) Bethune-Baker, Ent. Mo. Mag. vol. xxxiil. p. 203 (1897). Hab. Perak (Gunong Tjau). Expanse, ¢ & 2, 48-50 mm. This fine species, in the Hon. Walter Rothschild’s collection, is close to subfasciata Moore, but decidedly larger ; the differences will be better gathered by reference to the figure. ARHOPALA AUXESIA (Hewitson). (Plate III. fig. 23.) Amblypodia auxesia Hewitson, Ill. Diurn. Lep., Lye. p. 5, pl. i. figs. 1 9, 2&3 g (1865). Hab. Sumatra; New Guinea. Expanse 48-51 mmm. This species is nearest auzea de Nicéville, but the upperside colour is quite different ; the distinctions are detailed in the description of that insect. ARHOPALA FRUHSTORFER! Rober. Amblypodia fruhstorferi Rober, Ent. Nach. vol. xxii. p. 6 (1897). Hab. Java (Gede Mountain). Expanse 43 mm. This species is described from a single female; unfortunately the description is decidedly inadequate, but, so far as I can judge, not having seen the specimen, I believe it to be most probably auresia Hew. ARHOPALA AUZEA de Nicéville. Arhopala auzea de Nicéville, Journ. Bomb. Nat. Hist. Soc. vol. x. p. 169. n. 19, pl. S. figs. 29 ¢, 30 2 (1896). Hab. Java (Preanger). AMBLYPODIA GROUP OF THE LYCENID., 81 Expanse, ¢ & 2, 64 mm. This species is nearest auresia Hew., but the colour of the male is quite different, being somewhat silvery, shading into purplish at the borders, whereas awxesia Hew. is slightly greenish and very metallic, with a white spot at the end of the cell, and very broad black costa and outer margin in the primaries. ‘This species is paler in the female, with a dark spot closing the cell in the midst of a whitish patch, whereas Hewitson’s species is purplish, darker in its male, and has no spot at the end of the cell. ‘The underside pattern is likewise different, as described. ARHOPALA ANARTE (Hewitson). Amblypodia anarte Hewitson, Cat. Lycenide B.M. p. 5. n. 20, pl. il. figs. 16 & 17, g (1862) ; Druce, Proc. Zool. Soe. Lond. 1873, p. 353. n. 9; Doherty, Journ. A. S. B. vol. Iviii. pt. 2, p. 417 (1889). Arhopala anarte de Nicéville, Butt. Ind. vol. ili. p. 260. n. 819 (1890). Hab. Myitta; Burma; Malacca (interior); Borneo; Makassar. Expanse, ¢ 61-69, 2 65-68 mm. ¢. Upperside: both wings brilliant pale morpho-blue, tinged with lilac; primaries with very narrow brown costa and outer margins; secondaries with broad brown costa and very narrow outer margins. ‘Tail fine, long, tipped with white. Underside: both wings pale brown with a slight chestnut tinge, with darker spots palely encircled. Primaries with three increasing cell-spots, the second and third very large indeed, the second being the largest by its confluence with two spots over it extending up to the costal nervure, and it is also joined to the third spot by an additional one along the upper margin of the cell; below the third spot is a large one in the lower median angle, generally connected with the third by a small intermediate spot ; underneath the first and second, in the subdiscal area, is a very large spot half divided on the upper part of the centre; transverse band very unusual in form, divided into two distinct portions, the upper part forming a broad arc of five increasing spots, beginning over the third cell-spot and extending to the submarginal row; the lower part is composed of two large ovate spots below each other, shifted so far inwards as nearly to touch the third cell-spot ; submarginal row very indefinite and indistinct. Secondaries with a slight lilac suffusion in the upper half; at the extreme basal extremity of the costa are one or two spots, with four large basal spots below each other, second and fourth shifted inwards, followed by three larger spots below each other ; cell closed by a very large irregular spot, below which is a small one in the lower median angle; transverse band composed of four pairs of spots, the second spot shifted outwards, the second pair is shifted right outwards, the lower of which is the further out, the third pair well inwards, fourth pair inwards again, composed of the usual angular spot connected by a line or by a narrow neck with the long eighth spot; submarginal row fairly distinct, marginal row less so; a black velvety spot at the VoL, Xvi. Part I. No. 11.— August, 1903. M 82 MR. G. T. BETHUNE-BAKER: A REVISION OF THE anal lobe, which is slightly developed, preceded by two black spots, the middle one being almost entirely covered with pale bluish-metallic scales, and the other two broadly edged above in like manner; the submedian nervule is elongated, so as to form a tooth making a second short inner tail. @. Upperside: both wings pale lustrous violet-blue, more violet in hue at the outer edges; primaries with rather broad brown costa and broader brown outer margin; secondaries with broad brown costa and less broad brown outer margin, decreasing towards the anal angle. Underside just like that of the male, but without the lilac suffusion in the secondaries. This is a very distinct species and may be at once recognised by its lustrous lilac-blue above, and by its distinct markings below, especially its transverse band. Mr. Hewitson confused this species with agnis Feld. (which latter he correctly figures on pl. i. figs. 6 & 7 of Ill. Diurn. Lep.) ; he, however, figures his own species fairly well in the first instance in the Cat. B. M. pl. iii. figs. 16 & 17, but the transverse band of the primaries is not quite accurate; the spots should be separate, and the two lowest spots forming the lower portion of the band are quite distinct and far separated from the upper five; this, I believe, arises from the figure not being large enough, so that the markings are cramped to get them all in. ‘There should, however, be no difficulty whatever in recognising at a glance this lovely insect. ARHOPALA ACHELOUS (Hewitson). Amblypodia achelous Hewitson, Cat. Lye. B.M. p. 7. n. 80, pl. v. figs. 7, 8 ¢ (1862) ; Druce, P. Z. 8. Lond. 1878, p. 354. n. 13. Narathura achelous Distant, Rhop. Malay. p. 271. n. 18, woodeut n. 82, ¢ (1885). Arhopala achelous de Nicéville, Butt. Ind. vol. iii. p. 248 (1890); Druce, P. Z. 8S. Lond. 1895; p- 590. Hab. Singapore ; Borneo. Expanse, ¢ & 9, 46 mm. This species is very marked and should be recognised without difficulty, the spots beneath are very dark and prominent and the pale purplish subcostal stripe is a marked character. ARHOPALA HAVILANDI Bethune-Baker. Arhopala havilandi Bethune-Baker & Druce, P. Z. 8. Lond. 1896, p. 665, pl. xxx. figs. 8 $,9 ?. Hab. Kina Balu; Borneo. Expanse, ¢ & 2, 47-48 mm. I originally compared this species with aida de Nicéville, but having more recently seen a larger series of achelous Hewitson, I have no doubt the present is its proper place: it may be easily separated by the deeper purplish blue of the upperside, and by the quite uniform tone of the underside; it entirely lacks the pale subcostal streak AMBLYPODIA GROUP OF THE LYCZENIDZ. 83 of the secondaries as well as the large dark spots which are so prominent in Hewitson’s species. ARHOPALA AGABA (Hewitson). Amblypodia agaba Hewitson, Cat. Lycenide B.M. p. 8. n. 39, pl. iv. figs. 39 & 40, 2 (1862). Satadra agaba Moore, Journ. A. 8. B. vol. lili. pt. 2, p. 38 (1884); Elwes and de Nicéville, Journ. A. S. B. vol. lv. pt. 2, p. 430. n. 104 (1886). Arhopala agaba de Nicéville, Butt. Ind. vol. iii. p. 244. n. 797 (1890) ; Doherty, Journ. A.S. B. vol. lvii. pt. 2, p. 422. n. 20 (1889). Hab. India; Mergui Archipelago ; Upper Tenasserim ; Tavoy. Expanse, ¢ & 2, 44-47 mm. ¢. Upperside: both wings bluish purple; primaries with the costa having a very fine linear black line and a narrow black outer margin; secondaries with a broad blackish costa and narrow black outer margin; abdominal fold grey; tail long, black, tipped with white. Underside: both wings brown tinged with purple, with a super- imposed strong lilac gloss, with darker spots margined with very pale lilac. Primaries with the lilac gloss confined to the costal half of the wing; three increasing spots in the cell, above the third spot a small one close to the costa, below it (the third spot) a larger one; subdiscal area dark; transverse band oblique, outwardly almost straight, but occasionally slightly bent at or below the fourth spot, the band being composed of seven spots, not quite evenly placed, but in the main in a fairly straight direction ; submarginal row distinct, submedian area pale yellowish grey. Secondaries with the lilac gloss more or less all over, but strongest near the base and at the apex; four basal spots, the third shifted slightly outwards and the fourth well inwards, followed by three larger spots below each other; cell closed by a subquadrate spot, below which is a small one connecting it with the lowest of the three just mentioned, the spot closing the cell also touching the transverse band giving it a bifurcated appearance; transverse band composed of eight spots, the first two below each other and large, the third shifted well outwards, fourth further out, fifth slightly inwards, sixth outwards, seventh angular spot well inwards, eighth long spot outwards but inclined upwards to the internal nervure; submarginal band distinct, largish, somewhat indefinite, a black spot on the small anal lobe, beyond which a few greenish-metallic scales are sprinkled. 2. Upperside: both wings bright ultramarine-blue, tinged with purple in certain lights; primaries with costa broadly, outer margin more broadly, and apical area still more broadly black; secondaries with very broad margins all round, but decreasing towards the anal angle. Underside like that of the male, but often with much less of the lilac glossing. This is probably a somewhat local species, though not rare in its own localities. From havilandi it may be at once separated by its decidedly smaller size and its much M2 84. MR. G. T. BETHUNE-BAKER: A REVISION OF THE duller and purpler and non-lustrous blue, whilst below the pattern is quite different. It is rare in European collections, and as the male has never been figured, I have thought it well to do so now. ARHOPALA SELTA (Hewitson). Amb!ypodia selta Hewitson, Il. Diurn. Lep., Lye. p. 14. n. 92, pl. iia. figs. 36 & 37 (1869). Satadra selta Moore, J. A. S. B. vol. liii. pt. 2, p. 88 (1884). Arhopala selta de Nicéville, Butt. Ind. vol. iii. p. 245. n. 798 (1890). Hab. Moulmein; Mergui; Tenasserim. Expanse, ¢, 38 mm. ‘This species is not unlike a small specimen of agaba Hewitson, but it lacks the strong purplish tint of the underside, which is replaced by a much more subdued pink, and the pale lilac washing is totally absent; in the secondaries also the transverse band is quite disconnected from the upper two spots and is shifted well out from them and the spot closing the cell. Agada is also a decidedly larger insect and is more purple and duller in tone, whilst the black margins are much narrower, especially in the secondaries. In de Nicéville’s collection are several specimens from Tenasserim, and one female in which the blue is bright lustrous cobalt. ARHOPALA BROOKE! n. sp. (Plate I. fig. 29, 3.) Hab. Pulo Laut; Borneo. Expanse, ¢, 40-44 mm. ¢. Upperside: both wings very deep indigo-purple, nearly as dark as in dazalus Hew. ; primaries with a very fine black costa and the finest black linear outer margins ; secondaries with broad brown costa and fine blackish outer margins ; tail fine, white- tipped ; abdominal fold brownish grey. Underside: both wings dull brown with slightly darker markings, faintly and palely encircled. Primaries with all the markings very obscure, three small increasing cell-spots, below the third a smaller one in the lower median angle; transverse band composed of five spots, the first three outwardly oblique, the fourth shifted inwards, below which is the fifth, shifted outwards; sub- marginal row indistinct ; submedian area paler. Secondaries with an indistinct spot at the basal extremity of the costa, four basal spots below each other, the third shifted outwards, followed by three larger spots below each other; cell closed by a subquadrate spot, below which is a small spot in the lower median angle; transverse band composed of eight spots, the second one much larger than the first and touching the spot closing the cell, third shifted right outwards and detached, fourth outwards again but with an inclination inwards, fifth inwards, sixth outwards, seventh angular spot right inwards and detached, and practically confluent with the long eighth spot; submarginal row rather indistinct, darker over the pale sky-blue metallic scaling at anal angle. AMBLYPODIA GROUP OF THE LYCHNIDE. (o,9) 9) I have before me one specimen from Pulo Laut in which the brown surface below is quite lustrous and the markings of the primaries are much more distinct. : This species is nearest arow Hew., but can be recognised immediately by the exceptionally dark blue of the upper surface, in which it differs from all its near allies. ARHOPALA SCEVA, Nl. sp. Hab, Sumatra. Expanse, ¢ & 2, 40 mm. $. Upperside: both wings purple of a reddish tint, in a side light haying a brownish shade ; primaries with a narrow brown costa and outer margin; secondaries with a broad brown costa and narrow outer margin; the lower median nervule is elongated into a distinct tooth, but apparently there is no tail. Underside: both wings warm greyish brown with darker spots palely encircled. Primaries with three increasing cell-spots; below the second and third are two smaller ones divided by the lower median nervule; transverse band composed of six spots with a detached one below the sixth in the submedian area; first spot small and round, second, third, and fourth each shifted beyond its predecessor and increasing in size, of a broadish oval shape, fifth spot of same shape shifted inwards, sixth inclined slightly outwards, and below this the seventh detached spot already referred to; submarginal row very obscure, Secondaries with four basal spots, second and fourth shifted slightly inwards, followed by three large spots below each other, of which the second and third touch ; cell closed by a largish spot, below which is a very small one in the lower median angle; transverse band composed of eight spots, the second shifted well outwards, third and fourth confluent, shifted well outwards again, fifth narrow, wedge- shaped, shifted inwards, sixth outwards, seventh and eighth confluent and shifted right inwards ; submarginal row less obscure than in the primaries; anal spot deep black, preceded by two others, more or less covered with metallic bluish or greenish scales. 2. Upperside: both wings violet; primaries with broad brown costa, broader outer margin, and yet broader apical area; secondaries with very broad margins all round. Underside as in male. This species has a very distinct underside ; the spots are prominent and the transverse bands have an unusual look difficult to describe, especially in the primaries, in which the spots are of a broad flattened oval shape. ARHOPALA AROA (Hewitson). (Plate V. figs. 3 & 3a.) Amblypodia aroa Hewitson, Ill. Diurn. Lep., Lye. p. 13. n. 60, pl. ii. fig. 12, ¢ (1862) ; Butler, Trans. Linn. Soc. Lond., ser, 2, Zool. vol. i. p. 548 (1877). Narathura aroa Distant, Rhop. Malay. p. 266, pl. xxiii. fig. 17, g (1885). Arhopala aroa de Nicéville, Butt. India, vol. ui. p. 244 (1890). 86 MR. G. T. BETHUNE-BAKER: A REVISION OF THE Amblypodia pryerit Butler, P. Z. 8. Lond. 1892, p. 121. Arhopala aroa and pryeri Druce, P. Z. 8S. Lond. 1895, p. 590. Arhopala aroa de Nicéville and Martin, J. A. S. B. vol. Ixiv. pt. 2, p. 465, n. 894 (1895). Hab. Sumatra; Malacca; Java; Burma; Penang; Singapore; Borneo; Batchian ; and New Guinea. Expanse, ¢ & 92, 42-47 mm. This species is easily separated from se/ta by the plain brown underside, also by its larger size and narrow black margins above. It is an insect of very wide range and varies somewhat in tone of colour both above and below, the blue sometimes being quite blue and sometimes purplish; the underside markings are very constant. It comes within de Nicéville’s Indian region; I have specimens from Burma and from Tenasserim. In Sumatra it is common, and I have it also from Java, Malacca, Singapore, Batchian, Borneo, and New Guinea. The genitalia are of the ordinary pattern, the hooks being slender, but the clasps robust, whilst the penis has a some- what peculiarly shaped orifice, as will be seen from the figure (Pl. V. fig. 3a). ARHOPALA ELOPURA Druce. (Plate V. figs. 4 & 4a.) Arhopala elopura Druce, Ent. Mo. Mag. ser. 2, vol. v. p. 9 (1894) ; id. P. Z.S. Lond. 1895, p. 590, et 1896, p. 670, pl. xxix. fig. 6. Hab. Borneo. Expanse, ¢ & 9, 39-42 mm. This is evidently one of the commonest species of the genus in Borneo. I have specimens before me taken in every month from March to September inclusive. It is nearest to aroa Hewitson, but may be recognised by its more purple colour above, and easily by its pale underside, by the markings being no darker than the ground- colour, and by the transverse band of the primaries being always dislocated at the fourth spot. The genitalia are peculiar, the hooks being broadly spatulate, with an ample hooded tegumen and heavy clasps ; the orifice of the penis has a peculiar projecting upper lip, as shown in the figure. ARHOPALA RAFFLESI de Nicéville. Narathura amphinuta Distant (nec Felder), Rhop. Malay. p. 267. n. 2, pl. xxi. figs. 9 2,10 ¢ (1885). Arhopala rafflesii de Nicéville, Butt. India, vol. iii. p. 248. n. 803, pl. frontispiece fig. 136. Arhopala pseudomuta de Nicéville and Martin, J. A. S. B. vol. Ixiv. pt. 2, p. 465 (1895). Hab. Burma; Penang, Province Wellesley; Singapore; Sumatra; Mergui. Expanse, ¢ 43, 9 40-41 mm. This species is nearest pseudomuta Staudinger, but can be recognised at once by its much brighter, lighter, and bluer colour, and by the very decidedly broader borders of the upperside, and beneath by the difference in the transverse bands of both wings ; in the primaries the fourth spot is always shifted well out and dislocated (not so in AMBLYPODIA GROUP OF THE LYCANIDE. 87 pseudomuta), whilst in the secondaries the whole band is different. In Staudinger’s species spots 1 and 2 are quite isolated, and 3, 4, and 5 form an even arc, whilst in raffles 1 and 2 are confluent and almost quadrate, and 3 to 6 are nearly straight; the latter is also a smaller insect. De Nicéville and Martin (/. ¢.) sink this species under pseudomuta Staudinger, but they are quite distinct. ARHOPALA PSEUDOMUTA (Staudinger). (Plate I. fig. 30, 3.) Amblypodia pseudomuta Staudinger, Iris, vol. i. p. 125 (1889). Arhopala pseudomuta Bethune-Baker and Druce, P. Z. S. Lond. 1896, p. 662. Hab. Malacca; Borneo. Expanse, ¢ 46, 9 47 mm. This species may be recognised from raflesit de Nicéville by its larger size and blaer colour ; more detailed differences are given in the description of that insect. ARHOPALA ATOSIA (Hewitson). (Plate V. figs. 5 & 5a.) Amblypodia atosia Hewitson, Ill. Diurn. Lep., Lyc. p. 9. n. 37, pl. 11. figs. 8&9, 2 (1863) ; Druce, P. Z. S. Lond. 1873, p. 353. n. 4; Butler, Trans. Linn. Soe. Lond., ser. 2, Zool. vol. i. p. 548. n. 5 (1877). Narathura atosia Distant, Rhop. Malay. p. 265. n. 7, pl. xxiii. figs. 5 9,6 d (1885). Arhopala atosia de Nicéville, Butt. Ind. vol. ii. p. 241, frontispiece fig. 188, ¢ (1890); Druce, P. Z. 8. Lond. 1895, p. 589. Arhopala atosia de Nicéville, J. A. 8. B. vol. lxiv. pt. 2, p. 465 (1895). Amblypodia aricia Staudinger, Iris, vol. 1. p. 124, Taf. i. fig. 15 (1889). Hab. Sumatra; Borneo; Malacca; Mergui; Tenasserim Valley; Burma; Penang ; Sapanaya; Palawan. Expanse, ¢ & 2, 39-45 mm. 3. Upperside: both wings violaceous blue, with a linear black costa and posterior margin to primaries, and a large patch of differently arranged scales in the centre of the wing, showing darkly through in certain lights, but not altering the colour of the wing; a broad costa to secondaries and a linear black posterior margin ; abdominal fold greyish; tail black, tipped with white. Underside: both wings greyish brown with scarcely darker spots palely encircled. Primaries with three increasing discoidal spots; beneath the second and third are two spots in the submedian area and in the lower median angle respectively; above the third spot is often a small one near the costa; transverse band consisting of four confluent spots outwardly inclined, the fourth shifted further out than the others, followed by two confluent spots shifted right inwards; submarginal row just traceable. Secondaries with four basal spots inclined inwardly, but the third spot is projected outwards, followed by three rather larger spots below each other; cell closed by a subquadrate spot, below which is a small one in the lower median angle; transverse chain-band composed of two upper spots touching each other, detached from rest of band, the third and fourth spots are each shifted outwards, fifth inwards, its inner margin being far inwards, sixth very 58 MR. G. T. BETHUNE-BAKER: A REVISION OF THE slightly outwards, the seventh angular spot well inwards and confluent with the long eighth spot; submarginal row just traceable; a black spot at anal angle and between the submedian and lower median nervules, over which are scattered thickly bright metallic-blue scales which extend beyond the lower median nervule; over these metallic scales there is a dark edging, extending up to the internal nervure. @. Upperside: both wings pale violet; primaries with broadish blackish costa to near end of cell, where it increases suddenly into the exceedingly broad apical area ; outer margin very broad; secondaries with broad dark costa and decidedly less broad outer margin. I have before me three distinct forms of the female, viz.:—the type; another form from Penang, in which the colour is bright deep lustrous blue, not violet at all, with quite typical males; and a third form from Tenasserim, in which the colour is pale lustrous silvery blue, but here many of the males are paler, bluer, somewhat silvery, and more lustrous, approaching near to the colour of pastorella Doherty. ‘The under- side pattern of each is just the same, like the typical males. The species can only be separated from epimuta by the presence of tails. ‘The violet colour of the upper surface distinguishes the males from any close allies, and also the pattern beneath, _ especially the dislocated transverse band. Staudinger’s species aricia is identical with the pale silvery-blue form from Burma, showing the darker patch of scales, but less distinctly. The genitalia also help to separate this species from epimuta, the clasps being less broad at the base and rounder at the extremities; the hooks in both are broadish and inclined to be spatulate, not tapering narrower as is often the case (see Pl. V. fig. 5). ARHOPALA EPIMUTA Moore (nec Hewitson). (Plate II. fig. 9, ¢; Plate V. figs. 6 & 6a.) Amblypodia epimuta Moore, Cat. Lep. E.1.C. p. 42 (1857). Arhopala epimuta Druce, P. Z. 8. Lond. 1895, p. 593. Arhkopala antimuta de Nicéville, Butt. India, vol. iii. p. 277 (1890). Arhopala antimuta de Nicéville and Martin, J. A. S. B. vol. lxiv. pt. 2, p. 469 (1895). Arhopala epimuta Druce, P. Z. 8. 1895, p. 593. flab. Borneo Straits ; Perak ; Penang; Mergui; Tenasserim. Expanse, ¢ & 2, 44-46 mm. The species is an exceedingly close ally of atosia Hew.: in fact, almost the only character by which they can be separated is that Hewitson’s species has tails and this has none. This is the insect that de Nicéville identified as antimuta Feld., the type of which I have carefully examined, and I find that davisonii de N. is the antimuta of Felder. It will be observed that it has an exactly similar distribution as atosia Hew. I have specimens before me from each of the same localities. The genitalia of this species are referred to in the description of A. atosia Hew.; in both the penis is AMBLYPODIA GROUP OF THE LYCANIDA. 89 peculiar, being strongly curved near the centre, whilst the orifice is somewhat trumpet- shaped, but with a strongly projecting upper lip, which is best shown in the figure of the previous species (Pl. V. fig. 5 a). ARHOPALA MOOLAIANA (Moore). (Plate II. fig. 6, 3.) Narathura moolaiana Moore, P. Z. S. Lond. 1878, p. 835. Amblypodia epimuta Hewitson (nec Moore), Cat. Lyczenide B.M. p. 11. un. 51, pl. vi. figs. 59 & 60, 2 (1862). Arhopala pastorella Doherty, J. A. S. B. vol. lviii. pt. 2, p. 418, pl. xxiii. fig. 12 (1889) ; de Nicéville, Butt. Ind. vol. i. p. 274 (1890). Arhopala moolaiana de Nicéville, Butt. Ind. vol. ii. p. 274 (1890). Arhopala agelastus de Nicéville, Butt. Ind. vol. iii. p. 278 (1890). Narathura agelastus Moore, Journ. Linn. Soc. Lond., Zool. vol. xxi. p. 44 (1886) ? Hab. Burma; India; Tenasserim; Mergui; Penang; Perak; Bheeling. Expanse, ¢ 47-52, 9 47-50 mm. 3. Upperside: both wings morpho-blue, shading towards the outer margins into darker slightly purplish blue ; primaries with a narrow blackish costa and outer margin ; secondaries with a not very broad brown costa and narrow outer margin. Underside: both wings ochreous brown with darker spots palely encircled. Primaries with three increasing cell-spots, first and second rather small, third subquadrate, below the two last are two other spots separated by the lower median nervule; transverse band composed of six spots, the first four almost confluent and with an outward oblique inclination, forming a very slight arc, the first spot on the costa quite small, fifth spot shifted inwards, below which is the sixth; submarginal row distinct ; marginal row barely discernible ; submedian area pale; subdiscal area with a purplish-brown tinge. Secondaries with four small basal spots below each other; second and fourth shifted slightly inwards, followed by three large spots below each other, the middle one shifted slightly inwards ; cell closed by a slightly reniform spot, below which is a small one in the lower median angle; transverse band composed of eight rather distinct spots, the second shifted right out, just touching the outer edge of the first, third shifted right outwards again, touching the fourth below it, fifth shifted well inwards, sixth outwards, seventh angular spot right inwards, not disconnected, touching the irregular-shaped eighth spot ; submarginal row distinct, marginal row less distinct, a blackish spot at the anal angle preceded by two others which are almost hidden by the superimposed pale greenish-metallic scales, the spot at the angle having a broadish upper edging of these scales. 2. Upperside: both wings pale morpho-blue, darker at the edges ; primaries with broad black costa, increasing towards the apex, and very broad outer margin; the black apical area is extensive ; secondaries with a very broad blackish costa and narrow outer margins, Underside as in male. VOL. XVvil.—Part 1. No. 12.—August, 1903. N 90 MR. G. T. BETHUNE-BAKER: A REVISION OF THE In a male specimen from Perak the blue has a violet tinge, but is almost as lustrous as in the usual form. De Nicéville (. ¢.) has referred to this species under the name agelastus Hew. In his fine collection of this group, which he has so kindly lent me, is a large series of this species, the females preponderating by two to one male, taken mainly in Burma and Tenasserim, but there are specimens also from Mergui, Penang, and Bheeling. I have no doubt, from the single example of Doherty’s species pastorella before me, and from his figure of it, that this is identical with moolaiana. ARHOPALA AMPHIMUTA (Felder). (Plate II. fig. 7, 3.) Amblypodia amphimuta Felder, Wien. ent. Monatsch. iv. p. 396 (1860). Arhopala amphimuta Felder, Reise Novara, Lep. vol. ii. p. 232, tab. xxix. fig. 8 (1865). Arrhopala amphimuta de Nicéville and Martin, J. A. S. B. vol. lxiv. pt. 2, p. 469 (1895) ? Arhopala amphimuta Druce, P. Z. 8. 1895, p. 592. Hab. Malacca. Expanse, ¢, 50 mm. ¢. Upperside: both wings soft violet, with a brownish shade in side-lights; primaries with costa and outer margin very narrowly brown; secondaries with broad brown costa and very narrow outer margins. Underside clear ochreous brown with slightly darker spots palely encircled. Primaries with three increasing cell-spots, below the third a small one in the lower median angle ; transverse catenulated band composed of six spots fractured below the fourth, the upper four havea slight outward inclination, almost a curve, the first spot on the costa being very small, the fifth is shifted well inwards, and the sixth slightly inwards also ; there is the least trace of a submarginal row; the submedian area is slightly paler than the ground-colour. Secondaries with four basal spots below each other, the second and fourth shifted slightly inwards, followed by three larger spots with a slight inward inclination ; cell closed by a subovate spot, below which is a small one in the lower median angle; transverse catenulated band composed of eight spots in pairs, the first two quite separated and distinct, the Jower of which is shifted well outwards, the third and fourth confluent, shifted well outwards, fifth and sixth confluent, shifted inwards, seventh angular spot confluent with the eighth and shifted again inwards; submarginal row distinct, marginal row less distinct ; at the anal angle is a very large blackish spot extending into the next nervule space and well covered with a bright pale blue metallic spot, of which there is a slight trace in the second median nervule space. | This species can at once be separated from moolaiana Moore by the violet colour of the upperside, which in Moore’s insect is almost morpho-blue; the underside is also somewhat different as described. From inornata Feld., a still closer ally, it is more difficult to separate; it has, however, a much duller upper surface and is not so blue, whilst beneath the pattern is much more distinct and both transverse bands are more AMBLYPODIA GROUP OF THE LYCANID. 91 dislocated. Both Mr. Butler and Mr. de Nicéville are right in considering this species distinct from hypomuta Hew. It is quite distinct. I have examined the type of each most carefully and have no hesitation in saying that they are so distinct that probably both Mr. Hewitson, and certainly Mr. Distant, wrongly identified Felder’s species. I ~ am led to this conclusion from Mr. Distant’s figures, which do not resemble the species in either shape, colour, or pattern. It is quite different in shape from hypomuta, its costa being very strongly arched and the wing very much broader in the primaries; in the secondaries the outline is more angular, hypomuta having the outer margin a very even curve, it (amphimuta) is also half as large again and the underside pattern will be seen, on reference to my figures, to be also different. Amphimuta appears to be an exceedingly rare species, and as yet I have not seen a female of it. ARHOPALA INORNATA (Felder). (Plate II. fig. 8, 2.) Amblypodia inornata Felder, Wien. ent. Monatsch. vol. iv. p. 396 (1860); Hewitson, Cat. Lyczenide B.M. p. 12 (1862). Arhopala inornata Felder, Reise Novara, Lep. vol. ii. p. 234, pl. xxix. fig. 12, g (1865). Narathura inornata Distant, Rhop. Malay. p. 271, woodcut u. 81, g (1885). Hab. Malacca, interior ; Philippine Islands. Expanse, ¢ 49, 2 49-50 mm. é. Upperside: both wings rather bright violet-purple, with the linear brown costa and outer margin in the primaries ; secondaries with a broadish brown costa and linear outer margin. Underside: both wings greyish brown, with slightly darker markings palely encircled. Primaries with all the pattern very obscure, with three increasing cell-spots, below the second and third are two almost obsolete spots divided by the lower median nervule ; transverse band composed of five spots, beginning below the upper discoidal nervule, the first three confluent, with an outward inclination, the third spot being angled on its interior edge and so made almost confluent with the fourth and fifth, which are directly below each other and shifted slightly inwards; submedian area slightly paler than the ground-colour. Secondaries with pattern plainer than in primaries, with four small cramped spots very near the base of wings, followed by three decidedly larger ones below each other ; cell closed by a reniform spot, below which is a small one in the lower median angle ; transverse band composed of eight spots, the second being very much larger and directly below the first, the third and fourth confluent, shifted right out on to the outer edge of the second, fifth and sixth confluent with the fourth, but with their inner edges only shifted slightly inwards, seventh angular spot shifted right inwards, disconnected, but confluent with the eighth long spot ; submarginal row represented by a band of dark shading. 2. Upperside: both wings pale violet, with a broad brown costa and outer margins increasing towards the apex, which is largely brown in the primaries. In the Nn 2 92 MR. G. T. BETHUNE-BAKER: A REVISION OF THE secondaries the costa is broadly brown and the outer margin narrowly brown, increasing slightly close to the anal angle. The underside is exactly like that of the male, but rather paler and with the pattern rather plainer. This description is taken from Felder’s types from the Malacca interior, which the Hon. Walter Rothschild has kindly placed at my disposal. The species is very distinct and easily recognisable. I have, however, a specimen from Perak in which the blue is very much more extensive and is of a lustrous azure- blue with a violet tinge; the underside is of exactly the same tone and pattern. This species is very near amphimuta Feld.; it is, however, decidedly brighter and purpler, and its dark margins are but linear, whilst on the underside the pattern is very obscure, the transverse band of the primaries is not fractured, and the same band in the secondaries is different, the four middle spots being practically confluent ; it has also no dark spots or metallic scaling at all at the anal angle. The species is evidently very rare; I have only seen three specimens, viz. the types from Malacca interior and a female from Perak. ARHOPALA ASIA de Nicéville. Arhopala asia de Nicéville, Journ. Bombay Nat. Hist. Soc. vol. vii. p. 333. n. 9, pl. H. fig. 16, 3 (1892) ; id. Journ. A. S. B. vol. Ixiv. pt. 2, p. 470. n. 426 (1895). Hab. Quang, Malay Peninsula; Sumatra, Borneo, and Malacca. Expanse, ¢ & 2, 48-50 mm. I possess specimens of this species from Labuan, where Waterstradt obtained it in the year 1893 (I believe). The late Dr. Staudinger’s collection also has specimens from Borneo. All these are somewhat bluer and darker than the type, though evidently the same species. I notice also that de Nicéville’s Sumatra specimens are darker than his type from the Malay Peninsula. It is not difficult to separate them from tnornata Feld., as its wings are broader and the secondaries decidedly more ample and the brown margins rather wider. Beneath it is more easily separated, as the pattern is much plainer and is very distinctly spotted, whereas in inornata the spots are much more confluent ; this latter feature, of course, mainly refers to the transverse band. It is, I imagine, not very rare where it occurs; I have a fair series before me from Borneo. The genitalia are very similar to those of A. cator?, but are decidedly larger. ARHOPALA KURZI (Distant). Narathura kurzi Distant, Rhop. Malay. p. 682. no. 12, pl. xxi. fig. 1, ¢ (1885). Arhopala kurzi de Nicéville, Butt. Ind. vol. iii. p. 278 (1890). Hab. Malacca; Penang. Expanse, ¢, 44 mm. This is a smaller and slighter insect than asia of de Nicéville, and the colour above AMBLYPODIA GROUP OF THE LYCZENIDZ. 93 is decidedly different. On the underside the transverse band of the primaries will at once separate it, and the same band in the secondaries is also very different. I have only seen one specimen, kindly sent by de Nicéville. ARHOPALA AGESILAUS (Staudinger). (Plate II. fig. 10, ¢; Plate V. figs. 7& 7a.) Amblypodia agesilaus Staudinger, Iris, vol. ii. p. 127, Taf. i. fig. 16 (non 17) (1889). Arrhopala agesilaus de Nicéville and Martin, J. A. S. B. vol. Ixiv. pt. 2, p. 465 (1895). Hab. Palawan; Sumatra; Mindanao; Borneo; Perak; Malacca. Expanse, ¢ 44-45, 9 39-43 mm. This species may be separated from asia de N. by its much smaller size, the very broad borders to its wings, and by its bluer and paler upper surface. From kurzi Distant it may be separated by the difference in colour and by the very broad borders of the upperside, and beneath by the very different transverse band. It appears to be widely spread throughout the Malay Islands. I have many specimens before me from Palawan, and it also occurs in other islands of the Philippine group. I have examples also from Nias, Borneo (where it is not uncommon), Perak, and Sumatra, where the colour of the upperside is decidedly brighter blue. Staudinger suggested that the name v-major might be given to a larger insect which is paler beneath and also above; but this specimen from Sumatra does not belong to this species at all; and though the Bornean specimens are decidedly larger than those from most other localities, there is no other difference, and as it was not described the name may be allowed to drop. The genitalia are unusually small, the hooks have an upward inclination, and the penis is straight, with a trumpet-shaped orifice, as is shown in figure 7 a, Plate V. ARHOPALA CATORI, n. sp. (Plate II. fig. 11, g; Plate V. figs. 8 & 8a.) Hab. Borneo; Bilit; Palawan. Expanse, ¢ 44-46, 9 42-44 mm. 3. Upperside: both wings dark purplish sublustrous blue, in a side light the tone being very deep ; primaries with costa dark brown of medium width, outer margin very broad, wider towards the apex ; secondaries with very wide margins all round. Under- side : both wings warm ochreous brown (occasionally sub-olivaceous), with rather darker spots palely encircled. Primaries with three rather small increasing cell-spots, below the third an indistinct one ; trausverse band composed of five spots (very occasionally six by the addition of a small and very indistinct one just below the costa), the first (on the third subcostal nervule), second, and third almost confluent, with an oblique out- ward inclination, fourth and fifth shifted well inwards but having an oblique outward inclination ; submarginal row obscure; submedian area very slightly paler than the ground. Secondaries with four very small basal spots, the third shifted outwards, a rather larger spot in the centre of the cell, below which is a larger one; cell closed by 94 MR. G. T. BETHUNE-BAKER: A REVISION OF THE a subovate spot, below which is a very small indistinct one in the lower median angle; transverse band composed of eight spots, the first and second isolated, second shifted outwards, third and fourth shifted right outwards, fifth narrow inwards, sixth outwards, seventh right inwards and confluent with a long eighth spot ; submarginal row fairly distinct, edged inwardly with a light border; a black spot at the anal angle preceded by a blackish patch almost entirely covered with bright azure-blue metallic scaling which edges the anal spot broadly above. 2. Upperside: both wings paler violet-blue; primaries with broad brown costa and very broad outer margin, each increasing towards the apex; secondaries with very broad brown margins all round. Underside just like that of the male, but rather paler. This species is near agesilaus Stdr., but has rather broader borders above, whilst beneath it is at once recognizable from any of its near allies by the complete absence of any spot just below the costa between the upper basal and the first spot of the transverse band of the secondaries. This is quite a remarkable character. The spots are all smaller and more obscure. I have sixteen specimens before me from Borneo and Bilit (a small island off the mainland), and one from Palawan. I dedicate the species to Mr. Cator, who has brought home a good many examples of the genus and has kindly placed them at my disposal. I received it, however, in single specimens from other entomologists, but Mr. Cator brought the majority. The genitalia have an ample hooded tegumen with moderately strong hooks that are inclined downwards, the clasps are conical at their extremities; the penis of moderate length, but not quite straight, as shown in figures 8 & 8 a of Plate V. ARHOPALA SIMILIS Druce. Arhopala similis Druce, Proc. Zool. Soc. Lond. 1895, p. 592. Arhopala anila de Nicéville, J. A. S. B. vol. Ixiv. pt. 2, p. 469 (1895). Arrhopala anila de Nicéville, Journ. Bomb. Nat. Hist. Soc. vol. x. p. 6 (1896). Hab. Sandakan; Sumatra; Salangor. Expanse, ¢ 42-43, 2 44-40 mm. 3¢. Upperside: both wings bright blue with a violet tinge; primaries with a broad blackish costa and posterior margin, each of which increases rapidly towards the apex, so that the apical area is as wide asin the female of most species of the group ; secondaries with broad margins all round, the costa being broadest. Underside as in agesias Hewitson, except that there is no transverse band at all in the primaries. 9. Just like the male, except that the blue of the upperside is less violet in tinge. There’ is no difference in the extent of blue in males and females. ‘The species is very close to agesias, but may be recognised by the absence of the transverse band in the primaries. Dr. Martin obtained it in fair numbers in Sumatra. AMBLYPODIA GROUP OF THE LYCZHNIDA. 95 ARHOPALA AGESIAS (Hewitson). (Plate III. fig. 27; Plate V. figs. 9 & 9 a.) Amblypodia agesias Hewitson, Cat. Lyc. B.M. p. 2. n. 49, pl. vi. figs. 55 & 56 (1862). Amblypodia ovomaculata id. Ill. Diurn. Lep., Suppl. p. 22. no. 108, pl. suppl. viii. figs. 66 & 67 (1878). ' Arhopala agesias de Nicéville, Butt. Ind. vol. iii. p. 273 ; Druce, Proc. Zool. Soc. Lond. 1895, p. 592. Arrhopala agesias de Nicéville and Martin, J. A. S. B. vol. lxiv. pt. 2, p. 469 (1895). Hab. Borneo; Malacca; Sumatra; Pulo Laut. Expanse, ¢ & 2, 39-45 mm. This species is subject to considerable variation as to the size and colour of the spots beneath ; in Hewitson’s type these are small and dull, whilst in others they are very large and very dark, and the metallic scaling is often much extended towards the costa and differs in colour. I have a specimen from Kina Balu before me in which the colour is brilliant golden bronze, others are bright blue or green; the blue is also subject to variation, some being quite violet, others much bluer, whilst in the Kina Balu specimen the blue is brilliant azure. Hewitson describes a female in which the trans- verse band is absent ; this has been named similis by Druce and anila by de Nicéville. The male of this species appears to have been mistaken for the female, as both sexes are precisely the same. I have mounted the clasping-organs of one male, and have before me other specimens which I believe to be of the same sex. As I have specimens from Sumatra, Borneo, and Pulo Laut (a small island off New Guinea), it is not improbable that it may be more widely distributed over these islands than is yet recorded. : The genitalia are peculiar ; the tegumen small with extremely broad clasps, the clasps are long, with a waved outline, and the penis is bent a quarter from the orifice (see figs. 9 & 9a, Plate V.). A. ovomaculata I cannot regard as anything but a very handsome, darkly and very largely marked form of this species; the pattern is exactly the same, there being no difference whatever above or below, except that all the spots are very greatly exaggerated in the secondaries, as also is the metallic anal scaling. I have figured this form on Jellenee JUD shee, A ARHOPALA HESBA (Hewitson). (Plate III. fig. 28, 3.) Amblypodia hesba Hewitson, Ill. Diurn. Lep., Lye. p. 142, pl. iii b. fig. 47 (1865) Hab. Philippine Islands. - Expanse, ¢ & 2, 39-40 mm. A very distinct and easily recognisable species; the brilliant morpho-blue of the upperside and the decided set pattern and warm chestnut-brown of the underside give it quite a marked individuality of its own. It has hitherto been recorded from the Philippine Islands only. 96 MR. G. T. BETHUNE-BAKER: A REVISION OF THE ARHOPALA DAVAONA Semper. (Plate II. fig. 12, 3.) Arhopala davaona Semper, Schmett. Philip. Ins. Band v. p. 204. Hab. Mindanao. Expanse, ¢ & 2, 37 mm. 3. Upperside : both wings exceedingly deep purplish indigo-blue, with a dead lustre in certain lights ; in a side light the blue is scarcely visible at all: primaries with the brown costa not very broad, but the nervules appearing in the blue as brown lines; outer margins very broad brown, the blue being rather sharply defined on its outer edge: secondaries with the brown costa extending nearly to the cell, and the outer margins exceedingly broad, especially at the apex and the anal angle. Underside: both wings slightly rufous brown, with scarcely darker markings palely margined. Primaries with an indistinct spot in the middle of the cell and one at the end, below which is a very indistinct one in the lower median angle; transverse band almost straight, but with an outward inclination, composed of six small confluent spots of about equal width; occasionally there is a very slight bend at the fourth spot; sub- marginal row fairly distinct; marginal row almost obsolete. Secondaries with four small basal spots below each other, the third shifted outwards, followed by four rather large spots below each other, the fourth being really the eighth spot of the transverse band ; cell closed by a quadrate spot, below which isa small one in the lower median angle ; transverse band consisting of seven spots, the second the largest directly below the first, but with its outer edge shifted outwards and touching both the spot closing the cell and the third spot of the band, which last is shifted right out on to its outer margin, fourth spot shifted outwards again, fifth spot slightly inwards, sixth slightly outwards, seventh spot very slightly angled, shifted inwards ; submarginal row distinct, marginal row less so. @. Upperside: both wings mauvish brown; primaries with a small violet patch occupying the cell, the subdiscal area, and a very little of the submedian area; secondaries with the cell only occupied by some subdued indistinct violet scaling. Underside as in the male. This species may be at once distinguished from cator? mihi by its much smaller size, the very exceptionally deep purplish blue, with yet broader borders, and by its straight transverse band on the underside of the primaries. From avatha de Nicéville the exceptionally deep and dark colour of the upperside will at once separate it, as well as the more distinct markings of the under surface. . ARHOPALA AVATHA de Nicéville. Arrhopala avatha de Nicéville, Journ. Bomb. Nat. Hist. Soc. vol. x. p. 174. n. 23, pl. T. fig. 34, ¢ (1896) ; de Nicéville and Martin, Journ. A. 8. B. vol. lxiv. pt. 2, p. 469. n. 425 (1895). Hab. Sumatra. Expanse, ¢ , 32-36 mm. AMBLYPODIA GROUP OF THE LYCANID. 97 This species is very close to moore? mihi, but may be easily separated by the darker and more purplish colour and rather broader borders of the upperside, whilst beneath the transverse band differs, and in the secondaries the submarginal row is a very pronounced feature which is quite wanting in moore? ; it is also more “spotted” than my species. From davaona Semp. it can be distinguished by the purple being less dark, its borders less wide, its size smaller, whilst beneath the transverse band of the primaries is recurved, this in davaona being nearly straight. ARHOPALA MOORE Bethune-Baker. (Plate V. figs. 10 & 10 a.) Arhopala moorei Bethune-Baker and Druce, P. Z. S. 1896, p. 669, pl. xxxi. fig. 1, @. Hab. Borneo; Malacca; Sumatra. Expanse, ¢ & 2, 38-41 mm. Further material shows me that this species will not come next to metamuta, as I had originally placed it. In de Nicéville’s collection there is a very large series from Sumatra, where it is evidently a common insect; it can easily be separated from avatha de Nicéville as already mentioned. The genitalia are peculiar, the hooks are strongly curved downwards at the base, the clasps are broadish, particularly at the base, strongly excised in the centre of the upper margin; the penis is straight for the basal two-thirds, after which it is bent to a sharp upward curve. ARHOPALA WATERSTRADTI Bethune-Baker. Arhopala waterstradti Bethune-Baker and Druce, P. Z. 8. 1896, p. 668, pl. xxx. figs. 10 3, 11 2. Hab. Borneo. Expanse, ¢ 42, 2 40 mm. This species may be separated from moore’ Bethune-Baker by its decidedly larger size, very much brighter and bluer colour, and narrower margins on the upperside of the male. In the female the colour is brighter and more silvery than in the male, whilst in moore? it is darker and more violet. I have only seen the pair sent me by Dr. Staudinger, in whose museum the types remain. ARHOPALA GUNONGENSIS Bethune-Baker. (Plate II. fig. 13, 3.) Arhopala gunongensis Bethune-Baker, Ent. Mo. Mag. vol. xxxiii. p. 205 (1897). Hab. Perak (Gunong Ijau). Expanse, ¢ 36, 2 30 mm. This species is an obscure little insect, and is known to me only by the two types in Mr. Rothschild’s collection. VOL. Xvil.— Part 1. No. 13.—Augqust, 1903. 0 98 MR. G. T. BETHUNE-BAKER: A REVISION OF THE ARHOPALA DEVA Bethune-Baker. (Plate V. figs. 11, 11 a.) Arhopala deva Bethune-Baker and Druce, P. Z. 8. 1896, p. 669, pl. xxxi. fig. 3. Hab. N. Borneo. Expanse, ¢ & 2,40 mm. Mr. Cator brought home a considerable series from Borneo, in which the females predominated largely. I am inclined to think that it replaces antimuta Feld. in North Borneo. The genitalia are very unusual; the tegumen is very ample, hooded, and excavated in the centre, the hooks are stout with a slight downward curve, the tips very slightly spatulate; the clasps are large, excavated in the centre of the upper margin and also the outer margin, so as to make the upper extremity pointed; the inner portion of the clasps near the outer margin is studded with short spikes; the penis is long, irregular, with a globular orifice, the lower lip of which protrudes. ARHOPALA ANTIMUTA Felder. Arhopala antimuta Felder, Reise Novara, Lep. vol. ii. p. 233. n. 26 (1865). Amblypodia antimuta (?) Butler, Trans. Linn. Soc. Lond., ser. 2, Zool. vol. i. p. 548. n. 8 (1877). Nee Narathura antimuta Distant, Rhop. Malay. p. 266, pl. xxii. fig. 11, 9 (1885). Arhopala antimuta de Nicéville, Butt. Ind. vol. i. p. 277 (1890). Arhopala davisonii de Nicéville, Butt. Ind. vol. ii. p. 280, frontispiece fig. 185, 3 (1890). Arhopala antimuta Druce, Proc. Zool. Soc. Lond. 1895, p. 593. Arrhopala davisonii de Nicéville and Martin, J. A. S. B. vol. Ixiv. pt. 2, p. 469 (1895). Nec Arhopala antimuta, id. loe. cit. Hab. Malacca ; Singapore Straits; Perak; Burma; Tenasserim; Borneo; Mergui. Expanse, ¢ & 9, 38-40 mm. I have carefully compared the Felderian type with all de Nicéville’s specimens of davisonii de Nicéville, and find that they are identical. From deva mihi it may be separated by its darker and bluer colour (deva is bright purplish), and the shape of the wings is different, being narrower, the apex sharper and outer margins less rounded than in my species; beneath the transverse bands are different, and my species is very spotted, the transverse bands being composed of distinct spots. From hypomuta Hew. it differs in like manner as to colour; hypomuta is bright purplish blue, with exces- sively fine linear margins; beneath Hewitson’s species is also very much more spotted in pattern, and particularly so in the transverse bands. Antimuta is apparently one of the commonest species in Sumatra, and Mr. Doherty considers it one of the most ubiquitous insects in the Malayan regions, and he says it is abundant in Borneo ; I am, however, inclined to doubt this latter statement. I have not seen a dozen specimens from that locality, and Mr. Cator, who evidently collected somewhat assiduously there, did not bring home from British North Borneo a single example ; he, however, took a number of deva mihi, and I expect this is the species to AMBLYPODIA GROUP OF THE LYCANIDA. 99 which Mr. Doherty refers. The figure given by Mr. Distant, in his work ‘ Rhopalocera Malayana’ (/. ¢.), has nothing whatever to do with this species, and is quite distinct from it. The genitalia have scimitar-shaped hooks on the tegumen, the clasps are very narrow, with the extremity bluntly pointed; the penis is very small. ARHOPALA BRAHMA Bethune-Baker. (Plate IL. fig. 14, 3.) Arhopala brahma Bethune-Baker, Ent. Mo. Mag. vol. xxxiii. p. 203 (1897). Hab. Perak. Expanse, ¢ 42, 2 45 mm. This species can be recognised from hypomuta Hew. by its larger size and purpler colour, also by the obsoletion of the under-surface pattern, and by the lack of metallic scales at the anal angle below. ARHOPALA HYPomuTA (Hewitson). (Plate II. fig. 15, ¢.) Amblypodia hypomuta Hewitson, Cat. Lye. B.M. p. 11, pl. vi. figs. 63 & 64, g (1862) ; id. Ill. Diurn. Lep., Lyc. p. 12, pl. ii. fig. 13 (1863) ; Druce, P. Z. S. Lond. 1873, p. 354; Butler, Trans. Linn. Soe. Lond., ser. 2, Zool. vol. i. p. 549 (1877). Arhopala hypomuta de Nicéville, Butt. Ind. vol. i. p. 276 (1890). Hab. India; Borneo; Penang; Malacca. Expanse, ¢ & 2, about 42 mm. g. Upperside: both wings bright blue, with very fine linear black margins, the costa of the secondaries being broadish black, the abdominal fold greyish. Underside: both wings warm ochreous brown, with darker spots palely encircled. Primaries with three increasing oval cell-spots, and a spot below the third in the lower median angle; below the second is a trace of some darker shading; transverse band composed of six spots, the first very small on the costa, second and third shifted very -slightly inwards, fourth well out, fifth and sixth inwards, but inclined obliquely outwards (Hewitson being thus right in saying “ the last spot of this fascia projects outwardly ”) ; submarginal row fairly distinct, submedian area siightly paler. Secondaries with four basal spots below each other, followed by three larger ones ; cell closed by a subquadrate spot, below which is a small one in the lower median angle, and above it a small round one shifted inwards; transverse band composed of eight spots, the second placed obliquely and shifted right outwards, third shifted also well outwards, fourth out again, fifth inwards, sixth outwards, seventh angular, confluent with the eighth long spot shifted right inwards ; this band is distinctly spotted, though the spots are not completely encircled with a pale edging ; submarginal band distinct, marginal band obscure; a black spot at the anal angle edged above by brilliant pale blue metallic scales, which also cover the two black spots preceding it. 100 MR. G. T. BETHUNE-BAKER: A REVISION OF THE @. Upperside bright violet-blue, the primaries with a broadish purplish-brown costa and very broad outer margins, the apical area being broader still; secondaries with very broad costa, the outer margin being narrow, but increasing towards the anal angle; the median and the discoidal nervules are blackish. Underside as in the male. This species has rarely, if ever, been rightly identified with Hewitson’s type and figures. In his collection are two males labelled hypomuta, one of which agrees absolutely with the figure (pl. 11. fig. 13) in the ‘Ill. Diurn. Lepidoptera,’ and very closely with the previous figures (pl. vi. figs. 63 & 64) given in the ‘ Catalogue Lycenide B.M.’; but the other specimen is quite different: there is no difficulty whatever, there- fore, in settling which is the type, and I have taken my description of the male from the first type, whilst the female described is trom Singapore. There is, however, another specimen under this name in the Hewitson cabinet, which is labelled Hypomuta amphimuta, and this no doubt has caused the confusion, more especially as the labelled female specimen belongs to this male; both are, however, absolutely distinct from Hewitson’s species, but they agree fairly well with Felder’s figures of amphimuta, and as he probably knew Felder’s insect, we may correctly assume, I think, that they are that species. ‘The figures given by Distant in his ‘ Rhopalocera Malayana’ are probably taken from these two specimens, only his artist has thought it well to add tails to both sexes, which tney do not possess; were it not for this they would fairly well represent these insects. I have been very careful with the figures here given, and there should be no difficulty in future in recognising the species. ARHOPALA METAMUTA (Hewitson). (Plate V. figs. 12 & 12a.) Amblypodia metamuta Hewitson, Ill. Diurn. Lep., Lye. p. 18. n. 59, pl. i. figs. 14 & 15, g (1863) ; Butler, Trans. Linn. Soc. Lond., ser. 2, Zool. vol. i. p. 548. n. 10 (1877). Narathura metamuta Distant, Rhop. Malay. p. 267. n. 10, pl. xxi. fig. 19, ¢ (nec fig. 18 3, var. ?) (1885); Moore, Journ. Linn. Soc. Lond., Zool. vol. xxi. p. 44: (1886). Arhopala metamuta de Nicéville, Butt. Ind. vol. ii. p. 275 (1890). Arrhopala metamuta de Nicéville and Martin, J. A. S. B. vol. Ixiv. pt. 2, p. 469. Hab. Sumatra; Malacca; Mergui; Tenasserim; Penang; Burma. Expanse, ¢ & 9, 36-42 mm. This species is recognisable at once by the different blue of the two wings, the primaries being dark purplish blue, while the secondaries are lustrous azure. The genitalia are specialised, with ample tegumen and very strong short scimitar-shaped hooks; the clasps have bilobed extremities, the outer lobe having a somewhat pointed tongue-shaped tip, and the inner lobe being spatulate, the incision being marked and moderately deep; the penis is of medium size, with a globose orifice, the lower lip of which protrudes. AMBLYPODIA GROUP OF THE LYCZENIDA. 101 ARHOPALA MUTA (Hewitson). (Plate V. figs. 13 & 15 a.) Amblypodia muta Hewitson, Cat. Lycenide B.M. p. 11. n. 50, pl. vi. figs. 57 & 58, ¢ (1862) ; id. Il. Diurn. Lep., Lye. p. 12 (1865). Hab. Java. Expanse, ¢ 31-37, 9 31 mm. The genitalia have the hooks strongly curved downwards, with the attachment on the inside of the tegumen ; the clasps are very broad at the base and at a third along the upper margin are highly excavated, the rest of the clasp being very considerably narrower. This is a species that is recognisable at a glance. ARHOPALA ALACONIA (Hewitson). Amblypodia alaconia Hewitson, Ill. Diurn. Lep., Lye. p. 14, pl. i. figs. 52 & 53 (1865). Arhopala alaconia Druce, Proc. Zool. Soc. Lond. 1895, p. 593. Hab. Borneo. Expanse, ¢ & 9, 42-44 mm. 3. Upperside: both wings dull violet, with unusually broad blackish borders all round both wings. Underside: both wings dark and olivaceous brown, with darker spots edged with cream-colour. Primaries with three increasing cell-spots ; below the second and third are two others divided by the lower median nervule; transverse band composed of six spots, the first much compressed, third to sixth almost confluent, shifted outwards, and having a slight oblique outward tendency; the submarginal row distinct to the lower median nervule, a patch of creamy scaling between the upper three spots of the transverse band and the apex. Secondaries with spot at the basal extremity of the costa and four basal spots, the second and fourth shifted inwards, tollowed by four large ones below each other, the first and fourth of which touch the previous row; cell closed by an irregular spot; transverse band composed of eight spots, the first two excessively large, confluent, and irregular, third and fourth shifted outwards, fifth smaller inwards, sixth outwards, seventh subangular, confluent with the small eighth; submarginal row distinct sublunular, with the second to fourth marks very dark; a very small metallic-blue spot near the anal angle and a dark spot at the angle. The male and female of this species are precisely like each other. Oberthuri Ster. is its nearest ally, but it may at once be separated by the difference in colour of the upperside, and the underside of the secondaries is much more uniform than is Staudinger’s insect. ARHOPALA OBERTHURI (Staudinger). (Plate II. fig. 16, 3.) Amblypodia oberthuri Staudinger, Iris, vol. i. p. 132, pl. 11. fig. 4 (1889). Hab. Palawan; Burma; Upper Tenasserim. Expanse, ¢ 37-38, 2 40 mm. 102 MR. G. T. BETHUNE-BAKER: A REVISION OF THE I have before me, from the collection of Mr. de Nicéville, three specimens of this species from Tounghas (Central Burma), and one from the Dannat Range, Upper Tenasserim; they differ somewhat in that they are much greyer in the colour of the underside, and are also slightly smaller, but otherwise there is no difference whatever. From alesia Felder it differs in that it lacks a tail, and that its colour above is violet- blue, not pale azure as in Felder’s insect, whilst beneath it is much more uniform in general tone. ARHOPALA ALESIA Felder. Arhopala alesia Felder, Reise Novara, Lep. p. 235. n. 263, tab. xxix. fig. 18 (1864). Nilasera wimberleyi de Nicéville, Proc. Zool. Soc. Lond. 1887, p. 462, pl. xl. fig. 4, 2. Arhopala wimberleyi de Nicéville, Butt. Ind. vol. ii. p. 258 (1890). Hab. Philippine Islands ; South Andaman Isles. Expanse, ¢ & 2, 38 mm. This species has a more variegated appearance beneath than oberthuri Stgr., the dark spots being darker, and the ground-colour not quite so dark, whilst the pale edging to the spots is much more marked than in Staudinger’s insect, and it has also a good many pale scales scattered more or less over its surface. De Nicéville’s wimberleyt falls as a synonym to this. ARHOPALA OCRIDA Hewitson. Arhopala ocrida Hewitson, Il. Diurn. Lep., Lycenide, p. 146, pl. itia. figs. 88 & 39 (1865). Hab. Philippines. Expanse, ¢ & 2, 39-42 mm. On the upperside this species is almost exactly like alesia Feld., but is of a rather darker blue colour, while beneath the differences are greater. Ocrida is greyer and much more uniform; it has none of the light patches of pale scales, and its pattern is smaller and not confluent, as is the case in the secondaries of alesta ; there is also a distinct black lobe-spot and some metallic scaling in its neighbourhood, which is almost entirely (though not absolutely) wanting in Felder’s insect. Hewitson says “this species and A. corinda scarcely differ on the underside.” This is not quite accurate; there may be just at first sight a superficial resemblance, but it is exceedingly superficial. The two species are not really at all difficult to separate. ARHOPALA ADONIAS (Hewitson). (Plate IT. fig. 17, ¢.) Amblypodia adonias Hewitson, Cat. Lycenide B.M. p. 8, pl. iv. (2?) figs. 82, 33 (1862); id. Il. Diurn. Lep., Lye. p. 9 (1865). Amblypodia grynea Hewitson, Ill. Diurn. Lep., Lye. p. 20, pl. (suppl.) vii. ( ¢ ?) figs. 59 & 60 (1878). Hab. Java; Philippine Isles. Expanse, ¢ & 2, 42 mm. (3) AMBLYPODIA GROUP OF THE LYCANIDA, 103 3. Upperside: both wings brilliant shining metallic emerald-green, with a slightly brassy tint; primaries with costa narrowly and posterior margin somewhat broadly blackish brown ; secondaries with very broad brown margins all round, especially in the anal region ; tail brown, white-tipped. Underside greyish brown, with dark spots palely encircled. Primaries with three increasing spots in the cell; beneath the second and third two more spots separated by the lower median nervule; transverse band composed of seven spots and fractured below the fourth—first spot small, second and third shifted outwards, fourth further out, fifth spot shifted right inwards, followed by two rather long oblique spots ending on the submedian nervure; submarginal shading obscure; submedian area paler. Secondaries with a small spot at the costal base ; four basal spots, the fourth shifted well inwards, followed by three spots, the lowest of which is margined laterally only and is angulated in the centre; cell closed by a longish narrow spot margined laterally only, beneath which isa small triangular spot ; transverse band composed of eight spots margined laterally, the second being the largest and touching the spot closing the cell—the third spot is shifted outwards, but not detached, fourth further out, fifth well inwards, sixth outwards, seventh angular spot well inwards and touching the eighth, which is often almost broken into two spots; submarginal shading very obscure; a dark spot at the anal angle, and a trace of one between the second and lower median nervules, the space between and over which is sparingly filled with green-metallic scales, over which is a pale grey dusting. @. Upperside: both wings brilliant shining pale azure-blue ; primaries with costa broadly blackish brown and increasing to the apex, posterior margin very broadly blackish brown ; secondaries with very broad brown margins all round. Underside as in the male, except that the ground-colour is paler and yellower in the tint of brown. Hewitson described the female only. Herr Georg Semper has been kind enough to lend me many of his specimens, and fortunately among them is one male of this species from Java; it is precisely of the same pattern on the underside as the female of adonias. Hewitson described an insect as grynea likewise from a female, but there is absolutely no difference between the two species. The female of this species is very close to ocrida Hew., the uppersides of the females of these two species being almost identical, though the male has a greater expanse of blue in the secondaries: but beneath the difference is greater; the ground-colour of ocrida is paler and the. markings more prominent, and the two transverse bands different; in the primaries this band is not completely dislocated as in adonias, and its shape in the secondaries will be readily differentiated from the description. I have one female specimen from Mindanao, which I believe to be this insect, otherwise the species appears to be confined to the island of Java. 104 MR. G. T. BETHUNE-BAKER: A REVISION OF THE ARHOPALA EUMOLPHUS (Cramer). (Plate II. figs. 19, 20, 21,22 ¢ g, 23& 2429; Plate V. figs. 14 & 14 a.) Papilio eumolphus Cramer, Pap. Ex. vol. iv. p. 19, pl. ecxcix. figs. G, H, ¢ (1780) ; Herbst, Pap. pl. cexeviii. figs. 3 & 4, g (1804). Polyommatus eumolphus Godart, Enc. Méth. vol. ix. p, 652. n. 120 (1823). Amblypodia eumolphus Horsfield, Cat. Lep. Mus. E.I.C. p. 103. n. 35 (1829) ; Hewitson, Cat. Lycenide B.M. p. 8. n. 36, pl. vill. fig. 89, ¢ (1862). Amblypodia bupola Hewitson, Ill. Diurn. Lep., Suppl. p. 21. n. 102, pl. (suppl.) vii. figs. 64 & 65. ? (1878). Saiadra bupola Moore, Journ. A. 8. B. vol. liii. pt. 2, p. 38 (1884); de Nicéville, Journ, A. 8. B. vol. liv. pt. 2, p. 4. n. 302 (1885). Narathura farquhari Distant, Rhop. Malay. p. 264. n. 5, pl. xxiii. fig. 3, ¢ (1885). Amblypodia eumolphus Butler (nec Cramer), Trans. Linn. Soc. Lond., ser. 2, Zool. vol. 1. p. 548. n. 2 (1877). Arhopala farquhari Doherty, Journ. A. 8. B. vol. lviii. pt. 2, p. 421 (1889). Arhopala hellenore Doherty, Journ. A. 8. B. vol. lviii. pt. 2, p. 422. n. 18, pl. xxi. fig. 7, (1889). Amblypodia eumolphus de Nicéville, Butt. Ind. vol. ii. p. 263. n. 822 (1890). Amblypodia farquhari id. op. cit. p. 264. Amblypodia hellenore id. op. cit. p. 265. Narathura maxwelli Distant, Rhop. Malay. p. 263. n. 4, pl. xxiii. fig. 10, 2 (1885). Hab. Bengal coast; Java; Sikkim; Nepal; Assam; Chittagong Hill-tracts ; Burma; Malay Peninsula; Sumatra; S.E. Borneo; Palawan (Philippine Islands) ; New Guinea (Staudinger’s coll.). Expanse, ¢ 41-55, 2 43-51 mm. 2. Upperside: both wings deep emerald-green, with a strong brassy lustre in certain lights; primaries with a narrow brown costa and broad brown outer margin, this latter varies in width, and is sometimes very broad; secondaries with very broad brown margins all round; tail brown, tipped with white; lobe slightly developed. Underside: both wings palish ochreous brown, with darker brown markings encircled with whitish. Primaries with three increasing cell-spots; above the third is another small one near the costa varying somewhat in size, below it (the third) is a large spot in the lower median angle, subdiscal area dark; transverse fascia consisting of seven spots from the costa to the submedian nervure, the first three spots outwardly oblique, fourth spot slightly dislocated and shifted yet further out, fifth strongly dislocated inwards, fifth and sixth spots inclined very slightly outwards, the lowest being shifted slightly inwards, and sometimes it is quite small, not nearly touching the submedian nervure ; submarginal and marginal row narrowish in an almost continuous line ; submedian inter- space pale. Secondaries with a small dot at the extreme costal base ; four small basal spots followed by three larger paler spots below each other; cell closed by a long subquadrate spot, below which is a smaller one connecting it with the lowest of the AMBLYPODIA GROUP OF THE LYCAHNID. 105 three spots; transverse band composed of eight spots, the second much larger, and directly below the first and touching the cell-spot, the third shifted well outwards, but not detached, fourth shifted outwards again, fifth inwards, sixth outwards, seventh angular spot well inwards and almost confluent with the eighth long spot; submarginal row crescentic, fairly distinct ; marginal row indefinite, ending in a darkish spot on the anal lobe; no metallic scales at the anal angle. Tt will be well to describe hellenore Doherty and farguhari Distant, and show how they merge one into the another. A. hellenore Doh.—Upperside similar in colour to that of ewmolphus Cram., but with the posterior margin of the primaries black, narrow at the apex, increasing in width to the anal angle, where it is broad. Secondaries with rather more green than in the case of ewmolphus. On the underside the pattern is precisely the same as in Cramer’s species, but the ground-colour is variegated with pale patches and is dark greyish ; the lobe-spot is deep black and is preceded by green-metallic scales, which are often inclined to obsolescence. A. farquhari Distant.—Upperside similar in colour to that of ewmolphus Cramer, but with a rather narrow black posterior margin to the primaries; secondaries with rather more green than those of ewmolphus. Underside: both wings uniform dull ochreous brown, with rather darker spots palely encircled, the spots being of the same tone as the ground-colour but darker. In pattern this is similar to Cramer’s species, but it has a somewhat more “spotted” appearance, and the anal angle is very strongly irrorated with brilliant bluish-green metallic scales, with a dark spot at the anal angle and beyond the tail. The female is even more uniform in colour than the male. If we take specimens of the extreme of each form they appear satisfactorily distinct, and it is evident that this has unfortunately been done. But in examining a large series, connecting-links are shown in every grade, so that it would be impossible to say where the dividing-line should come. Hellenove Doherty is the extreme form in the one direction, with its rather more angular wing and straighter outer margin, the black border of which is quite narrow at the apex, but increases in a very defined straight line to the anal angle where it is broad, whilst beneath it is dark greyish, variegated with pale grey or bronzy-grey patches. Between this and the originally- described ewmolphus Cramer I have before me links of every grade, forming an absolute connecting-series between the two (there are about sixty specimens before me now, and if the two can be connected with so small a number, what would it be with a really large series ?). The same remark applies to the upperside, for the width of the dark border varies in a remarkable way, being in some specimens quite an eighth of an inch wide in the narrowest part, whilst in others it is very narrow. At first sight farqhari seems to be much more distinct : the outer margin of the primaries is somewhat produced, thus giving the wings a more rounded shape ; whilst the underside in extreme VOL. XviI.—PartT 1. No. 14.—August, 1903. P 106 MR. G. T. BETHUNE-BAKER: A REVISION OF THE specimens is different, the colour being very uniform darkish ochreous brown with the pattern more spotted in appearance. I have about thirty specimens, more or less typical, before me, in addition to which I have about a dozen connecting them with eumolphus Cramer, almost as closely as in the preceding instance, and one or two of the latter approaching the grey colour of the underside of farquhari Distant : the connecting- specimens are from Java, Palawan, New Guinea, Borneo, and one from Darjeeling ; the typical specimens before me being from Sumatra, Borneo (S8.E.), and New Guinea. I have not yet seen a typical ewmolphus from the Malay Islands, but I have a link very close indeed to farquhari Distant from Darjeeling. Under these circum- stances, and considering the fact that I can trace no difference whatever in the genitalia, I am forced to the conclusion that these three names represent one variable species, the female of which may be occasionally dimorphic, as in other species of this green group, for I have one or two very pale azure females which might be claimed by each of the authors as the female of his insect. De Nicéville pointed out (Butt. India, vol. iii. p. 265) that mazwelli Distant was probably the female of farquhari Distant ; and he is no doubt quite correct in this, for if we compare the underside of the figures of the two insects, they will be found to agree precisely. I have mounted in balsam several genitalia of each of the named varieties, and find they all agree precisely, and in this genus I consider this potent evidence that all are but one species. It is most probable that these very highly metallic-green colours and other such-like hues are a comparatively recent development, and we shall therefore not expect to find the female of a similar colour. I have no doubt, as de Nicéville has already pointed out, that the supposed female described by Horsfield is no other than Pagenstecher’s more recently described horsfieldi. The genitalia of this green group are very specialised, but all follow the same main lines: the tegumen is much less ample and distinctly angled where it resolves itself into its arm-like supports; the hooks are sharply bent at their muscular attachment inside the tegumen, whence they project horizontally in a slight curve; the clasps are bilobed, the upper lobe being anvil-shaped, the upper fore extremity being sharply pointed and very highly excavated internally, whence it is projected into the long finger-shaped lower lobe, which has an upward curve. The penis is very long, with a contrary angle fore and aft, with a large trumpet-shaped orifice much expanded on its outer lip. ARHOPALA AUREA (Hewitson). Amblypodia aurea Hewitson, Cat. Lycenide B.M. p. 8, pl. viii. figs. 87 & 88 (1862); id. Il. Diurn. Lep., Lyc. p. 9 (1863). Arhopala aurea Druce, Proc, Zool. Soc. Lond. 1895, p. 591. Hab. Borneo (Sarawak, Labuan, Tameang-Lajang) ; Sapagaya. Expanse, ¢, about 45 mm. AMBLYPODIA GROUP OF THE LYCANIDA. 107 This species may at once be recognised from the preceding by its very much smaller size, by its brassy-green colour, and by the quite different pattern of the underside. _ The female is, I believe, entirely unknown. It is apparently a rare species, being either not represented, or, if represented, in most collections it is only by two or three specimens. Mr. Cator brought home from the small island of Sapagaya more specimens of this butterfly than I had seen before in all the collections I had been through, yet there was not a single female among them. ARHOPALA BORNEENSIS Bethune-Baker. Arhopala borneensis Bethune-Baker and Druce, Proc. Zool. Soc. Lond. 1896, p. 666, pl. xxx. figev5;, ay. Hab. Borneo (Kina Balu, Tameang-Lajang) ; Malacca. Expanse, 3, 42-46 mm. From aurea Hew. this species may be recognised by the deeper tone of the green, by the decidedly paler and more uniform colour of the underside, the pattern being likewise paler, and by the pattern itself. In the primaries the transverse band is straighter and wider, in aurea it is rather sharply, angular, whilst in the secondaries this band is decidedly obscure. ‘The female of this species also is at present unknown to me. ARHOPALA TROGON Distant. Arhopala trogon Distant, Ann. & Mag. Nat. Hist. ser. 5, vol. xiv. p. 201 (1884) ; id. Rhop. Malay p. 275, woodcut 87, g (1885), and p. 464 (1886). Arhopala trogon de Nicéville and Martin, J. A. S. B. vol. lxiv. pt. 2, p. 467 (1895). Hab. Perak; Sumatra. Expanse, ¢ 40, 2 37-38 mm. This beautiful species may immediately be separated from its near allies by the purple tinge of its underside, and by the large and prominent pattern thereof. Like some others of this green group, it is apparently very local and very rare. ARHOPALA HORSFIELDI Pagenstecher. (Plate IT. fig. 26, 3.) Amblypodia horsfieldi Pagenstecher, Jahr. des Nass. Ber. fiir Naturk. vol. xii. pp. 96, 106 (1890). Nec Arhopala horsfieldi de Nicéville and Martin, J. A. S. B. vol. Ixiv. pt. 2, p. 467. n. 411 (1895). Hab. Java. ; Expanse, ¢ & 2, 39-40 mm. 3. Upperside: both wings dark brown, with a large bronzy metallic-green patch in the primaries from about the costal nervure to the internal margin up to half the median area, but ending abruptly at the close of the cell; in the secondaries this . metallic patch is very small and does not much exceed the cell itself; tail brown, white- tipped. Underside: both wings brown, very slightly lustrous, with dark spots palely P2 108 MR. G. T. BETHUNE-BAKER: A REVISION OF THE edged. Primaries with three increasing discal spots; below the second and third are two others divided by the lower median nervule, and above the third is a very small one on the costa; transverse fascia irregular, first spot small, second and third confluent, inclined outwards, fourth shifted well outwards, fifth and sixth confluent and shifted right inwards, but not detached, seventh shifted slightly inwards and_ oblique, sometimes quite small ; submarginal row distinct, but with obscure margins ; submedian area slightly paler. Secondaries with an obscure spot at the costal base; four basal spots, the fourth shifted inwards on to the inner margin, followed by three spots below each other; cell closed by a longish narrow spot, beneath which is a small triangular one; transverse fascia composed of eight spots, the second larger than the first and touching the upper outer extremity of the narrow spot closing the cell, third spot shifted right outwards but not detached, fourth further out, fifth inwards, sixth outwards, seventh angular spot inwards and touching the eighth long one; submarginal row rather obscure; lobe-spot black; a small black spot is also between the second and lower median nervules, the space between and over which is filled with bluish or greenish-metallic scales. 2. Exactly like the male, except that the green is replaced by a violet-blue patch on the upperside of both wings, and the underside is somewhat lighter. This species was originally described by Horsfield as the female of ewmolphus Cram. It can readily be separated from all others of this group, except basiviridis de N., by the green colour being limited to less than two-thirds of the primaries and to less than one-third of the secondaries. It is much closer to de Nicéville’s insect, in the description of which species the comparison will be found : horsfieldi appears to be entirely confined to the island of Java, and it is interesting to notice how very local all these green species are with the exception of ewmolphus. ARHOPALA BASIVIRIDIS de Nicéville. (Plate IJ. fig. 27, ¢; Plate V. figs. 15 & 15a.) Arhopala basiviridis de Nicéville, J. Bombay N. H. Soc. 1891, p. 373, pl. G. fig. 22, ¢. Arhopala horsfieldi, Druce, Proc. Zool. Soc. Lond. 1895, p.591 ; de Nicéville and Martin, J. A. S. B. vol. lxiv. pt. 2, p. 467 (1895). Hab. Borneo, Malacca, Burma. Expanse, ¢ , 42-44 mm. 3. Upperside: both wings dark brown, with a bluish-green metallic patch occupying the cell to the upper margin and one-third of the median area of the primaries. In the secondaries this metallic patch extends from just above the cell to the submedian nervule, but scarcely beyond the cell as to the outer margin. ‘Tail brown, white-tipped. Underside: both wings of a warm brown, with slightly darker spots and fascia palely margined. Primaries with three good-sized discal spots increasing in size; beneath the second and third are two other large spots separated by the lower median nervure, and above the third a small obscure spot; transverse band irregular, composed of AMBLYPODIA GROUP OF THE LYCANIDA., 109 seven spots, usually terminating on the submedian nervure, second spot larger and shifted very slightly outwards, third further out, fourth still further out, fifth inwards and confluent with the sixth, which is outwardly oblique, seventh spot shifted inwards and sometimes obscure ; submarginal row distinct and continuous, marginal row obscure ; submedian area paler. Secondaries darker in hue, with an obscure spot at the costal base, three dark basal spots, below which on the inner margin is a rather paler one, following these are three larger spots below each other; cell closed by a longish subquadrate spot, below which is a triangular one; transverse band composed of eight spots, the second larger than the first and touching the spot closing the cell, third spot detached from the second (usually but not always) and shifted right outwards, fourth shifted further out and confluent with the fifth, sixth shifted outwards, seventh spot well inwards and touching the long eighth spot; submarginal row fairly distinct. Three anal black spots, the middle one being almost entirely obliterated, with bluish or greenish-blue metallic scales, which also edge the other two spots above. All the spots are rather large and the fascia edged laterally. 2. Upperside: both wings rather pale, sublustrous, violet-blue, rather bright; cell closed by a black spot; all the borders of the primaries broadly brown, apical area very broad ; secondaries with costa and outer margin very broadly brown, decreasing at the anal angle, abdominal fold greyish. Underside exactly like that of the male in every respect. Mr. Druce has a male and female from Burma, taken respectively at Bankasoon and Meetan; the female has the blue of an exceedingly pale lustrous azure instead of the usual violet colour. This species is distinct from horsfieldi Pagenst. ; it is altogether a larger insect than the latter; the colour beneath is much browner and warmer in tone, making the spots much less prominent, though they are larger than in horsfieldi; the transverse fascia has a much more spotted appearance. The points of difference are difficult to describe, but are such as to separate the two species easily. The ground-colour of horsfieldi is much greyer in tone. In the female the blue of de Nicéville’s insect occupies a decidedly larger area in both wings and is brighter and rather lustrous. The figures will show the differences quite readily. The genitalia are similar to ewmolphus Cram., but the hooks are straight, and the anvil-shaped part of the clasp is deeply excavated fore and aft. ARHOPALA STAUDINGERI Semper. (Plate II. fig. 25, ¢ .) Arhopala staudingeri Semper, Schmett. Philip. Ins. Bd. v. p. 195. Hab. Mindanao. Expanse, ¢, 45 mm. ¢. Upperside: both wings brilliant metallic green; primaries with costa narrowly dark brown, posterior margin rather broadly brown, increasing considerably at the anal 110 MR. G. T. BETHUNE-BAKER: A REVISION OF THE angle ; secondaries with very broad brown margins all round. Underside : both wings brown, with spots and fascia edged with white. Primaries with three large increasing discal spots ; in the submedian area below the first two are two large confluent blotches making the whole basal area dark, with a pale mark showing the juncture in the middle ; below the third is another spot, and above it (the third, which is very irregular and large) are two small spots, one touching it, the other on the costa ; transverse band very irregular and very broad, touching the third spot and that below it and ending on the submedian nervure ; the first spot on the costa is small, the second and third each larger, fourth larger still and shifted outwards, fifth and sixth larger still and each shifted inwards, the sixth only as to its inner margin, seventh spot smaller, but still smaller at its lower extremity, area between this and the dark basal spots whitish; submarginal row very finely and obscurely edged with whitish; extreme margin white with an interior dark spotted row; internal area whitish grey. Secondaries with a spot at the costal base, four basal spots, the upper three touching a very large dark patch, occupying the central and upper part of the wing from the costal to the median nervures; the spots in the centre of the cell and closing it marking its lower extremity at the costal nervure, a white mark showing that the spots have become confluent; transverse band beginning on the lower discoidal nervure and touching the large patch; the second oval spot is shifted inwards, third outwards, fourth very large angular spot inwards and confluent with the fifth long spot; subapical area whitish grey ; submarginal row indefinite, the three lower spots being very dark with obscure internal definitions; lobe fairly developed, with a black spot followed by two subanal black spots, the middle one being almost obliterated with bright blue-metallic scales, which also edge the other two spots above ; tail brown, white-tipped, over the metallic spots is a greyish irroration ; margin finely whitish. I have only one male before me, but it appears to be quite a distinct species, the pattern beneath being very distinctive. ‘The very striking and handsome underside of this species should separate it without the least difficulty from all its allies. ARHOPALA ANELLA de N. Arhopala anella de Nicéville, Journ. Bombay Nat. Hist. Soc. vol. ix. p. 289, pL Ps fiesrs2 (1895). Hab. Perak, Malay Peninsula. Expanse, 2, 51 mm. The unique type specimen is before me and is certainly a distinct species, its nearest ally being irregularis Rober, also a three-tailed species; it has, however, a much larger area of blue on the upperside, whilst beneath the colour is very much paler and ochreous, and the spots fewer and more separated than in Réber’s species. AMBLYPODIA GROUP OF THE LYCAHNIDA, Hu ARHOPALA IRREGULARIS, n. sp. (Plate II. fig. 18, 9 .) Hab. Celebes, Banka. Expanse, ¢ 38-43, @ 35-38 mm. 3. Upperside: both wings very deep, slightly shining, violet-blue; the primaries with costa broadly, apical and posterior marginal areas very broadly purplish black ; secondaries with very broad margins all round, anteciliary line black, edged internally from the upper median nervule to the anal angle with white; fringes to both wings tipped with white. ‘Three tails fine brown, each white-tipped, central one very long. Underside: both wings greyish brown, with very dark spots edged with white. Primaries with a small spot at the extreme base of the cell, followed by two larger roundish ones, the third well before the end of the cell ; two small spots are over the second and in advance of it, the upper one touching the costa, one spot is over the third and in advance of it, and touches a very broad irregular fascia which crosses the end of the cell; this fascia is narrower on the costa and terminates very widely on the lower median nervule, and in its centre is a faint white dash indicating the end of the cell. ‘Trans- verse fascia very irregular, edged laterally only; the second spot with its outer margin shifted outwards, third with the inner margin shifted inwards, fourth outwards and angulated in the centre of its inner margin, fifth right inwards, with its outer margin slanting outwards and its inner very oblique and slightly angulated, sixth spot shifted slightly inwards and oblique; below this under the lower median nervule is an indefi- nite small oblique dash ; submarginal row distinct, the lower spots very much suffused inwardly and large, the three upper ones being quite >-shaped and much suffused inwardly with whitish ; marginal stripe dark almost obsolete in centre; anteciliary line dark, edged internally with whitish; submedian area greyish white, underneath the base of the cell! slightly darker. Secondaries darker than the primaries, with a good- sized spot at the costal base; a basal row of five spots all touching, the third and fourth shifted outwards, the fifth beyond the internal nervure unusually large, followed by three large spots, the upper one pear-shaped ; cell closed by a large spot with a pale central dash, below which is a narrow suboval spot: transverse band very irregular and edged laterally only, the upper spot having a double curve, followed by a larger one whose lower inner margin touches the large spot closing the cell; third spot shifted right outwards, fourth outwards again, the fifth spot oblique, shifted inwards, sixth merely a very dark narrow stripe, seventh and eighth also confluent, shifted right inwards and little more than a very dark narrow stripe ; outside the three upper spots there is a very dark suffusion up to the pale edging of the narrow scalloped submarginal stripe, which stripe is continued from the costa right round above the metallic scaling to the internal nervure; a black spot at the anal angle and between the second and lower median nervules, the space between and over which is broadly filled with brilliant bluish or greenish-bronzy metallic scales, which also extend into the upper median space; anteciliary line very dark, edged internally with whitish. 112 MR. G. T. BETHUNE-BAKER: A REVISION OF THE 2. Upperside: both wings warm brown ; primaries tinged with a shade of purple, with a good-sized bright violet patch up to half the median area; cell closed by a dark spot, this patch being smaller in the secondaries. I have before me one specimen from Banka in which the shade is as deep in tone as in the male, but it is more violet in hue. This species is very strikingly marked on its underside and should be recognised at once; it is perhaps nearest to abseus Hew., but can be very easily separated by its very much larger size, by the unusually broad borders of the male and the very dark blue of the upperside, whilst on the underside the spots are very much more distinct and are markedly outlined with whitish, whereas adseus is more uniform in tone and darkly clouded. The specimen figured is a female in my own collection; it would have been better had I figured the male, which is much more distinct, but I had returned the male specimens to Herr Ribbe before I was ready to figure it. ARHOPALA ABSEUS (Hewitson). Amblypodia abseus Hewitson, Cat. Lycenide B.M. p. 9.n. 40, pl. v. figs. 41 & 42 (if corrected in MS., then figs. 51 &52), 9 (1862); id. Il. Diurn. Lep., Lye. p. 10. n. 44. (1863) ; Druce, Proc. Zool. Soc. Lond. 1873, p. 353. un. 6. Satadra abseus Moore, Journ. A. S. B. vol. lili. pt. 2, p. 41 (1884). Arhopala abseus de Nicéville, Butt. India, vol. iii. p. 242. Arhopala amphea Felder, Reise Novara, Lep. p. 234, tab. xxix. fig. 19 (1865); Druce, Proc. Zool. Soe. Lond. 1895, p. 589; de Nicéville and Martin, J. A. 8. B. vol. Ixiv. pt. 2, p. 465 (1895). Hab. Sylhet ; Sikkim; Burma; Mergui; Singapore; Sumatra; Borneo. Expanse, ¢ 9, 36-41 mm. ¢. Blue areas distinctly fractured beyond the lower median nervule, where the fascia becomes very narrow ; below the upper discoidal nervule is a dark patch extending to the margin, submarginal row obscured. A black spot at the anal angle and between the second and lower median nervules, over and between which are some greenish sub- metallic scales more or less scattered, margined above with a dark dash up to the abdominal margin. The secondaries are glossy and the spots much darker than in the primaries. 2. Exactly like the male, except that the blue is paler and more restricted. All the male specimens from Borneo that I have seen are quite glossy purple. The species is very different from any others except irregularis and should be recognised without difficulty. ARHOPALA AMPHEA Felder. Hab. Philippines. Expanse, ¢ 38-45, 9 33-44 mm. This is, without doubt, a local race of abseus, being merely of a darker blue above, with the borders not so broad. AMBLYPODIA GROUP OF THE LYCANIDZ. 113 ARHOPALA KUNI (Rober). (Plate II. figs. 28 g, 29 2.) Amblypodia kuhni Rober, Iris, vol. i. p. 200, Taf. ix. figs. 5 & 8 (1887). Hab. Isle of Banka. Expanse, ¢ 47, 2 38 mm. This species should at once be recognised by the female, which is peculiar in the dark spots of the upperside closing the cell on both wings, which are specially noticeable on the secondaries, and by the three tails. The male is of a violet shade, not purple above, and is easily recognisable from diardi Hew. by the compact dark basal area and by having no purplish tinge below, and from fulgida by a similar difference in colour, and by its decidedly larger size, whilst the contour of its primaries differs from both, the outer margin of the apex being nearly straight, with a slight excavation in the median area. ARHOPALA DIARDI (Hewitson). (Plate V. figs. 16 & 16a.) Amblypodia diardi Hewitson, Cat. Lycenide B.M. p. 9. n. 43, pl. v. figs. 51 & 52 (if corrected in MS. then figs. 41 & 42), 3 (1862) ; Druce, Proc. Zool. Soc. Lond. 1874, p. 107. n. 2. Amblypodia? diardi Butler, Trans. Linn. Soc. Lond., ser. 2, Zool. vol. i. p. 549. n. 13 (1877). Satadra diardi Moore, Journ. A. 8. B. vol. liii. pt. 2, p. 41 (1884). Panchala diardi Distant, Rhop. Malay. p. 272. n. 1, pl. xxiii. fig. 14, ¢ (1885). Arhopala diardi de Nicéville, Butt. India, vol. iii. p. 256 (1890); de Nicéville and Martin, J. A. S. B. vol. lxiv. pt. 2, p. 466 (1895). Amblypodia capeta Hewitson, Il. Diurn. Lep., Suppl. p. 22, pl. viii. figs. 70 & 71 (1878). Amblypodia viardi Staudinger, Iris, vol. ii. p. 180 (1889). Hab. India; Assam; Siam; Penang; Malacca; Singapore; Sumatra; Palawan; Bosobosa (Philippines) ; Mindanao; Java; Borneo; Celebes. Expanse, ¢ 2, 40-48 mm. I have before me a specimen of this insect from Palawan of a most beautiful rich blue, almost as beautiful as anniel/a, but with a purple tinge in it; the underside, however, agrees entirely with this species, it has but little of the red at the base. Ihave compared Hewitson’s capeta very carefully with this species and can find no valid points of difference. It is very close to the next species, fulgida Hew., but is decidedly darker and more violet in the tone of the underside, and has the very finest possible dark linear outer margin and costa, whereas fulgida has both these margins decidedly wider though still narrow. Its distribution is extensive, occurring from the Khasia Hills (Assam) to the Philippine Islands on the east, to Sumatra on the west, into Java, where apparently it is not uncommon, and probably in suitable places throughout Borneo, as I have specimens from the south-east of the island and from the coast opposite Labuan. I have before me one extraordinary male from Lompa-Battan (Celebes), taken about 3000 feet high in March 1896, in which the underside is extremely dark; whilst it VOL, XVII.—Part I. No. 15.—August, 1903. Q 114 MR. G. T. BETHUNE-BAKER: A REVISION OF THE measures in expanse no less than 66 mm. In the genitalia the tegumen is very simple with hardly any hood and the sides very broad almost to the juncture with the clasps ; the hooks are long, angled almost at 45° about their centre, and fixed very close to the top of the tegumen; the clasps are small; the penis is rather like that organ in the eumolphus group, but much shorter, not having the narrow internal portion as in that section. ARHOPALA FULGIDA (Hewitson). Amblypodia fulgida Hewitson, Ill. Diurn. Lep., Lycenide, p. 11. n. 49, pl. v. fig. 31, 2 (1863). Nilasera? fulgida de Nicéville, Journ. A. 8. B. vol. lii. pt. 2, p. 80. n. 21, pl. ix. figs. 3 g, 3a 2 (1883). Satadra fulgida Moore, Journ. A. S. B. vol. liii. pt. 2, p. 41 (1884). Arhopala fulgida de Nicéville, Butt. India, vol. iii. p. 254 (1890). Panchala singapura Distant, Rhop. Malay. p. 273, woodcut 84 (1885). Hab. Philippines; N. India; Sikkim; Khasia Hills; Singapore. Expanse, ¢ 46-49, 9 45-46 mm. This species may be safely recognised from the preceding, A. diardi Hew., by the slight differences of the upperside as already enumerated under that species, but more especially by the fact that the subbasal broad dark band is always unbroken. A. fulgida has a much more restricted range than diardi. The genitalia are of just the same type as in diardi Hew., but the hooks are angled at a third from their attachment and the clasps are much more pointed. ARHOPALA ANNIELLA (Hewitson). Amblypodia anniella Hewitson, Cat. Lycenide B.M. p. 10, pl. viii. figs. 83 & 84, g (1862). Narathura anniella Distant, Rhop. Malay. p. 269, pl. xxi. fig. 20 (1885). Arhopala anniella Druce, Proc. Zool. Soc. Lond. 1895, p. 592; de Nicéville and Martin, J. A. S. B. vol. Ixiv. pt. 2, p. 468. Arhopala triangularis Stgr. (in litt.). Hab. Singapore; Province Wellesley; Perak; Sungei Ujong; Borneo; Batchian (Semper); Java; Philippine Islands; Sumatra; Tenasserim. Expanse, ¢ 40, 2 38-40 mm. Mr. Distant has severely strictured Hewitson’s figure of this species, yet his own is but little better; in shape it is worse than the original figure ; the costa and posterior margin of the primaries should be much straighter, especially the former, and the tip of the apex slightly produced, whilst the secondaries are absolutely wrong in shape and should be as described already, instead of rounded as figured by Mr. Distant. The rich beautiful blue of the male will separate it from any of its near neighbours, whilst below it is much more uniform in tone, and the secondaries are very dark rich purplish red, washed here and there with very pale lavender scales, and it has no AMBLYPODIA GROUP OF THE LYCANID. 115 distinct tail, only a very slight tooth. The specimens, however, from the Philippines and Batchian and Malacca have a slight tail and are not so uniformly dark as those from Borneo on the underside of the secondaries. ‘These specimens seem to form a sort of “transit” to fulgida Hew., as the upperside is not so intensely blue as in the Bornean specimens. It appears to be a local and rare insect, as only 19 specimens have passed through my hands, and I have not seen many in any collection, except that of the late Mr. de Nicéville, which contained 12, and added to our knowledge of the distribution of the species, there being specimens from Perak, Sumatra, Penang; and Tenasserim (Dannat Range). ARHOPALA ARTEGAL (Doherty). Flos artegal Doherty, J. A. S. B. vol. lvii. pt. 2, p. 423, pl. xxiii. fig. 5 (1889). Arhopala artegal de Nicéville, Butt. Ind. vol. iii. p. 225 (1890). Hab. Mergui. Expanse, ¢, 35 mm. I have not seen this species, but from the figure it is very close to anniella Hewitson- Mr. Doherty states that the two specimens he took were males; were it not that he is so accurate an observer I should doubt this, and if one can judge from the figure there is every appearance of a female in the shape of the wings. It is not impossible that these may prove to be females of Hewitson’s previous species. ARHOPALA APIDANUS (Cramer). (Plate V. figs. 17 & 17 a.) Papilio apidanus Cramer, Pap. Ex. vol. ii. pl. exxxvii. figs. F & G, § (1777); Fabricius, Lep. Ins. vol. ii. p. 119. n. 530 (1781) ; id. Mant. Ins. vol. ii. p. 69. n. 658 (1787). Hesperia apidanus Fabricius, Ent. Syst. vol. i. pt. 1, p. 280. n. 78 (1793). Polyommatus apidanus Godart, Enc. Méth. vol. ix. p. 652. n. 118 (1823). Amblypodia apidanus Horsfield, Cat. Lep. Mus. E.I.C. p.100. n. 32 (1829); Horsfield and Moore, Cat. Lep. Mus. E.1.C. vol. 1. p. 39. n. 53, pl.1. figs. 5 larva, 5 @ pupa (1857) ; Butler, Cat. Fabr. Lep. B.M. p. 180. n. 2 (1869) ; id. Trans. Linn. Soc. Lond., ser. 2, Zool. vol. i. p. 549. n. 12 (1877). Satadra apidanus Moore, Journ. A. 8. B. vol. li. pt. 2, p. 41 (1884). Panchala apidanus Distant, Rhop. Malay. p. 273. n. 3, woodcut n. 85, ¢ (1885). Thecla apidanus Horsfield, Cat. Lep. Mus. E.1.C. pl. iv. figs. 3 larva, 3 @ pupa (1829). Amblypodia apidanus Druce, Proc. Zool. Soc. Lond. 1873, p. 353. n. 7. Papilio dorimond Stoll, Suppl. Cramer, pl. xxxvii. figs. 4, 4a, 2? (1790). Arhopala apidanus de Nicéville, Butt. India, vol. iti. p. 253 (1890) ; Druce, Proc. Zool. Soc. Lond. 1895, p. 590. Arrhopala apidanus de Nicéville and Martin, J. A. S. B. vol. Ixiv. pt. 2, p. 466 (1895). Var. palawanus Staudinger, Iris, vol. 11. p. 130 (1889). Flos ahamus Doherty, J. A. S. B. vol. lx. pt. 2, p. 33, pl. i. fig. 6, 2 (1891). Hab. Mergui; Moulmein; Malacca; Singapore; Sumatra; Java; Borneo; Nias; Tenasserim; Philippines; Sambawa; Balabac; Assam. Expanse, ¢ 2, 38-50 mm. 116 MR. G. T. BETHUNE-BAKER: A REVISION OF THE The females of this species are subject to variation as to the colour of the blue; I have two with a very pale silvery-blue tinge, many are pale purplish, others are quite dark blue, and some pale violet. All my Java specimens are of the brilliant or silvery-blue forms, but the other forms appear to obtain indiscriminately in Nias, Sumatra, the Philippines, and Borneo. I cannot trace any difference between this and Dr. Staudinger’s palawanus; the colour of the under surface is rather paler, but otherwise they are essentially the same. I have two or three females entirely brown, with no blue area at all. ‘The red basal area varies considerably in the secondaries; I have specimens in which there is quite a suffusion of it, whilst in some it is almost absent, but I believe the streak on the costa of the primaries is always present. This strongly marked species is widely distributed, it having been taken in Moulmein, Burma (de Nicéville has sent me two from Tenasserim in his collection), Mergui, Penang (Herr Semper has sent me a pair from this island), Malacca, Singapore, Sumatra, Nias (Semper), Java, Sambawa, (my specimens from this island are of the ordinary form, Doherty says a local race occurs in the hills), Celebes (where a beautiful local race has apparently been set up), Borneo, and the Philippines. I have no doubt whatever that Doherty's Flos ahamus is merely a pale form of this species. The genitalia are specialised in every point, the tegumen with its extraordinary beak-shaped hook and projection below this, the hooks recurved at their tips; the clasps small and evenly oblong; the penis stout, curved, with a globose orifice, at the tip of which is a strong thorn-like spur. ARHOPALA Arca de Nicéville. Arhopala arca de Nicéville, Journ. Bomb. Nat. Hist. Soc. vol. vii. no. 3, p. 331. n. 8, pl. H. fig. 15, 2 (1892). Hab. Celebes. Expanse, 2, 49 mm. This may be the Celebes form of apidanus Cram. ?. The upperside is like the very pale silvery-blue form of Cramer’s insect, but the blue is more restricted. Below it is plain dark brown and whitish, all trace of red or pink or lilac is quite gone, there being also no trace of the red costal streak in the primaries nor of the basal streak in the secondaries ; in the primaries the. markings are the same, but greyish white takes the place of the lighter brown interspaces and the submedian area, and the lilac scaling at the apex is replaced by light grey; the same differences obtain in the secondaries, and the fringes are wholly tipped with white, which gives the insect a very striking appearance. I have before me the unique type from which de Nicéville described the species, and if all examples from Celebes agree with that specimen, it makes a very fine local race. AMBLYPODIA GROUP OF THE LYCANIDA. 117 ARHOPALA MORPHINA (Distant). (Plate II. fig. 30, ¢.) Panchala morphina Distant, Ann. & Mag. Nat. Hist. ser. 5, vol. xiv. p. 201 (1884) ; Rhop. Malay. p- 274, woodcut 86, ¢ (1885). Arrhopala morphina de Nicéville and Martin, J. A.S. B. vol. lxiv. pt. 2, p. 468 (1895) ; Druce, Proc. Zool. Soe. Lond. 1896, p. 666. Hab. Malacca; Sumatra; Borneo; Silam. Expanse, ¢ 46-50, 2 48 mm. 2. Upperside: both wings pale violet ; primaries with a broad brown costa up to the cell, a very broad and slightly increasing (as it nears the apex) outer margin, and yet broader apex; secondaries with very broad costa and broadish outer margin, increasing very considerably at the anal angle; underside exactly like that of the male. Until now the female of this species does not appear to have been recorded; but Mr. Cator was fortunate enough to capture one at Silam, an island off the coast of British N. Borneo, on the 26th of January, 1892, which is the type female and is in his collection. This species, though by no means the largest in size, is, | think, about the most magnificent of the genus, the intense richness of the deep ultramarine-blue of the upperside and the most beautiful softness of the tones of the under surface, coupled with the rich depth of the basal area, combine to make it a species so striking that once seen it will never be forgotten. It appears to be a rare insect; I have only seen a single specimen from each of the localities given, each of which is from a different collection. ARHOPALA ADRIANA (de Nicéville). (Plate V. figs. 18 & 18 a.) Nilasera? adriana de Nicéville, J. A. S. B. vol. lin. pt. 2, p. 79. n. 20, pl. ix. figs.5 ¢,5a 9 (1883). Arhopala adriana de Nicéville, Butt. Ind. vol. ui. p. 253 (1890). Hab, Sikkim. Expanse, ¢ 46-50, 2 46-48 mm. This species is quite distinct and easily separated from apidanus Cram. by the broad outer border to the primaries, and by the underside pattern, which is quite different ; the broad outer border of the primaries also separates it from asoka de N. It is a very local insect, only having been recorded from Sikkim; I have specimens also from Darjeeling. ‘The genitalia are most peculiar; the tegumen is elevated into a high saddle-back ridge, and then carried forwards into two strong, curved, pointed tusks, jointed at the base with the tegumen, below which are the usual hooks. The clasps are composed of two portions, the hinder portion extending upwards with a hammer- shaped head, the fore portion small, subtriangular, joined at its lower hind extremity to the other part. The penis is stout, strongly curved in its whole length. 118 MR. G. T. BETHUNE-BAKER: A REVISION OF THE ARHOPALA ASOKA (de Nicéville). Nilasera? asoka de Nicéville, J. A. S. B. vol. liii. pt. 2, p. 78. n. 19, pl. ix. figs. 6 g, 6a (1883). Satadra chola Moore, J. A. S. B. vol. liii. pt. 2, p. 39 (1884). Arhopala asoka de Nicéville, Butt. Ind. vol. iii. p. 253 (1890). Hab. Sikkim; Burma (Bhamo) (Staudinger). Expanse, ¢ 48-49, 9 46-48 mm. A close ally of adriana de Nicév.; besides the difference in the outer margins, as already indicated under that species, it may also be separated by the presence of metallic-green scales at the anal angle, which are not present in adriana ; the markings are more pronounced on the underside of the secondaries, especially the subbasal band, which is generally very dark and well marked. This is another very local species, though of wider range than the last, as besides specimens from Sikkim I have a couple from Bhamo (Burma). The genitalia are closely similar to those of the previous species, but the fore portion of the clasps is rather larger. ARHOPALA CHINENSIS Felder. (Plate III. fig. 1, ¢ ; Plate V. figs. 19 & 19a.) Arhopala chinensis Felder, Reise Novara, Lep. vol. ii. p. 231. n. 257, pl. xxix. fig. 10, ¢ (1865). Amblypodia chinensis Moore, P. Z. 8. Lond. 1865, p.774; Hewitson, Ill. Diurn. Lep., Lye. p. 14g. n. 96 (1869). Satadra chinensis Moore, J. A. S. B. vol. lit. pt. 2, p. 41 (1884). Nilasera? moelleri de Nicéville, J. A. S. B. vol. li. pt. 2, p. 80, pl. ix. figs. 4 ¢,4a % (1883). Satadra lazula Moore, J. A. 8. B. vol. lui. pt. 2, p. 40 (1884). Arhopala moelleri de Nicéville, Butt. Ind. vol. ti. p. 266 (1890). Hab. Shanghai; South China; Darjeeling; North India; Sikkim; Khasia Hills ; Sibsagar, Upper Assam. Expanse, ¢ 47-50, 9 43-47 mm. @. Upperside: both wings purplish brown, with a purplish-blue central patch, in the primaries extending over the discoidal cell, the submedian and about half of the median areas, in the secondaries extending just outside the cell and about halfway beyond towards the outer margin. Underside precisely as in the male. Felder’s species chinensis has for long remained practically unknown, and I was surprised in looking at this type in Mr. Rothschild’s Museum to recognise it as probably moelleri de Nicév. I have since carefully examined it with a number of de Nicéville’s species and I have no doubt it is that insect. There are two slight differences: in the type chinensis is a small light spot on the underside near the base in the cell of the primaries, this is not present in ordinary moelleri de Nicev., but there are indications of it in several specimens before me ; and in the secondaries the broad dark band at the base is more broken up on its outer margin than is the case with AMBLYPODIA GROUP OF THE LYCHNID#. 119 moellert, this, however, is decidedly variable in the specimens before me of the latter species. Under these circumstances I have no option but to sink de Nicéville’s moellert under chinensis Feld. My figure is taken from Felder’s type. The genitalia have the tegumen highly excavated, so that the apex is produced into a strong curved hook, the hooks themselves are sharply angled, the clasps are in two portions, the fore portion receding within the hinder portion to a large extent, the latter being rather small. The penis is very stout and strongly curved. ARHOPALA ARESTE (Hewitson). Amblypodia areste Hewitson, Cat. Lyc. B.M. p. 10. n. 44, pl. v. figs. 43 & 44, 9 (1862). Nilasera? areste de Nicéville, J. A. 8. B. vol. lii. pt. 2, p. 97. n. 247 (1888). Satadra areste Moore, J. A. S. B. vol. hii. pt. 2, p. 41 (1884). Satadra patuna Moore, tom. cit. p. 40. Arhopala areste de Nicéville, Butt. Ind. vol. iii. p. 271 (1890). Hab. Nepal; Sikkim; Khasia Hills. Expanse, ¢ & 9, 33-41 mm. This species is very close indeed to chinensis Feld., from which it may be separated by the different colour and pattern of the underside of the secondaries, by the absence of any metallic scales at the anal angle, and by the absence of any red at the base of the primaries, and in a lesser degree by the fact that chinensis has an exceedingly short tail, little more than a tooth perhaps, but in areste this is represented only by the outline of the wings being strongly scalloped. It has hitherto been taken only in Sikkim and Assam ; de Nicéville says it is a rare species, occurring at low elevations in August, September, and November, so that it is one of the most restricted species of the genus. The genitalia are very close to the previous species, but the clasps are more pointed and differently placed. ARHOPALA BAZALUS (Hewitson). Amblypodia bazalus Hewitson, Cat. Lyc. B.M. p. 8. n. 38, pl. iv. figs. 37 & 38, 2? (1862). Satadra bazalus Moore, J. A. 8. B. vol. liii. pt. 2, p. 39 (1884). Satadra teesta de Nicéville, J. A. S. B. vol. lv. pt. 2, p. 253. n. 6, pl. xi. fig. 8, § (1886). Arhopala teesta de Nicéville, Butt. Ind. vol. iii. p. 250. n. 806, pl. xxvii. fig. 197, g (1890). Arrhopala teesta de Nicéville and Martin, J. A. S. B. vol. Ixiv. pt. 2, p. 466. n. 400 (1895). Nee Arhopala bazalus de Nicéville, Butt. Ind. vol. iii. p. 249. n. 804 (1890). Hab. Sikkim; Sylhet; Upper Tenasserim ; Sumatra; Borneo ; Japan. Expanse, ¢ & 2, 46-54 mm. 3. Upperside: both wings excessively deep dull purple, suffused with a silky- brown lustre; primaries with the costa very narrowly, and the outer margin less narrowly black; secondaries with the costa very broadly and the outer margin less broadly blackish. Tail blackish, tipped with white, moderately long. Underside exactly like that of singla de Nicéy., except that in the primaries the transverse band 120 MR. G. T. BETHUNE-BAKER: A REVISION OF THE is usually extended to the submedian nervure instead of to the lower median nervule only, whilst in the secondaries there is usually a very distinct though sparing scaling of metallic-green colour at the anal area. 2. Indistinguishable from sing/a, except perhaps by the green-metallic scaling of the anal area. I have very carefully compared de Nicéville’s species A. teesta with this species of Hewitson’s and find they are identical in every particular. A. bazalus Hew., of de Nicéville’s ‘ Butterflies of India,’ no. 804, p. 249, vol. ii1., should be referred to A, bazaloides Hew. The ordinary form of this insect is intensely deep blackish purple, but this varies, even in specimens from the same locality. From the Khasia Hills I have specimens as dark as any I have seen, andI have also a few specimens not nearly so dark, of merely a dark sublustrous violet-purple. The width of the border varies also in like manner, some are almost linear, whilst others are broader, and one specimen, lent me by Mr. Druce, is most remarkable in this respect, the border being nearly an eighth of an inch wide and increasing at both the apex and anal angle. Turbata Butl., from Japan, is, without doubt, the same species, the only difference being that the underside is perhaps a little more uniform ; the pattern is precisely the same, and it varies in the blue of the upperside exactly in the same manner as do the Khasia Hill specimens. It is a widely distributed insect, though apparently local : the types are from Sylhet and Java; it also occurs in Sikkim, Assam, Upper Tenasserim, Sumatra, Java, S.E. Borneo (from where I have one female), and Japan. ARHOPALA SINGLA (de Nicéville). (Plate V. figs. 20 & 20a.) Satadra singla de Nicéville, J. A. S. B. vol. liv. pt. 2, p. 119, pl. ii. figs. 7 9,8 § (1885); id. Butt. Ind. vol. i. p. 250 (1890). Hab. Sikkim ; Burma (Shan States). Expanse, ¢ about 46, 2 46-50 mm. De Nicéville says: “I have no doubt whatever that A. singla and A. teesta (i. e. bazalus Hew.) represent two perfectly distinct species, as the males can be distinguished at a glance by the coloration of the upperside; but as the undersides of the two species are alike I do not know how their respective females from Sikkim are to. be recognised. As A. singla is only known from Sikkim, females from other localities may safely be placed under A. teesta (i. e. bazalus Hew.) (Butt. Ind. vol. iii. p. 259).” I have no doubt that this remark is quite correct. A. singla is evidently a local and not common species, at least A. bazalus is sent to this country in far greater numbers, but since de Nicéville wrote as above the species has been taken in the Shan States of Upper Burma. The genitalia have a good-sized tegumen, the hooks of which are strongly curved at their bases, the clasps have a long arm-like extension ; the penis is stout, rather long, with a trumpet-shaped orifice. AMBLYPODIA GROUP OF THE LYCANIDA. 121 ARHOPALA CORINDA (Hewitson). (Plate III. fig. 2, 3.) Amblypodia corinda Hewitson, Ill. Diurn. Lep., Lye. p. 14c. n. 81, pl. iia. figs. 34 & 35 g, 33 2 (1865). Arhopala acestes de Nicéville, Journ. Bomb. Nat. Hist. Soc. vol. vii. p. 330, pl. H. fig. 14, 3 (1892). Hab. Philippines; Malacca; Penang; Perak. Expanse, ¢ & 2, 42 mm. 3. Upperside: both wings shining brownish violet, central area brown. Costa and posterior margin of primaries very finely blackish ; costa of secondaries broadly brown, posterior margin finely black; tail black, white-tipped, and a short tooth-like projection at the end of the second median nervule. The exact hue of this insect is difficult to describe; the discoidal, submedian, and median areas are really brown, with a broad costa and narrow posterior and interior margins of a subdued lustrous violet, so suffused that the interior boundaries are not determinable; the secondaries are entirely of this subdued violet except the costa, whilst in certain lights the whole surface looks lustrous brown. Underside: both wings warm brown, with darker spots margined with whitish grey. Primaries with three increasing discal spots and below the third a smaller one; the basal area between the cell and the submedian nervure is darker brown margined outwardly with a pale line; area beneath the lower median nervule pale greyish; transverse band composed of five spots, confluent, the first to the third outwardly inclined, especially the third, fourth and fifth shifted slightly inwards, submarginal row very obscure. Secondaries with a small spot at the costal base, four basal spots, the third shifted slightly outwards, and fourth inwards beyond the internal nervure, followed by three larger spots below each other; cell closed by a larger spot, beneath which is a smaller triangular one; the two upper spots of the transverse band are quite disconnected from the others but touch the large spot closing the cell, third spot of transverse band shifted right outwards, fourth outwards again, fifth well inwards, sixth outwards, seventh angular spot inwards and confluent with the eighth long spot; submarginal row indefinite but distinct, followed by a marginal lunular row; a velvety-black spot is at the anal angle and between the lower and second median nervules, the space between and over which is filled with bright bluish-metallic scales. 2. Upperside: both wings blackish brown ; primaries with the discoidal, submedian, and two-thirds of the median area bright blue, which is invaded at the end of the cell by a black spot; secondaries with a much smaller and narrower bright blue area, also interrupted by a black spot closing the cell. Underside as in the male, but sub- marginal row distinct in primaries, and the upper two spots of the transverse band just connected with the others ; the colour is much greyer in parts than in the male. In certain lights the blue of the upper surface is a true blue, whilst in others it has a purplish tinge. VoL. XVII.—PART 1. No. 16.—August, 1903. R 122 MR. G. T. BETHUNE-BAKER: A REVISION OF THE I have very carefully compared de Nicéville’s acestes with this species and find it is identical; the upper surface of his two specimens (one from Perak and the other from Penang) is a shade purpler, but I have one from Dr. Staudinger from Malacca which is of the ordinary colour, and the underside pattern is precisely similar in all. The colour of the upper surface and the distinct markings of the underside at once separate it from any other species. Its distribution is peculiar, it hitherto having been recorded only from the Philippine Islands, Maiacca, and Penang. ARHOPALA BELLA Bethune-Baker. Arhopala bella Bethune-Baker and Druce, P. Z. 8. Lond. 1896, p. 664, pl. xxx. figs. 6 ¢,7 9. Hab. Kina Balu (Borneo). Expanse, ¢ 55, 9 48 mm. This very beautiful and (in the male) very unusual looking species will come next to A. corinda Hewitson, but may at once be recognised by its large size, its brilliant purple margins, and its lustrous brown inner areas; the transverse bands are also somewhat different. A. corinda below has a more spotted appearance than A. della. ARHOPALA BURUENSIS Holland. Arhopala buruensis Holland, Novit. Zool. vol. vii. p. 78 (1900). Hab. Buru Island. Expanse, ¢ 32, 2 30 mm. ; I have not seen this species, which Mr. Holland places near my kouwnga (the azinis de N.), and in my inability to compare the types it is impossible to do more than record the species, which in this section of the group are differentiated by so small or inconspicuous markings that it is not safe to make any remarks on the insect unless we have the species before us. ARHOPALA AZINIS de Nicéville. Arhopala azinis de Nicéville, Journ. Bomb. Nat. Hist. Soc. 1895, p. 3, pl. T. fig. 31. Arhopala kounga Bethune-Baker and Druce, P. Z. S. Lond. 1896, p. 662, pl. xxx. fig. 4, ? (nec fig. 3, 3). Arrhopala azinis de Nicéville and Martin, J. A. S. B. vol. lxiv. pt. 2, p. 466. n. 403 (1895). Hab. Sumatra (N.E.); Borneo (Kina Balu). Expanse, ¢ & 2, 34 mm. @. Upperside: both wings like the male, but blue slightly more violet and much less extensive, the primaries having a large black invading patch at the end of the cell. My description of kownga appeared just after de Nicéville’s had been published, and I find the male figured in my paper is only a curious form of aroa, the female being that sex of de Nicéville’s species. In that specimen, however, the black invading patch AMBLYPODIA GROUP OF THE LYCANIDA. 123 is not present in the primaries, and it will be interesting to note whether that is constant in the Bornean form. The species can readily be separated from A. ace de Nicéville by the broad borders to the primaries and the difference in the pattern beneath. My female from Kina Balu is apparently the only specimen of that sex yet taken. ARHOPALA ACE de Nicéville. Arhopala ace de Nicéville, Journ. Bomb. Nat. Hist. Soc. vol. vii. p. 329, pl. H. fig. 13 (1892). Arrhopala ace de Nicéville and Martin, J. A. 8. B. vol. Ixiv. pt. 2, p. 464. n. 390 (1895). Hab. Perak; Sumatra. Expanse, 3, 46-55 mm. The female of this species is at present unknown, but the insect is a strongly marked species and can be separated from azinis de N. by the quite different colour of the upperside and by the linear markings beneath. ARHOPALA AGRATA de Nicéville. Arhopala agrata de Nicéville, Butt. India, vol. iii. p. 251, frontispiece fig. 187, g¢ (1890). Hab. Singapore ; Java; Burma. Expanse, 2, 48 mm. I have not seen the types of this species, but Mr. de Nicéville has kindly lent his entire collection to me, in which are one specimen from Singapore and two from Burma; the types from Singapore are in the Raffles Museum at Singapore, which he (Mr. de Nicéville) describes as of the same colour as ¢eesta de Nicév. (i. e. bazalus Hew.) ; the specimens before me, however, are by no means so deep in hue, being of a more real purple with less of the brown shade; this, of course, may be an individual aberration, but it is advisable to notice it. The underside pattern is somewhat near that of A. ace de Nicéy., but the transverse band of the primaries is narrow and continuous, whilst in the latter species it is broad and distinctly fractured below the fourth spot; the markings are also very obscure indeed, which is not the case with ace, and the colour of the upperside of the latter is totally different. ARHOPALA DETRITA (Staudinger). Amblypodia detrita Staudinger, Iris, vol. 11. p. 129 (1889). Hab. Palawan. Expanse, ¢, 30 mm. This species was described by Staudinger from an unique specimen, but it appears to be fairly distinct from agrata de Nicév. ; it may be known by its much smaller size, its brighter blue, its broad borders, and beneath by its transverse bands being straighter and in the primaries entirely unfractured. R2 124 MR. G. T. BETHUNE-BAKER: A REVISION OF THE ARHOPALA @NOTRIA (Hewitson). (Plate III. fig. 3, ¢.) Amblypodia cenotria Hewitson, Ill. Diurn. Lep., Lye. p. 14a, pl. ic. fig. 56 (1865). Hab. Philippine Islands; Nias; Borneo. Expanse, ¢ & 9, 41-43 mm. 3. Upperside: both wings brilliant lustrous bluish violet with excessively fine linear black costa, no black posterior margin, the fringes only being blackish. The same is the case with the secondaries, except that the costa is broadly brown and the abdominal fold is brown; lobe-spot and tail brown, tipped with white. Underside: both wings brown, more or less with a chestnut tinge, with darker spots having pale borders. Primaries with the three usual discal spots, the second and third having each two small spots over them and one below each ; transverse chain-like fascia consisting of seven spots, one on the costa, the three next shifted right out, the two below inwards, below which is an indefinite spot varying in size; submarginal row distinct. Secondaries with a spot on the extreme basal edge of the costa, with four basal spots, the lowest shifted beyond the internal nervure, followed by three rather larger spots ; cell closed by an almost reniform spot, below which is a small triangular one ; chain- like fascia composed of four pairs of spots, the second pair shifted outwards, third inwards, fourth inwards again and confluent; submarginal row distinct, a velvety- black lobe-spot and another between the second and lower median nervules, each of which is edged above and in the intervening space with bright blue (occasionally green) metallic scales. @. Upperside: both wings brown, with a small purplish patch over the discal and central area of the wing, reduced in the secondaries almost to the cell itself. Underside as in the male, but the second and third discal spots are linked together by two spots touching the costal margin of each. The obliqueness of the first four spots of the transverse band in the primaries varies a good deal, but the species is well marked and easily distinguishable. It has often been confused with agnis Felder, which is a very much larger species. ARHOPALA DIAS (Hewitson). (Plate III. fig. 4, ¢.) Amblypodia edias Hewitson, Cat. Lycenide B.M. p. 6, pl. iv. fig. 36, g (1862) ; id. Ill. Diurn. Lep., Lye. p. 7 (1865). Hab. Java. Expanse, ¢, 51 mm. g¢. Upperside: both wings pale, very metallic, lilac-blue; primaries with linear brown costa and finest linear outer margins; secondaries with broad brown costa and very fine linear outer margins; abdominal fold pale brownish grey ; tail black, long, tipped with white. Underside pale chestnut-brown, with darker spots palely encircled. Primaries with three increasing cell-spots, between the second and third of which is AMBLYPODIA GROUP OF THE LYCANIDA. 125 another spot just over the cell, and below the third a larger spot in the lower median angle; transverse band very distinctly spotted, composed of six spots, first spot small below the costa, second, third, and fourth outwardly oblique and shifted right out from the first, below these two shifted right inwards, but inclined slightly outwards ; ‘submarginal row distinct but indefinite; submedian and subdiscal area pale greyish. Secondaries rather darker than primaries, with four small very dark basal spots, second and fourth shifted inwards, followed by three equally dark spots below each other; cell closed by a large subovate spot, below which is a smaller one in the lower median angle; transverse band very distinctly spotted, composed of eight spots, second separate shifted outwards, third shifted right out again, fourth slightly outwards, fifth well inwards, sixth slightly outwards, seventh angular spot well inwards, confluent with the long eighth spot ; submarginal row distinct, parallel with the outer margin; lobe very slightly developed, with velvety-black spot, and on each side of tail a dark spot with superimposed greenish-metallic scales, which also edge the lobe-spot above. The beautiful metallic pale colour above will separate this species from any of its allies, together with the very distinctly marked underside. It is a species that does not appear to have been recognised since Hewitson’s description, several identifications in collections I have seen being wrong. ARHOPALA AGNIS Felder. Arhopala agnis Felder, Reise Novara, Lep. vol. u. p. 228 (1865); Staudinger, Ex. Schmett. p- 281, pl. xevi., g (1888). Amblypodia agnis Hewitson, Il]. Diurn. Lep., Lyc. p. 14 a. n. 74 (1869). Narathura agnis Distant, Rhop. Malay. p. 262. n. 2, pl. xxi. fig. 29, 9 (1885). Amblypodia anarte Hewitson, Ill. Diurn. Lep., Lye. p. 4. n. 8, pl. i. figs. 6, 7, 2 (1863). Arhopala agnis de Nicéville, Butt. Ind. vol. iii. p. 236 (1890); Druce, P. Z.S. Lond. 1895, p. 589. Arrhopala agnis de Nicéville and Martin, J. A. S. B. vol. Ixiv. pt. 2, p. 466 (1895). Hab. Burma; Upper Tenasserim; Mergui; Malacca; Perak; Sumatra; Nias; Mindanao; Borneo. Expanse, ¢ 55-56, 2 53-64 mm. The colour of agnis is subject to some variation; I have one specimen that is almost blue in tone, and several others very pale mauvy-violet. This species is easily recognisable from any of the preceding near allies by its much larger size, by the difference in colour of the upperside; the pattern below is very similar, but the difference in tone of colour is very marked. ARHOPALA YENDAVA (Gr.-Smith). Amblypodia yendava Grose-Smith, Ann. & Mag. Nat. Hist. ser. 5, vol. xix. p. 297 (1887) ; id. Rhop. Exot., Lycenide, pl. xiv. figs. 8&9 g,10 ? (1898). Arhopala yendava de Nicéville, Butt. Ind. vol. lil. p. 238. n. 790 (1890). 126 MR. G. T. BETHUNE-BAKER: A REVISION OF THE Hab. Burma. Expanse, ¢ 58, 2 60 mm. The only difference between the two types of this species and agnis is that the blue in the female is of greater extent than is usual in Felder’s insect and that it is paler ana bluer, and I have but little doubt that the name should be sunk under agnis Felder. ARHOPALA TAMEANGA Bethune-Baker. Arhopala tameanga Bethune-Baker and Druce, P. Z.S. Lond. 1896, p. 568, pl. xxix. figs. 7 3, 8 9. Hab. 'Tameang-Lajang (S.E. Borneo). Expanse, ¢ 57, 2 54 mm. From agnis Feld. this species may be easily distinguished by the quite different colour of the upperside, which is very dark; the underside pattern is somewhat similar, but the spots of the transverse bands of both wings are more confluent, especially in the secondaries. In agnis both these fascia are very distinctly spotted. The types are now in my collection. ARHOPALA SEMPERI Bethune-Baker. Arhopala semperi Bethune-Baker and Druce, P. Z.8. Lond. 1896, p. 659, pl. xxix. figs. 9 g, 10 9. Hab. Borneo (Tameang-Lajang, Kina Balu); Baram River. Expanse, ¢ 54, 2 50 mm. This species is very near tameanga Bethune-Baker, but may be separated by the difference in the colour of the upperside, which is dull violet, and by the broadish borders ; whilst beneath the pattern is similar, but the spots are larger, especially those at the base and the cell-spots. The ground-colour is paler, making the pattern stand out much more prominently. I have, since my first description was written, seen two females from the Baram River (N. Borneo) which I believe to be this species ; they are in Mr. Rothschild’s Museum, but are very darkly marked. ‘The types are in my collection. ARHOPALA BARAMI, n. sp. (Plate IIT. fig. 5, ¢ .) Hab. Borneo; Perak. Expanse, ¢ & 2, 50-51 mm. g. Upperside: both wings purplish blue, in a side light dull and rather dark: primaries with the costa of medium width, dark brown, and a broad outer margin rather wider at the apex; secondaries with a very broad brown costa and rather less broad outer margin; tail brown, tipped with white, of moderate length. Underside: both wings ochreous-brown, with darker spots palely encircled. Primaries with three large increasing cell-spots, below each of which is another, the first two obscure, the AMBLYPODIA GROUP OF THE LYCANID. 127 second being small, the third is dark and large and occupies the lower median angle; transverse band very wide, composed of six spots, the first the smallest, the first four spots being confluent, inclined outwards, with a very slight curve at the fourth, spots 5 and 6 confluent, shifted slightly inwards; submarginal row obscure, marginal row more obscure; submedian and subdiscal areas pale. Secondaries with four very distinct basal spots below each other, fourth one shifted well inwards, followed by three largish spots below each other, second one shifted very slightly inwards; cell closed by a large spot, below which is a small one in the lower median angle; transverse band composed of eight spots, the first two large, the second very large and shifted outwards and touching both the spot closing the cell and the third spot which is shifted well outwards, fourth shifted again out, fifth slightly inwards, sixth very slightly outwards, seventh angular spot confluent with the eighth, which is almost bisected, shifted well inwards; submarginal row obscure, a deep black small spot at the anal angle preceded by two other black spots, which are well covered with pale azure-blue metallic scales, which also edge the anal spot above. 2. Upperside: both wings paler blue tinged with violet ; primaries with a widish brown costa, increasing suddenly just before the end of the cell, where a blackish patch is thus formed invading the blue; outer margin broad, increasing somewhat towards the apex. Secondaries with less broad margins than those of the male. Underside exactly like that of the male. This species can at once be separated from semperi Bethune-Baker by the purplish colour of the upperside, by the broad outer margins, and by the altogether paler tone of the underside, as also by the pattern being more confluent than that of semperi. I have before me a male from Perak, in which the outer margins are quite narrow. The male type is in the collection of Mr. Rothschild, and the female in that of the late Mr. de Nicéville, now in the Calcutta Museum. ARHOPALA DAJAGAKA (Staudinger, MS.). Arhopala dajagaka Bethune-Baker and Druce, P. Z.S. Lond. 1896, p. 660, pl. xxix. figs. 119,129. Hab. Kina Balu; Labuan; Tameang-Lajang. Expanse, ¢ & 2, 55-56 mm. This species may be readily distinguished from semperi mihi by the absence of a tail and by the colour of the underside, which is of a somewhat peculiar greenish brown, and by the isolated spot in the subdiscal area below the second spot in the discoidal cell of the primaries lying about midway between the median nervure and the inner margin. It may equally well be separated from barami mihi by the darker blue of the upperside, the narrow black outer margins, which are decidedly broad in barami, and by the under-surface differences just enumerated. 128 MR. G. T. BETHUNE-BAKER: A REVISION OF THE ARHOPALA ANAMUTA Semper. (Plate III. fig. 6, 3.) Arhopala anamuta Semper, Schmett. Philip. Ins. Bd. v. p. 208. Hab. Mindanao. Expanse, 3, 44-48 mm. g. Upperside: both wings dull brownish purple, looking in certain lights quite brown ; primaries with narrow black posterior margins and a narrower costa ; secondaries with a narrow dark posterior margin and a broad costa; abdominal fold brown. Underside: both wings warm brown, with darker spots palely encircled. Primaries with discal spots; between the second and third is often present on the discoidal nervure another minute spot, and below these two the usual spots are present in the pale internal area; transverse fascia composed of a curved row of six quite confluent spots ending on the lower median nervule, below which is a seventh spot shifted inwards; submarginal row rather indistinct, extreme margin finely dark. Secondaries with a half-spot at the extreme base of the costa; four basal spots inclined inwards, the third being projected outwards, followed by three larger spots ; cell closed by a subquadrate spot, below which is a third triangular one connecting it with the lowest of the previous three; transverse chain-band composed of four pairs of confluent spots, the lower of the first pair has its outer margin shifted outwards, the second pair is shifted further outwards, the third pair inwards, the fourth pair consists of the angular seventh shifted imwards and the long eighth spot; submarginal row fairly distinct; a black spot is at the anal angle and between the second and lower median nervule, the space between which is filled with metallic- green scales, with which the spots themselves are surmounted ; extreme margin finely dark. No tail, but the lower median nervule is elongated into a very short tooth-like projection, decidedly shorter than in dajagaka. This species is a close ally of dajagaka Stgr., but is easily distinguished by the uniformly broad and regular transverse band on the underside of the primaries. I have also a small-sized specimen of a much brighter blue, with no brown shading at all. ARHOPALA ALLATA (Staudinger). (Plate III. fig. 7, 3.) Amblypodia allata Stgy. Iris, vol. ii. p. 125, Taf. ii. fig. 1 (1889). Arhopala allata Druce, Proc. Zool. Soc. Lond. 1895, p. 590. Hab. Palawan, Mindanao, and Mindoro; Labuan. Expanse, ¢ 41-45, 9 42-44 mm. g. Upperside: both wings very dark, dull, almost brownish violet, in some lights looking almost brown ; primaries with broad blackish posterior margins and narrow costa; secondaries with very broad dark brown margins all round; tail tipped with AMBLYPODIA GROUP OF THE LYCZANIDA. 129 white. Underside: both wings chestnut-brown, with darker spots with pale en- circlings. Primaries with three discal spots; below the second and third is another spot, the second in the paler submedian area, that below the third at the angle of the lower median nervule ; the transverse band composed of five confluent spots, almost straight, but slightly inclined outwards; beneath this, under the lower median nervule, is sometimes a very minute dot; submarginal row somewhat indistinct, extreme margin excessively fine dark brown. Secondaries with four basal spots, the second and fourth shifted inwards, followed by three larger spots in a straight line below each other ; cell closed by a long narrow spot, below which is a very small one in the lower median angle; transverse band composed of two upper detached spots, the second shifted outwards, four confluent spots shifted right outwards, the lower two of which are each slightly shifted inwards, seventh and eighth spots quite confluent, the angular seventh being shifted well inwards; submarginal row fairly distinct, the upper white margin of which reaches the internal nervure; anal spot black, and a black dash is between the lower and second median nervules, the space between and above which is filled with metallic bluish-green scales. 2. Upperside: both wings dull bluish purple, brownish purple in certain lights. Primaries with very broad posterior and apical brown margins and a broad brown costa. Secondaries with very broad margins all round, in fact the purple only occupying an area a little beyond the discoidal cell. Underside like that of male, but in the transverse band in the secondaries the third of the confluent spots is shifted decidedly inwards; this is constant in the five females before me. The late Dr. Staudinger most kindly lent me his two original types of this species, which were figured in the ‘ Iris,’ vol. 11. pl. 11. fig. 1, 2 ; but there must have been a mistake with this figure, for it does not agree with the type female, nor yet with any female I possess, neither with any male. ‘here is no spot beneath the lower median nervule of the transverse band in the primaries, and in the secondaries the two upper spots of the transverse band are quite detached and are distinct and well defined, not confluent and almost obliterated as in the figure. ‘The spots and markings are also much sharper than the figure would lead one to suppose. The only specimen from Mindoro is bright almost ultramarine-blue, not dull purplish as usual. ‘The peculiar colour of the upperside, brownish violet, will separate this species from any of its neighbours. ARHOPALA KHAMTI Doherty. Arhopala khamti Doherty, J. A. S. B. vol. lx. pt. 2, p. 32, pl. i. fig. 5, ¢ (1891). Hab. Margherita, Upper Assam. I have not been able to see this species, but from the description it is evidently near anea Hew. VOL. XVII.—PaRT I. No. 17.—Augqust, 1903. S 130 MR. G. T. BETHUNE-BAKER: A REVISION OF THE ARHOPALA @NEA (Hewitson). (Plate V. figs. 21 & 21a.) Amblypodia enea Hewitson, Il. Diurn. Lep., Lye. p. 14e, pl. ite. fig. 55 (1869). Arhopala enea de Nicéville, Butt. Ind. vol. ii. p. 243 (1890). Hab. Sikkim; Khasia Hills. Expanse, ¢ & 2, 41-44 mm. This species is local, but not uncommon where it occurs; it may be readily distin- guished from its neighbours by its brighter colour, larger size, and the clear, clean-looking ochre-brown of the under surface. The genitalia have the tegumen excavated and open at the apex, and the hooks, jointed at the elbow, are from that point long and slender ; the clasps are very deep, excavated at the tips, forming a sort of jaw, the upper part of which is resolved into a long, curved, sharp hook. The penis is long and slender. ARHOPALA RAMA (Kollar). (Plate V. figs. 22 & 22 a.) Thecla rama Kollar, Hiigel’s Kaschmir, vol. iv. pt. 2, p. 412. n. 1, pl. iv. figs. 1, 2, ¢ (1848). Amblypodia rama Hewitson, Cat. Lyceenide B.M. p. 18. n. 59, pl. vi. figs. 69, 70 3, 71 ¢ (1862) ; Leech, Trans. Ent. Soc. Lond. 1889, p. 111. n.51. Panchala rama Moore, Proc. Zool. Soe. Lond. 1882, p. 252; Doherty, Journ. A. 8. B. vol. lv. pt. 2, p. 126. n. 122 (1886); Butler, Ann. & Mag. Nat. Hist. ser. 6, vol. i. p. 145. n. 38 (1888). Amblypodia querceti Moore; Horsfield and Moore, Cat. Lep. Mus. E.I.C. vol. i. p. 43. n. 64 (1857). Arhopala rama de Nicéville, Butt. Ind. vol. ii. p. 251 (1890). Amblypodia violacea Rober, Iris, vol. i. p. 72, Taf. v. fig. 6 (1886). Hab. Himalayas; Burma; China; Sikkim. Expanse, ¢ 35, 2 38-35 mm. Herr Rober has described his violacea, from “ the East Indies,” as a distinct species, but it agrees in every respect with this insect—there is not one small point in which 1 can differentiate it. It may be known from wnea Hew. by the broadish outer margins, by the pale colour of the underside, by the tail being little more than a long tooth, and by the absence of an anal black spot and any metallic scaling. The exceedingiy broad dark margins of the secondaries are also a marked charac- teristic of the species. ‘The genitalia are not unlike those of @nea Hew., but the tegumen-hooks are short and not elbowed; the clasps are broader, with the upper part of the jaw pointed but broader. The penis is decidedly stouter. ARHOPALA ATRAX (Hewitson). (Plate III. fig. 8, ¢ .) Amblypodia atrax Hewitson, Cat. Lyc. B.M. p. 18. n. 58 (nec pl. vii. figs. 80 & 82, 9) (1862) Moore, Proc. Zool. Soc. Lond. 1865, p. 774; Hewitson, Ml. Diurn. Lep., Lye. p. 13 (1869). Satadra atrax Moore, J. A. 8. B. vol. liii. pt. 2, p. 39 (1884). Arhopala atrax de Nicéville, Butt. India, vol. iti. p. 246 (1890). AMBLYPODIA GROUP OF THE LYCANID®, 131 Hab. India (distribution unknown to me); Tavoy Valley; Burma. Expanse, ¢ & 2, 42 mm. 3 (type). Upperside: both wings rather bright purplish blue, dull in a side light ; primaries with a narrow black costa and broad posterior margin ; secondaries with a spot on each side of the tail, which is black, rather long, tipped with white. Underside pale brown, washed with pale pinkish lilac, with darker spots palely edged, Primaries with three increasing cell-spots, the first and second of a narrow-oval shape ; below the second and third are two spots divided by the lower median nervule ; touching the tip of the inner edge of the third cell-spot is another one on the edge of the cell; on the costa are two small spots, one over the third and the other nearer the second cell-spot ; transverse band composed of seven spots dislocated below the fourth, first spot small, second larger, shifted outwards, third inwards, fourth outwards, fifth well inwards, sixth slightly out, seventh slightly inwards but with an outward inclination ; submarginal row narrow, rather obscure; submedian area pale. Secondaries with four small basal spots, second and fourth shifted inwards, followed by three larger ones below each other, the second of which is shifted inwards ; a large subquadrate spot closes the cell, below which is a small one in the lower median angle, where it joins the lowest of the series of three spots; transverse band composed of eight spots, the second very large and shifted slightly outwards as to its outer margin, but well inwards as to its inner margin, where it touches the spot closing the cell, third spot shifted well outwards, fourth further out, fifth inwards, sixth outwards, seventh angular spot shifted right inwards and touching the eighth spot (this transverse band has the appearance of a bifurcated band owing to the second spot touching the cell-spot, which is joined to the two spots below, as described); submarginal row very obscure; lobe-spot blackish (lobe scarcely developed), and a small blackish spot just beyond the tail well covered with pale bronzy-green scales, which fill up the space between it and the lobe-spot, and also edge the latter one. 2. Upperside: both wings bright rather pale blue, darker in a side light ; primaries with very broad costa, broader beyond the cell, outer margins very broad ; secondaries with very broad costa and rather less broad outer margin. Under surface exactly as in the male, but the spots are rather larger. In describing this insect Hewitson, as he intimates, confused two quite distinct species; fortunately his type male is preserved in the National Collection, and it is distinct from the female in the Hewitson Collection. Colonel Bingham has given a pair to the Museum which he captured in Burma (Tavoy Valley) in March 1893, so that I am thus enabled to describe the true female. My description of the male is from the original type. The female described with the male by Hewitson is entirely distinct, and I have therefore re-named and re-described it with its correct male under the name hewitsoni. 152 MR. G. T. BETHUNE-BAKER: A REVISION OF THE ARHOPALA HEWITSONI n. sp. (Plate III. fig. 9, ¢; Plate V. figs. 23 & 23 a.) Hab. India; base of Himalayas from Dehra Dun to Bhutan; Malda and Chota Nagpur Districts; Central Provinces; Calcutta; Orissa; Nilgiris; Burma. Expanse, ¢ 35, 2 38 mm. g. Upperside: primaries dull lilac, with costa narrowly dark brown, and a very dark brown posterior margin. Secondaries darkish brown, with a lilac-blue patch situated just over the discoidal cell. Tail brown, white-tipped. Underside pale ereyish brown, with a somewhat shining pale lilac tinge, spots darker, palely encircled. Primaries with three small discal spots increasing in size; beneath the second and third are two spots separated by the lower median nervule, and on the costa, a little forward of the third, is another small one; transverse fascia continuous, gently arched; fifth and sixth spots inwardly inclined, seventh spot detached and obscure ; submarginal row obscure but dark, and followed by a darker shading up to the margin; submedian area paler. Secondaries more tinged with lilac, with four very small dark basal spots, the fourth paler and obscure and shifted inwards, followed by three small spots, the upper two dark, the lowest paler and obscure ; cell closed by a larger spot, below which is a small triangular one; transverse fascia-spots—first and second almost confluent, the second rather the larger, the third and fourth confluent and shifted right outwards though not quite detached, fifth shifted inwards, sixth outwards, seventh shifted well inwards, detached from the eighth, which is small and really forms a fourth lowest spot beneath the three middle ones ; submarginal band darker, and followed by a darker shading up to the margin; a very dark spot on the lobe, followed by a scanty scattering of greyish scales mixed with a few metallic-bluish ones on the margin almost to the second median nervure. 2. Upperside: both wings brown: primaries with a darker patch at the apex of the cell, and a largish brighter lilac-blue patch occupying the discal, central, and sub- median areas, but not below the submedian nervure; secondaries uniform brown. Underside as in the male, but spots and fascia a little more distinct. This species has hitherto passed as 4. atraa: Hewitson. ‘That author described two distinct species under that name. ‘The male being described first, as usual, the name of atrav must attach to it, Hewitson’s type of which is in the Natioual Collection. The female described by Hewitson is this species, both sexes of which are now quite common, and are mentioned and re-described by de Nicéville in his ‘ Butterflies of India’ (vol. iii. p. 246). The colour of the upperside alone easily separates it from atrax, which latter is very decidedly brighter and bluer than the present species, which is of avery obscure violet colour. ‘The genitalia do not call for special remark, and are figured on Plate V. figs. 23 & 23 a. AMBLYPODIA GROUP OF THE LYCHNIDA. 133 ARHOPALA ALEMON de Nicéville. (Plate V. figs. 24 & 24a.) Arhopala alemon de Nicéville, Journ. Bomb. Nat. Hist. Soc. p. 371, pl. F. figs. 20 g, 21 9 (1891). _ Hab. Burma. Expanse, ¢ 40, ? 46 mm. De Nicéville likens this species to A. rama Kollar, but it is a much closer ally of A. hewitsoni mihi; indeed the males are difficult to separate ; alemon is somewhat larger, with the violet colour rather more subdued, slightly more extensive and indefinite as to its outer margins; the females are, however, more easily separated, being very much larger, with the whitish patch pupilled with black closing the cell, whilst below the ground-colour is without any violet tinge, decidedly paler than A. hewitsoni, the spots consequently showing proportionately darker. I have at present seen only this species from Burma, where it probably replaces A. hewitsoni in certain districts. The genitalia are slighter in proportion than those of the previous species, the hooks are longer, and the clasps smaller. The penis is nearly straight. ARHOPALA CANARAICA (Moore). Satadra canaraica Moore, Journ. A. 8. B. vol. liii. pt. 2, p. 39 (1884) ; Waterhouse, Aid, pl. elxv. figs. 5, 5a, 3 (1886). Arhopala canaraica de Nicéville, Butt. India, vol. i. p. 245. n. 799 (1890). Hab. Canara, 8. India (Moore); Travancore (de Nicéville). Expanse, ¢ 40-45, 9 44-45 mm. This species is somewhat near a/emon de N., but can be easily separated by the blue colour of the upperside and by the purplish tinge of the brown of the under surface, which is also decidedly darker than in alemon; the pattern is likewise smaller and darker. ARHOPALA ALEA (Hewitson). Amblypodia alea Hewitson, Cat. Lycenide B.M. p. 12. n. 57, pl. vii. figs. 79, 81, ¢ (1862). Satadra alea Moore, J. A. 8. B. vol. liii. pt. 2, p. 38 (1884). Arhopala alea de Nicéville, Butt. Ind. vol. ii. p. 246 (1890). Hab. Xndia. “ Male. Upperside: both wings violet-blue ; fore wing with the outer margin rather broad. Hind wing with the margins as broad as in the females of other species. Underside: both wings brown, tinted with lilac. Fore wing with the transverse band long and narrow, slightly curved. Hind wing with the band broken, ill-defined; the anal angle irrorated with white.” (J/ewitson, /. c.) This species Hewitson described from a specimen in the National Collection which 134 MR. G. T. BETHUNE-BAKER: A REVISION OF THE is now lost, there being no insect in the Museum that at all agrees with Hewitson’s figure. I possess one very close, but the transverse band of the secondaries is not absolutely broken, as shown in his figure. ARHOPALA DODONEA (Moore). (Plate V. figs. 25 & 25a.) Amblypodia dodonea Moore ; Horsfield and Moore; Cat. Lep. Mus. E.1.C. vol. i. p. 43. n. 65, pl. ia. fig. 8 (1857). Panchala dodonea Moore, Proc. Zool. Soc. Lond. 1882, p. 252. Panchala dodonea Butler, Ann. & Mag. Nat. Hist. ser. 6, vol. i. p. 144. n. 37 (1888). Panchala? dodonea Butler, Proc. Zool. Soc. Lond. 1886, p. 364. n. 32. Arhopala dodonea de Nicéville, Butt. Ind. vol. iii. p. 258 (1890). Hab. Western Himalayas. Expanse, ¢ & 2, 40-48 mm. This is a very distinct species, easily recognisable from all others; both sexes are exactly alike. It is perhaps nearest A. rama Kollar, but the very pale, almost silvery, blue, and the extremely broad borders of the upper surface separate it easily, whilst below it is quite different, being very pale, with the markings standing prominently out, the pattern being very similar. Mr. de Nicéville (/. ¢.) thought this might be a dimorphic form of rama. 1! have, however, examined several males and mounted dissections of the genitalia, which show, I think, that the species are quite distinct, though very close. It will be seen that the claspers differ somewhat, and all the organs are very much larger and stouter. ARHOPALA CoMICA de Nicéville. Arhopala comica de Nicéville, Journ. Bomb, Nat. Hist. Soc. vol. xiii. p. 170 (1900). Hab. Live Long, near Bhamo, Upper Burma (5000 feet). Expanse 43 mm. When describing this species from a single example, de Nicéville himself says it may be an aberration, but that he is at a loss to conjecture of what species. I have no doubt whatever that it is a sport of A.dodonea Moore. The shapes of the tail and of the lobe are almost identical; the outline of the wing is very similar. The colour of the wings above, together with the denticulate invasion of the blue patch, at once remind one of Moore’s insect, whilst below the ground-colour is very close in hue, and though the spots are aberrant, the general line of pattern is similar. This is an instance of the undesirability of describing a species from an unique example. ARHOPALA JAPONICA (Murray). (Plate ILI. fig. 10, ¢; Plate IV. figs. 21 & 21 a.) Lycena japonica Murray, Ent. Mo. Mag. vol. xi. p. 167 (1874). Hab. Japan. Expanse, ¢ & 2, 45-47 mm. AMBLYPODIA GROUP OF THE LYCANIDZ. 135 In both sexes the apex of the primaries is produced into a short hook tip. This species is somewhat near paramuta de N., but may easily be recognised by the paler and brighter blue, the cell being closed by a black spot on the upperside of the primaries, and by its larger size, whilst the colour of the under surface is much smoother and rather more uniform. The genitalia are very close to those of A. rama and dodonea, with the penis strongly bent near the orifice. The clasps are, however, broader, and the lateral supports of the tegumen straighter. ARHOPALA PARAMUTA (de Nicéville). (Plate IV. figs. 22 & 22a.) Panchala? paramuta de Nicéville, Journ. A. 8. B. vol. liii. pt. 2, p. 81. n. 23, pl. ix. figs. 7 g, 7a % (1883). Darasana newara Moore, tom. cit. p. 42 (1884); Waterhouse, Aid, pl. clxv. figs. 4, 4a, 3 (1886). Arhopala paramuta de Nicéville, Butt. Ind. vol. iii. p. 268 (1890). Hab, Nepal; Sikkim; Khasia Hills. Expanse, ¢ & 9, 34-36 mm. De Nicéville says (/.c.) this is a common species in Sikkim, occurring in April and May. It is very easily separated from those near it, being a very distinct little species. The genitalia are distinct, having a compact heavy tegumen, with short, nearly straight hooks; the clasps are just the shape of a ham cut off sharply at the knuckle end. The penis is nearly straight. ARHOPALA ZETA (Moore). Amblypodia zeta Moore, Proc. Zool. Soe. Lond. 1877, p. 590, pl. lviil. fig. 6, 2. Satadra zeta Moore, J. A. S. B. vol. li. pt. 2, p. 41 (1884). Arhopala zeta Moore; de Nicéville, Butt. India, vol. ui. p. 268 (1890). Hab. 8. Andaman Isles. Expanse, 2, 31 mm. 9. Upperside: both wings brilliant pale blue; primaries with costa broadly black up to apex of cell, where it suddenly increases in width up to the apex, outer margin very broadly black; secondaries with very broad margins all round. Under- side: both wings warm cinnamon-brown, with darker spots palely encircled. Primaries with three increasing cell-spots, below the third another spot in the lower median angle ; transverse band composed of five (? six) spots; both wings have been injured here, and it is not possible to say whether there is a small spot on the costa or not ; spots confluent, the first three forming a very slight are, and the fourth and fifth having a slight inward tendency, submarginal and marginal rows obscure; subdiscal area darker, submedian paler, than the ground-colour. Secondaries with four small basal spots, the third shifted slightly outwards, followed by four larger spots below each 136 MR. G. T. BETHUNE-BAKER: A REVISION OF THE other ; cell closed by a larger spot, below which is a small one in the lower median angle; above the spot closing the cell are two spots above each other, reaching to the costal nervure; these usually form the first two of the transverse band, which band is shifted slightly outwards, the second spot is shifted slightly out, the third inwards, fourth outwards, fifth inwards, the band being composed of only five spots ; submarginal and marginal rows fairly definite, no spots or metallic scaling at the anal angle. Mr. Moore has very kindly lent his type specimen (unique) to me, from which this description is taken. I have more to say about if under the next species. ARHOPALA ROONA (Moore). Narathura roona Moore, J. A. 8. B. vol. liti. pt. 2, p. 42 (1884). Arhopala roona Moore; de Nicéville, Butt. India, vol. ii. p. 268 (1890). Hab. Andaman Isles. Expanse, ¢ & 2, 33-39 mm. o & 9. Upperside: both wings rather bright violet-blue; primaries with costa broadly blackish up to the apex of the cell, where it suddenly increases up to the apex of the wine g, outer margin very broad ; secondaries with broad margins all round. Underside exactly like zeta Moore, except that the colour is paler and the markings larger, and also that in the secondaries the two spots usually forming the first two spots of the transverse band are just attached to the rest of the band; in the type of zeta they are detached. ‘There is no trace of the usual dark anal spots or metallic scaling. De Nicéville has very correctly diagnosed these two species. I keep them apart with great reluctance: the blue of the upperside is, however, quite distinct, being quite pale, almost silvery and lustrous in zeta; in roona it is quite violet. ‘The specimen of zeta has, however, a somewhat crippled appearance, and is decidedly smaller, and it gives me the impression of the larva having been starved. If this is so it is not at all unlikely that the colour is abnormal. Recent material has brought to our hands more voona from the Andaman Isles and also from Burma, but no zefa have come. I feel that they are probably one and the same species, but in the absence of sufficient material I refrain from definitely uniting them. ARHOPALA AGELASTUS (Hewitson). Amblypodia agelastus Hewitson, Cat. Lye. B.M. p. 12, pl. vi. figs. 61 & 62, 2 (1862). Narathura agelastus Moore, Journ. Linn. Soc. London, Zool. vol. xxi. p. 44 (1886). Arhopala agelastus de Nicéville (in parte), Butt. India, vol. ii. p. 278 (1890). Hab. India; Mergui; Burma. Expanse, ¢ ,47 mm. g. Upperside: both wings rather bright dark purplish blue; primaries with costa narrowly black, slightly increasing towards the apex, posterior margin rapidly decreasing (Sh) = AMBLYPODIA GROUP OF THE LYCAINID. ils towards the anal angle, apical area very broadly black ; secondaries with a broad costa and rather narrow posterior margin. Under surface ochre-brown, with darker markings palely but faintly encircled. Primaries with three small increasing cell-spots, below the third a small one in the lower median angle ; transverse band composed of six confluent spots, with a very slight curve in the three upper ones, the whole band being slightly oblique outwards ; submarginal and marginal row obscure; submedian area pale; sub- discal area dark. Secondaries with four, quite small, basal spots, the fourth shifted well inwards, followed by three larger ones below each other, the second shifted slightly inwards, and the third the largest of the three; cell closed by a subreniform spot, below which is a very small one in the lower median angle ; over the spot closing the cell are two others (usually forming the two upper spots of the transverse band) above each other, the lower one being the larger; transverse band composed of six spots beginning on the upper discoidal nervule, the first two confluent, the second shifted very slightly outwards, third with the inner margin only shifted inwards, fourth almost confluent with the third, fifth angular spot touching the sixth and shifted well inwards ; submarginal and marginal rows obscure ; a small black spot at the anal angle, preceded by two others, which are more or less covered with pale greenish-blue metallic spots. Arhopala perissa Doherty is a very close ally of this species; in fact, I am extremely doubtful if they are not one and the same, but as I have so little material I deem it best to keep them distinct for the present. Doherty’s species has the spots and fascia wider and larger, and the borders of the upperside broader. De Nicéyille says (Butt. Ind. /.¢.) that Mr. Doherty showed him certain specimens taken at Mergui and in Tenasserim, and that some, of them bore traces of a circular patch of differently shaded scales on the upperside of the fore wing, as in A. atosia Hewitson. I have these specimens now before me, named agelastus by de Nicéville ; but they are moolaiana Moore, and not Hewitson’s species: the reference to atosia should therefore be eliminated from the account of agelastus in de Niceéville’s work. ARHOPALA PERISSA Doherty. (Plate III. fig. 11, ¢.) Arhopala perissa Doherty, Journ, A. 8. B. vol. lvii. pt. 2, p. 419, pl. xxii. fig. 11 (1889) ; de Nicé- ville, Butt. India, vol. i. p. 278. n. 842 (1890). Hab. Myitta; Tenasserim Valley ; Burma. Expanse, ¢ & 9, 45 mm._ There appear to be two forms of this species, differing on the upperside only; the type is rich purplish blue, and I have several specimens of this colour, of both sexes, before me, but I have also in de Nicéville’s series several of both sexes which are brilliant pale lustrous blue, without any admixture of purple at all; the underside is precisely the same in every respect. I notice the same thing in townguva Gr.-Smith, but the difference is by no means so marked. VOL. XVI.— PART 1. No. 18.—August, 1903. T 158 Mk. G. T. BETHUNE-BAKER: A REVISION OF THE This species is an exceedingly close ally of townguva Gr.-Smith, but as I have only seen about a dozen of each species, I have thought it better to keep them distinct. It is quite possible, however, that further material may prove them to belong to one and the same insect; in fact I believe this will prove to be so. They obtain in almost exactly the same district ; both are local, this species being only recorded from Burma, and townguva from Burma and the Andaman Isles. ARHOPALA TOUNGUVA (Gr.-Smith). (Plate ITI. fig. 12, ¢.) Amblypodia tounguva Grose-Smith, Ann. & Mag. Nat. Hist. ser. 5, vol. xx. p. 268 (1887). Arhopala tounguva de Nicéville, Butt. India, vol. ii. p. 269. n.. 830 (1890) ; Grose-Smith, Rhop. Exot., Lyc. pt. 44, pl. xv. figs. 4, 5 (1898). Hab. Toungu; Burma; Andaman Isles. Expanse, ¢ & 9, 44-47 mm. In some specimens the underside is irrorated slightly with ochreous, whilst the upperside is occasionally of a violet-blue. Both sexes are almost the same. The markings of the underside are very close to perissa Doh., and also to asopia Hew. ; from the former it can be distinguished by the difference in the transverse band of the secondaries, and also by the colour of perissa being clear, smooth, cinnamon-brown beneath, whilst ¢owngwva has more or less of an irrorated or dusted appearance. Asopia Hew. is sufficiently distinct not to require differentiating. As mentioned under the previous species, I believe that ¢ownguva will prove to be one and the same as Doherty’s insect. ARHOPALA ASOPIA (Hewitson). (Plate IIT. fig. 15, ¢.) Amblypodia asopia Hewitson, Il. Diurn. Lep., Lye. p. 14f. no. 90, pl. ie. figs. 50, 51 (1869). Arhopala asopia de Nicéville, Butt. Ind. vol. ii. p. 270 (1890). d Hab. Moulmein, Upper Tenasserim, ‘Tilin Yaw, and Bhamo (N. Burma). Expanse, ¢ & 2, 46-48 mm. ¢. Upperside ; both wings dull sublustrous violet, rather dull in tone ; primaries with the costa narrowly, posterior margin broadly, apex more broadly dark brown ; secondaries with costa broadly, and posterior margin less broadly brown; abdominal margin greyish. Underside: both wings cinnamon-brown, with darker spots palely encircled. Primaries with spots darker than secondaries, and tinged with violet ; three discal spots, below which the submedian area is dark, below the third spot is another adjoining it; transverse band very regular and even in width, widish, slightly curved near the costa, and with a slight outward inclination; below the submedian nervure is an obscure spot beneath the band; submarginal band dark, but obscure ; submedian area pale. Secondaries with spots paler than in primaries, one on the costal base, four basal spots, the upper two very small and dark, followed by four larger ones; cell AMBLYPODIA GROUP OF THE LYCENIDA., 139 closed by an elongated spot, below which is a small triangular one ; two spots from the costa below each other, of which the lower one is slightly oblique and touches both the elongated spot and the upper one of the transverse band, which begins on the upper discoidal nervure and extends beyond the submedian ; the second spot, rectangular like the first, is shifted slightly outwards, the inner margin of the third is shifted inwards, fourth spot outwards, fifth right inwards, not detached, and confluent with the small sixth; submarginal row very obscure; the lower median nervure is slightly elongated, so as to form a short tooth-like projection. 2. Upperside: both wings bright blue; primaries with costa broadly, outer margins more broadly, and apical area still more broadly blackish ; secondaries with broadish borders all round, but decreasing towards the anal angle, the costa being the broadest. Underside as in the male. ; I have one female from Tilin Yaw (N. Burma) before me in which the colour of the upper surface is exactly of the same violet tint as in the male, not bright blue as usual. In de Nicéville’s fine collection I find several specimens of this insect, and it is from these that Iam enabled to add Tenasserim and Jilin Yaw to its distribution, whilst Haberhauer took a female in Bhamo, which is in Dr. Staudinger’s collection now before me. It is evidently a rare and local species, but is not difficult to separate from its near neighbours by the pale violet colour, and also by the underside being very decidedly paler than any of its allies. It should be easily recognised from the figure (Plate III. fig. 13, ¢). ARHOPALA @ETA de Nicéville. Arhopala weta de Nicéville, Journ. Bombay Nat. Hist. Soe. vol. vi. p. 333. n. 10, pl. H. fig. 17, 2 (1892). Hab. Burma. Expanse, 2,43 mm. I cannot help regarding this as merely an aberration of asopia Hew. ; the whole appearance of the specimen bears this out. In a difficult genus like this, where the species are so closely related and are separable by such small characteristics, it would be a wise plan never to describe a species from a unique example (as in this case) unless it should happen to be a palpably new and strongly marked insect. ARHOPALA PERIMUTA (Moore). Amblypodia perimuta Moore; Horsfield and Moore, Cat. Lep. Mus. E.1.C. vol. i. p. 42. n. 61 (1857) ; Hewitson, Cat. Lycenide B.M. p. 12. n. 55, pl. vi. figs. 65, 66, d' (1862) ; id. IIL. Diurn. Lep., Suppl. p. 21. n. 62, pl. vii. suppl. fig. 61, g (1878). Darasama perimuta Moore, Journ. A. 8. B. vol. liii. pt. 2, p. 42 (1884) ; id. Journ. Linn. Soe. Lond., Zool. vol. xi. p. 44: (1886). Arhopala perimuta de Nicéville, Butt. Ind. vol. iii. p. 270 (1890). 140 MR. G. T. BETHUNE-BAKER: A REVISION OF THE Hab. Sikkim; Sylhet; Khasia Hills; Cherra Punji, Margaree; Pegu; Burma; Mergui. Expanse, ¢ 30-32, 9 29-30 mm. This species can be easily recognised from asopia Hew. by its much smaller size and its blue colour (asopia being quite violet in tone), whilst the quite different colour beneath, yellow and purplish, will at once separate it. The genitalia have the tegumen short and of a heavy scimitar-shape; the clasps are bilobed, and the penis is a straight tube. ARHOPALA AZENTA (Hewitson). Amblypodia azenia Hewitson, Ill. Diurn. Lep., Lye. p. 7, pl. ii. figs. 22, 23 (1865). Hab. New Guinea. Expanse, ¢ 44-52, 2 46-48 mm. This very beautiful species has been exceedingly rare until within the last few years, but recent collections from New Guinea have contained considerable numbers. It has been somewhat difficult to give it and the following species (A. acron Hew.) their right position; but the latter is evidently near to admete Hew., so this will be their best position, though the brilliance of the blue of azenia Hew. at first makes it look out of place. There is no difficulty in recognising the species at a glance, the lustrous almost morpho-like blue of the upperside, and the beautiful white underside just tinged with faintest pink, beneath which the usual spots and bands (nearly allied in position to those obtaining in admete) show through as a sort of purplish-llac colour, should make the insect distinguishable at once. It is a beautiful species, but apparently with about the most limited range of distribution in the group. ARHOPALA ACRON (Hewitson). (Plate III. fig. 14, 3.) Amblypodia acron Hewitson, Cat. Lyc. B.M. p. 9, pl. v. figs. 538 & 54 (1862); id. Ill. Diurn. Lep., Lye. p. 7 (1865). Hab. Batchian. Expanse, ¢ & 2, 46 mm. This insect is a close ally of the preceding species, but can be distinguished at once in that the blue colour of the upperside is decidedly darker, whilst beneath the colour is uniform greyish brown, with the spots and bands a trifle darker, outlined sharply. with fine whitish edgings. It is, I believe, a very rare species; I have only seen about half a dozen examples in all the collections that have passed through my hands, and one of those is in my own collection. ‘Though Hewitson’s figure of the underside is not good, it is quite sufficient to enable the species to be recognized. Neither this nor the precediug species is likely to be confused easily with other species. I can trace no records of this butterfly elsewhere than from Batchian., AMBLYPODIA GROUP OF THE LYCENIDZ. 141 ARHOPALA ADMETE (Hewitson). Amblypodia admete Hewitson, Ill. Diurn. Lep., Lycenide, p. 7, pl. ii. figs. 18 & 19, 3g (1865). Amblypodia eucolpis Kirsch, Mitth. d. k. zool. Mus. Dresden, Heft ii. Taf. vi. fig. 6; Ribbe, Iris, vol. ii. p. 257 (1889). Arhopala admete Grose-Smith, Novit. Zool. vol. i. p. 581 (1894). Hab. Halmeheira; Ceram; Gilolo; New Guinea. Expanse, ¢ & 9, 44-50 mm. This species is at once separable from its near allies by its large size and by its underside pattern, which in the primaries is marked by the absence of any spot in the cell (I have specimens in which the least trace of one is just discernible), the first spot being that closing the cell, and by the broad dark transverse band, whilst in the secondaries the main characteristic is the very white edging (generally more or less suffused in the male) to the transverse band. Its distribution is somewhat confined and local, it being at present only recorded from Ceram, Gilolo, Halmaheira, and New Guinea. Var. WAIGEOENSIS, nov. var. (Plate III. fig. 15, ¢ .) é¢. Upperside: both wings bright rich lustrous blue, rather dark, but not at all dull, as in admete, but much brighter. Underside similar to admete in the markings of both wings, but the ground-colour quite pale; the apical area of the primaries is tinged broadly with pale lilac; the secondaries are entirely tinged with very pale pinkish lilac, aud the white external suffusion to the transverse band is very broad indeed, forming a most marked character. ‘The spots and markings are fully as dark as in admete, and therefore stand out very prominently in the pale ground. 2. Upperside: both wings very much paler and greener in tint, with the median nervules in the secondaries broadly black. Underside as in the male, but the tint of the secondaries is more pinkish ochreous, not shining, and there is no broad white suffusion to the transverse band. ‘There is a trace of a small and obscure spot in the centre of the cell of the primaries, both in this variety occasionally and also sometimes in addmete. ‘This variety forms a sort of link between epimete and admete. Hab. Waigece. Expanse, ¢ 57-40, 9 33-35 mm. I have three males and two females of this very pretty form from the island of Waigeoe, where I expect it-replaces adimete Hew. ARHOPALA EPIMETE (Staudinger). (Plate ILI. fig. 16.) Amblypodia epimete Staudinger, Iris, vol. ii. p. 128, pl. ii. fig. 2, ¢ (1889). [lab, Palawan. Expanse, ¢ & 2, 54-35 mm. 3. Upperside: primaries violet-blue, with costa broadly, posterior margin very broadly, 142 MR. G. T. BETHUNE-BAKER: A REVISION OF THE and apical area more broadly still, purplish brown ; secondaries dark brown, with an obscure blue patch occupying the cell or thereabouts ; no tail. Underside: both wings pale ochreous brown, tinged with pinkish ; spots and fascia darker, margined with whitish. Primaries with three increasing discal spots, beneath and above the third another spot, that above being close on the costa, subdiscal space up to the lower median branch dark ; transverse band composed of six laterally margined spots ending on the lower median nervule ; the first spot on the costa is small, second shifted well outwards is larger, third with its inner margin only shifted a little inwards, fourth large spot shifted well outwards and oblique (these spots are sharply margined and form a slight irregular curve), fifth and sixth spots are indefinite somewhat as to their margins, and the sixth is shifted outwards from the fifth; below this is usually another very indefinite and obscure spot, not reaching quite to the submedian nervule, and not always present. Secondaries more pinkish than the primaries, with four basal spots, the third shifted slightly out- wards, followed by four larger spots and occupying equivalent positions ; cell closed by an irregular spot, below which is another small one touching the third spot of the previous row above, and touching the spot closing the cell are two others, the upper of which is shifted outwards ; transverse band beginning on the upper discoidal nervule and composed of six small spots, the second shifted well outwards, but inclined inwards, the third right inwards, fourth outwards, fifth right inwards, the sixth very small, slightly inwards; these with the lowest of the two rows of four spots form an irregular inner half-circle round the lobe, posterior and abdominal margins; sub- marginal row as usual, the anal portion of which is broadly margined internally with greyish ; margin obscurely and darkly spotted; the usual submarginal row followed by a slight darker marginal spotting ; apical area slightly dusted with greyish ; submedian area from the transverse band pale. ?. Exactly like the male, except that the obscure blue patch in the secondaries is almost absent. This is apparently a rare insect, and at present is only recorded from the island of Palawan ; it is rather near admete, var. waigeoensis, but may be separated by its smaller size, its more violet colour, which is very much less extensive in the primaries and is almost absent in the secondaries, whilst beneath it is more uniform in tone and pattern and is much more spotted. ARHOPALA BELPH@BE Doherty. Arhopala belphebe Doherty, J. A. 8. B. vol. lviii. pt. 2, p. 421, pl. xxxiii. fig. 18 (1889) ; de Nicéville, Butt. India, vol. i. p. 272 (1890). Hab. Myitta; Tenasserim Valley ; Burma. Expanse 3 mm. [ have not seen this insect, but from the figure and description it evidently belongs AMBLYPODIA GROUP OF THE LYCANID#. 143 to this section. The resemblance to “ agesias” I cannot quite follow, as that is a very different-looking and strongly-marked species. _ ARHOPALA DUESSA Doherty. Arhopala duessa Doherty, Journ. A. 8. B. vol. lvii. pt. 2, p. 419, pl. xxii. fig. 6 (1889) ; de Nicéville, Butt. Ind. vol. iii. p. 271 (1890). Hab. Myitta; Dannat Range; Tenasserim. Expanse 36-38 mm. This species is near epimete Stgr., but the brilliancy of the blue of the upperside will at once separate it from that insect. ‘The underside of Staudinger’s species is so close to this that I believe the one is probably a local form of the other. ARHOPALA AMMONIDES (Doherty). (Plate III. fig. 31, ¢.) Acesina ammonides Doherty, J. A. S. B. vol. Ix. pt. 2, p. 34. n. 5 (1890). Hab. 'Tenasserim. Expanse, ¢ & 2, 30-31 mm. This species is very close to dwessa Doh. in the colour of its upperside, but is paler, and it has a tail, which duwessa has not; beneath it is deeper in general tone, whilst the light parts are whiter, and it has the marked whitish spot on the costa of the secondaries. It forms an excellent link from the tailless group we have just been considering with the tailed group named by Moore Acesina. 1 do not see my way, however, to retain this genus, as the characters given are quite inconstant in the various species that should form the subgenus, and would apply equally well to others widely separated from these. ARHOPALA AMMON (Hewitson). Amblypodia ammon Hewitson, Cat. Lycenide B.M. p. 9. n. 4, pl. v. figs. 49, 50, 9 (1862). Narathura ammon Distant, Rhop. Malay. p. 272. n. 19, woodcut n. 83, 9 (1885). Arhopala ammon de Nicéville, Butt. Ind. vol. in. p. 243 (1890). Arrhopala ammon de Nicéville and Martin, J. A.S. B. vol. Ixiv. pt. 2, p. 470 (1895). Hab. Malacca; Sumatra. i Expanse, ¢ & 9, 32-35 mm. This species may be easily separated from mnmonides Doh. by the lilac tone and the narrow borders of the primaries, and also by the uniform colour of the under surface. ARHOPALA MYRZALA Hewitson. Amblypodia myrzala Hewitson, Ul. Diurn. Lep., Lyc. p. l4c, pl. 11. figs. 41 & 42, g (1865). Hab. Mindanao; Philippine Islands. Expanse, ¢ & 2, 42 mm. 144 MR. G. T. BETHUNE-BAKER: A REVISION OF THE This species, from its size and general appearance, looks scarcely in keeping with its near neighbours, being about half as large again, but the underside pattern is so very similar that I cannot piace it elsewhere; it may at once be recognised by its large size. ARHOPALA BIRMANA (Moore). Panchala birmana Moore, Proc. Zool. Soc. Lond. 1883, p. 531. Arhopala birmana de Nicéville, Butt. Ind. vol. iii. p. 274 (1890). Acesina arisha de Nicéville, Journ. Bomb. Nat. Hist. Soc. vol. vi. p. 869 (1891). Arhopala birmana Grose-Smith, Rhop. Exot., Lyc. pt. 44, pl. xv. figs. 1, 2, 3 (1898). Hab. Toungu; Burma; Tenasserim. Expanse, ¢ 37-40, 2 32-40 mm. The width of the borders is variable. Specimens from Tilin Yaw (Burma) have the borders decidedly wider than those from Tenasserim, but in all cases they are much wider than in aberrans de N.; the blue is also darker and the size decidedly smaller. From paraganesa de N. it may be distinguished by its larger size, its darker and very much more extensive blue. I have the type specimens now before me, and they agree precisely with those of arisba de Nicéville. Referring to the original descriptions of this species, I have not been able to examine the genitalia of them, but I have no doubt in my own mind that the specimen with the white spot, described as a male, is the female, and that the one with the darker and larger blue area is the male which was described as the female; the same remark also applies to both the text and the plate in Mr. Grose-Smith’s ‘ Rhopalocera Exotica.’ ARHOPALA ABERRANS (de Nicéville). Acesina aberrans de Nicéville, Journ. A. 8. B. vol. lvii. pt. 2, p. 279. n. 8, pl. xiv. figs. 3 3, 4 9? (1888) ; de Nicéville, Butt. Ind. vol. iii. p. 281. n. 847, pl. xxv. figs. 142 ¢, 143 ¢ (1890). flab. Upper ‘Tenasserim ; Tilin Yaw (N. Burma). Expanse, ¢ & 2, 40-44 mm. This species on the upperside is very distinct, but on the underside it is almost an exact copy of paraganesa de N., only very much larger. It is also larger than birmana Moore, and of a much paler and rather brighter blue, whilst the female has a very marked patch of white at the upper extremity of the cell of the primaries. ARHOPALA PARAGANESA (de Nicéville). (Plate IV. figs. 23 & 23.4.) Amblypodia paraganesa de Nicéville, Journ. A. 8. B. vol. li. pt. 2, p. 63. n. 185 (1882). Panchala paraganesa Moore, Proc. Zool. Soc. Lond. 1883, p. 580. AMBLYPODIA GROUP OF THE LYCENID. 145 Acesina paraganesa Moore, Journ, A. 8. B, vol. liii. pt. 2, p. 42 (1884); de Nicéville, Butt. Ind. vol, 11. p. 281. n. 846 (1890). Hab. Northern India; Nepal; Sikkim; Assam; Java (Gede) (Hruhstorfer). _Expanse, ¢ & 2, 33-55 mm. This species has, I believe, been hitherto confined to North-eastern India, but recently Herr Fruhstorfer, of Berlin, has sent me some specimens of Arhopala to determine, among which isa single male of this species taken on Gede Mountain in Java; it is not, however, a typical specimen, the underside being more like ammo- nides in colour and general appearance—it is, however, most probably this species. The genitalia are quite typical of the genus Arhopala, with the clasps rather larger. than usual; the penis is bent in the middle. ARHOPALA ZEPHYRETTA (Doherty). Acesina zephyretta Doherty, J. A. S. B. vol. Ix. pt. 2, p. 33. n. 3 (1891). Hab. Margherita; Upper Assam ; Cherra Punji. Expanse, ¢ & 2,50 mm. 3 & 9. Upperside: both wings dark brown; primaries only with a very small diffused patch of pale blue extending but little over the cell. Under surface with the markings just like paraganesa de Nicéville, but the general tone is much more uniform without the strong white contrasts usually present in that species. This is probably only a local form of de Nicéville’s species. ARHOPALA ARIEL (Doherty). (Plate III. fig. 24, 3.) Acesina ariel Doherty, J. A. 8. B. vol. lx. pt. 2, p. 33. n. 4 (1891). Hab. Margherita; Upper Assam; Khasia Hills; Cherra Punji. Expanse, ¢ 31, 2 35 mm. g. Upperside: both wings very deep lustrous bluish purple; primaries with the purple colour having a strong, oval, well-defined outer margin, leaving the apex and the anal angle exceedingly broadly black, the outer margin less broadly so, and the costa up to the end of the cell rather narrowly black ; in the secondaries the purple has the same oval outer margin, but only extends to just beyond the cell. ‘Tail blackish, tipped with white; fringes dark, with scarcely a trace of light edging. Underside: both wings dark brown tinged with purple, with a slight lilac dusting. Primaries with the purple tinge confined to the costa and outer margin and the lilac dusting to the apex ; all the spots are precisely as in paraganesa de N., with the exception of the fourth spot of the transverse band which is shifted strongly outwards, and there is but little white on the interior of this band. Secondaries with all the markings precisely as in paraganesa de N., but the ground is very dark, strongly tinged all over with purple, with a slight lilac dusting especially near the apex, and a prominent white spot between the two outer spots on the costa. VOL. XVII.—PART I. No. 19.—August, 1905. U 146 MR. G. T. BETHUNE-BAKER: A REVISION OF THE 2. Upperside: both wings black ; primaries with avery small, pale blue, basal area ; secondaries with the least trace of a pale blue dusting close to the base, very much less in each case than in paraganesa de N. Underside: both wings darkish brown, with a slight violet tint and no hlac dusting; otherwise just like the male. This species is quite distinct from paraganesa de N.; the upperside is quite different, being very much darker blue, whilst the secondaries have a very much less extent of blue ; beneath it is fairly uniform in tone, with very little white indeed, almost confined to the costal spot of the secondaries. ARHOPALA GANESA (Moore). (Plate IV. figs. 24 & 24 a.) Amblypodia ganesa Moore; Horsfield and Moore, Cat. Lep. Mus. E.I.C. vol. i. p. 44. n. 66, pl. ia. fig. 9 (1857). Panchala ganesa Moore, Proc. Zool. Soe. Lond. 1882, p. 252; Doherty, Journ. A. 8S. B. vol. iv. pt. 2, p. 126. n. 121 (1886). Arhopala ganesa Staudinger, Ex. Schmett. p. 281, pl. xevi., ¢ (1888); de Nicéville, Butt. Ind. vol. iii. p. 282 (1890). Hab. Western Himalayas. Expanse, ¢ & 2, 28-31 mm. ‘This species may be recognised from all its neighbours by-the entire paleness of the underside of the secondaries; the primaries are also very much paler than any of its allies. he underside pattern is close to this subgroup of the genus, to which it is more nearly related than to dodonea Moore, near which it has usually been placed. De Nicéyille says (/. ¢.) it appears to be confined to an altitude of 5500 to 6500 feet in forests. It has a wide range, occurring in Kashmir on the west to Kumaon on the east, always in oak-forests, where it is rather common. The genitalia are similar to those of paraganesa de N., but the clasps are of a different shape, and the curve of the penis is nearer the base. ARHOPALA LOOMISI Pryer. Arhopala loomisi Pryer, Rhop. Nih, p. 11, pl. ii. fig. 15 (1886). Arhopala ganesa var. seminigra Leech, Entomologist, xxii. p. 44. Arhopala ganesa var. loomisi Pryer ; Leech, Butt. China, pt. 11, p. 848, pl. xxx. fig. 11 (1893), Hab. Japan ; Chang-yang (Central China) ; Chow-pin-sa (Western China). Expanse 32 mm. ‘“* Male. Primaries blue, paler beyond the cell ; costa and outer margin deeply bordered with black; the discoidal spot, which is within the costal border, is more intensely black. Secondaries black, suffused with blue, most pronounced along the central area. Fringes fuscous grey. Under surface (primaries) pale brownish grey, with four whitish wavy-edged darker bands, the first two are basal, the third is abbreviated, the fourth broad and the only one which is clearly defined; submarginal line slender and AMBLYPODIA GROUP OF THE LYCANID#. 147 interrupted: secondaries greyish brown, with four interrupted bands, these are of the ground-colour, but darker at the edges, and bordered with whitish; brownish sub- marginal line wavy, bordered with whitish, and followed by a few brownish linear dots on the margin; fringes pale grey, preceded by a darker line. Expanse 32 mm. “At the time I described the above form of A. ganesa, I was not acquainted with Pryer’s loomisi; I now find that the latter and my var. seminigia are identical ; as, however, Pryer’s name is the earlier one, it must stand for the Japanese and Chinese form of the Indian A. ganesa. ‘The type loomisi was taken at Kanozan, in Kadnoa, Japan; I received one example from Chang-yang, Central China, and another from Chow-pin-sa, Western China.” —Leech, 1. ¢. I have not seen a specimen of this insect. ARHOPALA ARYINA (Hewitson). Amblypodia arvina Hewitson, Ill. Diurn. Lep., Lyc. p. 12, pl. il. figs. 16 & 17, ¢ (1865). Arhopala aboe de Nicéville, Journ. Bomb. Nat. Hist. Soc. vol. ix. p. 281, pl. O. fig. 26, ¢ (1895). Hab. Java, Assam, 'Tenasserim. Expanse, ¢, 39 mm. 3. Upperside: both wings intense ultramarine-blue ; primaries with very fine black margins; secondaries with the costa broadly, posterior margin very finely black, abdominal fold blackish. Underside: both wings brown, suffused with lustrous purplish, entirely as to the secondaries, but only as to the upper two-thirds in the primaries; spots browner, with paler purplish edgings. Primaries with three discal increasing spots, the first two obscure, below which the submedian area is darker, below the third there is another spot; transverse band of moderately even width, outwardly oblique for the first four spots, followed by two spots below each other, shifted inwards; the band continuous, a small detached spot beneath the band is below the submedian nervule; submarginal lunular row rather obscure, followed by a marginal interrupted narrow band. Secondaries entirely purplish with reddish spots, four near the base, followed by three rather larger ones; cell closed by an elongated curved spot, below which is a triangular one; on the costa are two spots below each othex, the lower one being much larger than the first and touching both the elongated spot and the upper spot of the transverse band, which consists of three pairs of spots, each pair being quite confluent, the second pair is shifted inwards, the third pair well inwards, the latter spot of which is almost divided into two and touches the internal nervure ; submarginal row lunular, but rather indetinite. No tail. ?. Upperside: both wings pale sublustrous azure-blue; primaries with a broad brown costa, very broad outer margin, and yet broader apical area; secondaries with very broad margins all round, the costa being the broadest. Underside like the male, but much paler and not so rich in colour; pattern just the same, but inclined to be rather obscure. 148 Mk. G. T. BETHUNE-BAKER: A REVISION OF THE The male of this species is a beautiful insect, the intense richness and depth of the blue being very striking. I have one specimen from Java, in which the reddish purple of the underside is entirely replaced by a curious subdued steely bronzy green, something like the colour occasionally found in meander Boisd. In de Nicéville’s collection are two males and one female from ‘lenasserim, and in Staudinger’s collection there is one male from Assam. The female has not hitherto been described, but Herr Fruhstorfer has sent mea fine pair from Java for identification and I have described the female sent by him. ARHOPALA ADALA de Nicéville. Arhopala adala de Niceville and A. adulans de Nicéville, Journ. Bomb. Nat. Hist. Soe. vol. ix. pp. 282-284, pl. O. figs. 27& 28 g, 29 @ (1895). fab. Dannat Range and Ataran Valley; Tenasserim; Burma. Expanse, ¢ 43-44, 9 44-46 mm. I have very carefully examined this species and compared it most particularly with A. adulans described by de Nicéville at the same time, and I cannot believe that they are other than one and the same species. ‘The main differences stated by him are that the latter is darker and less reddish, and lacks the small costal spot over and beyond the third cell-spot. Now I find that in all species that have a reddish, or a reddish tinged with lilac or violet, underside, the colour is very liable to vary, and except in very fresh specimens the lilac or violet colour is apt to evanesce and leave a merely reddish hue behind ; then as to the lack of the costal spot, in one of de Nicéville’s specimens of adala there is a very small spot present on the costa of one wing, and I have noticed in this group of insects that this particular spot seems given to occasional disappearance, This species is very close indeed to arvina Hew., and will come next to it, indeed so close is it that on first examination I quite thought it was the female of that species, but closer inspection revealed that some of the specimens were males. ARHOPALA BUDDHA, n. sp. (Plate III. fig. 25, 3.) Hab. Java. Expanse 40 mm. 3. Upperside: both wings pale azure-blue; primaries with a broad black costa, broader outer margin, and yet broader apical area; secondaries with very broad margins all round, the costa being rather the broadest. ‘Tail short, fairly stout, tipped with white; at the end of the cell in the primaries is a distinct black spot invading the blue. Beneath both wings fairly dark purplish lilac, with rather darker spots tinged with red. The pattern is just the same as in adada de Nicéville, except that there is a distinct bend in the first four spets of the transverse band of the primaries and a very slight fracture below the fourth spot; there is also a slight trace of some submetallic scales at the anal angle. AMBLYPODIA GROUP OF THE LYCANIDA. 149 { have described this species from an example taken by Herr Fruhstorfer on Gede Mountain in Western Jaya at an altitude of about 4000 fect. _ ARHOPALA Ca&cA (Hewitson). Amblypodia ceca Hewitson, Il. Diurn. Lep., Lye. p. 14, pl. iv. fig. 28, J (1865). Arhopala ceca Druce, Proc. Zool. Soc. Lond, 1895, p. 591. Hab. Borneo. Expanse, ¢ & 9,40-42 mm. 3. Upperside: both wings submetallic violet-lilac, ina side leght more distinctly lilacine ; primaries with exceedingly fine linear brownish margins; secondaries with the costa broadly brownish and the posterior linear margin exceedingly fine. Under- side: both wings brown, with a somewhat bronzy lustre, with darker spots and fascice palely edged. Primaries with a minute spot beyond the centre of the cell, which is closed by a larger one; transverse fascia oblique and curved, ending on the lower median nervure, this fascia is quite broad at its end but tapers upward to a very narrow line at the costa; submarginal row very obscure; subdiscal area dark, two semi-confluent small dots are beneath the origin of the lower median nervule ; submedian area very slightly paler. Secondaries with all markings much darker, four very minute somewhat indis- tinct basal dots, followed by three spots below each other, the upper two being very small indeed, the lower one larger; cell closed by an elongated good-sized spot, from which up to the costa runs an equally broad straight stripe shifted slightly outwards and also slightly outwardly oblique; transverse band from the upper discoidal to the internal nervure composed of two very small spots, the third spot is larger and shifted inwards, fourth larger still and shifted outwards, fifth right inwards and confluent with the long sixth spot; anal angle witha small dark dot followed by a very slight brilliant ereen-metallic scaling, over which is an irregular whitish stripe. ail short, very stout, blackish, tipped with white; fringes scalloped with white on each side of the tail. @. Upperside: both wings purplish blue; primaries with a blackish spot closing the cell and a broad black costa, yet broader outer margin, and broader still apical area ; secondaries with extremely broad margins all round, the costa being the broadest. Underside like that of the male. This species is entirely distinct from any others that I know and appears to be confined to the island of Borneo. ARHOPALA CANULIA (Hewitson). (Plate III. fig. 26, 3.) Amblypodia canulia Hewitson, Ill. Diurn. Lep., Lye. p. 14f. n. 91, pl. ie. fig. 54, ¢ (1869). Hab. Philippines; Batchian. Expanse, ¢ & ?,37—-39 mm. g. Upperside: both wings deep purplish, with very fine black linear margins ; in the 150 MR. G. T. BETHUNE-BAKER: A REVISION OF THE secondaries the costais broad. Underside: both wings brownish grey. Primaries with a narrow, even, obscure, transverse band to the lower median nervule ; submarginal band darker and broadish at the anal angle, tapering away below the apex. Secondaries with an obscure narrow spot closing the cell and another obscure one below the cell; transverse band distinct, moderately even, lightly edged interiorly, but the outer paler edging diffused; submarginal row obscure, palely edged in the anal area; a small black spot at the anal angle, a larger one between the second and lower median nervule, the space between and over which is very pale bluish white. 2. Upperside: primaries blackish, with a paler and less purple-blue patch from the discoidal to the submedian nervule and about a third over the median area. The secondaries are bluish from the upper discoidal to the submedian nervure with a broad dark margin, all the nervures showing broadly black. Underside as in the male, but with all the markings very much obscured. This is a very distinct species and is recognisable at a glance; it is very near fudla Hewitson, the distinctions being recorded under that insect. ARHOPALA FULLA (Hewitson). Amblypodia fulla Hewitson, Cat. Lyceenidee B.M. p. 10. n. 47, pl. vi. figs. 67, 68, 3 (1862) ; Hewitson, Ill. Diurn. Lep., Lyc. p. 11 (1865). Hab. Boiron; Andamans; Amboina. Expanse, 6 37, 2? 36 mm. ¢. Upperside: both wings brilliant shining violet-blue; the primaries with an exceedingly narrow black costa and narrow black posterior margin; secondaries with the costa broadly and the posterior margin narrowly black, abdominal fold greyish, a slight tooth-like projection at the end of the lower median nervule. Underside: both wings pale, slightly rufous brown. Primaries paler in the submedian area, spotless, with a narrow transverse band to the lower median nervule somewhat obscure ; submarginal and marginal areas rather darker. Secondaries with an exceedingly obscure spot in the cell and a very narrow dash closing it; beyond this a broadish band of darker shading edged indefinitely rather palely, the exterior edging broadly pale; marginal and submarginal areas darker, divided at the anal angle by a pale curved line; a very small black dot at the extreme anal angle. @. Upperside: both wings paler and less violet than in the male; primaries with costa and posterior margin broadly and apex very broadly blackish ; secondaries with costa very broadly and posterior margin narrowly black. Underside as in the male. This species may be known from canulia Hewitson by the bluer and brighter colour of the upper surface and by the darker, redder, and warmer tone of the under surface. Canulia has a very cold greyish under surface and has also distinct black anal spots with metallic scaling, which are absent in this species. AMBLYPODIA GROUP OF THE LYCANIDA, 161 ARHOPALA (ANDAMANICA) W.-M. & de N. (Plate IV. figs. 25 & 25a.) Amblypodia (? Narathura) fulla var. andamanica W.-M. & de N. Proc. A. 8S. B. 1881, p. 143. n. 6. Narathura subfasciata Moore, Trans. Ent. Soc. Lond. 1881, p. 312. Arhopala andamanica de N. Butt. India, vol. 11. p. 267, pl. xxvii. fig. 199, ¢ (1890). Hab. Andamans. Expanse 40 mm. $. Upperside: both wings bluer than in A. fu/la Hew., and with less violet in it ; the costa and posterior margin more broadly and the apex much more broadly black in the primaries; the costa is also shorter, the apex sharper, and the posterior margin rounder than in fw/la Hew. Underside as in A. fu/la, but ochreous instead of slightly rufous. 2. Upperside: both wings paler and bluer than in the male ; primaries with a broad black costa, very broad outer margin, and very much broader apical area; secondaries with a very broad dark costa and rather narrow outer margin. This species is quite distinct from fwl/a Hew., and is a much smaller insect. The genitalia have the tegumen-hooks stoutish and the clasps are heavy and broad, terminating in short upturned points. ARHOPALA WILDEI Miskin. (Plate I. figs. 18 ¢,19 9.) Arhopala wildei Miskin, Ann. Queensl. Mus. no. 1, p. 71 (1891). Amblypodia cupido Staudinger, in litt. Hab. Queensland. Expanse 44—45 mm. ¢. Upperside: both wings sublustrous azure-blue of a roughish texture, with narrow black costa to the primaries and wide outer margins increasing towards the anal angle; secondaries with broad dark greyish costa, white above the costal nervure, and wide black outer margins increasing at the anal angle; tail black, tipped with white ; abdominal fold pale greyish. Underside: both wings white with brown spots and fascia. Primaries with costa broadly brown, extending between the discal spots, which thus show a white margin; two spots in the cell and one closing it, beyond which are three very oblique (outwardly) spots, forming a short very oblique row from the third subcostal nervule; submarginal row very distinct at the anal angle, tapering finely into obscurity at the apex ; submarginal band continuous. Secondaries with three basal spots followed by three rather larger ones below each other ; submarginal row distinct and lunulated; margin pale brown, moderateiy broad, lunulated as to its inner edge; extreme margin finely white; at the anal angle is a black spot, a very slightly metallic marginal dusting is between the lower and submedian nervures. 2. Upperside: both wings dark brown, with a large snow-white patch in the 152 MR. G. T. BETHUNE-BAKER: A REVISION OF THE primaries extending over the discal, submedian, and two-thirds of the median areas. In the secondaries the white patch is confined by the upper discoidal and submedian nervures and is not in the cell, except at its outer extremity, where it is closed by a dark spot. At the base of both wings there is a bright pale azure-blue scaling, extending half over the cell and more than halfway along the inner margin of the primaries, but almost confined to the cell in the secondaries. ‘Tail brown, tipped with white. This is a very distinct species and should be recognisable without the least difficulty. Dr. Staudinger sent me one female with his collection, and I found three males in . the Tring Museum, these being all the specimens I have seen. ARHOPALA DISPARALIS (Felder). Amblypodia disparalis Felder, Sitz, Akad. Wiss. Wien, xl. p. 454 (1860); Hewitson, Ill. Diurn. Lep., Lye. p. 12, pl. iv. fig. 25, g (1865). Arhopala disparalis Felder, Reise Nov., Lep. p. 230. n. 256, tab. xxviii. figs. 4 & 5. Amblypodia courvoisieri Ribbe, Iris, vol. xii. p. 335, Taf. vi. fig. 2, ¢ (1901). Hab. Amboina; Ceram. Expanse, ¢, 43 mm. 3. Upperside : both wings pale silvery blue not lustrous, with a lilac shade in a side light; primaries with a fine dark costa and rather wider outer margin; secondaries with a greyish costa not very broad, posterior margin narrow; tails black, white- tipped. Underside: primaries greyish white, becoming greyish brown on the outer marein, with no markings at all except a trace of a submarginal band; the subdiscal area is darker grey. Secondaries whitish, with a dark grey spot closing the cell> transverse band composed of eight dark grey isolated spots, with the exception that the seventh and eighth touch each other ; submarginal and marginal rows distinct, the latter fine; a dark spot on each side of the tail. 2. Upperside: both wings white, with very broad borders all round. Underside like that of the male. This species is near wilde Miskin, but is smaller, with narrower borders, rather different colour, and darkish grey beneath, not white, whilst the spots are also different, The species described under the name courvoisiert by Ribbe (/. ¢.) is without doubt Felder’s insect. ARHOPALA CRITALA (Felder). Amblypodia critala Felder, Lep. Amboin., Sitz, Akad. Wiss. Wien, xl. p. 453 (1860) ; Hewitson, Cat. Lyc. B.M. no. 67, p. 15 (1862) ; id. Ill. Diurn. Lep., Lye. p. 9, pl. iv. figs. 26, 27 (1865). Arhopala critala Staudinger, Exot. Schmett. p. 281 (1888). Hab. Amboina. Expanse, ¢ 48, 2 46 mm. g . Upperside: primaries exceedingly pale azure-blue, with white showing through AMBLYPODIA GROUP OF THE LYCANID”. 153 in the median area; costa and posterior margin very broadly dark brown; the outer margin of the blue is very strongly scalloped. Secondaries with extreme base dark brown, followed by a very broad band of white, extending along the costa to beyond the subcostal nervure and all down the abdominal fold; the rest of the wing to the margin is very dark blackish brown; the cell and about a quarter beyond it and the submedian area for a like distance are exceedingly pale azure-blue; tail dark brown, margin strongly scalloped on each side of it, tipped with white, each of the nervure-spaces from the anal angle to the upper median nervure has a broadish submarginal dash of very pale whitish blue, with which colour the fringes of both wings are tipped. Underside: primaries with the cell, the costa, and the posterior margin very broadly dark brown, the rest of the wing forming a large triangular white area. Secondaries, a broad dash of brown at the extreme base, followed by a very broad transverse band of white; rest of wing to posterior margin dark brown; from the internal to lower discoidal nervure is an exceedingly broad marginal band of very pale metallic greenish blue interrupted by the nervures, between each of which is a long black spot in the blue. 2. Upperside: like that of the male, but with no blue whatever, the blue area of the primaries being white ; the margins are also broader. There is a trace of a very scanty blue scaling at the base of the primaries. Underside as in the male. This is an instance of the very rare case of mimicry in this group, the species mimicking the genus Danais very beautifully. ARHOPALA AXIOTHEA (Hewitson). Amblypodia axiothea Hewitson, Il. Diurn. Lep., Lye. p. 7. n. 21, pl. ii. figs. 10 & 11, f (1865). Amblypodia strophe Grose-Smith, Novit. Zool. vol. iv. p. 366 (1897) ; id. Rhop. Exot., Lycznide, pl. xiv. figs. 6 & 7 (1898). Hab. New Guinea (Hewitson) ; Kapanu, Dutch S.W. New Guinea (Gr.-Smith). Expanse 46 mm. This species is like no other of the group; the broad and peculiar fascia of the underside will separate it immediately from every other species, VoL. XvIt,—ParT 1. No. 20,—Augqust, 1903, Xx Me ee : es AE yee eae Pee 2 HEA 156 ON THE AMBLYPODIA GROUP OF THE LYCANIDA. PLATE I. Fig. 1. Mahathala hainana, 2, n. sp., p. 23. 2. Lraota lazarena, 2, p. 18. 3. Surendra amisena, 3, p. 6. 4, 3 As type 2, p. 6. 5. Pe BY blue 2, p. 6. 6. om vivarna, type 3, p. 6. 7. Arhopala viola, 3, p. 51. 8. : dohertyi, 3, un. sp., p. 64. i) 5 hercules, var. herculina, 3, p. 28. 10. ik kiriwintt, 3, n. sp., p. 38. LiL, : eupolis, 3, p. 41. 12. » padus, 3, p. 46. is}, i blues?: p: 46. 14. rs eridanus, 3, p. 49. 15. 55 tephits, type 3, p. 93. » 6, from Celebes, p. 53. 17. Ps bazaloides, 3, p. 59. 18. 3 wildet, 3, p. 101. 19. 3 AS Oye toile 20. 55 sangird, 3, p. oT. 21. ul helius, 3, p. 58. 22: - », var. anthelius, 3, p. 58. 23. ns myrtha, 3, p. 61. 24. ms mindanensis, 3, 0. sp., p. 62. Zoe be myrtale, 3, p: 63. AD. 3 quercoides, 3, p. 68. 27. = phenops, 3, p. 68. 28. e yauensis, 3, p. 80. 29. eo brooke’, 3, p. 84. 30. pseudomuta, 3, p. 87. Mintern Broe,imp. s del.et chromo W Purkis EASE iy OF ARHOPALA AND ALLIED GEN] HClES SP PATE i, 158 ON THE AMBLYPODIA GROUP OF THE LYCEHNIDA. PLATE II. Fig. 1. Arhopala vihara, 3, p. 66. 2 malayica, 3, p. 71. 3 BA nicévillei, 3, p. 13. 4. i nobilis, 3, p. T4. Dees! a, 195 op O75 ps Ga: 6 5 moolaiana, 3, p. 89. 7 Pe amphimuta, 3, p. 90. 8 be inornata, 3, p. 91. i). #5 epimuta, 3, p. 85. 10. st agestlaus, 3, p. 93. hile 35 catori, 3, p. 99. 12. 5 davaoana, 3, p. 96. 13. * gunongensis, 3, p. 97. 14. s brahma, 3, p. 99. 15. m hypomuta, 3, p. 99. 16. Be oberthuri, 3, p. 101. iy. 5 adonias, 3, p. 102. 18. re wreqularis, 2, p. 111. 19. S eumolphus, 3, ex Darjeeling, p. 104. S e . S 3, ex Java, p. 104. 2 2. 5 2 3, ex Khasia Hills, p. 104. 22. F s os %) Ds LOLs 23. is a 2, type form, ex Sikkim, p. 104. 24, s * 2, pale blue, ex Palawan, p. 104. 25. AA staudingeri, 3, p. 109. 26. o horsfieldi, 3, p. 107. 27. fs basiviridis, 3, p. 108. 28. 4 hkuhni, 3, p. 115. 29. mo om LIES, 30. ; morphina, 3, p. 117. ManternBros.imp W.Purkiss del.et chromo SPECIES OF ARHOPALA AND ALLIED GENERA. (PLATE 1 ‘ yy) JS 160 Fig. Wy, 2 3 4, 5. 6 7 8 s) 10. ON THE AMBLYPODIA GROUP OF THE LYCANIDA, PLATE III. Arhopala chinensis, 3, p. 118. bb) 2? corinda, 3, p. 121. enotria, 3, p. 124. edias, 3, p. 124. barami, 3, p. 126. anamuta, 3, p. 128. allata, 3, p. 128. atrax, 3, p. 130. hewitsoni, 3, p. 132. japonica, 3, p. 134 perissd, 3, p. 137. tounguva, 3, p. 138. asopia, 6, p. 138. acron, 3, p. 140. admete, var. waigeoensis, 6, p. 141. epimete, 3, p. 141. acetes, 3, p. 49. anunda, 3, p. 76. 99 3 var., p. 76. + 2, pale blue var., p. 76. w@xone, 3, p. 39. auxesia, 3, p. 80. ariel, $, p. 145. buddha, 3, n. sp., p. 148. canulia, 3, p. 149. argesias, 3, p. 95. hesba, 3, p. 99. viola, type 2, p. Ol. vale Omap Olle ammonides, 3, p. 143. Transco Hoo Pl, AVM IC. Ml Mintern Bros.amrep. WPurkise de) et chromo SPECIES OF ARHOPALA AND ALLIED GENERA. be aa aa i Mi 7 i hoa nie ee uit ny PLATE Te . 21.— August, 1903. = a 162 Fig. ON THE AMBLYPODIA GROUP OF THE LYCANIDZ. or Ot Ora oS bo or oo PLATE IV. 1. Surendra amisena, genitalia profile; la, penis: p. 6. 2. Lraota timoleon, i AT op 1s Ye 3. Amblypodia narada, By gy Oa AL), 4. Mahathala ameria, LEGS | ep | eee 5. Thaduka multicaudata, genitalia profile: 5a, penis: p. : 6. Arhopala hercules, Be a OG op p- ie Be philander, Bs uh (RABIN Dies p- 8. e meander, a " SQ 1b 9. Re amytis, a ss 9 Gs 5 p- 10. a centaurus, 9 ee. LOGS p. Hak Be amantes, ataunietl el egaeynenne. p- 2 A padus, sf ee WAG © op p- 133 . eridanus, es aa AALS YCER Wine p- 14. Ms tephlis, 5 ss melARG Ss 3. p- 15. Bs bazaloides, ne Seen, nen S Hepes p- 16. a annulata, . Sept arenas p- 17. a theba, Ss fo SL eaSen es p- 18. 3 anthore, 3 oe Oia hee, p- 19. 4 aliteus, aa Bee el Diasec. p- 20. a dohertyt, a es SAOKGA sas p- 21. A Japonica, - a PALE aaa a p- 22. 50 paramuta, 35 Ss RackGy ee p- 23. 3 paraganesa, ss Se AACT Oe teers p: 24, - ganesa, a Si eZ ANGKe) a Pp. 25. os andamanica, ae a Oya p- lla GENITALIA OF dARHOPALA AND ALLIED GENERA. " Cn ue ci 164 ON THE AMBLYPODIA GROUP OF THE LYCANIDZ. PLATE V. Fig. 1. Arhopala alce, genitalia profile; 1 a, penis: p. 76. 2 3 anunda, genitalia profile ; 2a, penis: p. 76. 3 ae aroa, a be Bh op p. 85. 4 4 elopura, i ‘ AL Bs p. 86. 5 ¥3 atosia, pe < Gy op 1 6 ¥ epimuta, 4, Ap OG op p. 88. te a agesilaus, ,, Be Uae p. 93 8. a catori, vs - Sica p- 93. 9 a agesias, x a OiGpae wees Pao: 10. re moore, A Sed eORG aise 1 Oh 11 5 deva, a soa MADE Mog 105 Shs 12 x metamuta, genitalia profile ; 12a, penis: p. 100. 18 rs muta, * Ra US op DalOde 14 an eumolphus, ms AS Mies p- 104. 15, 3 basiviridis, ,, - IDG op p. 108. 16. r diardi, a RS NOG, op o> {hile}, WY, = apidanus, Bs *: ioe ee p- 115. 18. - adriana, . os WG op oe JULY, OD a chinensis, - 5 IDG op p. 118. 20. 3 singla, ie e AV Gs op p. 120. 21. A cenea, F 7 Dias eee p..130. 22. A rama, nS as WDA ly oy p. 150 23. A hewitsont, re a DS} hyp p- 132. 24, x alemon, ‘ Ss 24a, ,, p. 135. 25. a dodonea, ms rs PRG, p. 154. Franti Lok SeWel SY I4 be 16 EE TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON (continued). To Fellows. To the Public. ; Senha ; eG sa. is VOLUME XIV. (1896-1898, contaiming 47 Plates) . Price5 5 O 70 *O Parr]. (1896, containmg 2 Plates) . . .. . pa Oe Oa Opie. 0 8 O 5, 2 (S96. contamina Ovblates\ eee) 08 OO. 012 0 eho (USOC contamina OuPlates)s "eee OD Os, 1 Se) » 4. (1897, containing 10 Plates) apg kel ed 0) Os re 2-00 » Dd. (1898, containmg 2 Plates) . .... ep OorGe Opes 0 8 0 » 6. (1898, containing 7 Plates) . . .. . a OSs 50 1 4 0 » 7. (1898, containmg 6 Plates) . ... . yO LES 3 015 0 » 8 (1898, containing 5 Platesand Titleand Index),, 0 9 O O20 VOLUME XY. (1898-1901, containing 52 Plates) . Price5 15 6 714 0 Part 1. (1898, containing 8 Plates) . . . . . 4 Wl Ht) » 2 (1899, containing 5 Plates). . . . . sean OoeOa= 0) 0 12-0 » 8. (1899, containing 4 Plates) . . . . . epee) ee) 012 0 » 4. (1899, containmg 4 Plates) . ... . BoD IDO 012 0 5, Dp. (1900, containing 24 Plates) . . .. . pe 310 0 > 6. (1901, containmg 4 Plates) . . .. . ao OD lead 015 0 » 7. (1901, containing 3 Platesand Title and Index), 0 9 O 012 0 VOLUME XVI. (1901-19038, containing 38 Plates) . Priced5 8 Part 1. (1901, containing 6 Plates) . . . . . aye) alts) 2. (1901, containing 5 Plates),. . .. .» eee Ok abl: 8. (1901, containmg 9 Plates) 2 . . .'.* 3, O15 4, (1902, containing 5 Plates ie eiveteree arin. Bs AO) Sie) 5: (1902) containing «4 Plates) / 9.0935. = 5, 0118 6. (1902, containing 3 Plates) . . . . . 5 015 ic 8. “I oe oe) co (1902, containing 8 Plates) . . ... , Olt (1903, containing 3 Plates andTitleand Index),, 0 13 (rest) Woy top (eae oe) (dey IS) S) SC or OF CF OO & = r Co @ yi len eh fers 1S) Hey Rie) VOLUME XVII. Parr 1. (1903, containing 5 Plates) . . . . . Prcel 2 6... 110 0 CONTENTS. I. A Revision of the Ampiypopia Group of Butterflies of the Family Lycayipm. By Grorcs T. Beraune-Baxer, F.L.8., F.Z.S., &c. (Plates 1-V.). . . page 1 THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Tue scientific publications of the Zoological Society of London are of two kinds—“ Proceedings,” published in an octavo form, and “Transactions,” im quarto. According to the present arrangements, the “‘ Proceedings”’ contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended to be published in the “ Proceedings” by the Committee of Publication. A large number of coloured plates and engravings are attached to each annual volume of the “ Proceedings,” to illustrate the new or otherwise remarkable species of animals described in them. Among such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “Proceedings”? for each year are issued in four parts, in the months of June, August, October, and April, the part published in April completing the volume for the last half of the preceding year. Beginning from January 1901 they form two half-yearly volumes. The “ Transactions” contain such of the more important communications made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are issued at irregular intervals. Fellows and Corresponding Members, upon payment of a Subscription of £1 1s. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1871, if they exceed the value of five pounds. Fellows also have the privilege of subscribing to the Annual Volume of the ‘ Zoological Record’ for a sum of £1 (which includes delivery in the United Kingdom only), but this privilege only holds good if the subscription is paid before the First of December in each year. Such of these publications as are in stock may be obtained at the Society’s Office (3 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, H.C.), or through any bookseller. P. CHALMERS MITCHELL, rN t, 1903. ugust, Secretary. oe ee ee a? eo ree A= ae Lat a" oS. 2 a oP ; an x, - TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Vou. XVII.—Parr 2. LONDON: PRINTED FOR THE SOCIETY, SOLD AT THEIR HOUSE IN HANOVER-SQUARE ; AND BY MESSRS. LONGMANS, GREEN, AND CO., PATERNOSTER-ROW. August 1903. ’ a VLR asomtan Price 18s. ff i. SEI \ c ag A yseO” ee Taylor and Francis, Printers,] [Red Lion Court, Fleststreet 2) “"" —- TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. To Fellows. To the Public. & 5s. d. 25) Be le VOLUME I. (1833-1835, containing 59 Plates). . Price 313 6 . . . 418 O* VOLUME II. (1835-1841, containing 71 Plates). . »,» 400... 5 6 6* VOLUME. III. (1842-1849, containing 63 Plates). . ,, 3 8 3... 411 O* VOLUME IV. (1851-1862, containing 77 Plates). . 4, 6 2 0... 8 2 6+ VOLUME V. (1862-1866, containing 67 Plates). . ,, Ava de aeitee a OL OOF, VOLUME VI. (1866-1869, containing 92 Plates). . ,, 11 5 O . . .15 O OF VOLUME VII. (1869-1872, containing 73 Plates). . ,, 10 4 0 .. .1312 O+ VOLUME VIII. (1872-1874, containing 82 Plates). . ,, 9 8 3 .. .1211 OF VOLUME IX. (1875-1877, containing 99 Plates). . ee Gis See 1G Se O-- VOLUME X. (1877-1879, containing 95 Plates). . ,, 10 0 3 . . .18 7 OF GENERAL INDEX, Vols. I. to X. (1833-1879) . ,, Oe Gig as e010. O VOLUME XI. (1880-1885, containing 97 Plates). . ,, 912 0 .. .1216 0 VOLUME XII. (1886-1890, containing 65 Plates). . ,, DesOe, OR ce Coote 4a VOLUME XIII, (1891-1895, containing 62 Plates). . ,, G8 Bae ee Se sO * No copies of these volumes remain in stock. t Only complete copies of these volumes are left in stock. Continued on page 3 of Wrapper. ia) hs (or Or (aera) Il. Notes on the Transformations of some South-African Lepidoptera. (Continued from the ‘Transactions of the Zoological Society of London,’ vol. xv. pt. 6, April 1901.) By Lieut.-Colonel J. Maucomm Fawcert !, Received January 20, read March 18, 1902. [Puates VI.-VIILI.] SINCE writing my former paper on this subject (which was published in ‘Trans. Zool. Soc. of London, vol. xv. pt. 6, in April 1901), I have been prevented by the exigencies of the Service during the South-African campaign from devoting any time to it until now, as for the greater part of the intervening period I have been constantly on the march with different columns in the interior of the Transvaal. I had, however, one period of leisure during the months of November and December, 1900, and a few days in January 1901, while my regiment was in standing camp at Belfast (a station on the Delagoa Railway), and I was fortunate enough to discover and describe the wingless female of Metarctia meteus (Stoll) after rearing it from the larva, as well as the larve of Phyllalia patens (Boisd.), Estigmene dissimilis Distant, and other larve which feed on the various grasses and plants of the veldt, comprising two new species of both the last-named genera. During this time, however, my two friends in Durban, Mr. J. F. Quekett and Mr. W. Haygarth, who gave me so much assistance when preparing my former paper, were not idle, and the present paper is mainly due to their researches and observations. When I passed through Durban on my way home on sick-leave in April 1901, I found that a fine series of larvee and pupe had been collected by them and preserved for me in spirits, so that Lam now able to figure and describe the transformations of 35 more species. My thanks are also due to Mr. Alfred D. Millar, who kindiy presented me with the living larve of Ambulya postica (Walk.) and Chwrocampa orpheus Herr.-Schaff. * Communicated by Kt. Trimen, F.R.S., F.Z.5. VOL. XVII.—PART Il. No. 1.—Augqust, 1903. 166 LIEUT.-COL. J. M. FAWCETT ON THE OBSERVATIONS ON THE RHOPALOCERA. Family NYMPHALID. Subfamily NyMPHALIN& (Bates). Evuryrena wrarpas (Drury)!. (Plate VI. figs. 1, 2, larva; 3, 4, pupa.) Papilio hiarbas Drury, Tl. Exot. Ins. iii. p. 17, t. xiv. ff. 1, 2 (1782). Larva. Head fuscous brown, surmounted by two long, pale brown, divergent horns bearing branched spines, and surmounted at their apices by a tuft of smaller spines ; body grass-green, with paired subdorsal tubercles, each bearing a tuft of small brown spines ; a lateral line of small diagonal white streaks, one on each somite, coalescing to form a white lateral line on 10th, 11th, and 12th somites ; a fuscous dorsal patch between 6th and 7th somites; legs and claspers green. Feeds on common climbing stinging-nettle (Zragia sp.); the eggs are laid on the young shoots of the nettle, and covered with fine hairs which resemble the bristles of the plant. Mr. W. Haygarth has kindly supplied me with a drawing of this larva. Pupa. Pale flesh-coloured, marked with green dorsal and subdorsal stripes ; suspended by the tail from the stem of the food-plant. The larve of this species and of Byblia ilithyia figured here pupated on 5th Feb., and emerged on 12th Feb., 1899. The imago is common in Maritzburg and Durban in gardens, and is generally seen fo) near streams. BysuiA ILitHytA (Drury). (Plate VI. figs. 5, 6, larva; 7, 8, pupa.) Papilio ilithyia Drury, Ill. Exot. Ins. ii. p. 29, t. xvii. ff. 1, 2 (1778). Larva. Head fuscous brown, surmounted by two long, green, divergent horns, bearing brown branched spines, and surmounted at their apices by a tuft of smaller brown spines ; each somite bears six tubercles, each surmounted by a short brown branched spine ; a pale dorsal line and interrupted subdorsal and lateral lines of darker green streaks between the tubercles ; legs and claspers green. Feeds on common! climbing stinging-nettle (Tragia sp.), the eggs being laid on the young shoots of the plant in a similar manner to that detailed of the larva of E. hiarbas. I am indebted to Mr. W. Haygarth for a sketch of this larva. ' [For descriptions of both larva and pupa, see my ‘South-African Butterflies,’ i. pp. 259-260 (1887). Both Mr. W. D. Gooch and Capt. Harford note the Natalian pupa as green, with black streaks and dots; while Mr. Gooch records a variety of the larva in which all the body, except the spiracular stripe, is blackish.— R. T.] TRANSFORMATIONS OF SOME SOUTH-AFRICAN LEPIDOPTERA. 167 Pupa. Pale brown, with darker dorsal and lateral lines, and with a pinkish suffusion on the wing-covers ; it is suspended by the tail to the stem of the food-plant. The imago is common all over Natal and in the Transvaal. Both of the foregoing larvee present a striking resemblance to the larva of an Indian species, Pscudergolis wedah (Kollar) figured by me in the ‘Journal of the Bombay Natural History Society,’ vol. xi. plate U. figs. 9 a, 96 (1897), the latter larva differing from them only in structure in the absence of the spiny tubercles. CRENIS BOISDUVALI Wallengren!. (Plate VI. figs. 9, 10, larva.) Crenis boisduvali Wallengren, Rhop. Caftr. p. 30 (1857). Larva. Head fuscous; body pale brown laterally, merging into ferruginous dorsally ; a thin black dorsal line; each somite with six small black tubercles, furnished with small black spines, those on the 10th, 11th, and 12th somites well branched ; legs and claspers fuscous. Feeds on inkberry (Exeewcaria reticulata Miill.-Arg.), a very common tree on the Flat, Durban. I am indebted to Mr. W. Haygarth for a drawing of this larva. Pupa. Pale brown with dark fuscous markings. The time occupied by the transformations was 14 days in larval stage, and one week to 14 days in pupal stage. The imago is a common butterfly in gardens on Berea, Durban, where it may be seen flying amongst bushes quite late in the afternoon. CRENIS NATALENSIS Boisduval. (Plate VI. figs. 11, 12, 15, larva; 14, 15, pupa.) Crenis natalensis Boisduval, Delegorgue, Voy. Afr. Austr. ii. p. 592 (1847). Larva. Head and body bright ferruginous, head surmounted by a stout tuft of ferruginous hairs; black dorsal and spiracular lines; each somite bears six small tubercles, on which are planted short spines; legs and abdominal claspers ferruginous. The figures are drawn from sketches made by Mr. W. Haygarth. Pupa. Pale brown with dark fuscous markings; suspended by the tail from the stem of the food-plant. Larva feeds on inkberry (Kacecaria reticulata Miil.-Arg.). The transformations occupy a similar period to those of C. boisduvali. The imago does not, in my experience, seem so common as the latter insect, but its habits are the same. 1 [For a full description of both larva and pupa, see Rev. H. A. Junod, in Bull. Soc. Sci. Nat. de Neuchatel, xx. p. 21 (1892). M. Junod describes specimens observed in the Delagoa Bay district, and notes that the larva is of a greenish tint and gregarious habitually until after its second moult; and that the pupa presents two forms, one pearly green and almost without markings, while the other is brownish grey with intricate black linear markings.—R. T. | Z 2 168 LIEUT.-COL. J. M. FAWCETT ON THE CHARAXES VARANES (Cramer). (Plate VI. figs. 16, larva; 16 @, pupa.) Papilio varanes Cramer, Pap. Exot. i. t. elx. ff. D, E (1779). Mr. J. F. Quekett writes to me as follows :— “The eggs of C. varanes are laid on both sides of the leaves of the food-plant ; they are yellow when first laid, and subsequently become particoloured (yellow and brown) ; they are hatched in five days from date of laying. “The larve in Ist and 2nd skins are green, with black heads or marks. In 5rd moult they are green, with purple-brown mask and white dorsal spots, and bifid tail on anal somite yellow. “The full-grown larva is bluish green with dark green mask, the underside of which is blue; abdomen and claspers bluish white ; body thickly irrorated with minute white spots; on 6th, 8th, and 10th somites are cresceutic masks, white, with the green of the body showing through. Of these full-fed larvee, bred by Mr. Leigh, and brought down to me yesterday, two had three white dorsal marks, and one only two. My larva, in spirit, which is a fellow one to yours, has lost all trace of the white dorsal marks.” In this connection, I may mention that my full-grown larva, in spirit, has no trace whatever of dorsal marks, but agrees otherwise with the above description }. Family PAPILIONIDAS. Subfamily P1eER1N # (Bates). Pieris Gipica Godart. (Plate VI. figs. 17, 18, larva; 19, 20, pupa.) Pieris gidica Godt. Ene. Méth. ix. p. 131 (1819). Larva, Head and body pale bluish green, surface of body somewhat rough. Sub- dorsal paired white spots on each somite, those on the Ist somite being much larger than those on the others ; a bifid tail at the anal extremity. Feeds on the following four plants, viz.:—Capparis citrifolia (Lam.), Capparis quenezit (Sond.), Capparis zeyheri (Turez.), and Capparis corymbifera (BK. Mey.). Pupa. Pale yellowish green, with a fine reddish-brown dorsal line on the thoracic somites ending in a fine point on the head, and two lateral spines of the same colour above the inner angle of the wing-covers; a few blackish spots are sprinkled over the back, ‘These descriptions are taken from coloured drawings by Mr. W. Haygarth. The imago is a common insect in the Durban district during the summer. months, being replaced during the winter, or dry season, by the form abyssinica (Lucas). * (I figured (8.-Afr. Butt. i. pl. i. fig. 6, 1887) an apparently earlier stage of the larva, from a drawing by Mr. J. P. Mansel Weale of a Cape Colony example, with three crescentic dorsal marks. In a description and drawing of a Natal specimen sent me by Mr. C. N. Barker in 1896 the dorsal marks are represented as white, and are only two in number, on the 6th and Sth segments respectively._-R. T.] TRANSFORMATIONS OF SOME SOUTH-AFRICAN LEPIDOPTERA. 169 OBSERVATIONS ON THE H&ETEROCERA. Family SATURNIIDZ. NUDAURELIA BELINA (Westwood). (Plate VI. figs. 21, larva, half-grown; 22, larva, full-fed ; 25, pupa.) Antherea belina Westwood, Proc. Zool. Soc. 1849, p. 41, pl. viii. f. 2. Larva (early stage). Head and body velvety black, covered with pale bluish-white spots placed most thickly subdorsally ; each somite bearing six black spines of moderate length, the subdorsal spines being longer than the remainder; the spines are inclined to the rear, and each spine bears a few whitish hairs near its base; legs and abdominal claspers black; spiracles red. Final stage. In this stage the bluish-white spots have increased to such a degree that the larva has a bluish-grey appearance, the spines and spiracles alone showing out black. Feeds on Rhus longifolia (Sond.), Natural Order Anacardiacee, and other trees at Durban. Pupa. Red-brown with very large antennz-covers, and, like other species of this genus, the transformation is effected underground. ‘The moth is fairly common in Durban, being at least double-brooded. UstA TERPSICHORE Maass. (Plate VI. fig. 24, larva.) Usta terpsichore Maass. Beitr. Schmett. v. ff. 113, 114 (1886). Larva. Head and first somite velvety black, the head having steely-blue reflections, and bearing a small orange frontal spot between the eyes and near its apex. Body: eround-colour castaneous, each somite bearing two broad dark steely blue-black bands, which coalesce dorsally and laterally, and enclose two castaneous spots between them subdorsally ; each somite also bears (with the exception of the first) very small paired tubercles or spines subdorsally, laterally, and spiracularly, the subdorsal spines on the 2nd and 3rd somites being considerably larger than those on the others. Legs and abdominal and anal claspers black, with blue reflections; spiracles black, edged with a fine white line. Feeds on Commiphora caryefolia (Oliv.), Natural Order Burseracee. Pupa. Red-brown, and the transformation is effected underground. Mr. J. F. Quekett, who has reared it, informs me that this species is single-brooded, and that the perfect insect is rather scarce in the Durban district. 170 LIEUT.-COL. J. M. FAWCETT ON THE CopPaXxA FLAVINATA Walker. (Plate VI. figs. 25, larva; 26, pupa.) Copaxa flavinata Walker, Cat. Het. B.M. xxxii. p. 373 (1865). Larva. Head ferruginous; body black, marbled with a series of fine white lines arranged longitudinally and generally parallel to two dorsal white longitudinal lines, which enclose a dorsal black line. Each somite bears six paired bright ferruginous tubercles on paler raised bases, subdorsally, laterally, and spiracularly ; each tubercle bears on its apex five or six reddish hairs capable of inflicting a painful sting if the larva is handled incautiously. Legs, abdominal claspers, and anal somite bright ferruginous; spiracles black, defined by thin white lines. Feeds in companies on a species of Schmidelia (Linn.), Natural Order Sapindacez, and is a night feeder. The transformation is carried out underground, the pupa being red-brown and of the usual form, but furnished with a sharp spur or horn at its anal extremity. The moth is not uncommon in the Durban district. Mr. J. F. Quekett, who has reared the larva, informs me that this species is single- brooded. This larva presents a striking resemblance in structure to the Durban form of the larva of Ludia smilax (Westwood), although the perfect insects are so different in appearance, CirRINA rorDA (Westwood). (Plate VI. figs. 27, larva, early stage ; 25, larva, full-fed ; 29, pupa.) Saturnia forda Westwood, Proc. Zool. Soc. 1849, p. 52. Larva (early stage). Head red-brown ; body, ground-colour pale yellowish brown, with seven longitudinal red-brown lines, the dorsal line being broader and darker than the others; legs and spiracles red-brown, abdominal claspers pale yellowish. A few white hairs are scattered over the body. The later stages of the larva are very different :— Head and ground-colour of body velvety black, crossed on each somite by a silver band of moderate width composed of minute silver spots coalescing together; round the spiracles (which are black) is placed an irregularly-shaped bright yellow spot also composed of minute yellow spots coalescing together; in some specimens the band across the somites is yellow. A few silky-white hairs spring from the cross-bands and the bases of the abdominal claspers, and from small reddish tubercles situated just above them. Legs black, abdominal claspers red. Feeds on Rhus longifolia (Sond.), Natural Order Anacardiacez, Pupa. Dark red-brown or black, and of the usual Saturniid form, The transforma- tion is effected underground. I have not met with many examples of this moth, but it is not uncommon in the Durban district. TRANSFORMATIONS OF SOME SOUTH-AFRICAN LEPIDOPTERA. iL BuNa#A TYRRHENA Westwood. (Plate VI. figs. 30, larva, early stage, enlarged, x $; 31, 2nd stage; 32, 3rd stage; 33, full-fed; 34, pupa.) Bunea tyrrhena Westwood, Proc. Zool. Soc. 1849, p. 51, pl. vin. f 1. Larva. st stage. Head black ; body pale ochraceous, with paired conical humps from 2nd to 11th somite; each hump has a tuft of short black bristles on its apex; a subspiracular row of tubercles bearing tufts of short black bristles, and a tuft of bristles at the anal extremity. Spiracles black, surrounded by a narrow buff-coloured ring, which is again surrounded by a row of semilunular spots; legs and claspers buff coloured. 2nd stage. Wead and body green, the apices of the paired humps orange, with tufts of short black bristles as in the first stage ; spiracles red ; a subspiracwar yellow line bearing a small tubercle, with a tuft of black bristles on each somite; an oblong silver spot on 5th somite (but this character is often absent); legs and claspers green. 3rd stage. Head and body green, merging into light blue dorsally ; the conical humps have disappeared, the tops of the somites being much lower and each crowned by a small yellow tubercle ; four oblong silver spots (two on each side) on the 2nd, 3rd, 5th, 9th, and 10th somites, and two on the 12th somite; a roughish yellow wedge-shaped horn at the anal extremity, with some small red tubercles on it ; abdominal area darker and covered with minute silver spots; legs pale ferruginous ; claspers green. Ath and final stage. Head and body similar in colour to the last stage, being generally pale green above and darker green underneath. In some specimens the light blue of the dorsal area disappears, being replaced by pale yellowish green, in others it remains constant ; the dorsal prominences are much reduced and bear no tubercles on them ; there is usually an oblong silver spot on the 5th somite (but this character is not constant, some specimens having no silver spots at all) ; the minute silver spots on the abdominal area have also disappeared. This larva in the last two stages bears a marked resemblance to the larva of Gynanisa maia (Klug). Feeds on Celtis kraussiana (Bern.), Natural Order Urticacee. Pupa. Dark red-brown or black, of the usual Saturniid form. ‘Lhe transformation is effected underground. A common moth in the Durban district, but I did not meet with it in Mavitzburg. Henucua sminax Westwood. (Plate VI. fig. 35, larva.) Henucha smilax Westwood, Proc. Zool. Soc. 1849, p. 59. Larva, Head red; body, ground-colour pale ferruginous, covered by numerous irregularly-shaped, dark violet-blue spots, surrounded by narrow pale yellow lines; each somite bears six longish red tubercles with black apices, bearing a tuft of five or six short, black, stinging hairs or spines, which inflict a painful sting if the larva is handled carelessly ; legs red; claspers pale ferruginous. 2, LIEUT.-COL. J. M. FAWCETT ON THE I have already figured the larva of H. smilaw in Trans. Zool. Soc. vol. xv. pl. xlix. fig. 6, from the form prevalent in Maritzburg. The larva is extremely variable, as may be seen by a reference to the above-quoted paper, and the form now described is that prevalent in Durban and the coast-belt of Natal. It differs from the Maritzburg form in structure as follows :—(1) The body is much shorter and thicker ; (2) each somite bears six tubercles of some considerable length (as compared with the Maritzburg form), each bearing a tuft of five or six short hairs or spines, whereas in the latter form the tubercles are very small and short, the tuft of hairs is much thicker, and each tuft has four or five long hairs in addition, which are absent in the Durban form. The plan of coloration is the same in both forms, but the hues are darker and more intense in the Durban form. . It is remarkable that the Durban form is almost identical in structure with the larva of Copaxa flavinata, while the Maritzburg form agrees in structure with the larva of the allied species H. delegorquei. A reference to the figures in the Plates will demon- strate the similarity ; in the case of C. flavinata, however, the resemblance ends here, as it effects its transformation underground, whereas the larve of H. smilaa and H. delegorguei spin cocoons, the present species spinning its cocoon on the trunks of trees covered with lichen and other mosses and leaves. The moth is not very common in Durban, where the larva feeds on guava (Pstdium) ; but I met with a good many examples of it in Maritzburg, where the larva feeds on oak and Jasminum pubigerum (Natural Order Oleacez). I think the similarity of the plan of coloration of the form figured here to that figured in my former paper will sufficiently demonstrate that I was right in describing that larva as one of the forms of the larva of HZ. smlaa. Henucus DELEGORGUEI (Boisduval). (Plate VI. fig. 36, larva.) Bombyx delegorguei Boisduval; Delegorgue, Voy. Afr. Austr. i. p. 601 (1847). Larva. Head black; body, ground-colour pale ochraceous, each somite with six paired small tubercles, subdorsally, laterally, and spiracularly rather darker than the ground-colour, aud each bearing five or six short, black, stinging hairs or spines, and also a tuft of longer, pale yellow, silky hairs, which also grow irregularly over the body. A dark fuscous subspiracular line ; spiracles black, with some black spots round them ; legs and claspers reddish. Feeds on Micreglossa mespilifolia (Bth.), Natural Order Compositie, and is a day feeder, and only a few are found on the same plant. When full-fed it spins a cocoon with its hairs on the stem of the food-plant. The moth seems fairly common in the Durban district, but I did not meet with it at Maritzburg. (Sb) TRANSFORMATIONS OF SOME SOUTH-AFRICAN LEPIDOPTERA. li Family EUPTEROTIDZ. PHYLLALIA PATENS (Boisduval). (Plate VI. figs. 38, larva; 39, cocoon ; 40, imago.) Bombyx patens Boisduval; Delegorgue, Voy. Afr. Austr. i. p. 509 (1847). Larva. Head bright red ; body velvety black above, pale fulvous underneath, thickly clothed with chestnut hairs, which are thicker aud redder on the Ist somite over the head; the dorsal hairs are comparatively short and succeeded by two subdorsal rows of comparatively long hairs: in the last moult the body becomes pale fulvous, with a broad black band on each somite; legs and claspers same colour as hairs ; spiracles white. Feeds on the “veldt” grass, and is very common at Belfast, 6500 feet elevation in Eastern Transvaal. ‘The hairs have an irritating effect on the hand when the larva is handled. The larva constructs a sort of grotto on the ground under the leaves of the grass, in which it remains concealed during the heat of the day, emerging and feeding in the cool of the evening. Fig. 38 shows the attitude assumed by the larva while feeding. The larva takes a long time to grow. I noticed the larvee every evening feeding in the grass right through the summer months from November to February, when my specimens began to change into pup; they emerged as imagines in about one month. I think the species must be single-brooded. Pupa. Purplish red-brown. Formed in a reddish cocoon spun out of the hairs of the larva amongst the grass, or on the ground at its roots. The imago is a very common moth in the same locality, and is much attracted by light in the evening. PHYLLALIA FLAVICOSTATA, n. sp. (Plate VI. figs. 37, imago ; 41, larva; 42, pupa.) Female. Head, thorax, abdomen, and legs bright fulvous; legs thickly clothed with bright fulyous hairs on the femur, a wide black band on the tibia, and four narrow black bands on the tarsus; both wings pale cream-coloured, with a fulvous fascia on the costa, clothed with thick fulvous hairs, broad at base, narrowing to a point at apex ; a thin marginal fulvous line; cilia yellowish white; the antenne of the female specimen figured look filamentous, but are seen through the glass to be minutely serrated or pectinated. Larva. Head red; body velvety black, with subdorsal, lateral, and spiracular ereyish-white tubercles, bearing tufts of thick fulvous hairs of moderate leugth with a few longer hairs among them ; spiracles white ; legs and claspers red. Feeds on common “ veldt” grass. VoL. XViI.—Part U1. No, 2.— August, 1903. Lo > 174 LIEUT.-COL. J. M. FAWCETT ON THE Pupa. Red-brown, formed in a cocoon spun with the hairs of the larva. I reared the specimen figured from the larva during January, February, and March in Maritzburg. It differs from the allied species P. patens (Boisd.) in the larval stage in (1) the tubercles bearing the fulvous hairs being greyish white, (2) the hairs being not so long as in that species, (3) the legs and claspers being pinkish red. In the perfect state (1) the ground-colour of the wings is much paler; (2) the tibie and tarsi of the legs are banded with black. Family SPHINGIDZ. Subfamily AMBULICINA. AmBuLyx PosticA (Walker). (Plate VII. figs. 1, larva; 1a, front view of head of larva.) Basiana postica Walker, Cat. Het. B.M. viii. p. 237 (1856). Larva. Wead pale blue and conical (fig. 1 a), bearing two frontal white lines from the cone at its apex to the mandibles; body pale bluish green, cach somite from 4th to 11th bearing a white diagonal streak defined by a dark green streak above it, the last streak ending at the base of the horn; the lateral area below these streaks is pale blue; the 5rd somite has no streak, but the Ist and 2nd somites have a subdorsal white streak ; the whole body is irrorated with’ minute white spots, which become more numerous towards the dorsal area, giving it a whitish appearance; horn pale blue; legs and spiracles reddish ; abdominal claspers blue-green. Feeds on Chetachme aristata (Planch.), Natural Order Urticacee. This larva is much infested by ichneumons, and it is easy to discover on an individual the places where its epidermis has been pierced by the fly. The figure shows the favourite position assumed by the larva when in repose. The transformation is effected underground. The moth does not seem to be very common in Durban, and I have not seen it elsewhere. Subfamily SMERINTHINA. PotyprycHus GRAYI (Walker). (Plate VII. figs. 2, imago; 3, larva, early stage; 4, larva, full-fed; 5, pupa.) Smerinthus grayit Walker, Cat. Het. B.M. vin. p. 249 (1856). Larva (early stage). Head pale green, bearing superiorly a green serrated straight horn and thickly irrorated with minute white atoms; body grass-green, paling into TRANSFORMATIONS. OF SOME SOUTH-AFRICAN LEPIDOPTERA. 175 light blue dorsally, thickly irrorated with minute white spots; a pale yellowish-white subdorsal line and ill-defined pale diagonal streaks on each somite except lst and 2nd; spiracles whitish, defined by brown lines; horn fuscous, long and straight, and thickly irrorated with small tubercles of same colour ; legs fuscous ; abdominal claspers green, anal claspers defined by a black tubercular line. Final stage. Body as in early stage, but the head has no apical horn, and the anal horn is ochraceous, shorter, and curved downwards. Feeds on Cordia caffra (Sond.), Natural Order Boraginee. Pupa. Red-brown, of the usual shape, with a short spur at the anal extremity. The transformation is effected underground. The moth is rare in the Durban district, and I have not met with it elsewhere. Subfamily Cu #ROCAMPINA. CHa#ROCAMPA ESON (Cramer). (Plate VIII. figs. 20, larva, side view; 21, larva, dorsal view ; 22, pupa.) Sphine eson Cramer, Pap. Exot. in. pl. eexxvi. f. C (1779). Larva. Head olive-brown; body reddish ochraceous-brown above, dark red-brown underneath, covered by minute black striz ; a thin dorsal black line; a triangular dark brown or black patch on the 11th somite, the apex of which lies at the base of the horn ; another blackish patch below the horn and above the anal claspers; subdorsal dark brown lines, continuous on the first three somites, and continued faintly along the remaining somites to the base of the horn, their position being defined by the upper edge of the “eye- spots ” on the 4th and 5th somites, and by two small black spots on 5th, 6th, 7th, Sth, and 9th somites, and by the lower edge of the dark triangular patch on the 11th somite ; a dark laterai line on the first four somites, followed by oblique lateral lines on the remaining somites to the 11th, dark brown ; on 4th somite a black ‘“ eye-spot,” with a pale straw-coloured iris defined outwardly by a thin black line; on 5th somite a pale straw-coloured ‘eye-spot,” defined outwardly by a thin black line, thickening to a black crescent above the ‘“* eye-spot”; legs and claspers dark red-brown ; horn concolorous with the upper part of the body; spiracles dark brown. Feeds on common vine, fuchsia, and wild cucumber. Pupa, Pale brown dorsally, dark reddish brown on the breast and wing-covers, minutely mottled with black atoms and marked with ochraceous at the incision of the somites ; a short sharp-pointed horn at the anal extremity; is formed amongst leaves on the surface of the ground. Mr. J. F. Quekett, to whom I am indebted for specimens of this larva and pupa, informs me that in Natal there are two and even: three broods in the year. ‘The larva 2a2 176 LIBUT.-COL. J. M. FAWCETT ON THE feeds on the underside of the leaf, is fully exposed during the daytime, and suffers more from the attacks of ichneumons than the majority of larve he is acquainted with. ‘Three weeks are passed in the larval stage, and ten days in the pupal stage of the first brood, the second brood remaining in the pupal stage throughout the winter. CHROCAMPA ORPHEUS Herrich-Schaeffer. (Plate VII. figs. 6, larva, side view; 7, larva, dorsal view ; 8, pupa.) Cherocampa orpheus Herrich-Schaeffer, Aussereur. Schmett. fig. 104 (1854). Larva. Wead olive-brown. Body: first three somites pinkish grey, with a dorsal, two lateral, and two spiracular longitudinal brown streaks; the upper two broad on 3rd somite, and gradually tapering to narrow at the head; on 4th somite the usual “ eye ”-like spot, olive-green iris with dark brown pupil, inside which are two minute white spots, with one larger white spot above them situated on the iris ; dorsal area of body rich dark brown, with seven V-shaped paie streaks, olive-green, paling to white in the centre, one on each somite, the apex of the V pointing towards the head; below this a broad pale lateral line from 5th somite to horn, white superiorly, olive-green below, succeeded by an irregular, narrow, black spiracular line; ventral area and claspers pinkish grey ; horn pale olive-brown ; spiracles greyish. Feeds on species of orchid, Ansellia africana (Lindl.). The pupa, which is formed in a hollow of the trunk of the tree on which the orchid is growing, is similar to that of Cherocampa osiris, but much smaller, and is remarkable for its elongated snout (or haustellum-cover) and the spur-like projection at the apex of the wing-covers. It is usually of a pale shade of olive-green mixed with pale brown and minute black spots (fig. 8). When the larva is walking the horn moves forwards and backwards. During the daytime the larva generally remains motionless on the brown part of the stem of the food-plant, where its brown colour gives it protection; in the evening it comes out on to the leaves to feed: when resting the first three somites are retracted some distance into the 4th, and the whole larva bears dorsally what may be described as a striking resemblance to the head and neck of one of the smaller “ whip “-snakes, or ‘“ Kraits,” an analogy which I have before pointed out in the case of as they are called in India the larva of Cherocampa osiris. ‘The specimen figured was presented to me as a living larva by Mr. A. D. Millar; it pupated on board ship in the tropics on 15th April, 1901, and passed about one month as a pupa. The moth is found in gardens in Durban, and is considered a rarity. TRANSFORMATIONS OF SOME SOUTH-AFRICAN LEPIDOPTERA. 177 Subfamily SPHINGIN&. Hyoicus JunIPERr (Boisduval). (Plate VII. figs. 9, imago; 10, larva.) Sphina juniperi Boisduval; Delegorgue, Voy. Afr. Austr. i. p. 595 (1847). Larva. Head and body bright grass-green, thickly corrugated ; fine diagonal white lines defined superiorly by thin fuscous lines from 4th to 11th somites, the last line ending at base of horn; below each diagonal line is a triangular bluish space on each somite, which occupies the whole anal somite to base of horn, which is of the same colour; legs, claspers, and ventral area darker green than the dorsal surface of the body. Feeds on Tecoma capensis (Lindl.), Natural Order Bignoniacexe, which is a common shrub on the Natal coast. Pupa, which is formed underground, red-brown, with a proboscis-cover of moderate length closely applied to the breast. The moth is somewhat scarce in the Durban district. Mr. W. Haygarth, who has reared the larva, has kindly furnished me with a coloured drawing of it. NEPHELE VARIEGATA Butler. (Plate VII. fig. 14, pupa.) Nephele variegata Butl. P. Z. 8. 1875, p. 15. Pupa. Pale reddish brown, marbled with minute black spots and strie defining head and wings, &c.; proboscis-cover very long, and in shape resembling that of Cherocampa osiris, being boat-shaped instead of rounded, as is usually met with in the pupe of species of Sphingine. I am able to give a drawing of this singular pupa, which was found in a hole in the trunk of a wild fig-tree in Durban Town Square in August 1900. Subfamily MAcCROGLOSSIN &. Lopnuron Mareinatum (Walker). (Plate VII. figs. 11, imago; 12, larva; 13, pupa.) Darapsa marginata Walker, Cat. Het. B.M. viii. p. 185 (1856). Larva, Head bluish green; body pale yellowish green dorsally, darker green inferiorly ; an indistinct fine dorsal line, a fine fuscous lateral line from 3rd to 10th somites, and beneath it on each somite a whitish triangular area, small on 3rd and 4th somites, but increasing in size posteriorly, and defined inferiorly by a dark green band ; horn pale bluish green ; legs pink; claspers green. Feeds on Psychotria (= Grumilea) capensis (Sond.), Natural Order Rubiacer. 178 LIEUT.-COL. J. M. FAWCETT ON THE The pupa, which is formed amongst leaves on the surface of the ground, is reddish ochraceous with numerous fine dark strie. The moth is uncommon in the Durban district, where the larva figured was reared by Mr. W. Haygarth, who gave me coloured drawings of the larva and pupa. Family SYNTOMID.E. Mertarcrta meteus (Stoll). (Plate VII. figs. 15, male; 16, female; 17, larva; 18, cocoon; 19, pupa.) Bombyx meteus Stoll, Cramer’s Pap. Exot. iv. pl. ecexlvn. f. B (1782). Male. Head orange; body thickly clothed with orange-coloured hairs; antenne long, black, and pectinated; abdomen orange-coloured, with black rings on each somite ; fore wing dark slaty grey, costal area orange-coloured ; hind wing dark slaty grey, paler towards the centre, and with some orange hairs at the base; cilia orange- coloured. Exp. 38-48 millim. Female. Pale ochreous all over; antenne filamentous, of same colour as body; abdomen abnormally large. ‘The wings of the five specimens I bred never grew after emerging from the pupa, and remained undeveloped in a semi-folded condition. In some specimens the abdomen was so heavy that the legs were unable to support it in any position except on the bottom of the box. Larva. Head red; body black superiorly, reddish ochreous inferiorly, each somite bearing about eight tubercles, from which spring six tufts of long and short black hairs, the long hairs paling to grey at their extremities; the two lowest tubercles bear tufts of red hairs; spiracles black ; legs and claspers reddish. The larva figured is only three parts grown. ‘The hairs have an irritating effect on the hand when the larva is handled. When alarmed the larva rolls itself up into a ball. It feeds on species of Gnidia (Linn.) and Lasiosiphon (Fresen.), Natural Order Thymeliacee, also on species of Acalypha (Linn.), Natural Order Euphorbiacee, which are all very common plants amongst the grass of the ‘“veldt” in the Transvaal. Pupa. Red-brown, the female specimens with diminutive wing-covers and abnormally large abdomen ; the transformation is accomplished inside a strong web woven by the larva by means of its hairs and particles of earth, &c. Amongst my specimens the first web was formed about 6th or 7th December, 1900, and first imago emerged 15th January, L901. A common species at Belfast, 6500 feet altitude, in the Kastern Transvaal. TRANSFORMATIONS OF SOME SOUTH-AFRICAN LEPIDOPTERA, 179 Family ARCTIAD. Subfamily ARCTIAN &. ESTIGMENE DISSIMILIS Distant. (Plate VII. figs. 20, imago; 21, larva; 22, pupa in cocoon.) Eistigmene dissimilis Distant, A. M. N. H. (6) xx. p. 198 (1877). Larva. Head ferruginous, with a black A-shaped mark between the eyes; body black, with an indistinct dorsal line of brownish spots, the two spots nearest the anal extremity paling to yellowish; a brownish lateral line and an interrupted pale ochraceous subspiracular line ; each somite bears several tubercles crowned with black hairs of medium length, the spiracular line of tubercles bearing pale ferruginous hairs ; spiracles white ; legs and claspers pale ferruginous. The young larva differs in having the head, legs, and claspers black, and no pale ochraceous subspiracular line. Feeds on Vernonia gerrardi (Harv.), Natural Order Composite. Pupa. Pale red-brown, formed in a web spun amongst the leaves of the food- plant; the imago emerged on Christmas Day, 1900, after a nonage of three weeks. The larva is not uncommon on the “ veldt” at Belfast (6500 feet altitude) in the Kastern Transvaal, but I saw only the one imago which I had reared. ESTIGMENE ATROPUNCTATA, n. sp. . (Plate VII. figs. 25, imago, female; 24, larva, side view ; 29, larva, dorsal view ; 26, pupa in cocoon.) Male. Head and thorax creamy white; antenne black and pectinated; four black spots on the collar; a dorsal black stripe on thorax, and black stripes on the tegulie ; abdomen pale fulvous, with rows of dorsal and lateral black spots: fore wing creamy white, deepening to pale yellow on the costa; the veins white and more or less defined with black (this being a very variable character in different specimens) ; five black spots above and below vein 1 ; a black spot on the outer margin at the end of each vein, and an irregular number of black spots on the costa: hind wing white, with a black spot at the end of the cell. 2 Female. Differs in being much more heavily marked with black, vein | being broadly bordered with black, the costal black spots being much larger, and the hind wing having three elongated black spots in the first median interspace ; antennz filamentous. Exp. 46 millim. Larva. Wead bright polished red; body velvety black in some specimens, dark slaty grey in others, with velvety-black rings of tubercles on the somites; each somite bears eight black tubercles, each surmounted by a tuft of brush-like hairs, dark grey dorsally, 180 LIEUT.-COL. J. M. FAWCETT ON THE but paling to whitish laterally, giving the larva very much the appearance of the peculiar brush used to clean the inside of lamp-chimneys; legs and claspers bright red. ‘The young larva does not differ except in size. Feeds on Vernonia gerrardi (Harv.), Natural Order Composite, and Cyanotis nodifera (Kunth), Natural Order Commelinacee. Pupa. Pale red-brown. Formed in a slight web among the leaves of the food-plant, and very short and thick in proportion to its length. Of my examples the first specimen spun up its cocoon on 4th December, 1900, and emerged as an imago on 12th January, 1901. The imago is very common, coming freely to light, and buzzing round the mess-tent at night at Belfast (6500 feet altitude) in the Eastern ‘Transvaal. The figure represents a rather heavily-marked female. RHODOGASTRIA ASTREAS var. BAURI (Méschler). (Plate VII. fig. 27, larva.) Bombyx astreas Drury, Ill. Exot. Ins. ii. p. 49, pl. xxviii. fig. 4 (1773). Bombyz astreas var. bauri Moschl. Verh. zool.-bot. Ges. Wien, xxxiii. p. 289, pl. 16. f. 2 (1884). Larva. Head and body pale grass-green ; head with a black spot on crimson ground of moderate size at the apex of each eye ; body irrorated with minute yellow spots over dorsal region, and paired subdorsal crimson spots on 4th and 11th somites, and a crimson and black V-shaped mark pointing posteriorly, and enclosing a pale yellowish- green area on 7th and 9th somites; a yellow lateral line from 4th somite to anal extremity ; spiracles crimson ; legs and claspers green. Feeds on ‘‘ Waterboom.” The pupa, which is brown, covered with a white efflorescence like the pupe of some species of Hesperide, is formed amongst leaves on the surface of the ground in a light web. The specimens I saw when passing through Durban in October 1900 were forming pupe, which would indicate that the species is at least double-brooded. The moth is not uncommon on the coast of Natal. Family LASIOCAMPID. MeranastRia ACULEATA Walker. (Plate VII. figs. 29, larva; 30, society of pup in a common cocoon.) Metanastria aculeata Walker, Cat. Het. B.M. xxxii. p. 565 (1865). Larva. Head and body pale pinkish brown, marked by darix chocolate-brown lines and strie; the summit of each somite having an ovate mark of dark striz; the lines being most defined laterally, where two pale pinkish parallel lines run from head to anal extremity; a few sparse hairs are scattered over the dorsal region, and a spiracular TRANSFORMATIONS OF SOME SOUTH-AFRICAN LEPIDOPTERA. 181 line of tubercles each bearing a tuft of fairly long whitish hairs, and a tuft of similar hairs at the intersection of the somites ; legs and claspers pale pinkish; spiracles dark brown. Some specimens are much more heavily marked than others. Feeds on “Natal mahogany,” Trichilia dregeana (Harv. & Sond.), Natural Order Mehiacee. The larvee of this species are “social” in their habits, and feed in companies or societies of a hundred or so on one plant; they feed at night and sit in a cluster in hollows on the trunk of the food-plant during the daytime, when they resemble a stain on the bark. The pupe, which are purplish brown with pale bistre markings on the somites, are enclosed in social cocoons spun amongst leaves on the surface of the ground with the hairs of the larve. The imago is a common moth in Durban, where it has been reared by Mr. J. F. Quekett, who has kindly given me the above information about its habits. TRABALA PALLIDA, n. sp. (Plate VIII. figs. 1, female ; 2, larva, early stage ; 2 a, head of larva, front view; 3, larva, final stage; 4, cocoon.) Male and female. Head, thorax, and abdomen pale ochraceous; a yellow band on collar; antenn pale yellow, pectinated in both sexes: fore wing pale ochraceous ; a yellow spot at end of cell; a narrow, yellow, antemedial band, becoming obsolete at vein 1; an oblique, postmedial, narrow band olivaceous, reaching the inner margin, and continued across the disc of the hind wing: hind wing pale ochraceous, becoming almost white inwardly ; a yellow spot at end of cell; a faint indication subcostally of a lunular submarginal line, becoming obsolescent on first median interspace; cilia pale ochraceous. Underside: fore wing as on upperside; hind wing with a black spot at end of cell, and the costa thickly clothed with coarse, longish, black hairs. Larva (early stage). Head pale yellow, with four black spots, one at the apex, one just above the mandibles, and one on each eye: body, first somite concolorous with head, and bearing three black tubercles surmounted by tufts of blackish hairs, one dorsally and two laterally ; remainder of body pale ochraceous, covered by numerous minute black spots; a dorsal, subdorsal, and lateral line of small black tubercles, each bearing a few longish yellow hairs; a number of shorter yellow hairs scattered over the body; three or four blackish spots subdorsally at the intersection of the somites, where there are some narrow reddish bands on the somites; legs black; claspers ochraceous ; spiracles black. The final stage is very different from the early larval stage; the head remains the same, but the body is deep velvety black, with dorsal, subdorsal, and lateral small black tubercles, each bearing some yellowish-white hairs and a few black hairs ; each somite VOL. XViI.—ParT 1. No. 3.—dAugust, 1903. 2B 182 LIEUT.-COL. J. M. FAWCETT ON THE has two castaneous cross-bands, defined outwardly by pale ochraceous bands, the whole being thickly covered by pale ochraceous hairs ; legs and anal extremity pale ochraceous, marked with black; claspers black ; spiracles black, defined by fine white lines. Feeds on the “wild mango” tree, Selerocarya caffra (Sond.), Natural Order Anacardiacee. Pupa formed in an ochraceous cocoon composed of the hairs of the larva. This species has been reared by Mr. J. F. Quekett from the egg at Durban. Family HYPSID. CALLIORATIS BELLATRIX (Dalman). (Plate VIII. figs. 11, larva; 12, pupa.) ? Bombye bellatria Dalman, Anal. Ent. p. 50 (1828). Larva. Head ferruginous ; body, ground-colour dull yellowish white; each somite crossed by a broad black band or ring bearing small paired subdorsal, lateral, and spiracular tubercles ; legs, claspers, and spiracles steely blue-black ; a few black hairs on the black rings of the first three somites. The young larva differs only in size. Feeds on Crotalaria capensis Jacq., Natural Order Leguminose |. Pupa. Bright polished yellow, with paired black spots on the abdomen subdorsally and laterally, and black marks on the breast and wing-covers; spiracles black. Usually enclosed in slight white silky web on the stem of the food-plant. The moth is not uncommon in Durban. E@YBOLIs VAILLANTINA (Stoll). (Plate VIII. figs. 13, larva; 13a, front view of head of larva; 14, pupa.) Bombyx vaillantina Stoll, Suppl. Cram. Pap. Exot. pl. xxxi. f. 3 (1790). Larva. Head black, marbled with white fascie ; body, each somite pale yellow with subdorsal, lateral, and spiracular pairs of small black tubercles, each bearing a few black hairs, the somites being separated by broad black bands or rings, each band bearing also two fine white rings; the anal extremity black, crossed by two white bands; legs, claspers, and spiracles black. Feeds on Sapindus oblongifolius (Soud.), Natural Order Sapindacee. Pupa. Pale ferruginous, marked with numerous fine black lines defining eyes, antenne, veins on wing-covers, and abdominal somites; spiracles black. Like that 1 [So long ago as March 1858 I reared many larvee of this species at Knysna, in the Cape Colony, on the same food-plant. My description differs in a few points from that here given, as it notes that the bristles on the 2nd, 3rd, and 4th segments (and also on the 11th and 12th) are long and white, and that the tubercles on the black rings are steely blue, and that each bears a single long black bristle.—R. T.] TRANSFORMATIONS OF SOME SOUTH-AFRICAN LEPIDOPTERA. 185 of the last species, enclosed in a slight white silky web formed on the stem of the food-plant. The imago, which is known as the “ Peach Moth,” is common in Durban, and is a day flier. Family NOCTUID~. Subfamily QUADRIFIN&. OPHIUSA INDETERMINATA Walker. (Plate VIII. figs. 15, larva; 16, pupa.) Ophiusa indeterminata Walker, xxxiii. 955 (1865). Ophiusa demta Moschl. Verh. zool.-bot. Ges. Wien, xxxiii. p. 305, pl. xvi. £. 20 (1883). Larva. Head and body pale yellowish green, fading into pale blue dorsally. Head, 1st somite, interspaces between 4th and 5th and 5th and 6th somites marbled dorsally with minute black striz, which also appear in an irregular subspiracular line, most highly developed above the abdominal claspers; two small black tubercles on 11th somite. This larva is what is described as a ‘ half-looper,” from its mode of progression, though it has the usual number of abdominal claspers complete: when at rest the larva appears just like a twig. It feeds on Calpurnia lasiogyne (K. Mey.), Natural Order Leguminose. When full-fed the larva makes a cocoon just above the ground. Pupa. Rather pale reddish brown, covered by a beautiful plumbeous bloom, which I have endeavoured to reproduce in the figure, but, I fear, without much success. RHANIDOPHORA cINcTIGUTTA (Walker). (Plate VIII. figs. 17, imago; 18, larva, natural size; 19, larva, much enlarged.) Enydra cinctiguita Walker, Trans. Ent. Soc. (8) i. p. 77 (1862). Larva, Head ferruginous ; body dull yellowish white (or buff), with large irregularly shaped claret-coloured spots or bands across the somites ; on first somite and anal extremity the claret-coloured spots are small and disposed as in the enlarged figure ; on the 3rd, 4th, 5th, 6th, and 7th somites the spots form a broad irregular band, which encircles the body, while on the 2nd, 8th, 9th, 10th, 11th, and 12th somites these bands are broken up into large irregular spots; on 4th and 5th somites are six long, claret-coloured, spatulate filaments ; four filaments (two subdorsal and two lateral) on the 4th somite, and two (lateral) on 5th somite; a few short reddish hairs are scattered over the body ; legs reddish ; claspers yellowish white. When full-fed the larva makes a cocoon on the stem of the food-plant just above the ground. Pupa red-brown. Feeds on Thunbergia alata (Bojer), Natural Order Acanthacez. bo & bo 184 TRANSFORMATIONS OF SOME SOUTH-AFRICAN LEPIDOPTERA. The number and disposition of the long spatulate filaments seem to vary (as do the claret-coloured spots) in different specimens. ‘The life-size figure (fig. 18) of the larva was drawn by me from a larva at present in my possession, and has only six filaments on the 4th and 5th somites, while the enlarged figure was drawn by me from an enlarged sketch of another individual, and it will be seen that this specimen has, in addition to the filaments already mentioned, four pale whitish filaments subdorsally and laterally on 6th and 7th somites. In ‘ Fauna of British India: Moths, by Sir G. F. Hampson (vol. ii. p. 578), a figure is given of the larva of Tinolius edurneigutta Wllk., which is, so far as one can judge from a sketch, practically identical in structure with the larva of the South-African form ‘‘ cinctigutta.” Family GEOMETRID. PrroviaA DICHROARIA Herrich-Schaefter. (Plate VIII. figs. 9, larva; 10, pupa.) Petovia dichroaria Herrich-Schaeffer, Aussereur. Schmett. f. 189 (1854). Larva. Head dark fuscous; body pale green, marked by narrow, dark fuscous, transverse lines dividing the somites; a dark fuscous thin dorsal line, broadest where it unites with the transverse lines; a thin light brown spiracular line ; legs fuscous black ; claspers pale brown. Pupa. Pale fuscous-brown, with dark fuscous markings. Larva feeds on the “wild medlar,” Vangueria infausta (Burch.), Natural Order Rubiacee. This species was reared at Durban by Mr. W. Haygarth, who has kindly presented me with a coloured sketch of the larva and pupa. EvcuLoris DEvocatTa (Walker). (Plate VIII. figs. 5, imago; 6, 7, larva; 8, pupa.) Geometra devocata Walker, Cat. Het. B.M. pt. xxi. p. 514 (1861). Larva. Bright bluish green, paling to whitish laterally ; a fine dark green spiracular line, with an indistinct white line above it; abdomen, legs, and claspers darker green ; above the head on Ist somite two green tubercles. Fig. 7 shows the larva in the position in which it rests, simulating a stem of the food-plant. Pupa. Pale brown, of the usual shape. Larva feeds on the ‘wild medlar,”’ Rubiacez. This species was reared at Durban by Mr. W. Haygarth, who has kindly presented me with a coloured sketch of the larva and pupa. Vanqueria infausta (Burch.), Natural Order LIAB Wit, 186 TRANSFORMATIONS OF SOME SOUTH-AFRICAN LEPIDOPTERA. PLATE VI. Figs. 1, 2. Larva; 5, 4. Pupa of Ewrytela hiarbas, p. 166. Figs. 5, 6. Larva; 7, 8. Pupa of Byblia ilithyia, p. 166. Figs. 9, 10. Larva of Crenis boisduvali, p. 167. Figs. 11, 12, 138. Larva; 14, 15. Pupa of Crenis natalensis, p. 167. Figs. 16. Larva; 16a. Pupa of Charaxes varanes, p. 168. Figs. 17, 18. Larva; 19, 20. Pupa of Pieris gidica, p. 168. Figs. 21. Larva (half-grown); 22. Larva (full-fed); 23. Pupa of Nudaurelia belina, p69: Fig. 24. Larva of Usta terpischore, p. 169. Figs. 25. Larva; 26. Pupa of Copaxa flavinata, p. 170. Figs. 27. Larva (early stage); 28. Larva (full-fed); 29. Pupa of Cirina forda, p. 170. Figs. 30. Larva (early stage, enlarged); 31. Larva (2nd stage); 32. Larva (3rd stage) ; 33. Larva (full-fed); 34. Pupa of Bunwa tyrrhena, p. 171. Fig. 35. Larva of Henucha smilax, p. 171. Fig. 86. Larva of Henucha delegorquei, p. 172. Figs. 37. Phyllalia flavicostata, 2 ; 41. Larva; 42. Pupa of ditto, p. 173. Figs. 40. Phyllalia patens, 3 ; 38. Larva; 39. Cocoon of ditto, p. 172. Sranalool be. Vob XVI TEV J.M Faweett del West, Newman chromo E C. Knight lth TRANSFORMATIONS OF SOUTH AFRICAN LEPIDOPTERA cel esi) 188 TRANSFORMATIONS OF SOME SOUTH-AFRICAN LEPIDOPTERA. PLATE VII. igs. 2. Polyptychus grayi Walker; 3. Larva (early stage); 4. Larva (full-fed); 5. Pupa of ditto, p. 174. Figs. 6. Larva (side view): 7. Larva (dorsal view); 8. Pupa of Cherocamoa orpheus, 10s JUPAS, Figs. 9. Hyloicus gunipert; 10. Larva of ditto, p. 177. Figs. 11. Lophuron marginatum ; 12. Larva; 13. Pupa of ditto, p. 177. Fig. 14. Pupa of Nephele variegata, p. 177. Figs. 15. Metarctia meteus, 3; 16. Ditto, ?; 17. Larva; 18. Cocoon; 19. Pupaof @ of ditto, p. 178. Figs. 20. Estigmene dissimilis ; 21. Larva; 22. Pupain cocoon of ditto, p. 179. Figs. 25. Estigmene atropunctata, 2; 24. Larva (side view); 25. Larva (dorsal view) ; 26. Pupa in cocoon of ditto, p. 179. Fig. 27. Larva of Rhodogastria astreas var. baw, p. 189. Figs. 28. Metanastria aculeata, 6 ; 29. Larva; 30. Society of pupe of ditto im Figs. 1. Larva of Ambulyx postica Walker; 1a. Front view of head of ditto, p. 174. 8 a common cocoon, p. 180. Sram Loot Foo. Vl XVILSL, Vil RAL mia if Whi VNU) JIM. Fawcett del West, Newman, chromo. E.C Knight, lith TRANSFORMATIONS OF SOUTH AFRICAN LEPIDOPTERA. are Beal . t Or Ete ‘ "ch pee i Ve 7 190 TRANSFORMATIONS OF SOME SOUTH-AFRICAN LEPIDOPTERA. PCA E Valle . Trabala pallida, 2 ; 2. Larva (early stage); 2a. Front view of head of larva; 3. Larva (full-fed); 4. Cocoon of ditto, p. 181. . Huchloris devocata ; 6, 7. Larva; 8. Pupa of ditto, p. 184. . Larva ; 10. Pupa of Petovia dichroaria, p. 184. Larva; 12. Pupa of Callioratis bellatrix, p. 182. . Larva; 13a. Front view of head of larva; 14. Pupa of Egybolis vaillantina, p. 182. . Larva; 16. Pupa of Ophiusa indeternunata, p. 183. . Rhanidophora cinetigutta; 18. Larva (natural size); 19. Larva (much enlarged), p. 183. . Larva (side view); 21. Larva (dorsal view); 22. Pupa of Cherocampa eson, p. L785. SranshLoob Kee. Vel. xvi FE, Vir West, Newman, chromo TRANSFORMATIONS OF SOUTH AFRICAN LEPIDOPTERA. TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON (continued). To Fellows. To the Public. : &£ s. d. £ s. d. VOLUME XIV. (1896-1898, containing 47 Plates) . Price5 5 O 7 0 0 Part 1. (1896, containmg 2 Plates) . . . . . =e On 68-0 . 0 8 0 » 2 (1896, containmg 6 Plates) . . . .. » 0 9 O . 012 0 » 9 (1897, contaming 9 Plates) . . ... » 0 15 9 ae 0) » 4. (1897, containing 10 Plates) . . . . . 4 110 0 Se) > 9. (1898, contaming 2 Plates) . . . .. » O09 6 0 - 0 8 0 » 6. (1898, containing 7 Plates) i tee ty >» 018 O - 140 » 7. (1898, containing 6 Plates) : : » Oll 8 - 015 0 > 8 (1898, containing 5 Plates and Title and Index) ae OF OO - 012 0 VOLUME XV. (1898-1901, containing 52 Plates) . Price5 15 6 714 0 Part 1. (1898, containing 8 Plates) . . . . . » 9015 9 lige eal) » 2 (1899, containmg 5 Plates). .... #4, 09 0 012 0 » 9 (1899, containmg 4 Plates) ..... Ore GeO 012 0 >, 4. (1899, containmg 4 Plates) .... . » 9 9 O 012 0 », 5. (1900, containing 24 Plates) . . . . . rie ces 310 0 », 6. (1901, containmg 4. Plates) see Pees Aes >» O11 38 015 O » 7%. (1901, contammg 3 Platesand Title andIndex),, 0 9 0 012 0 VOLUME XVI. (1901-1903, containing 38 Plates) . Price5 8 O 7 4 0 Part 1. (1901, contaming 6 Plates) . . .. . » 918 O 1 4 0 » 2 (1901, containing 5 Plates) . . ... MUO cd Mee aig 015 O » 9 (1901, containing 9 Plates) . . . .. «4 O15 9 Pele 0 eae 902; containing. 5: Plates) earicger ie, 2.0.2.9) 0 012 0 » 9 (1902, containing 4 Plates) . .... «4 O18 6 - O18 O », 6. (1902, containing 3 Plates) . . . .. «4 O15 9 10 » 7%. (1902, containing 3 Plates) cena age Or Led On Lae 0 6 . 018 O 8. (1903, containing 3 Plates andTitle and Index), 0 13 VOLUME XVII. Panr I. (1903, containing 5 Plates) . .+. . . Pricel 26... 110 2. (1903, containing 3 Plates)... . . On Le 18 -0 for) i=) ”? CONTENTS. Il. Notes on the Transformations of some South-African Lepidoptera. (Continued from the ‘Transactions of the Zoological Society of London,’ vol. xv. pt. 6, April 1901.) By Lieut.-Colonel J. Maucoum Fawcert. (Plates VI—-VIII.). . . page 165 THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Tue scientific publications of the Zoological Society of London are of two kinds—“ Proceedings,” published in an octavo form, and “Transactions,” in quarto. According to the present arrangements, the “ Proceedings” contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended to be published in the “ Proceedings ” by the Committee of Publication. A large number of coloured plates and engravings are attached to each annual yolume of the “ Proceedings,” to illustrate the new or otherwise remarkable species of animals described in them. Among such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “ Proceedings”’ for each year are issued in four parts, in the months of June, August, Oct ober, and April, the part published in April completing the volume for the last half of the preceding year. Beginning from January 1901 they form two half-yearly volumes. The “Transactions” contain such of the more important communications made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are issued at irregular intervals. . Fellows and Corresponding Members, upon payment of a Subscription of £1 1s. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1871, if they exceed the value of five pounds. Fellows also have the privilege of subscribing to the Annual Volume of the ‘ Zoological Record’ for a sum of £1 (which includes delivery in the United Kingdom only), but this privilege only holds good if the subscription is paid before the First of December in each year. Such of these publications as are in stock may be obtained at the Society’s Office (3 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. P. CHALMERS MITCHELL, August, 1903. Secretary. TRANSACTIONS . OF THE ZOOLOGICAL SOCIETY OF LONDON. Vou. XVII.—Parr 3. LONDON: PRINTED FOR THE SOCIETY, SOLD AT THEIR HOUSE IN HANOVER-SQUARE ; AND BY MESSRS, LONGMANS, GREEN, AND CO,, PATERNOSTER-ROW. October 1904. Price 30s. Taylor and Francis, Printers.] {Red Lion Court, Fleet Street. TRANSACTIONS OF THE ZOOLOGICAL SOCIETY — OF LONDON. To Fellows. To the Public. sod. aa yah VOLUME I; (1833-1835, containing 59 Plates). . Price 318 6 . . . 418 O* VOLUME II. (1835-1841, containing 71 Plates). . 5 400... 5 6 6 VOLUME III. (1842-1849, containing 63 Plates). . , 8 8 3... 411 O* VOLUME IV. (1851-1862, containing 77 Plates). . ,, 6 2 0... 8 2 64° VOLUME _ V. (1862-1866, containing 67 Plates). . ,, 5 4 8 O19 OF VOLUME VI. (1866-1869, ee 92 Plates). scr jn Wd 10a lon One Os, VOLUME VII. (1869-1872, containing 73 Plates). . ,, 10 4 0. . .1312 OF VOLUME VIII. (1872-1874, containing 82 Plates). . ,, Sh eee ILS sp VOLUME IX. (1875-1877, aoe 99) Plates) recess lie) lO ter eG eae cp VOLUME X. (1877-1879, containing 95 Plates). . ,, 40 Orr ence cts LO. Ve OF, GENERAL INDEX, Vols. I. to X. (1833-1879) . ,, ORO a. ee O10 30 VOLUME XI. (1880-1885, See 97 Plates). . ,, DA Oo Ge soe IOS IGs 210) VOLUME XII. (1886-1890, containing 65 Plates). . , 5 8 0...7 4 0 VOLUME XIII. (1891-1895, containing 62 Plates) ae 63-8. 38 or 8 dO * No copies of these yolumes remain in stock. 7 Only complete copies of these volumes are left in stock. Continued on page 3 of Wrapper. [ en a III. 4A Monograph of the Fishes of the Family Loricariide. By OC. Tate Reean, B.A., F.Z.8. Received October 13, read November 17, 1903. [Piates IX.—X XT. | ‘THE Loricariidz are a family belonging to the suborder Ostariophysi, and are found only in the rivers of South America, ranging from Panama and Trinidad or Porto Rico to Montevideo. The Ostariophysi resemble the Malacopterygii, the most primitive of Teleostean Fishes, in having the air-bladder, if well developed, provided with a duct, the ventral fins abdominal in position, and a mesocoracoid element present in the pectoral arch; they are, however, distinguished by the modification of the anterior vertebre, which are usually fused, some of their lateral and superior elements forming a chain of ossicles (ossicles of Weber) connecting the air-bladder with the auditory organ. ‘The Ostariophysi are arranged by Boulenger in 6 families—Characinide, Gymnotide, Cyprinide, Siluride, Aspredinide, and Loricariide, the last being equivalent to the Siluride Hypostomatina of Gtinther, with the exception of the genera Callichthys, Sisor, Erethistes, Exostoma, and Pseudecheneis, or to the Loricariide and Argiide of Higenmann. The Loricariide share with the Siluride most of the characters which serve to distinguish that family from the more generalised Characinide, the most important of which are the reduced maxillary, the absent symplectic and suboperculum, and the union of supraoccipital and parietals to form a single parieto-occipital bone ; but they present certain features of still greater specialisation—such as the absence of parapophyses, the sessile ribs, and the compressed caudal vertebre—which warrant their separation as a distinct family. Most of the fishes of this family can be recognised by the armour of bony plates protecting the body, although the degenerate Argiine are naked ; the inferior sucker- like mouth is also characteristic !. It seems not improbable that the Loricariide may have evolved from the Siluridee in the neighbourhood of the Doradina, which they resemble in many respects. It appears that in nature these fishes fasten themselves to stones by means of the sucker-like mouth, whilst in captivity they have been observed to adhere to the bottom or sides of the vessel in which they are placed. lespiration seems then to be effected by taking in water through the gill-openings and expelling it again by the same passages in a reverse direction. VOL. XVII.— PART 11. No. 1.—October, 1904. 2D 192 MR. C. TATE REGAN ON THE The present paper deals mainly with the material in the British Museum, including the types of the species described by Giinther and Boulenger, but the author, during a visit to Paris, has been able to examine all the types of the species belonging to this family which have been described and figured by Castelnau and most of those of Cuvier and Valenciennes. In addition, some of the types of species described by Eigenmann have been received on loan, since the reading of this paper, from the Museum of Comparative Zoology, Cambridge, U.S.A., and descriptions of these are incorporated. The author wishes to express his gratitude to Mr. Boulenger for his ever-ready advice and help; to Prof. Vaillant, through whose kindness he was permitted to examine the specimens in the Jardin des Plantes; to Dr. Pellegrin for his great courtesy and personal attention during his visits to the Paris Museum; and to Dr. Garman, who has sent to the British Museum examples of several species previously unrepresented in its collection, and, in the case of those species which have been described from one or two specimens only, has lent the types. A tribute must be paid to Dr. Steindachner for the excellence and accuracy of his descriptions of so many fishes of this family, which have made it a comparatively easy matter to assign species described by him to their natural systematic position without having seen actual specimens of them. The paper of Prof. and Mrs. Kigenmann on the “South American Nematognathi ” has been of considerable assistance, especially proving of value as a ground-work and for purposes of reference. In the present paper 189 species are recognised as valid, 54 of which are described as new to science. Our knowledge of the geographical distribution of these species is so very incomplete that generalisations are of little value. Most of the genera seem to be represented in the principal river-systems of South America, the rivers south of the La Plata system excepted, and the species seem in many cases also to have a very wide distribution, as might be expected when it is considered that the Amazon system actually communicates with that of the Orinoco on the north, whilst the head-waters of the southern tributaries of the Amazon are in many cases only separated by a few miles from those of the rivers of the La Plata system. Whilst it is clear that the Magdalena, Orinoco, Amazon, and La Plata systems, and also the coast-rivers of Eastern Brazil, have each several characteristic species which do not extend into other river-systems, still in the present state of our knowledge it is difficult to say which are the species with their distribution thus restricted. Two important genera, Chetostomus and Arges, occur only in the Andes of Peru, Ecuador, Colombia, and Venezuela, and are characteristic of mountain-streams in which the other genera of this family are not found; both have been recorded from considerable altitudes, and it would almost seem that each newly explored valley may be expected to furnish some new form. 9 FISHES OF THE FAMILY LORICARIIDA. 195 In the following list, arranged in systematic order, a * indicates the locality from which a species has been recorded. The somewhat artificial divisions which are here established for convenience are:—(1) Western coast-streams of Peru and Ecuador ; (2) Rio Magdalena system (with Panama); (3) Venezuela and Guiana (Orinoco and Essequibo systems with Trinidad); (4) Upper Amazon (Amazon and its tributaries above its junction with the Yapura); (5) Middle and Lower Amazon; (6) Rio Paranahyba and Rio San Francisco and their tributaries, and smaller coast-streams in their neighbourhood; (7) Rio Parahyba, Rio Grande do Sul, and other coast- streams of S.E. Brazil; (8) Rio La Plata and its tributaries. 1 2 3 4. 5 6 7 8 Subfam. 1. Puecostomin», Gen. I. Precostomus Gronow. Subgen. a. Plecostomus. Il, 1, UGH ICAI, Socecannc0dd000000 We te * BA * *% a * 4 ley Gonmmensoarn CCP Ve soc000000000 Aa ve 96 a0 a * * *# Sh le jommcmims OL GP We csoascacoococd i 60 ys 00 o6 * * * 4. PB. spinosissimus Sidr. .-...2..--..-- Ra Gp Io testes JGR, odsocecng0s000000000 % GHD VELECSRGL Ae Ves ox. craray vererehopeno io she a Se * 30 * To Wo Caminenins SHER coacbonocon00o de pare ae as Fa &. 1B velllenntin Sua, coadeccooonnnd00u8 ec ee O06 oy: oe * . * @), 1% Grima OL 6? 1% coooooagcosec ae x *% * x OE aalatussC astel nema 50 50 O0 Se * * Il, 12 Galore Cut socccocccons000000 o0 D0 a¢ O10 ad. hoe ae * | 1), 1h, [oonellint ii so0c0cp0da00d0000c O° 60 00 O06 00 36 Be * | 1S), 12, Tenino aes) Ttsfis co000nh00080000000 oo 56 00 or: x | Heh Je, WING J “coascooooosod0obed 56 oe 30 06 a0 a0 a # | 15. B-macrops Hugm..-...-...... +s 60 20 p06 bo + * | 1G 12, GaN JG, “coocoodoadoooennds O° 90 4.0 a6 .- * AES PO DINING) Gt (Vala cision oessvav rene seis tere O° ae x BO oo oa 60 * IG, 1D, WED SHR cooscpoo oes oooHagadabe OC a6 50 a 90 * U@), 12, w@aeryn Cite sco0q0n000000G0e cn oie 50 oO oc 6 AY, IP, Moen SUR, nocccopoecunsodne0 BO oe o¢ as oo . * Zl, 12, ByveroeqintMNs ODI Go dooooogodoKN c * * Subgen. b. Pogonopoma Ren. | 222, Ie, vyeaiacmninenl Sei aocacaoaccdode Ap fs ae He a a * He}, Ie, jxallleeiinn Jes sd00enoodooeno oe Fa ML, Je. eeauiloemdns (6 Pp 4 ooeccooedoneos * * Subgen. c. Fhinelepis Spix. is, Ie, jemi Si coogoanudededdnS 50 a or a8 09 + * AS 1 OF SHOW oogaocachoovdeHouen 610 ae ae oe a9 * Gen. II. Hemipstzicutays Higm, ! Il, Te FON Lo ooocooocouccG0EcODKe Bip Re ye an 50 + * ] bo is) iS) 194 MR. C. TATE REGAN ON THE 2. 3 5 6. Gen. III. Ancistrus Kner. Subgen. a. Ancrstrus. 1. A.annectens gn. 3.2... 622s eee BD. INS WGN SES codccaboo0ao 4 lac : * SepAcvetentaculatusis 710) terest et ah ae * ALVA cibbiceps Aeris acc ene «lege eit eters: * OyrAvsliturabus Avery rpord-net lier ie rete oe = *% % 6. A. multiradiatus Hancock ...........- * * 6a. A. multiradiatus, var. alternans Rgn. . a ie Yo Ee GINS EYOR, coddosoooncoo00e | * * SipAL medians Aavens a meewiererrerieie cients | # Dy ASIA cloconocodnoonc00G0 KC peer a HO ASraspidolepisnGci 7 amr -leehy ae | * | 1blG AG Oe MANS COP Goiogcacgaacsoc0c | ae x | 12. A. scaphirhynchus Kner ............ * 13) A. oligospilus |Gthre VOU iis eee eels e Cg IBS AG aR Si GoooguSconudononeDE | * Hay AY EO Jy copvagnnoodeducs0a0n oo || 16. A. schomburgkii Gthr. .............- | x | 17. A. megacephalus Gthr. .............. | * 18. A. platycephalus Blgr, ..........-+-- | 30 Subgen. 6. Parancistrus Blkr. | HOF PAs miveatusn@ astelne ue vtec trolls * 20. A. punctatissimus Stdr. .........--. * 21. A. aurantiacus Casteln. ...........55: Subgen. c. Lasiancistrus Ren. 22, A. heteracanthus Gthr............... 23. A. pictus Casteln. ...........006- 0:6 . 24, A. mystacinus Kner ...............- * 25, A. guacharote C. dg V. 1.1.0... ee eee * | Subgen. d. Pseudancistrus Blkr. MO AG SEOs JUG sooo addcgdbooonces * Blo Xe Weds C5 CPG sa do0e50c0e0cas0 us se 28. A; depressus Gthr. ....:..-02--00-- * ODS Me UNA, Jes oo ooo b OO nH aoe 46 * 30. A. angulicauda Stdr.. ....,.....405- : 81. A, parahybe Higm. ..............-- Gen. IV. Panaguy Eigmn. 1. P. cochliodon Kner .............0-- * PE les OGnicns (CMOS Fouonsongsecn0gBeus a ob Jee Erle) SUR oo ganeadsn ue an0oKd % 4, P. nigrolineatus Pérs. .............. ne * * Gen. V. Cumrosromus Tschudi. 1. C. loborhynchus Pschudi ............ 2p Ob MANGE I> con000000ad0G900 3. C. marcapatee Rgn. .......... 06005. AMC abrevishitgn. eee errr ite FISHES OF THE FAMILY LORICARIID. 195 1, HH NEP SCONDAP WN He © bo oun NP iN oo Gen. V. Cumrosromus (con.). . dermorhynchus Blgr. .....-.---+- pAIscheri Sed775 | owe leselelelelsiiel ele) ye eum IAitime Soococoosoueococwe . guairensis Stdr, ......---+-+ ees . Marginatus gn, ......---+-2-+: MbranickawStd7seeresetlestete tte Ty .anomalus Ran. 1.1.1... eee eens Snudirostrise ici everett rere Lames CHG asoo00c0 0000000000 AQAQAQAqgaaeanana Gen. VI. Xenocara Regn. latifrons Gthr. ......+-+--++-+--e- gymnorhynchus Kner .....-.+---- dolichoptera Kner .......-...-.- rename OF 621% noocogoccodc hoplogenys Githr. ...-.2.-+se++e+- @umnare, OL a? We sooccconcob000 00 Chanresie/1gi mune einai kaye occidentalis Rgn............+00-- brevipinnis fign. .............+.. WANTING, JM Ga5b coeccuocoDCO DN Jomo, 06 GP We oe conotanasnon0u , stigmatica Bugm. .... 2.2... 00000- bd PA Pd bd be bd Pt dd Gen. VII. Psrvpacanruicus Blkr. _ 1p eomenms CLEP WG shacocceccnosoce , EL GomOss COAT, socccocc2c 0000s: } ID, onabtl Clim odcoowooad aco bod00d 5 Ie, loysinubs OL GP WG coccecca0000s006 Gen, VIII. Acanruicus Spix. Aro layients S09 oa0000000000000000 Subfam. 2. HypoproPpomMaTIN”®. Gen. LX. Hypopropoma Gthr. thoracatum Githr, .........+.... Gimimamin Suir Gooovecnacooo0de rnaldnyeal Jip, Sooolepddo00e 0a0e 3 OG! VEL ‘coonedooonodcond0G . sbeindachneri Blgr............... o fallen) Olf22 acovadecen0s50600000 Ft tf Gen. X. Orocrxcrvs Cope. O}, Gb, cooonceoo0 cngooNnD ac BEVALLALUN PLUG 1 ammmretearey vial elotadei els) siereists EVES ULUUS| OOP CMM aTonctelafere] «lcieinievsiaictenss: . taczanowskil Sidr. . ....-.---.005- | . thomsoni gn. .......5..--e esses ee KKK: 196 MR. C. TATE REGAN ON THE 4. 5. 6. Ue Gen, X. Orocrnctvs (con.). A® Os tilts Cyn? soogoccecduuooadDOOT * One OMmaculicanda Std joes els 6 siekeisiene * Gy Opinions Jeni, Gaecsncodogvo0dG0. Bs Ho O, miamonnncky Jilin, Saegoooouenooe oc se SOMpertoratus Hig... 6 vse. esse ea 9. O. lophophanes Higm. .......2.0+-5- * Subfam. 3. LoricaRtinz. Gen. XI. Loricarta L. Subgen.a. Rhineloricaria Blkr. il, Ibs stibymeawoy Suh Gooadoobooaaboos PS Ibs cabaakerRNI Stio Gooooo nb 00b0S0.00 Be Gh Ib eiMe ih Sonodcooconododgunood + . 4, LL. phoxocephala Higm. ............ ea OwmlilanceolatanGc/imalrerielererite eerie GIS EN WIT GooobuuoodsodaDoap oo 7. L. uracantha Kner & Stdr. .........- SaunemardalenceSzary eer teers Oli Read exe Be enSelerpereryictoctenetroretoretereree oe -: Eo NOS 1p eer SES caogaano0cngcKdcane aa *% 11. L. catamarcensis Berg ............6- : on OMe konopickyl Str merrrerierirerie * 13. L. steindachneri Rgn. .............. a * ES Da Maina PAC eres uscauerovatoeske evel cnet eee arses * Ba Hoy Unstrigilataceensel ot iplslfeiereisaelciele O10 * 16. L. microlepidogaster ign. .......... : * Mifegeli latirostri see ora. etetle vers colerel lel: oe . Subgen. 6. Pseudoloricaria Blk. lS lusileviusculatGsncumVeuene meter * * NSS Ib, UKE Goad oddandnnedoads o oe Subgen. ¢. Loricariichthys Blkr. RO, lip imAguMeyey let, ccogoasonsonpeavo0d * sk Tle Ih, thos JU 6heee soonauaauesccs Ra * OP). I avon O66? Va esonnenda coosobed * * OB lhe ICOMOS EWA Soyo 5occnoonooKT 30 * OL Iby Eis ii, Voolsedaddesbaoc000s 66 Dey. Ih, Goesbh Sai “Gocopeasoodcuceooman oe OX Wy chonney Gh Ce EN gomacmoe uci acco % Subgen. d. Loricaria. ii, IbyiGyelstl Ji oo aacaoonsgaoono OD K Meh lb, sucess (OL Ch 1% ogoeodossedboc ¢ 5.0 x De iy NGS es | cooncdooensaboogde bs ao wc SOmMnicataphiractaesamerwerttitretelert creer sk % Bil, Ih Cane, ChGhes sooboogansasauoe oe * SPE Ibs Shanlllviney Jeo Sopp oocodoAeonGoes * 33. I. apeltogaster Blom seo... 22. oe ac FISHES OF THE FAMILY LORICARIIDZA, 197 2. 3. 4. 6. rite 8. Gen. XI. Lortcarta (con.). Subgen. d. Loricaria (con.). CHL Ib, mmeeROeloN GOP 6050000000000 00r : : .- % OE by TEI, Che Ws nosccacnsocseco0es an ** * Qo, Ih, WEIMEEEAEY SHER 50 0000000000000000 % Sey leemalcromystax: GuiTe ener elite - * Beh Ib, lena CHER, soccocscvescon00ce * SQ), Ih, letmezeS IRR, oo0co0ccoc000000000 : % 40° LL: platycephala Aner! 30.5. saeee ee: . * Gen. XII. Hemroponticutuys Blkr. ils Jal GEPREIIS LGGP scococccoc0aKKe a 2. Hi. acipenserinus Kner .........---.. * Gen. XIII. Oxyrorrcarta Blkr. lo), loniemationans She, o50000000b000¢ o 6 oe * 2. O. platystoma Gthr. ......... eer nerorers * Sb O, wHAIMONiMS JAG socoococoosane .: * 4, Op SUOMINE JM, 505000000000000 06 * Sb O; ROMS JG sooocccccscce0vcns o oe * GA Orlyratehignwey were er eee * Go On rosin SG cogaocccosbo coon .- & QO, ernie kGagP Socacnecavccnoonve 50 * OO panenmnensiSeh2¢ 71a eae * 1G; O, tnenenn JM “sococo,00envc0dn00 Gen. XIV. Fartowerra Hiem. Ile Ito glexchins JGR, Good ocgdooodooReSE . * 2p din CHEMO Jy gonsssddooganodoas * Sh Wo Cxqplyaones KGP Goosoacsauosos . . * Al, Ito lane SUR, scococonco00Kp codoer .: * D> 18> CONS IER once do coodenoaGabasoe * Gp J, eniavoniinn CHIP, sogooccccesco06s .. * to Bs aston GHAR coscso50000d00¢d00 . * Subfam. 4. NeopLecostominm. Gen. XV. Neorrrcostomus Eigm. ile Ilo SEEMS (CL GP WS conn aopesobenean BS * Eg Subfam. 5. Angrin», Gen. XVI. Arens C. & Y. Ul, Ae Wanneslen Ji, soctocnsonooscuenc Ke 2, AX. WMT IBD 6 onaonscacooces 20 * Bp Aa uM Jl coancavcconaseen Ba A A eyclopumisHimboldt .5 2.6.0.5)... 198 MR. C. TATE REGAN ON THE 1 2 3 4 5. 6 7 8 Gen. XVI. Areus (con.). 5. A. whymperi Blgr. .......2--2-.--- % GHA el genManniwlgiay ele alee pip eter eels gs fe 2 Vanilletal 100s Guo oOpN DUE oon oS lien oe Me or 8. A. orientalis Blgr... 2.22... s eee ee a Bs % 9. A. brachycephalus Githr. ............ be ae oS ellie NOE ANS OND IMs co oopcoeouocaoon nebo * TALS iN sane) comtoy eh) JakebObean Gg den ayo 000 Wan ° uy Ps Py AG Iooeaithe Seis popoasdaeno a6 0c° 0° ee ae a0 * 1S}, AS EEO CS GP Wooo sonccodacdsudond RS Be o6 * 14. A. taczanowskii Blgr. ......--...... tee ae as * Wis, AG jorconercrillkey ORR Is Gecooccacobou. Ps NIGRMAVEtiscidenswitg7. erie rsi-roeknelc telecine Wiese To An HOSES Joi ago odo od 6064 0ojc0'0 bb a Ai * Sh Ay owes Mus eon ounooboud dob oN me Me bo k AO APA SLINONSIE/UGI. Mayenne riage tenee-Rerenv ier *% | Gen. XVII. Astrosiervs Humboldt. He PAUoni aly cowl csnO nerdy lai arent oe Before passing to the systematic portion of this work, it may be as well to summarise the more important sexual differences and the changes which take place during growth. Sexual Differences. The differences between the sexes in certain genera are very remarkable. In the forms with the body protected by bony scutes it is often the case that the little spines which cover these and the plates of the head are stronger in the males than in the females. This feature is greatly exaggerated in Plecostomus spinosissimus and P. feste, in which the males have the head and body covered with quite strong spines. In other cases specialisation has resulted in the great development of spines or bristles in certain areas only. Thus in Ancistrus setosus and its allies, and in many species of Oxyloricaria, Farlowella, and Loricaria, the sides of the head in the male are margined with bristles, which are either absent or considerably shorter in the female. In Loricaria lanceolata and its allies similar bristles occur in addition on the supra- occipital, the nuchal scutes, and the spine of the pectoral fin. In many species of Loricaria which have the sides of the head bristly in the males, the head is also some- what broader and the snout more obtuse in specimens of that sex, a feature which is perhaps more strongly marked in L. latirostris than in any other species (see Pl XoVas). In most species of Xenocara the naked margin of the snout is much wider in males than in females, and whilst the latter have a more or less distinct marginal series of FISHES OF THE FAMILY LORICARIIDA. 199 tentacles, the former have in addition a Y-shaped group of tentacles, the limbs of the Y starting in front of the nostrils and running forward to meet in the middle line on the upper surface of the snout (see Pl. XIV.). In Loricaria acuta and related forms the males are distinguished by the great development of the lower lip, which covers the entire under surface of the head behind the mouth. Steindachner, in describing L. spixii, states that this is due to the nursing- habits of the males, which are said to carry the eggs, during development, between the lower lip and the head. However, Dr. Siebenrock has kindly informed me that this statement was based only on the structure of the lip, which seems adapted for such a purpose, and that none of the specimens in the Vienna Museum show eggs in this position. In the Argiine the males are readily distinguishable by their elongate anal papilla. Changes during Growth. As in most Teleostean Fishes, the young have the head and eye proportionately larger than those of the adult, and as a general rule the head is relatively less broad, the snout shorter, and the interorbital space rather narrower in young specimens. In the Plecostomine the lower surface of the head and abdomen is naked in all very young specimens, and may remain so throughout life in some species, whilst in others a more or less complete covering of small granular scales may be acquired. In those species of Loricaria which have the abdomen covered in the adult with small granular plates or scales, the latter are acquired during growth in a similar manner. In the Plecostomine the dorsal fin is both higher and longer in the young than in the adult, the difference being very considerable in some species—e. q., in Plecostomus emarginatus the length of the base of the dorsal is equal to its distance from the adipose fin in small specimens, whereas in large ones it is contained 14 times in that distance. In the Loricariine the dorsal fin appears to become relatively higher during growth. Throughout the family the pectoral fin becomes relatively longer during growth, so that in the same species it may barely reach the base of the ventral in a small example and yet extend nearly to the posterior extremity in a large one. ‘There is no difference in the length of the pectoral in individuals of different sexes if they are of the same size. The length of the barbel is dependent on the size of the specimen; thus in Arges brachycephalus small specimens (80-110 mm.) have the barbel extending beyond the gill-opening, whilst in larger ones (200-220 mm.) the barbel extends only 2 of the distance from its base to the gill-opening. Considerable changes of colour and marking may take place during the growth of certain species. In many Loricariine the young have 4—6 broad dark cross-bands on the back, which are absent in the adult. In the Plecostomine, in species such as Plecostomus guacari, P. verres, P. punctatus, &c., which have the head, body, and fins covered with dark spots, these are much fewer and larger on the head and body VOL, XVIJ.—PART 111. No. 2.—October, 1904. 25 200 MR. C. TATE REGAN ON THE in the young, whilst the fins have a few cross-bars, which later on break up into spots, and these spots, like those on the head and body, seem to continually increase in number with increased size of the fish. Measurements. In descriptions of species the total length means the distance from the tip of the snout to the base of the caudal fin; but where the total length (in millimetres) of the largest specimen described is given, the length of the middle rays of the caudal fin is included. In the armoured forms the length of the head is measured to the posterior margin of the temporal plate, but in the naked-bodied Argiine to the edge of the bony operculum. In some cases it has been found useful to indicate the size of the mouth by giving the proportion of the length of the mandibular ramus—. ¢. of that box-like tooth-bearing portion of the mandible which borders the mouth—to the interorbital width. In counting the scutes, that one which is posterior to the clavicle is reckoned as the first, whilst the movable plates covering the bases of the caudal rays are not included. The length of the base of the dorsal fin is measured from the base of the first developed ray to the base of the last, the membrane behind the last ray being excluded. The sign f prefixed to the name of a species indicates that the author has not examined actual specimens, but has based his description on the published accounts. Family LORICARIID. Ostariophysi with parietals and supraoccipital united to form a single parieto-occipital bone, without suboperculum, symplectic, or metapterygoid; palatine articulating posteriorly with the prefrontal and anteriorly with the small rod-like maxillary, which bears a barbel; preemaxillaries not protractile, forming the entire upper border of the mouth. Caudal vertebra) compressed, with expanded and compressed neural and heemal spines ; preecaudal vertebree without parapophyses; coalesced anterior vertebre with outgrowths on each side which are suturally united to the cranium, forming, with the exoccipitals, two capsules enclosing the two halves of the reduced air-bladder ; the last (fifth) bearing a pair of strong ribs, articulating with the first free vertebre by a hinge which permits only of vertical movement ; ribs of the free precaudal vertebre, if present, sessile. Nostrils close together, a flap between them; mouth inferior, with expanded and reflected lips; jaws with teeth; palate toothless; lower and fourth upper pharyngeals with or without teeth; four gills, a slit behind the fourth; pseudo- branchie present or absent; gill-membranes broadly joined to the isthmus; four branchiostegals. Dorsal fin usually above the precaudal portion of the vertebral column; adipose fin often present; anal fin short; ventrals below or in front of the dorsal; anterior rays of dorsal, anal, pectoral and ventral, and outer rays of caudal simple, often spinous. FISHES OF THE FAMILY LORICARIIDE. 201 From anatomical considerations it seems best to recognise five subfamilies, the relations of which to each other are shown by the following scheme :— Neoplecostomine. Argiinee. Loricariine. | | Plecostomine. Hypoptopomatinee. | Analysis of the Subfamilies. I, Heemal spines all simple ; lower and fourth upper pharyngeals not toothed ; anterior rudimentary ray of ventral normally shaped, but internal and directed forwards ; upper part of the body protected by bony scutes. Lower transverse portion of clavicles and coracoids not exposed ; margin of snout naked or covered with small granular plates. . . 1. Precosrominm. Lower transverse portion of clavicles and coracoids exposed ; margin of snout composed of more or less firm plates. . . . . . . . 2. Hyporropomarinm. iI. Hzemal spines of the vertebra above the anal fin bifid; lower and fourth upper pharyngeals toothed. A. No anterior rudimentary ray of ventral; caudal peduncle elongate and strongly depressed ; upper part of the body protected by bony Scutestr gal cone it at Cueshe Lene wht ana eee men ead TOM ORTCA REINA B. Anterior rudimentary ray of ventral internal, plate-like ; caudal peduncle cylindrical or compressed. Upper part of the body protected by bony scutes . . . . . . . 4, NEoPLECOsTOMIN#. Toh Sree a) ai Mines ee eS Quake 2G aa gy Alcan Vee imita. lla Jamon iio Key to the Subfamitlies. I. Upper part of body protected by bony plates. A. Caudal peduncle compressed, cylindrical, or moderately depressed. a. Lower transverse portion of coracoids and clavicles not exposed. Abdomen naked or more or less completely covered with small granular scales, never as in the Neoplecostomme . . . . . Pxuxcosrominx (p. 202). Abdomen with a large, regular, nearly square patch of small granular plates, naked at the sides and anteriorly and posteriorly. . . Neoptecostomin” (p. 306). b. Lower transverse portion of coracoids and clavicles exposed, appearing as 2 pairs of plates extending between the bases of the pectorals. . . . Hyrorroromatin® (p. 262). B, Caudal peduncle strongly depressed, broad seloneate: lover transverse portions of clavicles and coracoids not exposed Loricartin& (p. 269). lila Bodyvanakcd mami mer me- nef?) 40 1-0 ee nme RGLINZE) (De SOM) aE 2 202 MR. C. TATE REGAN ON THE Subfamily I. PLECOSTOMIN &. Vertebre below the dorsal fin with bifid neural spines; none of the hzmal spines bifid; pharyngeals not toothed ; teeth in the jaws in a single series ; pseudobranchize present ; upper part of the body protected by bony scutes; caudal peduncle compressed, cylindrical, or slightly depressed, with 5 longitudinal series of scutes on each side for its whole length; lower transverse part of clavicle and coracoid not exposed ; margin of snout naked or covered with small granular plates ; anterior rudimentary ray of ventral internal and directed forwards. Key to the Genera. T. First scute of lower lateral series posterior to the temporal plate, with which the second is not in contact. A. Premaxillaries nearly equal in length to the dentaries and with about the same number of teeth. 1. Operculum and interoperculum little, and not independently, movable. Snout granular to its margm . . . . . . + + + - =. « « LI. Plecostomus Gronow. Snout witha naked margin . . . .. .. .. ... « « & HemipstlichthysMigmn. 2. Interoperculum more or less freely movable, usually spinate or bristly. a. Snout granular to its margin. Meeth mumerous bifid eee Pier Meio dile Metacritic ea mel ecistnusn Kener. Meeth few, spoon-shaped sii sift) er ee) fe) We elie) teed eo eaepeonegve Brom. 6. Snout with a naked margin. Mouth wide, the length of mandibular ramus nearly equal to the inter- orbital width; naked margin of snout not very wide, without WEN Gg oo 6 oo 6 86 6 6 6 of 5 6 oO 8 oO he Cigars Wscauah, Mouth moderate or narrow, the length of mandibular ramus considerably less than the interorbital width; naked margin of snout wide or with more or less distinct tentacles . . . . . . . . . . 6, Xenocara Regn. B. Preemaxillaries much shorter than the dentaries and with (ues) G Gg 6.6 6 6 0 a 0 o 5 0 0 0 05 6 Oo PRaminominGNS lle: IL. First two scutes of the lower lateral series below the temporal plate and in contact with it above . . . . . . . + « 8. Acanthicus Spix. 1. PLEcostomus. Plecostomus Gronow, Mus. i. p. 24 (1758), and Zooph. p. 127 (1763) ; Bleek. Ned. Tijd. Dierk. i. 1863, p. 77; Giinth. Cat. Fish. v. p. 230 (1864) ; Eigenm. & Higenm. Occ. Pap. Cal. Ac. i. 1890, p. 396. Hypostomus Lacép. Hist. Nat. Poiss. v. p. 144 (1803) ; Cuv. & Val. Hist. Nat. Poiss. xv. p. 489 (1840), part.; Kner, Denkschr. Ak. Wien, vil. 1854, p. 256. Rhinelepis Spix, Gen. et Spec. Pisc. Bras. p. 4 (1829) ; Cuv. & Val. Hist. Nat. Poiss. xv. p. 479 (1840), part.; Giimth. Cat. Fish. v. p. 252 (1864) ; Higenm. & Eigenm. Occ. Pap. Cal. Ac. i. 1890, p. 414. Body oblong or elongate; first scute of lower lateral series posterior to clavicle and FISHES OF THE FAMILY LORICARIIDA. 203 temporal plate, separating the second entirely from the latter; lower surface of head and abdomen naked, or with small granular scales ; snout covered with granular plates to its margin. Operculum and interoperculum with or without marginal bristles, the lower edge of the former contiguous and firmly connected by membrane to the upper edge of the latter, so that they are incapable of independent movement. Pre- maxillaries nearly equal in length to the dentaries, and opposed to them; teeth numerous, slender, curved, bifid, forming a nearly straight series in each jaw. Dorsal with I 7 rays, originating in advance of the ventrals, separated from the supraoccipital by 3 (rarely 2 or 4) scutes. Anal with I 3-5 rays, pectoral with I 6, ventral with I 5. Caudal emarginate or truncate. Adipose fin, if present, represented by a movable spine and a membrane attaching it posteriorly to the caudal peduncle. Ribs present, slender ; vertebre 5+9+16 (in P. verres). Rivers of S. America. It has seemed most convenient to retain this genus, although the difference from Ancistrus is so slight that some authors have proposed to unite them, a proposal which is greatly strengthened by the discovery of the species described below as Ancistrus annectens, which can only be specifically distinguished from Plecostomus guacari by 3 or 4 characters, one of these, however, being that on which the generic distinction is based—~. ¢., the structure of the operculum and interoperculum. The presence or absence of an adipose fin can scarcely be regarded as of generic importance in the light of the great variation which exists in Chetostomus anomalus, and in view of the fact that in the very natural subgenus which is here established under the name Pogonopoma tor P. wertheimeri Stdr., P. pellegrini Rgn., and P. genibarbis C. & V., an adipose fin is present in the first-named species, absent in the others. hinelepis Spix is therefore placed as a subgenus under Plecostomus. It is curious to note that in the fairly numerous examples of guacari, P. commersonii, and P. punctatus on the one hand, and of P. verres and P. emarginatus on the other, which have been examined by the author, the supraoccipital is constantly entirely bordered posteriorly by a single scute in the former, by a median scute and by one or more on each side in the latter, this feature seeming of considerable importance in showing the natural relations of the species. However, ina series of twelve examples of P. wuchereri, eight have the supraoccipital bordered by one scute, four by three, the character here proving not to be of specific value. Key to the Species. I. Adipose fin present, well-developed ; operculum and interoperculum not margined with. bristles. (PLEcostomus.) A. Length of mandibular ramus 22-4 times in the interorbital width. 1. Supraoccipital entirely bordered posteriorly by a single scute ; first dorsal ray about as long as the head; some of the scutes carinate. a, 25-26 scutes in a longitudinal series . . . .. =.» +. + +. + XJ. P. guacari Lacép. 204 MR, C. TATE REGAN ON THE 8. 28-80 scutes in a longitudinal series ; caudal peduncle 33-5 times as long as deep. Scutes mostly carinate ; length of base of dorsal greater than its distance from adipose fin . . . . 6) oo oO > Je Conan Cats WV Only the anterior scutes carinate; length of base of dorsal equal to (young) or less than (adult) its distance from PNG OOSS, HON GN ey SONG ig io (Oia Gg. G18 | Ooo lo Onl dio PromonanTS Oace WN y. 32-84 scutes in a longitudinal series. Caudal peduncle 7 times as long asdeep. . . . . . . . 4 P. spinosissimus Stdr. Caudal peduncle 11 timesaslongasdeep . ... . . . 5. P. feste Bigr. 2. Supraoccipital bordered posteriorly by a median scute and by one or more on each side; first dorsal ray about as long as the head. a. Caudal peduncle normally formed, the scutes of the fourth series not strongly angulated. 26-28 scutes in a longitudinal series, the upper and anterior Caninates ri cranes Sire ju Prenpaeecrey ish Ady sol rede OMwEs RUCNTCSHO ROGAN E 30 scutes in a eneiucinal series, all GHOMENS 5 5 6 6 0 o do (Po Ouro Swale, 27-28 scutes in a longitudinal series, none of them carinate . 8. P. vaillanti Stdr. 8. Caudal peduncle broad and flat below, the scutes of the fourth series strongly angulated. 28-30 scutes in a longitudimal series . . . . . . . . . QY. P. emarginatus C.& V. 5. Supraoccipital entirely bordered posteriorly by a single scute; first dorsal ray as long as the distance from the tip of snout to its base; scutes not carmate, 27-28 in a loncitudimallisertesy- mu cameniien iil icone men ete Omaealatuss@aste lime &. Length of mandibular ramus 14-24 times in the interorbital width. 1. Some of the scutes carinate, the keels sometimes very weak. a. Depth of body 64-8 times in the total length. Depth of head 2} times in its length; 30 scutes in a longitudinal series eA etepcelyulleney eniCidess Gee Moe vane . ll. P. cordove Gthr. Depth of head 12 times in its ienethe 28 scutes in a ean Series) Ho 1 Wee cree tee gece tao Ai Mev lis ten ewe tre atv Aaicamlla eae Gone ceca os 8. Depth of body 4-5 times in the total length ; 25-— scutes in a longitudinal series; diameter of eye 53-7 times in the length of head. First dorsal He not reaching adipose fin when laid back; barbel as Bil long as 3 the diameter of eye, orless. . . . . . 3 . 13. P. latirostris Ren. First dorsal ray reaching adipose fin when laid back ; harbel as iene AS GUENNGIER Or OE G6 ig 6 a 6.6 0.0/0 of oo 6 o 9 dik Jumma: iilesy, y. Depth of body 4-5 times in the total length ; 27-28 scutes in a longitudinal series ; diameter of eye 10 times in the length of head (in a specimen of AOR Ui) 6 9 Bb g oa 0 © oo 6 a ob a 0 se mode CL &s Wo 2. None of the scutes carinate. a. Supraorbital edges strongly raised; eye large. . . . 15. P. macrops Kigmn. FISHES OF THE FAMILY LORICARIIDA. 205 8. Supraorbital edges not raised; length of mandibular ramus 14-2} in the interorbital width. 26 scutes in a longitudinal series; diameter of eye 63 times in the irae of head (in a specimen of 260m) ee ¢ - . . . 16. P. garmani Ren. 26 scutes in a onenont cel series ; emieter of eye 7 times in the length of head (in a specimen of 130 mm.) ; caudal peduncle 23 times as long as deep. . . . 18. P. une Stdr. 26 scutes in a longitudinal series; diameter of eye 8 times in the length of head (in a specimen of 250 mm.) ; caudal peduncle 3 times as long as deep ; abdomen, in the adult, in great part covered with granular SHES 5 oo oo 0 6 8 op 6 5b 0 0 6 6 IQ I, anmeberard Cine, 27 scutes in a longitudinal series; abdomen, in the adult, almost entirely naked . . . . . 20. P. luetkent Stdr. y. Supraorbital edges slightly raised; rete of aenaibene ramus 14 times in the Feel width . . . . 21. P. auroguttatus Kner. II. Adipose fin small or absent ; operculum and interoperculum with well-developed marginal bristles. (Poconopoma, subgen. n.) AX, INGHOONS Win [EIN 5g 8 kg 8 BB og yy RRS IP anemiieninars Shale. B. No adipose fin. Length of snout 23 times in that of the head, which is 32 times ini the) totalilength)) =) 52): . 23, P. pellegrini Regn. Length of snout twice in that of the head, sift is 3-31 TMS Ti WAS WoO Wengen 5 5 6 6 19 6 0 6 6 5 et > ganioaniis Co & WW. III. Adipose fin absent; operculum and interoperculum not margined with bristles. (Rurneupis Spix.) Tnterorbital width 21-22 times in the length of head; opercular plate rather large ; abdomen almost entirely naked. . . . . . . 25. P. parahybe Stdr. Interorbital width 12 times in the length of head; opercular plate very large ; abdomen entirely covered with granular scales. . . 26. P. asper Spix. 1. PLECOSTOMUS GUACARI. Loricaria plecostomus Linn. S. N. i. p. 508 (1766) ; Bloch, Ausl. Fische, viii. pl. 374 (1794) ; Bl. Schn. p. 124 (1801). Hypostomus guacari Lacép. Poiss. v. p. 145 (1803). Loricaria flava Shaw, Zool. v. p. 38, pl. ci. (1805). Hypostomus plecostomus Schomburgk, Fish. Guiana, 1. p. 139 (1841) ; Kner, Denkschr. Ak. Wien, vil. 1853, p. 263. Plecostomus bicirrhosus Gronow, Cat. p. 158 (1854) ; Giimth. Cat. Fish. v. p. 231 (1864) (part.) ; Kner & Steind. Abhandl. bayr. Ak. x. 1865, p. 60; Hensel, Arch. Nat. 1870, p. 75; Steind. Denkschr. Ak. Wien, xliii. 1881, p. 109. Hypostomus robinii Gill, Aun. Lyc. N. York, vi. 1858, p. 46. Plecostomus brasiliensis Bleek. Nat. Verh. Holl. Maats. xx. 1864, Silures de Suriname, p. 7. 206 MR. C. TATE REGAN ON THE Plecostomus flavus Vaill. Bull. Soc. Philom. (7) iv. 1880, p. 155. Plecostomus plecostomus Bigenm. & Eigenm. Proc. Cal. Ac. (2) i. 1888, p. 168; Occ. Pap. Cal. Ac. 1. p. 406 (1890). Plecostomus seminudus Figenm. & Higenm. t. c. p. 169. Plecostomus boulengeri Eigenm. & Kennedy, Proc. Ac. Philad. 1903, p. 502. Depth of body 43-5 times in the total length, length of head 3-3} times. Depth of head 1% times in its length, breadth of head 1% times, length of snout 12 times, diameter of eye 54-94 times, interorbital width about 23 times. Length of mandibular ramus about 3 times in the interorbital width. Snout ovate, more or less narrowed anteriorly ; supraorbital margins raised; supraoccipital with moderately elevated median ridge; temporal plates carinate. Scutes spinulose, the anterior and upper scutes carinate, 25-26 in a longitudinal series, 7-S between dorsal and adipose fin, 13-14 between anal and caudal. Supraoccipital entirely bordered posteriorly by a single scute. Lower surface of head and abdomen, in the adult, almost completely covered with small granular scales. D. I 7, the first ray about as long as the head, the last 2-3 as long; length of base of dorsal equal to its distance from posterior 4 of spine of adipose fin. A. I 4. Pectoral spine extending well beyond base of ventrals. 9 Caudal emarginate, the inner rays 3—? the length of head, the outer produced. Caudal peduncle 5-3} times as long as deep. In the adult, numerous dark spots on head, body, and fins, those on the head smallest, those on the dorsal in 2 rows on each interradial membrane. In the young, spots on the head and body larger and fewer, fins with 4—6 transverse bars. Total length 350 mm. Twelve specimens: R. Amazon; Guiana; Venezuela; Trinidad. 2, PLECOSTOMUS COMMERSONII. Hypostomus commersonii Cuv. & Val. Hist. Nat. Poiss. xv. p. 495 (1840); Val. Voy. d’Orbigny, Poiss. pl. vii. fig. 2 (1847). Plecostomus commersonii Giinth. Cat. Fish. v. p. 232 (1864). Plecostomus spiniger Hensel, Arch. Nat. 1. 1870, p. 73. Plecostomus limosus Eigenm. & Eigenm. Proc. Cal. Ac. (2) i. 1888, p. 167, and Occ. Pap. Cal. Ac. i. 1890, p. 405. ? Plecostomus aspilogaster Cope, Proc. Am. Phil. Soc. xxxiii. 1894, p. 100, fig. 14. Depth of body 54-52 times in the total length, length of head 33-33 times. Depth of head 12 times in its length, breadth of head about 1§ times, diameter of eye 8-13 times, length of snout 1? times, interorbital width 23-23 times. Length of mandibular ramus 23-21 times in the interorbital width. Snout rather broad, rounded anteriorly ; supraorbital edges slightly or moderately raised; supraoccipital with median ridge ; temporal plates more or less distinctly carimate. Scutes spinulose, mostly carinate, 28-30 in a longitudinal series, 8-9 between dorsal and adipose fin, 14-15 between anal and caudal. Supraoccipital entirely bordered posteriorly by asingle scute. Lower FISHES OF THE FAMILY LORICARIIDA. 207 surface of head and abdomen, in the adult, almost completely covered with small granular scales. D. 17, the first ray about as long as the head, the last }-3 as long; length of base of dorsal equal to its distance from posterior part of spine of adipose fin. A.14. Pectoral spine extending beyond base of ventral. Caudal moderately emarginate, the outer rays scarcely produced. Caudal peduncle 33-4 times as long as deep. Numerous dark spots on head, body, and fins, those on the head the smallest, those on the dorsal in 2—4 series on each interradial membrane, those on the caudal usually indistinct. Total length 460 mm. Twelve specimens (including the type of the species and a co-type of P. limosus) : Rio La Plata and its tributaries; Rio Grande do Sul. 3. PLECOSTOMUS PUNCTATUS. Hypostomus punctatus Cuv. & Val. Hist. Nat. Poiss. xv. p. 493 (1840). Hypostomus subcarinatus Casteln. Anim. Am. Sud, p. 42, pl. xxi. fig. 1 (1855). Plecostomus punctatus Giinth. Cat. Fish. v. p. 233 (1864). Plecostomus affinis Steind. Sitzb. Ak. Wien, Ixxiv. 1876, p. 685. Plecostomus commersonii Kigenm. & Kigenm. Proc. Cal. Ac. (2) i. 1888, p.168, and Occ. Pap. Cal, Ac. 1. 1890, p. 403. Depth of body 5-6 times in the total length, length of head about 34 times. Depth of head 13-1? times in its length, breadth of head 14-1} times, diameter of eye 6-85 times, interorbital width 23-24 times, length of snout 13 times. Length of mandibular ramus 24-3 times in the interorbital width. Snout ovate, more or less narrowed anteriorly ; supraorbital edges not raised; supraoccipital with low median ridge; temporal plates not carinate. Scutes spinulose, the anterior carinate, 28-30 in a longitudinal series, 9 between dorsal and adipose fin, 14-16 between anal and caudal. Supraoccipital entirely bordered posteriorly by a single scute. Lower surface of head and abdomen, in the adult, almost completely covered with small granular scales. D. 1 7, the first ray as long as or a little longer than the head, the last 3-2 as long; length of base of dorsal nearly equal to its distance from the adipose fin. A. I 4. Pectoral spe extending to anterior + of ventral. Caudal emarginate. Caudal peduncle 33-4? times as long as deep. Numerous small dark spots on head, body, and fins, those on the dorsal (in the adult) in 2 series on each interradial membrane. Total length 500 mm. Twelve specimens (including the types of the species and of H. sudcarinatus): Southern and Eastern Brazil. Under the names P. commersonii affinis and P. commersonii scabriceps, Kigenmann distinguishes two varieties of this species, which are said to have larger spots than the typical P. punctatus, the second also having a large naked area on the head behind the VOL. XVII.—PAR? 11. No. 3.—October, 1904. 25 208 MR. C. TATE REGAN ON THE mouth. ‘These varieties scarcely seem worth recognition; in all young specimens the spots are large and the lower surface of head and abdomen partially naked, and the persistence of these features in the adult must be regarded as cases of individual variation only. 4, PLECOSTOMUS SPINOSISSIMUS. Plecostomus spinosissimus Steind. Denkschr. Ak. Wien, xl. 1880, p. 98, pl. v. fig. 1. Depth of body 63-64 times in the total length, length of head 33 times. Depth of head 14-2 times in its length, breadth of head 1§ times, length of snout 13 times, diameter of eye 7-8 times, interorbital width 2$ times. Length of mandibular ramus 3 times in the interorbital width. Snout narrowed anteriorly ; supraorbital margins raised; supraoccipital with median ridge; temporal plates not distinctly carinate. Scutes spinulose (in males strongly spinate), the anterior and upper scutes carinate, 32 in a longitudinal series, 10 between dorsal and adipose fin, 17 between anal and caudal. Supraoccipital entirely margined posteriorly bya single scute. Lower surface of head and abdomen, in the adult, almost completely covered with small granular scales. D.I1 7, the first ray nearly as long as the head, the last 3-2 as long; length of base of dorsal 14 times in its distance from the adipose fin. A.I4. Pectoral spine extending to base of ventral (in males armed with long curved bristles in its outer half). Caudal slightly emarginate. Caudal peduncle 7 times as long as deep. Numerous dark spots on head, body, and fins. Total length 250 mm. Two specimens: Guayaquil, W. Ecuador. 5, PLecostomus Festa#. (Plate LX. fig. 1.) Plecostomus feste Bouleng. Boll. Mus. Torin. xiii. 1898, no. 3829, p. 11. Depth of body 8 times in the total length, length of head 43 times. Depth of head twice in its length, breadth of head 14 times, length of snout 1§ times, diameter of eye 84 times, interorbital width 54 times. Length of mandible 2? times in the interorbital width. Snout narrowed anteriorly ; supraorbital margins raised; supraoccipital with median ridge; temporal plates not distinctly carinate. Scutes strongly spinate (in males only), those of the upper series carinate, 53 or 34 in a longitudinal series, 11 between dorsal and adipose fin, 19 between anal and caudal. Supraoccipital entirely margined posteriorly by a single scute. Lower surface of head and abdomen, in the adult, almost completely covered with small granular scales. D.1 7, the first ray nearly as long as the head, the last 2 as long; length of base of dorsal 12 times in its distance from the adipose fin, A.1I4. Pectoral spine extending to base of ventral, armed in its outer half with very long curved bristles (tin males only). Caudal slightly FISHES OF THE FAMILY LORICARIID®, 209 emarginate. Caudal peduncle 11 times as long as deep. Numerous dark spots on head, body, and fins. Total length 310 mm. One specimen: W, Ecuador. 6. PLECOSTOMUS VERRES. Hypostomus plecostomus (non Linn.) Cuv. & Val. Hist. Nat. Poiss. xv. p. 489 (1840). Hypostomus verres Cuv. & Val. t. c. p. 494. ? Hypostomus pantherinus Kner, Denkschr. Ak. Wien, vii. 1854, p. 267. Plecostomus bicirrhosus (part.) Giinth. Cat. Fish. v. p. 231 (1864). Depth of body 43-5} times in the total length, length of head 24-54 times. Depth of head 13-1? times in its length, breadth of head 11-1} times, length of snout 2-2 times, diameter of eye 6-94 times, interorbital width 2!-24 times. Length of mandibular ramus 53-32 times in the interorbital width. Snout narrowed anteriorly ; supraorbital edges raised; supraoccipital with strongly elevated median ridge; tem- poral plates carinate. Scutes spinulose, carinate except on caudal peduncle, 26-28 in a longitudinal series, 7-9 between dorsal and adipose fin, 13-14 between anal and caudal. Supraoccipital bordered posteriorly by a median scute and by 2 or 3 of a group of small scutes on each side. Lower surface of head and abdomen, in the adult, almost completely covered with small granular scales. D.I7, the first ray as long as or a little longer than the head, the last 2-2 as long ; length of base of dorsal equal to or greater than its distance from the adipose fin. A.I4. Pectoral spine extending a little beyond base or nearly to middle of ventral. Caudal deeply emarginate, the outer rays greatly produced. Caudal peduncle 53 (young) to 4+ times as long as deep. Numerous dark spots on head, body, and fins, smallest on the head, those on the dorsal in 2 series between each pair of rays. In the young, spots much fewer and larger. Total length 320 mm. Ten specimens (including the types of the species): Venezuela; Guiana; Marajo Island. 7. TPLECOSTOMUS CARINATUS. Plecostomus carinatus Steind. Denkschr. Ak. Wien, xlii. 1881, p. 108, pl. iv. fig. 2. Very closely allied to the preceding species, from which it is distinguished by the larger eye and the more numerous scutes, all of which are strongly carinate. Diameter of eye 6 times in the length of head. 30 scutes in a longitudinal series, 8-9 between dorsal and adipose fin, 16 between anal and caudal. The typical examples are 300-350 mm. in total length. R. Amazon. bo ty LS 210 MR. C. TATE REGAN ON THE 8. PLECOSTOMUS VAILLANTI. Plecostomus vaillanti Steind. Sitzb. Ak. Wien, Ixxvi. 1877, p. 225; Eigenm. & Higenm. Proce. Cal. Ac. (2) 1. 1888, p. 169, and Occ. Pap. Cal. Ac. i. p. 407 (1890). Depth of body 53-6 times in the total length, length of head about 3 times. Depth of head 14—2 times in its length, breadth of head 1% times, length of snout 1} times, diameter of eye about 64 times, interorbital width 2? times. Length of mandibular ramus about 24 times in the interorbital width. Snout rather broad, rounded ; supra- orbital margins strongly raised ; supraoccipital with moderately elevated median ridge ; temporal plates not carinate; barbel as long as diameter of eye. Scutes spinulose, not carinate, 27-28 in a longitudinal series, 7-8 between dorsal and adipose fin, 14 between anal and caudal. Supraoccipital bordered posteriorly by a median scute and by one or two on each side. Lower surface of head and abdomen almost com- pletely covered with small granular scales. D.I 7, the first ray as long as the head, the last $ as long; length of base of dorsal greater than its distance from the adipose fin. A.1I4. Pectoral spine extending to anterior $ of ventral. Caudal obliquely emarginate. Caudal peduncle 33 times as long as deep. Large dark spots on head, body, and fins, those on the fins sometimes forming cross-bars. Total length 200 mm. Two specimens: Paraguay: Rio Puty and Rio Preto, E. Brazil. 9, PLECOSTOMUS EMARGINATUS. Hypostomus emarginatus Cuv. & Val. Hist. Nat. Poiss. xv. p. 500 (1840); Kner, Denkschr. Ak. Wien, vil. 1854, p. 260. Hypostomus squalinum Schomburgk, Fish. Brit. Guiana, p. 142, pl. i. (1841). Hypostomus horridus Kner, t.c. p. 209. Plecostomus emarginatus Giinth. Cat. Fish. v. p. 233 (1864) ; Eigenm. & Eigenm. Proc. Cal. Ac, (2) i. 1888, p. 167, and Occ. Pap. Cal. Ac. i. p. 400 (1890). Plecostomus horridus Giinth. t.c. p. 232; Peters, Monatsb. Ak. Berl. 1877, p. 471. Plecostomus scopularius Cope, Proc. Ac. Philad. 1871, p. 55, pl. xvi. figs. 1 & 2 (nec Cope, t. c. p- 286). Plecostomus biseriatus Cope, t. c. p. 285. ? Plecostomus virescens Cope, op. cit. 1874, p. 137, and Proc. Am. Phil. Soc. xvii. 1878, p. 681, and ib. xxxiil. 1894, p. JO1. Plecostomus villarsi Liitk. Overs. Dan. Selsk. 1874, p. 211; Steind. Denkschr. Ak. Wien, xxxix. 1878, p. 42, pl. vil. Plecostomus tenuicauda Steind. t.c. p. 40, pl. vi., and ib, xli. 1880, p. 63. Plecostomus anne, Steind. op. cit. xliii. 1881, p. 112, pl. i. fig. 2. Depth of body 6-7 times in the total length, length of head 33-4 times. Depth of head 12-2 times in its length, breadth of head 13-1; times, length of snout 1}- 15 times, diameter of eye 5-11 times, interorbital width 23-23 times. Length of rnandibular ramus 3-4 times in the interorbital width. Snout more or less narrowed FISHES OF THE FAMILY LORICARIIDA, PALL anteriorly ; supraorbital edges moderately or slightly raised ; supraoccipital with median ridge; temporal plates usually not carinate. Scutes spinulose, usually only those posterior to the dorsal carinate, sometimes the anterior scutes also weakly carinate, sometimes none of the scutes with distinct keels, 28-30 in a longitudinal series, 8-9 between dorsal and adipose fin, 14-15 between anal and caudal. Supraoccipital margined posteriorly by a median scute and 1 or 2 on each side. Lower surface of head and abdomen, in the adult, almost completely covered with small granular scales. D. 1 7, the first ray as long as the head, the last $ as long; length of base of dorsal 13 (young) to 14 (adult) times in its distance from the adipose fin. A.I 4. Pectoral spine extending to base of ventral or beyond. Caudal strongly emarginate. Caudal peduncle 4-5 times as long as deep, broad and flat below, the scutes of the fourth series being strongly angulated, forming a ridge which margins the flat lower surface. Dark spots on head, body, and fins, those on the dorsal usually arranged in two series between each pair of rays. Total length 470 mm. Nineteen specimens (including the type of the species): Brazil to Colombia. 10. PLEcosToMUS ALATUS. Hypostomus alatus Casteln. Anim. Am. Sud, p. 41, pl. xx. fig. 1 (1855). Plecostomus alatus Giiuth. Cat. Fish. v. p. 234 (1864) ; Liitk. Vid. Selsk. xii. 1875, p. 144. Plecostomus francisci Liitk. Overs. Dan. Selsk. 1873, no. 3, p. 80; Vid. Selsk. xu. 1875, p- 143. Depth of body about 5 times in the total length, length of head 3} times. Depth of head 12 times in its length, breadth of head 1} times, length of snout 13-1} times, diameter of eye 64 times, interorbital width 22 times. Length of mandibular ramus 24 times in the interorbital width. Snout broad, rounded ; supraorbital edges not raised ; supraoccipital with low median ridge; temporal plates not carinate. Scutes spinulose, not carinate, 27-28 in a longitudinal series, 9 between dorsal and adipose fin, 15 between anal and caudal. Supraoccipital entirely bordered posteriorly by a single scute. Lower surface of head and abdomen, in the adult, almost completely covered with small granular scales. D.17; the first ray as long as the distance from its base to the tip of snout; length of base of dorsal equal to its distance from the middle of the spine of the adipose fin. A.1I4. Pectoral spine extending to anterior third of ventral. Caudal emarginate, the inner rays 3 the length of head, the outer much produced. Caudal peduncle 3} times as long as deep. Brownish, with yellow spots on head, body, and fins. Total length 300 mm. ‘Two specimens (including the type of the species): Araguay, San Francisco, and das Velhas Rivers. 212 MR. C. TATE REGAN ON THE 11. Priecostomus corpovz&. (Plate IX. fig. 3.) Plecostomus cordove Giinth. Ann. Mag. Nat. Hist. (5) vi. 1880, p. 11. Depth of body nearly 8 times in the total length, length of head 33 times. Depth of head 21 times in its length, breadth of head 1% times, length of snout 13 times, diameter of eye 10 times, interorbital width 5 times. Length of mandibular ramus 14 times in the interorbital width. Snout broad, rounded; supraorbital edges slightly raised; supraoccipital with low median ridge; temporal plates carinate. Scutes spinulose, those of the upper series and the anterior of the second series carinate, 30 in a longitudinal series, 10 between dorsal and adipose fin, 17 between anal and caudal. Supraoccipital bordered posteriorly by a median scute and by one on each side. Lower surface of head and abdomen almost completely covered with small granular scales. D.I7, the first ray nearly as long as the head, the last half as long; length of base of dorsal less than its distance from the adipose fin. 795 6 2 0 6 9 5 6 8 5 6 8 WL ZL Gpnetpo Ine. Interoperculum with about 15 spines; base of dorsal equal to its distauce from adipose fin; 26-27 scutes in a longitudinal SHED 5 5b 616 0 6 8 6B 610 0 6 o po 5 6 NE Al Gaaphaionas Wore, C. None of the scutes carimate ; dorsal with I 7-8 rays. 1. Supraoccipital with low median ridge. Length of mandibular ramus 2} times in the interorbital width; Jarge round dark spots on ert body, andfins . . . . . 18. A. oliyospilus Gthr. Length of mandibular ramus 24 times in the imterorbital width ; body and fins with a few broad dark bands . . . . . . . 14. A. vittatus Stdr. 2. Supraoccipital without median ridge. a. Length of mandibular ramus less than interorbital width. Interorbital width 1} times in the length of head . . . . . . 15. A. bachi Blgr. VOL. XVII.— PAR? 1. No. 5.— October, 1904. 2H 224 MR. C. TATE REGAN ON THE Interorbital width 22 times in the length of head . . . . . . 16. A. schomburghkit Gthr. Interorbital width 3} times in the length of head . . . . . . 17. A. megacephalus Gthr. b. Length of mandibular ramus greater than inter- orbitalwidth': «| -Wy-cmarietcn c.f lOmed en locycenialussblan: II. Interoperculum armed with slender spines with hooked apices, which can be everted and completely retracted beneath the operculum ; sides of head without bristles; last dorsal ray connected by a well-developed membrane either to the 3 or 4 scutes following it or to the spine of the adipose fin. (Parancistrus Blkr.) A. Last dorsal ray not connected to the adipose fin . . . 19. A. niveatus Casteln. B. Last dorsal ray connected to the adipose fin. Interorbital width 3 times in the length of head; caudal peduncle twice as Jong as deep » 6 09 8 6 6 6 6 6 6 6 5 0 Hh AL Osis Sale. Interorbital width 22-22 times in the length of head; candal peduncle about 1} times as long asdeep . . . . . . . 2). A. aurantiacus Casteln. III. Interoperculum armed with moderately stout spines, which can be everted and completely retracted beneath the operculum ; external to these some slender bristles ; sides of the snout with or without short bristles; last dorsal ray connected to the scute following it by an inconspicuous membrane at its base. (Lasiancistrus, subgen. nov.) A. Bristles longer than the interopercular spines. Interoperculum with about 12 spines; pectoral spine extending beyond middle of ventral; interorbital width 13 times in the les Orel “Gog 66 6 8 5 6 6 6 8 9 6 6 6 6 ey eh ldaareoradimny (Crate, Interopereulum with 12-15 spines ; pectoral spine elongate ; inter- orbital width 24 times in length of head; large round light spots on head, body, and fins . . . . - .. . . ~ . 28. A. pictus Casteln. Interoperculum with about 20 spines; pectoral spine extending a little beyond base of ventral; interorbital width 23 times in the length of head; body with dark spots or markings . . . 24. A. mystacinus Kner. B. Bristles shorter than the interopercular spines. . . . 20. A. guacharote C. & V. IV. Interoperculum armed with slender spines which cannot be completely retracted beneath the operculum, or with permanently everted bristles ; sides of the head with bristles which are always longer in the male than in the female and may be wanting in the latter; last dorsal ray connected to the scute following it by an inconspicuous membrane at its base or by a more developed membrane to the 2 scutes succeeding it. Abdomen completely naked. (Pseudancistrus Blkr.) A. Interoperculum armed with spines; dorsal with I 7-8 rays; length of mandibular ramus greater than the MeO OAL WAM G6 6 Gl Bek oO Ok 6 oo 0 Oth ey GargneNs Ilene, FISHES OF THE FAMILY LORICARITD#. 225 B. Interoperculum armed with bristles; length of mandibular ramus less than interorbital width. 1. Dorsal with I 7 rays. a. Length of base of dorsal nearly equal to its distance from the caudal. pee width 34-34 times in the length of head; caudal peduncle ri) 4 times as long as deep 7. A. barbatus C.& V. Interovbita width 4 times in the length of Teal caudal peduncle 2 aavimes) as) longsasideepmra tlie ions ine SubAnidennessusiGLnr: 6. Length of base of dorsal nearly equal to its distance irom the adipose nga eae aeE wey a ura 2 O-eeAergarentivenielvon 2. Dorsal with I 9-10 rays. D.19. 24 scutes ina longitudinal series. . . . . . . . . 80. A. angulicouda Stdr. D.110. 25-27 scutes in a longitudimal series . . . . . . . 31. A. parahybe Kigenm. J. ANCISTRUS ANNECTENS, sp.n. (Plate XI. fig. 2.) Depth of body 43-5 times in the total length, length of head about 34 times. Head 13-1; times as long as broad and 15-12 times as long as deep. Diameter of eye 7-8 times in the length of head, interorbital width 23-23 times, length of snout 1% times. Length of mandibular ramus 3 times in the interorbital width. Snout narrowed anteriorly; supraoccipital with strong median ridge; supraorbital edges raised; temporal plates carinate; interoperculum armed with 5 or 6 slender spines, the longest equal to 4 the length of head. Scutes spinulose, the anterior and upper ones carinate, 26-27 in a longitudinal series, 7-8 between dorsal and adipose fin, 13 between anal and caudal. Supraoccipital bordered posteriorly by a median scute and by one on each side; two series of scutes between the nuchal and posthumeral series. Lower surface of head and -abdomen completely covered with small granular scales. D. 17, the first ray equal to the length of head, the last } as long; length of base of dorsal greater than its distance from the adipose fin. A.I4. Pectoral spine extending beyond base of ventral. Caudal emarginate, the outer rays produced. Caudal peduncle 53-53 times as long as deep. Dark spots on head, body, and fins; dorsal with a series of especially large and distinct spots, one at the base of cant interradial membrane. Total length 280 mm. Two specimens from St. Javier and the Rio Durango, N.W. Ecuador, collected by Mr. Rosenberg, referred by Boulenger to Plecostomus bicirrhosus. I have given this species the name “ annectens”’ because it has the general form of Plecostomus guacari and of the species of the old genus Péerygoplichthys, and whilst agreeing with the former in the number of dorsal rays, has the movable inter- operculum of the latter. 2H 2 226 MR. C. TATE REGAN ON THE 2, ANCISTRUS UNDECIMALIS. Chetostomus undecimalis Steind. Denkschr. Ak. Wien, xxxix. 1878, p. 43, pl. viii., and xlii. 1880, p. 67. Plerygoplichthys undecimalis Kigenm. & Kigenm. Proc. Cal. Ac. (2) ii. 1889, p. 44. Depth of body 44 times in the total length, length of head 3} times. Head 1} times as long as broad and 13 times as long as deep. Diameter of eye 7} times in the length of head, interorbital width 27 times, length of snout twice. Length of mandi- bular ramus about 4 times in the interorbital width. Snout narrowed anteriorly ; supraoccipital with median ridge; supraorbital edges not or scarcely raised; temporal plates carinate ; interoperculum armed with 6-8 slender spines, the longest equal to + the length of head. Scutes spinulose, carinate except on the caudal peduncle, 28-29 in a longitudinal series, 8-9 between dorsal and adipose fin, 12-13 between anal and caudal. Supraoccipital entirely bordered posteriorly by a single scute; two series of scutes between the nuchal and posthumeral series. Lower surface of head and abdomen completely covered with small granular scales. D. I 10, the first ray equal to 8 the length of head, the last $ as long; length of base of dorsal equal to its distance from the caudal. A. I 3-4. Pectoral spine extending nearly to middle of ventral. Caudal emarginate, the shortest ray 3 the length of the longest, which is 1+ times as long as the head. Caudal peduncle 3 times as long as deep. Greyish, with dark spots on head, body, and fins, those on the head smallest, those on the fins in a single series between each pair of rays. Total length 239 mm. One specimen: R. Magdalena. 3. ANCISTRUS ETENTACULATUS. Hypostoma etentaculatum Spix, Gen. Spec. Pisce. p. 7, pl. iv. figs. 1 & 2 (1829). Hypostomus duodecimalis Cuv. & Val. Hist. Nat. Poiss. xv. p. 498, pl. eeccliv. (1840). Aneistrus longimanus Kner, Denkschr, Ak. Wien, vil. 1854, p. 283. Pterygoplichthys duodecimalis Giinth. Cat. Fish. v. p. 251 (1864). Pierygoplichthys etentaculatum Eigenm. & Eigenm. Proc. Cal. Ac. (2) ii. 1889, p. 44, and Occ. Pap. Cal. Ac. i. 1890, p. 428. Depth of body 43 times in the total length, length of head 3 times. Head 12 times as long as broad and 14 times as long as deep. Diameter of eye nearly 8 times in the length of head, interorbital width 24 times, length of snout 13 times. Length of mandibular ramus 3} times in the interorbital width. Snout narrowed anteriorly ; supraoccipital with median ridge ; supraorbital edges slightly raised ; temporal plates carinate ; interoperculum armed with 10-12 slender spines, the longest equal to 4 the length of head. Scutes spinulose, carinate except on the caudal peduncle, 28 in a longitudinal series, 6 between dorsal and adipose fin, 12 between anal and caudal. Supraoccipital entirely bordered posteriorly by a single scute; two series of scutes FISHES OF THE FAMILY LORICARIIDA. DOT; between the nuchal and posthumeral series. Lower surface of head and abdomen completely covered with small granular scales. D. 111; length of base of dorsal equal to 14 times its distance from the caudal. A.1I4. Pectoral spine extending to last + of ventral. Caudal emarginate. Caudal peduncle 3 times as long as deep. Irregular dark spots on head and body; fins dark with reticulating light lines. Total length 400 mm. ‘Two specimens (including the type of H. duodecimalis): Rio San Francisco, Brazil. 4, ANCISTRUS GIBBICEPS. Ancistrus gibbiceps Kner, Denkschr. Ak. Wien, vii. 1854, p. 284, pl. v. fig. 2. Liposarcus altipinnis Giinth. Cat. Fish. v. p. 239 (1864). Pterygoplichthys gibbiceps Giinth. t.c. p. 252; Eigenm. & Higenm. Proce. Cal. Ac. (2) ii. 1889, p. 44, and Oce. Pap. Cal. Ac. i. 1890, p. 429. Liposarcus scrophus Cope, Proc. Ac. Philad. 1874, p. 136. Chetostomus gibbiceps Steind. Denkschr. Ak. Wien, xlii. 1881, p. 114, pl. iv. fig. 1. Depth of body 44-42 times in the total length, length of head 23-3 times. Head as broad as long and 1} times as long us deep. Diameter of eye 63-73 times in the length of head, interorbital width 2-23 times, length of snout 1% times. Length of mandibular ramus 33-33 times in the interorbital width. Snout narrowed anteriorly ; supraoccipital with a strongly elevated median ridge; supraorbital edges scarcely raised ; temporal plates carinate; interopcrculum armed with a few slender spines, the longest less than $ the diameter of eye. Scutes spinulose, carinate except on the caudal peduncle, 28-30 in a longitudinal series, 5-6 between dorsal and adipose fin, j2-14 between. anal and caudal. Supraoccipital bordered posteriorly by a median scute and one on each side; 3-4 series of scutes between nuchal and posthumeral series. Lower surface of head and abdomen completely covered with small granular scales. D. 1 12-13, the first ray equal in length to the head; length of base of dorsal equal to about 14 times its distance from the caudal. A.I 4. Pectoral spine extending nearly to middle of ventral. Caudal emarginate, the outer rays produced. Caudal peduncle about 3 times as long as deep. Head, body, and fins with rather large dark spots separated by reticulating light lines; spots on the head smallest, on the occipital region more or less regularly radiating. Total length 280 mm. Three specimens: R. Amazon. 5. ANCISTRUS LITURATUS. Ancistrus lituratus Kner, Denkschr. Ak. Wien, vii. 1854, p. 285, pl. v. fig. 3. Pterygoplichthys lituratus Giinth. Cat. Fish. v. p. 252 (1864); Higenm. & Eigenm. Proc. Cal. Ac. (2) i. 1889, p. 45, and Occ. Pap. Cal. Ac. i. 1890, p. 433. Chelostomus lituratus Steind. Denkschr. Ak. Wien, xlii. 1881, p. 115. 238 MR. C. TATE REGAN ON THE Depth of body about 43 times in the total length, length of head 5 times. Head slightly longer than broad and 12 times as long as deep. Diameter of eye 7} times in the length of head, interorbital width 2% times, length of snout 14 times. Length of mandibular ramus 52 times in the interorbital width. Snout narrowed anteriorly ; supraoccipital with a strongly elevated median ridge; supraorbital edges scarcely raised ; temporal plates carinate; interoperculum armed with 12-15 spines, the longest 15 times the diameter of eye. Scutes mostly carinate, 28-29 in a longitudinal series, 6-7 between dorsal and adipose fin, 15 between anal and caudal. Supraoccipital bordered posteriorly by a median scute and one on each side; 3 series of scutes between nuchal and posthumeral series. Lower surface of head and abdomen completely covered with small granular scales, D. I 12-15; length of base of dorsal about 13 times its distance from the caudal. A.I1I4. Pectoral spine extending to anterior { of ventral; caudal emarginate. Caudal peduncle 23 times as long as deep. Dark, with light spots or vermiculations on head, body, and ae Total length 190 mm. One specimen: R. Amazon; Eastern Brazil. 6. ANCISTRUS MULTIRADIATUS. Aypostomus multiradiatus Hancock, Zool. Journ. iv. 1828, p. 246. HAypostomus pardalis Casteln, Anim, Am, Sud, p. 42, pl. xx. fig. 3 (1855). Liposarcus multiradiatus Giinth, Cat. Fish. v. p. 288 (1864). Liposarcus pardalis Giinth. t. e. p. 289. Liposarcus varius Cope, Proc. Ac. Philad. 1871, p. 284. Liposarcus jeanesianus Cope, ib. 1874, p. 135. Plecostomus pardalis Peters, Monatsb. Ak. Berl. 1877, p. 471; Steind. Denkschr. Ak. Wien, xliii. 1881, p. 112; ib. xlvi. 1882, p. 6. Pterygoplichthys pardalis Kigenm. & Higenm. Proc. Cal. Ac. (2) i. 1889, p. 45; Occ. Pap. Cal. Ac. 1. 1890, p. 431. Pterygoplichthys jeanesianus Figenm. & Eigenm. |. ec. et t. c. p. 433. Pierygoplichthys multiradiatus Kigenm. & EKigenm. Occ. Pap. Cal. Ac. i. 1890, p. 433. Depth of body 43-54 times in the total length, length of head 3-33 times. Head 1% times as long as broad and 13-13 times as long asdeep. Diameter of eye 63-9 times in the length of head, interorbital width 2-24 times, length of snout twice. Length of mandibular ramus 34-4 times in the Harned width. Snout somewhat narrowed anteriorly, entirely granular; supraoccipital with low median ridge; supra- orbital edges not or very slightly raised; temporal plates usually carinate; inter- operculum not notably spinate. Scutes spinulose, mostly carinate, 28-30 in a longitudinal series, 6 between dorsal and adipose fin, 11-13 between anal and caudal. Occipital bordered posteriorly by a median scute and by 1 or 2 scutes on each side ; 3 or 4 series of scutes between the nuchal and posthumeral series. Lower surface, in the adult, completely covered with small granular scales. D. I 12-13, the first ray less than the length of head in the adult; length of base of dorsal 14-14 times its FISHES OF THE FAMILY LORICARIID®. 229 distance from the caudal. A.I4. Pectoral spine extending to the anterior third or middle of ventral. Caudal emarginate, the outer rays produced. Caudal peduncle 3-37 times as long as deep. Dark spots or vermiculations on the sides of the body; in the young, fins with dark cross-bars, in the adult with large dark spots. Total length 590 mm. A.—Head with dark vermiculations and spots on a light background, those of the supraoccipital and interorbital regions with a more or less regular radiating arrangement. Abdomen with large dark spots on a light ground-colour. Seven specimens (including the types of the species and of H. pardalis): R. Amazon ; Guiana. B. A. multiradiatus var. alternans, var. nov.—Head and abdomen with light spots or vermiculations on a dark background. Three specimens: Paraguay and Southern Bolivia. 7. PANCISTRUS PUNCTATUS. Ancistrus duodecimalis (non Cuy. & Val.) Kner, Denkschr. Ak. Wien, vil. 1854, p. 281. Pieryyoplichthys punctatus Giinth. Cat. Fish. v. p. 251 (1864); Higenm. & Eigenm. Proc. Cal. Ac. (2) ii. 1889, p. 45; Occ. Pap. Cal. Ac. i. 1890, p. 431. Chetostomus punctatus Steind. Denkschr. Ak. Wien, xliti. 1881, p. 115. This species is stated to resemble A. gibbiceps in form. In the number of scutes and fin-rays, in the strongly carinate scutes, the low occipital crest, and the size of the eye it is similar to A. multiradiatus, from which it appears to differ notably in the narrower interorbital width (22-22 times in the length of head), the interoperculum armed with a few spines, about 2 of which are moderately elongate in the adult, and the pectoral spine extending only a little beyond the base of ventral. Head, body, and fins with rather small round blackish spots. Venezuela; Upper and Middle Amazon. 8. ANCISTRUS MEDIANS. Ancistrus medians Kner, Denkschr. Ak. Wien, vil. 1854, pp. 286 and 281. Chetostomus medians Giinth, Cat. Fish. v. p. 242 (1864). Hemiancistrus medians Kigenm. & Higenm. Proc. Cal. Ac. (2) 1. 1889, p. 43. Depth of body 43 times in the total length, length of head 2% times. Head as long as broad and 13 times as long asdeep. Diameter of eye 4—4} times in the length of head, interorbital width 2$-5 times, length of snout 13-1# times. Length of mandibular ramus 1$ times in the interorbital width. Snout broad and rounded ; supraorbital margins raised ; supraoccipital with median elevation ; temporal plates carinate or not; operculum reduced; interoperculum armed with 30 or more rather slender spines with curved tips, the longest equal to 1} times the diameter of eye. Scutes strongly spinulose, mostly carinate, 24 in a longitudinal series, 6-7 between 230 MR. C. TATE REGAN ON THE dorsal and adipose fin, 13 between anal and caudal. Supraoccipital entirely bordered posteriorly by a single scute. Lower surface of head and abdomen, in the adult, to a considerable extent covered with small granular scales. D. 1 7, the first ray $-3 the length of head, the last 3-3 as long; length of base of dorsal little less than its distance from the caudal. A. I 4. Pectoral spine extending nearly to middle of ventral. Caudal obliquely emarginate or truncate. Caudal peduncle 2$-3 times as long as deep. Rather large round dark spots on head, body, and fins. Total length 220 mm. ‘Two specimens: Surinam. 6 9, ANCISTRUS GIGAS. ? Chetostomus aculeatus Perugia, Ann. Mus. Genoy. (2) x. 1891, p. 637. Chetostomus gigas Bouleng. Proc. Zool. Soc. 1895, p. 526, and Trans. Zool. Soc. xiv. 1896, p. 30, pl. vi. Depth of body 43 times in the total length, length of head 25 times. Head 1} times as long as broad and 1# times as long as deep. Diameter of eye 13 times in the length of head, interorbital width 24 times, length of snout 13 times. Length of mandibular ramus 52 times in the interorbital width. Snout broad, rounded ; supraorbital edges not raised; occipital and interorbital regions flat ; interoperculum armed with very numerous rather slender spines with curved tips, the longest equal to ‘the length of head. Scutes strongly spinulose, all carinate, 25 in a longitudinal series, 5 between dorsal and adipose fin, 11 between anal and caudal. Supraoccipital bordered posteriorly by a single scute. Lower surface completely covered with small granular scales. D. 110; length of base of dorsal 14 times its distance from caudal. A.15. Pectoral spine extending to middle of ventral. Caudal truncate. Caudal peduncle twice as long as deep. Rather large dark spots on head, body, and fins. Total length 550 mm. One specimen: Asuncion, Paraguay. 10. ANCISTRUS ASPIDOLEPIS. Chetostomus aspidolepis Giinth. Proc. Zool. Soc. 1866, p. 603; Trans. Zool. Soc. vi. 1866, p. 477. Hemiancistrus aspidolepis Bigenm. & Eigenm. Proc. Cal. Ac. (2) 11. 1889, p. 43. Depth of body 5} times in the total length, length of head 3 times. Head 1} times as long as broad and twice as long as deep. Diameter of eye about 8 times in the length of head, interorbital width 22 times, length of snout 1} times. Length of mandibular ramus 14 times in the interorbital width. Snout broad, rounded; supra- orbital edges slightly raised ; supraoccipital convex, without distinct ridge; temporal plates carinate; interoperculum armed with a few short spines. Scutes spinulose, strongly carinate, 25 in a longitudinal series, 7 between dorsal and adipose fin, 12 between anal and caudal. Supraoccipital bordered posteriorly by a single scute. FISHES OF THE FAMILY LORICARIID®. Date Abdomen in great part covered with eranular scales. D.I 7, the first ray as long as the head; length of base of dorsal a little more than its distance from adipose fin. A. I 4. Pectoral spine extending beyond base of ventral. Caudal emarginate. Caudal peduncle about 23 times as long as deep. Numerous small dark spots on head, body, and fins. Total length 280 mm. One specimen (a skin): Veragua, Pacific Slope of Panama. 11. fANcISTRUS BRACHYURUS. Ancistrus pictus (non Casteln.) Kner, Denkschr, Ak. Wien, vil. 1854, p. 277, pl. iv. fig. 2. Ancistrus brachyurus Kner, t. ec. p. 279, pl. iv. fig. 1. Chetostomus pictus Giinth. Cat. Fish. v. p. 242 (1864). Chetostomus brachyurus Giinth. t. c. p. 243. Hemiancistrus pictus Kigenm. & Eigenm. Proc. Cal. Ac. (2) i. 1889, p. 43; Kindle, Ann. Ac. N. York, viii. 1895, p. 254. Hemiancistrus brachyurus Kigenm. & Eigenm. 1. c.; Kindle, t. c. p. 255. Apparently closely allied to A. aspidolepis, with which it agrees in the depressed head with broad snout and the strongly carinate scutes. Diameter of eye 5 times in the length of head (in specimens 150 mm. in total length). Interoperculum with about 20 setiform spines, the longest equal to $ the length of head. 23-24 scutes in a longitudinal series, 12 between anal and caudal. Abdomen naked. D. I 7, the first ray shorter than the base of the dorsal, which is considerably longer than its distance from the adipose fin. A. I 5. Pectoral spine extending nearly to the middle of ventral. Caudal obliquely truncate. Head and body usually marbled with light and dark, fins with two or three dark cross-bands; in the young, body with five vertical bars, fins with one or two cross-bands, Rio Negro. 12. ANCISTRUS SCAPHIRHYNCHUS. Ancistrus scaphirhynchus Kner, Denkschr. Ak. Wien, vil. 1854, p. 280, pl. ili. fig. 2. Chetostomus scaphirhynchus Giinth. Cat. Fish. v. p. 244 (1864). Hemiancistrus scaphirhynchus Kigenm. & Kigenm. Proc. Cal. Ac. (2) 11. 1889, p. 43, and Oce. Pap. Cal. Ac. i. 1890, p. 419; Kindle, Ann. Ac. N. York, viii. 1895, p. 254. Depth of body 62 times in the total length, length of head 3 times. Head 1+ times 73 as long as broad and 2} times as long as deep. Diameter of eye 73 times in the length of head, interorbital width 1? times, length of snout twice. Length of mandibular ramus 3? times in the interorbital width. Snout broad, rounded; head nearly flat above, without keels or ridges; interoperculum armed with about 15 slender spines, the longest 15 times the diameter of eye. Scutes spinulose, carinate except those of the upper series and the anterior ones of the second series, 26-27 in a longitudinal series, § between dorsal and adipose fin, 13 between anal and caudal. VOL. XVII.—PaRT U1. No. 6.— October, 1904, AT 232 MR. C. TATE REGAN ON THE Supraoccipital bordered posteriorly by a median scute and by one on each side. Lower surface of head and abdomen naked. D. 17, the first ray 3 the length of head, the last 2 as long; length of base of dorsal equal to its distance from the spine of the adipose fin, A. I 5. Pectoral spine extending a little beyond base of ventral. Caudal slightly and obliquely emarginate. Caudal peduncle 4% times as long as deep. Brown, with pale spots. Dorsal and caudal with alternate dark and light spots on the rays. JYotal length 170 mm. A single specimen: R. Amazon. 13. ANCISTRUS OLIGOSPILUS. (Plate XII. fig. 1.) Chetostomus oligospilus Giinth. Cat. Fish. v. p. 244 (1864). Hemiancistrus oligospilus Kigenm. & Figenm. Proc. Cal. Ac. (2) i. 1889, p. 43; Kindle, Ann. Ac. N. York, vii. 1895, p. 254. Depth of body 4} times in the total length, length of head 3 times. Head 1} times as long as broad and 1} times as long as deep. Diameter of eye 43 times in the length of head, interorbital width 24 times, length of snout 13 times. Length of mandibular ramus 22 times in the interorbital width. Snout narrowed anteriorly ; supraorbital edges slightly raised ; supraoccipital with low median ridge; temporal plates not carinate ; interoperculum with about 15 slender spines, the longest equal to 14 times the diameter of eye. Scutes spinulose, not carinate, 26 in a longitudinal series, 7 between dorsal and adipose fin, 13 between anal and caudal. Supraoccipital bordered posteriorly by a median scute and by one on each side. Lower surface of head and abdomen in the young naked, except for a granular strip between the pectorals. D. I 7, the first ray a little longer than the head, the last 2 as long; length of base of dorsal considerably more than its distance from the adipose fin. A.14. Pectoral spine extending well beyond base of ventral. Caudal emarginate, the inner rays 2 as long as the outer, which are equal to the length of head. Caudal peduncle 3 times as long as deep. Large round dark spots on head, body, and fins, those on the head and the lower surface somewhat smaller than the others. Total length 107 mm. One specimen: R. Capin. 14, ANCISTRUS VITTATUS. Chetostomus vittatus Steind. Denkschr. Ak. Wien, xliii. 1881, p. 115, pl. ii. fig. 5. Hemiancistrus vittatus Eigenm. & Eigenm. Proc. Cal. Ac. (2) 11. 1889, p. 44, and Occ. Pap. Cal. Ac. i. 1890, p. 421; Kindle, Ann. Ac. N. York, vii. 1895, p. 254. Depth of body 44 times in the total length, length of head 23 times. Head 1} times as long as broad and 1} times as long as deep. Diameter of eye 5 times in the length of head, interorbital width 24 times, length of snout 13 times. Length of mandibular ramus 24 times in the interorbital width. Snout narrowed anteriorly ; supraorbital FISHES OF THE FAMILY LORICARIIDA. 233. edges slightly raised ; supraoccipital with low median elevation; temporal plate not carinate ; interoperculum armed with a few slender spines, the longest equal to the diameter of eye. Scutes spinulose, not carinate, 26 in a longitudinal series, 7 between dorsal and adipose fin, 12 between anal and caudal. Supraoccipital bordered posteriorly by a median scute and by one on each side; lower surface of head and abdomen (in the young) naked, except for a granular strip between the pectorals. D. I 7, the first ray a little shorter than the head, the last } as long; length of base of dorsal considerably more than its distance from the adipose fin, A. I 4. Pectoral spine extending well beyond base of ventral. Caudal emarginate, the inner rays 2 as long as the outer, which are equal to the length of head. Caudal peduncle 24 times as long as deep. Body crossed by 5 dark brown nearly vertical bands; fins, except the anal, each crossed by 2 similar bands ; anal with a dark spot. Total length 57 mm. One specimen: R. Amazon. 15. ANCISTRUS BACIII. Chetostomus bachi Bouleng. Trans. Zool. Soc. xiv. 1896, p. 425, pl. xli. fig. 1. Depth of body 43 times in the total length, length of head 22 times. Head 1}; times as long as broad and 1# times as long as deep. Diameter of eye 6 times in the length of head, interorbital width 1? times, length of snout 1+ times. Length of mandibular ramus 33 times in the interorbital width. Snout narrowed anteriorly ; supraorbital margins not raised ; supraoccipital without median ridge; temporal plates not carinate; interoperculum armed with 12-15 slender spines, the longest equal to 14 times the diameter of eye. Scutes margined with short spines, not carinate, 25 in a longitudinal series, 8 between dorsal and adipose fin, 12 between anal and caudal. Supraoccipital bordered posteriorly by a median scute and by one on each side. Lower surface of head and abdomen almost completely covered with small granular scales. D. I 7, the first ray nearly as long as the head, the last half as long; length of base of dorsal a little more than its distance from the adipose fin. A. I 4. Pectoral spine extending to middle of ventral. Caudal very obliquely emarginate, almost truncate, the longest ray twice as long as the shortest. Caudal peduncle 24 times as long as deep. Uniformly coloured (in spirit) except for 4 or 5 faint dark vertical bands on the caudal. Total length 112 millim. One specimen: River Jurua. 16. ANCISTRUS SCHOMBURGKII. Chetostomus schomburgkii Ginth. Cat. Fish. v. p. 245 (1864). Hemiancistrus schomburgkii Higenm. & Kigenm. Proc. Cal. Ac. (2) ii. 1889, p. 43; Kindle, Ann. Ac. N. York, viii. 1895, p. 254. Depth of body 8 times in the total length, length of head 3 times. Head nearly as By it 234 MR. C. TATE REGAN ON THE . broad as long and 22 times as long as deep. Diameter of eye 43-53 times in the length of head, interorbital width 22 times, length of snout 14 times. Length of mandibular ramus 24 times in the interorbital width. Snout broad, rounded ; supra- orbital edges not raised; supraoccipital flat, without median ridge; temporal plates not carinate; interoperculum armed with 20-25 slender spines with curved tips, the longest equal to twice the diameter of eye. Scutes spinulose, not carinate, 25 ina longitudinal series, 6 between dorsal and adipose fin, 11-12 between anal and caudal. Supraoccipital bordered posteriorly by a median scute and by one on each side. Lower surface of head and abdomen naked (in the young). D. I 7, the first ray # the length of head; length of base of dorsal nearly equal to its distance from the adipose fin. A. I 4. Pectoral spine not reaching the base of ventral. Caudal peduncle 3 times as long as deep. Brownish, clouded with darker; dark spots on the fins. Total length 75 mm. ‘Two specimens: British Guiana. 17. ANCISTRUS MEGACEPHALUS. ? Hypostomus itacua Cuv. & Val. Hist. Nat. Poiss. xv. p. 505 (1840) ; Val. Voy. d’Orbigny, Poiss. pl. vii. fig. 1 (1847). ? Chetostomus itacua Giinth. Cat. Fish. v. p. 243 (1864). Chetostomus megacephalus Giinth. Proc. Zool. Soc. 1868, p. 232. Chetostomus macrops Liitk. Vid. Medd. 1873, p. 209; Steimd. Denkschr. Ak. Wien, xlii. 1881, peel2o,pliventioe. ? Hemiancistrus itacua Kigenm. & Higenm. Proc. Cal. Ac. (2) 11. 1889, p. 43; Kindle, Ann. Ac. N. York, viii. 1895, p. 253. Hemiancistrus megacephalus Wigenm. & Higenm. t.c, p. 44; Kindle, |. ¢. Depth of body 4-4} times in the total length, length of head 2% times. Head 14 times as long as broad and 13-1? times as long as deep. Diameter of eye 5-54 times in the length of head, interorbital width 3} times, length of snout 13-12 times. Length of mandibular ramus 1} times in the interorbital width. Snout rounded anteriorly ; supraorbital margins raised; supraoccipital flat, without median ridge ; temporal plates not carinate; interoperculum armed with 50 or more rather slender spines with curved tips, the inner longest and equal to 1} times the diameter of eye. Scutes margined, spinulose, not carinate, 24 in a longitudinal series, 7 between dorsal and adipose fin, 12 between anal and caudal. Supraoccipital margined posteriorly by a median scute and by one on each side. Lower surface of head and abdomen naked. D. 17, the first ray } the length of head, the last 2 as long; length of base of dorsal nearly equal to its distance from the caudal. A.I4. Pectoral spine extending to well beyond base of ventral. Caudal weakly and obliquely emarginate. Caudal peduncle 25 times as long as deep. Hach scute on the body with a large light spot, similar smaller spots on the head. Fins with bluish cross-bands. Total length 175 mm. i) wD (eo) | FISHES OF THE FAMILY LORICARITD#. Two specimens: Guiana. Examples in the Museum at Paris, brought from the La Plata River by d’Orbigny and named H. itacua by Valenciennes, proved to be Plecostomus commersonii C. & V. I have little doubt, however, that the species described and figured by Valenciennes is closely allied to Chetostomus megacephalus Gthr. and may be identical with it. 18. ANCISTRUS PLATYCEPHALUS. (Plate XII. fig. 4.) Chetostomus platycephalus Bouleng. Boll. Mus. Torino, xiii. 1898, no. 329, p. 9. Depth of body 53 times in the total length, length of head 2# times. Head a little broader than long and 21 times as long as deep. Diameter of eye 11 times in the length of head, interorbita¥ width 532 times, length of snout 12 times. Length of mandibular ramus equal to 14 times the interorbital width. Head without ridges or prominences, flat; snout very broad, rounded; interoperculum with 25-30 spines, the longest equal to about + the length of head. Scutes not carinate, the posterior rather strongly spinate, 24 in a longitudinal series, 5 between dorsal and adipose fin, 10 between anal and caudal; supraoccipital bordered posteriorly by a pair of scutes, which exclude the first median scute from contact with it; lower surface of head and abdomen naked. D. I 8, the first ray equal to 3 the length of head, the last 4 as long; length of base of dorsal a little less than its distance from the caudal. A. I 4. Adipose fin well developed. Pectoral spine extending to base of ventral. Caudal obliquely truncate. Caudal peduncle 25 times as long as deep. Olivaceous. Total length 140 mm. A single specimen: E. Ecuador. 19, ANCISTRUS NIVEATUS. Hypostomus niveatus Casteln. Anim. Am. Sud, p. 43, pl. xxi. fig. 3 (1855). Chetostomus niveatus Giinth. Cat. Fish. v. p. 243 (1864). Hemiancistrus longipinnis* Kindle, Ann. Ac. N. York, vii. 1895, p. 255. Depth of body about 44 times in the total length, length of head about 3 times. Head as broad as long and 14 times as long as deep. Diameter of eye about 7 in the length of head, interorbital width 24 times, length of snout about 12 times. Length of mandibular ramus about 13 times in the interorbital width. Snout broad, obtuse; supraorbital edges not raised; supraoccipital flat; temporal plates not carinate ; interoperculum armed with numerous long slender spines, with curved tips, the longest equal to } the length of head. Scutes spinulose, not carinate, 24 ina longitudinal series, 5 between dorsal and adipose fin, 13 between anal and caudal; times supraoccipital entirely bordered posteriorly by a single scute; lower surface of head and abdomen naked. D. 17, the first ray 3 the length of head, the last 3-} as long, ’ Professor §. H. Gage, of Cornell University, has kindly sent me a drawing of the type of H. longi- vinnis, which seems almost certainly identical with A. niveatus. . 236 MR. C. TATE REGAN ON THE connected by a membrane to the 3 scutes following it. Length of base of dorsal a little more than its distance from the caudal. A.I4. Pectoral spine extending to middle of ventral. Caudal obliquely and weakly emarginate or truncate. Caudal peduncle about 22 times as long as deep. Large round white spots on head, body, and fins. Total length 280 mm. : Two specimens, stuffed (types of the species): R. Araguay; R. Tocantins. 20. fANCISTRUS PUNCTATISSIMUS. Chetostomus punctatissimus Steind. Denkschr. Ak. Wien, xlii. 1881, p. 121, pl. ii. fig. 3. Parancistrus punctatissimus Higenm. & Higenm. Proc. Cal. Ac. (2) ii. 1889, p. 44; Occ. Pap. Cal. Ac. i. 1890, p. 423. Depth of body about 5 times in the total length, length of head 3 times. Depth of head about 1? times in its length, diameter of eye 6 times, interorbital width 3 times. Snout rounded anteriorly; supraorbital edges slightly raised; supraoccipital flat, without median ridge; temporal plates not carinate; interoperculum armed with slender spines, the longest equal to about twice the diameter of eye. Scutes spinulose, not carinate, 22 in a longitudinal series, 6 between dorsal and adipose fin, 11 between anal and caudal. Supraoccipital bordered posteriorly by a median scute and by one on each side. Lower surface of head and abdomen naked. D. I 7, the first ray equal to about $ the length of head, the last about 3 as long, and entirely connected by membrane to the spine of the adipose fin; length of base of dorsal greater than its distance from caudal. A.1I4. Pectoral spine extending beyond base of ventral. Caudal weakly and obliquely emarginate. Caudal peduncle twice as long as deep. Brown, with white or light blue dots on head, body, and fins. R. Amazon. The above is taken from Steindachner’s description and figure of a specimen 135 mm. in total length. 21. ANCISTRUS AURANTIACUS. Hypostomus aurantiacus Casteln. Anim. Am. Sud, p. 43, pl. xxi. fig. 2 (1855). Hypostomus nigricans Casteln. t. c. p. 44, pl. xxii. fig. 1. Hypostomus vicinus Casteln. t. ec. p. 45, pl. xxiii. fig. 1. Parancistrus aurantiacus Bleek. Ned. Tijd. Dierk. i. 1863, p. 79. Chetostomus aurantiacus Giinth. Cat. Fish. v. p. 246 (1864). Chetostomus nigricans Ginth. 1. c. Parancistrus nigricans Kigenm. & Eigenm. Proc. Cal. Ac. (2) ii. 1889, p. 44. Depth of body about 43 times in the total length, length of head 23-24 times. - Head broader than long and 1$-2 times as long as deep. Diameter of eye 6-8 times in the length of head, interorbital width 22-22 times, length of snout 12 times. FISHES OF THE FAMILY LORICARIIDA. 237 Length of mandibular ramus about 34 times in the interorbital width. Snout broad, rounded ; supraorbital edges not raised; supraoccipital flat; temporal plates not carinate. Scutes spinulose, not carinate, 22-23 in a longitudinal series, 5-6 between dorsal and adipose fin, 9 between anal and caudal. Supraoccipital bordered posteriorly by a median scute and by one on each side. Lower surface of head and abdomen naked. D.17, the first ray 2 the length of head, the last # as long, joined by a membrane to the adipose fin; length of base of dorsal greater than its distance from the caudal. A.14. Pectoral spine stout, bristly, extending to base (young) or anterior 4 (adult) of ventral. Caudal obliquely truncate. Caudal peduncle about 14 times as long as deep. Total length 240 mm. Three specimens (types of the species, of H. nigricans and I. vicinus respectively) : Peruvian Amazon. The largest specimen is stuffed and is orange-red in colour; the two smaller ones in spirit are blackish. ‘The one named /. vicinus has the spine of the adipose fin small, recumbent, and immovably joined to the scute below it. 22. ANCISTRUS HETERACANTHUS. Chetostomus heteracanthus Giinth. Proc. Zool. Soc. 1869, p. 425, figs. 3 & 4. Hemiancistrus heteracanthus Eigenm. & Eigenm. Proc. Cal. Ac. (2) 11. 1889, p. 44. Depth of body 5} times in the total length, length of head 23 times. Head 1} times as long as broad and twice as long as deep. Diameter of eye 83 times in the length of head, interorbital width 13 times, length of snout 13 times. Length of mandibular ramus 4 times in the interorbital width. Snout margined with very short bristles ; supraorbital edges not raised; supraoccipital flat; temporal plates not carinate; interoperculum armed with 12 stout spines, the longest equal to a little more than 4 the length of head; external to these, and overlying them, some long slender bristles. Scutes spinulose, not carinate, 24 in a longitudinal series, 7 between dorsal and adipose fin, 12 between anal and caudal. Supraoccipital bordered posteriorly by a median scute and by one on each side. Lower surface of head naked; abdomen with some small scattered granules. D.17, the first ray equal to 3 the length of head, the last 2 as long; length of base of dorsal nearly equal to its distance from the adipose fin. A.15. Pectoral spine extending to beyond middle of ventral. Caudal peduncle 2? times as long as deep. Olivaceous; a blackish spot ai the base of the first dorsal ray. Total length 180 mm. One specimen: Peruvian Amazon. 23. ANCISTRUS PICTUS. Hypostomus pictus Casteln. Anim. Am. Sud, p. 44, pl. xxii. fig. 2 (1855). Depth of body about 6 times in the total length, length of head 2} times. Head 258 MR. C. TATE REGAN ON THE 1; times as long as broad and twice as long as deep. Diameter of eye about 6 times in the length of head, interorbital width 24 times, length of snout 13 times. Length of mandibular ramus 2} times in the interorbital width. Head very similar to that of A. heteracanthus, the interoperculum armed with 12-15 moderately stout spines, which are covered by a group of long bristles ; sides of the snout also bearing bristles of moderate length. Scutes similar to those of A. heteracanthus both in structure and in number. Supraoccipital bordered posteriorly by a median scute and by one on each side. Lower surface of head and abdomen naked. D.17; length of base of dorsal equal to its distance from the adipose fin. A.I 4. Pectoral spine strong, elongate (broken in the type). Caudal peduncle 23 times as long as deep. Large, round, light spots on head, body, and fins. Length, to base of caudal, 108 mm. One specimen, in very bad condition (type of the species): Ucayale, Upper Amazon. 24. fANCISTRUS MYSTACINUS. Ancistrus mystacinus Kner, Denkschr. Ak. Wien, vii. 1854, p. 276. Chetostomus mystacinus Giinth. Cat. Fish, v. p. 244 (1864) ; Steind. Denkschr. Ak. Wien, xliii. 1881, p. 125. Hemiancistrus mystacinus Eigenm. & Eigenm. Proe. Cal. Ac. (2) 11. 1889, p. 43. Depth of body about 6 times in the total length, length of head 3 times. Head 1+ times as long as broad. Diameter of eye 7 times in the length of head, interorbital width 24 times, length of snout 13 times. Head very similar to that of A. hetera- canthus, the sides of the snout with very short bristles, the interoperculum armed with about 20 rather stout spines, external to these some long bristles which cover the spines when they are retracted. Scutes similar to those of A. heteracanthus in structure and number. Lower surface of head and abdomen naked. D. I 7, as high as long, the first ray scarcely longer than the ventral. A. I 5. Pectoral spine extending a little beyond the base of ventral. Caudal deeply emarginate. Olivaceous ; fins spotted or marbled with darker; lower lobe of caudal blackish. | The type, from Caracas, is (according to Steindachner) 250 mm, in total length. , 25. ANCISTRUS GUACIIAROTE. Hypostomus guacharote Cuv. & Val. Hist. Nat. Poiss. xv. p. 508 (1840). ? Ancistrus guacharote* Gill, Ann. Lye. N. York, vi. 1858, p. 409. Chetostomus guacharote Giinth. Cat. Fish. v. p. 245 (1864). ? Chatostomus trinitatis Giinth. t. ce. p. 246. 1 Dr. Gill has very kindly informed me that the examples which he recorded from Trinidad, and on which Dr. Giinther founded C. trinitatis, cannot be found, Therefore it is still uncertain whether or no the species C. trinitatis is valid. FISHES OF THE FAMILY LORICARIID®. 239 Hemiancistrus guacharote Eigenm. & Eigenm. Proc. Cal. Ac. (2) ii. 1889, p. 48, and Oce, Pap. Cal. Ac. i. 1890, p. 421. ? Hemiancistrus trinitatis Eigenm., & Eigenm. 1. c. et 1. c. Depth of body 6 times in the total length, length of head 2% times. Head 14 times as long as broad and twice as long as deep. Diameter of eye 5% times in the length of head, interorbital width 24 times, length of snout 13-14 times. Length of mandibular ramus 2} times in the interorbital width. Head similar to that of A. heteracanthus ; interoperculum armed with about 20 moderately stout spines, external to these some bristles which are shorter than the spines. Scutes similar to those of A. heteracanthus both in structure and number. Supraoccipital bordered posteriorly by a median scute and by one on each side. Lower surface of head and abdomen naked, except for a small granular patch behind each gill-cleft. D.1 7, the first ray equal to = the length of head, the last 2 as long; length of base of dorsal a little less than its distance from the adipose fin. A.1I4. Pectoral spine extending to anterior + of ventral. Caudal deeply emarginate. Caudal peduncle 2% times as long as deep. Uniformly olivaceous. Total length 100 mm. One specimen (type of the species): Porto Rico! ; (? Trinidad). 26. ANcIsTRUS SETOSUS. (Plate XII. fig. 2.) Chetostomus setosus Bouleng. Ann. Mag. Nat. Hist. (5) xix. 1887, p. 349. Pseudancistrus setosus Higenm. & Higenm. Proc. Cal. Ac. (2) 11. 1889, p. 45. Depth of body 6-64 times in the total length, length of head nearly 5 times. Head as broad as long and 24 times as long as deep. Diameter of eye 10-12 times in the length of head, interorbital width nearly 4 times, length of snout 1} times. Length of mandibular ramus 1¥ times greater than the interorbital width. Head without ridges or prominences, flat above; snout broad, rounded, its sides margined with bristles, which are scarcely visible in the female and equal to 7 the length of head in the male; interoperculum movable, armed with slender spines, which can be, to a considerable extent, everted and retracted, about 15 in the female and 25 in the male, the longest equal to nearly 2 the length of head. Scutes spinulose, not carinate, 25 in a longitudinal series, 5-6 between dorsal and adipose fin, 10 between anal and caudal ; supraoccipital bordered by a pair of scutes which exclude the first median scute from contact with it; lower surface of head and abdomen naked. D. I 7-8, the first ray equal to 2 the length of head; length of base of dorsal a little less than its distance from caudal. A.13. Pectoral spine extending to anterior } or nearly to middle of ventral, Caudal obliquely truncate. Caudal peduncle 23-24 times as long as deep. Olivaceous, body marbled, fins barred with dark spots. Total length 115 mm. ‘lwo specimens: Colombia. * The occurrence of this species at Porto Rico is remarkable and requires confirmation, VOL. XVII.—PakT 11. No. 7.—October, 1904. 2k 240 MR. C. TATE REGAN ON THE _ 27. ANCISTRUS BARBATUS. i Hypostomus barbatus Cuv. & Val. Hist. Nat. Poiss. xv. p. 506 (1840); Schomburgk, Fish. Guiana, p. 147 (1841) ; Kner, Denkschr. Ak. Wien, vil. 1854, p. 268, pl. i. fig. 2. Hypostomus guttatus Cuv. & Val. t. c. p. 508. Plecostomus barbatus Giinth. Cat. Fish. v. p. 237 (1864). Pseudancistrus guttatus Bleck. Nat. Verh. Holl. Maats. xx. 1864, Silures de Suriname, p. 10, pl. ii. fig. 2, and pl. ii. fig. 3. Pseudancistrus barbatus Bigenm. & Eigenm. Proc. Cal. Ac. (2) i. 1889, p. 45. . Depth of body 515% times in the total length, length of head 3-33 times. Head a little longer than broad and twice as long as deep. Diameter of eye 5-74 times in the length of head, interorbital width 33-3} times, length of snout 13 times. Length of mandibular ramus 14-12 times in the interorbital width. Head without ridges or prominences, flat above; snout not naked at tip, rounded, its sides margined with bristles, which are short in the female and equal to 2 the length of head in the male ; interoperculum moderately movable, armed with similar bristles, which are permanently eyerted, more numerous in the male. Scutes spinulose, not carinate, 24-25 in a longitudinal series, 6 between dorsal and adipose fin, 11-12 between anal and caudal. Supraoccipital bordered posteriorly by a pair of lateral scutes, and usually to some extent by a median scute. Lower surface of head and abdomen naked. D. I 7, the first ray equal to #-#+the length of head, the last $-$ as long; length of base of dorsal nearly equal to its distance from the caudal. A. I 4-5. Pectoral spine extending a little beyond base of ventral. Caudal slightly and obliquely emarginate or truncate. Caudal peduncle about 2} times as long as deep. Brownish, with small white spots on the head, and somewhat larger light spots on body and fins. Total length 260 mm. Twelve specimens: Guiana. 28. ANCISTRUS DEPRESSUS. Chetostomus depressus Giinth. Proc. Zool. Soc. 1868, p. 232; Steind. Denkschr. Ak. Wien, xlin. 188], p. 128, pl. v. fig. 1. Pseudancistrus depressus Rigenm. & Eigenm. Proc. Cal. Ac. (2) 11. 1889, p. 45. Depth of body 7 times in the total length, length of head 3 times. Depth of head 21 times in its length, diameter of eye 5 times, interorbital width 4 times. Length of mandibular ramus 1+ times in the interorbital width. Caudal peduncle 22 times as Jong as deep. Brownish, with small white spots on head, body, and fin-rays. In all other respects similar to the preceding species. Total length 130 mm. One specimen: Surinam. FISHES OF THE FAMILY LORICARITDA#, 241 29. ANCISTRUS GUENTHERI, sp. n. (Plate XII. fig. 3.) Plecostomus guttatus (non Cuv. & Val.) Giinth. Cat. Fish. v. p. 237 (1864). Depth of body 6 times in the total length, length of head 54 times. Depth of head twice in its length, diameter of eye 8 times, interorbital width 5% times, length of snout 14 times. Length of mandibular ramus 14 times in the interorbital width. Head, in other respects, similar to that of d. barbatus, except that the snout has a broad naked area at its tip. Scutes as in A. barbatus. Lower surface of head and abdomen entirely naked. D.1I 7, the first ray equal to = the length of head, the last 3 as long; Jength of base of dorsal nearly equal to its distance from the adipose fin. A. I 4. Pectoral spine not quite reaching the base of ventral. Caudal obliquely truncate. Caudal peduncle 24 times as long as deep. Colour uniform (in spirit). Total length 120 mm. One specimen: British Guiana. 30. FANCISTRUS ANGULICAUDA. Plecostomus angulicauda Steind. Sitzb. Ak. Wien, Ixxiv. 1876, p. 672, pl. xii. Delturus angulicauda, Kigenm. & Eigenm. Proc. Cal. Ac. (2) 11. 1889, p. 45, and Oce. Pap. Cal. Ac. i. 1890, p. 437. Closely allied to A. barbatus. Length of head about 33 times in the total length. Head a little broader than long. Diameter of eye about 53 times in the length of head, interorbital width about 54 times, length of snout 12 times. Head without keels or ridges. Sides of the head with short bristles in the male, without bristles in the female. Scutes spinulose, not carinate, 24 in a longitudinal series, 7 between dorsal and adipose fin, 10 between anal and caudal. Supraoccipital bordered posteriorly by a pair of scutes. A keel in front of the spine of the adipose fin, formed by 3 or 4 scutes. Lower surface of head and abdomen naked. D. I 9, the first ray longer than the head, the last 3 as long and connected by a membrane to the two scutes succeeding it. A. I 5, Pectoral spine extending beyond base of ventral. Caudal truncate. Caudal peduncle about 24 times as long as deep. Light brown, without spots. Specimens have been recorded up to 270 mm, South-eastern Brazil. 31. fANCISTRUS PARAHYBE. Delturus parahyse Kigenm. & Kigenm, Proc. Cal. Ac. (2) ii, 1889, p. 45, and Oce. Pap. Cal. Ac. 1. 1890, p. 438. Differs from the preceding species in having I 10 dorsal rays, 25-27 scutes in a longitudinal series, and all the fins with large round blackish spots. Rio Parahyba. 242, MR. C. TATE REGAN ON THE 4, PANAQuE. Cochliodon (non Cochlodon D’Orbigny) Kner, Denkschr. Ak. Wien, vil. 1854, p. 265; Higenm. & Eigenm. Proc. Cal. Ac. (2) ii. 1889, p. 44, and Occ. Pap. Cal. Ac. i. 1890, p. 424, Panaque Eigenm. & Eigenm. }. c. and t. c. p. 420. This genus is distinguished from Ancistrus by the teeth, which are. few in number and spoon-shaped. D.17. A. 14. Magdalena, Orinoco, Amazon, and La Plata Systems. Key to the Species. I. Interoperculum not notably spinmate. . . . . . . . . 1. P. cochliodon Kner. IL. Interoperculum armed with some slender ev salle spines. A. Supraoccipital nearly flat te P. dentex Gthr. B. Supraoccipital arched, elevated. Diameter of eye 9 times in the length of head, interorbital width twice (im specimen of 180mm.) . . . . Sie . 3. P. gibbosus Star. Diameter of eye 7} times in the length of eal interor pital width 1% times (in specimen of 210 mm.) ee P. nigrolineatus Ptrs. 1. PanaQuE COCHLIODON. Hypostomus cochliodon Kner, Denkschr. Ak. Wien, vil. 1854, p. 265, pl. ni. fig. 1. Plecostomus cochliodon Giinth. Cat. Fish. v. p. 238 (1864). Cochliodon cochliodon Higenm. & EHigenm. Proc. Cal. Ac. (2) 11. 1889, p. 44, and Occ. Pap. Cal. Ac. i. 1890, p. 425. Depth of body 4 times in the total length, length of head 3-34 times. Head 1§ times as long as broad and 1% times as pane as deep. Diameter of eye 64-6 times in the length of head, interorbital width 13-25 times, length of snout 13-1 2 times, Both jaws with 8 or 9 teeth on each side. Spo narrowed anteriorly, supraorbital edges moderately raised ; supraoccipital with median ridge, ending posteriorly in a pointed process; temporal plates obscurely carinate ; interoperculum scarcely more spinate than the other plates of the head. Scutes spinulose, the anterior carinate, 27 in a longitudinal series, 8 between dorsal and adipose fin, 14-15 between anal and caudal. Supra- occipital entirely bordered posteriorly by a single scute. Lower surface of head and abdomen completely covered with small granular scales. D.17; length of base of dorsal equal to its distance from the middle of the spine of adipose fin. A. I 4. Pectoral spine extending to anterior { of ventral. Caudal emarginate. Caudal peduncle 33 times as long as deep. Olivaceous, with some dark spots on head, body, and fins. Total length 200 mm. = 3 : Iwo specimens: Paraguay; R. Jungada, Matto Grosso; Rio Cujaba. FISHES OF THE FAMILY LORICARITD.®. 24: Qe 2. PANAQUE DENTEX. Chetostomus dentex Giinth. Proc. Zool. Soc. 1868, p. 233. Panaque dentex Eigeum. & Eigenm. Proce. Cal. Ac. (2) 11. 1889, p. 44, and Oce. Pap. Cal. Ac. 1. 1890, p. 427. Depth of body 5 times in the total length, length of head 3 times. Head 1} times as long as broad and 1% times as long as deep. Diameter of eye 5 times in the length of head, interorbital width 22 times, length of snout 12 times. Both jaws with 4 or 5 teeth on each side. Snout narrowed anteriorly, supraorbital edges moderately raised ; supraoccipital nearly flat, without median ridge, its posterior edge slightly convex, with scarcely distinct median point; temporal plates not carinate ; interoperculum armed with some slender spines, the longest equal to } the length of head. Scutes spinulose, not distinctly carinate, 26 in a longitudinal series, 6 between dorsal and adipose fin, 12 between anal and caudal. Supraoccipital bordered posteriorly by a median scute and one on each side. Lower surface of head and abdomen (in the young) in great part naked. D.17; length of base of dorsal equal to its distance from the middle of spine of adipose fin. A.I4. Pectoral spine extending to anterior + of ventral. Caudal emarginate. Caudal peduncle 22 times as long as deep. Olivaceous. Total length 85 mm. One specimen: Xeberos, Upper Amazon. 3. PANAQUE GIBBOSUS. Chetostomus cochliodon (sive gibbosus) Steind. Sitzb. Ak. Wien, Ixxx. 1879, p. 187; Denkschr. Ak. Wien, xlii. 1880, p. 63, pl. iv. fig. 1. Panaque cochliodon Higenm. & Eigenm. Proce. Cal. Ac. (2) it. 1889, p. 44, and Oce. Pap. Cal. Ac. i. 1890, p. 427. Depth of body about 4% times in the total length, length of head about 22 times. Head about 1} times as long as broad and 13 times as long as deep. Diameter of eye 9 times in the length of head, interorbital width twice, length of snout 12 times. Both jaws with about 8 teeth on each side. Snout somewhat narrowed anteriorly ; supra- occipital elevated and arched, with a nearly straight transverse posterior edge; inter- operculum armed with some slender spines, the longest equal to + the length of head. Scutes spinulose, not distinctly carinate, 26 in a longitudinal series, 6 between dorsal and adipose fin, 12 between anal and caudal; supraoccipital entirely bordered posteriorly by a single scute ; lower surface of head and abdomen completely covered with small granular scales. D.17; length of base of dorsal nearly equal to its distance from tip of spine of adipose fin. A. I 4. Pectoral spine extending to middle of ventral. Caudal strongly emarginate. Caudal peduncle 3 times as long as deep. Uniform greyish. Total length 180 mm. One specimen: Colombia (R. Cauca). 244 MR. C. TATE REGAN ON THE 4, PANAQUE NIGROLINEATUS. Chetostomus nigrolineatus Peters, Monatsh. Ak. Berl. 1877, p. 471; Steind. Denkschr. Ak. Wien, xliv. 1881, p. 7. Panaque nigrolineatus Figenm. & Eigenm. Proc. Cal. Ac. (2) 11, 1889, p. 44, and Oce. Pap. Cal. Ac. 1. 1890, p. 426. Depth of body 33 times in the total length, length of head 24 times. Head 1} times as long as broad and 1} times as long as deep. Diameter of eye 73 times in the length of head, interorbital width 14 times, length of snout 14 times. Both jaws with about 5 teeth on each side. Snout narrowed anteriorly ; supraoccipital elevated and arched, with a nearly straight transverse posterior edge ; interoperculum armed with some slender spines, the longest equal in length to the diameter of eye. Scutes spinulose, those of the caudal peduncle carinate, 25 in a longitudinal series, 7 between dorsal and adipose fin, 12 between anal and caudal; supraoccipital entirely bordered posteriorly by a single scute ; lower surface of head and abdomen completely covered with small granular scales. D.I 7; length of base of dorsal nearly equal~to its distance from caudal. A. 14. Pectoral spine extending to middle of ventral. Caudal obliquely truncate. Caudal peduncle twice as long as deep. Head, body, and fins with undulating longi- tudinal dark brown stripes, which are about as broad as the light interspaces separating them. Total length 210 mm. One specimen: Goyaz; Venezuela. 5. CiaTOsTOMUS. Hypostomus (part.) Cuy. & Val. Hist. Nat. Poiss. xv. p. 489 (1840). Chetostoma Tschudi, Fauna Peruana, Pise. p. 25 (1846). Chetostomus Kner, Denkschr. Ak. Wien, vii. 1854, p. 256 ; Bleek. Ned. Tijd. Dierk. i. 1865, p.78 ; Ginth. Cat, Fish. v. p. 240 (1864) (part.) ; Eigenm. & Eigenm. Occ. Pap. Cal. Ac. i. 1890, p. 441 (part.). This genus is distinguished from Ancistrus by the snout, which has a swollen naked margin. D. I 7-10. 0 bp 6 0 010 0 oo 8 eo 6b lo We Ze Uoragielis Shacks, 8. Ventrals extending to the anal opening, barbel 2-2 of the distance from its base to the gill- opening, pectoral spine to middle of ventral . . 13. A. sabaloC. & V. y. Ventrals extending 3 of the distance from their base to the anal opening, barbel to the gill- opening, pectoral spine to posterior } of ventral . 14. A. taczanowskii Bler. 2. Ventrals originating below the fifth dorsal ray . . . 15. A. prenadillaC. & V. B. Outer series of teeth of the preemaxillaries all acutely bicuspid; nasal flap not produced as a barbel ; adipose fin moderately developed, without trace ofa spine . . 16. A. fissidens Ren. C. Outer series of teeth of the preemaxillaries all bicuspid, or the lateral teeth unicuspid. Nasal flap produced as a lems 6 6 6 8 6 oe 58 6 6 6 66 6 6 6 Ml, AL jaye Bile. D. Outer teeth of both jaws compressed, expanded, some- times more or less distinctly bicuspid, usually distinctly incisor-like. Interocular width less than the distance from eye to posterior nostril; barbel not nearly reaching the gill-opening . . . . 18. A. perwanus Stdr. Interocular width equal to the distance from eye to posterior nostril; barbel extending nearly to the gill-opening . . . . 19. A. simonsi Ren. 1. ARGES HoMODON, sp. n. (Plate XXI. fig. 1.) Length of head 53 times in the total length. Interocular width a little less than the distance from eye to posterior nostril, 33 times in the length of head. Outer series of teeth of the premaxillaries all bicuspid, more or less Y-shaped ; mandibulary teeth bicuspid; barbel extending beyond the gill-opening. First dorsal ray equal to 2 the length of head; pectoral spine extending to middle of ventral; ventrals originating well in advance of the dorsal, extending # of the distance from their base to the anal opening. Spine of the adipose fin well developed, freely movable, inserted at a distance 510 MR. C. TATE REGAN ON THE from the caudal equal to 4 the length of the middle rays of that fin; adipose fin represented by a membrane connecting the spine posteriorly to the caudal peduncle. Distance from tip of snout to base of first dorsal ray 23 times in the total length, from base of last anal ray to caudal 63 times. Greyish; caudal with scme dark spots. Total length 52 mm. A single specimen from Villeta, Colombia, 5500 feet ; collected by Mr. Kay Thomson. 2. ARGES BOULENGERI, sp. n. Stygogenes humboldtii (non Swainson) Bouleng. Proc. Zool. Soc. 1887, p. 276, pl. xxi. fig. 2. Length of head about 33 times in the total length. Interocular width much less than the distance from eye to posterior nostril, about 5 times in the length of head. Lateral teeth of the outer series of the preemaxillaries unicuspid, median teeth bicuspid ; mandibulary teeth bicuspid; barbel extending to gill-opening. First dorsal ray nearly equal to the length of head; pectoral spine extending to posterior $ of ventral ; ventrals originating well in advance of the dorsal, extending 2 of the distance from their base to the anal opening; spine of adipose fin well developed, freely movable, inserted at a distance from the caudal equal to 2 the length of the middle rays of that fin. Adipose fin represented by a membrane connecting the spine posteriorly to the caudal peduncle. Distance from the tip of snout to the base of first dorsal ray 2192 times in the total length, from the base of last anal ray to the caudal 53 times. Brownish ; dorsal with some dark spots on the rays; caudal with dark spots forming 3-4 cross-bars. Total length 70 mm. Three specimens from Canelos, Ecuador, collected by Mr. C. Buckley. 3. ARGES GUENTHERI. (Plate XXI. fig. 2.) Stygoyenes guenthert Bouleng. Ann. Mag. Nat. Hist. (5) xix. 1887, p. 348. Cyclopium guentheri Higenm. & Kigenm. Occ. Pap. Cal. Ac. i. 1890, p. 350. Length of head 53-33 times in the total length. Interocular width 33-52 times in the length of head and nearly equal to the distance from eye to posterior nostril. Outer series of teeth of the premaxillaries more or less club-shaped, with a pointed lateral cusp, which is smaller and nearer the base in the lateral teeth than in the median ; mandibulary teeth bicuspid; barbel extending #3 of the distance from its base to the gill-opening. First dorsal ray equal to 3 the length of head; pectoral spine extending to middle of ventral; ventrals originating below the first dorsal ray, extending to or beyond the anal opening; spine of adipose fin well developed, freely movable, inserted at a distance from the caudal equal to $-3 the length of the middle rays of that fin. Adipose fin represented by a membrane connecting the spine posteriorly to the caudal peduncle. Distance from tip of snout to base of first dorsal FISHES OF THE FAMILY LORICARIID®. oll 9 2 times in the total length, from base of last anal ray to the caudal 53-64 times. Body spotted or marbled with blackish; dorsal and caudal with some ray 22-2 dark spots on the rays. Total length 85 mm. Kight specimens from Colombia. 4, ARGES CYCLOPUM. Pimelodus cyclopum Humboldt, Obs. Zool. i. p. 21, pl. vi. (1805). Cyclopium humboldtii Swainson, Nat. Hist. Fish. 11. p. 305 (1839). Arges cyclopum Cuv. & Val. Hist. Nat. Poiss. xv. p. 340 (1840). Stygogenes humboldtit Giinth. Cat. Fish. v. p. 223 (1864). Stygogenes cyclopum Giinth. t. ec. p. 224. Length of head nearly 4 times in the total length. Interocular width equal to the distance from eye to posterior nostril, 5} times in the length of head. ‘Teeth of the outer series of the premaxillaries mostly unicuspid; mandibulary teeth bicuspid ; barbel extending to gill-opening. First dorsal ray equal to > the length of head; pectoral spine extending to middle of ventral; ventrals originating below the first dorsal ray, extending to the anal opening. Spine of adipose fin fairly well developed, freely movable, inserted at a distance from the caudal equal to } the length of the middle rays of that fin. Adipose fin represented by a membrane connecting the spine posteriorly to the caudal peduncle. Distance from tip of snout to base of first dorsal ray 24-23 times in the total length, from base of last anal ray to caudal 4% times. Body spotted or marbled with dark brown; caudal dark at the base and in its posterior half. Total length 45 mm. Four specimens from Madame Ida Pfeiffer’s collection, locality unknown, agree fairly well with Humboldt’s figure and may be provisionally referred to this species, which certainly belongs to the section with short adipose fin and movable spine. A. cyclopum was originally recorded from the Andes above Quito, Ecuador. 5. ARGES WHYMPERI. Arges whymperi (part.) Bouleng. Proc. Zool. Soc. 1890, p. 451, pl. xli. fig. 2. Length of head 42 times in the total length. Interocular width nearly equal to the distance from eye to posterior nostril, 33 times in the length of head. Teeth of the outer series of the premaxillaries mostly unicuspid; mandibulary teeth bicuspid; barbel extending #4 of the distance from its base to the gill-opening. First dorsal ray equal to $ the length of head; pectoral spine extending to anterior 1 of ventral ; ventrals originating nearly below the first dorsal ray, extending 3 of the distance from their base to the anal opening, which is situated at 7 of the distance from the base of VOL. XVII.—PaART 1. No. 16.—October, 1904. 20 312 MR. C. TATE REGAN ON THE ventral to the origin of anal. Adipose fin long, very weakly developed, with a small imbedded spine. Distance from tip of snout to base of first dorsal ray 24 times in the total length, from base of last anal ray to caudal 5 times. Lrownish, closely spotted with darker. Total length 84 mm. A single specimen from Milligalli, Andes of Ecuador. 6. ARGES EIGENMANNI, sp.n. (Plate XXI. fig. 3.) Cyclopium cyclopum (non Humb.) Putnam, Am. Nat. 1871, p. 694; Eigenm. & Eigenm. Proc. Cal. Ac. (2) 1. 1888, p. 164, and Occ. Pap. Cal. Ac. i. 1890, p. 350. Length of head 4} times in the total length. Interocular width nearly equal to the distance from eye to posterior nostril, 3-33 times in the length of head. Outer series of teeth of the premaxillaries mostly unicuspid, mandibulary teeth bicuspid; barbel extending to 3-3 the distance from its base to the gill-opening. First dorsal ray equal to 2 the length of head; pectoral spine extending to anterior } of ventral; ventrals originating below, or nearly below, the first dorsal ray, extending #—{ of the distance from their base to the anal opening, which is situated at 7 of the distance from the base of ventral to the origin of anal. Adipose fin long, very weakly developed, with a small spine which is completely imbedded in the adult. Distance from tip of snout to base of first dorsal ray 23-23 times in the total length, from base of last anal ray to caudal 42_5 times. Greyish, sometimes spotted with darker; usually a series of whitish spots on the pores of the lateral line. Total length 100 mm. Several specimens from Cayambe and Machachi, Ecuador, collected by Mr. E. Whymper. , 7. ARGES VAILLANTI, sp. n. Length of head 4-41 times in the total length. Interocular width nearly equal to the distance from eye to posterior nostril, 33 times in the length of head. Outer series of teeth of the pramaxillaries mostly unicuspid; mandibulary teeth bicuspid ; barbel extending to, or nearly to, the gill-opening. First dorsal ray equal to = the length of head; pectoral spine extending to anterior + of ventral; ventrals originating below, or nearly below, the first dorsal ray, extending $—4 of the distance from their base to the anal opening, which is situated at 8 of the distance from base of ventral to origin of anal. Adipose fin weakly developed, with a small spine which is completely imbedded in the adult. Distance from tip of snout to base of first dorsal ray 23-2} times in the total length, from base of last anal ray to caudal 6 times. Brownish; a large dark blotch or bar at the base of the caudal; rays of all the fins with obscure dark spots. Total length 95 mm. FISHES OF THE FAMILY LORICARIID®. 313 Three examples from the Huamani Peak, Ecuador, 10,700 feet, received from the Paris Museum as Brontes prenadilla. 8. ARGES ORIENTALIS. (Plate XXI. fig. 4.) Arges orientalis, Bouleng. Ann. Mag. Nat. Hist. (7) xi. 1903, p. 601. Length of head 53-4 times in the total length. Interocular width less than the distance from eye to posterior nostril, 53—4 times in the length of head. Outer series of teeth of the premaxillaries unicuspid; mandibulary teeth bicuspid ; barbel extending 3-3 of the distance from its base to the gill-opening. First dorsal ray equal to 3—3 the length of head; pectoral spine not extending beyond base of ventral; ventrals originating below the first dorsal ray, extending about 3 of the distance from their base to the anal opening, which is situated at 3-7 of the distance from the ventral to origin of anal. Adipose fin, if present, very feebly developed, without or with a rudimentary imbedded spine. Distance from the tip of snout to the base of first dorsal ray 24 times in the total length, from the base of last anal ray to the caudal 6-63 times. Olive or yellowish above, spotted or marbled with dark brown; caudal with dark spots. Total length $0 mm. Ten specimens from the Albireggas and Milla Rivers, above Merida, Venezuela. 9. ARGES BRACHYCEPHALUS. (Plate XXI. fig. 8.) Arges brachycephalus Giinth. Proc. Zool. Soc. 1859, p. 92, and Cat. Fish. v. p. 222 (1864). Length of head 4-4} times in the total length. Interocular width equal to or a little less than the distance from eye to posterior nostril, 33-3 times in the length of head. Outer series of teeth of the preemaxillaries unicuspid; mandibulary teeth bicuspid ; barbel extending to beyond the gill-opening in the young, #—# of the distance from its base to the gill-opening in the adult. First dorsal ray equal to 3-4 the length of head; pectoral spine extending to anterior 3 of ventral; ventrals originating below the first dorsal ray, extending to the anal opening, which is situated at 2 of the distance from base of ventral to origin of anal; adipose fin long and low, not well developed, without trace of a spine. Distance from tip of snout to base of first dorsal ray 2% (young) to 2% (adult) times in the total length, from base of last anal ray to the caudal 44-44 times. Brownish, spotted or marbled with darker; some dark spots on the basal parts of the dorsal and caudal. Total length 218 mm. Ten specimens, including the types of the species from the Andes of Peru, and others from Cuenca, Ecuador. 10. ArGEs cHoT#, sp. n. (Plate XXI. fig. 5.) Length of head 4} times in the total length. Interocular width a little greater than the distance from eye to posterior nostril, 3} times in the length of head. Outer ZT 2 314 MR. C. TATE REGAN ON THE series of teeth of the premaxillaries unicuspid; mandibulary teeth bicuspid; barbel extending beyond the gill-opening. Longest branched ray of pectoral somewhat more than 2 the length of the spine, which extends a little beyond the middle of the ventral ; ventrals originating nearly below the first dorsal ray, extending a little beyond the anal opening; distance from the base of the ventrals to the anal opening twice that from the latter to the base of the first anal ray. Adipose fin well developed, originating at a distance from the dorsal nearly equal to the length of the base of that fin, extending on to the caudal, with a very small nodule, representing the spine, imbedded in the upper half of its posterior part. Distance from tip of snout to base of first dorsal ray 25 times in the total length, from base of last anal ray to the caudal 63 times. Greyish, indistinctly marbled with darker; some dark spots on the caudal. Total length 92 mm. A single specimen from the Chota Valley, Northern Ecuador, collected by Dr. H. Festa. J1. ARGES MARMORATUS, sp. n. (Plate X XI. fig. 6.) Length of head 4 times in the total length. Interocular width greater than the distance from eye to posterior nostril, 3 times in the length of head. Outer series of teeth of the preemaxillaries unicuspid; mandibulary teeth bicuspid; barbel extending beyond the gill-opening. First dorsal ray as long as the head; longest branched ray of pectoral equal to 2 the length of the spine, which extends nearly to the posterior end of the ventral ; ventrals originating nearly below the first dorsal ray, extending to a little beyond the anal opening, which is situated midway between the base of the ventrals and that of the first anal ray. Adipose fin wel] developed, extending from just behind the dorsal on to the caudal, with a nodule, representing the spine, imbedded in the upper half of its posterior part. Distance from tip of snout to base of first dorsal ray 24 times in the total length, from base of last anal ray to the caudal 6 times. Brownish, marbled with blackish; caudal with some dark spots. Total length 73 mm. ‘Two specimens, from St. Augustin, Andes of Colombia, 5000 ft., collected by Capt. Dowding. 12. fARGES LONGIFILIS. Se Arges longifilis Steind. Denkschr. Ak. Wien, xlvi. 1882, p. 19, pl. v. fig. 3. Length of head 33—4 times in the total Jength. Interocular width a little less than the distance from eye to posterior nostril, 34 times in the length of head. Outer series of teeth of the premaxillaries unicuspid ; mandibulary teeth bicuspid ; barbel extending to, or nearly to, the gill-opening. First dorsal ray equal to, or a little more than the length of head; pectoral spine extending to the extremity of ventral; ventrals FISHES OF THE FAMILY LORICARIIDA. 315 originating below the first dorsal ray, extending to the anal opening ; adipose fin long, well developed, without trace of spine. Distance from tip of snout to base of first dorsal ray about 22 times in the total length, from base of last anal ray to the caudal about 43 times. Body spotted or marbled with dark violet. The typical examples described and figured by Dr. Steindachner measure up to 180 mm. in total length. Rio Huambo and Rio Totora, Peru. 13. ARGES SABALO. Arges sabalo Cuv. & Val. Hist. Nat. Poiss. xv. p. 335, pl. 444 (1840) ; Giinth. Cat. Fish. v. p 222 (1864) ; Steind. Sitzb. Ak. Wien, Ixxii. 1876, p. 598 (part.), and Denkschr. Ak. Wien, xlvi. 1882, p. 17, pl. iv. fig. 2. Length of head 54 times in the totallength. Interocular width less than the distance from eye to posterior nostril, 42-42 times in the length of head. Outer series of teeth of the premaxillaries unicuspid ; mandibulary teeth bicuspid; barbel extending 3—3 of the distance from its base to the gill-opening. First dorsal ray equal to ? the length _ of head; pectoral spine extending to middle of ventral: ventrals originating below the first ray of dorsal, extending to the anal opening; adipose fin long, well developed, without trace of spine. Distance from tip of snout to base of first dorsal ray 23-22 times in the total length, from base of last ray of anal to the caudal 5 times. Body marbled with dark brown; fins with dark spots. The typical example described and figured by Dr. Steindachner is 200 mm. in total length. I have examined one small specimen, 55 mm. in total length, from Rozmain, Upper Peru, which differs only in the somewhat longer barbel. 14. ARGES TACZANOWSKII. Arges sabalo (part.) Steind. Sitzb. Ak. Wien, Ixxii. 1876, p. 598. Arges taczanowski Bouleng. Proc. Zool. Soc. 1890, p. 451, pl. xh. fig. 1. Length of head 4 times in the total length. Interocular width much less than the distance from eye to posterior nostril, 44 times in the length of head. Outer series of teeth of the premaxillaries unicuspid ; mandibulary teeth bicuspid; barbel extending to the gill-opening. First dorsal ray equal to # the length of head; pectoral spine extending to posterior + of ventral; ventrals originating nearly below the first dorsal ray, extending ‘a little more than 2 of the distance from their base to the anal opening. Adipose fin long, well developed, without trace of spine. Distance from tip of snout to base of first dorsal ray 22 times in the total length, from base of last anal ray to the caudal 5 times. Yellowish, marbled with purplish brown. Total length 84 mm. One specimen from Amable Maria, Peruvian Andes. 316 MR. C. TATE REGAN ON THE 15. fARGES PRENADILLA. Brontes prenadilla Cuv. & Val. Hist. Nat. Poiss. xv. p. 343, pl. 445 (1840) ; Ginth. Cat. Fish. v. p. 224 (1864). Arges prenadilla Steind. Denkschr. Ak. Wien, xlvi. 1882, p. 20, pl. vi. fig. 5. Length of head 4 times in the total length. Interocular width equal to the distance from eye to posterior nostril, 32 times in the length of head. Outer series of teeth of the premaxillaries unicuspid ; mandibulary teeth bicuspid; barbel not extending to the gill-opening. First dorsal ray equal to }-$ the length of the head; pectoral spine extending nearly to base of ventral; ventrals originating below the fifth dorsal ray, extending beyond the anal opening; adipose fin long, low, without trace of spine. Distance from tip of snout to base of first dorsal ray about 23 times in the total length, from base of last anal ray to the caudal about 43 times. The typical example described and figured by Dr. Steindachner is 75 mm. in total length. Andes of Ecuador. 16. ARGES FISSIDENS, sp. n. Arges whymperi (part.) Bouleng. Proc. Zool. Soc. 1890, p. 451. Length of head 5 times in the total length. Interocular width equal to the distance from eye to posterior nostril, 33 times in the length of head. Teeth of the outer series of the preemaxillaries all bicuspid, more or less Y-shaped ; mandibulary teeth bicuspid ; barbel extending a little beyond the gill-opening. First dorsal ray nearly equal to the length of head; pectoral spine extending to anterior 3 of ventral; ventrals originating nearly below the first dorsal ray, extending 3—} of the distance from their base to the anal opening ; adipose fin long, low, without trace of a spine. Distance from tip of snout to base of first dorsal ray 34 times in the total length, from base of last anal ray to caudal 54 times. Uniform brownish. Total length 89 mm. Two specimens from the Andes of Ecuador, collected by Mr. Edward Whymper. 17. Arges FrstH#. (Plate XXI. fig. 7.) Arges feste Bouleng. Bull. Mus. Torino, xiii. 1898, no. 329, p. 9. Length of head 41-42 times in the total length. Interocular width equal to the distance from eye to posterior nostril, 3} times in the length of head. Nasal flap produced as a barbel. Lateral teeth of the outer series of the premaxillaries unicuspid (rarely bicuspid), median teeth bicuspid; mandibulary teeth bicuspid ; barbel extending beyond the gill-opening. First dorsal ray nearly as long as the head; pectoral spine extending to middle of ventral or beyond; ventrals originating below, or somewhat in advance of the first dorsal ray, extending ? of the distance FISHES OF THE FAMILY LORICARIIDA. 317 from their base to the anal opening, or sometimes nearly to the anal opening. Adipose fin long, moderately developed, without trace of a spine. Distance from tip of snout to base of first dorsal ray 22 times in the total length, from base of last anal ray to caudal 63-7 times. Greyish; caudal with some dark spots. Total length 88 mm. Ten specimens from Eastern Ecuador. 18. ARGES PERUANUS. Arges peruanus Steind. Sitzb. Ak. Wien, Ixxii. pt. i. 1875, p. 601, pl. ix. figs. 3-6, and Denkschr. Ak. Wien, xli. 1879, p. 173. Length of head 84-4 times in the total length. Interocular width less than the distance from eye to posterior nostril, 33-33 times in the length of head. Outer series of teeth of the prxmaxillaries compressed incisors, with the upper edge slightly emarginate, notched; mandibulary teeth similar; barbel extending $ the distance from its base to the gill-opening. First dorsal ray equal to 3 the length of head ; pectoral spine extending to anterior + of ventral ; ventrals originating below or slightly in advance of the first dorsal ray, extending 3-2 of the distance from their base to the anal opening; adipose fin long, low, feebly developed, without trace of spine. Distance from tip of snout to base of first dorsal ray 25 times in the total length, from the base of last ray of anal to caudal nearly 5 times. Uniform brownish ; fins without spots. .. Total length 63 mm. Four specimens from Amable Maria, Peru. 19. ARGES SIMONSH, sp. n. (Plate XXI. fig. 9.) Length of head 4-43 times in the total length. Interocular width equal to the distance from eye to posterior nostril, 31-3 times in the length of head. Outer series of teeth of the premaxillaries incisors, with straight or notched edges; mandibulary teeth similar; barbel extending } + of the distance from its base to the gill-opening. First dorsal ray equal to 3-8 the length of head; pectoral spine extending to anterior 3 of ventral; ventrals originating below the first dorsal ray, extending 3-3 of the distance from their base to the anal opening; adipose fin long, low, moderately or well developed, without trace of spine. Distance from tip of snout to base of first dorsal ray 22-2 times in the total length, from base of last anal ray to caudal 03-6 times. Body sometimes spotted or marbled with dark brown; some dark spots on the dorsal. Total length 80 mm. Five specimens from Huaras, Peru, 10,700 feet, collected by the late Mr. P. O. Simons. 318 FISHES OF THE FAMILY LORICARIIDA. 17. ASTROBLEPUS. Astroblepus Wumboldt, Obs. Zool. i. p. 19 (1805). This genus, represented by a single species, is known only from Humboldt’s description and figure. It is allied to Arges, differing in the absence of ventral fins. It is possible that this latter feature is abnormal or accidental, and that the genus may be founded on a specimen of Arges brachycephalus or an allied species. ‘ 7 ASTROBLEPUS GRIXALV 4, Astroblepus grixavlii Humb. Obs. Zool. i. p. 19, pl. vil. (1805) ; Cuv. & Val. Hist. Nat. Poiss. xv. p- 349 (1840) ; Giinth. Cat. Hish. v. p. 224 (1864) ; Eigenm. & Eigenm.-Occ. Pap. Cal. Ac. i. 1890, p. 351. In form and proportions, in the position of the fins and in the number of their rays essentially similar to Arges brachycephalus Gthr. Dorsal with I 6 rays, anal with I 6, pectoral with I. 9. Coloration uniform. Total length 550 mm. Popoyan, R. Cauca; known to the inhabitants as the pescado negro, and eaten by them. _] have compared a specimen of Arges brachycephalus of 218 mm. with Humboldt’s figure and cannot see any differences except the absence of ventral fins. INDEX TO GENERA AND SPECIES. Page FAGEVUNEE G060000000000000 261 TM BAS coon h00660000 219 AY SHIR. io 6 oye eseyete aime rovers 262 acanthicus, Rhinelepis........ 262 ACTIN cooponoabosno0seoco0n 302 EYE) ocopooanoocoeccKde 304 ——— EREVAN, Gogaog000000 305 —— gladiolus ..,........0. 305 —— AEGIS ooc0o0g000000000 303 = IHR Gog 0000090000000 304 oxyrhymchusi ley. .trvjeieler: 304 acipenserina, Loricaria........ 296 acipenserinus, Hemiodon .... 296 —, Hemiodontichthys...... 296 aculeatus, Cheetostomus ...... 230 ACUS PACESLLAielarereycistelovsisiaret ts 304 , Farlowella ...... Rodan 304 POM, ICCA, Sondonuendce 287 affinis, Otocinclus............ 267 » IAGCOMOMUS Gogooooone 207 agassizli, Rhinelepis.......... 219 alatus, Hypostomus.......... 211 pPlecostomus) enya 211 alga, Cheetostomus .......... 255 alternans, Ancistrus.......... 229 altipinnis, Liposarcus ........ 227 amazonica, Loricaria ........ 286 amazonum, Acestra .......... 305 > elOnGIE soocooceoooo 305 INSIGHTS oodcoospoasone 221, 251 —alternans .............. 229 ANS UMCAIG Deereieyell--)-leie 241 —— annectens.............. 225 aspidolepis ............ 23 aurantiacus ............ 236 > |WIN CowoooobeO sooo OBy 233 <—— INS Gaboonaeeanenn 240 —— hrachyurus ............ 231 GAEKIND, poogpadbeoonec 258 ——chagresi .............. 256 =———CITTHOBUBseetete nies) lees LOO GETING aoocosectbodoos 240 dolichopterus .......... 254 —— duodecimalis .......... 229 D3 Ancistrus etentaculatus ...... 226 FAITE. caocoocacegden 227 IGS so00dgcc0eo000G000 230 guacharote ............ 238 FUCA 55 950000000000 241 — gymnorhynchus ........ 254 heteracanthus .......... 237 hoplogenysisemeaseicriae: 255 leucostictus ............ 255 INOERS ~Go000n00000000 227 lomeateMS Gooosaoenoos 226 ——— MeEdIaNS 5.05 aoe sis cielo 229 —— megacephalus .......... 23: —— multiradiatus .......... 228 —— mystacinus ............ 2388 TUNG Jo angoa00o0H000 235 oligospilushee tere ace 232 —— parahybe .............. 241 PICTUSH ere eet 231, 237 —— platycephalus .......... 235 —— punctatissimus ........ 236 —— JUNIE) oocccopdo0000 229 secaphirhynchus ........ 231 — schomburghii .......... 233 EAWOME) cogocagsoosce0ND 239 Stigmaticus) jj jyree re 259 —temminckii ............ 254 —— undecimalis ............ 226 VOUS bo oocbodDDO0D00 232 angulicauda, Ancistrus........ 241 DAWN Soocccooqd00ac 241 ElecostomuUsie eyelets 241 anisiti, Pterygoplichthys ..., 221 année, Plecostomus .......... 210 annectens, Ancistrus ........ 225 anomalus, Cheetostomus ...... 250 NUS wMOLICATIAI er eylelteeteretee: 289 apeltogaster, Loricaria........ 292 INTHE cog ood 0d 00000000000 307 bouleng erie eityeriledls 310 brachycephalus ........ 313 NOE) ooodooou0D00K008 313 ——cyclopum.............. 311 CIQONMANNIU rele pepelstetetolel sts 312 VOL. xvil.— Part 11. No. 17.—October, 1904, Page INEESIGIED od occpcoaene000K 316 —— fissidens .............. 316 FUMIO, 5 oo000c0Ga0006 310 homod on aeee ences 309 Horeeatillis) “oo o.000K0000000 314 —— marmoratus ............ 314 Orientalisy jaja riers 313 WERENT: 5 Go50060000000 317 prenadillapeere aes 316 sabalomanan cerca 315 WONT. so conoooon0d000 317 ——— taezanowskil) “ys sleaseecele 315 VENER oscogeoooa000s 312 qwihiy Perle ./etepelcleta 311, 316 asper, Plecostomus .......... 220 aspera, Rhinelepis .......... 220 asperatus, Hypostomus ...... 217 aspidolepis, Ancistrus ........ 230 , Cheetostomus .......... 230 ,» Hemiancistrus ........ 230 aspilogaster, Plecostomus .... 206 ATTOVGWES ooococcocdcddoad 318 GELENNED! cooc0so 0000008 318 atropinnis, Plecostomus ...... 215 aurantiacus, Ancistrus........ 236 , Chetostomus .......... 236 —, Hypostomus .......... 236 pLarencistlusmmaeretatrets 236 SWE, Ie “Soooccocca0 301 auroguttatus, Hypostomus ..,. 217 , Plecostomus .......... 217 bachi,yAmncistrus) y.ii-riieielelele 233 ——., Chetostomus .......... 233 barbata, Loricaria .......... 301 » Oralosieriet) Googooso0n 301 barbatus, Ancistrus .......... 240 , Hypostomus ........ -. 240 ——, Plecostomus .......... 240 , Pseudancistrus ,....... 240 bicirrhosus, Plecostomus ,, 205, 209 bilobatum, Hypoptopoma . 263, 265 biseriatus, Plecostomus ...... 210 borellii, Plecostomus ........ 212 2U boulengeri, Arges......... , Plecostomus bransfordi, Loricaria brasiliensis, Plecostomus ... brevicauda, Plecostomus...... 4 brevipinnis, Xenocara ..... brevirostris, Loricaria ..... , Oxyloricaria brevis, Cheetostomus Brontes brunnea, Loricaria bufonia, Xenocara bufonius, Hypostomus ..... cadez, Loricaria calamita, Ancistrus ....... , Chetostomus ....... ——, Hypostomus ....... brachycephalus, Arges..... brachyurus, Ancistrus ..... , Cheetostomus ....... , Hemiancistrus ..... branickii, Chestostomus ... prenadillay vst. cre. eras. FISHES OF THE FAMILY LORICARITDA. see aU see eu 560 2) ... 258 caracasensis, Hemiloricaria .... 270 carinata, Farlowella.......... 305 ——., Loricaria......... Bab ee carinatum, Hypoptopoma .... 264 carinatus, Plecostomus........ 209 castanea, Loricaria ...... 7, 288 catamarcensis, Loricaria ...... 280 cataphracta, Loricaria ........ 291 @heetostomancinaecteutlseraceit 244 Cheetostomus.... 222, 244, 251, 259 = HONINDE) GogouunodedooN 230 see DUR ala ata stoners ste lecereproue 255 EMMIS ooneagcodeson 250 aspidolepigy W.irastetsiaclveiste 230 —— anrantiacus ............ 236 Sa IDACHIS ei cre, s,accadececoe traps 233 —— brachyurus .,........... 231 ranicktlsy ev tieya cee 250 ———=4DYOVIS! corecaiafecnyin icere aceiedoys 247 ——CalaMitA) soci cee ge cece 258 CirrhOSsusmerysrereeier 56. 257 —Sqaambonkyn odoogasaasae 243 den tex Miaernmctesreesas 243 =e pressuls mauslelcyeteteteleresiaie 240 dermorhynchus ..... 248 dolichopterus .......... 254 Ais Cherimersante eevee rere 248 fordilis ae coe e eee esate 261 = FRANCIS Sooindooacconce 227 —— gibbosus ......... 243 Wo asomadopoday sono 230 —— guacharote .......... 238 Page Cheetustomus guairensis ...... 249 gymnorhynchus ........ 254 heteracanthus ........... 237 hoplogenys | Viracjscrsre creer 255 ——— USER ie eielenstcioreteted telat 261 TLEVGOE). “aag sep obin's pig dda 234 JOSS Gorda douodl coon 258 latitrons:! aeteevejersch clei eishe 253 leucostictusiey-tremen tidee 255 ——— Hituratys) We ici ses sels 227 loborhynchus) 4.5.2... 246 = MACTOPS\sueretehelsiolelsie ticles 23 —— maculatus ............ 246 —— malacops ............., 255 —— marcapate ............ 246 —— marginatus ............ 249 —— medians................ 229 —— megacephalus .......... 234 —— microps....... , 247, 249, 251 —— mystacinus ............ 238 —— nigricans .............. 236 —— nigrolineatus .......... 244 ——— MANE “oscooquagdeds 235 ——nudiceps .............. 254 —— nudirostris ............ 251 —— oligospilus ............ 232 ———) ATE Conseocdeduoqons 231 —— platycephalus .......... 235 —— punctatissimus.......... 236 PUD CLACUS) cio) terersinssorsi sls 229 scaphirhynchus ........ 23 SCTICOUS? Beers Heo iee 252 SETTALUS craven feticastertisiel ets 260 SONS Gooaneunnooccdy ao 239 —— spinosus................ 260 StANNIY yore ese 248 taczanowskil............ 247 — tectirostris ............ 255 WON, Guacionooscoe 6 260 BNE) noo oooGonGD ads 293 UNndecimalistyn race 226 VON: “Koncoocesosoue 256 VItLAtUsiteiemitkmemienie 232 chagresi, Ancistrus .......... 256 AGU aoanobecagonee 256 Ones); AWASS 6 oooncoonbudcpes 318 cirrhosa, Loricaria ...... 286, 291 VGN 6 once ogsa00600 256 cirrhosus, Ancistrus .......... 256 ——, Chetostomus ...... 256, 257 ——, Hypostomus .......... 256 Cochhiodonteyercterrricrlerr 242 cochliod onee ape) ariserers 242 cochliodon, Chaetostomus . 243 TC ochliodontnmmrett reer 242 ——, Hypostomus .......... 242 Page cochliodon, Panaque...... 242, 243 , Plecostomus .......... 242 commersonii, Hypostomus .... 206 , Plecostomus ...... 206, 207 cordove, Plecostomus ........ 212 Cyclopiumlacmreyeeeeeer ini: 307 ey clopum ir yareiastyer-tt: 312 guentherlenrcmiriieeer- 310 humbol dingy eevee 3ll cyclopum; Argesy jery-tnietaeiele 311 Oy.clopiumirereriaet 312 ——, Pimelodus ............ dll =——7 Stygogenes\.\cees eee: 3ll Welturis se tereeotustelstale eevee .. 222 angulicauda ........ Nn. . 241 parahy brevets device 241 dentex, Cheetostomus ........ 245 EAN AGUO NAR Tateccm ican 243 depressa, Loricaria .......... 296 depressus, Ancistrus.......... 240 , Chetostomus .......... 240 —, Hemiodon ......... poo es) ——, Hemiodontichthys ...... 296 , Pseudancistrus ........ 240 dermorhynchus, Cheetostomus , 248 dolichoptera, Xenocara ...... 254 dolichopterus, Ancistrus ...... 254 » Cheetostomus .......... 254 duodecimalis, Ancistrus ...... 229 , Hypostomus ......... . 226 , Pterygoplichthys ...... 226 dura; Hortcariaiaqermscsi ceri 291 eigenmanni, Arges, .......:.. 312 emarginatus, Hypostomus .... 210 , Plecostomus ....... aes 2LO erinaceus, Hypostomus .. 251, 252 etentaculatum, Hypostomus 226 , Pterygoplichthys ...... 226 etentaculatus, Ancistrus ...... 226 evansil, Loricaria ............ 290 Harlowwell agveruaci-tterccile rac 302 BCUS Reparers alesse eects nie 304 -——amazonum ........ . 305 CATING LAs eee ee oes nielee ole. 305 eladiolavmnmrrtdr erence 305 GILG 5 ao ponqsedoanodd 303 Sea PTACIIAS i. sicitin ass ober 303 AMOR Shee ene 304 —— oxyrhynchus ........ .. 304 ISHS INES 56 o00000n9 0b 005 516 FIM CMON Avoogonsed 208 filamentosa, Loricaria .... 274, 292 fimbriatus, Otocinclus ........ 267 fischeri, Cheetostomus ........ 248 INDEX TO GENERA AND SPECIES. fissidens, Arges.............. flagellaris, Plecostomus ...... 291 Havas Morcariance cee codes 205 flavus, Plecostomus .......... 205 flexalis;@tocimclus’ <3... .22.- 267 fordii, Chetostomus.......... 261 ——,, Hemiancistrus.......... 261 ——,, Pseudacanthicus........ 261 francisci, Plecostomus........ 211 frenata, Loricaria............ 302 , Oxyloricaria .......... 302 garmani, Plecostomus ........ 214 genibarbis, Acanthicus........ 219 pplecostomusirr-yaetrcierrsr. 219 p JMNMEL HT ooncaoocecaa 219 gibbiceps, Ancistrus.......... 227 , Chetostomus .......... 227 ——,, Pterygoplichthys ...... 227 gibbosus, Chetostomus ...... 248 ) PERS 66 cocdn0000000 243 gigas, Ancistrus ............ 230 , Chetostomus .......... 230, gladiola, Farlowella.......... 305 gladiolus, Acestra............ 305 gladius, Acestra ...........- 305 sy Harlowella; wie sj. ae + « 303 gobio, Hemipsilichthys ...... 221 , Menomystus .......... 221 gracilis, Farlowella .......... 803 granosus, Hypostomus........ 306 , Neoplecostomus...... .. 306 SPEIECOSLOMUSIN terelel-taliiclere 306 erixalve, Astroblepus ........ 318 guacari, Hypostomus ........ 205 guacharote, Ancistrus ........ 288 , Chetostomus .......... 238 ——., Hemiancistrus.......... 28) , Hypostomus .......... 238 guairensis, Cheetostomus...... 249 guentheri, Ancistrus.......... 241 > INES) Ggoooc0ccob00000 310 ——, Cyclopium ............ 310 , Hypoptopoma.......... 264 ——,, Oxyloricaria .......... 299 ; Sty GOKeNes «22.20... 310 culare, Hypoptopoma ........ 265 guttatus, Hypostomus........ 240 , Plecostomus' :......... 241 , Pseudancistrus ........ 240 gymmorhynchus, Ancistrus.... 254 ——, Chetostomus .......... 254 ——, Xenocara.............. 204 [SEMIN sop00000p00000000000 ASU loricariiformis .......... 298 Page Hemiancistrus .......... 221, 259 aspidolepis .::.:....... 23 lbrachyuTuswrreiriersiereler= 25 ——— lbs coooncboasboogsDe 261 UCI soodsoooeade 239 —— heteracanthus .......... 237 ———= ]thHDS GoqgcodscooboGOuE 261 MIACMED: yo daadodudisn ooo 234 —longipinnis ............ 235 — medians ........ pbanod 229 megacephalus .......... 234 ——— Mh ARAGUNES 165 oano00000 238 —— ClligoymIlMS Gooscooccc0s 252 {ROMS soododsacoondeon 231 — scaphirhynchus ........ 231 —— schomburekii ....... (ee 200 SCLTALUSHustepelerstelelelsTeleieysicls 260 ROUGE” Gossccocooacd0 260 (MUM ‘G\ecloacoq0500 .. 239 WRENS. 6G boGoGb000D000 232 EVvemiloricarigeacrieeer ert 270 caracasensis ............ 27 iclemiod one ereeeeeerra .. 296 acipenserinus .......... 296 depressusiencnrrr ric: 296 platycephalus .......... 295 Hemiodontichthys .......... 296 acipenserinus .......... 296 depressustarrieernriiertr 296 Hemipsilichthys .......... ». 220 FOMOscosodsvdcocoodoos 221 heteracanthus, Ancistrus...... 237 ——, Cheetostomus .......... 23 , Hemiancistrus ........ 237 heylandii, Plecostomus ...... 221 lENONONINS Soaccosspccoo0d006 265 —— levior ................ 268 leptochilustiacrmrritletter 268 MUNG oooocccpc0000000 268 homodon, Arges ........ so50 Bhi) hoplogenys, Ancistrus ........ 255 , Chetostomus.......... 255 DREMOEMD 5 5 000000000000 255 horridus, Hypostomus........ 210 ——, Plecostomus ........ .. 210 humboldtii, Cyclopium ...... 311 , Stygogenes ........ 310, 311 Hypoptopoma .............. 268 bilobatum ........ 263, 265 CHM § So50000R0000 264 RAVINE. Hog aeco00000000 264 WER Gooose ooec00d0D~ 265 FOL noosccoccdc00000 265 steindachneri ..... coon Aas) thoracatum ......., 265, 265 Hypostomus, .202, 221, 244,251, 259 Page Hypostomus alatus .......... 211 = OSE ooganc0e0ce bao PALM —— aurantiacus ..:......... 236 —— auroguttatus..........:. 217 —— barbatus .............. 240 —— | KONING “Bosoouocsscoas 258 CHEM, 9 gob 05000000000 258 GHANOSNS’ coaspodacocdos 256 COONHECON socecoocodes 242 commersonil............ 206 durodecimalitsiuesnmrry ie 226 —— emarginatus............ 210 ChINaCeUS Meee 251, 252 etentaculatum ..2:...... 226 @TLANOSUS Wren e 306 UAGME oocoddsasecdscso 205 URNS. gnoc0c000000 238 guttatus ........... ... 240 ————horridusy irae cieieisieee 210 WEED, oddoqosc000%o00s 284 INNO ooodccococgden 236 TUNES) “Sodaccen00c0ss 239 MCHCTOS “Goocacddo0ea0c 254 (MUNIN oascccaocce A0Y) WECM G goaducdodennes 243) ——— HHS) “Gonoodousocddsa0 237 —— plecostomus........ 205, 209 ———= PUNGENT) Gogacacoconned 207 OMA Gos cooocn cos 205, 215 SCXTALUS rrr Rnvir- terete 260 ——spinosuS :.2...-...25-- 260 Squalinumly jae LO. —— subcarinatus............ 207 —— temminckil ............ 254 =————-AVACILLUS ray feu sieelensvereventi ot Tove 236 hystrix, Acanthicus .......... 262 , Chetostomus ........ +. 26] ——, Hemiancistrus.......... 261 ——,, Pseudacanthicus........ 261 IIMMIEOS 5650600000000 261 itacua, Chetostomus ........ 234 ——, Hemiancistrus ........ 234 ——, Hypostomus .......... 254 jeanesianus, Liposarcus ...... 228 , Pterygoplichthys ...... 228 jelskii, Cheetostomus ........ 258 joberti, Hypoptopoma ........ 265 LO tocinclusteenreeinreeieets 265 johnii, Plecostomus .......... 217 Ubataeluoricaridiyrrtrireierrs 278 juvens, Pterygoplichthys...... 221 Knexit,pAcestralenyy.)vsrerlererretace 304 ——, Farlowella ...... Pat 304 20 2 konopickyi, Loricaria labialis, Loricaria ......... leevior, Hisonotus......... leviuscula, Loricaria .... lamina, Loricaria lanceolata, Loricaria....... Lasiancistrus...........-- laticeps, Loricaria latifrons, Cheetostomus , Plecostomus leptochilus, Hisonotus ..... leucostictus, Ancistrus..... , Chetostomus .....-. lima, Loricaria , Plecostomus limosus, Plecostomus Liposarcus.........++0++ altipIMNis! sree: jeanesianus multiradiatus — scrophus varius , Cheetostomus , Pterygoplichthys loborhynchus, Cheetostomus longifilis, Arges lophophanes, Otocinclus , Rhinelepis Woricaria), ayiecciee) sil siedel-'= anus —— cade —— castanea............ — catamarcensis -—— cataphracta Cirrhosasseeee een depressa ~—— dura 294, 205 HPRCNOCAL D1 eielorere = 1-12 latirostris, Loricaria....... —— pardalis...........4. lituratus, Ancistrus ....... longimanus, Ancistrus..... longipinnis, Hemiancistrus ... acipenserina ........ ACULABerlsnoneterlerepecietel> == CANINA LA aye veyelene elekeietsies FISHES OF THE FAMILY LORICARIIDA. 277 224 295 253 eh 208 ... 283 286 289 292 301 301 270 299 sae OF ... 279 ... 292 287, 288 280 56 Pk 236, 291 296 291 290 Page Loricaria flava .........+8+6 205 EVONALA seroyervelchessiovelsesheher=t= 302 —— jubata ......... Perinat 278 —— konopickyi ...... ditiegoo 281 —— rey, sooo once no0es 288 ——leewiusculay rls jeri 284, 285 Inmina on iene 294, 295 danceolataysercterercrarersnepeiete: 277 —— laticeps ...........0-00. 295 Latirostriswey re trieltietelstalck: 285 —— lima .........6.4. 281, 282 — macrodon.............. 293 —— macromystax .......... 294 —— MacropSs ....... eee ee 290 —— maculata ...........005 286 ——magdalene ...... nono 279 —— microlepidogaster ...... 283 —— nigricauda ..........6. 275 —— nudirostris ............ 288 —— nudiventris ............ 290 —— panamensis ............ 301 —— parahemiodon .......... 286 WANBagnodooogocuunned 276 UE Shocospdoooondoo 283 —— phoxocephala .......... 276 —— platycephala............ 295 —— platystoma ............ 298 —— platyura .......... osc CHK) —— plecostomus ...........- 205 —— punctata .......-...0-. 285 —— rostrata........ 299, 300, 501 —— setifera is... ccsvceeee 291 —— simillima .............. 292 ——— SPIK eee ee eles cee 289 —— steindachneri .......... 281 —— strigilata ............., 283 ———stubelil. jsicceesineaset 287 —— teffeana............ . 280 —— typUS..........06.. 286, 287 LIracant happen eicrrs 278 —— valenciennesi .......... 286 VATICLALA careers shale oleae 293 VORMES Guocuocscooppen 293 Loricariichthys .......... 270, 273 —— maculatus.............. 286 loricariiformis, Harttia .,.... 298 ORY LOTICALIA Was etrtactnets 298 luetkeni, Plecostomus ........ 217 lyra, Oxylovicaria..........9. 300 macrodon, Loricaria.......... 298 macromystax, Loricaria ...... 294 macrops, Chaetostomus ...... 234 PIGOVICALI OTe rrettonerr 290 ——, Plecostomus .......... 214 maculata, Woricariasin. «s+ 286 Page maculatus, Cheetostomus...... 246 , Loricariichthys ...... 286 , Plecostomus ........ 286 maculicauda, Otocinclus ...... 268 , Parotocinclus ........ 268 magdalenz, Loricaria ....... 279 malacops, Chetostomus ....,, 255 marcapate, Chetostomus .... 246 marginatus, Chetostomus .... 249 marmoratus, Arges ........ . 3l4 medians, Ancistrus .......... 229 , Cheetostomus .......... 249 ——., Hemiancistrus..... g0cnn 22) megacephalus, Ancistrus...... 234 , Cheetostomus ........ . 234 , Hemiancistrus ..... .. 284 Microlepidogaster............ 265 perforatus ......... .. 269 microlepidogaster, Loricaria .. 283 microps, Cheetostomus. 247, 249, 250 , Neoplecostomus ........ 306 ——,, Plecostomus .......... 306 montana, Xenocara .......... 258 multiradiatus, Ancistrus ...... 228 j LUIPOSATCUS! a) -jaferers/eloleletele 228 , Pterygoplichthys ..... . 228 mystacinus, Ancistrus ........ 238 , Chetostomus ..... mondG ete) —, Hemiancistrus ........ 238 Neoplecostomus ........ . 806 HENOSE)! poooopooto0G6 306 == MICLOPS).teferelotspsfsi ojovaleiale 306 nigricans, Cheetostomus ...... 236 , Hypostomus ....5.. 26. 236 i earancistrusmerriieiine . 236 nigricauda, Loricaria ........ 275 WOtocinclusimyserieecer 268 nigrolineatus, Chetostomus .. 244 HAMS. cooccococgb 90s 244 niveatus, Ancistrus .......... 235 , Cheetostomus .......... 235 , Hypostomus .......... 235 notatus, Hisonotus .......... 268 7 Otocinelus) .yaeerels pogo bts) nudiceps, Cheetostomus ...... 254 >» ELV POStOMUS, oie eeies 264 nudirostris, Chetostomus . 251 SGOTICATION Eyer eters 288 nudiventris, Loricaria ........ 290 occidentalis, Xenocara........ 257 oligospilus, Ancistrus ...... . 232 ——,, Chetostomus .......... 232 j emiancistrusi e).iyeiiever 232 orientalis, Arges .........05. 318 INDEX TO GENERA AND SPECIES. 323 Page Otocinclusmeres erence 265 AIS | SecooadeboosouOS 267 == WMT OHENWS. 5 6500000000000 267 == 11.0) Sane OAD OO aeOon 267 == JONG. sopovcoanooo0500 265 = Ophophaneseeie riers 269 = TCHR coonssso006 268 = THANE. oo cccodgueoce 268 MOLAMS caosoonqoc00000008 268 ——= SHOTS ooonsonc9a006 269 WES ooosocoooecedaad 267 wattabusisc.myeaseca eer 267 Oxyloricariayaa)-seeeeeeeen 297 barbatay:- imei 301 brevirostris: ...-...+-..- 299 frena tavern eee OO —— guentheri .............. 299 loricariiformis .......... 298 Wath Igoooapeoecejoob One 300 panamensis ....-......-. 301 —— platystoma ............ 298 MOWUSLAA ee ctveeraaicte 300 TROSTENIEs GoooodnoDOHEOOS 300 oxyrhynchus, Acestra........ 304 spiarlowellamryirrie iia: 504 Oxaropsisieerrcteclelsirciie rile 263 wrightiana ............ 264 panamensis, Loricaria ........ 301 Oxyloricariaygee/-leei- ter 301 JPEG WO > cocosagnc0000090000 242 Go IOCN. Gogenacc 249, 243 > (MEE GoooosoocbangbEH 245 —— gibbosus .............. 245 niorolineatus .......... 244 pantherinus, Hypostomus .... 209 , Plecostomus ........ .. 213 iRarahemiodonwenrrtrierccine 270 iby PUS a staierereya er eistets se) o's 286 parahemiodon, Loricaria...... 286 parahybee, Ancistrus.......... 241 > IDEN oS occcb0eoo0EC 241 , Plecostomus .......... 219 Parancistrus| 5 .....0.0-... 221,222 aULANtIACUS) sje)... LO —— nigricans .............. 236 punctatissimus.......... 236 pardalis, Hypostomus ........ 228 7 LEIPOSAECUS) =.) \e)01-1+/0\0 5 « 228 ——,, Plecostomus .......... 228 , Pterygoplichthys ...... 228 Parotocinc luster tattle 265 maculicauda............ 268 parva, Loricaria ............ 276 paulina, Loricaria............ 283 pellegrini, Plecostomus ...... 218 Page perforatus, Microlepidogaster.. 269 Ococinclusteeeetistlll- 269 peruanus, Arges ..........4. 317 phoxocephala, Loricaria ...... 276 pictus, Ancistrus ........ 231, 237 , Chetostomus .......... 231 ——, Hemiancistrus.......... 231 ——, Hypostomus .......... 237 Pimelodus cyclopnm.......... 311 platycephala, Loricaria ...... 295 platycephalus, Ancistrus ...... 235 , Chetostomus .......... 235 ——,Hemiodon ............ 295 platystoma, Loricaria ........ 298 wOxyloricariayenyyseitelele 298 platyura, Loricaria .......... 270 Plecostomus ........ 202, 203, 221 afhiniseyaveriaieicielaavecareuerelste 207 ——= HENS) ce6c0q9000090000 211 —— angulicauda ............ 241 ——= AIMED. Gis000 0000009800006 210 ———= AVE ooodnoand0000000000 220 as pilomastermsyteleitelcleyeiers 206 ———ALTOPINMISh ya (eset eestor 215 —— auroguttatus .......... 217 —— barbatus .............. 240 bicirrhosus ........ 205, 209 —— biseriatus .............. 210 borelliay civ ceiesaesssersie.e 212 boulencerigerrrtrl errr 206 brasiliensistemeryreiirecierr: 205 IDC, sooccdoo0aed 216 CEisiMERIE) Gooogoon0dod0d 209 GHONTOCNS 5500000000000 HH commersonil........ 206, 207 CONE oanodco0000c0c 212 emarginatus ............ 210 fostaes se, Sasintinosshisers 208 Hacellanistrteeeteeeerr 291 ——— HEA baoocugooddo000C 205 ——=AnGEINGED poocecaegoccod 211 garment gear 214 genibarbisiaeee eerie 219 $= {EIN Gooc00cc0n0000 306 IMENES soncooo0c0s00K6 24) WENCH soaccooccoa008 221 IQAHGIS GooooncoGcono0 210 iM, Soo0000000000000 217 ONS oooc00c0an0000 213 —— ima! aaa heisiee 214, 216, 217 —— limosus.....,.......... 206 IGANG congogn00d0000¢ 217 MACTODSeetettmirreerrrets 214 AAGMEUS 5 500000000000 286 = MICLOPS).) feleoerenedthereate 306 —— pantherinus ............ 213 Page Plecostomus parahybe........ 219 > EEN 554 ogop000000600 228 pelleorini eres 218 —— plecostomus........ 205, 206 ——> DUNES 6 sooocvaccntbeo 207 = RON Sooqoccv0cc0de0c 215 SCADTICEPS I yeeros 207 See SCODLUATIUS Macys 210 > HEMMINGS, sn cobooooooes 206 ——= SPINE» acaoooostooocce 206 ——— SOMONE - 55 50¢g0000 208 —— TOMEI, oracscooocen 210 UN, cos bo0000c00D00Ne 213 WE) oococacc0nbo0000C9 215 venlllertn cocoaoccccsc0$ 210 = VOMIMTEWMETES o conocaacoec 216 VERE) “dooonosocodoo000 209 —— Wiles soogoassosocacsc 210 === REIN. oooondopoanady 210 WHINY, 555500000000 218 = VGN s soon onc0 000s 216 plecostomus, Hypostomus . 205, 209 = VOICES, 6 5500000d00000 205 , Plecostomus........ 205, 206 IAW ONO NING, occocoonoosoaces 205 prenadilla, Arges ............ 316 s_Drontes'® steeritacieiete 318, 316 Pseudacanthicus ............ 259 fordiie aay nee eae 261 ———_ hy Sstrixn ante eee 261 = ENTS 5 oococccvcvnscen 260 PUGH. coscocccasanous 260 Pseudancistrus .......... 221, 222 ———— baxbatusie-eere eae naae 240 depressise nner ner enn 240 ——= UUW, oopoancoogoocon 240 —— SAWS guocooocogduabao 239 TWINS voooccoo0oen 218 Pseudohemiodon ............ 270 Pseudoloricaria.......... 270, 272 Pterygoplichthys ............ 221 —— FM, soenovoconos0000 LAl —— duodecimalis............ 226 ——etentaculatum .......... 226 —— gibbiceps .............. 227 IPMETAIS socaoccoocce LAS WEA) poneoraocancd see 221 IES soocooaccc000e 227 —— multiradiatus .......... 228 pardalis hearin yeye rere 228 = DUNCLALUSE ieee oO) undecimalisiaeaeeeennnee 226 punctata, Loricaria .......... 285 punctatissimus, Ancistrus .... 236 , Cheetostomus .......... 236 ——, Parancistrus .......... 236 punctatus, Ancistrus.........4 ——, Chetostomus .......... p ——, Hypostomus............ ——,, Plecostomus............ ——,, Pterygoplichthys ...... Rhinelepis........ ,. 202, 205, ACANTNICUS reece ide dsta AP ASSIZII ee letersi el seolece ee ERIE) cobsumousnoentad <2 2 genibarbis.......5..+-6 2 = hfs] 48) Geritigioan OMe ounue —— lophophanes............ SUTIZOSA 6.56. eee ee ees Rhineloricaria .......... 270, robini, Hypostomus ...... PISIOCOSLOMUS, Geisler shaaerste robusta, Oxyloricaria ........ rostrata, Loriearia .. 299, 800, , Oxyloricaria. so 055.0250. sabalo, Arges........+. . seabriceps, Plecostomus ...... scaphirhynchus, Ancistrus .... » Cheetostomus 2 .ris. iar , Hemiancistrus.......... schomburgkii, Ancistrus ...... , Hemiancistrus.......... scopularius, Plecostomus...... 2 scrophus, Liposarcus ........ 22 seminudus, Plecostomus ...... sericeus, Chestostomus........ serratus, Cheetostomus........ selemiancistrushe sear 2 ——, Hypostomus .......... ——, Pseudacanthicus........ setitera, Woricariay se -ictecs: : setosus, Ancistrus............ 23 7 Ohestostomus! etc cia 23 , Pseudancistrus ........ simillima, Loricaria .......... SIMONSI,PATSCS a jarvernienit ae ce: spiniger, Plecostomus spinosissimus, Plecostomus ... . spinosus, Chaetostomus........ , Hemiancistrus.......... y ——,, Hypostomus , Pseudacanthicus........ spixii, Loricaria squalinum, Hypostomus stannii, Cheetostomus steindachneri, Hypoptopoma .. SMLOXICArlayensleeyeraistanaecrey y stigmatica, Xenocara ........ y stigmaticus, Ancistrus strigilata, Loricaria strigosa, Rhinelepis stubelii, Loricaria subcarinatus, Hypostomus .... taczanowskii, Arges , Chetostomus .......... tectirostris, Cheetostomus teffeana, Loricaria temminckii, Ancistrus .. , Hypostomus tenuicauda, Plecostomus ternetzi, Plecostomus thomsoni, Cheetostomus thoracatum, Hypoptopoma. 263, 265 trinitatis, Cheetostomus , Hemiancistrus.......... typus, Loricaria une, Plecostomus undecimalis, Ancistrus........ , Cheetostomus FISHES OF THE FAMILY LORICARIIDA. undecimalis, Pterygoplichthys . uracantha, Loricaria.......... 27 vaillanti, Arges..........06+5 spEleCOStOMIUS mviiainsie e 2 valenciennesi, Loricaria ...... Z variegata, Loricaria variolus, Cheetostomus........ 256 varius, Liposarcus............ 225 vermicularis, Plecostomus verres, Plecostomus .......... Z vestitus, Otocinclus .......... VetularMoricarlanrcirttrcters vicinus, Hypostomus ........ villarsi, Plecostomus.......... virescens, Plecostomus........ vittatus, Ancistrus .......... 233 ——, Chetostomus .......... ——,, Hlemiancistrus.......... ——,, Otocinclus ............ wertheimeri, Plecostomus ——., Pseudancistrus ........ whymperi, Arges ........ wrightiana, Oxyropsis ........ 26 wuchereri, Plecostomus ....,. 2 OXEHOETE Sagccespnoodaocoun 2 brevipinnis: Vs. s i. as. 2 ——bufonia so... cins ese e es 26 === cChagresi. . 66. see eens 256 Citrhosa versus eeiicnclere —— dolichoptera............ gymnorhynchus ........ —— hoplogenys ............ 255 —— latifrons .............. —— montana .............. 268 loecidentalisueasy heer Aina oonosacdeoso 2 (AMINO Goocneaoonor Kenomystus .1.... 60. s eens —— GObIO Peeve eee eee eee es \ Rourions ple 326 FISHES OF THE FAMILY LORICARUDA, Jeb 1D. Fig. 1. Plecostomus feste (x3), p. 208. Dh Ee borellit (x $): 2a, head seen from above, p. 212. 9 oe - cordove (xX $): 3a, head seen from above, p. 212. “WAOGHOOAS %IITITHORa dG WISHA SNNOLSOONTdA'1 ‘dure soag ura pig TEMES E)D VOR ae AL Ab LN D0 POOL hit we UN (eas ARG dh Ta aay a ie ene, Agente Sore PLATE X. yOL. XVII.—PaRT 11. No. 18.—October, 1904. 328 FISHES OF THE FAMILY LORICARIIDA. UPLATE X. Fig. 1. Plecostonvwus garmani, p. 214. 2. 5 wucherert (Xx $), p. 216. THaeqHonM' d's] INVNGVD SNNOLSODXTd T ‘hha soug usaquIyy “WIL 38 [ep wean pr VBE UM Pele? PRE ROL oan Annoy Theta ek a Aye - rua out ae Oo FISHES OF THE FAMILY LORICARIIDA. PLATE Xi Fig. 1. Plecostomus latirostris, p. 213. 2. Ancistrus annectens (X 3%), p. 225 ‘SNHLOANNY SMUYLSIONYVG@ ‘SIULSOUILVT SONOLSOORTd I dua soag wsa3uy UGH 32 "TSP weerg ¢ LN 26 MAK 070° G7 IO \j ne Are Dy aa [2 hed FISHES OF THE FAMILY LORICARIIDA. PLATE XII. Fig. 1. Ancistrus oligospilus: 1a, head seen from above, p. 232. nN 7) mm © setosus: 2 a, head seen from above, p. 239. guentheri: 3a, head seen from above, p. 241. platycephalus: 4a, head seen from above, p. 235. Trarsdood. Ko Vl MUU F%. XM. Mintern Bros, imp. J.Green del.et lith. 4.A.PLATYCEPHALUS. 2.A.SETOSUS. 3.A.GUENTHERI. 1.ANCISTRUS OLIGOSPILUS. Aah si 1 iil as Ns Nae OAL GAD ea Ss ening Seg Hy r — by = bd i=) =I = — A 334 FISHES OF THE FAMILY LORICARIIDA. PLATE XIII. Fig. 1. Chetostomus marginatus, p. 249. 2. By anomalus, p. 259. 3. : brevis, p. 247. TrambZi0k eo Yl XV PY AM MinternBros.imp. J.Green delet hth. 3.¢ BREVIS. 2.C.ANOMALUS. 1.CHATOSTOMUS MARGINATUS. fee ea} tare iat Cy Fy ; Ne PGA II CIN VoL. XvViI.—ParT 11. No. 19.— October, 1904. 396 FISHES OF THE FAMILY LORICARIIDA. PLATE XIV. Fig. 1. Chetostomus marcapate, p. 246. 2. S thomsont, p. 250. 3. 5 microps, p. 251. 4, maculatus, p. 2406. [p. 257. 5. Xenocara occidentalis: 5 a, heads of male ( ¢ ) and female ( 2 ) seen from above, 6. montana, p. 258. ; le 5 chagresi, heads of male (¢ ) and female ( ? ) seen from above, p. 256. Tranbsool. He Vl MVM S¢, XIV. MinternBros.imp. J.Green del.et lth. 4.Cc.MACULATUS. 3.C.MICROPS. 6.X. MONTANA. 7.X.CHAGRESI. 2.C. THOMSONI. AT AS . Pp 1.CHATOSTOMUS MARCAI 5.XENOCARA OCCIDENTALIS. eerie ; Baeble se th a Tai a TO ie gs t PIAL GM Ly ae eae ty Ne UP caal (Sy) (os) Fi ~ Be onw FISHES OF THE FAMILY LORICARIIDA. PLATE XY. . Xenocara latifrons: 1a, head seen from above, p. 253. . Hypoptopoma quenthert, p. 264. . Otocinelus vittatus, p. 267. . Loricaria microlepidogaster : 4a, head seen from above; 40, head and abdomen seen from below, p. 288. . Loricaria jubata: 5 a, heads of male (¢) and female (2) seen from above, p-. 278. ‘VIVE TS SALSVOOGIAHIOMOIN VIEVIlacl 7 ‘SOIVLLIA SNTIONIOOLO§ “lGSHLINEND VNOGOLdOaAH “SNOGHILVGD VAaVOONAX ft + durt Soag waaay £ “UAL a2 2p w8eIg -¢ nigh A ee AN 26 MAN 9% BHO 340 FISHES OF THE FAMILY LORICARIIDA. PLATE XVI, Fig. 1. Loricaria latirostris: 1 a, heads of male ¢ ¢ ) and female ( 2 ) seen from above ; 1 4, head and abdomen of male seen fiom. below, p. 283. 2. Loricaria nigricauda: 2a, head seen from above; 2%, head and abdomen seen from below, p. 275. ViGMAVOTES IN Ds “SISLSORLIVG VIEVoOIeaOT I Yat 4e'jePp useeug UL OE MN 720° 97S OE » Wi es ‘ nee y Pap Wise 3 iV phat Tie Wane wha (Se) ree ih) FISHES OF THE FAMILY LORICARIID. PLATE XVII. Fig. 1. Loricaria punctata: 1a, head and abdomen seen from below, p. 285. 2. Loricaria simillima: 2a, head seen from above; 24, head and abdomen seen from below, p. 292. . Loricaria macrops: 3a, head seen from above, p. 290. “‘SdOHOVIN TE “VWWITTINIS “1G ‘VIVILONONd VIEYVOIEAOTTI ‘duat- soag usaqurpy VWBLI32 [ap usean ¢ | { 1 \ HAN Ab MAY 220° BOS" To PLATE XVIII VoL, XVII.—PART 111. No. 20.—October, 1904. bo Ss 344 FISHES OF THE FAMILY LORICARIIDA. PLATE XVIII. Fig. 1. Oxyloricaria quentheri: la, head seen from above; 16, head and abdomen seen from below, p. 299. 2. Oxyloricaria frenata: 2a, head seen from above; 26, head and abdomen seen from below, p. 302. Wal NAN BLEL el © (©) G ‘THUSEHLNGEND VIEVOIAHOTAXO I “Wat qe TPP weezy’ ¢ duu soug mre Sarpy WAL 26 MAY 7X0 PB OSes | i si} a as ahs a 7 ie} Aha “iNecttee H a. a Ta Ahi nly i CUTE RIK, | are: beeen 346 FISHES OF THE FAMILY LORICARIID. PLATE XIX. Fig. 1. Oxyloricaria robusta: 1a, head seen from above ; 16, head and abdomen seen from below, p. 300. 2. Oxyloricaria lyra: 2a, head seen from above; 26, head and abdomen seen from below, p. 300. ‘VaAT OC WLISNEGOU VIEVOLHOTAXO 1 ‘dur soag usa qurpy YERISE [eee ie AU bo UL G VOPR GOT | i qh e ; at , t iy i ) i i i ih ie ; i 1 V i aS i i yun} \ i ‘ i ‘ Ss tng, ian y ery et (3) TSG [oa) Fig Bl 9 Oo FISHES OF THE FAMILY LORICARITID. PLATE XX. . Loricaria laticeps: 1a, head and abdomen seen from below, p. 295. . Farlowella gladiola: 2a, head seen from above (X14); 26, head and abdomen seen from below (X13), p. 305. . Farlowella gracilis: 5a, head seen from above; 30, head and abdomen seen from below, p. 303. “SMUD tal) GL + dust: soug we purpy TN AG MME GV0PR OO De MO WGN TD) VITEMOTEVsS @ Saal Winehy OlslOW I “YAW Ie TEP USeTy LP he yp pve PLATE 3X1 (xi Fig. Neo ie 2) TS Or oo DD FISHES OF THE FAMILY LORICARIIDA, be) 9 3” be) PLATE XXI. . Arges homodon, p. 309. guentheri, p. 310. eigenmann, p. 312. orientalis, p. 313. chote, p. 315. marmoratus : 6a, head seen from above, p. 314. feste: Ta, head seen from above, p. 316. brachycephalus: 8 a, head seen from above, p. 315. simonsit: 9a, head seen from above, p. 517. Sand Loo SooPl AVI. FYAAT Mintern Bros imp. 4,4.0ORIENTALIS 5.A.CHOTAS. ‘PHALUS. 9.A.SIMONSI. J. Green del et lith. 1L.ARGES HOMODON 2.A.GUENTHERI. 3.A.EIGENNA 6.A.MARMORATUS, 7.A.FESTZ. 8.A ACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON (continued). To Fellows. To the Public. EN Ei de £ s. a. VOLUME XIV. (1896-1898, containing 47 Plates) . Price5 5 0 GOO Parr 1. (1896, containing 2Plates) . .... » 0 6 0 2038s 0 » 2. (1896, containing 6 Plates) . . |. . sabes 9) 0 ao ORO », 3. (1897, containing 9 Plates) - . - . . 5 Oma 9, gto les a0) » 4. (1897, containing 10 Plates) . . . . . pre Os 0 2 0 0 » 5. (1898, containing 2 Plates) . . ... a Oe 6900 S10 S20 » 6. (1898, contaming 7 Plates) . . .. . 3.0 180 1 4 0 » 7. (1898, containing 6 Plates) : ; 53 ONLI 015 0 » 8. (1898, containing 5 Plates and Title and Index) , OFS O 012 0 VOLUME XV. (1898-1901, containing 52 Plates) . Price5 15 6 714 0 Parr 1. (1898, containing 8 Plates) . . . . - ye Onlomeo Jere O » 2. (1899, containmg 5 Plates). . .. - ee Oe oO 012 0 » 98. (1899, containing 4 Plates) . . . . . Be OMe Ore O 012 0 » 4. (1899, containing 4 Plates) . ... . SO OnE 012 0 »» D. (1900, containing 24 Plates) . . . . . ee Lea 310 0 » 6. (1901, containing 4 Plates) : 5 OR Mligee 015 0 » 7. (1901, containing 3 Plates and Title and inden 09 0 012 0 VOLUME XVI. (1901-1903, containing 38 Plates) . Price5 8 0 7 4 0 Part 1. (1901, containing 6 Plates) . . .. . 55) OLSiO 1 4 0 » 2 (1901, containing 5 Plates) . . ... , Ol 8 Oulbas0 » 3. (1901, containing 9 Plates) . . .. . sap OF, LO dba [aX 0) » 4 (1902, containing 5 Plates) . . - ... fh, O22 OU @ 012 0 » 5. (1902, containing 4 Plates) . . . . . Sa OnlownG 018 O », §6. (1902, containing 3 Plates) . . . . - 53 Onl Mee Leas » 7. (1902, containing 3 Plates) : “ pee OF Lees 015 O » 8 (1908, containing 3 Plates andTitle and Index), 013 6 018 0 VOLUME XVII. Panr 1. (1903, containing 5 Plates) . . . . . Pricel 2 6... 110 » 2. (1908, containing 3 Plates). ... . ss Ole LS eh inner On Levee O 3. (1904, containing 13 Plates). . . . . oes" ENE OaeO morc weet Bc Let. OF ”? CONTENTS. III. A Monograph of the Fishes of the Family Loricarupa. By C. Tats Reean, BACH ZS... (Plates LX.-X XI) 908 ye ees ee ee pa se adioll THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Tue scientific publications of the Zoological Society of London are of two kinds—“ Proceedings,” published in an octavo form, and ‘ Transactions,” in quarto. According to the present arrangements, the “ Proceedings”’ contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended to be published in the “ Proceedings” by the Committee of Publication. A large number of coloured plates and engravings are attached to each annual volume of the “ Proceedings,” to illustrate the new or otherwise remarkable species of animals described in them. Among such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “Proceedings”’ for each year are issued in four parts, in the months of June, August, October, and April, the part published in April completing the volume for the last half of the preceding year. Beginning from January 1901 they form two half-yearly volumes. The “Transactions ”’ contain such of the more important communications made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are issued at irregular intervals. Fellows and Corresponding Members, upon payment of a Subscription of £1 1s. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. ‘They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1871, if they exceed the value of five pounds. Fellows also have the privilege of subscribing to the Annual Volume of the ‘ Zoological Record’ for a sum of £1 (which includes delivery in the United Kingdom only), but this privilege only holds good if the subscription is paid before the First of December in each year. Such of these publications as are in stock may be obtained at the Society’s Office (3 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. P. CHALMERS MITCHELL, October, 1904. Secretary. i A oe NS wee eS Flare TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Vou. XVII.—Parr 4. LONDON: PRINTED FOR THE SOCIETY, SOLD AT THEIR HOUSE IN HANOVER-SQUARE ; AND BY MESSRS, LONGMANS, GREEN, AND CO., PATERNOSTER-ROW. October 1905. Price 30s. Taylor and Francis, Printers. ] [Red Lion Court, Fleet Street. TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. To Fellows. To the Public. BE esi aaals £ 5. d. VOLUME I. (1833-1835, containing 59 Plates). . Price 313 6 . . . 418 O* VOLUME _ II. (1835-1841, containing 71 Plates). . 5 400... 5 6 6* VOLUME III. (1842-1849, containing 63 Plates). . 4, 8 8 8... 411 O* VOLUME IV. (1851-1862, containing 77 Plates). . 5, 6 2 0... 8 2 6f VOLUME _ V. (1862-1866, containing 67 Plates). . , 5 4 38 . . . 619 OF VOLUME VI. (1866-1869, containing 92 Plates). . , 11 5 O .15 0 OF VOLUME VII. (1869-1872, containing 73 Plates). . 5, 10 4 O . . .1812 OF VOLUME VIII. (1872-1874, containing 82 Plates). . ,, Oe Sia is Le dee Orr VOLUME IX. (1875-1877, containing 99 Plates). . ,, 12 1 6. . .16 2 OF VOLUME X. (1877-1879, containing 95 Plates). . , 10 0383 .. .18 7 , OT GENERAL INDEX, Vols. I. to X. (1833-1879) . ,, Or Og ae OO) VOLUME XI. (1880-1885, contaming 97 Plates). . ,, 912 0... 12 16 0 VOLUME XII. (1886-1890, containing 65 Plates). . ,, DB HOw eke or an VOLUME XIII. (1891-1895, containing 62 Plates) . # 6 8 38 811 0 * No copies of these volumes remain in stock. + Only complete copies of these volumes are left in stock. Yontinucd on page 3 of Wrapper. IV. A Contribution to our Knowledge of the Varieties of the Wall-Lizard (Lacerta muralis) in Western Europe and North Africa. By G. A. Bounencnr, 7.R.S., V.P.ZS. Received November 22, 1904; read February 21, 1905. [Puates XXIJ1.-X XIX, and Text-figures 1-6.] Turrry years have now elapsed since special attention began to be paid to a systematic study of the variations of the Wall-Lizard of the Mediterranean basin, our knowledge of which has been so greatly advanced by the publications of Bedriaga, Eimer, Braun, Werner, and others. It is Bedriaga who, bya valuable series of papers, issued between 1874 and 1883, has most contributed to this advance, from the systematic point of view, Eimer’s researches being more of a theoretical kind, and almost entirely confined to Italy and Malta. And to Bedriaga also we are indebted for a complete summary of the state of the question in 1886, at which date appeared his great monograph of the genus Lacerta, published by the Senckenberg Society. Since that time, however, a much larger material has been examined, and this has resulted in new views on the definition of the races and on their mutual relationships. I have always been keenly interested in the question, and have missed no opportunity of adding to the collection in the British Museum, which is now a very imposing one, both from the point of view of the number of specimens and of the variety of localities. Of late a tendency has sprung up to greatly multiply the species and thus destroy the old conception of Lacerta muralis. 1 doubt whether such attempts will conduce to a better understanding of the subject, and I have often been struck by the want of discrimination in the selection of characters on which many of these supposed species are founded. Characters of form and coloration are given as distinctive which or examination of even moderately large series of specimens prove to be worthless, whilst others of greater importance have been overlooked or neglected !. The object of this paper is to supply detailed descriptions of the specimens in the British Museum, and by drawing special attention to the individual variations, which are so frequently passed over in systematic definitions, to furnish a sounder basis for future work. I have endeavoured at the same time to review the work of my predecessors, so far as I ‘ Kimer, for instance, attempted to classify the variations according to the pattern of coloration and the shape of the head, practically neglecting everything else. His work, therefore, although extremely interesting and containing much that is true as regards the derivation of markings, cannot be said to have adyanced our knowledge from the systematic and zoogeographical points of view. VOL. XVI.—ParT Iv. No. 1.— October, 1905. © — 392 ‘MR. G. A. BOULENGER ON THE VARIETIES OF have been able to have access to type specimens. For this purpose I have been fortunate in being allowed to utilise the collection of M. F. Lataste, which has been entrusted to my care, and to compare many of the type specimens described by Dr. de Bedriaga, which have been acquired for the British Museum since the publication of the ‘ Catalogue of Lizards.’ Iam also greatly indebted to Count Peracca for friendly criticism and much important material in connection with this work, and to Dr. de Bedriaga, Prof. Camerano, Dr. Gestro, Dr. Werner, Dr. Vinciguerra, Prof. Monticelli, M. F. Doumergue, the Rey. Fathers G. Fournier and 'T. Neve, Dr. Gadow, M. Dollo, Sr. A. Ferrer, Hr. Lorenz Miller, Dr. J. Roux, Mr. Bryan Hook, and Prof. I. Bolivar for loan or gift of specimens. This paper does not profess to be a monograph: it is simply a contribution to our knowledge of these Lizards studied in relation to their distribution, a continuation of the excellent accounts given in similar manner by Bedriaga in the seventies of last century—with this difference, however, that the lepidosis is described with greater precision. It deals only with the western parts of Europe, including Italy, and with North Africa; the Lizards of Europe east of the Adriatic, and of South-western Asia, are at present being studied by Dr. Werner, Hr. Lehrs, and Prof. v. Méhely, and it is better to await the conclusion of their labours before preparing for publication the results of my examination of these Lizards, which I propose, however, to furnish ere long. When this is done, I hope to be able to summarise the whole subject in a strictly systematic order. In order to ensure absolute accuracy, the figures illustrating this paper are reproductions of photographs, on which great pains have been bestowed by my excellent artist, Mr. J. Green. The coloured Plates have, of course, been prepared from living specimens. I—CENTRAL EUROPE. (Plates XXIV. & XXV.) All the specimens from which the following description is drawn up belong to what may be called the typical form in its narrowest sense, the species Lacerta muralis having been established by Laurenti! upon specimens from the vicinity of Vienna, where, as is well ascertained 2, no definable varieties occur. In order to have before me ‘‘ topotypes ” of the species, I applied to my friend Dr. Werner, who, with his usual kindness, at once sent me two specimens, a male and a female. Very curiously, the latter turns out to be the most aberrant example, so far as the scaling is concerned, which I have found among the hundreds of specimens examined by me from France, Belgium, Western Germany, Switzerland, and various parts of Austro-Hungary, and 1 Syn. Rept. p. 61, pl. i. fig. 4 (1768).—Seps muralis. * Werner, Rept. u. Amph. Oesterr. Ung. p. 40 (1897). (oN) or wo LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. adds considerably to the ascertained amplitude of variation in the number of dorsal scales and femoral pores. However much all these specimens vary in scaling and coloration, I feel satisfied that further subdivision into local varieties is unfeasible, and that even insular forms such as have been named by R. Blanchard! are not capable of definition. Form and Proportions. Head about once and a half or once and three-fifths as long as broad, more or less depressed, its depth, in the tympanic region, equal to the distance between the anterior corner, the centre, or the posterior corner of the eye and the anterior border of the tympanum; snout obtusely pointed. Neck as broad as the head, or slightly constricted. Body rather strongly depressed. Hind limb, in the males, reaching the axil, the shoulder, or the collar; in the females, reaching the wrist or the elbow of the fore limb laid against the body; foot once and one-tenth to once and one-third the length of the head, rarely (in males) not longer than the head. Tail cylindrical or cyclotetragonal, often slightly flattened at the root, once and two- thirds to twice and one-sixth the length of head and body. Scaling. Nostril (text-fig. 1) pierced between nasal, postnasal, and first upper labial ; exceptions, in which the rostral enters the nostril, are extremely rare >. Nasals usually forming a Text-fig. 1. End of snout in two male specimens of Lacerta muralis from the Bies Glacier, near Randa, Switzerland. median suture, variable in extent; rostral rarely in contact with frontonasal?; latter some- times in contact with frontal*+; an azygous shield sometimes separates the prefrontals (Pl. XXYV. fig. 2); frontal usually as long as its distance from end of snout, often a * Mém. Soc. Zool. France, iv. 1891, p. 502. * One specimen from Fontainebleau (Lataste Coll.), and one from the Bies Glacier, near Randa, Switzerland (altitude 6500 feet). * Four specimens from Eaux-Bonnes, Pyrenees ; 3 (out of 4) from near St. Malo (Pl. XXV. fig. 3); 2 from near Dinant, Belgium; single specimens from Paris, Bordeaux, Fontainebleau, Glenan Isles, and Bazias (S. Hungary). * Single examples from Bordeaux, Haux-Bonnes, Glenan Isles, Herkulesbad, Bozen. 3Aa2 304 MR. G. A. BOULENGER ON THE VARIETIES OF little shorter, sometimes a little longer, as long as or a little longer than the fronto- parietals!; parietals once and one-fourth to once and a half as long as broad, usually in contact with the upper postocular ?, the outer border never distinctly concave for the reception of the supratemporal. A female specimen from Voéslau, near Vienna, is remark- able for the very regular division of each parietal into two by a transverse suture (Pl. XXV. fig. 4). Occipital usually shorter than, and as broad as or narrower than, the interparietal, but sometimes as long as the latter and not rarely broader 3; a small additional shield not unfrequently intercalated between the two. A series of granules between the major (second and third) supraoculars and the supraciliaries; the first or the first and second supraciliaries usually in contact with the second supraocular (Pl. XXV. fig. 1), but the series of granules sometimes complete, extending from the first supraocular to the upper postocular. The specimen from near Vienna, noticed above as having the parietal shield divided, is further anomalous in the total absence of the first supraocular (Pl. XXV. fig. 4). Postnasal single+; nasal usually separated from the anterior loreal by the postnasal®. Temporal scales small, often granular; masseteric disk and tympanic shield nearly always present®; masseteric disk usually large, round or oval, and separated from the parietal by a supratemporal shield or one or two, rarely three, series of scales. Four upper labials anterior to the subocular, very rarely three or five, and on one side only ; subocular very variable in shape’. 20 to 30 scales and granules in a straight line between the symphysis of the chin-shields and the median collar- plate; gular fold usually distinct, indicated by one, two, or three series of minute granules. Collar without, or with merely a trace of, serration, composed of 7 to 12 plates, usually 9 to 11 (not including the outer, smaller plates which are separated from the edge by granules). Scales on body granular, juxtaposed, round or oval-subhexagonal, more or less distinctly keeled, rarely smooth ®, often a little larger on the back than on the ‘ Shortest frontal, measuring only three-fifths its distance from end of snout, in a male from St. Malo (Pl. XXY. fig. 3). * Exceptions in three specimens from Glenan Isles and in three from Eaux-Bonnes. JI must here correct a lapsus in my description of Lacerta depressa (P. Z.8. 1904, ii. p. 333) where “upper supraocular” should read ‘“‘ upper postocular.” * Occipital very minute, almost reduced to a granule, in one specimen from Dinant. * Absent, fused with the anterior loreal, in a specimen from St. Malo; divided into two in a female from Paris (Lataste Collection). ° Exceptions, in which the nasal joins the anterior loreal above the postnasal, are not at all unfrequent ; { have noticed them in specimens from St. Malo, Glenan Isles, Eaux-Bonnes, Odilienberg, and Randa. ° Absent in a specimen from St. Malo and in one from Eaux-Bonnes. 7 Its lower border sometimes as long as its upper border, a character which has been given as diagnostic of LL depressa. 8 In two specimens from Dinant, Belgium. LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. 359 sides; 42 to 62 scales across the middle of the body, usually 47 to 591; 3 or 4, rarely 2 or 3? or 4 or 52 transverse series correspond to one ventral plate, 29 to 45 to the a length of the head. Ventral plates in 6 longitudinal and 23 to 52 transverse series; the plates of the two median series as broad as or narrower than those next to them. Text-fig. 2. a. Preanal region of Lacerta muralis. a. 9, Eaux-Bonnes. 6. ¢, Eaux-Bonnes. c. ¢, St. Malo. Preanal plate (text-fig. 2) very variable, but usually large, bordered by one semicircle of small plates, often with an outer semicircle of smaller plates or scales. Scales on the upper surface of the tibia granular, more or less distinctly keeled, usually smaller than those on the back, rarely nearly as large. 22 to 28 lamellar scales under the fourth toe. Femoral pores from 13 to 25 on each side, usually 16 to 214. Caudal scales (in the basal fourth) usually narrow, truncate or more or less obtusely pointed, the upper more or less strongly keeled, the lower smooth or feebly keeled; the whorls alternately longer and shorter, this feature more distinct in some specimens than in others; the fourth or fifth whorl behind the postanal granules contains 25 to 32 scales, exceptionally 20°. Coloration. Upper surface varying from grey to brown, sometimes with a tinge of greenish; two light, whitish, yellow, or greenish streaks or series of spots on each side, the upper from the outer edge of the parietal shield to the upper surface of the base of the tail, the lower from the upper lip, through the ear, to the groin (Pl. XXIV. figs. 3 & 4); between these two streaks, on each side, a black or dark brown band, which may be- spotted with whitish, yellow, bluish, or greenish ; the back often spotted or marbled with dark brown or black; frequently a dark brown or black vertebral stripe or series of more or less confluent spots. Some specimens, especially males, entirely marbled ‘The minimum (42) occurs in a female from Véslau, near Vienna, the maximum (62) in a female from Dinant (Pl. XXY. fig. 5). * In a male from St. Malo and in a male from Fontainebleau. ° In a male from Dinant, in a male from St. Malo, in a female from Haux-Bonnes. * The minimum (13) occurs in a female specimen from Voslau, near Vienna. I cannot agree with Werner (Zool. Jahrb., Syst. xiv. 1903, p. 333) when he says that the number of femoral pores increases with age. At least this is contrary to my experience in this lizard, quite young specimens from France and Germany having 16 to 19, as in adults. * Again in the aberrant specimen from Véslau. 506 MR. G. A. BOULENGER ON THE VARIETIES OF or reticulate with black, without or with mere traces of the light lateral streaks (Plate XXIV. fig. 1). Upper surface of head usually more or less dotted or spotted with black. Sometimes a well-defined black ocellus, with white or green centre, above the shoulder or axil!. Black and white spots on the sides of the tail, usually forming regular bars. Lower parts white or yellow in females, white, pink, yellow, orange, or brick-red in males; blue spots on the outer row of ventral plates, especially in males; belly usually immaculate, or with scattered black dots in females, more or less spotted with black in males. In some specimens from Dinant, in Belgium (Plate XXIV. fig. 2), the black spots are confluent into longitudinal bands on the sides of the belly. Young, just out of the egg, are dark grey above with round or vermicular lighter spots; a black vertebral line is often present; the two lateral streaks are usually very distinct and formed of more or less confluent white spots ; the blackish band between them, continued from the temple, bears small round white spots ; limbs blackish, spotted with white ; lower parts white. (See text-fig. 4a, p. 369, and text-fig. 6 a, p. 399.) Measurements of adults (in millimetres) :— 1 2. os 4, 5 6 TOMA KOKR WOES 6 5 a 6 5 a Ue 67 62 60 yy 59 ay » fore limb . gets 22 25 21 24 21 Ibe GENER! G @ cis 6 o o 9 Ie 14 16 14 15 13 Width of head 12 S i 9 9 9 Depth of head Syma tune 7 8 7 7 7 orevimbh muster men Pa ee Oe eee 20 23 19 20 18 Fandelimib) eee en SONS 5 mee ONES OMG ROGE fecal, Aye i eet 0k Roly el Ome Game wl aa See ete See SIC TIE NOR HOR IO: Bile * Tail reproduced. Glenan Isles, Brittany (Giard). 22 22 3) Eaux-Bonnes, Pyrenees (Lataste). a) a) 9? Herkulesbad, Hungary (Méhely). 3? 39 2 er. Os fe) LS) 40 Oy 40 Oy 40 OS In the following table I give the size and record variations in the lepidosis in a number of adult specimens in the British Museum or in the Lataste Collection :— te 2. 3. 4. 5. 6. 7. Nr. St. Malo . Si OGN 47 6260 lew) Oral Ope ‘ 2 6 Bh 8 © BF oD Tf 3 OG BO Bh © BW 1 Fl _ 9 153) 50) 29) 959) 925 2022 2mmO 1 Very well marked in some specimens from Paris, the Pyrenees, and the Alps near Randa. 1. Length (in millm.) from snout to vent. 3. Transverse series of ventral plates. granules between symphysis of chin-shields and median collar-plate. (on right and left sides, if differing). LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. Great Chausey Id. 3) 39 Glenan Isles ” . . Bouron, Fontainebleau . 2) 2) Kaux-Bonnes, Pyrenees . 3d 3) Dinant, Belgium 37 3) Lower Ahr Odilienberg, Vosges . bP) 2? Randa, Switzerland 9 3 i! Miesenbach, nr. Vienna Voslau, nr. Vienna Herkulesbad, Hungary . > 2? Bozen, 8. Tyrol 33 >) 3) supraocular. iy) ey Ce Eee Gs.) ee) Oe Bil ee Ose ae 9 OG ay eo) G65) 560 24) 10 9 638 54 2% 9 BGS aie 2a 5 Ch OF Gin Od 25 1 sy es © 65 48 29° Ia » G8 EB 23 1@ >» © Bl BD © , © BO Py © oO G0) 540 25 no vats 25. TO 2 esa 55h 1250 aS EO) i OO) 9 59 54 29 7 a » 16) 62g 27 9 y 6B 8D) 1p a SS EE FD © 62 55 27° «9 6 64 48 26 8 OS a 27 Til Gn be noe) 826889 5 8 HD BS © Se @) ki) BA 1@ 9 56 42 28 9 3 57 59 25 10 9 59 55 2 9g $ 63 56 2 » 62 56 2% 12 2) 67 57 830F= 10 4. Number of plates in collar. oo oO et | 5 6. Us 26 «620-21 #41 20 19 1 22 18-19 1 22 18 1 23 rts 27 19 1 25 17-19 O 27 Wy 1 26 16-17 0O 26 «21-20 1 2b 16-18; 2 23 Wi 1 20 18-19 0 20 16 1 21 18-20 1-2 23 «621-20 «21 23 «621-20 =I 20 20 2 25 20-21 0 24 «621-18. 24 19-20 1 24 21-20 1 27 = 21-20 -—O 20 18-20 1-2 22 19 1 24 =20-19 2 27 20 2 22 18-16 J] 22 13 J 30 19-20 2 27 = «18-202 28 19-20 1-2 26 21 1 28 21 1 2. Number of scales across middle of body. 5. Number of scales and 6. Number of femoral pores 7. Number of supraciliary shields in contact with second For detailed descriptions applying exclusively to the typical form, ¢f. Fatio, Faune des Vert. de la Suisse, iii. p. 92 (Switzerland); Lataste, Faune Herp. de la Gironde, p. 71 (Gironde); Bedriaga, Arch. f. Nat. 1878, p. 271 (W. Germany, Paris, Riviera, Tyrol). 358 MR. G. A. BOULENGER ON THE VARIETIES OF Specimens from the Pyrenees at or near Kaux-Bonnes (700-1100 metres), of which I have examined over 50, are in no way distinguishable from the Wall-Lizard of the plains of France. But specimens from a greater altitude (1545 m.) in the neighbourhood of the Pic du Midi have been alluded to by Bedriaga (Arch. f. Nat. 1879, p. 290) as remarkable for their strongly depressed body and coarse dorsal scaling, and their general appearance when in the open had suggested Lacerta vivipara to their collector, F. Lataste. The dorsal scales are described as very flat, but with a more or less distinct keel; and, although the caudal scales are only feebly keeled, the whorls project on the sides to form a very distinct serration. ‘The masseteric disk is usually broken up into 5 or 6 large scales. Dark ocelli are usually present above the axil. From a brief allusion again made to these lizards by Bedriaga in his monograph Abh. Senck. Ges. xiv. 1886, p. 196), Méhely (Ann. Mus. Hung. ii. 1904, p. 371) is inclined to suspect the existence of a new species in the Pyrenees. As there is fortunately one of these specimens in the Lataste Collection, I can confidently say that a specific separation from L. muralis is out of question, and I append a few particulars of this unique example, a female, which is figured on Pl. XXIV. fig. 5 and Pl. XXYV. fig. 6. Head rather short and but moderately depressed, its greatest depth equalling the distance between the anterior border of the eye and the tympanum. Hind limb not quite reaching the wrist of the adpressed fore limb ; foot not longer than the head. Frontal shield a little shorter than its distance from the end of the snout!; first supraciliary in contact with the second supraocular; occipital shorter, but a little broader, than the interparietal ; nasal reaching or nearly reaching the anterior loreal ; four anterior upper labials; masseteric disk very small on one side, indistinct on the other. : Only 19 gular scales and granules on the median line? ; collar with entire edge, formed of 9 plates. Dorsal scales granular, flat, rounded, faintly keeled, larger on the back than on the sides, where 3 correspond to a ventral plate; 50 scales across the middle of the body ; 32 transverse series correspond to the length of the head. 28 transverse series of ventral plates. Scales on upper surface of tibia feebly keeled, much smaller than on the back (8 rows of these scales corresponding to 5 dorsals). Femoral pores 17-16. 24 lamellar scales under the fourth toe. Caudal scales truncate and rather feebly keeled, the whorls very distinctly longer and shorter alternately and detached at the sides, forming a strongly serrated outline ; 24 scales in the fourth whorl. Brown above, with two white, black-edged streaks on each side, and a vertebral series * Quite in the same proportion as in Méhely’s Lacertu horvathi. * Just below the minimum recorded in the typical form from other localities. LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. 309 of partly confluent black spots ; two black ocelli with white centres above the shoulder ; only very small spots on the sides of the tail; belly greyish white, with small black spots on the sides. millim, IN GAO WO VEMh 6 5 o 0 o o 0) 55 9 foresimby seeds eck lemeNO@Elnert! 5 5 6 6 5» o a Ale Niecliin@ilec 5 56 o o@ 6 o 5» oo} 0) IDndn@r iene! 5 o o 6 6 » 6 & 8 Fore limb Me ads as BOO) Jabpecllbim) Gs % 5b o oo waldo vo. Hex POG ches pe a ok ea PL Maile (reproduced)iae emer eLOS In the same part of the Pyrenees, but at a much lower altitude, I captured in June 1878, on a wall in the town of Bagnéres de Bigorre, a rather remarkable specimen, preserved in the Brussels Museum, which my colleague M. Dollo has had the kindness to send me for re-examination. It is a female of large size (72 millim. from snout to vent), which attracted my attention by the presence of a large black ocellus with blue centre, above the axil, well displayed as the lizard sunned itself on the old wall. Head once and two-fifths as long as broad, much depressed !, its depth equal to the distance between the posterior border of the eye and the tympanum. Hind limb reaching the elbow of the adpressed fore limb ; foot once and one-fourth as long as the head. Frontal shield nearly as long as its distance from the end of the snout; series of supraciliary granules complete ; occipital much smaller than the interparietal ; masseteric disk rather small. 21 gular scales and granules on the median line; collar with entire edge, formed of 10 plates. Dorsal scales granular, flat, smooth, faintly keeled towards the tail, a little larger on the back than on the sides, where 3 correspond to a ventral plate ; 55 scales across the middle of the body ; 28 transverse series correspond to the length of the head. 29 transverse series of ventral plates. Scales on upper surface of tibia distinctly keeled, smaller than dorsals. Femoral pores 17 on each side. 26 lamellar scales under the fourth toe. Caudal scales obtusely keeled, very obtusely pointed behind; the fourth whorl behind the postanal granules contains 36 scales. Colour grey-brown above, with a dark brown lateral band bordered above and below by a white streak ; a few biackish dots irregularly scattered on the back ; above the axil, the lateral band is interrupted by the large ocellar spot mentioned above; belly without spots. A side view of the head and anterior part of the body of this specimen (enlarged 24 times) is given on Pl. XXV. fig. 7. ' Quite as much as in a typical LZ. depressa, (Su) & VOL. XVU.—Part iv. No. 2.— October, 1905. Os (oF) S MR. G. A. BOULENGER ON THE VARIETIES OF Il.—SPANISH PENINSULA. (Plates XXIV. & XXV.) The typical form, in no respect distinguishable from that inhabiting Central Europe, certainly occurs in the northern parts. The British Museum contains specimens from Corunna (Seoane) and from Barcelona (Thomas and Pocock). Specimens from Arnao (Bedriaga) are in the Lataste Collection 7. The belly of the adult males is spotted with black, and the masseteric plate is well developed, except in one specimen from Barcelona, in which it is altogether absent. Larger, more handsomely marked specimens occur on the islet La Deva, opposite Arnao, and have been described by Bedriaga as Var. Rasquinetil, Arch. f. Naturg. 1878, p. 260, pl. x. figs. 1 & 2. Dark olive-green to dark brown above ; sides of body sky-blue with a black network ; a series of black spots forms an interrupted stripe along the middle of the back ; large black spots on the upper surface of the tail; lower parts bright red, much spotted with black ; the whole of the outer row of ventral shields and the outer half of the adjacent ventral shields sky-blue. The type specimen, male, of which the life-coloration has been described and figured by Bedriaga, is now preserved in the British Museum. Dimensions (in millimetres) and scaling are as follows :— ROSIN SHO WO WOME 5 > ba 6 10 & eo 40,6 0. 00 7 Fore samy je yell ops Mees (a aia et meet RSS hen'etheofneads iy ung Neale -ole suis ar ae Oma nn asc Heart Momus, Wadthtotehead ies 2 taps crach cn Meare wera, mane heats ine: IDendn Hee 6 6 6 6 4 6.6 6 ova 6 g o 9 JO 'Roresinibieie 28 yet ee ee) ee ene cu ticu sy cymers, cote mmean2 As Hind lum bie ee GO a By Ne ey ee SOG TROYES SAS, out oh, Eman anand Alomar ciated he toa ty alle) Atel Geqocheenl ae So 6 6 a 6 5 5 6 a ol Le iINumberjofrscaleskacrossibodyiimaur) | -neeeeetn nace) OS os 5, corresponding to a ventral plate . . 8-4 es >) from chin-shields\to collar’. 2 = 3 5 25 i wees marcel 5 6 5 6 6 6 4.0.0 6 9 5 transverse rows of ventral plates. . . . . 26 a Hae WOES ¢ 5 6 @ o 6 6 6 0 4 o JEANS Dorsal scales very distinctly, rather strongly keeled, the keels disappearing on the sides of the body; 38 transverse series correspond to the length of the head. Edge of ‘ Specimens from Arnao have been described from life by Bedriaga, Arch. f. Naturg. 1878, p. 267. LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. 561 collar yery slightly serrated. Temporal scales rather large, with quite distinct tympanic and masseteric shields, the latter separated from the parietal by two series of scales. The rostral shield very nearly enters the nostril. 25 scales under the fourth toe. In my opinion, Rasquinet’s lizard does not constitute a definable variety, any more than the lizards on the Glenan Isles in Brittany, and I regard it as almost certain that were any large number of specimens collected some would prove to be absolutely unseparable from the typical form as occurring on the mainland of the Asturias. In North-western and Central Spain and in Portugal there occurs a distinct form which has been insufficiently defined by Seoane as Var. BOCAGEI, in a pamphlet issued in 1885 (antedated 1884) under the title of “ _Identidad de Lacerta Schreiberi (Bedriaga) y Lacerta viridis, var. Gadovii (Boulenger) i investigaciones herpetoldgicas de Galicia.” “ Lacerta muralis, Laur., var. Bocagei, nov. var. ComunisimoentodaGalicia. Esta forma, intermedia entre muralis fusca y muralis pytiusensis, se diferencia de la primera per el dorso que es de color verde, y de la segunda por que los flancos son pardos, ademas de los cardcteres que espondré al describirla.” I here give details of specimens from Galicia, received from the late V. L. Seoane and preserved in the British Museum and in the Lataste Collection, from Madrid, received from the Madrid Museum, and from the Serra de Gerez, collected by Dr. Gadow. Specimens from Pontevedra (Bosca) are in the Lataste Collection. ile 2. 3. 4. 3. 6. omeGaliciay es. 57 ~=6«GO——s«éYd 9 29 16-15 e * 45 54: 27 11 23 18-17 ae 429 55) 1 2A LON 24a Les Ont 35 60 o7 29 8 20 16 ile Ree ee sis) GS) op Wa) by ae ©. Corrinne on os eames pop bho: oct! 58 32 11 25 15 6. Escorial, Madrid 56 65 26 1! 30 19 i 5 _ 56 62 27 11 27 ~=19-20 oF: sp 3 59 56 29 10 29 17-16 6. S. de Gerez 55 58 26 9 26 17 2. be 55 5d. 26 9 28 19-18 = oe 55 57 23 11 26 17 a iy SoM MmCCRMECSIn On AST ilo 1. Length (in millim.) from snout to vent. 2. Number of scales across middle of body. 3. Transverse series of ventral plates. 4. Number of plates in collar. 5. Number of scales and granules between symphysis of chin-shiclds and median collar-plate. 6. Number of femoral pores. 3B2 362 MR. G. A. BOULENGER ON THE VARIETIES OF The dorsal scales are sometimes distinctly keeled, sometimes quite smooth; 36 to 50 transverse series correspond to the length of the head; the ventrals form 6 longitudinal series. 28 to 34 caudal scales in the 4th or dth whorl behind the postanal granules, those on the upper surface strongly keeled. 21 to 29 scales under the fourth toe. The head (Pl. XXYV. figs. 8 & 9) is often rather more depressed than in the ‘Central-European specimens, and may remind one of L. bedriage from Corsica; its depth at the occiput equals in some specimens the distance between the tympanum and the posterior border of the eye, in others that between the tympanum and the centre of the eye. Out of 5 specimens from the Serra de Gerez, 2 have the parietal excluded from contact with the postocular!?. The masseteric disk is often very small or absent. The markings vary much, but there is never any indication of a dark vertebral stripe; the light dorso-lateral streaks, if present, are closer together, or at any rate not wider apart, on the body than on the nape ; the sides, and often the whole of the body, may show a black network enclosing small roundish pale green spots; the lower parts are uniform, or but scantily spotted with black. Young specimens from Oporto (Gadow) and Cintra (Yerbury) may be referred to the same variety. Two specimens, male and female, from Galicia, collected by M. Seoane, are figured, Pl. XXIV. figs. 6 & 7, and the anterior part of the body of a spotted male from the Serra de Gerez, collected by Dr. Gadow, is represented on Pl. XXIV. fig. 8. Specimens from Lisbon agree with those from Galicia, except that the caudal scales are less strongly keeled. The scales are smooth or nearly so, and the masseteric disk is small or absent. i 2. 3. 4, 5 6. Oe dus =o 6 6 5 6 6 5 5 iS) 60 26 9 29 21 ¥ ” 48 60 30 9 23 16-18 ? ” 46 55 30 10 29 19-18 Boettger (Sitzb. Akad. Berl. 1887, p. 155) also finds the Lisbon and Gerez specimens very similar, and describes the ground-colour as of a greenish grey. Specimens from Coimbra have been described by Bedriaga in 1890 (‘ Amphibiens et Reptiles recueillis en Portugal par M. A. F. Moller,’ p. 52). The femoral pores are stated to vary between 15 and 19. Unlike the Portuguese specimens examined by me, the tail does not equal two-thirds of the total, and the scales on the upper surface of the leg may be as large as those on the body. * A young lizard from Ciudad Real (Bosca) in the Lataste Collection, which agrees well with the Serra de Gerez specimens, also shows this character. LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA, 363 Measurements of specimens from N.W. Spain and Portugal :— 1. 2s 3. 4, 5. 6. From snout to vent 57 55 55 55 48 46 3 2) forelimb ion saat 19 21 21 18 17 Weng throlib ead) igs. uracil 12 14. 13 Il 10 Wikia @mincel Soe 56 2 eo 6 o LO 8 9 9 8 i IDeA OEE GB 56 6 6 s 4 6 6 8 i/ 7 6 64 54 INST eo 6, o- bo oid bo. y cv) W/ 20 18 16 14 Handclimb: <.2¢ 44. A eee cen oO) 24, 30 27 26 21 Dail oc vy ong. Soy OME Hd OO 92 103 93 102 60 l. g. Galicia (Seoane—Lataste Coll.). 2. 2. ” 2» 83. ¢. Serra de Gerez (Gadow). Ate . Bi 5. g. Lisbon (H. O. Forbes). Gs Os 5 a Tail reproduced. Var. LIOLEPIS. Bedriaga (Arch. f. Naturg. 1879, p. 292) has briefly alluded to a small form which he observed alive at Valencia, on the road to the Dehesa de la Albufera, and which struck him as remarkable for the extremely small dorsal scales and smooth caudal scales, the middle dorsals of which are more or less enlarged. ‘The colour is described as a pale yellowish brown above, with a brown, white-edged lateral band, greyish or yellowish beneath. Length from snout to vent 49 millimetres. Among numerous specimens from Valencia in the British Museum (Lilford, Bosca, Bolivar) some agree with Bedriaga’s description in having the upper caudal scales very faintly keeled, almost smooth, and the granular dorsal scales are very minute, scarcely larger than those on the upper surface of the tibia. Some specimens, however, have the dorsal scales larger, and the caudal scales distinctly, though rather feebly, keeled; I do not think it possible to sharply define this form, which may be said to connect the typical form with the var. hispanica, and also with the var. pityusenis, I have received specimens of the same form from Seville (Prof. Calderon). The scales across the body vary from 52 to 69; 3 or 4 correspond to one ventral plate, and 42 to 56 to the length of the head; they are smooth or faintly keeled on the back ; ventral plates in 6 longitudinal rows, except in one 2, which has 8; collar- plates usually smaller than in the typical form, 9 to 14 in number; 25 to 39 gular scales and granules on the median line. Upper caudal scales feebly or faintly keeled, sometimes nearly smooth; 28 to 34 scales in the 4th or 5th whorl behind the postanal granules. Femoral pores 15 to 22 on eachside. 23 to 27 scales under the fourth toe. Nearly constantly four anterior upper labials!. Masseteric disk more or less * One specimen has five on one side. 364 MR. G. A. BOULENGER ON THE VARIETIES OF developed (separated from the parietal by 1 to 4 series of scales) in 10 specimens out of 30, this plate in the others being just distinguishable or absent. In one specimen from Seville the parietal is not in contact with the postocular. No specimen exceeds a length of 60 millim. from snout to vent, and females of 44 millim. are gravid. In their markings the very young (see text-fig. 4 c, p. 369) are quite similar to those of the typical form from France, and a black vertebral streak (which I have never observed in any of the specimens from North-Western and Central Spain and Portugal) is likewise often present. The adults are pale grey or brown, and the blackish spots, if present, are small; a median series of black spots or traces of a blackish vertebral line are often present ; ventral spots are absent or small and confined to the sides. One male specimen from Valencia is exceptional in having the lower parts dotted all over with black. I append particulars of 12 specimens from Valencia, and of 2 from Seville. iL, 2 2 A.B, 6. Valencian, noted Gace sits 60 57 27 10 33 18 Pf ¥ GO. BS M9 1D BS WAG i 1 BS GD PE a BA 20 s il WEBS BS OR TO) 9S 16 i ih BD RY ye ey es) =p i i BU BOY OS) Ml BR TES , by 498 G4 723i Ome ne 7 18 9 i BO 17 50 60 ol Wit 30 16-17 50 62 30 Bl 33 17-18 6 + 46 53 28 Il 28 16 ee a Mee at! oe ASG Ra Qcen Groene soz ml oen0 Sevalllenty ey Peete san i | aes 52 62 28 11 26 19-20 Sch AO sete cme teeeneae ts | SD 50 57 29 iil 28 19 Measurements of largest male and female from Valencia :— 3. g MromysnOutstOnvenitay.s les) i mitcn wnt memo. 54. 3 fp atone lbwm) 96 GG 5 kl gE 19 Ibe One 6 6g ol 6 & 6 1s 16 12 Wyachla @e inert saa le! go 6 as 6 6 ) HO 8 Wenn OE INN) 5 5) gn 8 6 Ge 8 6 Io NM) 5 6 6 6 & 6 8 6 6 6 6 & 17 Find glib yey) ai ak ie We ip es eee oO 26 HOOtRES Br roy ce at ox, tee: us, ope Tam) oe all 18 Tail pape yee damn cioke:, ley npc edioe cade, WebeamalleO) 60 * * Reproduced. Male and female specimens from Valencia are figured on Pl. XXIV. figs. 9 & 10, and Pl. XXYV. fig. 10. LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. 365 Var. VAUCHERI. A single male specimen of a lizard agreeing in most respects with the Tangier variety described further on as var. vaucheri was collected at Salir, Algarve, by Dr. Gadow, who has presented it to the British Museum. The dorsal scales are feebly keeled and not larger than those on the upper surface of the tibia; they number 76 across the middle of the body, 3 and 4 transverse series correspond to the length of one ventral plate, and 57 to the length of the head. The ventral plates form 27 transverse series, in most of which the outer plate is divided, so that the longitudinal rows may be described as 8 in number. 10 plates in the collar, and 31 gular scales and granules in a line in the middle. Femoral pores 18-17; 28 lamellar scales under the fourth toe. 30 scales in the fourth whorl behind the postanal granules. The specimen is anomalous in having only three anterior labials on the right side ; the masseteric shield is small and divided into two or three pieces. The back is greyish brown with a black network, the sides are black with round light spots; small black spots are present on the ventral shields. millim, Hromend of snout to vente =. 96 2 4 5 4 47 ss 3 my won lr 6 5 5 5 5 NO Ibeacin Orne!’ 6 6 6 6 6 of 8 a oto MD Wadthtof heads, var ln te veem nee ete 8 Depth of head. 6 Fore limb RATS SET GE WY omega BLO Hinds limbic ie es nee ee ee eG Foot . : 15 Tail (reproduced) . 73 Compared with the Tangier specimens on which this variety is primarily founded 1, the lizard here described differs in its still finer lepidosis and in the division of the outer ventral plates—characters which probably could not be maintained as distinctive should a series of specimens be available. Var. MONTICOLA. A single female specimen collected by Dr. Gadow in the Serra Estrella, Portugal, at an altitude of about 6000 feet (Pl. XXIV. fig. 11), first struck me by its general resemblance, in form and markings, to the Caucasian vars. chalybdea and saxicola. It has also much in common with the newly described ZL. horvathi from §.W. Hungary. Among the Lizards of the Madrid Museum, which Prof. Bolivar has kindly sent to me 1 Below, p. 418, Pl. XXIX. fig. 9. 366 MR. G. A. BOULENGER ON THE VARIETIES OF for study, I was pleased to find two further specimens of this form, both females—one marked “ Galicia” (Pl. XXIV. fig. 12), the other from an unknown Spanish locality (Pl. XXIV. fig. 15). A young specimen labelled “ Spain (Bosca) ” is preserved in the Lataste Collection (see text-fig. 4d, p. 869). The head (Pl. XXV. figs. 11 & 12) is strongly depressed, and the supratemporals are well visible from above; the parietal does not touch, or scarcely touches, the upper postocular, and its antero-lateral border is slightly concave !; in one specimen the nasal Text-fig. 3. a. b. @ Upper and side views of heads of :—a. Lacerta horvatht ; b. L. muralis, var. monticola ; c. L. muralis, var. depressa. forms a suture with the loreal?; four anterior upper labials; masseteric disk large ; rostral forming a suture with the frontonasal ; frontal as long as or a little shorter than its distance from the end of the snout; supraocular granules forming a complete series. 1 Exactly as figured for Z. mosoriensis by Méhely, Ann. Mus. Hung. ii. p. 366, fig. 2, A.—‘ Diese Eigenschaft, die ich fiir einen blutsverwandtschaftlichen Character einer mit Lacerta savicola Kversm. im Kaukasus beginnenden und mit Lacerta horvathi abschliessenden, yon der Entwickelungsreihe der Lacerta muralis grundverschiedenen phyletischen Reihe betrachte.”—Méhely, 0. ¢. * As in ZL, horvathi, Méhely, t. c. p. 367, fig. 3, C. LACERTA MURALIS IN. WESTERN EUROPE AND NORTH AFRICA. 367 22 to 25 transverse series of gular scales; collar entire, with 9 or 10 plates. Dorsal scales (Pl. XXY. fig. 15) large and flat, smooth or faintly keeled, larger than those on the sides, which are larger than those on the upper surface of the tibia, 5 or 6 of the former corresponding to 8 of the latter; 46 to 50 scales across the middle of the body, 27 to 34 transverse series corresponding to the length of the head!. Ventral plates in 28 or 29 transverse series, some of which may contain 8 plates. Femoral pores 17 to 20. 25 or 24 lamellar scales under the fourth toe. 28 to 30 caudal scales in the front whorl behind the postanal granules, the upper truncate and very feebly keeled; the whorls very markedly longer and shorter alternately. The hind limb reaches the wrist or the elbow of the adpressed fore limb. The following are the particulars of the three female specimens :— 1 2. 3. 4, 5. 6. Spain . : 76 50 28 10 25 18-17 Serra Estrella 4 ee RY BG 47 29 9 25 19-20 Galiciagay-s |. isd. 3 Ge! i ee G2 46 28 9 22 Wi Greyish or pale olive above (in spirit), with large black spots forming two series along the middle of the back; sides darker, bordered above by a series of black spots or by a black wavy streak enclosing light round spots; a few dark ocelli with light centres may be present above the shoulder and axilla; lower parts whitish or greyish, with or without scattered black dots on the belly and under the hind limbs. Measurements (in millimetres) :— inom Grom WOVE 6 5 5 0 5 5 6 OO 62 2 i i Hola). 5 6 6 co o By 22 24: lbesn@t Ingles o ¢ 6 56 6 0 op o, LG 15 15 Wadthvotiheady .) “ak alk wales, cca coll 10 10 Depth of head . taNi | aoie sayy 7 6 i Horevlimibi sy Laos an nod oe See ere eye oe. 20 21 angdulamib’, 6) siyt e ey ae ee ee 31 31 IBOObH Kee a at Os mah ren oe | gee en 7 7 17 Figures are given on page 566 of the upper aspect of the head of the largest female specimen from Spain, in one of the types of LZ. horvathi Méhely, and that of a * These scales are strikingly similar to those of Z. horvaihi, which, according to Méhely, number 40 to 46 across the middle of the body. In a specimen which Prof. y. Méhely has kindly sent me, the scales number 43 across the middle of the body, and 31 transverse series of dorsal scales correspond to the length of the head ; the Professor kindly informs me that the statement made by him that 10 to 12 transverse series of dorsal scales correspond to the length of one centimetre is the result of a lapsus, 18 to 22 being the right nnmber. The scales are represented as much too large in the figure of Z. horvathi given in the Hungarian journal ‘ Allattani Kozlemények,’ iii. 1904, pl. v. VOL. XViI.—ParT 1V. No. 5.— October, 1905. Bio 368 MR. G. A. BOULENGER ON THE VARIETIES OF male from Trebizond, one of the types of LZ. depressa Camerano!, in order to show the striking resemblance in the form of the head-shields in these geographically remote varieties. A very distinct variety, which is, however, connected with the var. liolepis, is that first described by Dr. Steindachner as a form of Lacerta oxycephala, and for which the name hispanica, proposed by him, should be retained. Its synonymy is as follows :— Var. HISPANICA. Lacerta oxycephala, var. hispanica Steindachner, Sitzb, Ak. Wien, Ixii. 1870, 1, p. 336, pl. 1. figs. 3-6; Boettger, Abh. Senck. Ges. xii. 1881, p. 376. Lacerta oxycephala, part., Schreiber, Herp. Eur. p. 404 (1875). Lacerta muralis fusca, part., Bedriaga, Arch. f. Nat. 1879, p. 293. Lacerta muralis, subsp. steindachneri Bedriaga, Abh. Senck. Ges. xiv. 1886, p. 256. Lacerta muralis, var. hispanica Boulenger, Cat. Liz. ii. p. 33 (1887). This form, which in its sharply pointed snout and conspicuously striated body reminds one of a young Acanthodactylus, has been very well figured by Steindachner and fully described by Bedriaga. Its characters have been summarised by me as follows :— Head and body much depressed; snout moderately long, acutely pointed. Fre- quently five upper labials anterior to the subocular ; no masseteric shield. Collar-plates very small; gular fold scarcely indicated. Scales small, flat, smooth, roundish, 50 to 60 across the middle of the body. Grey above, with five or seven longitudinal blackish streaks usually bearing small irregular light spots; the lateral streak, passing through the eye, broadest, the vertebral narrowest; in the young, the lateral black streaks separated by pure white ones; lower surfaces uniform whitish. I have nothing to modify in this definition, as the British Museum has only added one to the two small specimens mentioned in the Catalogue. The femoral pores number 15 to 19 (14-18, Bedriaga). ‘The series of supraocular granules is complete or the first supraciliary is in contact with the second supraocular ; one specimen has 6 anterior upper labials on both sides, a second has 5 on one side and 4 on the other, whilst a third has 4 on both sides ; the parietal does not touch, or barely touches, the upper postocular ; gular scales in 26 to 29 transverse series. Upper caudal scales feebly keeled. 25 or 26 scales under the fourth toe. The hind limb reaches the axilla or the shoulder. This lizard does not appear to exceed a length of 50 millim. from snout to vent. Figured on Pl. XXIV. fig. 14 and Pl. XXYV. figs. 14 & 15. Habitat. S.E. Spain (Alicante, Murcia, Cartagena, Almeria). ? Of. Boulenger, P. Z. 8. 1904, ii. p. 332, pl. xxii. fig. a. LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. 369. Text-fig. 4. ee B Bt % HY 2 % f Young specimens of :—a, f. typica, Arnao, Asturias ; 6, var. bocaget, Madrid ; ¢, var. liolepis, Valencia ; d, var. monticola, Spain ; e, var. hispanica, Almeria. III. BALEARIC ISLANDS. (Plates XXIII. & XXVI.) A good example of the neglect of the lepidosis in defining races of the Wall-Lizard, of which I have complained above, is offered by the lizards inhabiting the Balearic Islands. Although the form and coloration have been described by Bedriaga and by Bosca, the difference in scaling by which the Iviza lizard can be sharply separated from the so-called varieties inhabiting the two more eastward islands has hitherto entirely escaped attention. Yet, with the help of these characters, the Balearic lizards stand out as the best-definable of all the subspecies or varieties which have been included under Lacerta muralis, and certainly have stronger claims to specific separation than have Z. bedriagw and L. depressa. But, then, it is necessary to entirely put aside the colour—whether brown, yellow, green, or black—and to place together, as var. pityusensis Bosca, the Lviza specimens referred by Bosca to vars. balearica and pityusensis, and to group under the older name of yar. lilfordi Gthr. all the lizards from Majorca, Minorca, and neighbouring islands, which have been described by Bedriaga as vars. balearica, gigliolii, and lilfordi, the numbers of the scales counted across the middle of the body (595 to 68 in the former, 70 to 90 in the latter) being sufficient for their discrimination. ‘The important fact that in the Iviza lizards the rostral shield constantly enters the nostril, as only exceptionally happens in other forms of L. muralis, affords a diagnostic character by which to separate them from those of Spain, which agree in the number of scales across the body. ac 2 370 MR. G. A. BOULENGER ON THE VARIETIES OF Since this paper was written Hr. Lorenz Miiller has reported (Zool. Anz. 1908, p. 502) the occurrence of the var. cettii (= tiliquerta) at Binisaida, S.E. coast of Minorca, and he has kindly sent me one of his specimens, which agrees entirely with the Sardinian var. ¢iliguerta. Whether these lizards are really indigenous or were introduced by human agency is a question which Hr. Miiller himself has suggested for consideration, but which cannot at present be solved. Var. PITYUSENSIS. Lacerta muralis, var. balearica Boscdé, An. Soc. Espan. H. N. xii. 1883, p. 245. Lacerta muralis, var. pityusensis Bosca, |. c. p. 246. Lacerta muralis balearica, var. pityusensis Bedriaga, Abh. Senck. Ges. xiv. 1886, p. 254. Head (Pl. XXVI. fig. 1 a) once and a half to once and two-thirds as long as broad, moderately depressed, its depth in the tympanic region equalling the distance between the anterior border or the centre of the eye and the anterior border of the tympanum ; snout rather pointed. Neck as broad as or a little broader than the head. Body mode- rately depressed. Hind limb reaching the shoulder or the collar in the males, the shoulder, the axil, or the elbow of the adpressed fore limb in the females; foot as long as or a little longer than the head. Tail once and a half to a little over twice the length of head and body, rounded in section, cyclotetragonal at the base. Rostral shield entering the nostril, often largely ; nasals forming a suture behind the rostral, separated from the anterior loreal by the postnasal, which is usually single '. Frontal as long as or shorter than its distance from the end of the snout; a series of granules between the supraciliaries and the principal supraoculars”, the first or first and second of which are usually in contact with the first supraciliary. Occipital usually as long as and broader than the interparietal, sometimes much smaller. Parietals once and one-third to once and two-thirds as long as broad, in contact with the upper post- ccular. Upper temporal shield entirely absent or broken up into a series of small shields. ‘Temple covered with granular scales ; tympanic and masseteric shields distinct, the latter separated from the parietal by from one to three series of scales. Four upper labials anterior to the subocular °, 26 to 34 scales and granules on a line between the symphysis of the anterior chin- shields and the median collar-plate; gular fold distinct; collar even-edged, formed usually of rather small plates, 9 to 14 in number. 1 Two postnasals on the left side in two of the specimens in the British Museum, two on both sides in a specimen in the Madrid Museum. 2 These granules are reduced to a few in one of the specimens in the British Museum and in four in the Madrid Museum. ° Five on one side in two specimens, on both sides in one. LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. 371 Dorsal scales (Pl. XX VI. fig. 14) granular, roundish-hexagonal, more or less feebly but always distinctly keeled, 55 to 68 across the middle of the body; 3 and 4 trans- verse series (rarely 4 throughout) correspond to one ventral plate, 35 to 47 to the length of the head. Ventral plates in 6 longitudinal and from 26 to 30 transverse series. Anal plate bordered by one or two semicircles of small plates. Seales on upper surface of tibia finely granular, much smaller than dorsals, more or less distinctly keeled. 19 to 24 pores under each thigh. 26 to 31 lamellar scales under the fourth toe. Upper caudal scales strongly keeled, truncate or very obtusely pointed behind; the whorls subequal in length, the fourth behind the postanal granules containing 30 to 38 scales. Brown, yellow, or green above, with black or reddish-brown spots or vermiculations, usually arranged in five or seven more or less well-defined longitudinal stripes; the black markings sometimes forming chains enclosing roundish light spots, or an irregular reticulate pattern, or reduced to mere dots or vermiculations ; females often with a light dorso-lateral streak, beginning from the supraciliary edge, sometimes bordered by fine black lines, sometimes with a black vertebral streak ; upper surface of head spotted or vermiculated with black; tail with dark and light markings, forming more or less regular longitudinal series. Young with six light streaks on the body. Belly yellow, orange, or red, without markings, except on the outer row of ventral plates, which bears large blue spots. : Measurements (in millimetres) of two of the type specimens in the Lataste Collection :— OF MGA FAODIn OWN Bf o o 6 6 5b 0 6 c 72 64: - 3 tone dims) 95 6 boo b 6 86 30 23 Bene thyotheadan ecm isin.) alanine: 18 15 Width of head POM On bie or hs ect ee 13 9 Wein OF eAGL 5 6 5 6 5 6 0 0 0 6 8 Wil 7 Roreplimbi, 3.0 <2 vayecc, co es Goa ORE 28 19 Jalal bhalos Nmee emer eon ic. ol Bin me noe sous c 40 32 HOOTERS chs, \c. uksih ion) SUR ueh een: | cameePEnrere 21 17 Maleate 5 ee a tr ooo eer sees cme Al 85 * * Tail reproduced. I have examined 79 specimens, all collected by Sr. E. Bosca on Iviza, three of which belong to the Lataste Collection, the rest being in the British and Madrid Museums. eo “I bo MR. G. A. BOULENGER ON THE VARIETIES OF I append particulars as to size and scaling of some of them :— 1a ORI nS Phel ial OEE e 6. 3 EO CO oy aay) TIP) 1830) 21 * 75 Olin 127 OUN Wel mesEon 4 735 (UG 7ie 27) wlan MAS Au noe 3 7 7201620) 228 8) GA P 7 80) Of 1D BB. seeoe ss WO EY OR TOG 20 " Gf G)) 98 1 3 120 - B C9 WP cil i 21 By GR 93 1O 8) 23 Mi GQ eNGS eo NET E32, 23 . a Gh 3 Th ob geo © aly alae Heedeetted an nts Onan CO ueRee TSO. 9 26 19 sid drte Lage ae tie ad OA Cag OS geNe29 Oh) | OLE sola baci Rete ae EG Olea 7amer vO aN ied TS 3 eRe SOD elie tie son eesg Soules Sian 7; abe COMn pes OR Memo See 20 =2 1 5 LORE eae oD ban MOG UM aro Semnl OL eau STa Mos =20 1. Length (in millim.) from snout to vent. 2. Number of scales across middle of body. 3. Transverse series of ventral plates. 4. Number of plates in collar. 5. Number of scales and granules between symphysis of chin-shields and median collar-plates. 6. Number of femoral pores (on right and left sides, if differing). \ This form was found by Bosca on Iviza and on the small neighbouring islands La Grosa, La Redonia, and El Malvi. Several specimens are figured on Pl. XXVI. figs. 1-5. Var. LILFORDI. Zootoca lilfordi Giinther, Ann. & Mag. N. H. (4) xiv. 1874, p. 158. Lacerta lilfordi Braun, Arb. Zool. Inst. Wiirzb. iv. 1877, p. 4. Lacerta muralis Braun, l. c. p. 26. ; Lacerta muralis fusca Bedriaga, Arch. f. Naturg. 1878, p. 275. Lacerta muralis, var. gigliolit Bedriaga, op. cit. 1879, p. 247, pl. xvii. fig. 2. Lacerta muralis, vars. balearica, giglolii, lilfordi Bedriaga, Bull. Soc. Zool. France, 1879, p. 221, pl. ix. figs. 1-3. Lacerta muralis balearica, part., Bedriaga, Abh. Senck. Ges. xiv. 1886, p. 248. Variations in colour in this lizard are solely responsible for the several names in the above given synonymy. ‘The black colour of the specimens from the Isla del Ayre, near Minorca, was the sole ground on which Gunther established his Zootoca lilfordi, and he showed his neglect of structural characters by associating with it the black lizard of the Filfola rock near Malta. The lizard of Isla del Ayre differs, 1 repeat, in no other LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. 373 respect but the ground-colour from the brown or green lizards of Minorca and Majorca, the markings of which may be detected in certain lights in some adult black specimens, and which are distinctly shown by all young specimens. Brown and olive- green specimens from Minorea were first referred by Bedriaga to his subspecies fusca, and afterwards to a distinct variety named dalearica; whilst specimens with a blue or blue-green back, from the Dragoneras Island, near Majorca, were described as another variety named after Prof. Giglioli of Florence. Having recently received from Sefior J. Ferrer, of Mahon, numerous living specimens (some brown, some green, some bluish green on the back, and the green ones having after a time become brown '), I cannot regard the coloration as affording any justification for the establishment of varieties; and as the structural characters are the same in these so-called varieties, | have no hesitation in throwing them all together under the name of var. lidfordi Giinther, which has priority over that of balearica. The form on which L. lilfordi was established I regard as a mere case of melanism, fixed on certain islands, through influences which are as yet unexplained; but there would, in my opinion, be no more reason for allowing it the rank of variety, in the sense in which I take the term, than in the case of the black individuals of Vipera berus and V. aspis or of Lacerta vivipara. Besides, if all the individuals on Ayre Island and other islets in the Balearic are affected with melanism, the name of the variety coupled with an indication of the locality is amply sufficient to avoid confusion with any other form. On the other hand, the Iviza lizard, which has been treated by previous authors as a race equivalent to the vars. balearica, gigliolii, and lilfordi, can be well defined and constitutes a quite distinct form. Having explained my reasons for throwing together the so-called varieties mentioned in the synonymy, I will proceed to a description of the material before me, consisting of 32 specimens from the following localities: —Mahon (Braun in Lataste Coll., Bedriaga, Ferrer), I. del Rey, near Minorca (Braun in Lataste Coll., Ferrer), I. del Ayre, near Minorca (Lilford, Braun in Lataste Coll., Ferrer), I. Dragoneras, near Majorca (Bedriaga), I. la Guardia, near Majorca(Bosca). ‘The specimens from Is. del Ayre and la Guardia are black or blackish, and represent the L. lilfordi, sensu stricto. Head (Pl. XXVI. figs. 7a, 76) once and a half to once and two-thirds as long as broad, moderately depressed, its depth in the tympanic region equalling the distance between the anterior border or the centre of the eye and the anterior border of the tympanum ; snout usually decidedly pointed. Neck as broad as or broader than the head. Body moderately depressed. Hind limb reaching the shoulder, the collar, or a little beyond the latter in males, the axil or the shoulder in females; foot once to once and one-third the length of the head. ‘ail once and a half to once and four- fifths the length of head and body, rounded in section, cyclotetragonal at the base. 1 As happens in the males of our Sand-Lizard, Lacerta agilis. O74 MR. G. A. BOULENGER ON THE VARIETIES OF Rostral shield usually narrowly separated from the nostril, but sometimes entering it; nasals forming a suture behind the rostral, separated from the anterior loreal by the postnasal, which is single. Frontal as long as or shorter than its distance from the end of the snout ; a series of granules between the supraciliaries and the principal supra- oculars, the first of which is often in contact with the first supraciliary. Occipital very variable in size, often longer and broader than the interparietal!. Parietals once and one-third to once and two-thirds as long as broad, in contact with the upper postocular, Upper temporal shield entirely absent or broken up into a series of small shields. ‘Temple covered with granular scales ; tympanic and masseteric shields usually distinct °, the latter separated from the parietal by two to four series of scales. Four upper labials anterior to the subocular. 30 to 40 scales and granules on a line between the symphysis of the anterior chin- shields and the median collar-plate; gular fold distinct; collar even-edged, formed usually of rather small plates, 9 to 15 in number. Dorsal scales granular, round or oval, smooth, 70 to 90 across the middle of the body (Pl. XXVI. fig. 7 c); 4 and 5 transverse series (rarely 3 and 4) correspond to one ventra} plate, 42 to 62 to the length of the head. Ventral plates in 6 longitudinal and 26 to 30 transverse series. Anal plate bordered by one or two semicircles of small plates. Scales on upper surface of tibia finely granular, smaller than dorsals, smooth; 17 to 25 pores under each thigh. 27 to 32 lamellar scales under the fourth toe. Upper caudal scales more or less strongly keeled, truncate or very obtusely pointed behind; the whorls subequal in length, the fourth behind the postanal granules containing 30 to 46 scales. The colour of the back varies from brown to olive, bright grass-green, or blue-green ; the sides are brown or reddish brown with paler brown or pale green spots. Three stripes of dark brown or blackish spots or vermiculations extend along the back, but it frequently happens that these stripes lighten in the centre and that their borders become converted into lines, such specimens having six fine dark lines along the back ; some specimens have dark marblings or vermiculations all over the back, and such is the case in the single very young specimen I have examined *. The lower parts vary from orange to salmon-red, copper-colour, or brick-red, without or with small black spots which may form regular longitudinal series; a series of turquoise-blue spots on each side of the belly; throat spotted or marbled with reddish brown or blackish. In the specimen from Dragoneras Island, Minorca, named by Bedriaga var. gigliolii, the back, at least posteriorly, and the upper surface of the tail are described as of a dark blue-green or peacock-blue, and the outer row of ventral shields lapis-blue. Preserved 1 Twice as long and twice as broad as the interparietal in a ¢ from Mahon. Hardly distinct in single specimens from Mahon, Dragoneras Island, and Ayre Island. ® From Mahon, in the Lataste Collection. LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. 375 in spirit, as they are now, such specimens cannot be distinguished from ordinary Minorca lizards. The Ayre Island specimens are described by Giinther (Zootoca lilfordi) as uniform deep shiny black above and of a beautiful sapphire-blue beneath. But some specimens from the same island sent to me alive by Sr. Ferrer are only of a very dark brown, with the dark markings (lines and spots) of the main island form perfectly distinguishable ; pale bluish-green or turquoise-blue spots on the sides of the lapis-blue belly, which may bear small black spots. Now that they have been in spirit for many years, Lord Lilford’s specimens also show very distinctly the dorsal markings. Some of the black specimens from La Guardia Island, near Majorca, collected by Sefior EK. Bosca, have numerous pale bluish-green spots on the sides of the body and on the limbs. The young of these melanotic insular forms are described by Braun as brown above, with dark spots and lines, dark brown or black on the sides, with numerous blue spots, the tail dark green; the belly is greenish in the middle, black spotted with blue on the sides !. Measurements (in millimetres) :— Mahon. Ayre Id. (types). (aN ———) 3 Q. J @, ION HOON WOVEN 2 6 6 o 5 o o A 62 64: 61 a i none lit) 2 5 4 6 o 5 BS 22 28 23 Iban OGG! 5 Gg 6 «2 0 o 6 o 5 US 14: 17 15 Widihofsheadii tg tpifen cers, se tumeaeer semen 9 12 g IDeyaln OE INE! Gg gg oo oh 9 7 9 7 Doral) sos so Go fb 6 a no 6 (28 22 20 23 ind slim by py caes chic 0). cen eae AO) 36 37 35 HOOT MR Mek ol. “itis, idyistu ny Cae Seema eae 19 1g 18 HUEVOS, Stairs Mme petty akin ues bys ty, heen b eal 0) 100 110 Gon * Tail reproduced. Particulars of some of the specimens examined (as in preceding table) :—— il. 2. 3. 4, iy, 6. Bo Melo BTS PSone, A) 78 27 15 37 = 24-25 o aa 16 ah at eo og 8 ND 90 28 iil 85 =. 25-22 D5 og = pen pero A)

) 212902 en tod Pare eee ame Te UmtaToy™ » Rioteh: Met 9 28 22-21 * ee wes to. ee pe ROC OM OOPme LO E23: S220) meas eee, ohne % SON GONeGN 9) 28 19 % Sestri Ponente,nr.Genoa .. ¢ 65 60 28 9 25 23 ] » ” ” et a's 60 62 250 10 20 23-22 10) 53 5 a 4 bg ae GON Pt aN pi Bleep al [Sb) io) [a) MR. G. A. BOULENGER ON THE VARIETIES OF ile Dy, 3. 4 by, 6 firs Rapallo a Oey BB) Be ORG 21 2 B Sie less te tel al leas USS Oss OOH 51/5 sO ib (0 ea aim eek Lan Portofino, nr. Rapallo . . . g 65 58 24 Sy ay es al ., Hs ON DA ae Oe 26 8 2577 2 20219 2 Spezia (type) C164 O24 1 SO e320 Lerici . po eS Ge GS Doe wed tell On? Gre b=) mn, BURL RERUN bi hy). A0E ecm mn ORD Shien oO mama) manta O: 23 1 % Og Se aoe ose SG OR BS ee NO 2) lee Slo eee es eee oo ay OR OO ey 7 Ie ee ee Bologna, 4) 5. eq) oy es eee mm OS um Ota Onnmn Onto ieee 20 eae fs Sd ol SRS Gee ghee Re a OO Se Oia 2.0) 22 ile) Mit; Meta, Terra di Wavoro 39g) 8) Sl yO 2732 2 The dorsal scales are granular and distinctly keeled; 535 to 52 transverse series correspond to the length of the head, and 4, or 3 and 4, to one ventral plate. The scales on the tibia are appreciably smaller than those on the back. 24 to 28 lamellar scales under the fourth toe. ‘The scales on the upper surface of the tail are rather strongly keeled, and are truncate or end in a very obtuse angle; 30 to 36 scales in the fourth or fifth whorl behind the postanal granules. Single specimens from Sestri Ponente, Portofino, Livorno, Bellagio, and Lerici are remarkable in the rostral shield forming a sutnre with the frontonasal; in four specimens! the granules between the supraciliaries and the supraoculars ave reduced to from two to five. ‘The masseteric shield is well developed, or even large, except in Bedriaga’s type specimen, which is further aberrant in having the parietal shields transversely divided, as in the female from Véslau noticed above. Two of the Lerici specimens have five anterior upper labials on one side, a third has five on each side. Measurements :— all; 2, 3. 4, 5. 6. From end of snout to vent . . . . 65 60 64 62 68 65 a ie foresumbyy ss) | eh 21 26 24 27 25 IbeaVOE awl 5 yo 6 io, o a dy 18 17 15 18 16 Witcher Go 6 4 aio i 6 5 dill 9 iit 9 11 11 Depthkotiteadiermers i aar-eeer ee age ann 8 6 8 6 9 8 Horedimib, is. sake eae eo ee 19 24: 22 238 25 JeGpovclibivnoye CNemye slo o,f. cae ee | 31 37 34 35 38 TROOG ie ee tetevk sj des Ss ast eo LO) 16 2) 18 19 2) RENT Pernt ee sh PR meee Ply tee tard scoala vdull(O) Fo 98h 130) AS 25 + Tail reproduced. 1. go. Rapallo. 4. 9. Lerici. Ph Ye r 5. ¢. Bologna. 3. ¢. Spezia (type). 6. g¢. Sestri Ponente. 1 Marked * on the table. LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. 385 As may be seen from the above-given measurements, the proportions do not differ from those of the typical form ; nor does, as a rule, the shape of the head, although it may be very strongly depressed—its depth, in certain males from Liguria and Pianosa, not exceeding the distance between the eye and the tympanum. The hind limb reaches the shoulder or the collar in males, the axil in females. As to the coloration, the passage from the typical form! is so complete that certain specimens may be referred almost equally well to the one form as to the other. For instance, one of the females from Lerici has the black network on the back much broken up and a well-marked light dorso-lateral streak, and would be unhesitatingly classified as f. typica but for the rather large black spots on the belly, arranged in regular longi- tudinal series. For it is frequently the case in this variety that the black spots run together to form longitudinal bands. The dorsal markings are very variable, usually affecting the form of a close network, but sometimes appearing in disconnected blotches or vermicular lines ; a vertebral chain of black spots may be present. Black and white spots form very regular and conspicuous bars on the sides of the tail. Dr. Gestro has very kindly sent me a large number of specimens of the Wall-Lizard collected indiscriminately in the immediate vicinity of Genoa; this series goes far to show the instability of the characters on which the distinction between the var. bruegge- mannt and the typical form rests. Some specimens are not at all separable from the latter, while others, of the most vivid grass-green on the back (Pl. XXII. fig. 2), answer to the former; and between these two extremes every possible variation in shade may be followed. A female with the markings as in the typical form, but remarkable for its green back and yellow lower parts, is figured on Pl. XXII. fig. 3. The back may be coppery brown, or greyish brown, or greenish grey or olive, or green in front and brown behind. ‘The colour of the lower parts varies equally, and irrespective of that of the upper parts. Most specimens are white on the belly, often with a brownish, pinkish, or greenish tinge, uniform or more or less spotted with black; while in one male specimen with olive-green back the lower parts are copper-red, with black spots confined to the collar-plates and to the two outer rows of ventral plates (P]. XXII. fig. 1). An interrupted or continuous series of turquoise-blue spots is always present on the outermost row of ventral plates. ‘The back, in the males, is always much spotted or reticulated with black or dark brown, and a black vertebral line or regular series of spots is often present; the sides are brown or purplish, with black network, often enclosing round blue or green spots; a blue, black-edged ocellus is often present above the axil. One of these specimens (Pl. XXIII. fig. 3) is so strikingly similar in its markings to certain males of the var. guadrilineata that it might easily be mistaken for that variety were it not for the flatter head; and it is noteworthy that its exceptionally 1 Especially from such individuals as have been named var. nigriventris (non Bonaparte) by De Betta, Erp. Venet. p. 154 (1857). VOL. XVIl.— PART 1V. No. 5.—October, 1905. E poh) 384 MR. G. A. BOULENGER ON THE VARIETIES OF small scales (65 across the middle of the body) is a further approximation to the Corsican-Sardinian form. Camerano has already pointed out that some of the Ligurian specimens examined by him, with a much flattened head and reticulate markings, are suggestive of the var. bedriage from the mountains of Corsica. I am much inclined to think that the vars. guadrilineata and bedriage are directly derived from the var. brueggemanni!. I may add that quite young specimens from Lerici, which I provisionally refer to this variety, lack the dorsal reticulation and the ventral spots, and cannot be distinguished from French specimens of similar size. Var. NIGRIVENTRIS. Podarcis muralis, var. nigriventris Bonaparte, Icon. Faun. Ital., Anfib. pl. —. fig. 6 (1836). Lacerta muralis neopolitana, var. nigriventris Bedriaga, Arch. f. Nat. 1879, p.277, and Abh. Senck. Ges. xiv. 1886, p. 213. Lacerta muralis fusca, var. nigriventris Bedriaga, Arch. f. Nat. 1879, p. 288, pl. xvil. fig. 3. Lacerta muralis neapolitana, var. ventromaculata Bedriaga, Bull. Soc. Zool. France, 1879, p. 205. Lacerta muralis, var. flaviundata Bedriaga, 1. ce. p. 218. Under this name Bonaparte has figured a Wall-Lizard, from Rome, measuring 210 millim., which has been regarded by Bedriaga as a variety of his subspecies neapolitana, on account of its large size and the green colour of the spots which appear between the meshes of the black network. Similar lizards occur, according to Bedriaga, at Arezzo in ‘Tuscany, in the Roman Campagna, on a rocky islet, La Scuola di Pianosa, close to Pianosa, in the Adriatic, and on Paimajola and Salina, Lipari Islands. In describing his LZ. muralis fusca, var. nigriventris, from Rome, Bedriaga distinctly disclaimed any identification with Bonaparte’s var. nzgriventris, and he soon after altered the name to flaviundata. This variety is described, from a single male specimen (now in the British Museum), as deep black above, with numerous golden-yellow wavy cross- streaks ; the white ground-colour of the belly to a great extent disappears under the large black spots; the ventral shields of the outer row are green. On comparing the figure with that given by him of L. brweggemanni on the same plate, a striking resem- blance in form and markings will be observed between the two lizards, the chief difference being that the black is more developed in var. flaviundata and the green colour is replaced by yellow. I owe to the kindness of my friend Dom Grégoire Fournier, O.S.B., an interesting little series of these black lizards collected by Dom Theodore Néve, O.S.B., on the outer walls of Rome, near St. Peter’s, and the series is of importance as completely bridging over the gap between the var. flaviundata and the var. nigriventris, which are seen to represent merely individual differences. The specimens reached me quite fresh, and 1 Since this paper was set up in type, I have received from Prof. Camerano a male specimen from Pianosa, which may well be regarded as intermediate between the vars. brueggemanni and bedriage. It has 70 scales across the body. Specimens from Elba, communicated to me by Count Peracca, are true var. brueggemanni with 58 to 65 scales across the body. LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. 385 showed the ground-colour betiween the black network to be lemon-yellow in some, yellowish green to bright grass-green in others. specimens, also from Rome, through Dr. Vinciguerra. 3 Measurements :— ile . Rome (flaviundata, type) . . 64 ¢&.- Outer walls of Rome . 76 » ” 73 . “6 70 . 66 67 2 5 65 33 5 65 3) PP) 64 op = 60 3) 33 66 ” ” 66 op Fi 52 . Castelfranco, Ostia 62 . Campagna,nr. Rome .. . 67 ie From snout to vent . 64. 5 Wore Ine 6 o o Be Length of head 17 Width of head 11 Depth of head 8 Fore limb . 24 Hind limb . 36 Foot d 20 Tail (reproduced) 105 40 +0 Of OY 2, 64: I have since received a few living In addition to these, the British Museum possesses one specimen from Castelfranco, Ostia (near woods), presented by Dr. Sambon, and one from the Roman Campagna, received from Prof. Collett. Ww Ww w (oy) (xt 9 www WW WD WO WD NODE EB OF OL eS Ct 107 4, 10 Rome (type of var. flaviundata). Walls of Rome. » Castelfranco. wnwwu COMCTUMCT CoN Onn GUN GOs or © Oe SE 2m NHNwnnwnvnvnwvnw nv wo o> 6. 18-19 19-20 =I ee Oe ee) Head (Pl. XXVIII. fig. 1) once and a half to once and two-thirds as long as broad, moderately depressed, its depth, in the tympanic region, equal to the distance between the anterior border or the centre of the eye and the anterior border of the tympanum. Hind limb reaching the shoulder or the collar in males, the axil in females. Foot not more than once and one-third the length of the head, sometimes but slightly longer than the head. 35 9 -/! 386 MR. G. A. BOULENGER ON THE VARIETIES OF Rostral not entering the nostril!; frontal as long as its distance from the end of the snout or a little shorter ; occipital often as long as, and sometimes much broader than, the interparietal (Pl. XXVIII. fig. 1); series of granules between the principal supraoculars and the supraciliaries incomplete, the first (or the first and second) supra- ciliaries being in contact with the second supraocular?; temporal scales granular, the tympanic and masseteric shields distinct, the latter more or less enlarged and in contact with the upper temporal shield or separated from it by one or two series of granules; supratemporal shield often small, sometimes broken up into small scales? ; usually four upper labials anterior to the subocular, five on each side only in one specimen, five on one side in three. 22 to 29 scales and granules between the symphysis of the chin-shields and the median collar-plate; gular fold distinct. Collar without serration, composed of 8 to 11 plates. Scales on body granular, oval or oval-subhexagonal, strongly keeled; 55 to 71 scales across the middle of the body ; 4 (or 3 and 4) transverse series correspond to one ventral plate, 37 to 51 to the length of the head. Ventral plates in 6 longitudinal and 24 to 29 transverse series. Preeanal plate large, with one or two semicircles of small plates. Scales on the upper surface of the tibia granular, strongly keeled, smaller than the dorsals. 18 to 23 femoral pores on each side. 27 to 31 lamellar scales under the fourth toe. Upper caudal scales strongly keeled, narrow, very obtusely pointed behind ; 30 to 38 scales in the fourth whorl behind the postanal granules. The coloration, which is the same in both sexes, is an exaggeration of the bruegge- manni variety, the black predominating. Some specimens approach very. nearly the latter in having the middle of the back green with a black network, whilst in general the black has so invaded the upper parts of the body that the yellow or green ground- colour is reduced to isolated spots, which may be as small as in the ji/folensis variety ; these spots may be round or wavy and transverse, as in Bedriaga’s var. flaviundata, with every passage between the two forms. The sides are black, with round cream- coloured, yellow, or blue spots. The head in some specimens may be described as brown with black spots or vermiculations, in others as black with yellow or green variegations. The limbs are black, with yellow or green round spots. ‘The lower parts are black and white, the black usually preponderating and often disposed in longitudinal bands; some specimens are nearly entirely black beneath, with mere remains of the white ground-colour. Large blue spots on the outer row of the ventrals, but never forming a continuous band as in Bedriaga’s figure of the var. flaviundata. ‘Tail with the black and white markings, usually so conspicuous on the sides in the typical * Forming asuture with the frontcnasal in one specimen. Another is remarkable in haying two superposed postnasals. * In the type of Bedriaga’s var. flaviundata the supraciliary granules are reduced to 3 or 4. * Tn one specimen the parietal does not touch the postocular. LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. 387 form, very strongly defined and forming nearly complete annuli, separated by brown interspaces; the regenerated portion, as in all Wall-Lizards, with black and light longitudinal streaks, strikingly contrasting with the annular markings of the primary tail. The markings are the same in the young as in the adult. Male specimens from Rome are figured on Pl. XXII. figs. 4-6, and a young one on text-fig. 6 6, p. 399. Replying to my enquiries respecting the distribution of LZ. muralis typica and var. brueggemanni in Italy, my friend Count Peracca kindly informs me that he received alive, many years ago, a lizard, apparently answering to the var. brueygemanni, from Naples, which specimen has appeared in Camerano’s list, but has since been mislaid’. He has since made many excursions about Naples without rediscovering such a variety, nor has he found any other than the serpa or neapolitana variety south of Naples, even at altitudes of 1000 metres or above. In Liguria he found the var. brueggemanni, with lemon-yellow or green back, only near the sea; on the hills, from 100 metres upwards, only the ordinary grey-brown typical form. In the Arno Valley the var. brueggemanni follows the plain which extends inland rather far from the sea, and it even penetrates as far up as Florence, where, however, specimens inter- mediate between it and the typical form occur. Much remains to be done in ascertaining the exact range of the typical form, the var. brueggemanni, and the var. nigriventris in Italy, and how far the three are capable of satisfactory definition. My own impression is that the first two intermix, in Liguria at least, where the southern and northern forms meet in the same localities, and that the same will be found to be the case between the second and third forms when we come to know more of the lizards occurring between Florence and Rome. It has often been stated that the serpa or neapolitana form can be readily distin- guished from the typical form by the shape of the head and the general proportions. How vague these differences are in most cases is well shown by these large lizards of the nigriventris variety, referred by some to the typical form, by others to the “neapolitana” form. And the resemblance to the latter is emphasised by a com- parison with the black Filfola lizard, which is regarded as derived from the Maltese “neapolitana.’ And yet there can be no doubt that the var. nigriventris is only an exaggeration of the var. lrueggemanni, which is itself completely connected with the typical form of Northern Italy and Central Europe. As often happens in these lizards, the var. nigriventris has acquired with its larger size a finer scaling, which approaches 1 As noticed above, a specimen from Mount Meta, not far from the Province of Naples, has been sent to me on loan by Prof. Monticelli. 388 MR. G. A. BOULENGER ON THE VARIETIES OF that of the var. serpa, as may be seen by a tabulation of the number of scales in specimens of the different varieties and local forms :— ie 2: 3. 4, 5. Var. brueggemanni . . « » « - 51-65 85-52 22-80) 17-25 4-88 Var. nigriventris, Rome. . . « + 55-71 87-51 22-29 18-23 27-31 Var. serpa, Rome, Naples . . . - 58-76 88-55 22-81 18-27 28-33 PA Malta © 20). oie Ol 741i 40S i223 eel 2 mao Ua Var. jilfolensis, Filfola . . . . . 68-82 48-62 29-88 18-27 31-36 1. Number of dorsal scales in a transverse series. 2. Transverse series of dorsal scales corresponding to length of head. 3. Number of gular scales in a longitudinal series, 4. Number of femoral pores. 5. Number of lamellar scales under fourth toe. B. Vars. cAMPEsTRIS Betta and serpa Raf. (Plates XXIII, XXVII., & XXVIII.) The forms to be described here have frequently been confounded, or grouped together under the same general designation, as by Bedriaga (L. muralis neapolitana), Camerano (L. serpa), and myself (var. tiliguerta). But it is really possible to dis- tinguish a nortbern and a southern form—the former, var. campestris, being more sharply differentiated from the typical form, with its vars. brueggemanni and nigri- ventris, than the latter. While concentrating their attention on the colour, the markings, and the shape of the head, characters far less stable than one would gather from their writings, Bedriaga and Eimer have somewhat neglected the lepidosis, which, however subject to fluctuations within very wide limits, affords a safer means of defining forms. If the lepidosis be taken into consideration, together with the other characters, it will be found that, after elimination of the typical form and the two varieties into which it gradually transforms from north to south, two further varieties can be separated with sufficient precision. If we could put aside the more northern of the two latter, the var. campestris, we should feel perfectly justified in saying, that the Wall-Lizard passes gradually, at least so far as structural characters are concerned, from the typical form in the north to the var. serpa in the south, the drift of variation from north to south being in the direction of a larger size, a shorter body, smaller scales, longer toes, and smaller scales. But this continuous series is broken, or obscured, by the presence, from the plains of North Italy to the Roman province, of the var. campestris, which, further south, gradually passes into the var. serpa, especially as regards the coloration. I fully agree with Himer in regarding the striated type of the var. campestris as the most primitive of all the Wall-Lizards, and I am quite prepared to admit, so far as such speculations based on theoretical conceptions are allowable, that it has given rise, or at least is most nearly related, to the form that has become modified into both the LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. 389 typical form and the spotted or reticulate types of the var. serpa. This is all the more plausible from the fact that the striated pattern is preserved, in females and young only, in the less divergent types of both the typical form and of the var. serpa. Some specimens of the latter are not at all to be distinguished, so far as markings are concerned, from the var. campestris. Eimer’s theory of the derivation of patterns of markings being accepted as correct, we are led to regard the var. campestris as the most ancient form from which the others are descended; and this, I think, is also supported by the structural characters, which differ less from what we may assume to be the more normal or generalised form of Lacerta before adaptation to climbing petrophilous habits had been reached. The var. campestris has often, and rightly, been compared to such species as J. viridis, L. agilis, and L. taurica. De Filippi and De Betta long ago pointed out that, although both the typical form and the var. campestris coexist in the plains of Northern Italy, they keep aloof from each other, the former being a true petrophilous lizard, while the latter has its abode in such localities as are frequented in Central and Northern Europe by L. agilis. According to Lorenz Miller, who has recently published an interesting popular article on these lizards!, the Wall-Lizard lives principally on walls and rocks, while the var. campestris (his L. serpa) prefers meadows, borders of woods, and other grassy localities; he tells us how sharply the territories of both forms are divided in a locality in Piedmont where he had an opportunity of observing them. It was on the road from Baldichiero to Montafia. On one side a steep rock with low prickly vegetation, on the other a grassy border separating the road from cultivated fields. ‘This road forms an absolute boundary between the abodes of the two forms. ‘The var. campestris is a form of the plain, rarely found, in isolated examples, in the mountains, which they exceptionally ascend along grassy valleys. At Imola, in Romagna, De Betta found “le due varieta muralis e campestris conviventi, sulle mura la prima, nella adjacenti campagna la seconda.” But in the south of Italy and in Sicily, where both the typical form and the var. campestris have practically disappeared, the var. serpa inhabits indifferently all localities, and runs up the walls and bare cliffs as well as among the grass. What the factors are that have preserved the purity of the two races (f. typica and var. campestris) over so great a part of Italy, we do not know, but it is within the limits of direct observation that south of Rome a marked distinction between the two has ceased to exist, and this has caused much divergence of opinion among herpe- tologists as to the naming of certain specimens, as I have pointed out below. I am myself inclined to the conclusion that the vars. brueggemanni and nigriventris gradually lead to the southern form, as suggested by Bedriaga, when describing the former, and that, climatic or other conditions having changed, these extremes of the typical form ‘ Bliitter f. Aq.- u. Terr.-Kunde, xiii. 1902, pp. 158, 169. 390 MR. G. A. BOULENGER ON THE VARIETIES OF have mixed with the var. campestris and formed together the assemblage which we now designate as var. serpa. ‘This explanation may account for certain reappearances of the characters of the typical form in Southern Italy, Sicily, and Malta, which have so greatly perplexed the best observers. ‘The Sicilian specimens which have been distinguished as var. sicula have retained much of the general habits and coloration of the var. campestris, but the lepidosis is quite the same as in the var. serpa. Var. CAMPESTRIS. Podarcis muralis, var. campestris De Betta, Atti Acc. Verona, xxxv. 1857, p. 152; Faun. d’Ital., Rett. Anf. p. 28 (1874); Atti Ist. Venet. (5) iv. 1878, p. 897, and v. 1879, p. 388. Lacerta tiliguerta De Filippi, N. Ann. Se. Nat. Bologna, (3) v. 1852, p. 69. Lacerta muralis neapolitana, vars. campestris and livornensis Bedriaga, Arch. f. Nat. 1879, pp. 274 & 279, and Abh. Senck. Ges. xix. 1886, pp. 209 & 228. Lacerta muralis, var. campestris Leydig, Deutschl. Saur. p. 228 (1872); Eimer, Zool. Stud. Capri, Ips 205 ple tl. owe. Lacerta muralis striata, part., Eimer, op. cit. p. 25. Lacerta muralis striata campestris Kimer, Arch. f. Nat. 1881, p. 328, pl. i. figs. 1 & 2. Lacerta serpa, part., Camerano, Mon. Saur. Ital. p. 56 (1885). Lacerta muralis, var. multifasciata Positano Spada, Boll. Soc. Rom. Zool. i. 1892, p. 154, fig. Head (Pl. XXVIII. figs. 2 & 2a) once and a half to once and two-thirds as long as broad, moderately depressed, its depth in the tympanic region equal to the distance between the anterior border or the centre of the eye and the anterior border of the tympanum ; snout obtuse. Neck as broad as or a little narrower than the head. Body moderately depressed. Hind limb reaching the shoulder, the collar, or a little beyond the collar in males, the axil or the elbow of the adpressed fore limb in females; foot once and one-sixth to once and one-third the length of the head. ‘Tail cyclo-tetragonal at the base, once and two-thirds to twice as long as head and body. Rostral shield narrowly separated from the nostril, or entering it’; nasals always forming a suture behind the rostral ; frontal usually as long as its distance from the end of the snout, sometimes considerably shorter; a series of granules between the supraciliaries and the two principal supraoculars, the first of which is usually in contact with the frst supraciliary, rarely with the second also; in a few specimens ® the series of granules beginning from the first supraocular; parietals once and one-third to once and a half as long as broad, usually in contact with the upper postocular, but sometimes not®; occipital very variable, usually shorter but often broader than the 1 This I observe in several specimens from Turin, in one from Florence, in two from Castelfranco, near Rome. 2 One from Venice, one from Bologna, one from Lake Trasimene, Perugia. * In four specimens from Turin, in two (out of four) from Lake Trasimene, in five (out of six) from Castelfranco, near Rome. LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. 391 interparietal ; temporal scales usually larger than in the typical Z. muralis, with the tympanic shield always distinct and the masseteric frequently so, the latter separated from the parietal by a narrow supratemporal or by one to three series of scales ; nearly constantly four upper labials anterior to the subocular!. Collar-edge more or less distinctly serrated (Pl. XXVIII. fig. 24), sometimes as much as in ZL. viridis; 8 to 11 plates in the collar; gular fold very distinct ; 21 to 32 scales and granules between the symphysis of the chin-shields and the median collar-plate. Scales on the back roundish-hexagonal or oval-hexagonal, very distinctly, often sharply, keeled; the vertebral line usually with a more or less regular band of much smaller granular scales (Pl. XXVIII. fig. 2¢); lower lateral scales as large as or a little smaller than the dorso-laterals, smooth or faintly keeled; 53 to 65 scales (usually 57 to 61) across the middle of the body; three or four transverse series of scales correspond to one ventral plate, 53 to 50 to the length of the head. Ventral plates in 6 longitudinal ? and 25 to 51 transverse series. Anal plate smaller than usual in L. muralis typica, usually with two or three semicircles of small plates, but sometimes with one only. Text-fig. 5. a. b. (@ Anal region of Lacerta muralis, var. campestris. ad, Turin. 6. 2, Turin. c. o, Ancona. Scales on upper surface of tibia smaller than the dorsals, more or less strongly keeled; 24 to 30 (usually 25 to 29) lamellar scales under the fourth toe; 15 to 24 (usually 17 to 22) femoral pores on each side. Upper caudal scales (Pl. XXVIII. fig. 2 ¢) strongly keeled and obtusely but distinctly pointed behind; 30 to 40 scales in the fourth whorl behind the postanal granules. Typical examples of this race, as represented by specimens from Verona, Venice, Treviso, and Turin (Pl. XX VII. figs. 1 & 2), have the upper surface of the head, a vertebral band, the sides of the body, the limbs, and the upper surface of the tail pale brown or olive-brown, the sides of the back bright grass-green; the brown vertebral band bears two parallel series of small black spots, or a single straight or zigzag series of large black spots; these spots, as a rule, do not begin until some way behind the occiput or nape; the sides, in females, bear two more or less distinct, straight or wavy, whitish streaks (the upper proceeding from the supraciliary border, the lower passing through the ear), which may be edged with black or bordered by small black ’ Five on one side in one specimen from Lake Trasimene, three on each side in one from Ancona. 2 The outer ventral plates are divided into two in a female specimen from Lake Trasimene. VOL. XViI.—PaktT lv. No. 6.—October, 1905, 3F 392 MR. G. A. BOULENGER ON THE VARIETIES OF spots, while in adult males they are spotted with black and white, the spots often forming ocelli; a large black spot or an ocellus with blue or green centre often present above the axil; black spots on the upper surface of the head present or absent ; white black-edged spots on the hinder side of the limbs. Large male specimens have the upper surface of the head bright green. Lower parts white, greenish, yellowish, or reddish, without spots or with a series of black spots on the outer row of ventral plates, which may also bear blue or green spots, usually of small size. The young are always very distinctly streaked, and the two green dorsal stripes are narrower in proportion to the brown ones (text-fig. 6 ¢, p. 399). Further specimens answering to the above description are in the British Museum from Bologna, L. Trasimene (Perugia), Florence, Ancona, Castelfranco near Ostia, Elba, and Corsica. De Betta himself had recorded his var. campestris from Corsica and Rome. More aberrant specimens are the following :— Green colour absent!, or reduced to a narrow streak on each side of the vertebral stripe (males and females from Turin and Trieste). Olive above with three reddish-brown stripes, but no spots (female from Turin). Pale brownish grey, with a pale green stripe on each side of the vertebral line; a trace of a whitish dorso-lateral line; no spots (female from the sandy shore at Ostia, near Castelfranco, Rome). Five black stripes along the back, with narrow light streaks between them (var. multifasciata Positano Spada, from Ponte Salario, near Rome). Green above, with a dorso-lateral series of large black spots; another lateral series of large black spots from axil to groin, and a third on the outer row of ventral plates, which are blue; no vertebral stripe or series of spots (var. divornensis Bedr., from Colambrone, near Leghorn). Measurements :— 1. 2, 3. 4 From end of snout to vent. . . 70 62 60 60 iy i My fore limb . 26 21 23 21 eno thvorshea dia ii. satiety aL (4 13 15 13 Minatvorlnegl > G o 46 o o o | dll 9 10 9 IDgada Oe lglg 5 6 6 56 6 o LO 63 8 7 oma lin 4 6 6 6 6 6 o au oe 19 20 19 lett Ibi o 6 6 6 6 6 0 6 CY 32 34 29 HOOb mA Dea meee et, <2 craps date ora O 7 18 15 Marlon ial sawn cat des be OR ee a8 118 100 102 1. g. Verona. 2. 9. Treviso. 3. ¢. Turin. 4, 2. Turin. 1 One cannot, however, too much insist on the chameleon-like changes that take place in these and other lizards. We know how the male Z. agilis may entirely lose its green colour towards the end of summer, and I have observed a brown lizard of the var. campestris, which in life showed no trace of green, turn to pale green on the anterior part of the back after a few hours’ immersion in spirit. LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. Particulars of some of the specimens examined :— Verona. +” Venice . Treviso . Bologna be) oP) Florence 50 eae ce somes nares Calambrone, Leghorn . Ancona 6 6 L. Trasimene, Perugia Rome 1. Total length (in millimetres). ils g 70 ee G3 ¢ 50 3 76 E72 ay vy oy) 70 9 62 S$ 64 pe Bits » 58 re 50 9 59 a” 60 » 96 3) 55 g 69 > 67 Q@ 52 3 67 9 64 SUA 3 65 S$ 82 » 68 @ 70 68 EY eee ae 70 as GS 9 61 3 66 2. 3. 4 }. 6. 25 9 224 19-18 LOM le 2S 18 ZOme el oO 19 27 © 21 30 10 26 #£«17-18 32 O23 17 3l 8 24 15-16 30 iS) eB 16 26 10 28 #£17-18 24 Il 28 22-21 29 10 25 20-19 23 Il 23 21-20 29) 10) 23° 18=1'9 26 9 2 19-20 2 10 22 20 26 10 2 22-23 26 10 25 22-19 BO) Ls} 21 G0) QB 22 25 8 27 22-20 Pek NO) Bas 18 26 Il 27 24-23 0) aL 18 28 Il 26 21-19 vo 2. Number of scales across middle of body. 3. Transverse 5. Number of scales and granules between series of ventral plates. 4. Number of collar-plates. chin-shields and collar. 6. Number of femoral pores (right and left, if differing). lamellar scales under fourth toe. bo 7. Number of 394 MR. G. A. BOULENGER ON THE VARIETIES OF Var. SERPA. Lacerta serpa Rafinesque, Caratt. alc. u. Gen. Anim. p. 8 (1810). Lacerta sicula Rafin. 1. ¢. Lacerta olivacea Ratin. |. c. Lacerta puccina Rafin. |. e. Podarcis muralis albiventris et siculus Bonap. Icon. Faun. Ital., Anf. (1836). Lacerta muralis cerulea Eimer, Verh. phys.-med. Ges. Wirzb. i. 1872, Sitzb. p. ix, and Zool. Stud. Capri, ii. p. 5, pl. i. (1874). Lacerta muralis neapolitana Bedriaga, Enst. d. Farb. b. d. Hidechs. p. 15 (1874). Lacerta faraglionensis Bedriaga, op. cit. p. 16, pl.—., and Die Faraglione-Kidechse (1876). Lacerta muralis, vars. striata, elegans, modesta, maculata Eimer, Zool. Stud. Capri, ii. p. 24, pl. ii. Lacerta muralis, var. viridiocellata Bedriaga, Arch. f. Nat. 1877, p. 113. Lacerta muralis neapolitana, part., Bedriaga, Arch. f. Nat. 1878, p. 285, Bull. Soc. Zool. France, 1879, p. 202, Arch. f. Nat. 1879, p. 274, and Abh. Senck. Ges, xiv. 1886, p. 204. Lacerta muralis, vars. faraglionensis, latastii, viridiocellata Bedviaga, Bull. Soc. Zool. France, 1879, pp. 207, 209, 211, pl. ix. fig. 4. Podarcis muralis, vars. albiventris, sicula, doderleini, maculata, viridiocellata, faraglionensis De Betta, Atti Ist. Ven. (5) v. 1879, pp. 3889-393. Lacerta muralis maculato-striata, punctato-striata, punctato-fasciata, striato-maculata, maculata, reticulata (part.), modesta, concolor, elegans, ceeruleo-ccerulescens, caerulescens monaconensis, caerulescens gallensis Kimer, Arch. f. Nat. 1881, pp. 875-389, pls. xili.—xv. Lacerta serpa, part., Camerano, Mon. Saur. Ital. p. 56 (1885). Near Rome the var. serpa occurs in addition to the vars. nigriventris and campestris, but further south the var. campestris is replaced by this very closely related form, which reaches a larger size, and which in many respects differs little from the typical L. muralis, especially if compared with the southern var. nigriventris. As in the latter, the head is generally more flattened, resembling less that of ZL. viridis, the temporal scales are smaller, more granular, the collar has the edge even, or, if serrated, less distinctly so than in var. campestris, the difference in size between the vertebral scales and the dorso-laterals is less marked, the subdigital lamelle are usually fewer, and the caudal scales are more truncate posteriorly, their posterior border rarely forming a well-marked angle. But it is distinguished from the typical form and its vars. brueggemanni and nigriventris by generally smaller scales, more numerous femoral pores, and longer toes with more numerous lamelle under the fourth. The complete passage between LZ. muralis typica and the var. serpa, so far as scaling is concerned, is well expressed by the following tabulation ! :— ale 2. 3. 4, 5. Forma typica 5 6 o o CREO 29-45 19-30 13-25 22-28 Var. brueggemanni . . . 51-65 35-52 22-30 17-24 24-27 Var. nigriventris. . . . 55-71 37-51 22-29 18-23 27-31 28-36 Weis Sans “6 1d 6 @ a WeHwy/ 30-60 25-385 19-29 * 1. Number of scales across middle of body. 2. Number of transverse series of scales corresponding to the length of the head. 3. Number of scales along middle line of gular region. 4. Number of femoral pores, o. Number of lamellar scales under fourth toe. LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. 395 As regards the coloration, which is highly variable in individuals associating together, the resemblance to var. campestris is, as a rule, greater than to any other form, and some specimens approach the latter so closely that it is not surprising that most authors should have failed to draw a distinction between the two forms. The head is often comparatively larger than in the typical form and the var. campestris ; its width is contained once and a half to once and two-thirds in its length ; its depth equals the distance between the anterior border or the centre of the eye and the anterior border of the tympanum ; the snout is obtusely pointed and as long as or a little longer than the postocular part of the head. The hind limb reaches the collar, or between the collar and the ear in males, the axilla, the shoulder, or the collar in females; foot once and one-sixth to once and a half the length of the head. The tail is rounded in section and measures once and two-thirds to twice and four- fifths the length of head and body. The rostral never enters the nostril and is nearly always separated from the frontonasal by the nasals!; postnasal single; frontal as long as its distance from the end of the snout, or shorter; series of granules between the supraciiiaries and the principal supraoculars rarely complete °, the first supraciliary, and sometimes also the second, being in contact with the second supraocular ; parietals once and one-third to once and a half as long as broad, nearly always in contact with the upper postocular 3 ; occipital usually shorter but frequently broader than the interparietal ; temporal scales granular, usually not much larger than the dorsals; tympanic shield distinct; masseteric shield usually present +, more or less developed, separated from the parietal or from the upper temporal by one to three series of scales; almost constantly four upper labials anterior to the subocular °. Collar with even edge (Pl. XXVIII. fig. 5) or very feebly serrated, composed of 7 to 12 (usually 9 to 11) plates ; 25 to 53 scales and granules in a longitudinal series between the symphysis of the chin-shields and the median collar-plate ; gular fold distinct. Dorsal scales (Pl. XXVIII. fig. 8 a) very small, convex, round or roundish-hexagonal, more or less distinctly keeled, sometimes smaller still and irregularly disposed on the vertebral line; 58 to 77 scales across the middle of the body, usually 62 to 70; 3 and 4, 4, or 4 and 5 transverse series correspond to one ventral plate, 35 to 60 to the length of the head. Ventral plates in 6 longitudinal and 23 to 30 transverse series. Preanal plates usually smaller than in the typical form and with two semicircles of small plates. ' The rostral forms a suture with the frontonasal in a female from Modica and in a male from Rome. 2 Two specimens from Procida Id., one from Casamicciola Id., one from Ischia, one from Syracuse. * Exceptions in three specimens from Catania, in two from Lipari Id. (Plate XXVIL. figs. 5, 5a), in one from Giglio Id., in one from Ponza Id., in one from Capri, in one from Pompeii, and in one from Messina. 4 Absent in two specimens from the Faraglioni, in one from Capri. * Five on each side in one specimen from Pompeii and in one from Messina. 96 MR. G. A. BOULENGER ON THE VARIETIES OF 3 Scales on upper surface of tibia minute, keeled, smaller than dorsals; 19 to 29 femoral pores on each side (usually 20 to 26); 28 to 36 lamellar scales under the fourth toe. Upper caudal scales (Pl. XXVIII. fig. 3a) strongly keeled and truncate or very obtusely pointed ; 30 to 38 scales in the fourth whorl behind the postanal granules. The enormous amount of variation in colour and markings which this form undergoes, often irrespective of localities and among individuals living together, may be gathered from the above synonymy, and especially from the general accounts given by Eimer. Suffice it to say, for the purposes of the present work, that some specimens, females and young, closely agree in colour and markings with de Betta’s var. campestris (Kimer’s striata), which may be regarded as the primitive type from which the spotted, reticulate, and uniformly coloured forms are derived. The colour of the head and back varies from greyish brown to olive, grass-green, or bluish green; striated specimens are the exception, most of the specimens being spotted or reticulated with black (Pl. XX VII. figs. 4 & 6), the black spots often forming longitudinal series (Pl. XXVII. fig. 3), sometimes wavy cross-bands; some specimens show only faint traces of darker markings and lead to others which are uniformly olive or yellowish brown (LL. olivacea Raf., modesta Kimer). A blue or green ocellar spot is often present above the axil,and sometimes acquires a large size. ‘The lower parts are usually white or greenish white, and without spots, except on the outer row of ventral plates, which often bear blue or greenish-blue spots. In some specimens from Sicily (var. sicwla Bp.), which in structural characters entirely agree with the var. serpa, the belly is orange or red in the males, yellow or white in the females, these specimens being further remarkable for the brilliant green colour of the back, bordered on each side by a more or less distinct whitish streak, as figured on Pl. XXVII. sites ho The most remarkable colour-variation is that exhibited by the lizards found on the Faraglioni rocks close to Capri (var. cwrulea Eimer, faraglionensis Bedr.): black or bluish black above, with more or less distinct markings of a more intense black, sapphire-blue or greenish blue beneath; pale blue spots usually present on the sides; the supra-axillary blue ocellus always distinct (Pl. XXIII. fig. 2). Specimens with blue belly and blue or blackish back, with or without lighter blue or ereen spots, are reported from the Mezzagiorno rock near Palmarola, and Santo Stefano Island near Ventotene. ‘The vars. cwruleo-cwrulescens, cerulescens monaconensis, and cerulescens gallensis of Kimer, from other small rocks near Capri, also with blue belly, are practically intermediate in the dorsal coloration between the var. ca@rulea and the lizards of Capri. * J. Scherer, Blitter f. Aquar.-u. Terrar.-Kunde, xiv. 1903, p. 288, regards L. sicula as a species distinct from L. serpa, but Iam unable to confirm the distinctive characters (collar, supraciliary shields, &c.) which he adduces in fayour of such a separation. LACERTA MURALIS IN WESTERN EUROPE AND NORTIT AFRICA. Measurements :— il Dy 3h 4, OD. 6. From end of snout to vent. . . . 90 70 79 66 65 66 5 op 3 forelimb . . 386 25 31 25 26 24 Ibemeua@Dexls 6 6 6 a 0 46 0 oe 16 20 15 16 15 Wid thuotineadiss . S)).5 aise ramet 9 13 9 10 9 Wepthyotehead® se: 3. & eee eee 7 10 8 8 7 Borewimbaeesna 2 | 3 Se OO 21 28 23 24. 22 Eindslim beeen. 2) cs Gk en ee 37 45 37 39 i IRON GS oll oko) hese wo. Cees 21 25 21 21 21 Tai) Meee ak} sn) 5) See lUl Seen Say eels Gro ieee SSeS * Tail reproduced. 1. g. Rome. 4. 9. Faraglioni. 2. 9. Pompeii. 5. g. Palermo. 3. ¢. Faraglioni. 6. @. Palermo. Particulars of some of the specimens examined ! :— ie Nee a ag 6. 7. Romcmepeee ts oe cs: co SRR SOME ETGM O78 10Ne -oik moon otian30 Beye a on 54:) 2 oak, Su BOM TONee Qn ml OM OOM mono emEOS FP a EN c. (hose hy: 21) COMMCONGIO Iga Ty 5728) .apoSwe ns 30 oman SU See eee es) (SO GD 82 © br Boe aT DP hfe a Ne tO GGE BOGOR Ova MICOMMIMNSD b : ne) G8 A Oye epee) i 98} GS oR) il) eT ¢ S70 GE OR OG DD) i Sele 5S) 26 plOme2s et 19221 ames rae — (=) ie) — cw) ii w oo oo <>) Pov NW 5s oo 6 0 po GS @ wll 2) ra fon) ee) ~w wo w CO or i oo oo © Ow WO %~ (0) ic) (ey) 3) ” Ae 9 Ventotene Id. Se es EE -eG Sam SigenoOies 6 6 6 6 6 © (SO) ea 22 12 36 28 33 Tsclivapimdae beak wine isl hie 2 Wire NG 4 6 6c 3 3 Sees tac, Wek ie Rrocidapldsmwsae-) aaa nen) TA) (oe) us 9 ~ QI ~) is ri) [es] oo © > See ee. co 69 65 28 12 381 21-23 «30 5 ty Smite, MOA eno) DHOdmem 29 8 26 2421 33 @asamicciola ld; . =» «. = G ¢4 60) 25.510) 2% 26-27 31 Chriiaty (oy A oh ibe ash ey OU Gah Be 9° 27 821-20 «632 aravlioniees sar ys) eesti) OC) vO mee aomell ~32) (28-24 lor, Chi YP) PY/ 9 86 22-23 31 ” . . . . . . ” ‘ Tabulated as in var. campestris, above, p. 393. 398 MR. G. A. BOULENGER ON THE VARIETIES OF 1 2 4, 5 6 Mi Faragloni GA 69 27a ae 2 Sa 0. pias’ © 74 .68 80 De PAD 1) PN? AD) e Ge GO 80 om Be Bea sp Be : Sei OO acuad! Po NIk 0) 22 30 Giannutri Id.. 3 72 465) 2b) LON 28a 242229) Reggio, Calabria 3 268) 926, A Sle 262829) es os 2 69 68 29 11 29 28-24 32 Lipari Id. Ge GOT 738 e226 9 28 24-28 33 3 F 2 635 37/5030) 9 382 26 36 Stromboli Id . g 5S 167) 28 9 32 2425 33 Messina ref Ome Ae Oe a hee 25 34 yi eG BS ID BY BR Il - g 68 70 27 9 25 24-23 32 a » @ Gs 2 NO 4e (2eeee all Catania. 3 SAN 76) 24a om ee 7 25 31 5 ON (nn (ee a> ne O42 OS O a Sone Alann 2 Omen (ange) 2 me O Rs At 567.) (60) 26 9° 25 22-23 29 = g 71. 66 - 28 9 27 21-22 31 i oy) GG GD BX) IO. ay 24 30 ee nS po OSe 69 27, OM 27123-22730 Syracuse 6 64 65 26 .18 -27 25-23 32 Modica . Bo tho eG OS a UI PR IL FI LRA ASME aie) URRO TUE -CoRinams) df dene 665.65), 29) I 2) 2322) Sil TEGO Sie bb, eld ol ht, GRE) Bel 2) 8 6 19-20 28 Bosco di Marineo, Palermo g¢ 70 65 27 Jl 25 22-23 30 (Pl. XXVIII. figs. 4, 4a). Bosco di Marineo, Palermo . @ 63 638 25 .10 29 21-22 30 ” » 7 Of GO. AY) 9 27 21-19) “730 Bedriaga, in his Monograph of 1886 (p. 190), records the typical form (his fusca) from Messina, but without particulars. Specimens from Messina, bearing a consider- able resemblance in form and markings to the typical L. muralis, have been presented to the British Museum by Count Peracca, but in their scaling they agree with var. serpa. An aberrant male specimen from Cosenza, Calabria, has been sent to me by Count Peracca, with the request that I should decide whether it must be referred to the var. serpa or to the var. nigriventris. It is, in fact, exactly intermediate between the two, both in form and coloration, and it approaches the specimens from Malta which will be described hereafter. Its lepidosis, as expressed in the above table, is :— 1. 28 3. 4. 5. 6. Ue Cosenza mena n- mC 76 66 24 9 27 22 30 ‘This would apply equally well to var. nigriventris and to var. serpa. The head is once and two-thirds as long as broad and its depth equals the distance between the centre of LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. 399 the eye and the tympanum. The hind limb extends to midway between the collar and the ear; the foot measures once and one-third the length of the head. The scales on the back are granular, oval-hexagonal, distinctly keeled; 3 and 4 transverse series correspond to one ventral plate, and 45 to the length of the head. The body and the limbs are olive above (in spirit) with black spots forming a wide- meshed network; no distinct ocellar spot above the axil; the head is much spotted with black above, and a black streak runs along the sides, passing through the eye and above the tympanum, and continued on the neck; tail with blackish transverse bands with sharply defined white spots on the sides. The throat and belly are greyish white, with small black spots. Text-fig. 6. Young specimens of :—a, f. typica, Courmayeur, Piedmont ; 6, var. nigriventris, Rome; c, var. campestris, Turin; d, var. serpa, Pompeii; e, var. serpa, Messina. V.—MALTA. (Plate XX VII.) The Wall-Lizard of Malta has been identified by Bedriaga and myself with the Sicilian and South-Italian form (var. serpa), by Camerano with the typical form, while Eimer regards it as intermediate between the two. I regard EHimer’s view as perfectly correct if size and coloration alone are taken into consideration, but the scaling is decidedly that ‘of var. serpa. Eimer, Arch. f. Nat. 1881, p. 570, who has observed the lizard in Malta, remarks as follows :— “Tt is remarkably small for a Southern form, not larger than the Corsican [var. quadrilineata], which is not much larger than the German. Besides, it is platycephalous {like the typical form] and belongs to the reticulata type, more or less nigriventris, the neck, the sides, and the posterior part of the belly being more or less spotted with black. However little the latter character may be marked, it is, however, highly note- worthy that it should, as so frequently in North Italy, coincide with the reduction of VOL. XVU.—Part ly. No. 7.—October, 1905. 3G 400 MR. G. A. BOULENGER ON THE VARIETIES OF the large dorsal spots in a locality so far removed from the home of the platycephale reticulate, and so completely isolated. It is also striking that the shape of the head should alter simultaneously with the size, the colour, and the markings, for it must be noted that the Maltese form shows less green but more brown and yellow in its livery. We have therefore apparently to deal here with correlative conditions between the form of the head and the other factors. As the race derived from the Maltese lizard, the large, robust, black lizard of the Filfola rock, is pyramidocephalous!, we feel inclined to conclude that a correlation exists between the shape of the head and the robust. build.” I have not examined a very large number of lizards from Malta, but I find that, if they agree in the shape of the head with the typical form, as is also the case in many Sicilian and South-Italian specimens of the var. serpa, they certainly agree with the latter in the lepidosis, as may be seen from the following tabulation :— ls 2. 3. 4. 5. 6. fe Gis 71 61 27 Wil 35 25 34 Ges 63 64 20 9 30 23-24 30 Sis 59 74 27 8) 31 19 30 Oey. 57 68 29 ) 3l 20-22 33 ee 57 64 30 wal 29 24-22 32 The depth of the head equals the distance between the centre of the eye and the anterior border of the tympanum. ‘The hind limb reaches the collar, or a little beyond, in males, the axil in females; the foot measures once and one-third the length of the head. The rostral shield does not enter the nostril, and in two specimens it is in contact with the frontonasal; the series of granules between the supraciliaries and the supraoculars is incomplete, the first supraciliary being in contact with the second supraocular; the parietal is in contact with the upper postocular; occipital as large as or smaller than the interparietal, in one specimen a little larger and twice as broad; supratemporal narrow or broken up; temporal scales small, but much larger than dorsals ; tympanic and masseteric shields usually well developed; usually four upper labials anterior to the subocular ?. Collar even-edged; gular fold well developed. Dorsal scales granular-subhexagonal, faintly keeled; 4 or 4 and 5 correspond to one ventral shield, 40 to 58 to the length of the head; scales on upper surface of tibia smaller still. Upper caudal scales rather strongly keeled, truncate. * This is true only of adult males.—G. A. B. * A young specimen, received from Prof. Giglioli, has five anterior upper labials on both sides and lacks the saasseteric disk. LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. A(1 The upper parts are greyish olive, in spirit, with a darker lateral band, with or without a vertebral series of small blackish spots, or reticulate all over with brown or black, the network enclosing roundish pale spots; on the posterior part of the back the dark markings may form wavy transverse bands; the head is sometimes uniform olive-brown above, sometimes much spotted with black; the lower parts are unspotted, or with rather small black spots, forming regular longitudinal series along the rows of ventral plates. On the whole, as noticed by Eimer, the coloration bears much resemblance to that of LZ. muralis (typ. and var. brueggemanni) as occurring near Genoa. The young is covered with dark and light spots, with a dark lateral band edged above and below with whitish. According to Eimer, the colour of the back in life is yellowish green or yellow; the belly saffron-yellow or red. Measurements :— g Q. IO GiClOrGnomn wo yess 5 0 o 5 o ll 57 2; 5 5 fore limb 28 21 Length of head Sy Poa mee ch ee ecw ag 12 Wien Cielo 5 5 6 6 0 6 oo 6 ill Oss (Honeshimib; © enn agale Sch eee ok Sep Soe 20 lnordilumib; ja gee sy ao Steet ee ES 30 Otol Re Mn REM ae TOS is, Us gust org 0 nel 18 Tail (reproduced) . 107 85 The Wall-Lizard of Linosa (Pl. XX VII. fig. 8.), referred by Camerano and myself to the typical form, agrees in all essential respects with that from Malta, judging from the single specimen at my disposal, received from Prof. Giglioli. It has 68 scales across the body, 28 gulars in a longitudinal series, 25 femoral pores on each side, and 34 lamellar scales under the fourth toe. ‘The belly is spotted with black. The following form, from the Filfola Rock, south of Malta, is evidently derived from the Maltese lizard, just as the Faraglioni lizard is derived from that living on Capri. But while in the latter case the distinction rests entirely upon the darkening of the ground-colour—size, scaling, and markings having remained exactly the same,—we here find that the melanism is obtained by an invasion of the black markings, and this difference is accompanied by an increase in size and, on an average, smaller scales. The Filfola form is therefore entitled to the same rank as the var. nigriventris, which stands to the typical form exactly in the same relation as the var. fi/folensis does to the var. serpa. Var. FILFOLENSIS. Zootoca lilfordi, part., Giinth, Ann. & Mag. N. H. (4) xiv. 1874, p. 158. Lacerta jfilfolensis Bedriaga, Die Faraglione-Hidechse, p. 19 (1876) ; Braun, Arb. zool. Inst, Wiirzb. iv. 1877, p. 49, pl. 1. fig. 14; Bedriaga, Arch. f. Nat. 1878, p. 297. 402 MR. G. A. BOULENGER ON THE VARIETIES OF Lacerta muralis, vay. filfolensis Fimer, Arch. f. Nat. 1881, p. 408, pl. xv. fig. 24; Bedriaga, Abh. Senck. Ges. xiv. 1886, p. 218. Podarcis muralis, var. filfolensis De Betta, Atti Ist. Ven. (5) v. 1879, p. 392. Giinther had only two specimens of the Filfola lizard before him when he published his remarks on Zootoca lilfordi, and both Bedriaga and Kimer do not appear to have had access to any very large material when writing their descriptions. The figures accompanying Braun’s remarks were taken from one of Giinther’s specimens in the British Museum, but are very inaccurate, as the author himself suspected. A few years ago, my friend Mr. Norman Douglass visited the Filfola Rock, and collected a good number of specimens, which he has kindly presented to the British Museum. The following description is based on 15 specimens, measuring from 56 to 82 millimetres from snout to vent. Eimer’s largest specimen (male) measured 90 millimetres without the tail. The depth of the head (Pl. XXVIII. fig. 5) equals the distance between the centre or the anterior border of the eye and the tympanum ; its width is once and a half to once and three-fifths in its length. The hind limb reaches the shoulder, the collar, or a little beyond in males, the elbow of the adpressed fore limb, the axil, or the shoulder in females ; the foot measures once and one-fourth to once anda half the length of the head. Except in two specimens, the rostral shield does not enter the nostril; the nasals form a suture behind the rostral; the frontal is as long as or considerably shorter than its distance from the end of the snout; a series of granules between the supraciliaries and the two principal supraoculars, this series complete in five specimens, whilst in nine the first supraciliary is in contact with the second supraocular, and the second supraciliary also in one specimen ; parietal once and one-third to once and two-thirds as long as broad, in contact with the upper postocular, except in one specimen; occipital very variable in size, as long as or shorter than the interparietal, sometimes wider than the latter; supratemporal narrow or broken up; temporal scales small, but much larger than the dorsals; tympanic shield distinct; masseteric shield usually well developed; four upper labials anterior to the subocular. Collar even-edged ; gular fold distinct. Dorsal scales granular-subhexagonal, feebly or faintly keeled; 68 to 82 scales across the middle of the body ; 4 and 5 correspond to one ventral plate, 43 to 62 to the length of the head; ventral plates in 6 longitudinal and 26 to 31 transverse series. Anal plate rather large, bordered by one or two semicircles of small plates. Scales on upper surface of tibia smaller than dorsals; 18 to 27 femoral pores on each side; 31 to 36 lamellar scales under the fourth toe. Upper caudal scales strongly keeled, truncate ; 34 to 42 scales in the fourth whorl behind the postanal granules. Eimer describes the colour of the living male lizard as intensely black above, with small roundish spots or dots which are greenish yellow on the middle of the back, more LACERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. 403 green or bluish on the sides; lower down on the sides the spots are quite blue; a cobalt-blue spot in the axil; the ventral shields are reddish brown or rusty red in front, blackish grey or bluish black behind; a series of blue spots on the outer row of ventrals. This coloration, which, as pointed out by Eimer, may easily be derived from that of the Maltese lizard, is strikingly similar to that described by Costa in his L. acrolampra from Aspromonte, Calabria!, and is also very similar to that of Bonaparte’s var. nigriventris. ‘The female is described as having the upper parts often more purplish brown, especially along the vertebral line, and the belly of a paler rusty red, fading to flesh-colour or pale reddish brown under the limbs and tail; the throat pale blue, spotted with black and reddish brown. Specimens in spirit are black above, with pale greenish-blue dots or small round or oval spots, which are usually larger on the sides; the upper surface of the head is brownish black, uniform or with small pale spots. A female specimen is entirely black above. ‘The belly varies from dark grey to bluish black; in some of the males, black spots form regular longitudinal bands on the ventral plates. The tail sometimes bears light spots, above and beneath. Particulars of specimens examined :— 1 2: os 4, 5 6. Ub. Gun 82 78 26 10 29 18-20 33 x 82 79 28 12 36 23 ? 53 80 75 26 10 33 23 ? 5 76 80 26 9 38 if 32 35 75 82 27 1 36 24, 34 25 75 80 28 10 33 25 33 m 74 Ye 27 10 85 26-25 36 DS 67 71 26 10 35 24 31 fe) 7A 69 3l 10 33 24 35 a 70 7 29 10 382 22 31 a 69 72 29 10 31 24 32 a 65) 77 ON LON 36) 20" 2INr as? 55 63 68 28 11 37 25 36 7 56 79 28 11 33 21-23 32 eh 56 HO 28 8 382 22 33 Measurements :— Ole fe) From end ofsnout to vent ...... 82 70 a3 ot 5 oe bbi) 56 6 « o Cid) 25 Hencth) ofshiendieay suse vetee Manet mei oe 20 15 WidthvotheadmenWentocn mute cit am tse LO 10 ’ Fauna di Aspromonte, p. 75 (1828). 404 MR. G. A. BOULENGER ON THE VARIETIES OF Depuln.or lel Goo so) 66 o 5 6 «0 IO 7 Koreslamibey. \

  • Py a 0) 65 6 26 10 30 20-19 Bh a g 52 64: 6 31 8 30 20-21 3 33 39 50 58 6 30 8 26 17 3 i = 48 55 6 29 9 28 18 3 55 5 A7 5d 6 28 8 27 19-17 0 * 5 46 57 6 30 9 25 16 D a 55 45 56 6 32 10 28 Wig Tlemsen Ss 56 56 6 25 11 24 18-17 b Fr 54 61 6 27 8 Die 18 5 3 52 65 6 27 10 29 21 55 ee 48 60 6 25 10 27 19-18 x 56 57 6 30 9 27 19 a5 By 56 55 6 30 iil 26 18-19 Mascara iene» -9 51 61 6 29 9 25 Ne Aas’ NELSON cy een = 46 61 6 25 10 D7 19-21 ne fc) 48 60 6 28 9 24 16-17 Aumale A ki 57 57 6 50 12 26 17-16 35 Pa ee een B05 153 50 65 6 30 Il 27 17-18 SE ES SB en onc 56 59 6 27 11 26 18-19 5 5 Le Cece ee 47 65 6 28 Tul 27 20-22 sRebesaiea is. ta eh eS 42 56 6 30 TLL 26 Y/ 1. Length from snout to vent (in millimetres). 2. Number of scales across middle of body. 3. Longitudinal series of ventral plates. 4. Transverse series of ventral plates. 5. Number of plates in collar. 6. Gular scales and granules on a line from symphysis of chin-shields to collar. 7. Number of femoral pores (right and left). Measurements :— ale 2: b 4, From end of snout to vent . .. . 51 46 52 56 op 3 » worl, 5 6 20 Me 21 21 Ibo @enenGl>s 6 6 6% 59 6 6) dhe 10 14 12 iWiadthvothead auntie we oe cits: 8 vA 9 8 Depthvotheadieemers ss iss) fs) 2. 6 5 7 6 Horewimbymey siih ei a keg i ce LS 15 20 Ly/ Elid Minnis eri ceesir spit yay cre ee? 22 30 26 GOMES EL uA) ear se) LO 13 17 14 Fares: wee is : 89 83 95 87* * Tail reproduced. 1. g. Tamaruth Valley. 2. Qe Ditto. 3. d. Tlemsen. 4, 2. Ditto. 312 418 MR. G. A. BOULENGER ON THE VARIETIES OF Specimens from the Atlas of Morocco have been figured by me in ‘ Novitates Zoologice,’ xii. 1905, pl. ii.; one from Tlemsen, near Oran, is represented on PI. XXIX. vag) JUD Var. VAUCHERI. Like the preceding, this variety differs but little from the typical form, the principal distinction being in the smaller scales. The head is rather short, not more than once and a half as long as broad, and moderately depressed, its depth, in the tympanic region, equalling the distance between the eye and the tympanum. ‘The hind limb reaches the shoulder in males, the elbow of the adpressed fore limb in females; foot once and one-sixth to once and one-third the length of the head. Head-shields as in the typical form, but parietals rather shortge: usually but little longer than broad; series of granules between the supraciliaries and the principal supraoculars complete, or first supraciliary in contact with the second supraocular ; parietal in contact with the upper postocular, except in one specimen; masseteric shield present, except in one specimen; four upper labials anterior to the subocular (one specimen has five on one side, another has three). Dorsal scales finely granular, distinctly keeled, 61 to 731 across the middle of the body, 3 and 4 transverse series corresponding to one ventral plate, 36 to 50 to the length of the head. Ventral plates in 6 longitudinal series. Scales on upper surface of tibia a little smaller than dorsals. 23 to 26% lamellar scales under the fourth toe. Upper caudal scales moderately or strongly keeled, truncate or very obtusely pointed posteriorly. Olive-grey above, with small black spots or reticulations; a more or less defined dark lateral band, bordered above by a whitish streak or series of white spots; no vertebral line or median series of spots ; hinder side of thighs with round white spots ; two series of white, black-edged spots along each side of the tail; belly uniform white in both sexes, or with a few scattered black dots; throat with black dots. Particulars of specimens examined (as in preceding table ):— ale 2. 3. 4 5. 6 lis Se Seer Nae Oe 68 6 26 10 30 18-19 Swe Jr emelbsyaie,, Wee. 70 6 24 10 20 17-18 OF ee aut neopets OL 61 6 30 &) 26 18-17 rime its marcia Mura 67 6 32 il 29 18-17 sf 50 73 6 30 10 24 18-15 50 68 6 30 10 28 15-16 1 76 in the specimen from Algarve, described above, p. 365, ? 57 in the specimen from Algarve. 3 28 in the specimen from Algarve. LAOCERTA MURALIS IN WESTERN EUROPE AND NORTH AFRICA. 419 Measurements :— 3 9. From end of snout to vent . 32 50 % x ee torenlimbinem ein tate aie 19 Length of head SP Pen aes: arco uae sare aees 1S 10 Whidthtotihteadue sy ma head Mae ne a on tee 9 ui Depth of head . ALE ROME Ret Ae 6 5 Horedimbye ae wea) ose Van ae he ee 1G 15 Hind alibi eaton Weugoee ek Ss css, Shwe oe, oO 24 OOH ELAS Laat Se seis- ota, 6 13 [Taj eR OAR hays ade ole arae sy), = Pee gw BIG 83 A female specimen is figured on Pl. XXIX. fig. 9. Var. QUADRILINEATA. Camerano long ago (Atti Acc. Torin. xiii. 1877, p. 87) recorded the occurrence of this form, described by him under the name of JZ. tiliguerta, in the vicinity of ‘Tunis, where two specimens are reported to have been collected by the late Marquis O. Antinori in 1865. These specimens, preserved in the Turin Museum, have been kindly entrusted to me for study by Prof. Camerano, and I find one of them to agree in every respect with specimens of the var. quadrilineata from Sardinia and Corsica, while the other is referable to the Sardinian var. tiliguerta, as pointed out to me by Count Peracca. ‘The association of these two specimens points to Sardinia as their country of origin, and as, since 1865, no one has ever found either of the Sardinian varieties in any part of Tunisia, I cannot help regarding it as highly probable that some mistake has taken place in the labelling of the specimens in the Antinori Collection. It may perhaps be useful to give particulars of the two specimens preserved in the Turin Museum :— Il 2. : 4, 5. ; te g, var. guadrilineata . . 60 63 27 ll 31 22-23 31 6, var. tiliguerta . . . 68 72 27 10 28 23 30 The Lataste Collection contains three specimens, belonging to the var. quadrilineata, received from M. Ed. Taton as from Cyrenaica. ‘These lizards agree entirely with Corsican specimens, and the fact that a Lacerta vivipara bearing the same indication of locality was received along with them makes me regard the information as doubtful, to say the least. I have thus been unable to convince myself of the occurrence of either the var. quadrilineata ox the var. tiliguerta in North Africa, and I consider the report of their presence there as based on erroneous indications of localities. 490) VARIETIES OF LACERTA MURALIS IN W. EUROPE AND N. AFRICA. Tabular Synopsis of the Lepidosis in the Forms described. il 2 3: 4, 5. 6. th 8. Fitypica . 9. . . . 42-62 2-3, 3-4, 4-5 28-45 6 PBB) (Aw 19-30 13-25 WROD 5 9 Sioa 6 Sach) 3, 3-4 36-60 6-8 25-32 8-12 23-31 138-19 WiltOlepis! els) 2) 2 OZ—GO 3-4 42-56 6-8 25-31 10-14 25=35 15222 v.uvauchern . . . . . 61-76 3-4 36-57 6-8 24-32 9-11 24-31 13-19 vy. monticolla . . . . 46-50 3, 3-4 27-34 6-8 28-29 9-10 22-25 17-20 y. hispanica . . . . 50-60 3-4 45 6 26-31 9-10 26-29 15-19 Vv. pityusensis . . . . 55-68 3-4, 4 35-47 26-30 9-14 26-34 NQ=O4 We CLEORES oo lea 6 i CAVESXD 3-4, 4-5 49-62 6 26-30 9-15 30-40 eS y. breviceps. . . . . 46-55 2-3, 3 24-36 6 23-28 8-10 20-25 14-16 vy. brueggemanni . . . d1-65 3-4, 4 o0-02 6 23-27 9=112 22-30 W725 v. nigriventris. . . . do-71 3-4, 4 37-51 6 24-29 8-12 22-99 18-23 v. campestris . . . . 63-65 3, 3-4 33-50 6-8 23-32 8-11 21-32 15-24 To CMe 4 6 6 no BRERUe 3-4, 4, 4-5 35-60 6 24-30 7-12 25=30) 18-29 V.jufolensis . . |. . 68-82 4-5 43-62 6 26-31 8-12 29-38 18-27 v. quadrilineata . . . 56-75 3-4, 4-5 40-57 6 24-32 9-14 26-37 18-28 v.tiiguerta . . . . 65-82 3-4, 4, 4-5 45-58 6-8 25-31 8-12 27-33 21-26 v.bedriage .. . . . 62-78 3-4, 4, 4-5 30-56 6 24-28 7-13 26-39 19-26 Teisandoduie) © 6) ei) GO=76 4, 4-5 33-47 6 24-26 12-13 30-385 23-26 1. Number of scales across middle of body. 2. Number of transverse series of scales corresponding to one ventral plate. 3. Number of transverse series of scales corresponding to the length of the head. 4, Number of longitudinal series of ventral plates. 5. Number of transverse series of ventral plates. 6. Number of collar-plates. 7. Number of scales and granules in a longitudinal series between the symphysis of the chin- shields and the median plate of the collar. 8. Number of femoral pores on each side. 9. Number of lamellar scales under the fourth toe. wo to lo to Pree ho bo bo ht Ce co b 2 Ww | | @o =I ary f : nee ry i i af io ine. Mi) RR a Att ie re ani 422 VARIETIES OF LACERTA MURALIS IN W. EUROPE AND N. AFRICA. Mea, SO:GRE Lacerta muralis. Var. brueggemanni Bedr., ¢. Genoa (p. 383). Natural size. a. Ventral aspect of same. Var. brueggemanii Bedr., ¢. Genoa (p. 383). Natural size. a. Ventral aspect of same. Var. brueggemanni Bedr., 2. Genoa (p. 383). Natural size. 3a. Ventral aspect of same. 4, 5. Var. nigriventris Bp., 6. Old walls of Rome (p. 387). Natural size. 6. Var. nigriventris Bp., ¢. Old walls of Rome (p. 387). Ventral aspect. bo Nw eS eH (J) D.MAcCBETH Sc PHOTO GREEN LACERTA MURALIS. _ xvil—Part iv, No. 10,—October, 1905. 424 VARIETIES OF LACERTA MURALIS IN W. EUROPE AND N. AFRICA. PLATE XXIII. Lacerta muralis. Fig. 1. Var. lil/fordi Gthr., ¢. Mahon (p. 376). Natural size. la. Ventral aspect of same. 2. War. serpa Raf., ¢. Faraglioni, near Capri (p. 396). Natural size. 2a. Ventral aspect of same. 3. Var. brueggemanni Bedr., ¢. Genoa (p. 383). Natural size. 4, War. quadrilineata Gray, 3. Flumini di Quarto, Sardinia (p. 407). Natural size. Var. bedriage Camer., ¢. Corsica (p. 412). Natural size. 3) 5a. Ventral aspect of same. “OS HLAdOV'G STIVYONW VLYsaOoV!) MAN RO INK 726 °F PL OML 7 a a a 426 VARIETIES OF LACERTA MURALIS IN W. EUROPE AND N. AFRICA. PLATE XXIV. Lacerta muralis. Fig. 1. F. typica, ¢. Fontainebleau (Lataste Coll.) (p. 356). Natural size. 2s os .» Dinant, Belgium (p. 356). fs 3 Hua icee o i (ds BD) ” 4, H : Domodossola, Piedmont (p. 355). 5 dD. “A »» High Pyrenees (Lataste Coll.) (p. 358). My 6. War. bocagei Seoane, ¢. Galicia ‘ (p. 362). 55 The _ 96 os 34 3 (p. 362). » 8. . ss 3. Serra de Gerez, Portugal (p. 362). ,, 9. Var. liolepis Bigr., ¢. Valencia, Spain (p. 364). 5 10. i Ait eos 5 » (p. 364). 3 11. Var. monticola Blgr., 2. Serra Estrella, Portugal (p. 365). ,, 19), - a » Galicia (p. 366). 35 13), Bs * »» Spain (p. 366). 14. Var. hispanica Stdr., 9. Spain (p. 368). 9 t Sve. Vol MU FU XM. Ge. TONS. £00 u @ a —_ ¥, PHOTO. J. GREE D. Macaetn, AS) URAI LACERTA PLATT, SORT Fig. 1. © 5 15. VARIETIES OF LACERTA MURALIS IN W. EUROPE AND N. AFRICA. Forma fypica, 3. 2. PLATE XXV. Lacerta muralis. Turin (p. 354). Upper view of head. x 23. Ferriere, Apennines (pp. 393, 378). Upper view of head. X 24. St. Malo (p. 354). Upper view of head. x 23. Voslau, Vienna (p. 354). Upper view of head. x 23. Dinant, Belgium (p. 355). Posterior part of body and base of tail. x 24. High Pyrenees (p. 858). Upper view of head. x 23. Bagnéres de Bigorre (p. 559). Side view of head and anterior part of body. x 2d. Var. bocaget Seoane, ¢. Serra de Gerez (p. 562). Upper view of head. Var. hispanica Stdr., 2. Spain (p. 568). Upper view of head. x 3. Var '. liolepis Bl 29 9 er., 39 99 ( ( . breviceps Blgr., ¢. Italy (p. 379). Upper view of head. X 23. ( ( 5 is (p. 362). Side view of head. g. Valencia (p. 364). Posterior part of body and base Ofgtallaeancoe r. monticola Blgr., 2. Spain (p. 866). Upper view of head. x 23. 5 » (p. 366). Side view of head. %X 23. 3 » (p. 967). Posterior part of body and base Oitall ae 9 % » (p. 368). Side view of head. X 3. p. 879). -Side view of head. x 24. p. 879). Posterior part of body and base of tail. x 23. Trans Loot. Soo, Vol, MM F¢ AMV. D.MACBETH 80, J.GREEN PHOTO LACERTA MURALIS Mirvac % PLATE - ORL 430 VARIETIES OF LACERTA MURALIS IN W. EUROPE AND N. AFRICA. PLATE XXVI. Lacerta muralis. Fig. 1. Var. pityusensis Bosca, 6. Iviza (p. 372). Natural size. 1a. He BS i » (p. 370). Side view of head. x 23. 10. a Pe Ni » (p. 371). Posterior part of body and base of tally << 25. 2. y a 2 » (p. 372). Natural size. 3, 4 % ; » (p-372). Head and anterior part of body. Natural size. 3) v5 D 2 » (p. 372). Head and anterior part of body. Natural size. 6. Var. lilfordi Gthr., ¢. Mahon (p. 876). Natural size. 7. as » 5 1.del Ayre (one of the types) (p. 376). Natural size. CG se BAe, en pis (p. 373). Upper view of head. X 24. CB 3 He s a " (p. 873). Side view of head. X 26. TC. % a. WD . (p. 574). Posterior part of body and base of tail. x 24. elspa /, ir SUN et ” 17 A ie D.MAcBETH sC. LACERTA MURALIS. J.GREEN PHOTO. ene ae Ra AUR e eXONGVS INN, e Ai? oe xvit.—Part iv. No. 11.—October, 1905. po VARIETIES OF LACERTA MURALIS IN W. EUROPE AND N. AFRICA. PLATE XXVIII. Lacerta muralis. _ 1. Var. campestris Betta, ¢. Treviso (p. 391). Natural size. 2. ss ms p> (de BOIL). bs 3. Var. serpa Raf., ¢. Pompeii (p. 396). Natural size. 4, = of Gre eomel(jpn od). 3 Oy OCs » ¢. Liparild.(p.395). Anterior part of body. Natural size. 6. ; ys. Messina *(p.9296)): Bs re vu Uo ; » o. Catania (p. 396). 5 Fe He 8, 8a. oi See laimosalid (ps 40h): 4 ae Py 9,9a. Var. fiifolensis Bedy., ¢. Filfola Rock (p. 404). Natural size. CMH FY AVL fo oe LO0 ae Sian \ we Cra Jie \ ' ' Pal ed 4 4 2 in ai Py <4 oe stad Kea. x sc. D.MAcBeTH J.GREEN PHOTO TA MURALIS > ‘ LACEE GAR Eo MORGVGIGY 434 VARIETIES OF LACERTA MURALIS IN W. EUROPE AND N. AFRICA. PLATE XXVIII. Var. wigriventris Bp., 3 . Lacerta nuralis. Rome (p. 885). Upper view of head. (Doc) aside 2. 2. Var. campestris Betta, ¢. Turin (p. 390). Upper _,, 3 3 2as a Fe 5 (oe SIO), sucks a i 4s 2b. M ss a 5 (De GUN) Ibowre op BS 26 * Fs Fi »» (p. 391). Posterior part of body and base of taille x32 3. Var. serpa Raf., ¢. Rome (p. 395). Lower view of head. X 2. BI. Hs els » (p. 295). Posterior part of body and base of ills Se 4, ee » >> Bosco di Marineo, Palermo (p. 398, table). Upper view of head. xX 4a. - ieee Me 3 a (p. 398, table). Side view head. x 2. 5. Var. filfolensis Bedr., 6. Filfola Rock (p. 402). Side view of head. X 6. Var. quadrilineata Gray, 2. Ajaccio (p. 406). Upper ,, mi x 7. Var. tiliquerta Gm., ¢. Cagliari (p. 409). ~ a 3 x Td. “ sete 2 (p. 409). Side _,, an 8. Var. bedriage Camer., 3. Corsica (p. 411). Upper ,, am 8a. RA C nF soem pes alels): Side ,, i 9. Var. sardoa Peracca, ¢. Gennargentu (p. 415). Upper ,, 5 CMH IY AVI, 1c 7) OC. 00. 5 Lo WNd, 5 * Vane a D.MAcBETH sc. J.GREEN PHOTO. LACERTA MURALIS Dek Avie eK cm (5%) 5. Var. téliquerta Gm., 3. 6. 9 9 o? 7. Var. bedriage Cam., ¢. 8,8a. Var. sardoa Peracca, 9. Var. vaucheri Blgr., 9. 10. Var. bocaget Seoane, °. VARIETIES OF LACERTA MURALIS IN W. EUROPE AND N. AFRICA. PLATE XXIX. Lacerta muralis. len Vier: quadrilineata Gray, ¢. Flumini di Quarto, Sardinia (p. 407). Natural size. 9. Lanusei, Sardinia (p. 407). Natural size. 3. Orezza, Corsica (p. 407). Anterior part of body. Natural size. ¢. Flumini di Quarto, Sardinia (p. 407). Lower view. Natural size. S. Sardinia (p. 409). Natural size. BS (p. 410). Anterior part of body. Natural S1Ze. Corsica (p. 412). Natural size. @. Gennargentu (p. 415). Natural size. Tangier (p. 419). Natural size. Tlemsen, Oran (p. 418). Natural size. Y AAD. Si CMM. ui 100. ans L000, « Sf) > Ea WG J.GREEN PHOTO. BETH SC. D.Mac LACERTA MURALIS TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON (continued). To Fellows. To the Public. AB le Sar’ cle VOLUME XIV. (1896-1898, containing 47 Plates) . Priced 5 O 0 Part l. (1896, containing 2 Plates) . ... .- ina Olan O97, 0 OD SiO » 2. (1896, containing 6 Plates) . . . . . One OnnO: 012 0 »» 8. (1897, containing 9 Plates) . . + = . ph On Loe D Wer lin O », 4. (1897, containing 10 Plates) alu OO) 2070 » D. (1898, containing 2 Plates) . . . = - 5) oe ow) 0 0 » 6. (1898, containing 7 Plates) By Dradlts) 40) 1 0 » 7. (1898, containing 6 Plates) . . . = - eo OR Ob 0 Cea 0 9 0 012 0 . (1898, containing 5 Plates and Title and Index) ,, VOLUME XV. (1898-1901, containing 52 Plates) . Price5 15 6 714 0 Parr 1. (1898, containing 8 Plates) . . . - - ey Oley) Lemley O , 2. (1899, containing 5 Plates)... - + » 0 9 0 012 0 »» 3. (1899, containing 4 Plates) . . . - - POMBO wa) 012 0 » 4. (1899, containmg 4 Plates)... - - SOMO 012 0 »» D- (1900, containing 24 Plates) . . - - - ste oO 3 10 0 » 6. (1901, containing 4 Plates) : : es OL ATS ah ORO » 7. (1901, containing 3 Plates and Title Pinas) On OE.O 0 12°70 VOLUME XVI. (1901-1903, containing 38 Plates) . Price5 8 0 7 4 0 Pant 1. (1901, containing 6 Plates)... +--+ » 018 0 15 re O » 2. (1901, containing 5 Plates) - . . «> Se pss Oen Bi thats} Oo 0 » 98. (1901, containing 9 Plates) uimie cues eens oat Os Lore Le Lag G » 4. (1902, containing 5 Plates) . . - . . eee Onis Oia O 012 0 » 5. (1902, containing 4 Plates) . . . - - ee Olona 6 018 0 ,» 6. (1902, containing 3 Plates) . . . - > ey Obie 9 aD get Beh 0) » 7 (1902, containing 3 Plates) Paes Onis 6, 015 0 » 8. (1903, containing 3 Plates and Title and vinden 013 6 018 0 VOLUME XVII. Parr 1. (1903, containing 5 Plates) » 2 (1908; containing 8 Plates) . . . 3. (1904, containing 13 Plates) . . . + . yin 4, (1905, containing 8 Plates) ” w fon) — — =~ = cS LOG ~ ro) lor) ” CONTENTS. 1V. A Contribution to our Knowledge of the Varieties of the Wall-Lizard, Lacerta muralis, in Western Europe and North Africa. By G. A. Boutznerr, F.R.S., VePZSe (Plates X X11. eX EX) Se pace aol THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Tux scientific publications of the Zoological Soe of London are of two kinds—“ Ereceedings,,: published in an octavo form, and “‘ T'ransactions,’” in quarto. According to the present arrangements, the “ Proceedings” contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended to be published in the “ Proceedings”? by the Committee of Publication. A large number of coloured plates and engravings are issued in the “ Proceedings,” to illustrate the new or otherwise remarkable species of animals described in them. Among such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “ Proceedings” for each year are issued in four parts, in the months of June, August, October, and April, the part published in April completing the volume for the last half of the preceding year. Beginning from January 1901 they form two half-yearly volumes. The ‘Transactions?’ contain such of the more important communications made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. ‘They are issued at irregular intervals. Fellows and Corresponding Members, upon payment of a Subscription of £1 1s. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 26 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1881, if they exceed the value of five pounds, Fellows also have the privilege of subscribing to the Annual Volume of the ‘ Zoological Record’ for a sum of £1 10s. (which includes delivery), but this privilege only holds good if the subscription is paid before the First of December in each year. Such of these publications as are in stock may be obtained at the Society’s Office (8 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. P. CHALMERS MITCHELL, Cctcber, 1905. eteber, 1905 Secretary. TRANSACTIONS OF THE ZOOLOGICAL SOCIETY [OF LONDON Vou. XVII.—Part 5. LONDON. PRINTED FOR THE SOCIETY, SOLD AT THEIR HOUSE IN HANOVER-SQUARB ; AND BY MESSRS. LONGMANS, GREEN, AND CO., PATERNOSTER-ROW, December 1905, * . Price 20s. Taylor and Francis, Printers.) {Red Lion Court, Fleet Street, oi ‘ et TRANSACTIONS OF THE ZOOLOGICAL SOCIETY | OF LONDON. VOLUME I. (1833-1835, containing 59 Plates) . VOLUME _ II. (1835-1841, containing 71 Plates) . VOLUME III. (1842-1849, containing 63 Plates) . VOLUME IV. (1851-1862, containing 77 Plates) . VOLUME _ V., (1862-1866, containing 67 Plates) . VOLUME VI. (1866-1869, containing 92 Plates) . VOLUME VII. (1869-1872, containing 73 Plates) . _. VOLUME VII. (1872-1874, containing 82 Plates) . “VOLUME IX. (1875-1877, containing 99 Plates) . VOLUME X. (1877-1879, containing 95 Plates) . GENERAL INDEX, Vols. I. to X. (1833-1879) VOLUME XI. (1880-1885, containing 97 Plates) . VOLUME XII. (1886-1890, containing 65 Plates) . VOLUME XIII. (1891-1895, containing 62 Plates) . . Price 2d 3) 39 33 * No copies of these volumes remain in stock. To Fellows. Obs 25 & oiler = — Or oS Nt (on (es) 6 eCiaase To the Public. d. Ox 6* o* 6* Continud on page 3 of Wrapper. V. On the Intestinal Tract of Mammals. By P. Cuanmers Mrrcueu, W.A., D.Sc., Secretary to the Society. Received and read June 6th, 1905. [‘Text-figures 1-50. ] ConTENTS. Page Page IMMER ocgoacdscenoggsacadauGnnoo08 438 TV GHGOCED, nou oo00000b0000900080" 484 Morphology of the Intestinal Tract.......... 438 ONOWICGE cooanoccoacondodoovouRe 488 Systematic Description.................... 442 WeWOGED Sonooccoodoncacco0e 0000 488 INKOIOIMPITMUNIMA Sd oo gacogoosanebeoo0 442 IysEcTrvora. MARSUPIALIA. IMMA ooldsoboosocssoss > e98en0 490 DyGlaaAGES oo 2 webaooosescopoonoocKT 445 Macroscelidideoeriieicrisieerie cieikaat 490 IDPEVATOCED oo occ ucbooabo uD DO OoUDOaeO 446 WAGES 5 ccoccbncccpo0nboeaKD000d0 490 IHeaMNGGED concauaqcousaboducogaeEs 446 Soricidsamaepeeeee reer ese aerate: 490 JARECO ENGAGED a5cc co bocogooDGOGoOCNC 449 IDaME@RCHS soedacccoodoccne bo n0C 491 IDiGIEMERAGED! oo cooodcsosongHOUaDUeS 449 C@entebideeres cnccyec sc sarc mares a revere 492 MONODELPHIA. CHIROPTERA. AMAURY oS oGondoansooonnDadadc 453, Moecachiropterany sitll sittin 492 Onyeteropodid eerie tri 453 Microchiropteras-j-rleiii eter ir= 492 PHoripora. (CMON soaonggogocanonedenc 493 WEMCES 665550600 05 009d00usn0000C 454 CaRNIVORA. XENARTHRA. Oimimiiky gooococdgooobbovoudoes og 493 Bradypodidseemerier rity errrest err: 455 Mrichechid spe aneee rere ena 493 Myrmecophagide .............-.--.% 455 IPhocidsehreverschni eee oer 493, Wasypodidee ey acres create citer 458 IMustelidseramer yi iesici tare ce 495 Hyracorpna. [Brocyonid commer tierce terrae 496 Hyracidj ........ Seah achagecs avast ese 460 WiSGES 5 Gados cose ceobM0o ond cl0.D6 498 Sirenta. (Chink cage gos oeacadn Ungosodcucot 498 Mian atid emer s armrests Stevie. 464 TE EaACD cco oooc coo acvodD00NROCE 498 TEINGHAGED. os oacccoadeseoo0o00dco" 465 WinGisntits) Saodovacdsco os 0b oc gopDC 500 IMnnahinGks «ces doosodsocsqc co anocee 465 INGIGEDS) Gece dane ssid os cdobooTS 501 PRoBOSCIDFA. PROsIMLE, WieremnGkS ooojctededoaoccoctonde 466 Nine, Goookdbaoooooveabo woeoss 503 Crracra. Ie PERBNED Ga Sogosanonodnoooevb SoC 503 IeiiinGkS Se booaboogdcone dnoodc 467 - llnGhiGinEs Sa gdanadsosooencdcedouos 503 ARTIODACTYLA. (Chinomiyinceieerterrer retin aoietats 505 Hippopovaraidee) rs.) elles eee ao 469 Callan: gogaccoudooooneesubo0005 507 SCE oo.coocnoo Mauno DOD TUM On OOUD 469 IonisinGy codons pcoonp ano bodo eho 508 ThintgiliGhs; 5 o6ece0a%e cobonon00000 aC 471 Sma. Wamelidwee eee cicirecck ie cise 473 Mapalidoueerciierreirictr eke taser: 508 (Oath a6 ooacumeso deo coodnD oboe 474 (GEE) chddbadtioosonebbeconoces oc 509 LiKE eno c oslon oo womb Moco Gap ccuE ATA @ercoprthecid) 3. -)) se es 511 Ginathdsem masseter rot sisteiers cuayactansret ste ATA Tein) ba pkaono od ddodeocs co0- 512 PERISSODACTYLA. Anthropomorphe ...........-...+-. 5135 AUG) Bavocunecooenune one ooo 05 476 lakspainED 5 oooaocbodoDe oO eC aodnR DO OC 515 RAGE ooaocoececpHopees sec 478 Homology of the Mammalian Caecum ........ 515 LO NCEN ooo boncade cophaoogupaoct 480 Form and Function of the Caecum .......... 522 LODENTTIA, | Secondary Relations between the Proximal and Seiuroideaye pea iola see scien tery 480 Distal Portions of the Intestinal Tract .... 524 Cagstoroid cami atitieiter te der) vite arr 482 Homoplastic Modifications ..............+: 526 Mis omord eter reierntette titi icy via 482 Summary of Systematic Inferences .......... 528 Wipodord oder a coebare t-te reietaiencr os 482 General Conclusion) “fies ee tele e ls ss cl 534 Myoidea. 22-5 +2 52st ee essences even 482 MEIBEL OSC ONCMLCEB ny ia viele! ented cue oleysdeketels}sheke os 535 VOL. XVIL.—Part vy. No. 1.—December, 1905. 3.N 458 DR. P. CHALMERS MITCHELL ON THE INTRODUCTION. ALTHOUGH there is a considerable literature on the Intestinal Tract of Mammals, much of it gained from observations made from animals in the Collection of this Society, it has been impossible hitherto to gain a survey of the modifications presented by the tract in the whole group. Some observers have described with varying detail the complete abdominal viscera of a larger or smaller number of animals. Others have limited their subject and extended their area. But, even in the latter case, it has frequently happened that in one animal the stomach, in another the liver, in a third the cecum has absorbed the attention of the anatomist. The result is that whilst there is a great deal of scattered information, the changing interest of single observers and the diverse points of view of different observers have made the assemblage of facts extremely difficult to compare. From time to time, in the last six or seven years, I have made what use I could of the rich material in the Prosectorium of the Society, and, following the method which I found of service in the case of the Intestinal Tract of Birds, I have limited my observations to a definite set of facts, hoping that the examination of a continuous series by one observer, from one point of view, would yield more information than might be derived from a wider range of work over a smaller number of animals. I have thus been able to cover a very considerable proportion of the existing mammalian types. The figures which illustrate this Memoir are prepared from a much larger number of laboratory drawings. Naturally they are to a certain extent diagrammatic, but I believe that in the main they give a faithful representation of the essential facts. In the systematic part, I have followed closely the classification given by Weber (27). As for the nomenclature of individual animals, I have not attempted to follow the names on which modern systematists are beginning to agree. In nearly every case I have used the name given by my predecessor, Dr. P. L. Sclater, in the ninth edition of the ‘List of the Vertebrated Animals now or lately Living in the Gardens of the Zoological Society of London.’ Moreuo.oey oF THE INTESTINAL TRACT. The work of Toldt (26) chiefly on Man, of Klaatsch (11) chiefly on Reptiles and Mammals and relating in particular to the mesenteries, of Mathes (13) on Amphibia, and my own work on Aves (15) have paved the way for the general comprehension of the morphology of the intestinal tract in Mammals. ‘The primitive gut ran backwards from the stomach to the cloaca suspended from the dorsal wall of the body-cavity by a dorsal mesentery. ‘This tract, ontogenetically and phylogenetically, becomes longer than the straight length between its extreme points, and consequently is thrown into INTESTINAL TRACT OF MAMMALS. 439 a series of folds. The mesentery grows out with these folds, but the presence of adjacent organs, the disturbance due to the outgrowth of the liver, and the secondary relations brought about between different portions of the gut, as the outgrowing loops invade each other’s localities, with occasional “ short-circuiting ” of their blood-vessels, disturb the primitive simplicity. Three definite regions, however, require separate treatment. The first region is the duodenum. In the vast majority of Birds, and in many Reptiles and Amphibia, the portion of the gut immediately distad of the stomach grows out into a long and narrow loop, the proximal and distal ends of which are close together, whilst the loop itself may remain long and narrow, or may develop minor loops on its course. In many of the simpler Mammals (vide the systematic part of this memoir, infra, p. 442) the duodenum is a simple fold, but in others it is thrown into a large series of minor loops which pass insensibly into the next portion of the gut. Taking Mammals generally, they may be said to be characterised by a tendency to less definite separation of the duodenum as a specialised portion of the gut. The second portion is what I have termed Meckel’s Tract (Mitchell, 15). It consists of the part generally known as the small intestines and stretches from the distal end of the duodenum to the cecum or ceca. It is the chief absorbing portion of the gut, and in nearly all Birds and Mammals is the longest portion. It represents, however, only a very narrow part of the original straight gut, corresponding to not more than two or three primitive somites of the body. This narrow area grows out to form the greater part of what is known as the “ pendent loop” in mammalian embryology. Its anterior, or proximal, end is close to the approximated proximal and distal ends of the duodenal loop, and its distal end. passes into the rectal portion. In the embryos of all Birds and Mammals the median point of Meckel’s tract, the part of the loop which has grown out furthest from the dorsal mesentery, is marked by the diverticulum cecum vitelli, the primitive connection of the cavity of the gut with the narrowing stalk of yolk-sac. Naturally, in Birds and Reptiles, where the yolk-sac is of great functional value, this diverticulum is large and important. The main factor of the portal vein, that known as the anterior mesenteric division, runs from this to the dorsal edge of the mesentery near the duodenum, receiving on its way tributaries from the whole course of the small intestines. In perhaps a majority of the families of Birds a remnant of this diverticulum persists throughout life, and in some cases even blossoms out into a second or deuteroblastic growth and activity. In Mammals, no doubt in association with the reduction of the yolk-sac, this diverticulum, which is known as Meckel’s diverticulum, has less importance. It is seldom the starting-point of the largest tributary of the middle mesenteric vein, and its presence in the adult must be extremely rare. Birmingham (2), it is true, states that it occurs in 2°2 per cent. of adult human subjects. I myself have examined the intestinal tract of over two hundred Mammals, including many examples of several species, and although I have always carefully looked for it, have found it in no single instance. It is not present even in aN 2 440 DR. P. CHALMERS MITCHELL ON THE the Monotremes, although, from the large size of the yolk-sac in these Mammals, I had some expectation of finding it. The only possible record of its occurrence in a mammal other than Man occurs in the work of that wonderful old naturalist and observer, Pallas. In his monograph on the Glires (20) he describes and figures the presence of a small diverticulum as the cause of Meckel’s tract, some inches proximad of the ceecum, in a species of Ochotona, and remarks of it that it is peculiar and “ ut in avibus quibusdam,” as in certain birds. I have myself examined two examples of Ochotona rufescens and did not find the diverticulum in either. In Birds, Meckel’s tract falls into minor folds or loops, the characters of which present an interesting series of some systematic importance. In Mammals, Meckel’s tract is much more uniform; it may remain relatively short, or increase enormously in length, but in either case it falls into a fairly symmetrical shape, suspended at the circumference of a nearly circular expanse of mesentery. Where it is short, it is thrown into very simple minor loops; where it is long, these minor loops form a huge convoluted mass, the details of which it is impossible to portray exactly in a diagram. The distal portion of Meckel’s tract in Mammals, as in Birds, is usually nearly straight, and runs back parallel with the caecum or ceca towards the point where the original pendent loop passed into the rectal portion of the gut. But here there is a considerable difference between the condition in Birds and Mammals. In the series of birds which I examined and described (15) I found that the distal portion of Meckel’s tract frequently assumed a certain independence and entered into relations with the duodenum. To a parallel, but not identical, disposition developed in some Mammals, I shall call attention later (infra, p. 524). In Birds the ceca almost invariably arise at the point where the distal end of Meckel’s tract returns to the dorsal line of the primitive straight gut, in fact at the distal root of the pendent loop. In the embryo of Man, and certainly in some other Mammals, the origin of the cecum is some little way from the extreme distal end of the pendent loop. On the other hand, in some Marsupials and a certain number of other forms scattered irregularly through the groups, the cecum arises in the bird- like position. I shall return to the morphology of the mammalian cecum later in this memoir (énfra, p. 515), after the systematic description of the modifications of the intestinal canal in the groups. For the present, it is enough to say that it seems to me indubitable that the mammalian cecum is homologous with one of the paired czeca in Birds. The third portion of the gut extends from the ceca to the cloaca or anus. To avoid ambiguous terms in their application to the portions of this region, I shall refer to it INTESTINAL TRACT OF MAMMALS. 441 as the Hind-gut. It corresponds with a much larger part of the primitive straight gut than the duodenum and Meckel’s tract together, and its proximal portion, in conse- quence, is extremely close to the origin of the duodenum. In the vast majority of Birds, this portion is relatively extremely short, often being only from one-tenth to about one-thirtieth of the whole length of the gut. Birds, in fact, have anticipated the suggestion of Metchnikoff, and have got rid of the large intestines and rectum almost entirely. A certain number of primitive forms, such as Struthious Birds, alone haye retained them. Mammals may be contrasted with Birds as a group in which tke hind-gut is always relatively long, sometimes extremely long, amounting to almost one-half of the total length of the intestinal tract. The Ostrich is the only bird that can be described as mammalian in the character of its hind-gut. Moreover, in Mammals the hind-gut shows a strong tendency to differentiation into regions the characters of which are of systematic importance. The first region is the colon or large intestines. This region presents very varied modifications, and a general feature of these is a tendency to close relationship with the duodenum, in position and sometimes by short-circuiting blood-vessels. In fact this region corresponds with the supra-duodenal loop of Birds in relationships and possibly in function, although it is differentiated from a region of the gut morphologically distinct. Like the whole of Meckel’s tract, it is an outgrowth of a relatively very short portion of the primitive dorsally placed straight gut. The second region of the hind-gut runs from the distal end of the colon to the anus. It may be called the rectum, and although it is some- times lengthened, it is typically not very much longer than the portion of the primitive straight gut which it represents. It is most dorsally placed in the body-cavity, and its proximal end is primitively very near to the duodenum. In some Mammals there is no differentiation of the hind-gut into colon and rectum; in others the differentiation is extreme. A final contrast between Mammals and Birds (in the morphology of the gut) may be stated as follows:—In Birds the portion of the gut known as Meckel’s tract in the course of its lengthening tends to fall into subsidiary loops, the number, position, and patterns of which are of systematic value, whilst the hind-gut tends to remain simple; in Mammals, Meckel’s tract, even although greatly lengthened, does not fall into conspicuously different patterns, but remains a simple coil or series of coils at the periphery of a circular expanse of mesentery, the hind-gut tends to lengthen, and the number, position, and patterns of the subsidiary loops thus formed are of systematic value. 449 DR. P. CHALMERS MITCHELL ON THE SysteMAtTic DESCRIPTION. Class MAMMALIA. Sub-Clas MONOTREMATA. Order MONOTREMATA. Ornithorhynchus (fig. 1). I am indebted to my friend Dr. Wilfrid Harris for the opportunity of examining the intestinal tract of a specimen of Platypus. There is no specialised duodenalloop. Meckel’s tract is nearly circular in disposition, being suspended in very regular minor loops at the circumference of a circular expanse of mesentery. The cecum (fig. 1, C) is of moderate length and slender, but is hollow to the tip. There is a well-marked colic loop (fig. 1, C.L.), and the rectum is straight. The veins show no feature of special interest. A single main tributary of the portal vein drains the whole area of Meckel’s tract and receives a tributary from the colic loop and a posterior mesenteric vein from the rectum, the latter being shown divided in the figure at z. The original position of the mesentery is little disturbed, but in association with the outgrowth of Meckel’s tract and the colic loop a certain amount of twisting has taken place, and in the undisturbed condition the colic loop is attached closely to the duodenal portion. Echidna (fig. 2, p. 444). There is no specialised duodenal loop. Meckel’s tract is circular, as in Ornitho- rhynchus, but is thrown into more complex minor loops, the length of which, as well as the number, is rather underestimated in the diagram. ‘The cecum (fig. 2, C) is very short, much shorter than in Ornithorhynchus, but is hollow to the tip. The colic loop (fig. 2, C.L.) is less differentiated from Meckel’s tract and is itself thrown into a series of minor folds. The rectum is short and relatively wider; it runs a straight course, following the dorsal mesentery. In the figure it is drawn twisted round. The portal vein is diagrammatically simple: a single large vessel receives a number of regularly- disposed tributaries from Meckel’s tract, and is joined posteriorly by a rectal or posterior mesenteric vessel. The patterns of the gut in the Monotremes are not much removed from a primitive reptilian simplicity. At first sight, the figures recall the arrangement in many Birds, but there are two marked differences. In the first place, practically every Bird has a specialised duodenal loop; this loop in the Monotremes, as in Mammals generally, is not separated off from the proximal portion of Meckel’s tract. Secondly, the posterior INTESTINAL, TRACT OF MAMMALS. 443 Hig. 1. Intestinal Tract of Ornithorhynchus anatinus. S. Cut end of duodenum next stomach. R. Cut end of rectum near cloaca. C. Czecum. C.L. Colic loop. x. Cut ends of posterior mesenteric vein. straight portion of the gut (that from the lower portion of “a” to R in fig. 1) corre- sponds only superficially with the rectal portion in Birds, in which, except in a few very isolated forms, such as the Struthious Birds and the Screamers, the paired ceca mark the beginning of the straight portion, whereas in the Monotremes, as in Mammals al ob) > generally, the homological equivalent of the straight portion of Birds extends to “ the single representative of the paired ceca. ‘Thus, in the morphological features underlying the pattern, the bird-like gut of the Monotremes is mammalian. he slight differentiation of the colic loop is as primitive in Mammals as the absence of a specialised 444 DR. P. CHALMERS MITCHELL ON THE supra-duodenal loop in Birds. There is no special resemblance between the gut-pattern of the Monotremes and those of other mammalian groups. ‘There is a general Intestinal Tract of Echidna hystrix. 8. Cut end of duodenum next stomach. R. Cut end of rectum near cloaca. C. Cecum. C.L. Colic loop. resemblance, due rather to absence of specialisation than to the presence of definite characters with the gut-patterns of the Insectivora. INTESTINAL TRACT OF MAMMALS. 445 Sub-Clas MARSUPIALIA. Order MARSUPIALIA. Sub-Order POLYPROTODONTIA. Family Dipetpuyipa. Didelphys virginiana (fig. 3). Fig. 3. Intestinal Tract of Didelphys virginiana. §. Cut end of duodenum next stomach. R. Cut end of rectum near anus. C. Cxecum. VOL. XV1I.—Part Vv. No. 2.—December, 1905. 30 446 DR. P. CHALMERS MITCHELL ON THE In the Opossums the duodenal loop is well differentiated. Meckel’s tract is a nearly regular circular stretch of gut at the circumference of its mesentery, and is thrown into only comparatively slight minor loops, but at its distal extremity, towards the recurrent part of the circle, it is nearly straight, as in most Birds and Mammals. The single ceecum is very large, wide, and long (C), and marks the passage of the gut into the hind-gut, which is not differentiated into colon and rectum and is only slightly expanded in length, being thrown into a few minor loops. The portal system is extremely simple, consisting of a single anterior mesenteric branch which curves round Meckel’s tract, receiving tributaries from the duodenum and from the circular expanse. One of these tributaries, specially enlarged, drains the cecum, and the rectal or posterior mesenteric vein joins the anterior mesenteric at the edge of the mesentery. Probably this pattern is as nearly primitive as is to be found amongst Mammals. The hind-gut with its mesentery is represented in the figure (fig. 3) as being twisted over, the specimen having been put in that position for convenience in drawing; but the rectal vein marks the line of the original straight dorsal mesentery, and it will be seen that the gut displays in a simple form a duodenal fold, a circular Meckel’s tract, and a hind-gut in almost diagrammatic simplicity. The posterior mesenteric vein is in its primitive condition as the great longitudinal vein of the gut, whilst the anterior mesenteric departs from its primitive simplicity only in so far as its original main termination on the yolk-sac has dwindled, with the disappearance of the yolk-sac, into a simple tributary, whilst the main vessel appears to follow the contour of the gut. Family Dasyuripa. Dasywrus viverrinus (fig. 4), Phascogale penicillata. In these small carnivorous forms the duodenal fold is little differentiated, and Meckel’s tract is a very simple circular coil thrown into evenly disposed minor loops. There is no trace of a ceecum, and the hind-gut, which is undivided into colon and rectum, is relatively short and simple. ‘The portal system appears to correspond closely with that in the Opossum, except for the absence of the czecal branch. It is impossible not to regard the presence of a cecum as a primitive feature in Mammals, and hence in the Dasyuride, as in most of the Insectivora, which the Dasyuride resemble in habits, the absence of the cecum must be regarded as secondary. It is equally plain that the gut is degenerate from an extremely simple type. Family Nororyceripa. I have not been so fortunate as to examine an example of this rare type. Family Peramenipa&. Peragale lagotis (fig. 5, p. 448). The gut of the Bandicoot is less degenerate than that of the Dasyuride. The duodenal loop is well separated; Meckel’s tract is relatively short, but has the normal INTESTINAL TRACT OF MAMMALS. 447 Intestinal Tract of Dasyurus viverriius. S. Cut end of duodenum next stomach. %. Cut end of rectum near anus. 448 DR. P. CHALMERS MITCHELL ON THE disposition. The single cecum is extremely large, although short, and passes directly into the very wide hind-gut. The hind-gut is differentiated into a distinct but short colic loop (C.L.) and a wide rather short rectum. In this case also, partly on account of the condition of the specimen, I have only scanty notes of the veins, the general course of which seems to correspond with that of the similar structure in Didelphys. Fig. 5. Intestinal Tract of Peragale lagotis. 8. Cut end of duodenum near stomach. H. Cut end of rectum near anus. C. Cecum. C.L. Colic loop. The Polyprotodont Marsupials show a primitive type of intestinal tract in which the main features of the mammalian gut are indicated with diagrammatic simplicity. INTESTINAL TRACT OF MAMMALS. 449 Family EPANORTHID®. I have not been able to examine any example of this family Sub-Order DIPROTODONTIA. Family PHAscoLaRcTID&. I have not been able to examine a Koala or a Wombat. Flower (5), however, has described the alimentary canal of Phascolomys, and has figured the cecum, the figure being reproduced by Oppel (16, p. 567). According to him, the whole intestinal tract is long, in relation to the vegetable food, and the hind-gut is longer and more voluminous than the fore-gut. The proximal portion of the colon is sacculated, and Flower identifies as the true ceecum a forwardly projecting, wide, but short, sac of the colon. In addition to this, however, he figures and describes a second cecum, larger than the first, but very narrow and vermiform, although provided with a separate aperture into the colon. The most obvious interpretation of the structure is that the Wombat has a pair of ceca, each member of which is degenerating, the normal cecum being short and thin-walled, the second member of the pair being elongated and probably glandular. It has been repeatedly referred to as the vermiform appendix of the Wombat, but it is obviously not that portion of the normal cecum. Family Paataneprip2. Phalangista vulpina (fig. 6, p. 450), Dromicia nana, Petaurus sciureus (fig. 47, Il, p. 518), Bettongia penicillata (fig. 7, p. 461), Bettongia cuniculus, Petrogale penicillata, Macropus bennetti (fig. 8, p. 452) I follow Weber (27) in not attempting to separate the Phalangers from the Kangaroos. In Phalangista vulpina (fig. 6) the duodenum is hardly separated as a specialised loop. Meckel’s tract is relatively short, occupying only part of the circumference of a circle. The cecum is extremely long and wide, and passes directly into the hind- gut, the straight termination of Meckel’s tract entering the continuous cecum and hind-gut almost at a right angle. The hind-gut shows a very definite colon subdivided into a series of minor loops, whilst the rectal portion is short and straight. The anterior mesenteric vein appears to have as its chief tributary the large cecal vessel, whilst it receives minor tributaries from the duodenum and from Meckel’s tract, and is joined by a large vessel which represents the colic branch and the posterior mesenteric or rectal vessel. It is plain that the primitive simplicity of the portal system has been distorted in appearance by the disproportionate growth of the cecum and the colon, but the underlying type is not dissimilar to that in Didelphys and the Polyprotodonts. ‘The intestinal pattern in Dromicia is essentially similar, but Meckel’s tract is more nearly circular, and the hind-gut is much shorter and very wide, with the result that the colon is represented by a single wide loop. ‘he condition in Petaurus resembles 450 DR. P. CHALMERS MITCHELL ON THE that in Dromicia closely, with the peculiarity, to which I shall call attention later, that the distal end of Meckel’s tract is inserted at an acute angle into the junction of the cecum and hind-gut, so disposed as to suggest the former existence of a pair of ceca (fig. 47, Il, p. 518). Intestinal Tract of Phalangista vulpina. Letters as in fig. 5. In Bettongia penicillata (fig. 7) there is no specialised duodenum. Meckel’s tract is relatively short and with slight minor loops; it enters the conjoined very wide, rather long cecum and hind-gut at an acute angle, with the same suggestion as in Petaurus. The hind-gut shows a marked differentiation into an enlarged colon and a straight rectum. ‘The portal system is diagrammatically simple, the vessel to the colon being distinct from the posterior mesenteric or rectal branch. INTESTINAL TRACT OF MAMMALS. 451 Bettongia cuniculus shows a pattern essentially similar, but both the cecum and the hind-gut are relatively shorter and wider. Intestinal Tract of Bettongia penicillata. Letters as in fig. 5. Petrogale penicillata, of which I examined a fcetus, was almost exactly like Bettongia penicillata, but Meckel’s tract was relatively longer. 452 DR. P. CHALMERS MITCHELL ON THE In Macropus bennetti (fig. 8) the pattern is essentially similar to that in Bettongia, with two exceptions. The less important point of difference is that the colon (C.L.) Fig. 8. Intestinal Tract of Macropus bennetti. 8. Cut end of duodenum near stomach. R. Cut end of rectum near anus. C. Normal cecum. Ce. Vestigial cecum. C.L. Colic loop. passes insensibly into the rectum, both portions of the hind-gut forming a continuous series of long subsidiary loops. The presence of a separate posterior mesenteric vein, INTESTINAL TRACT OF MAMMALS. 453 however, shows that only a relatively short part of the hind-gut is to be regarded as rectum. ‘The second point of difference is more important. The distal straight end of Meckel’s tract joins the conjoined caecum (C) and hind-gut at an acute angle, as in Petaurus and Bettongia. But exactly opposite the normal cecum there is a second cecum (C2), much shorter than the first, but quite distinct. Some years ago I examined the ceca of all the Kangaroos that died at the Gardens, and was able to satisfy myself that this relic of paired ceca is quite a common occurrence. I believe it to be of considerable theoretical importance. The Diprotodont Marsupials are typically herbivorous forms, and, especially in the larger Kangaroos, show an approximation in general character to the herbivorous Eutheria. This resemblance or convergence displays itself in the intestinal tract, first by a great increase in length in the portion known as Meckel’s tract, an increase which, although it is not sufficiently represented in the figures, is never so great asin the Kutherian type. Next, it reveals itself in the complexity of the hind-gut. Here, also, the essential difference between the convergent resemblances of a homoplastic nature and resemblances showing affinity is manifest. In both Marsupial and Eutherian herbivorous animals the hind-gut is lengthened, and the lengthening is associated with the fundamental mammalian division of the hind-gut into colon and rectum. But in no case is there the slightest approach to the special peculiarities assumed by the long colon of Kutherian Herbivora. The gut-pattern remains that of the simplest Mammalian type, and differs in no important morphological respect from that of the carnivorous and insectivorous Polyprotodonts. The gut of Marsupials shows, under the differences associated with different habit, a fundamental identity. It is simply a diagram of the primitive mammalian gut, and this primitive character is further accentuated by the abundant traces in the Diprotodonts of an originally paired state of the ceca. Sub-Clas MONODELPHIA. Order TUBULIDENTATA. Family ORYCTEROPODID#. I have not had the opportunity of examining Orycteropus. Flower (5), however, dissected a specimen that died in the Gardens. He described the whole gut as being extremely long, the duodenum and Meckel’s tract together measuring thirty-two feet, whilst the hind-gut was capacious and seven feet in length. ‘There was a long single cecum, with an expanded globular extremity. VOL. xvit.—PaArtT v. No. 3.—December, 1905. 3 P 454 DR. P. CHALMERS MITCHELL ON THE Order PHOLIDOTA. Family Manip#. Manis tricuspis (fig. 9). The duodenum in Manis is a well-marked distinct loop. Meckel’s tract is very long Intestinal Tract of Manis tricuspis. S. Cut end of duodenum near stomach. R. Cut end of rectum near anus. (The hind-gut is marked by longitudinal striz.) and composed of very thickly-set minor coils regularly disposed save towards the distal end, where the circular course of the tract nearly completes itself. There is no trace INTESTINAL TRACT OF MAMMALS. 455 of a cecum. The hind-gut is relatively short, very wide and strongly marked with longitudinal strie. It shows no trace of division into colon and rectum. The portal system is simple, consisting of a single anterior mesenteric vein curving round Meckel’s tract and receiving numerous tributaries, first a duodenal branch and then many branches from the tract. There is a single rather large posterior mesenteric or rectal vein. The pattern of the intestinal tract of Manis is of striking, almost reptilian, simplicity. The peculiarities are the loss of the cecum and the longitudinally striped hind-gut. But for these two features and the loss of any special orientation of the anterior mesenteric vein with regard to the yolk-sac diverticulum, the pattern would closely resemble that found in many primitive Birds. ‘There is very little that is distinctively Mammalian about it. Order XENARTHRA. Family Brapypopips. Bradypus infuscatus (fig. 10, p. 456). The duodenum forms a distinct loop of rather wide calibre. Meckel’s tract is unusually short, occupying only a small portion of the usually circular area of mesentery. It is separated from the wide hind-gut by a vestigial cecum (C), and opposite this there is (C2) a still smaller vestige of another caecum, showing that these structures were originally paired. The hind-gut is differentiated into a long, wide colon and a shorter, nearly straight rectum. The position of the colon is somewhat unusual, inasmuch as it occupies a considerable part of the mesenterial expanse generally occupied by the posterior portion of Meckel’s tract. The mesenterial out- growth from which the greater portion of the gut is suspended is unusually long and narrow. ‘The portal system consists of a main anterior mesenteric vein which distally curves beyond Meckel’s tract and drains part of the colon. ‘There is a long and normal posterior mesenteric vessel. Family Myruecoruacipe. Tamandua tetradactyla (fig. 11, p. 457). In this Anteater there is a loop which may be recognised as the duodenum, although it is not well separated from the proximal portion of Meckel’s tract. Meckel’s tract is long, thrown into a regular series of minor loops, the whole area occupying the circumference of a nearly complete oval expanse of mesentery. ‘The usual cecum is absent, but on the other side of the gut there is a vestige (C2) of the other member of the primitive pair. The hind-gut is very short and straight and cannot be regarded as divided into colon and rectum. ‘The portal system is simple: an anterior mesenteric vein, with the usual curved configuration, drains the duodenum and the whole range of Meckel’s tract, and there is a well-marked posterior mesenteric vein. 9 ¢ dP 2 456. DR. P. CHALMERS MITCHELL ON THE 8. Cut end of duodenum near stomach. R. Cut end of rectum near anus. C. Caecum. C2. Vestigial execum. C.L. Colic loop. INTESTINAL TRACT OF MAMMALS, 457 J PAS Intestinal Tract of Tamandua tetradactyla. S. Cut end of duodenum near stomach. R. Cut end of rectum near anus. C2. Vestigial caecum; the normal cecum is absent. 458 Dk. P. CHALMERS MITCHELL ON THE Family Dasypopipa. Tatusia keppleri (fig. 12), Xenurus unicinctus, Dasypus villosus (fig. 13), Dasypus minutus. In Tatusia (fig. 12) the duodenum is a distinct loop. Meckel’s tract is very long Intestinal Tract of Vatusia keppleri. S. Cut end of duodenum near stomach. R. Cut end of rectum near anus. C. Caecum. Ce. Vestigial cecum. C.L. Colic loop. INTESTINAL TRACT OF MAMMALS. 459 and thrown into elaborate subsidiary loops, but ultimately pursues a straight course along the distal portion of its oval expanse of mesentery to join the hind-gut not far from the duodenum. It enters the hind-gut between two very short but symmetrically placed ceca (C and C2). The hind-gut is differentiated into a colic loop (C.L.) and an expanded rectum. The anterior mesenteric vein curves round Meckel’s tract in the normal fashion and is joined by a large posterior mesenteric vein. Fig. 13. L - ~ CAPR io ay , nS, = Intestinal Tract of Dasypus villosus. Lettering as in fig. 12. In Dasypus villosus (fig. 13) the pattern of the gut and the disposition of the portal veins are practically identical with the condition in Tatusia, although the duodenum is less separated from Meckel’s tract. The ceca, however, are definitely paired (C and C2). The same pattern with paired ceca occurs in Dasypus minutus, whilst the condition in Xenurus unicinctus is again similar, except that the ceca, although paired, are vestigial. 460 DR. P. CHALMERS MITCHELL ON THE The order Xenarthra, containing the Sloths and Anteaters, shows a definite and characteristic pattern of the intestinal tract. The typical mammalian characters, the rather incomplete separation of the duodenum, the simplicity of .Meckel’s tract, and the division of the hind-gut into colon and rectum, are all present, without any special elaborations. ‘The typical veins are similar to those in the Marsupials, an anterior mesenteric curving round Meckel’s tract, and a posterior mesenteric or rectal vein. The inference seems to be obvious that in this respect the Xenarthra have retained a primitive arrangement, possibly the arrangement that existed in the Reptilian stock which gave rise to Mammals. The special feature of the gut of the Xenarthra is the actual presence of paired ceca in the Dasypodide and the presence of traces of an original paired condition in the Myrmecophagide and Bradypodide. As similar traces of a paired condition occur amongst the Marsupials, I think it a probable inference that in the primitive Mammal the ceca were paired. The orders Tubulidentata, Pholidota, and Xenarthra are usually grouped together as Edentata, although there are no very definite characters uniting them. The evidence afforded by the patterns of the intestinal tracts gives no further reason for regarding the group as coherent. I have not myself examined Orycteropus, but as that animal is stated to possess a long gut and good cecum, it obviously differs in this respect from other Edentates. Any resemblances between Manis and the Xenarthra are just as clear between the Xenarthra and many Marsupials; they depend merely on the fact that in all these animals the pattern of the gut is primitive, and the common possession of a primitive condition of a set of organs, although it may be of convenience for artificial classification, gives no evidence that can be applied to the-construction of a classification based on affinity. Order HYRACOIDEA. Family Hyracipa. Hyraa capensis (fig. 14). The intestinal tract of this animal is very peculiar. The duodenal loop is definite, more definite than in the figure, being longer and rather narrower. The next area of the gut is easy to interpret, if for the moment consideration be omitted of the peculiar expansion marked A.C, in the figure. Neglecting that, Meckel’s tract may be described as fairly normal, thrown into regular minor coils that lie at the circumference of an approximately circular outgrowth of mesentery, corresponding to the pendent loop of mammalian embryology. At its distal extremity this passes into the hind-gut, the latter occupying, as in many Mammals, a portion of the mesentery of the pendent loop. The transition from Meckel’s tract to the hind-gut is marked by a pair of cxca, which again, if we still omit consideration of the cxcal expansion on Meckel’s tract, we should interpret at once as a well-developed pair of ceca, homologous in position and relations with the less developed pair of cxca of the Xenarthra. The hind-gut is INTESTINAL TRACT OF MAMMALS. Fig. 14. t f f; EX Fee. 5 Intestinal Tract of Hyraw capensis, S. Cut end of duodenum near stomach. R. Cut end of rectum near anus. AC. Accessory cecum. ' C.2 } Paired ceca. C.L. Colie loop. VOL. xvit,—part Vv, No. 4.—December, 1905, 3Q 461 462 DR. P. CHALMERS MITCHELL ON THE subdivided into a distinct colic loop (C.L.) and a long straight rectum. ‘The portal system consists of an anterior mesenteric vein which arises at the periphery of Meckel’s tract, in a fashion similar to the most common disposition in Birds; it runs straight through the suspensory mesentery of Meckel’s tract, receiving tributaries on either side, one of which is the duodenal vein. There is a middle mesenteric vein draining the distal portion of Meckel’s tract and the paired ceca, and this is joined posteriorly by a colic vein and a large posterior mesenteric or rectal vein. The difficulty in the interpretation of this intestinal tract is the presence on Meckel’s tract of a large expansion, which I propose to call the accessory cecum, and which is marked A.C. in fig. 14. This structure has been described and figured frequently, and, so far as 1 am aware, has always been set down as the normal Mammalian cecum, the pair lower down being regarded as appendages peculiar to Hyraz. I think that a comparison of the figure of the tract of Hyrax with the figures of the intestinal tracts of the other Mammals described in this memoir is enough to suggest at once that the paired ceeca are the representatives of the normal cecum of Mammals. Their relations to the rest of the alimentary canal, and, in addition, the evidence that I have been able to adduce from the conditions in Marsupials, Sloths, and Anteaters as to a paired condition of the ceca being a primitive mammalian character, make these organs in Hyrax fall into line with the general disposition of the mammalian intestinal tract. Moreover, apart from the abnormality of the position of the unpaired cecum, were it to be regarded as the representative of the normal mammalian cecum, there is nothing about its structure or relation to its entrance and exit making it similar to any Mammalian cecum with which I am acquainted. The character of the intestinal canal before and after the accessory cacum has been referred to by George (8), who, in his elaborate monograph on Hyrax, gave a figure of the stomach, part of the alimentary tract, and the unpaired cecum seen from the outside and after being opened. Following Cuvier, he took for granted that the unpaired cecum was the representative of the normal mammalian cecum, and noticed, as a remarkable circum- stance, that the paired ceca resembled those of Birds. He states correctly that the small intestines, both anterior and posterior to the unpaired cecum, have the interior surface increased by a series of folds of the mucous membrane, these folds, however, unlike the normal valvule conniventes, being longitudinally directed. On the other hand, after the paired czca, the intestinal tract becomes wider in calibre and with a smoother internal surface. Dr. Seligmann (the Society’s pathologist) was kind enough to prepare sections for me from pieces selected from the gut proximal to the unpaired cecum, between the unpaired and the paired ceca, and distal to the paired ceca. Although the material was somewhat macerated, not having been preserved for “nicroscopical purposes, it was sufficiently good to show the general nature of the wall of the gut in these different regions. In the anterior portion the cavity was con- siderably reduced by the folds in the walls, and by closely set extremely long villi, as, INTESTINAL TRACT OF MAMMALS. 463 indeed, Meckel had originally described. Distad of the unpaired cecum, the less elaborate thickening of the mucous membrane left a wider central cavity. The longitudinal folds were much more regularly arranged than in the anterior portion, and the villi were less numerous and much shorter, in considerable portions being practically absent. What I take to be the colon proper, however, the region distad of the paired ceca, shows a condition which I have not found described by any of my predecessors in this field. The calibre of the central cavity is relatively much greater, and although, as George stated, the internal surface is relatively smooth, there are present numerous, but widely scattered, extremely long papille which extend far into the cavity of the gut, and in the sections frequently lay doubled up. Notwithstanding the very large amount of work that has been published on the alimentary canal, I am disposed to think, and examination of Oppel’s (16) careful compilation of the literature confirms me, that we have not sufficient systematic information as to the characters of the lining membrane of different regions of the mammalian alimentary canal, to draw morphological conclusions from the evidence of particular cases. In the case of Hyraa, there is a gradual change from the duodenum towards the paired ceca, the change consisting in a reduction of the complexity of the mucous membrane and an increasing preponderance of the glandular element over the villi, whilst there is an abrupt and greater change behind the paired ceca, of which the most marked feature is the sudden appearance of the very long scattered papille. So far the microscopical structure, if it does not give active support, at least is not opposed to my view that the paired ceca are the representatives of the normal mammalian cecum. I have no idea if any significance from the systematic point of view is to be attached to the presence of the very long villi in the colon; there is practically no recorded information on this subject. Of my own knowledge I can say that I have noticed the absence of colic villi in a good many Primates and Carnivora and have not found them in these groups, whereas I have frequently found them in Ungulates and Rodents. The general pattern of the intestinal tract in Hyrax, however, suggests no affinity with the patterns exhibited by Rodents and Ungulates. The simple duodenum, the nearly circular Meckel’s tract, and the hind-gut divided into a simple colon and rectum merely conform to the fundamental mammalian plan. The presence of the paired ceca, on my view that paired ceca are a primitive mammalian feature, does not help us with the placing of Hyrax. ‘The most striking resemblances are resemblances with, on the one hand, the patterns displayed by the Xenarthra among the Edentates, and, on the other hand, with the Manatee among the Sirenia. But it must be remembered that the common possession of a primitive simplicity is no guide to affinity. Otherwise, it might be said at once that, in the character of its intestinal tract, Hyrax stands nearest to the Edentates, Xenarthra, Marsupialia, and Sirenia. 464 DR. P. CHALMERS MITCHELL ON THE Order SIRENIA. Family Manatipa. Manatus inunguis (fig. 15). In the Manatee the duodenal loop is not differentiated from the proximal loops of Intestinal Tract of Manatus inunguis. S. Cut end of duodenum near stomach. R. Cut end of rectum near anus, C. 2 Co, | Paired caca. : ae X. Cut and separated ends of mesenteric vein. INTESTINAL TRACT OF MAMMALS. 465 Meckel’s tract. Meckel’s tract is long and suspended at the periphery of a nearly circular expanse of mesentery. It passes into the hind-gut by a short straight portion at the point where the ceca occur. The caca are most interesting. There is a widely expanded common portion terminating in a pair of short conical ceca with the points turned proximally, whilst distally the unpaired portion passes into the hind-gut. I do not think that there can be any doubt but that these paired structures represent the primitive paired ceca of Mammals. It is specially interesting to notice that in the Dugong and in Rhytina there is a single cecum (Owen, 17). ‘The hind-gut is very long, almost equal in length to the fore-gut, and is differentiated very slightly into what may be regarded as colon and rectum. ‘The portal system is simply a single long trunk which may be regarded as the posterior mesenteric vessel, giving off branches to Meckel’s tract and to the hind-gut. 7 Family Haticorip&. I have had no opportunity of examining a Dugong, but a good deal of information regarding the alimentary canal can be obtained from Owen (17), Rapp (22), and Home (9). ‘The general character is similar to that of the Manatee. The tract consists of an enormously long, much-looped, combined duodenum and Meckel’s tract, and a still longer and more capacious hind-gut. At the junction of the two there is a fairly large, conical, thick-walled cecum. Family Rayrinipa. Naturally I have had no opportunity of examining /hytina; Steller (25) himself, however, has given a general description of the intestinal canal, from which it appears that it was enormously long, with a distinct hind-gut, the proximal portion of which, as well as the large single cecum, was sacculated. The pattern of the intestinal tract in the Sirenia, at least so far as Wanatus is a guide, is extremely interesting. In the drawing (fig. 15), for the sake of convenience in arranging the actual specimen, the mesentery is represented as twisted on itself in the region where the hind-gut joins Meckel’s tract. In reality, the gut is simply two large loops suspended from the dorsal mesentery, one loop representing the duodenum and Meckel’s tract, the other representing the colon and rectum. ‘The paired ceca mark the point where each loop returns to the primitive dorsal line. If the ceca were long, the gut would closely resemble that of the Ostrich, one of the most primitive of birds, and it is difficult to regard the Manatee in any other way than as having retained a primitive Reptilian pattern of the alimentary tract. ‘The descriptions by older authors, which I have cited, make it plain that, except as regards the cecum, the intestinal tract of the Dugong and of Steller’s Sea-Cow were essentially like that of Manatus. The single caecum in its allies links this Reptilian condition with the condition most usual in Mammals. Amongst other Mammals, the most nearly allied patterns are those displayed by /yrax and by the Proboscidea. 466 DR. P. CHALMERS MITCHELL ON THE Order PROBOSCIDEA. Family Evernantipa. LHlephas africanus (fig. 16). By the kindness of my friend Mr. R. H. Burne, of the Royal College of Surgeons Museum, I have been able to examine the intestines of a young African Elephant. The duodenum is a distinct loop. Meckel’s tract is relatively short and thrown into a Fig. 16. a. SS Rab &S Intestinal Tract of Elephas africanus. S. Cut end of duodenum near stomach. R. Cut end of rectum near anus. C. Caecum. series of minor loops up to the recurrent terminal portion where it passes at an acute angle into the continuous caecum and proximal portion of the hind-gut. The actual tubular edge of the small intestine passes some little way into the cavity of the cecum INTESTINAL TRACT OF MAMMALS. 467 and large intestine, suggesting a primitively paired condition of the ceca. The actual cecum is large, long, and wide. ‘The hind-gut is wide in calibre and relatively very long. It is thrown into a continuous series of folds in which a colon and rectum, not well separated, may be distinguished. The portal system consists of an anterior mesenteric vein draining Meckel’s tract round which it curves, and receiving a large duodenal branch. ‘There is a strong posterior mesenteric vein, draining colon and rectum. The pattern displayed by the Proboscidea is of primitive simplicity, consisting of the duodenum, the pendent loop transformed into Meckel’s tract, and a second large loop forming colon and rectum. ‘The most striking character is that the hind-gut is very little shorter than the fore-gut. There is no trace of any of the special Ungulate characters. Obviously this pattern is very closely similar to that of Manatus. If it be remembered that the Dugong and Rhytina have only a single cecum, then it is plain that in the pattern of the gut the Sirenia and Proboscidea are not far apart and that neither group shows a trace of the Ungulate specialisation. Order CETACEA. Sub-Order ODONTOCETI. Family DeLpHiniD®. Phocena communis (fig. 17, p. 468). This is the only example of the Cetacea of which I have been able to dissect the alimentary canal myself. There is no separate duodenum, the complicated stomach being immediately succeeded by an excessively long gut disposed in a series of very regularly arranged coils forming an enormous mass, which tapers gradually to the hind-gut. There is no distinction between the hind-gut and Meckel’s tract and there is no cecum. The whole length of the gut is suspended by dorso-ventral mesentery stretching in almost primitive simplicity fore and aft. A single large mesenteric vein runs along this, receiving, at almost diagrammatically regular intervals, a series of veins from the coils. ‘The calibre of the gut, as in most fish-eating animals, is rather small, very regular throughout the length of the gut, and the wall is relatively thick. The simplicity of the arrangement of the gut in Phocwna appears from the recorded observations to be present in at least most of the other Odontoceti, and, at first sight, it would seem to be a fair interpretation that it was a primitive feature and that it pointed to a separation of the Cetacea and other Mammalia before the special characters of the mammalian gut had been formed. I have not myself had the opportunity of examining the intestinal tract of any of the Mystacoceti, but they are described as possessing both duodenal and colic loops, and a single forwardly directed cecum. Moreover, in many of them the gut is relatively short, and from the descriptions appears to have the normal mammalian peculiarities. I interpret the 468 DR. P. CHALMERS MITCHELL ON THE Intestinal Tract of Phocena communis. S. Cut end of duodenum near stomach. R. Cut end of rectum near anus. INTESTINAL TRACT OF MAMMALS. 469 simplicity of the gut of Phocena, therefore, as being secondary. I hope that some anatomist who has an opportunity will make a careful examination of the gut-pattern in some of the Mystacoceti, as it is quite possible that thereby some clue to the affinities of the group would be given. Flower (5) states that in the Whalebone Whales a short conical cecum is present, and that, unlike the Toothed Whales, the former display a differentiated hind-gut, with a distinct colic loop. Order ARTIODACTYLA. Sub-Order NON-RUMINANTIA. Family HrpropoTaMip#. I have had no opportunity of dissecting a Hippopotamus, but Flower (5) and Garrod (6) have published some information, from which it appears that the general pattern is not unlike that of the Swine. The duodenum and Meckel’s tract, which are not sharply marked off from each other, are enormously long and with a narrow calibre. ‘There is no cecum. The hind-gut displays first a series of irregular loops, corresponding in position to the spiral coil of the Pig, and somewhat similar to that when the latter, as in the figure (fig. 18), has been dissected out; then follows a “* transverse and descending colon,” and then the rectum. ‘The literature, however, on this subject is scanty, and more information is much wanted. Family Suip#. Sus scrofa (fig. 18, p. 470), Phacocherus wthiopica, Babirussa babirussa, Porcula salvania. In the Domestic Pig the duodenum is a well-separated loop or occasionally a double loop. Meckel’s tract is extremely long and disposed in a number of closely-packed simple folds. The cecum is single, wide, relatively short and tapering to a point. The hind-gut exhibits, first of all, an enormous double loop, more or less coiled into a spiral. In the drawing, which was made from an embryo Pig, the relative size and capacity of this loop is much less than in an ordinary adult example. The growth of the loop has caused the rupture of the mesentery, the loop with its very large blood- vessel being supported by a special fibrous band with some muscular fibres. In the figure the thinner distal limb of the loop is represented as partly dissected away from the main mass, and the edge of the mesentery supporting the rectum is torn. There is a single, large, anterior mesenteric vein draining the duodenum and Meckel’s tract. ‘The cecum is drained by a separate branch, which enters the anterior mesen- teric vein near the edge of the mesentery. The huge vessel from the great loop of the hind-gut is immediately proximal to the last-named vessel. ‘The posterior mesenteric or rectal vein is slight and insignificant; in the diagram, owing to the displacement caused by dissecting out the great loop, it is represented as cut across at X and X. VOL. XVII.— Part v. No. 5.—December, 1905. aR 470 DR. P. CHALMERS MITCHELL ON THE In Phacocherus ethiopicus, the Wart-Hog, the pattern is fundamentally similar to that displayed by the Common Pig. The duodenum is not well separated from Meckel’s tract, but the latter is exactly like the corresponding area in the Pig. The cecum is more globular, and does not taper to a point. The anterior portion of Fig. 18. \W) SP i X iR Intestinal Tract of Sus scrofa. ie q ey, LF q S. Cut end of duodenum near stomach. R. Cut end of rectum near anus. C. Caecum. 8.P. Spiral coil. X. Cut ends of posterior mesenteric vein. the hind-gut is an enormously roomy loop, the proximal limb of which has a much wider calibre than the distal, and which is partly saccular. The rectum is short and nearly straight. The relations of the mesentery and the general arrangement of the veins are similar to those in the Pig. INTESTINAL TRACT OF MAMMALS. 471 In Babirussa babirussa (fig. 48, 1, p. 520) the structure still more closely resembles that of the Pig. The duodenum is a distinct simple loop. Meckel’s tract is enor- mously long and thrown into regular closely-packed folds; the proximal portion of the hind-gut and the rectum are practically identical with the corresponding regions in the Pig. To the peculiarities of the cecum I refer later (p. 519). In the Pigmy- Hog, Porcula salvania, the pattern is in the main similar to that of the Common Pig, the spiral coil being extremely well marked. Excluding the Hippopotamide, on which I have not sufficient information, it is clear that the gut-pattern of the non-ruminant Artiodactyla is definite, the charac- teristic features being the length and simplicity of Meckel’s tract, the short wide cecum, and the expansion of the anterior portion of the hind-gut into a very long and extremely capacious single loop, coiled into an elongated spiral and with the proximal portion of the loop much wider than the distal portion. Sub-Order TRAGULOIDEA. Family Tracutipm. Tragulus meminna (fig. 19, p. 472). In the Ruminant Artiodactyles the total length of the intestinal tract is relatively great. The increase compared with the groups which I have already considered is almost entirely in the hind-gut, the complications of which are characteristic. The increased length of this region and its mode of folding in the body-cavity has led to considerable alteration of the mesentery, with the result that when the tract is laid out in a fashion to display its pattern the mesentery has been cut or torn in a number of regions. The Chevrotain, of which I have examined several individuals all belonging to the common Indian species, displays the Ruminant Artiodactyle pattern in a simple form. There is no distinct duodenal loop. Meckel’s tract is much elongated, and is suspended in a number of simply-placed folds at the periphery of a semicircular expanse of mesentery; distally these folds shorten and the recurrent portion of the tract is nearly straight, passing nearly at right angles into the relatively short and wide cecum (fig. 19, C). In the actual figure the width of the cecum and of the portion of the hind-gut with which it is continuous is slightly underestimated. The hind-gut displays three well-marked regions. ‘The most proximal (S.P. in fig. 19) is a short narrow loop coiled into a loose spiral of about two turns. This is the simplest form in which the spiral watchspring-like coil of the hind-gut of Ruminants appears. It lies folded back over the central portion of the mesentery of Meckel’s tract; but in the Chevrotain the connection is very slight, and the loop may be folded back with almost no dissection. Next follows a longer loop (S.F. in fig. 19). In the living animal this is applied closely to Meckel’s tract lying just where the minor loops of that start from the mesentery. It corresponds with the similar fold which I have described in other Mammals, and it is in intimate connection with the mesentery and ” . 3R2 472 DR. P.. CHALMERS MITCHELL ON THE blood-vessels of Meckel’s tract, so that when it is dissected away, as in the figure, it appears to have neither mesentery nor blood-vessels. The rectum or most distal portion of the hind-gut is relatively short and very little convoluted, and is supported by the usual mesentery. The anterior mesenteric branch of the portal system follows the course of Meckel’s Intestinal Tract of Tragulus meminna. 8. Cut end of duodenum near stomach. R. Cut end of rectum near anus. C. Ceecum. 8.P. Spiral coil. S.F. Supra-Meckelian fold. X, X. Cut ends of posterior mesenteric vein. tract, receiving factors from the subsidiary loops with almost diagrammatic simplicity. Distally a separate and slightly enlarged factor drains the cecum. ‘The posterior mesenteric or rectal vein has the normal course, and receives the vein which drains the spiral fold. INTESTINAL TRACT OF MAMMALS. 473 Sub-Order TYLOPODA. Family Cametipm. Camelus dromedarius (fig. 20), Lama huanaco. A single glance at the figures will show the close similarity between the pattern in this sub-order and in the Traguloidea. There is no specialised duodenum in the Camel. Meckel’s tract is simply a repetition of that in Zragulus. ‘The cecum is longer and relatively not so wide. The spiral coil is similar in position, but is very much longer, Fig. 20. Intestinal Tract of Camelus dromedarius. Lettering as in fig. 19. coiled into several turns, and much flatter. In the undisturbed condition it lies folded over Meckel’s tract. ‘The supra-Meckelian fold has precisely the same relations, lying in the undisturbed condition very closely applied to the origin of the minor loops of Meckel’s tract. ‘The rectal portion has a normal course. The blood-vessels show only slight differences. ‘Lhe factor of the anterior mesenteric vein which drains the caecum AT4 DR. P. CHALMERS MITCHELL ON THE joins the main vein about the middle of the course, running across the semicircular expanse of mesentery. The rectal or posterior mesenteric vein, represented as cut across at X and X, receives the factors which drain the spiral coil and enters the anterior mesenteric vein close to the duodenum. In the Guanaco the pattern can hardly be distinguished from that of the Camel. The cecum is relatively much shorter, and the spiral coil is rather longer and less tightly coiled. Sub-Order PECORA. Family Cervip#. Cervus avis and many others. Family Bovins. Ovis tragelaphus (fig. 21), Gazella dorcas, and many others. Family Girarripa. Giraffa camelopardalis (fig. 22). Taking advantage of the material at the Gardens, I have examined a large number of Deer, Sheep, Cattle, Goats, Antelopes, and one specimen of Giraffe, and have found Intestinal Tract of Outs tragelaphus. Lettering as in fig. 19. a remarkable uniformity in the very elaborate patterns displayed by these specialised animals. The duodenum is a well-marked loop in the Giraffe (fig. 22) and in the Cervide ; it is less well separated in the Bovide (fig. 21). Meckel’s tract is in all INTESTINAL TRACT OF MAMMALS. 475 very long, forming, in most cases, a very regular series of closely-packed minor loops at the periphery of the usual semicircular expanse of mesentery. On the whole, these coils are more regular and more closely packed in the Cervidze and Giraffide than in the Bovide, but 1 doubt if there is enough distinction between the groups to lay any stress upon it. The cecum is long, but varies from genus to genus. The hind-gut always displays the three regions to which I have called attention in describing Tragulus and Camelus; but besides this there is sometimes an additional minor loop between the cecum and the spiral coil, which I term the postcecal loop Intestinal Tract of Giraffa camelopardalis. S. Cut end of duodenum near stomach. R. Cut end of rectum near anus. C. Caecum. P.C.L. Postescal loop of hind-gut. S.P. Spiral loop of hind-gut. S.F. Supra-Meckelian fold. X, X. Cut ends of posterior mesenteric vein. X2, Xz. Cut ends of vein from spiral loop. (fig. 22, P.C.L.) and which is well marked in the Giraffe and some Deer, but not so distinct in the Bovide. ‘The spiral loop is always long and flatly coiled in a number 476 DR. P. CHALMERS MITCHELL ON THE of turns. The supra-Meckelian loop is also very long, and is so closely applied to the course of Meckel’s tract that it is difficult to dissect it off. The rectal portion varies in length, being relatively longest, most capacious, and most convoluted in the Giraffe. The portal system presents a number of minor variations which seem not to be distributed in accordance with the systematic divisions of the animals in question. The anterior mesenteric vein always follows the course of Meckel’s tract, closely receiving more or less regularly arranged minor factors from the subsidiary loops. Where, as in the Giraffe, the course of the tract nearly completes a circle, a large vessel may run across the diameter of the mesentery. The cecum is drained by a vein which joins the anterior mesenteric usually some distance from the edge of the mesentery. The spiral loop is drained by a large vessel which sometimes joins the cecal vein, sometimes enters the anterior mesenteric vein proximad of the cecal vein. ‘The posterior mesenteric or rectal vein joins the anterior mesenteric vein very near the edge of the mesentery. The Ruminant Artiodactyles, then, display a pattern peculiar to the group, and characterised by the enormous length, special modification, and arrangement of the hind-gut. In all, the hind-gut displays three well-marked regions: a spiral loop simplest in Zragulus, in the others forming a closely-coiled, flat, watchspring-like arrangement, folded over on the mesentery that supports Meckel’s tract; a supra- Meckelian fold, which in the characteristic and most specialised cases is stretched round Meckel’s tract just at the line where the minor folds leave the mesentery, and which is drained by branches from the vessels of Meckel’s tract; and a rectal portion, the degree of convolution of which varies nearly directly with the size of the animal. The Non-Ruminant Artiodactyles display a pattern fundamentally similar to, but less complicated than, that of the Ruminant forms. Meckel’s tract is almost identical in its disposition. The spiral coil of the hind-gut is present and is very large, but its calibre is wider in proportion to its length and the coiling is not so flat. There is no more than a trace of the supra-Meckelian fold, so that the hind-gut, although long, is less differentiated. Order PERISSODACTYLA. Family Taprripa. Tapirus americanus (fig. 23). In the Tapir the duodenum is a long and well-differentiated narrow loop. Meckel’s tract is comparatively short, and consists of a small number of relatively large loops suspended at the periphery of a nearly circular expanse of mesentery. The cecum is long, very capacious, and irregularly sacculated. The hind-gut is divided into two well-marked regions. ‘The first region consists of an enormously long and capacious colic loop, the proximal and distal limbs of which le close together, and of a relatively long and nearly straight rectum. ‘The anterior mesenteric vein curves round the INTESTINAL TRACT OF MAMMALS. Fig, 23. Cie. Intestinal Tract of Tapirus americanus. S. Cut end of duodenum near stomach. R. Cut end of rectum near anus. C. Crecum. C.L. Colic loop. X, X. Cut ends of posterior mesenteric vein. X2, X2. Cut ends of colic vein. VOL, XVII.—PART y. No, 6.—--December, 1905. =I | 478 DR. P. CHALMERS MITCHELL ON THE mesentery of Meckel’s tract and receives a very large factor from the cecum. More proximally it is joined by a large vessel (represented in fig. 23 as cut across at X2) from the colic loop and by the posterior mesenteric vessel. Family Ruocerotip#. Rhinoceros unicornis (fig. 24). I had the opportunity of examining the intestinal tract of the large Indian Rhinoceros which died in 1904 after having lived many years in the Gardens. As in Fig. 24, Intestinal Tract of Rhinoceros wnicornis. Lettering as in fig, 25. fo) ‘s many large Mammals, the length of the gut is small in proportion to the size of the animal, ‘There is no distinct duodenal loop Meckel’s tract is relatively simple, consisting of a number of simple minor loops at the periphery of an elongated expanse of mesentery. ‘The cecum is short, but very wide and capacious. ‘The hind-gut repeats the pattern described in the Tapir, consisting of a very long and 479 INTESTINAL TRACT OF MAMMALS. Intestinal Tract of Hquus grant. a) Lettering as in fi ms | nm ovr) 480 DR. P. CHALMERS MITCHELL ON THE extremely capacious colic loop and a wide, sacculated, but nearly straight rectal portion. The arrangement of the veins is essentially identical with that found in the 'Tapir. Family Equipa. Equus granti (fig. 25, p. 479). I have examined the intestines of several Horses and Asses, but, except in minor details, the pattern displayed is identical. ‘The duodenum is not so well marked as in the Tapir, but is more specialised than in the Rhinoceros. Meckel’s tract is relatively longer than in the two other Perissodactyle families, and consists of a number of regularly arranged minor loops round the circumference of an oval expanse of mesentery, united with the general mesentery only by a narrow neck. ‘The cecum is very long and capacious. ‘The hind-gut displays an enormously capacious colic loop exactly similar in its relations to that of the Tapir and Rhinoceros, and a relatively long and slightly folded rectum. The arrangement of the mesenteric veins is similar to that in the other Perissodactyles, save that the anterior mesenteric vein is made up of a number of branching factors. The almost essential similarity of the gut-pattern displayed by the three Perissodactyle families, although the members of these differ much in structure and habits, is very striking. Except so far as both Artiodactyles and Perissodactyles have the whole gut relatively long and the hind-gut especially long, there is no special connection between the patterns displayed by Artiodactyles and Perissodactyles. In the former group the lengthening of the proximal portion of the hind-gut assumes the form of a narrow and very long loop with no trace of spiral coiling; in the latter the colic loop is always coiled and relatively narrower. Order RODENTTA. Sub-Order SIMPLICIDENTATA. Group HaApLopoNToIDEa. I have had no opportunity of examining members cf this group. Group Sciuroipua. Xerus capensis (fig. 26). ‘This animal shows the characteristic Rodent configuration of the alimentary canal in a relatively simple form. ‘There is a fairly distinct duodenal loop. Meckel’s tract consists of a number of simple minor loops at the periphery of the usual, nearly circular expanse of meseutery, and terminates in a relatively straight recurrent portion which enters the cecum at right angles. The cecum in Xevus is a very capacious sac, much expanded distad of the small intestine and tapering to a point anteriorly. ‘The hind-gut is divided into two regions. There is a much lengthened colic portion, variously disposed, but never showing any trace of the spiral coil which is so well marked in the Artiodactyles ; in Xerws this portion consists of two specialised loops . INTESTINAL TRACT OF MAMMALS 481 nearly equal in length, which are folded over on Meckel’s tract in the undisturbed The second portion of the hind-gut is a relatively long and nearly straight condition. rectum. The anterior mesenteric vein sweeps round the mesentery of Meckel’s tract, and is Intestinal Tract of Xvrus capensis. Cut end of duodenum near stomach, S. It. Cut end of rectum near anus. C. Czeecum. CL. | ; CL f Colic loops. X. Cut ends of posterior mesenteric vein. ; ’ ‘The posterior mesenteric vein, which drains the rectum joined by a large colic vein. joins the anterior mesenteric near the edge of the mesentery. 482 Dk. P. CHALMERS MITCHELL ON THE Group CastoroinEa. Castor fiber. In the Beaver, one example of which I have examined, the general pattern is similar to that of Xerus. ‘The duodenum consists of a number of loops not well separated from the anterior part of Meckel’s tract. The latter region is almost identical with that in Xerus. The cecum is very capacious. The portion proximad of the entrance of the small intestine is extremely long and tapers to a point. Posteriorly, it passes into a much-sacculated expanded portion, the line of which is continuous with a sacculated, much-dilated anterior portion of the hind-gut. The general configuration of this region strongly suggests that the existing czecum is one member of an original pair. The colic region of the hind-gut consists of two loops, the distad of which is very long and irregular, bearing several minor loops, the whole being folded over the Meckelian mesentery. ‘The rectal region is relatively short and nearly straight. The general arrangement of the veins is similar to that in Xerus. Groups GEOMYOIDEA, ANOMALUROIDEA, and MYoxoIDEA. I have not examined any of these. Flower (5) states that the caecum is absent in Muscardinus avellanarius, a condition recalling the Insectivora and obviously secondary in Rodents. Group Dipopoipna. Alactaga indica, Dipus egyptius. In the Jerboas, of which I have examined several specimens, the pattern is similar to that in other Rodents. ‘The duodenum consists of a number of short loops not well marked off from Meckel’s tract. The latter is suspended by an elongated oval expanse of mesentery and is relatively rather short and wide, being thrown into numerous very short, regularly disposed loops, and ending in a straight recurrent portion. The cecum is not so capacious as in the Squirrels and Beavers, but seems to vary considerably in different individuals. It is long, and gradually tapers to a point. The hind-gut consists of a colic region formed of three very definite loops, folded over the Meckelian region. ‘The rectum is relatively long and straight. ‘The anterior mesenteric vein has the normal course. It is joimed by a very large cecal vein, and by three veins from the three colic groups, and at the edge of the mesentery unites with the posterior mesenteric vein which drains the rectum. Group Myoiwra. Mus decumanus (fig. 27), Mus arboricola, Cricetus frumentarius. In the Rat the duodenum is a well-marked separate loop. Meckel’s tract is relatively long and is composed of a number of long, irregular, minor loops forming a contorted mass at the periphery of a semicircular expanse of mesentery, and terminating in a straight, recurrent portion. ‘The cecum is capacious, tapers to a blunt point, and has a well-marked spiral curvature. The hind-gut is relatively short and is divided into a colic loop which lies loosely folded over the Meckelian mesentery and is generally INTESTINAL TRACT OF MAMMALS. 483 marked by ovoid masses of feeces, and a short straight rectal portion. The anterior mesenteric vein sweeps round the Meckelian mesentery, receiving factors from the duodenum and from the minor coils of Meckel’s tract. Near the edge of the mesentery it is joined by a large cecal vein, a vein from the colic loop, and the posterior mesenteric vein. In the Australian Bush-Rat (J/ws arboricola) the arrangement is much the same. Intestinal Tract of Mus decumanus. Xe. Cut ends of colic vein, Other letters as in fig. 26. The spiral curvature of the capacious cecum is less marked and the colic region tends to display two loops. In the Hamster (Cricetus frwmentarius) the duodenum is not so distinet as a separate loop. The cecum is relatively very much longer, is irregularly convoluted, 484 DR. P. CHALMERS MITCHELL ON THE and shows signs of spiral twisting. ‘The hind-gut is composed of two rather irregular, narrow colic loops, and the rectum is relatively long. Group BaTHYERGOIDEA. I have not been able to examine any example of this group. Group HystRicorpEa. Family Hysrercipm. LHystrix cristata, Sphingurus prehensilis, Atherura africana. In the Porcupine there is a distinct duodenal loop. Meckel’s tract is long and composed of numerous, rather short, irregular loops, disposed in the usual fashion round the edge of an elongated, hemispherical expanse of mesentery with a very short, straight, terminal portion. ‘The caecum is very long and rather narrow. ‘The proximal portion of the hind-gut is expanded and forms a kind of accessory cecum. The colic portion of the hind-gut is composed of two long, distinct, but rather irregular groups, and the rectum is relatively short and straight. The anterior mesenteric vein is short and very wide, and receives a number of radiating veins from the loops of Meckel’s tract and a large czecal vein, as well as two veins from the colic loops, whilst in the usual fashion it is joimed by the posterior mesenteric vein from the rectum. In the Tree-Porcupine (Sphingurus prehensilis) the pattern is closely similar. The minor loops of Meckel’s tract are rather less numerous; the cecum is very long and capacious, but terminates in a long, slender, finger-like process, which may be described almost as a vermiform appendix. The hind-gut has precisely the disposition of that region in the Common Porcupine, and the arrangement of the mesenteric veins is similar. In the Brush-tailed Porcupine (Atherwra africana) the duodenum is less distinct ; Meckel’s tract is relatively considerably longer, with more numerous minor loops. The ceecum is large and tapers more gradually to a point. The hind-gut has the same pair of colic loops and a straight rectum. The mesenteric veins are practically identical in their disposition. Family ERETHIZONTIDA. I have not examined any members of this family. Family Caviip®. Cavia porcellus (fig. 28), Hydrocherus capybara (fig. 29, p. 486), Dolichotis patachonica. In the Guinea-pig (Cavia porcellus) there is a distinct duodenal loop. Meckel’s tract is relatively long and is composed of a number of long, irregularly arranged, and contorted minor loops at the periphery of an oval expanse of mesentery, The terminal INTESTINAL TRACT OF MAMMALS. 485 portion is straight except that it is involved in the spiral twisting of the caecum. ‘The cecum is very large and sacculated ; it is folded over the Meckelian mesentery and is spirally twisted, the twist involving not only the terminal portion of Meckel’s tract, but the proximal portion of the hind-gut, including practically the whole of one of the two Intestinal Tract of Cavia porcellus. X2, X2. Cut ends of colic vein. Xs, Xs. Cut ends of cecal vein. Other letters as in fig. 26. cecal loops that are found in most other Rodents. There is an expanded portion of . . i ? -) rn . > £ > } the hind-gut distad of the entrance of Meckel’s tract into the caecum, and sacculated like the caecum. ‘Lhe hind-gut shows two definite regions. ‘I'here are two colic loops, VOL. XVi1.—PART v. No. 7.—December, 1905. 3 7 486 DR. P. CHALMERS MITCHELL ON THE one of which, as already explained, being coiled up with the cecum, and the second being free. The rectal portion is very long, its anterior region being expanded to form a series of minor loops. The anterior mesenteric vein curves round Meckel’s tract and receives a large, branching colic vein represented in fig. 28 as divided at Xs, and a vein from the colic loops (shown divided at X2 in the figures). It is jomed by a normal posterior mesenteric vein. In the Capybara (fig. 29) there is no well-separated duodenum. Meckel’s tract has Fig. 29. > Intestinal Tract of Hydrocherus capybara. Lettering as in figs. 26 and 28. the usual disposition, being composed of a large number of closely-set minor loops. The cecum is extremely capacious, but is not spirally twisted to any appreciable extent, but is sacculated, and has an interesting peculiarity described by Garrod (7). The normal czecum is, in reality, not a direct continuation of the hind-gut, but a lateral outgrowth from such a continuation. The peculiarity was not so marked externally in INTESTINAL TRACT OF MAMMALS. 487 the example I examined as in that figured by Garrod, but is well seen on opening the cavity. It is obviously the same kind of condition as occurs in the Hare, and is to be explained by the supposition that originally the cca were paired. The hind-gut is divided into a colic region consisting of two loops closely twisted together, and showing a spiral folding, most unusual in this region in the Rodents, and recalling superficially the spiral coil of Artiodactyles. It is essentially dissimilar, however, as it is formed of two loops folded together. The rectum is very long, its proximal portion being considerably expanded. The arrangement of the veins is normal. In the Patagonian Cavy (Dolichotis patachonica) there is no distinct duodenum. Meckel’s tract is very long, and consists of much contcrted minor loops at the edge of an elongated oval mesentery. The czecum is extremely long and capacious and is spirally twisted, the twisting involving the proximal portion of the hind-gut. The latter consists of a colic region, the first loop of which is involved in the coiling of the czecum, the second being large and extremely irregular. The rectum is very long and displays a large series of minor loops. The arrangement of the mesenteric veins is essentially similar to that in the Capybara. Sub-Family Cuincnintp&. Lagostomus trichodactylus. In the Viscacha there is no distinct duodenal loop. Meckel’s tract is long, and consists of a series of minor loops and a long straight recurrent portion suspended at the periphery of an expanse of mesentery nearly circular. The cecum is extremely capacious and sacculated. ‘The hind-gut consists of two very long, rather simple loops. The rectum is greatly expanded, consisting of a series of proximal loops and a straight distal portion. The anterior mesenteric vein curves round the Meckelian mesentery and receives a large vein which drains the cecum and the first loop of the hind-gut and, separately, a vein from the second colic loop, whilst it is joined near the edge of the mesentery by a normal posterior mesenteric vein. Sub-Family Capromyipz. Capromys pilorides. The pattern of the intestinal tract of this animal is extremely like that of the Guinea-pig. ‘The cecum is rather less sacculated and is relatively longer, the spiral twist being rather greater. Otherwise the diagram given for Cavia (fig. 28, p. 485) might equally well represent Capromys. Sub-Families OcropontimDa and CHENODACTYLID&. I have not had the opportunity of examining members of these sub-families. 488 DR. P. CHALMERS MITCHELL ON THE Sub-Order DUPLICIDENTATA. Family Ocnoronipa. Ochotona rufescens. In the Mouse-Hares the general pattern of the alimentary canal is similar to that about to be described in the Rabbit. There is a well-marked duodenum. Meckel’s tract is composed of a large series of contorted minor loops at the circumference of a nearly circular expanse of mesentery. ‘The cecum is enormously long and capacious. It is coiled tightly and its extremity forms a long vermiform appendage. Just at the point where the small intestine enters the cecum (fig. 46, II, p. 517) there is a second cecum, in the form of an elongated, slightly curved, hollow papilla, which appears to be, with- out doubt, the vestigial representative of the second of an original pair of czeca. The coiling of the czecum has involved the greater part of the first colic loop. The second colic loop is long and distinct, and the rectum is relatively long. I have seen only preserved viscera of this genus, and I was unable to follow out the veins in detail : they appeared to be similar in arrangement to those of the Rabbit. Family Leporip®. Lepus cuniculus (fig. 30). In the Rabbit the duodenum is a well-marked simple loop. Meckel’s tract is very long and is composed of a series of minor loops nearly equal in size, and a terminal recurrent portion suspended at the margin of a long, oval, narrow-necked expanse of mesentery. ‘The czecum is very long, capacious, and ends, as is well known, in a finger- shaped, narrow, thick-walled vermiform appendix. It is spirally twisted and the twisting has involved a large portion of the end of Meckel’s tract as well as the first portion of the hind-gut. In the figure the latter is represented after having been dissected off the spirally-coiled caecum. The second loop of the hind-gut les folded over the Meckelian mesentery and is also represented in the figure as dissected away. The rectum is relatively short and straight. The anterior mesenteric vein curves round the Meckelian mesentery, receiving tributaries from the duodenum and from the minor loops of the tract. It is joined by a very large vein which branches over the coiled czecum and the first loop and part of the second loop of the hind-gut. There isa short, straight posterior mesenteric vein draining the rectum. In the embryo Rabbit the duodenal loop is relatively much longer, forming a narrow simple loop. ‘The minor loops of Meckel’s tract are much shorter and less numerous. The coiling of the caecum is much slighter, so that that organ appears as a long, com- paratively simple diverticulum, which at its proximal end has just begun to be spirally twisted. The first colic loop is nearly free from the cecum, and the second loop is relatively longer. ‘The appearance of the region in which the sacculus rotundus occurs suggests that that organ is the remnant of a second caecum, the other member of an original pair of ceeca. In the Hare the second caecum appears in a much more definite form (fig. 46, I, p. 517). INTESTINAL TRACT OF MAMMALS. 489 The character of the pattern displayed by the Rodents as a whole is remarkably homogeneous, and lends little support to the subdivisions of the group, which rest on the characters of other organs. The gut is elongated, no doubt in relation to the vegetarian habit of the members of the group. ‘The duodenal loop is generally well separated. Meckel’s tract, as in Mammals generally, varies only in the degree to Fig. 30. Intestinal Tract of Lepus cwniculus. S. Cut end of duodenum near stomach. R. Cut end of rectum near anus. C. Czeum. CLi. Colic loop. CL2 [S.F.}. Second colic loop acting as supra-Meckelian fold. X, X. Cut ends of posterior mesenteric vein. §.R. Sacculus rotundus. which its elongation has led to the elaboration of its minor loops. The special characters are to be found in the hind-gut. ‘The cecum is capacious, and shows a marked tendency to being spirally coiled, the coiling involving part of Meckel’s tract 490 DR. P. CHALMERS MITCHELL ON THE and more or less of the colic loops of the hind-gut, according to the extent to which it has taken place. ‘The elongation of the hind-gut is in the form of two loops, which, in most cases, although folded over the Meckelian mesentery have remained fairly free from it. The rectum shows a tendency to be elongated in its proximal region. It is interesting to notice that vestiges of an originally paired condition of the ceca appear to be specially characteristic of the whole group, but particularly of the Duplicidentata : in Ochotona two ceeca actually occur ; in the Hare they are present, the second caecum being less marked, whilst in the Rabbit the second ceecum is reduced to a sacculus rotundus. Order INSECTIVORA. Division MENOTYPHLA. Family Tupagipa. I have not been able to examine any example of this family myself. According to Flower (5) in Tupaia tana the intestinal tract is short and simple and there is no trace of a cecum. Flower, however, quotes Hunter’s authority for the existence of a cecum in one species of Tupaia. Family MAcROScELIDIDé. T have not examined any of these Shrews myself. Flower (5) states that the intestinal tract is much more like that of the Lemurs than that of other Insectivora. It is relatively long, and a capacious cecum marks the position where the anterior part of the tract passes into a capacious hind-gut which is differentiated into a looped colon and a rectum. Division LYPOTYPHLA. Family Tarpipa. Talpa europea. In the Mole the duodenum is not well separated from the proximal portion of Meckel’s. tract. The latter is relatively very long, and consists of a number of regularly arranged, closely packed minor loops, at the periphery of a long oval expanse} of mesentery, the recurrent portion of the tract being nearly straight. The hind-gut consists of a narrow, slightly twisted colic loop and a long straight rectum. There is no trace of a cecum. ‘The anterior mesenteric vein curves round the mesentery of Meckel’s tract and is joined by a straight posterior mesenteric vein. Family Soricip&. I have examined no example of this family. Hunter (10) and Flower (5) have described the shortness and simplicity of the intestinal tract of these animals, calling attention to the absence of the caecum, and in particular to the Reptilian fashion in INTESTINAL TRACT OF MAMMALS. AQ] which the simple gut is suspended along its entire course from stomach to anus by a simple fold of mesentery. Family Ertnaceia. Erinaceus algirus. The pattern of the intestinal tract is practically identical with that found in Centetes, which I describe below. Family PoraMoGaLID&. I have examined no specimen of Potamoga/e, and can find no recorded anatomy. Intestinal Tract of Centetes ecaudatus. §. Cut end of duodenum near stomach. R. Cut end of rectum near anus. 4992 DR. P. CHALMERS MITCHELL ON THE Family Centeripe. Centetes ecaudatus (fig. 31, p. 491). The duodenum cannot be distinguished from Meckel’s tract as a separate loop, The latter is relatively long, and is suspended in very regular minor loops at the periphery of a nearly circular expanse of mesentery. ‘There is noceecum. ‘The hind-gut consists of a short, straight, and rather wide rectal portion. ‘The anterior mesenteric vein curves round the Meckelian mesentery and is joined by a posterior mesenteric vein, Family CHRYSOCHLORIDS. I have not examined any members of this family. Flower (5) states that the caecum is absent. As the cecum is absent in at least Tupaia tana, it is plain that the names Menotyphla and Lypotyphla are not well chosen for the two divisions into which the Insectivora are divided. I have already called attention to the extreme, almost Reptilian, simplicity which characterises the gut-pattern of most of the Insectivora. It cannot be assumed, however, that this simplicity is primitive. The absence of the ceecum is probably secondary, and the differentiation of the hind-gut in the Macro- scelididee and in the Mole affords evidence of a formerly more complex condition in the other forms. Order CHIROPTERA. Division MEGACHIROPTERA. Family Preropopis. Pteropus gouldi, P. medius, Cynonycteris collaris. In the Fruit-Bats the intestinal tract is relatively much longer than in the Micro- chiroptera, a condition which is rather in contrast to that found in Birds; fruit-eating birds practically invariably have the intestinal tract much shorter and of relatively wider calibre than occurs in their nearest insectivorous or seed- and grain-eating relatives. In the Pteropodide the duodenum is not a distinctloop. Meckel's tract has a disposition almost exactly similar to that shown in fig. 31. The hind-gut consists merely of a short, rather wide rectum, and there is no cecum. Division MICROCHIROPTERA, I have examined a number of Bats, chiefly Vespertilionide. In all these the intestinal tract is extremely short, consisting merely of a very short and slightly folded loop, representing the duodenum and Meckel’s tract, and suspended by a slightly expanded mesentery, and a hind-gut consisting of a very short straight rectum. I myself have seen no cecum, but Owen states that a short cecum is present in Megaderma, one of the Rhinolophide, and in Rhinopoma, one of the Emballonuride. It is plain that the very simple gut of the Chiroptera cannot safely be asserted to be primitive. INTESTINAL TRACT OF MAMMALS. 493 Order GALEOPITHECID. I much regret that I have been unable to examine a specimen of this order. Flower (5) states that the cecum is very long, and that the hind-gut exceeds in length the fore-gut. Leche (12), in his monograph on the genus, does not figure the whole intestinal tract, but in his short description gives sufficient indications that the gut is very unlike that of the Insectivora or Chiroptera, and much more like that of the Lemurs. It appears that the duodenum is more or less distinct and that Meckel’s tract consists of a number of loops of narrow calibre. The cecum is long and wide. The hind-gut is very long, longer than the fore-gut, and is divided into three regions. Proximally there is a very wide region, continuous with the caecum and similar to it in structure. Next follows a long colic portion, the arrangement of which I cannot follow from the description, and finally a rectal portion. Order CARNIVORA. Sub-Order PINNIPEDIA. Family Orartp®. Otaria californiana (fig. 52, p. 494). The duodenum is not well separated from Meckel’s tract. The latter is enormously long, rather thick-walled, and of even and small calibre, recalling the character of the intestinal tract in fish-eating birds; it is arranged in a very large number of irregular loops which form a contorted mass at the circumference of an elongated, oval expanse of mesentery. The distal, or recurrent, portion is nearly straight. There is a short cecum. The hind-gut can hardly be separated into colic and rectal portions, but the proximal part forms a somewhat enlarged loop. ‘The anterior mesenteric vein curves round the Meckelian mesentery, receiving numerous, regularly disposed tributaries from the minor loops. It is joined near the edge of the mesentery by a posterior mesenteric vein. Family TRIcHECHID&. I have had no opportunity of examining any example of this family; but Flower (5) states that the intestinal tract is similar to that of the Phocide, Family Poocip®. Phoca vitulina. There is no difference of importance between the patterns of the gut displayed in the Sea-Lions and in the Seals. In the latter, the duodenum is rather more distinct ; Meckel’s tract is perhaps rather longer. ‘The caecum is longer and narrower, ‘The hind-gut and the arrangement of the blood-vessels are similar in the two families. The character of the gut in the Pinnipedes shows not the smallest resemblance to that of the Sirenia, It has a superficial similarity to the gut of the Cetacea, but this VoL. Xvil.—Part Vv. No, 8.—December, 1905, 30 494 DR. P. CHALMERS MITCHELL ON THE is confined to the enormously elongated, thick-walled Meckelian portion; the hind-gut and the cecum show a close resemblance with what is found amongst Carnivora generally. The resemblance between the gut of the Cetacea and of the Seals and Sea-Lions is therefore purely adaptive, and is in relation to the fish-eating habit. Intestinal Tract of Otaria californiana. =) ’ S. Cut end of duodenum near stomach. Rk. Cut end of rectum near anus. C. Ceeeum. X. Divided ends of posterior mesenteric vein. INTESTINAL TRACT OF MAMMALS. 495 Sub-Order FISSTPEDIA. Division ARCTOIDEA. Family Mustanipa. Mustela subpalmata, Galictis barbara, Ictonya zorilla, Lutra vulgaris. In these genera the pattern of the intestinal tract is much alike. The duodenum is not well separated from Meckel’s tract. ‘The latter region is of moderate length and Fig. 33. So BEA Seki hok \) BY, SESGTLORD DAY an ') Intestinal Tract of C reoleptes caudivoluulus. Letters as in fig. 32. ° is suspended ina series of simple loops with a terminal, nearly straight, recurrent portion, at the circumference of a mesenteric outgrowth which is usually nearly circular. There is no trace of a cecum. ‘The hind-gut is not divided into colic region and rectum, but there is usually a slight expansion of this region just distad of Mecicel’s tract, and 9 ¢ aU 2 496 ° DR. P. CHALMERS MITCHELL ON THE suggesting the possible former existence of a colic loop. The anterior mesenteric vein runs nearly diametrically across the Meckelian mesentery, receiving the usual tributaries from the loops of Meckel’s tract, and, near the dorsal edge of the mesentery, is joined by a posterior mesenteric vessel. Fig. 34. NEED. [S ogy a NE Pera} 2 ks aS Co 5 AV pee ere 6 FEW TTTLT « Intestinal Tract of Ursus malayanus. Lettering as in fig. 32. C.L. Colic loop. Family Procyonipm. Bassaris astuta, Nasua vittata, Procyon lotor, Cercoleptes caudivolvulus (fig. 33, p. 495). In this group there is not much difference from the characters shown by the intestinal tract of the Mustelidea. The duodenal loop is generally rather better separated. Meckel’s tract is almost similar in character. There is no trace of a ceecum, and, as INTESTINAL TRACT OF MAMMALS. @ Se EP : a os j ) YS La] : v5 Epes md RCE, Sweewenee ATL Se SN oS OT ed Gt 5 eS dD Toe Po Aimee Intestinal Tract of Canis vulpes. S. Cut end of duodenum near stomach. lt. Cut end of rectum near anus C, Caecum, 498 DR. P, CHALMERS MITCHELL ON THE shown in fig. 32, p. 494, the hind-gut consists of a straight rectal portion with occa- sionally a slight proximal looping. There is a peculiarity, however, which is not well marked in the figure of Cercoleptes, but which is conspicuous in Bassaris, Naswa, and Procyon. Immediately distad of the terminal or recurrent straight portion of Meckel’s tract the calibre of the intestinal tract suddenly increases. ‘The anterior mesenteric vein is bent rather sharply on itself and receives very numerous, long, radiating tributaries from the minor loops of Meckel’s tract. It is joined by the usual posterior mesenteric vein. Family Ursinz. Ursus malayanus (fig. 34, p. 496). The duodenum is represented by a fairly distinct loop. Meckel’s tract consists of a series of rather regular open loops arranged at the periphery of a nearly circular expanse of mesentery, the whole tract being relatively rather long. There is no cecum. ‘The hind-gut consists of two portions, a slightly looped colic region (C.L. in fig. 54) and a long, wide, and nearly straight rectal portion. The anterior mesenteric vein curves round the Meckelian mesentery and receives a large colic branch, whilst it is joined by the posterior mesenteric vein from the rectum. Family Canipa. Canis familiaris, Canis vulpes (fig. 39, p. 497). I follow Max Weber (27) in placing the Canide in the Arctoid division of the Carnivora. In this family the duodenal loop is well distinct from the next portion of the gut, this distinctness being better marked in puppies than in adults. Meckel’s tract is relatively shorter than in the other Arctoids, and is composed of rather short, irregular, minor loops and a terminal, recurrent, straight portion at the circumference of an elongated oval expanse of mesentery. ‘The cecum is present and of moderate size; it is twisted in a cork-screw fashion, the twisting being less marked in young puppies than in adults. The hind-gut is divided into a colic portion, usually better marked than in fig. 835, which was drawn from a Fox, and a relatively long, nearly straight, rectal portion. ‘The anterior mesenteric vein is sharply bent on itself as in the Procyonidz, and is joined by a posterior mesenteric vein which drains both the colic loop and the rectum. Division HERPESTOIDEA. Family Hyanipz. Hyena crocuta (fig. 36). In the Hyena there is a distinct duodenal portion. Meckel’s tract is relatively very long, and consists of a closely-set mass of contorted, irregular, minor loops and a long, straight, recurrent portion, suspended at the circumference of an elongated oval expanse of mesentery. There is a comparatively long, straight cecum. . The hind-gut is much wider in calibre and consists of a distinct colic loop and a long wide rectum. ‘The anterior mesenteric vein is very large and is bent on itself and receives numerous regularly placed affluents from the course of Meckel’s tract. It is joined by a short colic branch and a long posterior mesenteric vein from the rectum. INTESTINAL TRACT OF MAMMALS. 499 Intestinal Tract of Hycena crocuta. Lettering as in fig. 35. C.L. Colic loop. 500 DR. P, CHALMERS MITCHELL ON THE Family Viverripa2. Herpestes griseus (fig. 37), H. albicauda, H. ichneumon, Suricata tetradactyla, Cynictis penicillata, Nandinia binotata, Arctitis bin- turong, Cryptoprocta feron. The general arrangement of the intestinal tract is similar throughout this family. Fig. 37. Intestinal Tract of /erpestes griseus. Lettering as in fig. 35. The duodenum is a distinct loop; in fig. 87 it had been straightened ont before the drawing was made. Meckel’s tract is relatively short and consists of a small number INTESTINAL TRACT OF MAMMATS. 501 of irregular scattered loops, by no means closely set together, and a straighter recurrent portion suspended on an irregularly circular expanse of mesentery. ‘The caecum varies considerably, and shows evidence of being degenerate. In Herpestes griseus (fig. 37, C), for instance, it tapers to a point and is slightly twisted. In H. albicauda it is short and rather wide; in H. ichnewmon relatively longer and very slender. In Suricata it is of moderate length and of even calibre. In Cynictis it is much as in Suricata. In Nandinia it is absent. In Arctitis I found it present, but in a vestigial form, consisting of a minute globular sac, joined to the gut by a hollow narrow stem, whilst, according to Weber and Beddard, it is sometimes absent. In Cryptoprocta it is small and tapers towards the extremity. The hind-gut is usually rather short and shows only traces, which vary in the different genera, of differentiation into colic and rectal portions. ‘The anterior mesenteric vein curves round the Meckelian mesentery, receiving irregularly placed tributaries from the wall of the gut; it is joined by a posterior mesenteric vein from the whole of the hind-gut. Family Fevipa. Melis caracal (fig. 38, p. 502), Felis serval. In the Felidee the duodenal loop (the figure of the Caracal was drawn after the duodenum had been partly dissected out) is distinct. Meckel’s tract is relatively short and is composed of a few very irregular well-separated minor loops, and a straight recurrent portion suspended at the periphery of a nearly circular expanse of mesentery. ‘The cecum is present, but is obviously degenerate. It is always short, and in the Caracal is slightly curved; in the Serval it is still shorter and is bluntly rounded. ‘The hind-gut is relatively very short indeed, and shows practically no trace of differentiation into colic and rectal regions. ‘The anterior mesenteric vein curves round the Meckelian mesentery, receiving irregularly placed tributaries, and is joined by a posterior mesenteric vein from the hind-gut. The general features of the Carnivora, as a whole, are the great preponderance of the fore-gut over the hind-gut, and the reduction of the latter region and of the cecum. The Pinnipedia are clearly marked off by changes, probably adaptive, in the Meckelian tract. Between the Arctoid and Herpestoid divisions it is difficult to discriminate so far as the characters of the intestines are concerned. ‘The cecum, which is obviously degenerating in the whole group, cannot be taken as a useful character from this point of view. It is absent, it is true, in all the Arctoids except the Canide, but in the latter it is larger than in any other Carnivora, whilst in some of the Herpestoids it is always absent (Nandinia) or occasionally absent and always small in Arctitis. In the Arctoids, as a whole, it may be said that Meckel’s tract is relatively longer and more complicated, but, on the other hand, in //y@na it is conspicuously long. ‘The hind-gut, which appears to be in process of reduction in the group, is without doubt shorter and less differentiated in the Herpestoids than in the Arctoids, VOL. XViUl.—PaAr? v. No. 9.—December, 1905. hx 502 DR. P. CHALMERS MITCHELL ON THE Intestinal Tract of Felis caracal. Lettering as in fig. 35. INTESTINAL TRACT OF MAMMALS. 305 Order PROSIMLA. Sub-Order TARSIID A. Family Tarsiina. I have had no opportunity of examining Jarsivs. Burmeister (3), however, in his well-known monograph, has described and figured the intestinal tract. There appears to be no distinct duodenal tract. Meckel’s tract is relatively long, and is disposed in irregular loops at the circumference of an oval expanse of mesentery. ‘There is a capacious, spirally twisted cecum. The hind-gut is extremely small, but much larger in calibre than the fore-gut; there is a slight trace of a colic loop and a very short, straight rectum. Sub-Order LEMURID/A. Family Lemurina. Lemur mongoz var. nigrifrons (fig. 39, p. 504), L. rufifrons, Chirogaleus cocquereli (fig. 40, p. 505). In these Lemurs there is no very distinct duodenal loop. Meckel’s tract varies in length and in the complexity and number of its minor loops. In Lemur mongoz var. nigrifrons (fig. 59) these are present on the proximal half of the nearly circular expanse of mesentery, absent on the distal half; in L. rwfifrons, there are rather more minor loops, and in Chirogaleus (fig. 40) the minor loops occur all round the expanse of mesentery. The cecum is always present and is capacious ; in the species of Lemur that I have examined the cecum is extremely long; in Chirogaleus it is relatively much shorter. ‘The hind-gut shows two distinct regions. There is a colic loop, which is long and narrow in Lemur, but relatively much shorter in Chirogaleus, and a rectal portion nearly straight, which varies in length. ‘The anterior mesenteric vein receives a number of rather irregularly placed tributaries from Meckel’s tract, and a large cecal vein, and is joined by a posterior mesenteric vein that drains both the colic loop and the rectum. Family InpRIsiIna&. I have had no opportunity of examining any members of this family. Flower (5), however, has published an account of the intestinal tract of Indris brevicaudatus. He describes the whole intestinal tract as being of extraordinary length, comparable in relation to the size of the animal with that of grass-eating Rodents. No special description of the duodenum and Meckel’s tract is given, but these would appear to differ little except in length from the corresponding structures in true Lemurs. The cecum is enormously long and of great capacity. ‘The hind-gut is differentiated into a very long irregularly looped colic region and a rectal region. No account is given of the mesenteric veins. 504 DR. P. CHALMERS MITCHELL ON THE Intestinal Tract of Lemur mongoz var. nigrifrons. S. Cut end of duodenum near stomach. R. Cut end of rectum near anus. C. Caecum. C.L, Colic loop, INTESTINAL TRACT OF MAMMALS, 505 Family CHIROMYIN. IT have had no opportunity of examining the Aye-Aye, but the monographs of Owen (18) and of Peters (21) give a considerable amount of information as to the structure of the intestinal tract. ‘There is a distinct duodenum; Meckel’s tract is long Intestinal Tract of Chirogaleus cocquereli. Lettering as in fig. 39. and convoluted ; the cecum is of fair size, but is in course of degeneration, and in its terminal three-quarters is much reduced in calibre. The hind-gut is very long and is differentiated into a long, narrow, colic loop anda rectal portion. No clear information is given as to the mesenteric veins. 506 DR. P. CHALMERS MITCHELL ON THE Fig, 41. Intestinal Tract of Galago garnetti. Lettering as in fig. 39. INTESTINAL TRACT OF MAMMALS. 507 Family Gatacine. Galago garnetti (tig. 41). In Galago there is no distinct duodenal loop. Meckel’s tract is relatively short and is suspended in a series of minor loops at the circumference of the usual elongated expanse of mesentery. The cecum is long and capacious. The hind-gut is extremely long and capacious, and is divided into a well-marked compound colic loop, which, in Intestinal Tract of Perodicticus potto. S. Cut end of duodenum near stomach. R. Cut end of rectum near anus. C. Cxecum. C.L,1 & 2. Colic loops. 508 DR. P. CHALMERS MITCHELL ON THE the normal position, lies folded against the Meckelian mesentery, and a long and wide rectal portion. The anterior mesenteric vein is nearly straight and is joined by a large colic vein and a separate tributary from the colic loop. The large posterior mesenteric vein, which drains the rectum, unites with the anterior mesenteric near the dorsal edge of the mesentery. Family Lorisinx. Perodicticus votto (fig. 42, p. 507). In the Potto there is no distinct duodenal loop. Meckel’s tract is relatively short and consists of a series of irregular loops occupying only about the anterior two-thirds of the circumference of an oval expanse of mesentery. The cecum is long and capacious. he hind-gut is extremely long, and consists of a colic region composed of one or two short kinks and two very large loops, which, in the undisturbed condition, are folded together and applied to the Meckelian mesentery, and a straight rectal portion. ‘The anterior mesenteric vein is nearly straight and receives a cecal factor and two large factors from the colic loops. It is joined by the posterior mesenteric vein from the rectum. The gut-pattern of the Prosimie is fairly definite in character. The duodenum is seldom well distinct from Meckel’s tract; the latter varies in length, probably in relation to the diet. ‘The caecum is always present and usually very capacious. ‘The hind-eut (except in Zarsius, where it is obviously reduced and still shows signs of former differentiation) is relatively extremely long, sometimes being as long as, or longer than, the fore-gut. It is, moreover, of greater calibre. It shows a well-marked colic region, which may be a long narrow loop or a complex set of folds, and a distinct rectum. Order SIMI. Sub-Order PLATYRRHINA. Family Haranipa. LHapale penicillata, Midas rufimanus. In the Hapalide the whole intestinal tract is relatively short, although Weber's (27) statement that it is only a quarter the length of the body is certainly an under- estimate. ‘The duodenal loop is not well separated. Meckel’s tract consists of a series of regular loops and a terminal straight portion arranged round the periphery of a nearly circular expanse of mesentery. ‘The cecum is long and very capacious. The hind-gut is relatively long and capacious ; it consists of a wide colic loop which passes insensibly into a straight rectal portion. ‘The anterior mesenteric vein stretches diametrically across the Meckelian mesentery, receives a large colic vein, and is joined by a posterior mesenteric vein which drains both the rectum and colon. INTESTINAL TRACT OF MAMMALS. 509 Family Cesipa. Sub-Family Nycripiruecina. Callithrix cuprea (fig. 49, I, p. 521). There is no distinct duodenal loop. Meckel’s tract is relatively not very long and is composed of a simple series of loops at the periphery of an elongated expanse of mesentery, the distal portion of which also acts as a mesocolon. The cecum is large and capacious and of even calibre throughout. The hind-gut is very long and is differentiated into a colic loop composed of a number of minor loops, a sigmoid flexure, and a short straight rectum. The anterior mesenteric vein runs diametrically across the Meckelian mesentery, receives a large colic vein, and is joined by the posterior mesenteric vein which drains both rectum and colon. Sub-Family PiITHEcinaé. I have examined no examples of this sub-family. ‘The description given by Owen (19) shows that the pattern of the gut is similar to that found in the Nyctipithecine, with the difference that Meckel’s tract is relatively longer and the hind-gut relatively shorter. The cecum is long and capacious. Sub-Family Mycetina. Mycetes beelzebul (fig. 48, II, p. 520), M. seniculus. There is no distinct duodenal loop. Meckel’s tract is relatively short and composed of a small number of simple minor loops. It is suspended along two-thirds of the circumference of an oval expanse of mesentery, the distal one-third acting as a mesocolon. The cecum is very short but extremely capacious, forming a bluntly rounded cone, the base of which is nearly as wide as the height. The hind-gut consists of a much distended proximal portion continuous with the cecum, a short colic loop, and a relatively long rectum. ‘There is no distinct sigmoid flexure and the hind-gut as a whole is relatively shorter than in the Nyctipithecine. ‘The arrangement of the mesenteric veins is much as in the last-named sub-family. Sub-Family Cxpina. Ateles variegatus (fig. 43, p. 510), Cebus fatuellus, Ouacaria rubicunda. In Ateles the duodenal loop (fig. 43) is rather wider in calibre than the succeeding intestinal region, and the duodenum is more distinct than in most of the Platyrrhine Monkeys. Meckel’s tract is relatively rather long and is arranged in a series of closely- set irregular loops at the periphery of an oval outgrowth of mesentery, the distal part of which, as in many other Platyrrhines, acts as a mesocolon. The distal extremity of Meckel’s tract passes into the junction of the cecum and colon at an acute angle, and in such a fashion as to suggest the vestigial presence of a second cecum. ‘The cecum is relatively short and forms a pointed elongated cone, which, in the fresh condition, is slightly coiled. _ The hind-gut displays first an expanded region continuous with the yOL. XVII.—PAkT v. No. 10.—December, 1905. aY 510 DR. P. CHALMERS MITCHELL ON THE cecum; next, a rather short colic loop which passes insensibly into a long rectum. The anterior mesenteric vein curves round the contour of Meckel’s tract, receiving Intestinal Tract of Ateles varieqatus. S. Cut end of duodenum near stomach. KR. Cut end of rectum near anus. C. Ceecum. C2, Vestigial cecum. C.L. Colic loop. irregular tributaries as well as a large cecal and a branched colic vein, ‘I'he posterior mesenteric vein drains both the colic loop and the rectum. In Cebus fatuellus (fig. 49, I, p. 621) the general pattern of the intestinal tract is INTESTINAL TRACT OF MAMMALS. 511 similar to that in Afeles ; but the duodenal loop is less distinct, the caecum is relatively a little larger and does not taper, whilst there is no external indication of a second cecum. The hind-gut is practically identical with that of Afe/es in its arrangement. In Ouacaria rubicunda there is no distinct duodenal loop. Meckel’s tract Js relatively much longer and is composed of irregular loops very closely set together. The single cecum is capacious, tapers to a point, and is slightly spirally twisted. The hind-gut, which is relatively short, consists of a well-defined colic loop passing insensibly into a straight rectal portion. The arrangements of the mesenteric veins are much as in Afeles. Sub-Order CATARRHINA. Family CERcoPITHECID &. Sub-Family CeErcopitHecIna. Cercopithecus niger, C. stairsi, C. werneri (fig. 44, p. 512), Macacus arctoides, M. rhesus (fig. 49, IIT, p. 521). In these Monkeys the duodenal loop is seldom distinct. Meckel’s tract consists of a number of simple minor loops and a terminal unconvoluted portion suspended at the periphery of a nearly circular expanse of mesentery, the terminal portion of which, however, acts also as a mesocolon. Meckel’s tract as a whole is relatively rather short, but is relatively longer in most Cercopithecin, and especially in the Macaci, than is represented in fig. 44. The cecum, which is always present, varies considerably. The condition shown in fig. 44, where the cecum is rather large, capacious, and slightly curved, I have found in other species of Cercopithecus, but it is usually relatively shorter and wider. The hind-gut is always long and extremely capacious. It displays first a portion continuous with the cecum; then a very large colic loop always thrown into a number of wide minor loops, the last of which may be regarded as a sigmoid flexure and which passes into the relatively short, straight rectum. The anterior mesenteric vein curves round the Meckelian mesentery and receives irregular branches from the minor loops, a large cecal vein, and one or more vessels from the proximal portion of the hind-gut. The posterior mesenteric vein forms a complex branching system which drains both colon and rectum. Sub-Family SemNopitHecina. Nasalis larvatus. In this sub-family the duodenum appears not to be separate. Meckel’s tract. is relatively very much longer than in the Cercopithecine, its minor loops being more numerous, closer set, and longer, but the general pattern formed is similar. The cecum is relatively smaller and is conical, but shows no trace of vermiform appendix. The hind-gut, except that in proportion to the fore-gut it is shorter, is similar in its differentiation. ‘The mesenteric veins repeat the arrangement found in the Cercopithecine., ia) . a YZ 512 DR. P. CHALMERS MITCHELL ON THE Family HyLoBatTipa, T have not myself had an opportunity of examining the intestinal tract of a Gibbon, but, from the published accounts of several authors, it is clear that the pattern does not differ in any important respect from that of the Anthropomorphe. The duodenum is not distinct as a separate loop. Meckel’s tract is relatively long and forms a series Fig. 44. Intestinal Tract of Cercopithecus werneri. Lettering as in fig. 43. of closely aggregated minor loops with a terminal unconvoluted portion arranged round three-quarters of the periphery of a nearly circular expanse of mesentery, the proximal portion of the colon occupying the terminal quarter. The cecum is short, nearly globular, and terminates in a very long twisted vermiform appendix. The hind-gut is also long, but is said to vary considerably in the different species. INTESTINAL TRACT OF MAMMALS, 513 It is differentiated into a proximal portion continuous with the cecum, and forming the ascending colon of human anatomy, and a very large and capacious colic loop which forms a series of minor loops, the most distal of which may be identified with the sigmoid flexure of human anatomy. Finally, there is a short rectum. Family ANTHROPOMORPH®. (rorilla savagei (fig. 45, p. 514), Anthropopithecus niger, Simia satyrus. In the Gorilla (fig. 45) there is no single loop to represent a separate duodenum. Meckel’s tract is relatively very long and consists of a large number of irregular minor loops very closely set together and forming a compact mass arranged round three- quarters of the circumference of a circular outgrowth of mesentery, the last quarter being occupied by the ascending colon. The cecum is short, forming an elongated oval globe with a very long twisted vermiform appendix. The hind-gut is capacious and its calibre is much larger than that of the fore-gut. It is differentiated into three regions—a proximal portion continuous with the cecum; a very large loop displaying three rather irregular wide loops, the most distal of which may be regarded as the equivalent of the sigmoid flexure; and a relatively short, straight rectum. The extensive attachment of the mesocolon to the great omentum has produced complexities in the veins. The anterior mesenteric vein is enormous and curves round the Meckelian mesentery, its distal end forming a loop with a factor of the right colic vein; it receives irregular branching tributaries from the minor loops of Meckel’s tract. A very large tributary of the anterior mesenteric corresponds to the right colic vein of human anatomy and drains by various branches the distal part of Meckel’s tract, the cecum, and a considerable portion of the colon. The posterior mesenteric vein forms a branching system draining the rectum proper and the remainder of the colic region of the hind-gut. In the Chimpanzee the arrangement of Meckel’s tract is similar to that in the Gorilla. The cecum is relatively larger and more capacious, whilst the long vermi- form appendix is less twisted. ‘The hind-gut has the same general size, capacity, and relations, but is less divided into definite minor loops. The general arrangement of the mesenteric veins is similar. In the Orang there is no distinct duodenum. ‘The minor loops of Meckel’s tract are arranged in a proximal and distal set, the expanse of mesentery on which they are suspended being deeply divided between these two regions. ‘The csecum and vermiform appendix differ very little from those of the Gorilla and Chimpanzee. The hind-gut shows three distinct regions. ‘There is a very wide portion forming one large loop corresponding to the ascending, transverse, and descending colons of human anatomy; then a very large region which may be regarded as the sigmoid flexure and is composed of a number of minor loops, and finally a short straight rectum. I find that I have not detailed notes on the mesenteric veins. al4 DR. P. CHALMERS MITCHELL ON THE ~ o>, ‘ a i Se, BS, = ibe Y \ 50 Intestinal Tract of Gorilla savager. S. Cut end of duodenum near stomach. Rk. Cut end of rectum near anus. C. Caecum. X, X. Cut ends of posterior mesenteric vein, 2. Cut ends of colic vein. 3. Cut ends of posterior colic vein. INTESTINAL TRACT OF MAMMALS. 515 As a group, the Simie display a quite definite pattern of the intestinal tract. The duodenum and Meckel’s tract together form a series of loops which differ from group to group in their relative complexity, arranged round about three-quarters of the circular outgrowth of mesentery. The cecum is always present and appears to have been originally capacious and of nearly equal calibre throughout its length; but it is in process of shortening throughout the group, being, as a rule, shorter in the Old World Monkeys than in the New World Monkeys and the Anthropoid Apes (if in the latter case the vermiform appendix be reckoned with the cecum). The state of the case may be put in another way. ‘The originally long capacious cecum of the Simie is retained by the greater number of the Platyrrhine Apes; in the Catarrhine Apes, except the Hylobatide and the Anthropomorphe, it tends to become shorter without the formation of a vermiform appendix. In the two last-named groups its proximal portion has remained capacious, but the greater part of its original length has been transformed without shortening into the thick-walled vermiform appendix. Family Hominipa. Human anatomy is so well known that it is unnecessary to do more than refer to the essential similarity between the pattern of the human intestinal tract and that of the Anthropomorphe. Probably primitive man was of omnivorous, with a tendency to carnivorous, diet, but the result has not yet been any marked adaptive change in the character of the gut as compared with that of the Anthropoid Apes, although in the latter the diet is omnivorous with the strongest leaning towards the vegetable side. Homo Loey or tHE MAMMALIAN Ca&cuM. Notwithstanding the existence of a voluminous literature on the cecum, I cannot find that any attempt has been made to institute a comprehensive comparison of the cieca in Mammals and Birds or to decide as to whether or no these structures are to be regarded as homologous. 1 have no doubt myself but that the structures in the two groups are homologous, and that any anatomist who had the opportunity of studying them in a large series of Birds and of Mammals would come to the same conclusion. ‘The anatomical evidence for the conclusion may be examined conveniently under three heads—the position of the ceca in the course of the intestinal tract, the occasional occurrence of paired ceca in Mammals, and the relations of the adjacent extremities of the ileum, ceecum, and colon in many Mammals. Position of the Ceca. In the vast majority of birds the hind-gut is relatively very short, and the paired ceca, accordingly, lie relatively close to the cloaca. In a Fowl and Pheasant, for instance, where the length of the intestinal tract from the stomach to the cloaca may 516 DR. P. CHALMERS MITCHELL ON THE be taken as about sixty inches, the ceca open into the gut between three and four inches from the cloaca. In the human being, where the length of the whole tract from the stomach to the anus may be taken as being about twenty-six feet, the cacum is placed at a distance of about five to six feet from the anus, whilst in many Mammals, as, for instance, a Rabbit or Sheep, the caecum lies very much nearer the middle of the length of the gut. In a study of the intestinal tract of Birds (15) | was able to show that in some Birds, ¢.g. the Struthious Birds, the hind-gut was relatively very much longer than in the majority of Birds, with the result that the caecum in such forms lies much more nearly in the common mammalian position—in the Ostrich, for instance, almost midway in the course of the gut. It may be that in the Ostrich the hind-gut has increased in length in the history of the transformation of a flymg bird into a large terrestrial form, but, on the other hand, it is equally clear that in the marjority of birds, and especially in the Passeres, there has been a phylogenetic shortening of the hind-gut with a consequent approximation of the ceca to the cloaca. On the other hand, if one looks through the figures 1 to 45 which illustrate this memoir, it will be seen that whilst in the majority of the mammalian types the cecum is nearer the posterior end of the intestinal tract than in the case of the majority of Birds, the relative position varies, and in many forms, especially Carnivora, it approaches the avian position. ‘The morpho- logical position, in fact, varies very little; the apparent position varies in proportion to the extent to which the hind-gut has departed from its primitive straight condition to be differentiated into a more or less complex series of loops. 1n the vast majority of Birds the hind-gut has apparently become shorter instead of lengthening; in a very considerable proportion of Mammals it has been lengthened. Occurrence of Paired Coca. In the vast majority of Birds the ceeca form a pair of ventro-lateral outgrowths from the wall of the gut. In certain cases, however, é. g. the Ardeidz normally, in Columbide, Vulturide, and Plotus as individual variations, one of the pair may be completely absent. In Mammals, on the other hand, the normal state of affairs is that one cecum only is developed. The exceptions are, however, more numerous than is generally supposed. Amongst the Edentata the existence of a pair of symmetrical ceca is well known. — I have found them myself in Tatusia keppleri (fig. 12, p. 458), Dasypus villosus (fig. 13, p- 459), Dasypus minutus, and Xenurus unicinctus. Rapp (23) has recorded their presence in Myrmecophaga didactyla and Dasypus sexcinctus, comparing them with the similar organs in Birds. Meckel (14), Owen (19), and Flower (5) have stated that they exist in Cyclothurus didactylus, whilst Flower has figured them in the last-named genus and in Dasypus sexcinctus. As I have already described and figured (fig. 14, p. 461), a symmetrical pair is present in Hyraa, in the paired ceca being obviously comparable with the normal cecum of Mammals in position. Finally, paired ceca occur in Manatus (fig. 15, p. 464). INTESTINAL TRACT OF MAMMALS. D517 Much less attention has been given to the equally interesting fact that there not infrequently occurs a paired condition of the ceca in which the usually suppressed member of the original pair is vestigial. Meckel (14) states that this occurs in some Marsupials. I have found it in several species of Wacropus (fig. 8, p. 452) andin Petaurus sciureus (fig. 47, Il, p. 518). I have already discussed its occurrence in the Wombat (supra,p. 449). Amongst the Edentates I found it in Bradypus infuscatus (fig. 10, p. 456). In the Common Hare (Lepus timidus) (fig. 46, 1) the distal end of Meckel’s tract is Fig. 46. C Cxeea of Lepus timidus (1) and Ochotona pusilla (11). C. Cut end of normal cecum. If. Cut end of colon. 2. Vestigial cecum. §.I. Cut end of small intestine. inserted between the ordinary very large cecum and a vestigial globular second caecum, A similar, but not so strongly marked, condition occurs in the Beayer (Castor fiber) and VOL. XViI.—Part y. No. 11.—December, 1905. 3 518 DR. P. CHALMERS MITCHELL ON THE inthe Capybara. A condition closely siniilar, but in which the facts of the case are more striking, occurs in Ochotona (fig. 46, I], p. 517), where the accessory cecum is a long papilliform hollow process. Since observing this in a specimen of Ochotona rufescens that died in the Zoological Gardens, I find that it was described and figured by Pallas (20) in the case of three species of what is now known as Ochotona. It is to be noticed, however, that in one of the three figures (Lagomys ogotona) a third unpaired ceecal diver- ticulum is figured. ‘The presence of the latter may, if confirmed as a normal occurrence, throw some doubt on my interpretation of the second cecum in Ochotona as the Fig. 47. Tleo-cxecal regions in Dasypus minutus (1), Petawrus sciureus (II), and Gazella maria (LI). 8.1. Cut end of small intestine. . Cut end of colon. a . Normal cecum. C2. Second cecum, symmetrical in I, vestigial in II, reduced to a mass of lymphoid tissue in ILL. W. Edge of ‘ window ” cut into wall of gut. I.A. Free tubular aperture of ileum to colon. B. Dotted lines show position of flaps of ileo-ceecal valve. vestigial representative of the second member of a primitive pair. In several Antelopes, e.g. Gazella marica (fig. 47, III), the vestigial presence of a second cecum is strongly suggested by the configuration of the junction of the ileum and colon and by the presence of amass of lymphoid tissue (at C2 in fig. 47, III) at the point where the second member of a pair would lie. Comparison of the three drawings in fig. 47 will, I think, lead any anatomist to the INTESTINAL TRACT OF MAMMALS. 519 conclusicn that the same structures are involved in each case. The drawing labelled with the Roman numeral I represents the junction of the ileum, colon, and symmetri- cally paired ceca in Dasypus minutus. A “window” was cut in the wall of the gut to show the relation of the posterior end of the ileum with its projecting, tubular orifice placed between the orifices of the ceca. The apparent discontinuity between the ileum and the colon is seen to be a simple consequence of the outgrowth of the paired ceca. The drawing labelled II in the same figure (47), which represents the condition in Petaurus sciureus, is obviously identical morphologically with the condition in Dasypus, with the difference that the cecum marked C2 is much reduced. In the Gazelle (fig. 47, III) the condition is again obviously repeated, but one cecum is still further reduced, whilst the aperture of the ileum instead of being projecting and tubular is a shit between two valvular flaps, the position of which is represented by dotted lines. Relation between the Ileum, Cecum, and Colon. The very peculiar valvular arrangements that occur where the ileum, cecum, and colon meet have been described and studied in a number of cases, but, so far as I am aware, the attention of observers has been directed chiefly to the functional peculiarities of this region. The structures are generally adapted to a greater or less extent to allow free exit of the contents of the ileum into the lower bowel and free communication between the cecum and colon without regurgitation of the contents of the latter organs into the ileum. A further adaptive series of modifications varies with the degree to which the ceecum is modified as a glandular rather than an absorbing or digesting organ. Behind the adaptive modifications, however, lies the question of the morphological material from which these modifications have been produced. ‘The latter vary very much from genus to genus, sometimes from species to species, but throughout the series there is the common feature that they are easy to understand ifit be supposed that originally the ileum opened into the colon between symmetricaily placed ceca, and difficult to understand if the traditional view be accepted that the mammalian cecum is an unpaired structure. In fig. 48, p. 520, the condition is represented as it occurs in two animals very far removed in the mammalian series, fig. 48 | being drawn from Badbirussa and fig. 48 IL from Mycetes beelzebul. In both cases the colon, cut across at H, widens considerably before it joins the dilated conical cecum C. Windows have been cut in the wall of the gut to show the relation of the tubular free edge of the ileum, which projects into the expanded end of the colon precisely as if a hole had been cut in the wall of the latter and the tubular end of the ileum pushed through into the cavity. Just anterior to this in each case is the aperture of the cecum with a raised oval lip. If it be supposed that in the region of the colon vertically under the letters LA. in the drawings there was formerly a second cecum, the anatomical relations at once become simple. BAR] 520 DR. P. CHALMERS MITCHELL ON THE In fig. 49 the same region is represented as it occurs in three Monkeys, fig. 49 I being a Callithrix, 11 being a Cebus, and III a species of Macacus. In these the only dissection that was made was the cutting of windows in the wall of the gut, The Fig. 48. Tleo-caecal regions in Babirussa babirussa (1) and Mzycetes beelzebul (11). C, Caecum. S.I. Cut edge of small intestine. H. Cut edge of colon. I.A. Free tubular aperture of small intestine to colon. C.A. Aperture of cecum to colon, C. Caecum, W. “ Window” cut in wall of gut. aperture of the ileum in each case is a tubular projection; in Callithrix the aperture into the cecum is also narrow and tubular, in the others it is in wide INTESTINAL TRACT OF MAMMALS. 521 communication with the colon. Here, again, it is obvious that the condition is easy to understand if it be supposed that in the ancestral form the ileum opened into the colon between a pair of symmetrically placed ceca. The diagrammatic outlines in fig. 50, p. 522, show what appear to be the obvious interpretation of the Mammalian cecum. In the upper figure the primitive paired condition is shown; in the lower figure the only modification is the suppression of one member of the pair. Fig. 49. eo-ceeal regions in Callithrix cuprea (1), Cebus fatuellus (II), and Macacus rhesus (III). CinI. Cut end of cecum, C.A. Projecting tubular aperture of cecum to colon. Other letters as in fig. 48, I conclude, then, that the ceca in Birds and Mammals are homologous. Primitively they are paired outgrowths of the hind-gut situated near the point where the primitive straight rectum joins what is known as the pendent loop. In Birds they appear, in the adult and in the embryo, nearly at the point where the pendent loop joins the straight rectal portion, this position being due to a phylogenetic shortening or failure of expansion of the hind-gut. In Mammals they lie some way on the posterior limb of the dorsal pendent loop, and this position is due to an expansion of the hind-gut. In Birds, typically, both cca develop symmetrically, but there are cases where only one develops. In Mammals, typically, one caecum degenerates, but there are cases where both are developed symmetrically, and very numerous cases where one remains in a more or less vestigial condition, ‘The irregularity of the distribution of symmetrically 522 DR. P. CHALMERS MITCHELL ON THE paired czeca, and of vestigial remains of a second ceecum, through the Orders and Sub- Orders of Mammals, is in harmony with the view that all Mammals are descended from creatures with paired ceca, and that for some unknown reason there has been a tendency throughout the group for one of the members of the pair to degenerate. ForM AND FUNCTION OF THE CCUM. In analysing the facts known as to the condition of the ceca in Birds, I have been able to show (15) that there is no definite invariable connection between the nature of the food and the degree of cevelopment of the ceca. Whilst, on the whole, it may be Fig. 50. C2 1A / S| n Diagrams of Mammalian Ieo-cwcal Region. I. Primitive stage with paired cxca. IL. Stage in which one cecum has been lost. C, C2. Ceca. S.I. Cut end of small intestine. L.A. Free tubular aperture of small intestine to colon. H. Cut end of colon. said that in herbivorous forms the ceca are generally large, and generally small in insectivorous, frugivorous, carnivorous, and piscivorous forms, there are many exceptions and difficulties in the way of accepting such a generalisation. Thus the Owls and the Falconide have a diet that is closely similar, and, none the less, the Owls have a pair of very long czeca, whilst in the Falconide these organs are very much reduced and apparently functionless. The insectivorous and omnivorous Rollers, Motmots, and Bee-eaters have a pair of large ceca, whilst in Passerine Birds of similar habit the ceca are vestigial glandular nipples. It is impossible to doubt that family-history dominates function in this matter. Certain families tend to retain the ceca; others INTESTINAL TRACT OF MAMMALS. 025 to lose the ceca, and direct adaptation to diet appears only to accelerate or to retard these inherited tendencies. So also in Mammals, no more than a general relation between diet and cecal development can be shown to exist. First, there is no relationship between diet and the complete or partial presence of a pair of ceca. ‘The occurrence of such a pair appears to be rather an “accident” of inheritance than imany relation to function, an accident more common in less specialised forms. The large size of a single cecum, in Mammals, is rather more definitely correlated with a herbivorous as opposed to a frugivorous, carnivorous, piscivorous, or omnivorous diet than is the case with Birds. Thus it is small in the Monotremes ; large in all the Marsupials except the Carnivorous Dasyuride. In the Ungulates the cecum varies in size considerably, and Maurice Schiff (24) endeavoured to establish the existence of a correlation between the complexity of the stomach and the size of the cecum. In the Ruminant Ungulates, for instance, where the stomach is large, the cecum is relatively small; in the Horse and Tapir, where the stomach is simple, the cecum is relatively large; inthe Pigs, which have a simple stomach, the cecum is again capacious; whilst in the Hippopotamus, where the stomach is complex, the cecum is absent. I find it difficult, however, to follow this comparison far. Many of the Antelopes and Gazelles have relatively large ceca, although in them the stomach is complex. ‘The Rhinoceros, which has a simple stomach, has a cecum not relatively very large. ‘The Tragulide, with a complex stomach, havea capacious cecum. In the Rodents, the diet ranges from being herbivorous to practically omnivorous, and the cecum is always relatively large ; ] find it impossible, however, to see any correlation between the comparative size and the extent to which the diet is strictly herbivorous. Outside the limits I have mentioned the correlation between size of the ceecum and diet breaks down completely. ‘here is no distinction between the plant-eating and insect-eating Edentates in this matter, The Sirenia are all strictly herbivorous, but amongst them /thytina alone seems to have possessed a really capacious cecum. In the small Carnivora, many of which are practically vegetable-feeders, there is no relation between the diet and the cecum. Amongst the Primates, the cacum is always large in Lemurs, whether these be omnivorous, frugivorous, or insectivorous; and in the Monkeys and Apes there is always a cecum, varying from group to group in capacity and without direct relation to the nature of the food. Where the cecum is normally small, there is no direct relation between diet and complete absence. It is absent in the piscivorous Dolphins, present in the equally piscivorous Seals. So far as I can find out, those Insectivora and Chiroptera in which the caecum is present do not differ, so far as diet is concerned, from the members of these Orders in which the caecum is normally absent. In the case of reduced ceca, it is necessary to remember, as Birmingham (2) has suggested, that the formation of a vermiform appendix may be not degeneration, but the elaboration of a new highly glandular organ. An elaborate comparative study of 524 DR. P. CHALMERS MITCHELL ON THE the functions of the different parts of the intestinal tract is still necessary before the extent to which adaptation has modified the cecum can be properly estimated, but, so far as I can judge from the anatomical data assembled here, it seems a reasonable inference that ancestral history is at least as important as adaptation to present function. 6 SECONDARY RELATIONS BETWEEN THE PROXIMAL AND DistaL Portions OF THE INTESTINAL TRACT. In my memoir on the intestinal tract in Birds, I called attention to a peculiar relation between the proximal and distal portions of the gut. In the natural condition the duodenum, which in Birds is almost invariably a separate, distinct loop, lies ventrally to the rest of the gut. Cuvier called attention to the close interconnection between the duodenum and a distal loop of the gut, terming the latter loop the colon. The term is confusing, because the Mammalian colon is a region distad of the caecum or ceca, whereas in Birds the portion of the gut which comes into intimate relation with the duodenum is proximad of the cca, and is indeed the most distal part of Meckel’s tract. In Birds, Meckel’s tract displays a strong tendency to be divided into distinct minor loops, and it is the most distally placed of these loops that enters into relations with the duodenum, and to which I therefore gave the name “ Supra-duodenal Loop.” In the more primitive types of Birds this interconnection is but little marked and has but little importance. In the higher members of each Avian series, however, there is a tendency for the interconnection to be very highly developed. In the Passeres, for instance, the supra-duodenal loop is very accurately moulded on the duodenum, these two regions of the gut lying in close application, with a partial fusion of their mesenteries, and with a complex interchange of blood-vessels and nerves. The possibility of this interconnection having been formed is due to the mode of growth of the gut, the greater part of the length in the adult being formed from only a very narrow portion of the primitive straight antero-posterior tube, so that loops which may be separated by a long stretch of gut in the adult may come into close contact as they hang suspended from a very narrow attachment to the dorsal wall of the cavity. But although the possibility of developing such interconnections exists for many regions of the fore-gut in Birds, the interconnection has been developed so constantly and progressively between the duodenum and the supra-duodenal loop, and so slightly and so rarely in any other region, that there is a strong suggestion of some functional adaptation. In Mammals, as in Birds, the greater part of the length of the gut has similarly arisen as the outgrowth of only a very small part of the primitive straight gut, and there is thus a similar opportunity for the establishment of interconnections being formed between the closely bunched loops. In Mammals, as in Birds, the actual interconnection is limited to a definite region. The region, however, is different in the two cases. In Mammals the duodenum is very seldom a definite loop; Meckel’s INTESTINAL TRACT OF MAMMALS. 525 tract, moreover, is never arranged in a series of definite loops, the number, position, and shape of which are of systematic importance, as is the case in Birds. Thirdly, a narrow portion of the originally nearly straight hind-gut, distad of the ceca, a region which in Birds practically invariably remains primitively short, may grow out to form long colic loops. The interconnection that takes place in Mammals is between one of these colic loops, a portion of the hind-gut distad of the ceca, and a portion of Meckel’s tract not limited to the duodenum. Moreover, the interconnection is different in its nature in the two groups. In Birds the supra-duodenal loop is applied directly, and with an accuracy that becomes more perfect in the higher members of each series, actually to the fore-gut itself, as it forms the duodenal loop. In Mammals the colic loop is applied to the mesentery that suspends Meckel’s tract along the bases of the minor loops of that tract, so that the colic loop is applied to an area of Meckel’s tract very much shorter than itself, following the sweep of the mesentery, but not imitating the actual loopings of the small intestines. However, the interconnections of the mesentery and of the blood-vessels may be elaborated in Mammals as in Birds. Thus, out of materials that are not homologous, a very remarkable and apparently adaptive relation has been produced in the two groups. I may now give a systematic summary of the extent to which this interconnection of the gut occurs. Monorremata. Slight, and with the duodenal region in Ornithorhynchus; practically absent in Echidna. Marsupratia. Practically absent throughout. Tusuttpentata. No information. Puourwora. Absent. Xenartura. Absent. Hyracorpea. Absent. Srrenza. Absent. Progoscrpea. Absent. Ceracea. Absent. ARTIODACTYLA. Non-Rominantia. Absent. TracutorpEa. A large colic loop, distad of the spiral coil, is folded against the mesentery of Meckel’s tract, but the connection is not so intimate as in the Pecora. Tytoropa. A colic loop of moderate length folded over against the Meckelian mesentery, but the interconnection not so definite as in the Traguloidea, and not nearly so definite as in the Pecora. Pecora. In all these the interconnection is extremely definite, bemg much more elaborate than in any other Mammalian group. The very large colic loop is folded so closely against the Meckelian mesentery, following the latter along the curved line of origin of the minor loops of Meckel’s tract, and the mesenterial fusion and interconnection of blood-vessels is so complete, that an elaborate dissection is required to separate the two regions of the gut. Anyone who observes this in a number of Pecora must be VOL. XVI.—PaART vy. No. 12.— December, 1905. 4 526 DR. P. CHALMERS MITCHELL ON THE impressed by the probability of there being some physiological explanation of this elaborate arrangement. Ropentra. In this group two sets of interconnections exist. The colic region is divided in most cases into two loops. Of these, that nearer to the cecum is involved in the spiral twisting of the cecum, and a certain amount of fusion of mesentery and blood- vessels takes place, the amount varying with the length and degree of twisting of the cecum. The second colic loop is the representative of the region that forms the inter- connection in the Pecora. In different Rodents almost every degree of fusion between this loop and the Meckelian mesentery occurs, reaching, in Lepus, a stage which is extremely like that found in the Pecora, and forming a striking case of convergent resemblance. Insectivora. Practically absent. Curroprera. Absent. Carnivora. Practically absent. Prostmiw. In the Lemurs there is the beginning, and not much more, of a connection between the usually large colic loop and the Meckelian region. Smrz. In this group, whilst it never becomes so elaborate as in the Pecora and higher Rodentia, there is almost invariably an interconnection between the colic loop and the Meckelian region, the mesenterial fusion varying considerably, but the interconnection of the blocd-vessels by special loops being sometimes considerable. Homopuastic MopIFICATIONS. Before attempting to decide to what extent the characters of the gut-patterns in Mammalia throw light on the affinities of the mammalian groups, it is necessary to consider briefly how far the characters of these patterns are modified in accordance with function. The mammalian gut is, of course, the same organ throughout the group, and its general function and mode of action are alike ; it is to be expected that where Mammals belonging to different Orders have become adapted to similar food, the gut will have become modified in similar directions. Obviously such similarities can be no guide to affinity; they are what I have termed “ multi-radial apocentricities,” peculi- arities of structure that may have been independently repeated in different cases; and, before drawing conclusions as to affinities, it will be necessary, so to speak, to correct the gut-patterns in these respects, to eliminate from them the convergent resemblances. It is important to notice, however, that in the mammalian gut, however great may be adaptive resemblances, the inherited element dominates the structure. Take the broadest example of this dominance of the inherited factor, of the nature of the material on which the adaptations have to be impressed. The intestinal tract of Birds must be closely similar in its origin to that of Mammals, and different groups of Birds have become adapted to very much the same different sets of food as in the case of Mammals. None the less, as I have already pointed out, underlying these adaptations INTESTINAL TRACT OF MAMMALS. 527 there is an essential difference between the general mode of modification of the Avian intestinal tract and that of Mammals. In Birds the region known as Meckel’s tract tends to be so modified as to display individual loops, the number, position, and character of which differ in the different groups. In Mammals such a modification or series of modifications of Meckel’s tract never takes place. In Birds, when the hind- gut becomes long it forms irregular minor loops like those of Meckel’s tract in Mammals; in Mammals, when the hind gut becomes long it tends to form minor loops, the number, position, and character of which differ in different groups, just as is the case with the fore-part of the gut in Birds. In Birds great increase of size of the czeca is associated, on the whole, with particular diets; in Mammals it is similarly associated with particular diets ; but whilst in Birds both ceca elongate, in Mammals it is only one ceecum that elongates. I may turn now to examples within the group of Mammals. There is a general similarity in the nature of the food of the herbivorous Marsupials and the herbivorous Eutherian Mammals. A Kangaroo and a Sheep in the state of nature live on food of a very similar kind, and in captivity thrive on practically identical food ; and yet, notwithstanding the adaptations to a strictly herbivorous diet in the two cases, there is a profound difference in the gut-patterns. Similarly, Dolphins and Seals live on very much the same kind of food, and yet the character of the intestinal tracts in the two is profoundly different ; the limits of the homoplastic modification are definitely set by the inherited material. Finally, to take an example of an opposite nature. ‘The terrestrial Carnivora, from the Lion to the smallest Suricate, display a pattern of intestinal tract essentially similar. And yet, although we give the name Carnivora to the group, almost every possible kind of diet is found amongst them—purely carnivorous, piscivorous, omnivorous, frugivorous. The most important of the homoplastic modifications that have to be allowed for are as follows :-— Lengthening of the Gut in Herbivorous Forms. Precisely as. happens in the case of Birds, so in Mammals there is a direct association between the herbivorous habit and a great length of gut. The explanation of this, no doubt, is simply that the vegetable matter which such creatures devour is in a form which requires not only prolonged digestive action, but from the intimate admixture of indigestible material a very large absorbing surface. In the case of Meckel’s tract, the lengthening results simply in a greater number and closer grouping of the minor loops. With regard to the hind-gut, every herbivorous Mammal has a relatively long and capacious hind-gut, but this relative lengthening is less conspicuous in the Marsupials, and is not associated witha specialisation into regions. 4a 2 528 DR. P. CHALMERS MITCHELL ON THE Lengthening of the Gut and Thickening of its Walls in Piscivorous Forms. This homoplastic modification is as well marked in Mammals as in Birds. The gut of fish-eating forms is very long, with very thick wall and relatively small calibre. The explanation, no doubt, is similar to that of the herbivorous modification. Fish, as it is eaten by wild animals, contains a large bulk of indigestible matter, and so requires an extended absorbing surface; moreover, no doubt as a protection against wounding by bones, the wall is specially strong and the calibre narrow. The modification applies to Meckel’s tract. Where it is possible to distinguish the limits of that region and the hind-gut, it is seen that the hind-gut of piscivorous Mammals is never specially long or capacious. Shortening of the Gut in Frugivorous Forms. Curiously enough, whilst in frugivorous Birds the gut is always strikingly short, wide, and simple, a similar change has not taken place in Mammals. Frugivorous Bats have the gut rather longer than insectivorous forms, and amongst Lemurs and Monkeys there is no evidence for the existence of such a modification. Increase of Length in Large Forms. In Birds the relation of the length and calibre of the gut to the size of the whole creature is striking. If two birds of similar habit and of the same group be compared in this respect, it will be found that the gut of the larger bird is relatively longer rather than relatively wider. The same general rule applies to Meckel’s tract in Mammals ; when a smaller Mammal is compared with an allied Mammal larger in size but of the same general habit, it will be found that the larger animal has Meckel’s tract relatively longer but not relatively wider. In the case of the hind-gut, however, on the average, increase of capacity is given by increased calibre rather than by increased length. SUMMARY OF SYSTEMATIC INFERENCES. When an assemblage of anatomical facts has been simplified, so to speak, by eliminating the element due to convergent modification, 7. ¢. those resemblances that they share because of adaptation to similar functions, a further scrutiny is necessary before drawing, from the likenesses and differences displayed, inferences as to systematic alliance. Likenesses which are due to the common possession of primitive features cannot be regarded as evidence of near relationship ; that certain members of a group have retained what was once the property of all the members of that group can be no reason for placing such creatures close together in a system, if the system is to be based on blood-relationship. Resemblances that are due to the loss or reduction of parts that once were the property of the ancestral stock cannot afford a clear ground INTESTINAL TRACT OF MAMMALS. 529 for inferring systematic affinity, since there is no difficulty in supposing the same organ to have been reduced or lost in independent branches of the common stock. Resemblances, on the other hand, that are new acquisitions, that depend on definite anatomical peculiarities, must be the most likely field for the discovery of clues to affinity, simply because it appears to be less probable that the same anatomical device should have been produced independently than that it should have been acquired only once. The value of the clue afforded by such resemblances will be in direct relation to the complexity of the anatomical structure. It will be convenient to set out from a conception of what may be taken to be the most primitive, generalised, or archecentric condition of the mammalian intestinal tract. From a general review of the characters found in the different existing groups, I infer that in its most generalised, and probably ancestral, condition the mammalian intestinal tract consisted of a tube not very much longer than the length of the body. It was suspended in simple folds at the periphery of a dorso-ventral mesentery stretching from the stomach to the cloaca. It was of fairly even calibre throughout its length and was divided into two regions, probably not far behind the centre of its length, by a pair of ceca of moderate size. It was drained by a longitudinal vein running a nearly straight course in the mesentery from the cloaca to the liver and receiving minor tributaries from the loops of the tract suspended at the edge of the mesentery. Of these minor tributaries, one of conspicuous size came from the primitive yolk-sac and afterwards formed the main anterior mesenteric vein which in most Mammals, in association with the growth of Meckel’s tract, appears to be the principal vein of the portal system, the primitive longitudinal vein being reduced to the comparatively unimportant posterior mesenteric vein. In the Monotremata, Meckel’s tract has become lengthened in the characteristic mammalian fashion, whilst the hind-gut is relatively long and shows in a simple fashion the typical mammalian specialisation into colic loop and rectum. The single cecum is degenerate. ‘There are no positive indications of affinity. In the Marsupials, when allowance has been made for the increase of length due to vegetarian habit, there is no important difference between the groups. The intestinal tract is more generalised than that of Monotremes, this being shown in the frequently great proportional length of the hind-gut without marked differentiation of that region into distinct portions, and in the frequent traces of a paired condition of the ceca. The so-called Edentata show characters as diverse as is to be expected from the heterogeneous nature of the group. The Xenarthra cannot be divided in accordance with vegetable or animal diet, but display a primitive simplicity very like that in the Marsupials. Meckel’s tract is lengthened, but the hind-gut remains conspicuously long without well-marked differentiation into colic and rectal portions. ‘Traces of the primitive paired condition of the ceca are even more conspicuous than amongst 530 DR. P. CHALMERS MITCHELL ON THE Marsupials. The Tubulidentata appear to have a simplicity like that of the Xenarthra, but still more like that of Marsupials. The Pholidota stand far removed from any other Mammals, their modification of the primitive simplicity being peculiar to the group and recalling the pattern in some birds. The Hyracoidea show a pattern which, except for one peculiarity, has moved little from the primitive mammalian type. Meckel’s tract is long and simple but for the presence upon it of the accessory cecum. ‘The hind-gut is relatively long and has no very marked differentiation. The paired ceca are primitive. There is a total absence of the special peculiarities found in the Ungulata or Rodentia. The Sirenia are almost diagrammatic in their primitive Mammalian simplicity. Meckel’s tract and the hind-gut are almost equal in length, and are suspended by a practically continuous mesenteric fold. ‘The two regions are marked off by the ceca, which in one member of the group are paired. ‘There is no trace of any of the Ungulate specialisation. : The Proboscidea are almost as primitive as the Manatide, the intestinal tract consisting of a Meckel’s tract and a hind-gut not very different in length suspended on a continuous mesentery. The duodenum, however, is more distinct, and there is a single cecum. ‘There is no sign of any special Ungulate character. In the Cetacea, the Odontoceti show a much elongated gut of great apparent simplicity, no distinct regions being displayed. From the evidence afforded by the Mystacoceti, however, it is plain that the simplicity of the Toothed Whales is secondary, and that the Cetacean stock contains the ordinary mammalian pattern of the intestinal canal in a simple form, from which no suggestion of affinity can be derived. In the Artiodactyla an interesting set of conditions is displayed. In all the sub- groups of the order Meckel’s tract is relatively long and simply arranged. The hind- gut, however, is also long, being either as long as, or relatively longer than, in the primitive mammalian type. It displays, however, a tendency to be modified in a definite direction, peculiar to the Order. The cecum varies, but in a number of cases shows evidence of a primitive paired arrangement. ‘The special Artiodactyle peculiarity consists in the tendency of the colic region to be greatly elongated in the form of what is really a single loop, in the course of which minor loops may be developed. This loop outgrows the primitive mesentery, and is frequently supported by special bands, or is folded against, and attached secondarily to, the Meckelian mesentery. This colic loop, moreover, tends to assume a spiral arrangement. In the Non-Ruminantia the colic loop is enormously long and capacious, its proximal limb being especially large in calibre. It is arranged in a rude conical spiral, the tip of the loop forming the apex of the spiral. In the 'Traguloidea the spiral coil is very short and of even calibre, consisting of not more than two or three turns, which, moreover, are arranged in an irregular pyramid. 31 or INTESTINAL TRACT OF MAMMALS. Distad of this, the hind-gut displays a second peculiarity which I have described as the supra-Meckelian fold. In the Tylopoda the spiral coil is enormous, of even calibre, and arranged as a flat watchspring. The supra-Meckelian fold is well developed. In the Pecora the spiral coil is extremely long, very narrow in calibre, and the flat watchspring-Jike disposition is typical. The supra-Meckelian fold is equally typically developed. Additional extensions of the hind-gut appear irregularly as loops proximad of the spiral coil, or distad of the supra-Meckelian fold in the rectal region. The Artiodactyle colic region, then, displays a definite case of an anatomical peculiarity, so well-marked and complex as to be a safe guide to affinity. This consists in the elongation and spiral coiling of the proximal portion of the colon. In the Non-Ruminantia it has developed in one direction; in the Traguloidea it has remained in the primitive Artiodactyle stage. In the Tylopoda and Pecora it has developed in a characteristic direction different from that in the Non-Ruminantia. The evidence given by the supra-Meckelian fold, however, which is common to the Traguloidea, Tylopoda, and Pecora, would seem to associate these three more closely, to the exclusion of the Non-Ruminantia. In the Perissodactyla, as in the Artiodactyla, the characteristic differentiation is in the hind-gut. That region is long and very capacious, and is marked by the enormous development of the colic loop. In all three, the latter is a very long, narrow, simple loop, the proximal and distal limbs of which are closely approximated and are of huge calibre. ‘There is perhaps a remote resemblance with the Non-Ruminant Artiodactyles, but in the Perissodactyles there is no trace of the minor expansions of the colic loop or of the spiral coiling which is the characteristic feature of the Artiodactyle arrangement. The Tapirs, Rhinoceroses, and Horses, far apart as they are in many ways, have retained an identical mode of adaptation to vegetable diet, thus showing a dominance of the inherited element over the purely adaptive element. The Artio- dactyla and Perissodactyla have in common little more than that in both the primitive pattern of the hind-gut has been modified by adaptation to a herbivorous diet. This modification is essentially an elongation and specialisation of the colic region, but the mode of the elongation differs in the two groups. The Rodentia have moved away from the primitive mammalian type in a definite and characteristic fashion. Meckel’s tract tends to be elongated, but retains a comparatively simple arrangement. ‘There are abundant traces of a primitive double condition of the ceca. The hind-gut has increased in length and this increase tends to assume the form of a pair of long colic loops, which in some of the larger forms are extremely complicated in the adult, and which in some of the smaller forms are partly merged in one another. ‘The characteristic peculiarity, however, is that the normal cecum, which usually becomes very capacious, tends to be spirally coiled, the coiling involving 532 DR. P. CHALMERS MITCHELL ON THE in extreme cases the proximal colic region and the distal part of Meckel’s tract. This disposition of the increased length of gut, consequent on adaptation to vegetable diet, is peculiar to Rodents. In most of the living Insectivora the pattern of the intestinal tract is of reptilian simplicity, consisting of a series of very simple short loops suspended by a continuous mesentery, without ceca or differentiation into regions. There is evidence, however, that this simplicity is secondary. Some genera have a cecum, and in some there is a differentiation into fore-gut and hind-gut, whilst the latter again can be divided into colic loop and cecum. On the evidence, we can say only that the Insectivora have departed from the ancestral mammalian type by reduction of the gut. The Chiroptera display the same: conditions as the Insectivora, save that the reduction has gone further, and the traces of former complexity are rarer. The Galeopithecidze, so far as I can gain information from other authors, are much less reduced than the Insectivora or Chiroptera, and display at least the primitive mammalian differentiation of the gut, recalling the Prosimie rather than the Insectivora. The Carnivora show a reduction from the pattern of the primitive mammalian intestinal tract so far as the hind-gut and ceecum are concerned, whilst Meckel’s tract is frequently elongated. ‘The varying conditions of the ceecum from complete absence to moderate length, and the indications of former greater length of the hind-gut, are scattered throughout the group in the familiar mode of degenerate structures. In a broad way, however, it may be said that those Carnivora which are most specialised in other directions are most degenerate as regards the ceecum and hind-gut. The intestinal tract of the Prosimiz is not far removed from the primitive mammalian pattern, Meckel’s tract remaining simple and not relatively long, whilst the hind-gut, in nearly all the members of the group, is very long and capacious, and divided into colic loop and rectum. ‘The colic loop tends to grow out in a single, long and narrow loop, sometimes simple, sometimes with subsidiary folds on its course, suggesting, perhaps, a remote connection with the Ungulate stock. The Simize, with which, for the present purpose, the Hominidee may be reckoned, display a type of intestinal tract which in many respects is simpler than that of the Prosimiee and would appear to have been independently derived from the primitive mammalian pattern. Meckel’s tract varies, but tends to lengthen and to be arranged in a closely set series of long minor loops. Traces of a primitive paired condition of the cca are frequent, but the normal single cecum is not so long as that found in the Prosimiz. The hind-gut is usually relatively long and capacious, but shows no signs of the formation of the long narrow colic loop found in the Prosimize. In the simpler cases, 1t is composed of a series of minor loops suspended by a mesentery which is practically continuous with the Meckelian mesentery, so that the configuration of the whole intestinal tract (e.g. fig. 44, p. 512) recalls that of some of the simplest Mammalian INTESTINAL TRACT OF MAMMALS. 530 types. In many of the New World Monkeys this simple hind-gut has degenerated, becoming shorter. In the Old World types the primitively undifferentiated simple loops of the hind-gut tend to appear as the definite structures known to human anatomists as the very large loop composed of ascending, transverse, and descending colon, and the irregular kinks known as the sigmoid flexure. A characteristic feature of this group, which does not appear in any marked form in the Prosimiz, is the elaboration of the connection between the colic region and the mesentery of Meckel’s tract. If, then, we were to arrange the existing Mammalia solely on the evidence to be derived from the characters considered in this Memoir, we should recognise an ancestral group the members of which would be associated merely because of their possession of an ancestral type of intestinal tract, and for the separation of which consideration of other characters would be necessary. This group would contain :— MarsuPIALIA. XENARTHRA. TUBULIDENTATA. GALEOPITHECIDE. From this primitive centre eleven radiating streams have been formed with only in one or two cases any indications of the relative position of the streams. Hach stream sets out from the primitive centre and proceeds in a definite direction. HYRACOIDEA. 1.) SIRENIA. PROBOSCIDEA. This first stream has moved very little from the primitive type, aud its members are associated rather by common absence of differentiation than by peculiarities. 2. MoNOTREMATA. Not far removed from the centre, but along a definite radius. 3. PHOLIDOTA. The group has moved far, and in an absolutely definite way. 4, CETACEA. More complete information wanted, but, so far as known, the group has moved along a definite radius. 5. UNGULATA. These, again, are on a very definite radius, the most specialised members having moved very far from the centre. The Perissodactyla first branched off from the radius near the centre; next the Non-Ruminantia have separated, and further out the Traguloidea, and, finally, towards the furthest point of the radius, the Tylopoda and Pecora, VoL. xvit.—part v. No. 13.—December, 1905. 43 534 DR. P. CHALMERS MITCHELL ON THE 6. Prosi. There are suggestions that the radius on which the Prosimie lie is confluent at first with the Ungulate radius or at least parallel with it, but the group has not moved far from the ceutre, and certainly leaves the Ungulate radius before the specialisation of the Perissodactyla. 7. SIMLA. The radius of this group is close to that of the Prosimize and Ungulata, but its least specialised members have not moved so far from the common centre as the lowest Ungulates and Prosimize, whilst the more specialised members move outwards along their own radius. @ 8. RopENTIA. A distinct radius. along which the more specialised members have moved far from the centre. 9. INSECTIVORA. 10. CuIRoprera. 11. CaRNIVoRA. These three groups, so far as their lower members are concerned, have not moved far from the common centre. The higher members show a similar mode of speciali- sation, which consists of reduction of the hind-gut, and whilst probably, on the evidence of the intestinal tract, they would be placed close together, the evidence is not conclusive, as it may well be supposed that the reduction could have taken place independently. GENERAL CONCLUSION. The positive systematic inferences to be drawn from this assemblage of anatomical data are obviously extremely limited, but I am of opinion that a similar scrutiny of any other assemblage of data, conducted so as to exclude in the same way the drawing of inferences based on the common possession of primitive characters, would give equally limited results. Approach to correct classification, based on affinity, can be made only by combining the few positive results that examination of each series of anatomical data would yield. On the other hand, it appears to me that an examination such as I have endeavoured to make in this memoir has an important bearing on the theory of evolution. We have no information as to why the different groups of animals came to occupy the particular radiating streams of differentiation on which, with regard to any particular set of organs, they are found to lie. The initial deciding factor may have been direct adaptation, the selection of small variations or of large variations, or correlation with some such adaptation or selection. But once any set of animals has come to occupy a particular radius, then, in that respect, its possible range of variation On os (oni INTESTINAL TRACT OF MAMMALS. - is limited and defined. Whatsoever influences come to work upon it, however the selection or the direct moulding of the environment may act upon the stock, the possibilities are limited and defined by the past history. ‘The changes must run along certain grooves. Whatever forces operated on the diverging members of the Ungulate stock, for instance, had, so to speak, to reckon with the definite Ungulate modification that had taken place in the pattern of the intestinal tract. The changes and elaborations that could take place in the intestinal tract of the Ungulates were so far definitely limited; the direction of the movement was limited and could lead only to further specialisation of the limited material, or, if that were impossible, to extinction. And the further along each radius a set of organisms moved, so the inevitable unlike- ness from organisms on other radii must have increased. ‘The existence of such a principle of the formation of definite grooves along which the specialisation of organisms must necessarily move has been recognised by many writers; but amongst these not a few have endeavoured to read into the phenomena some mysterious, ultra-physical directive force, or, on the other hand, to seek out “selection-value” for each anatomical disposition. It is at least simpler first to exhaust the obvious interpretations, to be certain that there is a mystery, before turning to a mysterious explanation. List oF REFERENCES. 1. Bepparp, F. E—Mammalia. Cambridge Natural History. 1902. 2. Brrmincuam, A.—Some Points in the Anatomy of the Digestive System. Journ, of Anat. XXxV. pp. 33-36. 3. Burmeister, H.—Beitriige zur naheren Kenntniss der Gattung Tursius. Berlin, 1816. 4. Cuyrer, G.—Legons d’ Anatomie comparée. 2nd edition, 1835. 5. Frower, Sir W. H.—Lectures on the Comparative Anatomy of the Organs of Digestion of the Mammalia. Medical Times and Gazette, 1872. 6. Garrop, A. H.—On the Brain and other Parts of the Hippopotamus. ‘Trans. Zool. Soc. mi. pp- 11-17. . Garrop, A. H.—On the Cecum coli of the Capybara (Hydrocharus capybara). Proc. Zool. Soc. 1876, pp. 19-22. 8. Grorce, M.—Monographie anatomique des Mammiféres du Genre Daman. Ann. Sci. Nat., ~ Zoologie, 47 ann. ser. vi. vol. 1. p. 252. g. Home, Sir E.—Anatomy of the Dugong. Phil. Trans. 1820, p. 315. 10. Hunrer, W.—Hssays and Observations, vol. 11. 11, Kxaarscn, H.—Zur Morphologie der Mesenterialbildungen am Darmkanal der Wirbelthiere. Morph. Jahrb. 1892, pp. 385-450, 609-716. 12. Lecue, W.—Ueber die Saiugethiergattunz Galeopiihecus. Kong. Svenska Vet.-Ak. Hand). xxi. 1886. Dale ON THE INTESTINAL TRACT OF MAMMALS. . Maruzs, P.—Zur Morphologie der Mesenterialbildungen bei Amphibien. Morph. Jahrb. 1895, pp. 256-288. . Mecxet, J. F—Traité général d’Anatomie comparée, vol. viii. (French translation by Sanson and Schuster.) Paris, 1838. . Mrrcuett, P. Coarmers.—On the Intestinal Tract of Birds; with remarks on the Valuation and Nomenclature of Zoological Characters. Trans. Linn. Soc., 2nd ser. Zool. vol. viii. pp. 173-275. 1901. . Orret, A.—Lehrbuch der vergleichenden mikroskopischen . Anatomie der Wirbelthiere. Partai, sor . Owen, Sir R.—On the Anatomy of the Dugong. P. Z. S. 1888, p. 28. . Owen, Sir R.—On the Aye-Aye (Chiromys, Cuvier). Trans. Zool. Soc. v. 1862. . Owen, Sir R.—On the Anatomy of Vertebrates. 1868. . Patras, P. S.—Novee Species Quadrupedum e Glirium Ordine. 2nd edition, 1784. . Prrers, H. W.—Ueber die Siiugethiergattung Chiromys (Aye-Aye). Abh. kén. Ak. Wiss. Berlin, 1865. . Rapp, W. von.—Die Cetaceen. 1837. . Rapp, W. von.—Anatomische Untersuchungen tiber die Edentaten, Tubingen, 1852. . Scuirr, M.—Lecons sur la Physiologie de la Digestion. 1867. . Steuer, G. W.—Beschreibung von sonderbaren Meerthieren. 1753. . Torpr, C_—Bau und Wachstumsveranderungen der Gekrése des menschlichen Darmkanales. Wiener Denkschr. 1879, Abth. 2, pp. 1-56. Weser, Max.—Die Siingetiere. 1904. TRANSACTIONS OF THE ZOOLOGICAL SOCIELY OF LONDON (continued). To Fellows, To the Public. ens asl Oi Shi Oe VOLUME XIV. (1896-1898, containing 47 Plates) . Price5 5 O Ce OuaO Parrl. (1896, containing 2 Plates) . . ... eNO Ono 0-76 F.0 2. (1896, containme 6 Plates)’. 200 |... OF, 9-60. O-I2E0 » 98. (1897, containing 9 Plates) . . . . . ea On Laie 9) Tees » 4. (1897, containing 10 Plates) . . . . . are LOO z 2 0 0 » 5. (1898, containing 2 Plates) OF. Ges 0) (0) » 6. (1898, contaming 7 Plates) . . . . . a On Leo 14 0 Bm cs (tees, contamme 6 Plates)? ss. 8s. ere Opel wey 015 0 » 8. (1898, containing 5 Platesand Titleand Index),, 0 9 O 012 0 VOLUME XV. (1898-1901, containing 52 Plates) . Price5 15 6 714 0 Parr 1. (1898, containing 8 Plates) . . . . . Se ONT Gra 9 ae 0; » 2. (1899, containing 5 Plates). . .. . we. ee ues O) Oni2=-0 » 9. (1899, containing 4 Plates) . . . . . Be Oe a0) 012 0 » 4. (1899, containing 4 Plates) . . . . . OREO 012 0 » 5. (1900, containing 24 Plates) . . . . . ah ele G 83810 0 », 6. (1901, containing 4 Plates) Uses aE e's Sore Oa lae 015 0 » 7. (1901, containing 3 Platesand Title and Index),, 0 9 0 012 0 VOLUME XVI. (1901-1903, containing 38 Plates) . PriceS5 8 O 7 4 Parr 1. (1901, containing 6 Plates) . . .. . Bo OLS 10 1 4 0 » 2. (1901, containing 5 Plates) . pO Lee 0. 152Q » 3, (1901, containing 9 Plates) . . . . . fied Ud Raa?) Deltas. » 4. (1902, containing 5 Plates) ay) Oe OPO 012 0 » 9. (1902, containing 4 Plates) pe Ov on6 SOP Siee 0. » 6. (1902, containing 3 Plates) . . . . . nO et 9 A Ge TSK) » 7. (1902, containing 3 Plates) hon ftes ruciapemeero ng OLA OS Be *-OF1br.0 » 8. (1908, containing 3 Plates andTitle and Index),, 0 13 6 018 0 VOLUME XVII. Parr 1. (1903, containing 5 Plates) . . . . . Pricel 2 6 BiOn 0 » 2. (1903, containing 3 Plates). . . . . Iie On 1d 6 018 O » 9. (1904, containing 18 Plates). . . . . penal Geral) 110 0 » 4, (1905, containing 8 Plates). . . . . Farol ope Ole Aine ee ceuake LOI O 5. (1905, containing 50 Text-figures) . . . pote Ol bonis O L070 CONTENTS. V. On the Intestinal Tract of Mammals. By P. CHALMERS Mircuent, M.A., D.Se., Secretary to the Society . . . ee aa Mey acc aye ore. 2ST) THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Tun scientific publications of the Zoological Society of London are of two kinds—“ Proceedings,” published in an octavo form, and “ Transactions,”? in quarto. According to the present arrangements, the “ Proceedings” contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended to be published in the “ Proceedings” by the Committee of Publication. A large number of coloured plates and engravings are issued in the “ Proceedings,” to illustrate the new or otherwise remarkable species of animals described in them. Among such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “Proceedings” for each year are issued in four parts, in the months of June, August, October, and April, the part published in April completing the volume for the last half of the preceding year. Beginning from January 1901 they form two half-yearly volumes. The © Transactions’ contain such of the more important communications made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are issued at irregular intervals. Fellows and Corresponding Members, upon payment of a Subscription of £1 1s. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1881, if they exceed the value of five pounds. Fellows also have the privilege of subscribing to the Annual Volume of the ‘ Zoological Record’ for a sum of £1 10s. (which includes delivery), but this privilege only holds good if the subscription is paid before the First of December im each year. Such of these publications as are in stock may be obtained at the Society’s Office (3 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, H.C.), or through any bookseller. P, CHALMERS MITCHELL, December, 1905. ? Secretary. TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Vou. XVII.—Parr 6. LONDON: PRINTED FOR THE SOCIETY, SOLD AT THEIR HOUSE IN HANOVER-SQUARE ; AND BY MESSRS, LONGMANS, GREEN, AND CO,, PATERNOSTRR-ROW, October 1906, Price 30s. an : ansoriah Institution [Red Lion Court, Fleot Atreuf: NOV 2 1906 Taylor and Francis, Printers. ] —— \\ ‘\ \ bb Mla wu y TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. To Fellows. To the Public. £s. d. eo Gh VOLUME I. (1833-1885, containmg 59 Plates). . Price 313 6 . . . 418 OF VOLUME TI. (1835-1841, containing 71 Plates). . 5 4 00... 5 6 6% VOLUME III. (1842-1849, contains 63 Plates) as ee tOnGh he: eae ene len VOLUME IV. (1851-1862, Loans 77 WlateS\Uys ere ein CB Onw 2 OM aa Mauna a Guo inl Om VOLUME V. (1862-1866, containing 67 Plates). . ,, Die Atenoto nh tan cnneresO i 0 VOLUME VI. (1866-1869, containing 92 Plates). . ,, 11 5 0 15 0 O VOLUME VII. (1869-1872, containing 73 Plates). . 4 AU O Sinha ceca eral ea then O) VOLUME VIII. (1872-1874, containing 82 Plates). . ,, SLRISht 2h ig com erdle) AME (0) VOLUME IX. (1875-1877, containing 99 Plates) . Sele al OR eee olGi 280) VOLUME X. (1877-1879, containing 95 Plates) . ee LOO 18 7 O GENERAL INDEX, Vols. I. to X. (1833-1879) S 0 7 6 010 0 VOLUME XI. (1880-1885, containing 97 Plates). . ,, 9 12 G ee oN ANG) VOLUME XII. (1886-1890, containing 65 Plates) . 5) 5 8 0 7 4 0 VOLUME XIII. (1891-1895, containing 62 Plates) : 3 rere Scars que) sbi @) *No copies of these volumes remain in stock. Continued on page 3 of Wrapper. VI. Fourth Contribution to the Ichthyology of Lake Tanganyika1.—Report on the Collection of Fishes made by Dr. W. A. Cunnington during the Third Tanganyika Expedition, 1904-1905. By G. A. Bounnnerr, F.R.S., V.P.Z.S. Received February 6, read March 6, 1906. [Puares XXX.-XLI1.] THE last expedition organised out of the funds remaining in the hands of the Tanganyika Exploration Committee was entrusted to Dr. W. A. Cunnington, who left London in March 1904, and returned home in June 1905. The total length of time spent on and around Tanganyika was about 8 months, and collections were made at a great number of localities, marked on the sketch-map on the next page, some of which appear here under a spelling somewhat different from that adopted by Mr. Moore and followed in the previous reports dealing with his results. Speaking of the fishes alone, I may say that Dr. Cunnington has been highly uccessful in his efforts, and has fully justified the expectations entertained by the Committee who entrusted him with the mission of forming as complete a collection as possible of the animals and plants inhabiting the great lake. The series brought home by him is larger than any of those from that lake with which it has hitherto been my privilege to deal. The preservation of the specimens is excellent and their value is enhanced by a careful labelling of every one of them, in most cases accompanied by either sketches or notes concerning the life-coloration, on which I have largely drawn in preparing the present account. I have in every case reproduced the native names (mostly of the Ujiji fishermen) consigned in these notes, although in many cases such names appear to be applied regardless of the real specific characters. Dr. Cunnington has also been so fortunate as to considerably extend the list of Cichlid fishes in which the parents protect their offspring by giving them shelter in the mouth and pharynx. This mode of nursing is illustrated, in the present collection, by examples of seven species, viz.:—Paratilapia pfefferi, Pelmatochromis polylepis, Pelmatochromis pleurospilus, Enantiopus longianalis, Simochromis diagramma, Tilapia dardennii, and Eretmodus cyanostictus. Mr. Moore had previously made us 1 For tlie third contribution, ¢f. Tr. Z. 8. xvi. 1901, p. 137. VOL. XVIJ.—PART VI. No. 1.—October, 1906. 4c 538 MR. G. A. BOULENGER ON A COLLECTION fren Mshale a \ Toa OKavara Is Tembwi aX Kibwesi Mpale' \ Sumbua Mrumbi Karema g-soKombe Utinta , GoKirando | Msamba i V. Molire, { Cae Sumbu eee Boy Se chanikal a Wp) Kasanga (Besmerkaurg) acquainted with similar habits in #Hnantiopus longianalis and Tropheus moorii. I find in Dr. Cunnington’s notes an interesting entry referring to this matter: ‘“‘The natives say it is always the female, in the cases where one of the parents takes the eggs in the mouth.” This is not only confirmed by the specimens in the col- lection, whenever I have been able to ascertain the sex, but agrees entirely with a statement made by me four years ago! to the effect that so far as I could speak from personal observation, having tested the sex of a great, number of specimens of Tilapia nilotica and strigigena from the Nile, it is invariably the female who thus carries the eggs. This was in contradiction to statements made by Lortet and by Giinther, who ascribed this habit to the male in the species of the same genus with which they had dealt. I have, how- ever, since had occasion to examine the specimen of Chromis (Tilapia) philander sent by Mr. Nendick Abraham to Dr. Giinther 2, and was able to satisfy myself from autopsy that it is a female and not a male; whilst Dr. Pellegrin has ascertained the female sex of a specimen with eggs in the mouth pre- sented to the Paris Museum by Dr. Lortet as his Chronis paterfamilias (= Tilapia simonis). Further observations by Dr. Pellegrin? on Tilapia galilwea and Pelmatochromis lateralis, by Mr. Schoeller on Paratilapia multicolor, and by myself on Tilapia natalensis have led to the same result. It therefore remains unproved whether in any of the African or Syrian Cichlids the buccal “incubation,” as it has been called by Dr. Pellegrin, devolves on the male; the instances previously adduced 1 ¢The Field,’ c. 1902, p. 33. ~2 Ann. & Mag. N. H. 1901, viii. p. 321. -3 Mém. Soc. Zool. France, xvi. 1904. OF FISHES FROM LAKE TANGANYIKA. 539 being either controyerted or unsupported by the only reliable evidence—an exami- nation of the genital glands !. In my last report I expressed the opinion that, notwithstanding the exertions of Mr. Moore and the officers in the service of the Congo Free State, not more than one- half of the species of Fishes existing in Lake Tanganyika was then known to Science. The valuable collection formed by Dr. Cunnington goes a long way towards the fulfilment of this prediction. Mr. Moore’s first collection, made in 1895-96, contained about 90 Fishes, referred to 33 species, 25 of which were described as new. His second collection, made in 1899-1900, contained 180 examples of 48 species, 22 of which were new. Dr. Cunnington succeeded in bringing home 300 specimens, referred to 84 species, 27 of which are here described for the first time. No doubt many more species remain to be discovered. The following Table, giving a complete list of the Tanganyika fishes, will show the order in which they have been discovered. Column 1 indicates the species discovered by Mr. Coode Hore and described by Dr. Giinther in 1895; column 2 the species in Mr. Moore’s first collection; column 3 those in the various small collections formed by the officers of the Congo Free State; column 4 those in Mr. Moore’s second collection ; and column 5 those in Dr. Cunnington’s collection. The names of endemic genera and species are printed in italics. ‘ [Since the reading of this Report I have been able to considerably extend my observations on the nursing- habits of these fishes, and in eyery case the specimen carrying eggs or young in the mouth has proved to be of the female sex. First, Mr. E. Degen has brought back from Lake Victoria a large and yaluable series of Cichlid fishes, many of which represented new species which haye been described in the ‘ Annals and Magazine of Natural History’ for May 1906. The maternal solicitude was observed in three species, viz. :—Haplochromis ishmaelt, Paratilapia serranus, and Tilapia martini. Secondly, Lord Walsingham, when at Biskra in February and March last, was so kind as to collect for me a large number of Haplochromis desfontainesii in a tepid spring at Ain Oumash, and most of the females, which do not appear to reach so large a size as the males, had the mouth and pharynx full of eggs or embryos. The ege of this fish is not perfectly round, the upper pole being somewhat pointed ; the vitelline sac, when hanging from the embryo, is pyriform in shape. And, thirdly, Dr. Giinther has been so good as to hand over to me, for the British Museum, a further series of Tilapia philander, together with some 7’. natalensis, which he had received from the Rey. N. Abraham, several of which haye eggs in the mouth as well as in the genital glands.—23.6.06.] 4c 2 Co W~ID . Labeo cylindricus Peters, 1852 . Varicorhinus tanganice Blgr. 1900 . Barbus platyrhinus Blgr. 1900 . Barilius moortt Blgr. 1900 . Neobola minuta Blgr. 1906 . Clarias robecchii Vincig. 1893.. . Dinotopierus cunningtoni Blgr. 1906 . Chrysichthys cranchii Leach, 1810 . Auchenoglanis occidentalis C. & V. 1840 . Amphilius platychir Gthr. 1864 . Synodontis granulosus Blgr. 1900 . Malopterurus electricus Gm. 1789 . Haplochilus pumilus Blgr. 1906 MR. G. A. BOULENGER ON A COLLECTION PoLYPTERID A. . Polypterus congicus Blgr. 1898 .......... LrprposiRENID&. Mormyrip. . Marcusenius tanganicanus Blgr. 1906...-.. . Mormyrus longirostris Peters, 1852 ...... CLUPEID&, . Pellonula miodon Blgr. 1906 ............ CHARACINID®. 3 rhodopleura Bler. 1906) tee 3. . Citharinus gibbosus Blgr. 1899 .......... CypRINiD&. Fy Grae dallenes USO) SoS eS o ules 5 eae » tropidolepis Blgr. 1900 Pr tanoanicce Blan: OO Meer elo: SILURIDZ. Wocephalus Blan 1698.0... a myriodon Blgr.1900........ es brachynema Blgr. 1900 ...... sianenna Blgr. 1906 Phyllonemus typus Blgr. 1906 multipunctatus Blgr. 1898 us melanostictus Blgr. 1906 CYPRINODONTID&. i tancanicanus Blgr. 1898 . Protopterus ethiopicus Heck. 1851 ...... . Hydrocyon lineatus Blkr. 1863 .......... . Alestes macrophthalmus Gthr. 1867 a eR RES * * * * Ke KK * * * OF FISHES FROM LAKE TANGANYIKA. 541 feet 2 3. 4 5 SERRANID 2. 34. Lates microlepis Blgr. 1898 ............ yee ese lees Seal) | nen se ON weangustimons ler. LOO Greaeee eee eo he edie a S58 ae een CicHLID&. | | 36. Lamprologus brevianalis Blgr.1906 ......) .. | .. eT hice ete 37. x tetracunthus Blgr. 1899 ...... ty otiel| : cont 38. elongatus Bley. 1898 ........ | a | x ee | x 39. A cunningtoni Blgr. 1906 ...... ie Ae a a | * | 40. 3 tretocephalus Bler. USRE). oe sce * ~! * ce | x 4]. 5 mondabu Blgr. 1906 MP Ae cia xe x 42. i modestus Bler. SOS ae ea | x 2, he 43. Ly multifasciatus Blorse OO Geen bt * 44, yy lemairic Bler. 1899 ........ 2 a x 45. es hecqui Blgr. 1899 ie Beye lies Mss * 46. 5 maori: Wailers, WEBS) oo ucoods 6 oe, 47. is callipterus Blgr. 1906 ...... a nt a ne | ne 48. 5 reticulatus Bler. 1906 ...... aa on | es ae | Ps 49, oH calliurus Bler. 1906 ........ er Sontisny! aS * 50. aS brevis Blgr. 1899 ..... tea Ws ae * * ol. 5 compressiceps Blgr, 1898 | alle sais * 52. py fasciatus Blgr. 1898 ........ | ye 53. furcifer Bler. NSO SRA AR Ae. Sees | Be | ee 54. Julidochromis ornatus Bler TSN). ae oe | 55, Telmatochromis vittatus Blom USI Saas aie * | 56. temporalis “Bler. EGIS. o ooo | * * x * 57. Paratilapia frontosa Blgr, SUG Myo | o-6 se x 58. A preffert Bler. TSO Sie yest sae | +e “a * 59. 35 calliura Bler. DUS Oc iste tee Geta ane x 60. ny macrops Blgr. 1898 .......... * * 61. i ventralis Blgr. 1898 .......... * a * 62. Fe dewindti Blgr. 1899 .......... te 5 * * 63. 5 funcufer, Biers 1898) se sae eae 38 * oh ry 64, = stenosoma Bler. 1901.......... Bre * 65. 3 leptosoma Blgr. 1898.......... * | es 66. nigripinnis Bler. HNO Satis oe oe 50 Pe 67. Bathybates. ferox Bler. 1898 Sapecticcarsbara rests sean) all hate © ke 68. £6 fasciatus 1k, ISOM ooo ac doo Vrs + 69. Be nuinor Bler. “1906 ne EEG dm EE x * 70. Haplotawodon microlepis Bler. 1906 ...... a 50 ag: as o, 71. Pelmatochromis polylepis Blgr. 1901 ...... ahs ae | * * Be 72. PA melanostigma Blgr. 1906 ry | 5 1 73. of pleurospilus Bler. USO Giese & 74. ies rhodostigina Bler. 1906)... Se | ve dc Salat: * 75. auritus Blgr. HISTO asset ae i, oo pee tS * 76. Betodus ‘descampsi Blgr. LOGE REM er oi an eee * 77. Enantiopus melanogenys Blgr. 1898 ...... we a * * 78. 5 longianalis Bler. Nees omipt. 5.0 oe at | ales ae Se * K 79. Xenotilapia sima Blgr. 1899 ............ bey Ui /A ce aes sk 80, a ornatyinnis Blgr. 1901 ......, * on a6 Pr | 81. Grammatotria lemairii Blgr. 1899 ........ ie ers * re * 82. Trematocara marginatum Blgr. 1899 ...... * * 83. rr unimaculatum Blgr. 1901 « * 84. re nigrifrons Bler. 1906 ........ a | * | 542 MR. G. A. BOULENGER ON A COLLECTION | 1 2 3. 4 5 CicHLIp@ (continued). 85. Gephyrochromis moorti Blgr. 1901 ........ 6 Festi Pea ane % 86. Tropheus moorti Blgr. 1898.............. aa * ob * 87. a annectens Bler. 1900 .......... te oD Fa 88. Simochromis diagramma Gthr. 1893 ...... * Ba ae * | * | Sonetilapiamnil oficaslnel iO mearia err e rrr Hewes: avo ny * 90 Re eeburtoninG.threls 93 saan er ree ae ra | a0 * *% Dae lations: Blot sel. 0. Grwarinieien rir: ae * Pe gy TAO ENBIDR IUEIS), Sn go bb oo oe saaaee * cs Ea oc Hep OS a eadardennisplonlo 9) ne ney eet ah ats * * ea | ee ey 1 Uedoamdellere, USNS) a5. b-o.00-a og ou 6 ae fe % * | * 955 3) plewrolenia BlerallQON evant een: x | 96. ,, étrematocephala Blgr. 1901 ........ iG bo Be a OT = microlepis Blors WSO9 Mean sone ip Bi * So ule tae | lek gy 4) -eawas Ane UBOM yo. un bos naceeoc * | DOs ees, onandoculissBlon wl SO OM victim sie ye uN * ae | * 100. Petrochromis polyodon Blgr. 1898 ........ His cio * a o8 * 101. i tanganice Gthr. 1893........ | * | 6.0 | ros * | x 102. Cunningtonia longiventralis Blgr. 1906 .... ne | io, anvil ae | Be % 103. Asprotilapia leptura Blgr. 1901 .......... Ss, Nene . P| 104. Hretmodus cyanostictus Blgr. 1898 ........ aE x * ac | * 105. Spathodus erythrodon Blgr. 1900 ........ ae Feige | * | 106. Perissodus microlepis Blgr. 1898.......... Sy Sel eae lah 107. Xenochromis hecqui Blgr. 1899 .......... mG ' ay % % * | 108. Plecodus paradowus Blgr. 1898 .......... Fits pa % ye * MAstacEMBELIDZ. | 109. Mastacembelus frenatus Blgr. 1901........ oo eye tei % 110. - moor), Blors i898. heen | ae * a ths ( oe WiLL, = cunningtont Blgr. 1906 ....| av a0 ae x 112. ‘ ellipsifer Blgr. 1899 ...... so || we % eS} ne tanganice Gthr. 1893...... 4 j 114. i. teniatus Bley. 1901 ...... Pabievotss ai vung iae Se rea eee * 115. 5 ophidrum Gthr, 1893 ...... Pie al aces ne se * A glance over this list shows that one of the six species discovered by Mr. Coode Hore (Mastacembelus tanganice) and six of those in Mr. Moore’s first collection are unrepresented in the collections subsequently made. Apart from many important additions to our knowledge, the collection now worked out calls for the following modifications to the conclusions previously arrived at as to the general character of this remarkable fish-fauna :— Two new generic types of Siluride are here described, all previously known autochthonous genera falling under the family Cichlide. One species of Cichlids, Tilapia nilotica, of wide distribution (Jordan, Nile, East and West Africa), previously known from Lake Kivu, is, with 7. burtonz, also found in Lake Kivu, the only non-endemic species of that family the occurrence of which has been ascertained. OF FISHES FROM LAKE TANGANYIKA. 043 One species widely distributed in East Africa, Zabeo cylindricus, is added to the list of Cyprinide, all previously known species being endemic. Neobola, ranging over North-east Africa and Lake Victoria, has yielded a third species representing this genus of Cyprinide in Lake Tanganyika. An Alestes previously regarded, from insufficient material, as identical with a Nile and West-African species, is now found to be distinct. The Characinide can therefore no longer be said to be all non-endemice. A small Cyprinodontid (Haplochilus pumilus) was discovered almost simultaneously by Dr. Cunnington in Lake Tanganyika and by Mr. E. Degen in Lake Victoria. And, finally, representatives of two families widely distributed in tropical Africa, the Mormyride and the Clupeide, have been added. The discovery of Mormyride is of particular interest, as their supposed absence was regarded as a striking negative feature of the great lake. These data, however, only further emphasize the truly African freshwater and modern character of the fauna of Lake Tanganyika, a subject with which I have fully dealt, from the ichthyological point of view, in an Address read at the British Association Meeting in South Africa last summer, and I have nothing to alter in the general conclusions I had arrived at before studying Dr. Cunnington’s collection. Dr. Cunnington supplies the following note on the methods of fishing observed by him :—‘* Almost the whole shore of this big lake is inhabited by a fishing population, and large quantities of fish of the most diverse kinds are obtained from the well-stocked water. The lake-shore inhabitants belong, however, to a number of different tribes, so that the methods of fishing adopted, and the kinds of fish most highly prized, vary somewhat in the different regions. While most, if not all, of the Tanganyika fish are other hand, the large Siluroids, with their rich oily flesh, are usually much valued by the natives for eating. Again, towards the north end of the lake, the small whitebait- like Clupeid, Pellonula miodon, is a particularly favourite dish. “‘ Perhaps the commonest method of fishing is by means of a net laid out from the shore in a semicircle and drawn in from both ends. In some cases this may be merely a flat wall of net-work, buoyed above and weighted below, though even then not more than three or four feet in depth, while in other cases it may be still shallower, but furnished at intervals with a series of pockets. On parts of the western shore a small and singularly inefficient net of this type is employed, consisting of but two such pockets of small dimensions in the middle, and on each side a long-shaped ring converted into a grating by flat strips of bark. “Fishing is also carried on in many parts of the lake by the aid of a long line bearing a series of baited hooks at intervals. ‘This is essentially in the nature of a surface-line, being buoyed at each end, and the line is baited overnight and hauled in in the morning. Fishing with rod and line is but little practised, and for this the natives 544 | MR. G. A. BOULENGER ON A COLLECTION appear never to use a float, although that is the general custom on Victoria Nyanza. When using hook and line alone a bait of stiff paste is used, and it is especially Characinids which are caught in this manner. “Torch-light fishing goes on to a considerable extent in various parts of the lake, the larger fish being speared and the smaller caught in hand-nets as they are attracted to the light. At the north end, in particular, an important fishery is carried on for ‘dagaa,’ a young Clupeid which swims in shoals. Owing to the scarcity of firewood, large bundles of dried reeds are tied together to form a gigantic torch, which projects over the bows of the canoe and is gradually pushed forwards as it burns away. The fish are netted and brought in large quantities to the native markets, while some are sun-dried and disposed of at considerable distances from the lake. “Finally, there are different kinds of traps used around Tanganyika. Small conical wicker traps are placed in the rivers and streams, while across the latter, not far from the lake, may be built an elaborate system of traps and barriers to catch the fish proceeding down the stream. Out in the lake itself, particularly towards the southern end, large trilobed basket-traps are used for the purpose of catching the big Silurid ‘singa.’ The trap is suspended from a float, a few feet under the surface, and is baited with some vegetable substance. In certain districts a great number of such floats may be seen dotting the surface of a bay or inlet.” POLYPTERID 4S. 1. PoLYPrERus coneicus Bler. A single male specimen, 620 millim. long, from Kituta. Native name Munkunga. Dr. Cunnington measured a larger specimen, 710 millim. long. The fish occurs also in the rivers. 14 spines in the dorsal; 56 scales in a longitudinal series, 46 round the body, 13 between the occiput and the origin of the dorsal. The azygous shield between the nasal bones is present. Dr, Cunnington describes the colour as greenish grey above, shading off to lighter below, belly with pinkish tinge; 8 or 9 black blotches along side of body; black spots on side of head; paired fins with dark spots. LEPIDOSIRENID A. 2. PROTOPTERUS £THIoPICUS Heck. One specimen, measuring 485 millim., taken in a native-made wicker trap, baited with vegetable matter, in the Kakonde River, near its mouth in Mtondwe Bay, was brought home by Dr. Cunnington, who also observed another specimen, measuring 970 millim., in a river at Moliro. The fish has not been found in the lake itself, and the natives, who call it Sompo, catch it in small muddy streams and say (at Mpala) OF FISHES FROM LAKE TANGANYIKA. 945 that they sometimes come across it when digging graves. It is much feared for its bite. Although captured by vegetable bait (cassava, a sort of tuber), a specimen opened by Dr. Cunnington contained partially digested fish in the stomach. This is the first specimen from the borders of Lake Tanganyika which falls under the eyes of an ichthyologist, and it confirms the provisional identification of the species (see Boulenger, Poiss. Bass. Congo, p. 38). The specimen figured by Moore in his ‘Tanganyika Problem, p. 153, is from Lake Albert Edward. The following notes are taken from Dr. Cunnington’s specimen :— Depth of body 84 times in total length; length of head 4¢ times in distance from end of snout to vent; diameter of eye 15 times in length of head; dorsal fin originating a little nearer to vent than to occiput ; 60 scales in a longitudinal series from the gill- opening to above the vent, 42 round the body; vent on the left side. Dr. Cunnington describes the coloration as ‘mottled grey and black, darker above; sinuous lines of brown on the dorsal region and on the tail; tail with large black blotches on a grey ground; belly mottled grey and flesh-colour, with a pinkish tinge all over.” MorMYRID&. 3. MARCUSENIUS TANGANICANUS, sp. n. (Plate XXX. fig. 1.) Depth of body 34 to 5} times in total length, length of head 43 to 43 times. Head as long as deep, not quite twice as long as broad; snout as long as or a little shorter than the eye, rounded, projecting beyond the mouth, which is small and a little anterior to the vertical of the nostrils; teeth bicuspid, 5 in the upper jaw, 6 in the lower; nostrils on a line with lower border of eye, nearer the latter than the end of the snout; eye 44 to 43 times in length of head, about once and 3 in interocular width. Dorsal 51-34, considerably shorter than its distance from the head, anterior rays longest. Anal 23-25, originating below 8th to 10th rays of dorsal, equally distant from base of ventral and from root of caudal; anterior rays longest, especially in the males, in which the anterior half of the fin forms a rounded lobe. Pectoral pointed, as long as or a little shorter than the head, extending beyond base of ventral. Caudal fin scaly at the base, with pointed lobes. Caudal peduncle twice and + to twice and 2 as long as deep, a litfle shorter than the head. 65 to 67 scales in the lateral line, Ss in a transverse series on the body, ae in a transverse series between dorsal and anal, 12 round caudal peduncle. Uniform greyish above, silvery white beneath ; dorsal and anal fins whitish at the base, grey distally. Total length 160 millim. Two specimens from Sumbu, and one taken in a river at Msamba. Native name: Zoveruz. This species is very closely allied to, and exactly intermediate between, JZ. disco- rhynchus Peters, from the Zambesi and Lake Nyasa, and MW. petherici Blgr., from ,the VOL. XVII.—PART VI. No. 2.—October, 1906. 4p 546 MR. G. A. BOULENGER ON A COLLECTION Nile. It differs from the first in the mouth being a little more anterior, in the shorter dorsal fin with 31-34 rays instead of 34-56, and in the depth of the caudal peduncle being contained less than twice and a half in its length; from the second in the caudal peduncle being more than twice as long as deep and surrounded by 12 scales instead of 14. 4, MoRMYRUS LONGIROSTRIS Peters. A single specimen, 530 millim. long, from Sumbu. Native name: Domodomo. Said to grow to nearly three feet. Dorsal 75, 54 times as long as anal; anal 18; 26 scales round caudal peduncle. Dr. Cunnington describes the coloration as ‘‘ Grey above, lighter to white below, darker on top of head, with darker blotches along middle of side; tail darker, with some patches of red; two brown spots below the eye; ventrals reddish, grey distally.” M. longirostris, which is known from the Zambesi and the Congo, is so closely related to M. caschive, from the Nile, that it should perhaps be regarded as merely a variety of the latter. CLUPEIDA. 5. PELLONULA MIODON, sp. n. (Plate XXX. fig. 2.) Depth of body 43 to 53 times in the total length, length of head 3% to 4 times. Snout obtusely pointed, as long as the eye, the diameter of which is contained 33 to 35 times in length of head and equals once and 4 to once and 2 interorbital width; chin projecting beyond the snout; maxillary extendgjng to below anterior border or anterior fourth of eye; teeth small, canines confined to anterior extremity of mandible. Gill-rakers long and slender, closely set, 35 to 40 on lower part of anterior arch. Dorsal 15-16, equally distant from centre or posterior border of eye and from root of caudal; longest rays about $ length of head. Anal 16-17, a long way behind the dorsal, nearer to root of caudal than to base of ventrals. Pectoral 2 length of head, widely separated from the ventral. Ventral inserted below anterior third or middle of dorsal. Caudal deeply forked, with pointed lobes. Caudal peduncle once and } to once and 2 as long as deep. 45 to 48 scales in a longitudinal series, 14 in a transverse series; 19 or 20 keeled scutes between isthmus and ventrals, 10 or 11 between ventrals and anal. Olive above; silvery below, with a blue indistinct lateral band ; dorsal and caudal greyish, other fins white. Total length 140 millim. Several specimens from Ndanvie, Tembwi, Vua, Mbete, Niamkolo Id., and Kasakalawa. Native names: Lumbu, Sembe, and Dagaa. This is, no doubt, the little fish “ very like Whitebait ” mentioned by Livingtone (‘ Last Journals,’ vol. ii. p. 17) as the Dagala or Nsipé of Lake Tanganyika, which, in my previous ignorance of any sort of Clupeid in OF FISHES FROM LAKE TANGANYIKA. 547 that lake, I had suggested might be the Cyprinodont Haplochilus tanganicanus. Whether the statement of the natives that it carries its eggs in its moutk is true or is based on confusion with some other fish cannot at present be decided ; but the material at hand offers no evidence in its favour. Stanley (‘ How I found Livingtone,’ p. 530) also speaks of it as the Dogara, a species of whitebait, which is caught in great nets by the thousand and spread out in the sun to dry, and exported, salted, even as far as Unyanyembe. The Clupeids are represented in the fresh waters of tropical Africa by three closely related genera: Pellonula Gthr., Odaxothrissa Blgr., and Microthrissa Blgr. Of the first, to which the Tanganyika fish belongs, two species only were known, inhabiting West Africa and the Congo. The principal characters distinguishing the three species with which we are now acquainted are contrasted in the following “ key ” :— I. Dorsal 15-17, originating at equal distance from end of snout and from base of caudal fin, or a little nearer the latter; lower jaw projecting beyond the upper. Depth of body not more than 4 times in total length ; keeled ventral scutes 13-15 + 9-10; anal originating a short distance behind vertical of dorsal. P. vorax Gthr. Depth of body 42 to 53 times in totallength; keeled ventral scutes 19-20 + 10-11; anal originating a long way behind dorsal . . . . . . . P. miodon Bigr. II. Dorsal 13-14, originating much nearer end of snout than base of caudal fin; upper jaw projecting beyond the lower; keeled ventral scutes 12-13+12; anal originating a long way behind dorsal . . . P. acutirostris Blgr. CHARACINIDA. 6. Hyprocyon LINEATUS Blkr. Kituta. Native name: Sokonto. 7. ALESTES MACROPHTHALMUS Gthr. Several specimens from Niamkolo, Mtondwe Bay, and Kituta. Native names: Manji, Mianga, and Mowanzi. Grows to a length of 560 millim. 8. ALESTES RHODOPLEURA, sp.n. (Plate XXX. fig. 3.) Depth of body 4 times in total length, length of head 4 to 43 times. Head longer than deep, nearly twice as long as broad, its width much less than its length to the occiput; snout rounded, feebly projecting beyond the lower jaw, once and 3 diameter of eye, which is 4 times in length of head; adipose eyelid very short ; interorbital width nearly half length of head ; maxillary not extending to below anterior border of eye; 18 teeth in the upper jaw, 8 in the outer row of the lower jaw ; length of lower border of second suborbital equal to or a little more than diameter of eye. 4p2 548 MR. G. A. BOULENGER ON A COLLECTION Gill-rakers moderately long, 18 or 19 on lower part of anterior arch. Dorsal II 8, behind vertical of base of ventrals, originating much nearer root of caudal than end of snout, longest ray about # length of head. Adipose fin small, about twice as far from rayed dorsal as from root of caudal. Anal II] 12-13. Pectoral as long as head, not reaching ventral. Caudal forked. Caudal peduncle once and 3 to twice as long as deep. Scales 28-29 ah 1 between lateral line and ventral. Olive-green above, silvery below, with an ill-defined pink band along each side, beginning on the gill- cover ; a large blackish (in life dark greenish) patch on caudal peduncle and root of caudal. Total length 500 millim. Several specimens from Niamkolo, Mtondwe Bay, and Lofu. Native names: Kolo- kolo and Kologo. This is the fish previously recorded by me, from a single badly preserved specimen, as A. macrolepidotus C. & V. But that species, of which I have examined over 100 examples from the Nile and from West Africa, has never more than 26 scales in the lateralline. A. batesii Blgr., from South Cameroon, agrees with the Tanganyikan species in the number of scales in the lateral line as well as in the number of anal rays, but there are two series of scales between the lateral line and the root of the ventral fin, the head is shorter in the adult, and the caudal peduncle is shorter, as in A. macro- lepidotus. A. grandisquamis Blgr., from the Congo, and A. brevis Blgr., from Guinea, are other close allies of the Tanganyikan form, but both have a shorter head and a shorter caudal peduncle, and the first has 24 to 27 scales in the lateral line and 13 or 14 anal rays, the second 21 to 23 scales in the lateral line and 16 to 18 anal rays. 9, CITHARINUS GIBBosUS Bigr. Niamkolo. Native name: Jmbaza. Dr. Cunnington measured a large specimen, which had been speared by the natives ; its total length was 540 millim. CYPRINIDA. 10. Lasgo cyLinpricus Peters. Two specimens, both with 10 branched rays in the dorsal fin, from Utinta. Native name: Bimanda. Grows, according to Dr. Cunnington, to a length of 360 millim. ‘Dark olive-green to almost black above, lighter below, with the belly greyish ; scales with darker outlines; all fins dark greenish black, darker between the rays.” L. cylindricus, which is very nearly related to the Nilotic L. forskalii, has a wide distribution in Kast Africa, from Abyssinia to the Zambesi. OF FISHES FROM LAKE TANGANYIKA, 549 In my first Report I had recorded a Laveo from Lake Tanganyika, on the evidence of a sketch made by Mr. Moore; but as no specimen had been preserved, the species could not be determined. 11. VaRricorHINUS TANGANIC Blgr. Two specimens from Niamkolo. Native name: Jmbara. Dr. Cunnington measured a specimen 463 millim. long. ‘‘ Greenish brown above ; paired fins and anal with a pinkish tinge.” This is the fish described by me as Capoéta tanganice. I have since had occasion to show that Capoéta is a synonym of Varicorhinus of Riippell, which has priority. 12. BARBUS SERRIFER Blgr. Kituta, ina swamp. Native name: Palala. “ Olive-green above, shading to silver below; a red spot on gill-cover immediately behind eye; blue-black spot at root of caudal; dorsal and caudal fins with distal third orange; paired fins and anal yellowish.” 13. BaRBUS TROPIDOLEPIS Bley. Niamkolo, Mbete, and Uvira. Native names: Kaluka Matanda, Kitumbi Mbalaga, and Birigi. Dr. Cunnington notes that this fish grows to a length of 500 millim., and is often taken in rivers!. The colour in fresh specimens is olive above, shading off to silvery or to light brownish-grey and silvery; paired fins and anal yellowish; the largest specimen preserved (measuring 290 millim.) had the raised lines on the body of a lighter olive and a red blotch at the tip of the spine of the dorsal fin. A smaller specimen (195 millim.) has a short anterior barbel on one side only. 14. Baritivs Moort Blgr. Sumbu. Native name: Airiabakawba. The largest specimen measures 160 millim., and has 11 very distinct vertical dark bars on the side. 15. Neopota Minura, sp.n. (Plate XXXII. fig. 1.) Depth of body 6 times in the total length, length of head 5 times. Snout pointed, not projecting beyond the mouth, shorter than the eye, which is 35 times in length of head and equals interorbital width; mouth extending to below anterior border of eye; suborbital bones covering greater part of cheek. Dorsal II 7, originating further back than anal, at equal distance from head and from root of caudal, longest ray about * Of. M. Questiaux’s notes on the habits of the Mbirihi, in Boulenger, Poiss. Bass. Congo, p. 228. 550 MR. G. A. BOULENGER ON A COLLECTION 2 length of head. Anal II 18. Pectoral acutely pointed, as long as head, reaching root of ventral. Caudal deeply forked. Caudal peduncle twice as long as deep. Scales 37 Ae 1 between lateral line and root of ventral, 12 round caudal peduncle. Silvery, olive in the upper third. Total length 26 millim. This, according to Dr. Cunnington, is the full size of the fish. Several specimens were taken at Mbete, in shallow water, by means of a piece of calico held by two men. Native name: Salala. The genus Weoéola Vinciguerra is intermediate between Barilivs Ham.-Buch. and Chelethiops Blgr., differing from the former in the more posterior position of the dorsal fin, from the latter in the absence of a ventral keel and in the wider separation of the pectoral fins. Two species were known: JV. botiegi Vincig., from Lake Rudolf, Galla- land, and Somaliland, and WV. argentea Pellegr., from Lake Victoria. The three species may be distinguished by means of the following characters :— Anal 17-19, its origin corresponding to that of dorsal; lateral line 40-45 ; mouth extending to beyond vertical of anterior border of eve . . . UN. botiegi Vincig. Anal 17-18, its origin corresponding to that of dorsal; lateral line 48-52 ; mouth extending to below anterior border ofeye. . . . N. argentea Pellegr. Anal 20, originating in advance of dorsal ; lateral line 37 ; mouth extending to) belowsanteriorsborderoLeyeny a0) eee ne NSCinucomB lors SILURIDA. 16. CLaRiAs RoBECccHII Vincig. Niamkolo. Native name: Pongo or Kambali. DINOTOPTERUS, g. n. Intermediate between Clarias and Heterobranchus. Dorsal fin divided into a rayed and an adipose portion, but the latter not supported by produced neural spines. Temple unprotected by bone. 17. DINoTopTERUS CUNNINGTONI, sp. n. (Plate XXX1.) Depth of body 7} times in total length, length of head 32 times. Head extremely flattened, once and 3 as long as broad, its upper surface covered with very thin skin and rugose with striz radiating from the occipital region ; frontal fontanelle large, twice and 3 as long as broad, + length of head; occipital fontanelle very small; a rather long and narrow occipital process ; snout broad, truncate, not projecting beyond OF FISHES FROM LAKE TANGANYIKA. 551 the mouth ; eye small, 5 times in length of snout, 74 times in interorbital width ; nasal barbel 3 length of head; maxillary barbel a little longer than head, reaching posterior third of pectoral; outer mandibular barbels once and $ as long as inner and nearly as long as head. Jaws with bands of minute villiform teeth; similar teeth form a rather broader crescentic band on the vomer. Gill-rakers very long and slender, closely set, about 110 on anterior arch. Dorsal with 53 rays, the posterior of which are the longest and measure about 4+ length of head; distance between first dorsal ray and occipital process 3 length of head; adipose dorsal low and short, in contact with caudal. Anal with 45 rays, separated from the root of the caudal by a space equal to twice diameter of eye. Pectoral 4 length of head; spine rather weak, not serrated, a little more than half length of fin. Ventral equally distant from end of snout and from root of caudal. Caudal rounded. Dark brown, almost black above, pinkish white beneath. This remarkable fish is described from a single specimen, 500 millim. long, caught at Mpala in a floating basket-trap. Native name: Singa. Dr. Cunnington saw larger specimens, measuring 1020 and 1290 millim. A young specimen, 90 millim. long, was taken in a native trap at Mbete and is here figured (Plate XX XI. fig. d). The head is less depressed and the eyes are proportionally larger than in the adult; the ventrals are a little nearer the end of the snout than the root of the caudal. This fish, although until now unknown to zoologists, has been alluded to by Burton (‘The Lake Regions of Central Africa, ii. p. 67) and by Stanley (‘ How I found Livingstone, p. 530). The latter says: “The Silurus called by the Wajiji Singa, which grows, according to native report, to four and even six feet in length. The one I sketched was 384 inches long, and weighed 102 Ibs., but was considered to be a small one. It is an extremely flat fish, of a dark brown colour on the back, and light brown, inclined to whiteness, on the belly.” Coode Hore (‘ Tanganyika, p. 150) also alludes to it as “the oily Singa (Silurus) of six feet in length.” Cameron, on the other hand (‘ Across Africa,’ i. p. 270), applies the name Singa to Protopterus: “In the after- noon I shot a large Lepidosiren, called by the natives Singa ; but it was so loathsome to look at that no one would touch it, and the people declared it was poisonous.” 18. CHRYSICHTHYS BRACHYNEMA Bler. A single specimen from Niamkolo. Native name: Mondi or Nvulu. 19. CiRYSICHTHYS SIANENNA, sp.n. (Plate XXXII. fig. 2.) Depth of body 5} to 6 times in total length, length of head 4} times. Head moderately depressed, once and 2 to once and 4 as long as broad, smooth above ; snout rounded, projecting a littie beyond the lower jaw, } length of head, about 2 width of mouth; eye 4 to 44 times in length of head, once to once and 4 in interorbital width; nasal barbel very thin, 4 diameter of eye; maxillary barbel 552 MR. G. A. BOULENGER ON A COLLECTION once and } to once and 4 length of head; outer mandibular barbel nearly twice as long as inner and half length of head; vomero-pterygoid teeth forming a long and narrow crescentic band; occipital process narrow, narrowly separated from the small interneural shield. Dorsal I 6, nearly twice as distant from the root of the caudal as from the end of the snout; spine serrated behind, 4 length of head; longest soft rays 2 to 3 length of head. Adipose dorsal small, 2 to 3 times as long as deep its base 3 to 4 times in its distance from the rayed dorsal. Anal IV 10. Pectoral spine about 2 length of head, feebly serrated on outer side, more strongly on inner side. Ventral not reaching anal. Caudal deeply forked, with pointed lobes, outer rays 3 to 34 times as long as the median. Caudal peduncle twice as long as deep. Colour in life clive above, shading to light grey and silver below; anterior edge of dorsal fin red, the tip black ; maxillary barbels and ventral and caudal fins pinkish. Total length 220 millim. Said not to grow larger. Two specimens from Niamkolo and one from Mbete. Native name: Sianenna. A very distinct species, to be placed near C. furcatus Gthr. PHYLLONEMUS, g. n. Allied to Chrysichthys Blkr., but without nasal barbels and with the maxillary barbel fringed on both sides by a leaf-like membrane in its distal part; only vomerine teeth on the palate. 20. PHYLLONEMUS TYPUS, sp.n. (Plate XXXII. fig. 3.) Depth of body 54 times in total length, length of head 34 times. Head smooth, once and } as long as broad; occipital process very short, widely separated from the dorsal; snout flat, broadly rounded, subtruncate, projecting very slightly beyond the mouth; eye + length of head, equal to interorbital width; maxillary barbel once and 3 as long as head, extending beyond extremity of pectoral; outer mandibular barbels as long as head, twice as long as inner; vomerine teeth in a short trans- verse series. Dorsal I 6, a little nearer end of snout than root of caudal; spine short and feeble; longest rays 2 length of head. Anal 12, opposed to adipose dorsal, which is about 3 as long as its distance from the rayed dorsal. Pectoral spine 4 length of head, strongly serrated on its inner border. Ventral inserted a little behind vertical of dorsal, not reaching anal. Caudal peduncle a little longer than deep. Caudal deeply notched. Brown above, white below; maxillary barbels blackish at the end; dorsal dark brown, adipose dark brown, white behind; caudal brown, each lobe tipped with white; anal white, dark brown at the base; ventrals white. Total length 73 millim. A single specimen from Niamkolo. Native name: Kapondo. OF FISHES FROM LAKE TANGANYIKA. 503 21. AUCHENOGLANIS OCCIDENTALIS, var. TANGANICANUS. (Plate XX XIII.) Depth of body 4 to 44 times in total length, length of head 3 times. Head once and 4 as long as broad, coarsely granulate above; snout pointed, about 2 length of head; diameter of eye 7 to 9 times in length of head, 2 to 23 times in interorbital width, which equals 4 to 2 length of snout; width of mouth about half that of head; lips thick, papillose; maxillary barbel about 3 length of head, extending to posterior border of eye ; outer mandibular barbel as long as or but slightly longer than maxillary, inner mandibular + to + length of head; preemaxillary teeth forming two small groups, which are longer than broad; mandibular teeth forming a band interrupted in the middle ; occipital process broader than long, in contact with the large interneural shield. Humeral process small, acutely pointed. Dorsal I 7; spine rugose in front, not quite halt length of head. Adipose fin 34 to 53 times as long as deep, narrowly separated from the rayed dorsal and from the caudal. Anal IV 8. Pectoral spine as long as dorsal, feebly serrated in front, strongly behind. Ventral not reaching anal. Caudal rounded or slightly notched. Olive or brownish grey above, with more or less distinct round darker spots, uniform white beneath; dorsal and caudal fins with numerous round dark spots. Total length 550 millim. I have previously recorded this species under the name of Auchenoglanis biscutatus. But the Tanganyikan specimens differ in the longer snout and the narrower interorbital space. ‘They are hardly distinguishable from A. occidentalis C. & V., which inhabits the Nile, the rivers of West Africa, and the Congo, and has also been confounded with A, biscutatus. The above-given description is taken from two specimens from Kalambo aud one from the middle of the lake (Moore Collection) and from one from Mtondwe Bay (Cunnington Coll.), Native name: Poloko or Kavungwe. 22. SYNODONTIS MULTIPUNCTATUS Blgr. Niamkolo. Native name: HKamwengquwe. 23. SYNODONTIS MELANOSTICTUS, sp. n. (Plate XXXIV.) Body feebly compressed, its depth nearly equal to length of head and 4 times in total length. Head as long as broad, strongly granulate above, the rugosities extending on the snout; frontal fontanelle narrow; snout rounded, slightly longer than post- ocular part of head; eye supero-lateral, nearly 6 times in length of head, twice and + interorbital width ; occipito-nuchal shield obtusely tectiform ; mouth with moderately developed lips; maxillary barbel once and } as long as head, without fringe or border; mandibular barbels with long simple barbs, the outer nearly as long as the maxillary; premaxillary teeth forming a broad band, about 15 irregular transverse series; movable mandibular teeth about 30, the longest 2 diameter of eye. Gill-opening not extending below base of pectoral. Humeral process acutely pointed, without VOL. XVII.— PART VI. No. 3.—October, 1906. 4p 04 MR. G. A. BOULENGER ON A COLLECTION keel, granulate above, striated below, not extending beyond vertical of occipito-nuchal shield. Skin slightly villose on the sides of the body. Dorsal I7; spine strong, curved, nearly as long as head, striated, smooth in front, with a few very feeble and indistinct serre behind; a filament terminates the spine. Adipose dorsal 4 times as long as deep, twice and } as long as its distance from the rayed dorsal. Anal IV 7. Pectoral spine striated and strongly serrated on both sides, nearly as long as head. Ventral not reaching anal. Caudal deeply forked, upper lobe longer. Grey above, white beneath, closely dotted with black on the body as well as on the fins. Total length 260 millim. Grows to 520 millim. according to Dr. Cunnington. A single specimen from Lofu. Native name: Aiyegeyege. Allied to S. multipunctatus Blgr., zambesensis Peters, and punctulatus Gthr. Distinguished by the almost total absence of serration on the back of the dorsal fin-spine, the presence of strong serre on the outer border of the pectoral spine, and the villosities on the body. The eye is much smaller than in S. multipunctatus. This species is also nearly related to 8. schal/ Bl. Schn., in which the upper surface of the snout is smooth and the outer edge of the pectoral spine more feebly serrated. 24, MALOPTERURUS ELECTRICUS Gm. Niamkolo. Native name: Auwnta. CyPRINODONTIDA. 25. HAPLOCHILUS PUMILUS, sp.n. (Plate XXXII. fig. 4.) Depth of body 33 to 43 times in total length, length of head 3} to 4 times. Upper surface of head flat; snout shorter than the eye, the diameter of which is 3 times in length of head; lower jaw projecting beyond upper; interorbital space } length of head. Dorsal 10-11, originating nearer to root of caudal than to occiput; posterior rays longest, $ to 3 length of head. Anal 14-15, originating at equal distance from eye and from root of caudal ; posterior rays longest, about as long as dorsals. Pectoral 2 length of head, reaching beyond base of ventral. Caudal rounded, nearly as long as head. Caudal peduncle once and 4 as long as deep. Scales with the concentric striae very indistinct, 27 to 29 in a longitudinal series, 7 in a transverse series. Yellowish, more brownish on the back, where the scales have darker outlines; fins purplish, without any markings. Total length 34 millim. Three specimens, taken shore-wading at Kituta. Native name: Msipa.—The same species has been obtained in Lake Victoria by Mr. E. Degen. 26, HAPLOCHILUS TANGANICANUS Bler. Several specimens from Niamkolo, Moliro, and Mpala. Native name: AMishia. Grows to 135 millim. Dorsal and anal fins striated with dark and light lines. Or or or OF FISHES FROM LAKE TANGANYIKA. SERRANIDS. 27. LatEs MIcROLEPIS Blgr. (Plate XX XV. fig. 2.) Several specimens, measuring from 60 to 410 millim., from Niamkolo and Kituta. Native name: Pamba or Sangala. Dr. Cunnington has taken a photograph of a specimen measuring 1°34 in. 28. Lares ANGUSTIFRONS, sp.n. (Plate XXXV. fig. 1.) Depth of body 34 times in total length, length of head 5 times. Upper profile of head nearly straight; diameter of eye 44 times in length of head, once and 3 in length of snout; interorbital space narrow, only 3 diameter of eye; lower jaw projecting; maxillary extending to below centre of eye, the width of its distal extremity 4 diameter of eye; pre- and suborbitals finely serrated; cheeks, opercles, and occiput covered with small scales; preopercular border forming nearly a right angle, finely toothed on its vertical limb, with three widely separated spines on its lower limb, and with a very strong spine at the angle; two strong opercular spines. 4 or 5 strong spines on the clavicle. 14 gill-rakers on lower part of anterior arch. Dorsal VII, 112; third spine very strong, longest, about $ length of head; longest soft rays + length of head. Anal III 8; spines short, second and third equal. Pectoral not quite 4 length of head. Caudal rounded. Caudal peduncle once and 2 as long as deep. Scales 125 = Dr. Cunnington describes the coloration as of a yellowish tinge all over, darkish grey above, shading to almost white below ; middle region of body faintly mottled with darker grey ; ventral fins yellowish. Total length 265 millim. A single specimen from Niamkolo. Native name: Pamba ekundu (Red Pamba). This species, although very closely related to the preceding, is easily distinguished from it by its much narrower interorbital region! and by the rounded caudal. According to Dr. Cunnington, the natives regard this fish as distinct in kind from LI. microlepis, and they say it does not grow so large, the specimen here described being regarded by them as of the full size. CIcHLIDA. 29. LAMPROLOGUS BREVIANALIS, sp. n. (Plate XXXVI. fig. 1.) Depth of body 33 times in total length, length of head 3 times. Snout pointed, a little longer than the eye, the diameter of which is contained 3} times in length of head and equals interorbital width; maxillary extending to below anterior border ' Tn the young Z. microlepis the interorbital width equals at least 3 the diameter of the eye, and it equals 2 in a specimen nearly equal in size to the type of L. angustifrons (total length 245 millim.), An upper view of the head of this specimen is given on Pl. XXXV. fig. 2, for comparison. In a specimen 450 millim. long the interorbital width equals once and 4 the diameter of the eye. 452 506 MR. G. A. BOULENGER ON A COLLECTION of eye; 8 rather large canine teeth in front of the upper jaw and 4 in the lower, followed by a moderately broad band of minute, villiform teeth; lateral teeth small ; cheek naked, operculum scaly; pits of sensory canals very large. Gill-rakers short, 9 on lower part of anterior arch. Dorsal XVITI—XIX 11, spines subequal from the eighth, 2 length of head; longest soft ray 2 length of head. Anal IV 7, fourth spine as long as longest dorsals. Pectoral rounded, 3 length of head. Ventral reaching origin of anal. Caudal rounded. Caudal peduncle slightly longer than deep. Scales 40-43 eo lateral lines a Brown, somewhat lighter beneath, with darker and lighter longitudinal streaks on the sides and five broad dark transverse bands ; vertical fins with numerous round whitish spots. Total length 83 millim. A single specimen from Niamkolo. Native name: Sietela. A young specimen was obtained in Komba Bay. Very closely allied to L. tetracanthus Blgy., with which it agrees in the short anal fin with only four spines. Distinguished by the larger mouth extending to below the anterior border of the eye, and by the absence of a whitish marginal band on the dorsal fin. 30. LAMPROLOGUS ELONGATUS Bler. This species, which was established on young specimens, proves to be the giant of the genus, exceeding L. lemairii in size. Three large specimens were brought home by Dr. Cunnington. 1. Niamkolo, off Mtondwe Island. Nativename: Avbua. Total length 325 millim. 10) OVIOTU ILS A OR Stay, CR ee 16, 1h 25 ? 2. Sumbu. Total length 120 millim. D. XVII 10; A. V 8; Sq. 85 3 L. 1 a 3. Niamkolo Bay, taken from an empty Meothawma shell, dredged in a few fathoms. Total length 63 millim. D. XVIII10; A. VI8; Sq. 925; L123 The cheek, which is naked in the young, is much deeper (3 times diameter of eye) and covered with very small scales in the adult. The depth of the body is contained from 33 to 4 times in the total length, the length of the head 22 to 5 times. Diameter of eye, in the adult, 8 times in length of head, 3 times in length of snout, once and 4 in interorbital width; the band of minute teeth behind the canines is much broader than in the young, and the longest rays of the dorsal and anal fins measure nearly 2 the length of the head. Dr. Cunnington describes the adult, when fresh, as black with white mottling. The vertical fins are dark brown, with round light spots. Dr. Cunnington has noted that the fish readily takes artificial spinning-bait. OF FISHES FROM LAKE TANGANYIKA. 557 31. LAMPROLOGUS CUNNINGTONI, sp. n. (Plate XXXVI, fig. 2.) Depth of body 3? to 44 times in total length, length of head 52 times. Snout obtusely pointed, twice as long as eye, the diameter of which is contained 5 times in length of head, and once and } in interorbital width; maxillary not extending quite to below anterior border of eye; 6 large canines in front of upper jaw, 4 in front of lower, followed by a rather broad band of minute, villiform teeth; lateral teeth small; cheek naked, its depth once and + diameter of eye; operculum with a few scales. Gill-rakers short, 8 on lower part of anterior arch. Dorsal XX 10-11, spines increasing in length to the last, which measures about $ length of head; soft dorsal acutely pointed behind, longest ray measuring 3 length of head. Anal V 7, last spine longest, as long as last dorsal ray; the fin pointed like the dorsal. Pectoral rounded, 2 length of head. Ventral reaching vent or a little beyond. Caudal rounded. Caudal peduncle once and § as long as deep. Scales 85-87 — lateral lines = Greyish brown above, white beneath; small light spots on the sides; very indistinct traces of five or six dark bars across the back; vertical fins brown, with rather indis- tinct small roundish light spots. Young with 7 or 8 regular dark bars on the body. Total length 190 millim. Two adult specimens from Moliro, native name IMsheshe; and two young from Mbete, native name Twinanzoli. This species is allied to L. elongatus, but easily distinguished by the rounded caudal fin, the shorter and fewer gill-rakers (8 instead of 12), and the more numerous spines in the dorsal fin (20 instead of 17 or 18). 32. LAMPROLOGUS TRETOCEPHALUS Blgr. A single specimen from Kigoma. Native name: Ndubu. Total length 53 millim. 1D) SAVVAUE TOS: PAULIN, Ws ister G17 as 1B, = Snout not longer than eye; canine teeth very small; ventral fin reaching beyond origin of anal. 33. LAMPROLOGUS MONDABU, sp. n. (Plate XXXVI. fig. 3.) Depth of body 32 times in total length, length of head 3 times. Snout pointed, once and 2 to twice as long as eye, the diameter of which is contained 43 to 5 times in length of head and equals or is slightly less than interorbital width ; maxillary extending to halfway between verticals of nostril and eye; 6 large canines in front of each jaw, followed by a rather broad band of minute, villiform teeth; lateral teeth small; cheek naked, its depth once and + diameter of eye; operculum with a few scales. Gill-rakers very short, 7 or § on lower part of anterior arch. Dorsal XIX 9; spines increasing in length to the last, which measures 3 length of head; soft dorsal acutely pointed behind, longest ray measuring } length of head. Anal V 7, last spine longest, 1 length of head ; the fin pointed like the dorsal. Pectoral rounded, 3 length of head. 558 MR. G. A. BOULENGER ON A COLLECTION Ventral reaching origin of anal. Caudal slightiy emarginate. Caudal peduncle as eo 25-27 13-14’ 9-14" mens, as noted by Dr. Cunnington, varies from grey to olive, rather lighter below, but very uniform; dorsal edged with yellow, bluish in its upper portion, with yellow long as deep. Scales 42-46 lateral lines The coloration, in fresh speci- spots; upper half of caudal minutely spotted with yellow, lower half darker. Total length 105 millim. Two specimens. Native name: Mondabu. Allied to LZ. modestus. Distinguished by the slightly emarginate caudal fin and the shorter mouth. 34, LAMPROLOGUS MODESTUS Blgr. A young specimen from Komba Bay. Di XX°S= AVI 6.) Sc. 36: 2-71. 1 = Yellowish, with 7 brown lines across the 13 o) back. 35, LAMPROLOGUS MULTIFASCIATUS, sp.n. (Plate XXXV. fig. 3.) Depth of body 35 to 3} times in total length, length of head 22 to 3 times. Snout obtusely pointed, as long as the eye, the diameter of which is contained 34 times in length of head and equals interorbital width; maxillary extending to below anterior border of eye, or a little beyond; 8 canine teeth in front of each jaw, outer very large, followed by a narrow band of villiform teeth; lateral teeth small; cheek naked, operculum scaly. Gill-rakers moderately long, 8 on Jower part of anterior arch. Dorsal X VII-X VIII 8-9, spines subequal from the eighth or ninth, about 2 length of head; longest soft ray } length of head. Anal VI-VII 6, last spine slightly longer than longest dorsal. Pectoral rounded, 2 length of head. Ventral nearly reaching origin of anal. Caudal rounded. Caudal peduncle as long as deep. Scales 35-38 ca upper lateral line short, 8-15, lower absent. Pale brown above, white beneath, with 16 or 17 very regular dark brown bars, the last 3 or 4 of which are on the caudal fin; dorsal fin with oblique dark brown bars. Total length 34 millim. Five specimens from Niamkolo Bay, dredged among shells in about 12 fathoms. This little fish is easily distinguished from all its congeners by its elegant markings and by the total absence of the lower lateral line. 36. LAMPROLOGUS LEMAIRIT Bler. Eight specimens from Niamkolo, measuring 102 to 220 millim. Native names: Kibua, Miseke or Moseke, Sietela. Seven specimens have 19 spines and 7 soft rays in the dorsal fin, one has 18 spines and 8 soft rays; in the largest specimen the dorsal spines increase in length to the OF FISHES FROM LAKE TANGANYIKA. 559 last, which measures half the length of the head, and the pectoral fin measures two- thirds the length of the head. ‘The anal spines number constantly 8; one specimen has 6 soft anal rays, the others having 5. The adult specimen is of a uniform dark brown; the other specimens, according to the collector’s notes, were grey-brown or olive above, shading to white beneath, the scales of the upper parts dark-edged; the dark bars on the body less distinct than in the type specimen ; the dorsal fin is tipped with black, the pectorals are olive and the ventrals black. 37. LAMPROLOGUS HECQUI Bler. Two specimens from Niamkolo Bay, taken in an empty Neothawma shell, and three from Mtondwe Bay. ‘Total length 45-60 millim. Regarded by the native fishermen as adult, and called by them Mitota, Nyonga, and Mpama. Depth of body 53 to 52 times in total length. Dorsal XVIII-XIX 8-9; anal VI-VII 7-8. Scales 52-56 nae lat. 1. — Gill-rakers rather long and slender, 12 to 15 on lower part of anterior arch. Grey above, lighter below, with 8 or 9 dark vertical bars on each side, or large blotches alternating with narrow bars; a more or less distinct dark round spot at the root of the caudal fin ; vertical fins with dark bars ; dorsal fin and upper part of caudal with a pale border faintly edged with black, as in L. callipterus. 38. LAMPROLOGUS CALLIPTERUS, sp. n. (Plate XXXVI. fig. 4.) Depth of body equal to length of head, 33 to 3} times in total length. Snout obtusely pointed, once and 4 as long as eye, the diameter of which is contained 4 times in length of head and hardly equals interorbital width ; maxillary extending to below anterior border of eye; 6 or 8 large canines in front of each jaw, followed by a moderately broad band of minute, villiform teeth; lateral teeth small; cheek entirely naked or almost so; operculum with large scales. Gill-rakers rather long, 10 to 12 on lower part of anterior arch. Dorsal XVIII—-XIX 9, spines increasing in length to the last, which measures nearly $ length of head; soft dorsal acutely pointed, longest ray 2 to? length of head. Anal VIII 7-8, last spine nearly as long as last spine of dorsal; the fin pointed like the dorsal. Pectoral rounded, about 3 length of head. Ventral extending to origin of anal, or beyond. Caudal rounded. Caudal peduncle 4. 24-25 10-11 ? 10-13" Brown or olive above, paler or whitish below, the scales edged with darker; dorsal fin with a yellow border finely edged with blackish, this border continued on the upper third of the caudal; vertical fins with numerous small round whitish spots. Three specimens, measuring from 90 to 125 millim. The largest is from Mpala, the once and + to once and 4 as long as deep. Scales 58-40 lateral lines two others from Niamkolo. Native name: Zagezi and Miseke or Moseke. 560 MR. G. A. BOULENGER ON A COLLECTION Allied to Z. lemairti and L. hecqui. Distinguished from the former by the longer gill-rakers, the broader interorbital space, fewer scales in a vertical series, and more numerous soft rays in the anal fin; from the latter by larger scales and 8 anal spines instead of 6 or 7. 39, LAMPROLOGUS RETICULATUS, sp.n. (Plate XXXYV. fig. 4.) Depth of body 4 times in total length, length of head 35 times. Snout pointed, as long as the eye, the diameter of which is contained 34 times in length of head and equals interorbital width ; maxillary extending to below anterior fourth of eye ; canines very small, followed by a narrow band of minute, villiform teeth; a few scales on the cheek; operculum with large scales. Gill-rakers short, 14 on lower part of anterior arch. Dorsal XIX 10, spines increasing in length to the last, which measures 4 length of head; longest soft dorsal ray but slightly longer than longest spine. Anal VIII 8, last spine as long as last dorsal spine. Pectoral rounded, # length of head. Ventral extending to origin of anal. Caudal truncate, slightly emarginate. Caudal peduncle 4 BL jg; lateral lines we Brown, the scales edged with darker, forming a reticulate pattern ; dorsal and anal fins dark brown, with oblique rows of round whitish spots; caudal fin brown, with numerous, rather indistinct light spots at the base, and an oblique whitish streak in the upper corner. Total length 62 millim. A single specimen from Vua. Native name: Zagezi, like L. callipterus. Distinguished from L. callipterus, the young of which it resembles in coloration, by the truncate caudal fin and the smaller teeth. once and } as long as deep. Scales 39 40. LAMPROLOGUS CALLIURUS, sp. n. (Plate XXXV. fig. 5.) Depth of body 23 to 3 times in total length, length of head 3 times. Snout obtuse, a little shorter than the diameter of the eye, which is } length of the head and a little greater than interorbital width ; maxillary extending to below anterior fourth of eye; outer canines strong, curved; a narrow band of minute, villiform teeth; cheek naked ; operculum scaly. Gill-rakers long and slender, 14 to 16 on lower part of anterior arch. Dorsal XVII-XVIII 7; spines subequal from the eighth or ninth, measuring about } length of head; soft rays a little longer. Anal VIII 6-8, last spine nearly as long as longest dorsal spines. Pectoral rounded, 3 length of head. Ventral extending beyond origin of anal. Caudal truncate. Caudal peduncle a little deeper than long. Scales 35-36 24 ; lateral lines 8. Yellow, brownish on the back, with rather 12-13 ’ S210 indistinct darker bars across the back and across the caudal fin; a sharply defined black. opercular spot, edged with silvery in front. Total length 50 millim. OF FISHES FROM LAKE TANGANYIKA. 561 Three specimens, dredged among shells in about 30 fathoms at Tembwi. Nearly allied to Z. brevis. Differing in the truncate caudal and in the longer and more numerous gill-rakers. 4], LAaMpPRoLoets BREVIS Blgr. This species was founded on a single specimen, 35 millim. long, obtained by Captain Hecgq at Albertville. The Cunnington collection contains two specimens, measuring 45 and 38 millim. respectively, from Rumonge, which agree very closely with the type, but show variation in the fin-rays, viz.:—D. XVII-XVIII 7; A. VII-VIII 6. Depth of body equal to length of head, 3 times in total length. 30 to 35 scales ina longitudinal series ; scales on nape very small. Unitorm brown above, whitish beneath ; fins brown, without spots, dorsal tipped with black, ventrals black at the end; opercular black spot strongly marked. 42. LAMPROLOGUS COMPRESSICEPS Bler. A single young specimen from Niamkolo. The number of species of Lamprologus now known from Lake Tanganyika amounts to 18, the principal characters of which are tabulated on p. 962. 43. TELMATOCHROMIS TEMPORALIS Bley. 11 specimens from Kaboge, Mshale, Kituta, Mtondwe Bay, and Niamkolo Island. Native names: Mébulila, Fisia (young), and Kamfoto (young). Dorsal) KVM -XX 7-10), Anal WIV 627 Sq 40-47% ==) 1) ee Dr. Cunnington describes the adult as olive, with dark patches in dorsal region and irregular blotches on body and root of tail; the temporal band orange; dorsal fin spotted with yellow and white, and tipped with red; caudal and anal spotted with yellow and white. In the young the body may be crossed by dark bars; the temporal band is very distinct, and the bar at the base of the pectoral fin is edged with whitish behind. 44, PARATILAPIA FRONTOSA, sp. u. (Plate XX XVII. fig. 1.) Depth of body twice and } in total length, length of head nearly 3 times. Forehead gibbose !, the head at the interorbital region as deep as the body; snout twice and 3 as long as the eye, the diameter of which is contained 5} times in length of head and nearly twice in interorbital region; maxillary not extending to below anterior border of eye; 3 series of teeth in each jaw, outer large; 5 series of scales on the cheek ; large scales on the operculum. Gill-rakers rather long, 13 on lower part of anterior arch. Dorsal XIX 8; spines increasing in length to the last, which measures a little ‘ Probably in adult males only. VOL, XVII.—PART VI. No. 4.—October, 1906. 4F 562 MR. G. A. BOULENGER ON A COLLECTION Principal Characters of the Species of Lamprologus. Dorsal. Anal. Scales. ee Caudal. L. brevianalis...... XVITI-XIX 11 IY 7 40-43 = 9 Rounded. L. tetracanthus ....| XX 10 po song Ad 10 | Rounded. L. elongatus ...... | XVII-XVIII 10-11) v-vr7-9 | 85-95 wee | 1 Truncate. Wormintre a.c4) soc | TT 95-87 9 | 8 Rounded. L. tretocephalus ....| XVI-XVII 10-11 | V6 37-42 es 9 Rounded. L. mondabu ...... XIX 9 v7 2-46 | 7-8 [Slightly emarginate. L, modestus........ XIX-XX 810 | v-vI6-7 | 36-40 a 7 Truncate. L. multifasciatus ..| XVII-XVIII 8-9 | VI-VII6 | 35-38 34 8 Rounded. L. lemairii ........ XVIII-XIX 7-8 | VIII5-6 | 40-48 35, | 9-10 Rounded. Te hecgut sn sete oso | XVIUL-XIX 8-9 | VI-VIT7-8 | 52-57 ct, | 12-16 Rounded. Tb, GUO 56 646600 XIX-XX 8-9 VII-VIII 6-7| 33-35 nam 9-10 Rounded. L. callipterus ...... | XVIII-XIX 9 VIII 7-8 | 38-40 4 | 10-12 Rounded. DL, retwculatus ...... XIX 10 VIII 8 39 * 14 Truncate. L. calliurus........ KVIII-XIX 8 VIII 6-8 | 35-36 <4. | 14-16 Truncate. ES breuis es Nene XVII-XVIII 6-7 | VII-IX 5-6| 33-35 3 9 Rounded. L. compressiceps.... XX-XXI 6 Xx 5 32-33 3 15 Rounded. L. fasciatus ...... XIX 8 X 6 46 = 12 Rounded. Oy UGHREP ooobeo ae XIX-XXI 7-9 VI-VII 5-6 | 50-60 ea 14-16 | Deeply notched. less than 4 length of head; soft dorsal much produced, longest ray nearly as long as head. Anal III 7, third spine as long as, but stronger than, last dorsal, soft rays like those of dorsal. Pectoral acutely pointed, as long as head, reaching origin of anal. Ventral produced into a long filament, extending beyond origin of anal. Caudal rounded. Caudal peduncle as long as deep. Scales 36 - ; lateral lines =p Brownish abeve, whitish beneath, with six broad dark cross-bands, the last two on the caudal peduncle; a dark brown opercular spot and a large dark brown spot at the base of the pectoral fin ; fins without markings. Total length 255 millim. OF FISHES FROM LAKE TANGANYIKA. 563 A single specimen from Kigoma. Native name: Ndubu. This species occupies an isolated position. The number of spinous rays in the dorsal fin exceeds that of any species previously described of the genus Paratilapia. 45. PARATILAPIA PFEFFER! Bler. Two specimens from Sumbu. Native name: Kilomo. Dorsal XVI 9-10; Anal III 7-8; Scales 32-335; Lateral lines 443. 3 or 4 series of scales on the cheek. Dr. Cunnington has noted that the female nurses the embryos in her mouth. One of the specimens, measuring 72 millimetres, has a few large ova (24 millim. in diameter) in the pharynx. 46. PARATILAPIA VENTRALIS Bler. Two specimens from Sumbu. Native name: Lala. Dr. Cunnington describes the coloration as follows :—Blackish all over, but with blues and greens showing in places; dorsal fin light blue-green, tipped with blackish ; anal similar, but with broader band of black; ventrals black, bright yellow an bifurcate at the end. 47. PARATILAPIA DEWINDTI Bler. One specimen from Sumbu and one from Lofu River. Native name: Avsangi. Depth of body equal to length of head, twice and # in total length. Lateral lines 30-34 I-18" longitudinal stripes of violet alternating with stripes of yellowish; a black blotch on the gill-cover behind the eye ; gular region yellowish ; dorsal fin with brownish stripes and tipped with yellow; caudal with olive vertical bars; ventrals and anal blackish at Described by Dr. Cunnington as dark olive above, silvery below, with five the ends. 48. PARATILAPIA FURCIFER Bler. Two specimens from Niamkolo and two from Sumbu. Native names: Liukonko (plural Wakonko) and Lala. The largest specimen measures 135 millim., the types not exceeding 110 millim. In these large epegmicns the length of the eye equals that of the snout and the interorbital width and 4 the length of the head. Dorsal XIJI—-XIV 12-14; Anal III 9; Scales 58 5-6 50- 60-64; et lines 53-35 Dr. Cunnington describes the colour as olive above, lighter to silvery below, with bright iridescent blues and greens; gill-cover yellow; dorsal and anal fins light iridescent violet, with oblique stripes of yellow passing into dark stripes posteriorly, 42 564 MR. G. A. BOULENGER ON A COLLECTION the rays tipped with black; caudal yellow, with dark dorsal and ventral margins; ventrals light bluish with some yellow stripes, the produced filaments tipped with yellow. 49, PARATILAPIA STENOSOMA Bler. A single specimen, measuring 240 millim., from Kaboge. Dorsal XV 13; Anal III 14; Scales 69 os ates s the upper extending un- interrupted from above the operculum to the root of the caudal fin. 28 gill-rakers on lower part of anterior arch. 50. BATHYBATES FEROX Bler. Two specimens from Niamkolo and two from Kasakalawe. Native names: Lilembela (plural Malambela) and Molirdi. These specimens measure 160 to 275 millim. and have the fin-formula: D. XIV 15, A. If 16-17. 14 or 16 gill-rakers on lower part of anterior arch. The length of the pectoral fin may not exceed 3 that of the head. 51. BaTHyBaTEs Frasciatus Blgr. A single specimen, measuring 3 Dorsal XVI 17; Anal III 17; 150 ae Lat. 1. = Length of head 34 times in total length. Caudal peduncle twice 39 millim., from Niamkolo. Native name: Mishipa. ] oD 9 gill-rakers on lower part of anterior arch ; Scales as long as deep. The black basal band on the dorsal fin is absent. 52. BATHYBATES MINOR, sp.n. (Plate XX XVIII. fig. 1.) Depth of body 34 to 33 times in total length, length of head 22 to 3 times. Snout with convex profile, slightly longer than the eye, the diameter of which is contained 34 to 32 times in length of head and slightly exceeds interocular width ; maxillary extending to below anterior border of eye; teeth in 2 or 3 series in the upper jaw, in 2, in the lower; 5 series of small scales on the cheek. Gill-rakers rather short and thin, 12 on lower part of anterior arch. Dorsal XIII 15-16, deeply notched, the eighth to eleventh spines being only about half as long as the fifth and last, which are the longest and measure about ? length of head; longest soft rays 2 to} length of head. Anal IIT 16-17, longest soft rays 2 to } length of head. Pectoral pointed, 3 length of head. Ventral not reaching vent. Caudal forked. Caudal peduncle 13 to 13 times as long as deep. Scales small and very irregular, 90-95 aa lateral lines eae Brown above, silvery on the sides and below; spinous dorsal black, white at the base; a black streak along the middle of the soft dorsal. Total length 160 millim, Two specimens from Kituta and Lofu. Native names: Nswepa and Lulambi. Also OF FISHES FROM LAKE TANGANYIKA. 565 some quite young specimens, 30 to 55 millim. long, from Mbete. Several specimens bad previously been obtained by Mr. Moore on the Usambara market ; but they were badly preserved, and I took them to be the young of B. ferow. Pellegrin has placed Bathybates in his division “ Opisthospondyli”’. I can see no justification for this course, the vertebral column not differing notably from such forms as Paratilapia robusta and longiceps. Bathybates is more nearly related to Paratilapia than to any other genus, differing chiefly in the longer anal fin and in the dentition ; this consists of an outer series of ankylosed teeth, behind which are one to three series of hinged depressible teeth. We are now acquainted with three species, the principal characters of which may be contrasted as follows :— I. Spinons dorsal fin not notched, the posterior spines subequal; 3 or 4 series of teeth in the upper jaw, 3 in the lower. D. XIV-XV 15-16; A. IIT 16-17; Sq. 78-100 en 13-15 gill-rakers on lower part of anterior arch ; interocular width less than diameter of eye. B. ferow Blgr. D. XVI 16-17; A. III 17-18; Sq. 140-150 = 18-19 gill-rakers on lower part of anterior arch ; interocular width greater than diameter of COE ne! En lo Os CMP SPE el! ints ov MoM ol be ce ce en earaunon dellge, II. Spinous dorsal notched, the last spine twice as long as the antepenultimate ; 2 or 3 series of teeth in the upper jaw, 2 in the lower. D. XIII 15-16; A. III 16-17; Sq. 90-95 2%; 12 gill-rakers on lower part of anterior arch ; interocular width slightly less than diameter of CVC ES ET Nt ce ek a ase Ot te Pere Rem elma topeeert LSP 700720 eS I may here add, as a correction to the original diagnosis of the genus, that well- preserved specimens show the maxillary bone not to be exposed when the mouth is fully closed. HAPLOTAXODON, g. n. Mouth nearly vertical, directed upwards, with a single series of small, equal, conical teeth. Maxillary exposed at the end. Body elongate, strongly compressed, covered ‘So far as I am acquainted with the number of vertebrz in the African Cichlide, the following species only are to be referred to this division :— Enantiopus longianalis ..........-+-- +200 eee eens 144 24=38. XGHEIT MGUY sonccgdupGonodabcenononjopeoboso be 134 22=35. ss CHULOMYES So bdo 0 cDGoDoCUGcdEEONOOUnOOD 14+ 24=38. Graninalotrra lematnits oo eae desler s ee bate la etre 14422=36. Trematocara wnimaculatum .. 16... ee ee ec ee eee ne ne eees 12+19=31. - HEPAT dao sood boon bbe 6050000009000 00 13+-16=29. FUMING NUD O66 OD OdbD OR bOG0 DUC OC0Ua UN OD OUEE 15+20=35. All’generic forms restricted, according to the present state of our knowledge, to Lake Tanganyika. 566 MR. G. A. BOULENGER ON A COLLECTION with small ctenoid scales; two lateral lines, the upper complete. Dorsal with 17 or 18 spines, anal with 5. Vertebre 38 (19+ 19). The presence of a single series of teeth distinguishes this genus from Paratilapia, with which it is connected by P. stenosoma. 53. HaPLOTAXODON MICROLEPIS, sp. n. (Plate XX XVIII. fig. 2.) Depth of body 34 to 3} times as long as deep, length of head 8} to 33 times. Snout much shorter than the eye, chin very prominent; eye very large, 2? to 3 times in length of head (chin included), its diameter much greater than interocular width ; maxillary not extending to below anterior border of eye; two series of scales on the cheek; opercle scaly. Gill-rakers long and slender, closely set, 18 to 20 on lower part of anterior arch. Dorsal XVII-X VIII 11-12; spines strong, equal in length from the 9th or 10th, 4 length of head ; longest soft rays about 3 length of head. Anal III 9; spines very strong, 3rd longest, } length of head. Pectoral acutely pointed, slightly shorter than head, not reaching origin of anal. Ventral produced into a filament extending a little beyond origin of anal. Caudal peduncle twice as long asdeep. Caudal deeply 65-77 notched. Scales 70-77 ae lateral lines 77,; 4 or 5 series of scales between the lateral lines in front. Olive or grey above, shading off to silvery white below; dorsal and caudal grey-brown, with yellowish-white large round spots; anal and ventrals white, tinged with yellow. Total length 210 millim. One specimen from Niamkolo (native name Liukonko, plural Makonko), one irom Kasawa, and one from Kasanga (native name Lukoko). 54, PELMATOCHROMIS POLYLEPIS Bler. (Plate XXXVIL. fig. 2.) Mtondwe Bay and Niamkolo, four specimens, measuring from 102 to 500 iniilim. Dr. Cunnington measured one of 435 millim. Dorsal XV 15-15; Anal III §; gill- rakers 12-14 on lower part of anterior arch ; Scales 58-67 3; Lateral lines 16-19 the largest specimen the snout is once and two-thirds as long as the eye, which is 4 times in the length of the head. The smallest specimen (here figured) is remarkable in having the 35rd to 5th spines of the dorsal longest, half the length of the head. The 30-33 ia in coloration is described by Dr. Cunnington as greyish or brownish grey above, white below, with or without five or six inconspicuous vertical darker stripes on the body; the markings on the dorsal fin brown; patches of red at the angle of the jaws and at the roots of the paired fins; a red vertical stripe on the gill-cover; anterior rays of dorsal tipped with red. According to Dr. Cunnington, the female carries eggs in her mouth. Native names: Sanga and Kanga. OF FISHES FROM LAKE TANGANYIKA. 567 55. PELMATOCHROMIS MELANOSTIGMA, sp.n. (Plate XX XIX. fig. 1.) Depth of body 23 to 54 times in total length, length of head 22 to23 times. Snout with more or less curved upper profile, as long as or a little shorter than the eye, which is 2? to 3 times in length of head and exceeds interocular width; mouth extending to below anterior border of eye; teeth forming a narrow villiform band, some on the sides of the lower jaw strongly enlarged; two series of scales on the cheek. Gill- rakers short, 10 to 12 on lower part of anterior arch. Dorsal XV-XVI 11-12; spines rather strong, equal in length from the sixth or seventh, 2? to 3 length of head; longest soft rays } length of head. Anal III 6-7; third spine a little shorter than longest dorsals. Pectoral acutely pointed, slightly shorter than head, reaching origin of anal. Ventral reaching origin of anal or beyond. Caudal peduncle once and 3 as long as deep. Caudal deeply notched. Scales finely denticulate, 34-48 as lateral lines Bo regular longitudinal series of small black spots on the back and sides; a blackish opercular spot; a reddish spot below the eye; dorsal fin yellow, tipped with blackish, with black spots, which may form oblique series, and an orange marginal band in Dark olive above, yellow on the sides, lighter to silvery below; more or less front; pectoral yellow ; gular region and ventral fins orange. In other specimens the black spots are absent and the dorsal and anal may be broadly edged with black. The vertebree number 34 (16-418). Total length 115 millim. Two specimens from Uyira (native name Mariongo), one from Kaboge (same native name), five from Niamkolo (native name MWalenda or Marenda), one from Lofu (same native name), and one from Kituta. 56. PELMATOCHROMIS PLEUROSPILUS, sp. n. (Plate XX XIX. fig. 2.) Depth of body 34 times in total length, length of head 22 to 3 times. Snout with curved upper profile, as long as the eye, which is 3 times in length of head and a little exceeds interocular width ; mouth not extending quite to below anterior border of eye ; teeth forming a narrow villiform band; two series of scales on the cheek. Gill-rakers short, 11 or 12 on lower part of anterior arch. Dorsal XI—XII1 12-15 ; spines equal in length from the fifth, } length of head, soft rays not longer. Anal III 7-8; third spine as long as longest dorsals. Pectoral acutely pointed, slightly shorter than head, reaching origin of anal. Ventral reaching origin of anal. Caudal peduncle twice and 3 \ y . 96 rp 2h as long as deep. Caudal deeply notched. Scales finely denticulate, 32-35 7); lateral lines oe Olive above, silvery white beneath ; a black opercular spot, followed by a series of six black spots along the side of the body; the back of the male shot with red; dorsal with blackish or brownish spots, tipped with black; dorsal and anal edged with orange, more conspicuously in the male than in the female; pectoral 568 MR. G. A. BOULENGER ON A COLLECTION yellow or orange; caudal with an orange band, with (¢ ) or without ( 9 ) white and grey markings. Total length 75 millim. Three specimens from Mpala and three from Tembwi. Native name: Sembe. The females carry eggs in the mouth. 57. PELMATOCHROMIS RHODOSTIGMA, sp. n. (Plate XX XIX. fig. 3.) Depth of body 34 times in total length, length of head 3 times. Snout with slightly curved upper profile, as long as the eye, which is 3 times in length of head and slightly exceeds interocular width ; mouth extending to below anterior border of eye; teeth forming a narrow villiform band, some enlarged and curved on the side of the lower jaw ; two series of scales on the cheek. Gill-rakers short, 10 on lower part of anterior arch. Dorsal XIII 13-14, spines equal from the fifth, about 3? length of head, soft rays not longer. Anal IIT 7-8, third spine as long as longest dorsals. Pectoral acutely pointed, as long as head, reaching origin of anal. Ventral reaching origin of anal. Caudal peduncle once and }as long asdeep. Caudal deeply notched. Scales finely denticulate, 34-35 ip lateral lines = Olive above, silvery white below; longitudinal series of pink dots on the body; throat pink; dorsal streaked with pink; pectorals and ventrals pink ; a pink band on each lobe of the caudal. Total length 75 millim. Two specimens from Mshale. Native name: Kasalu. This species is closely allied to the preceding. It differs in the enlarged teeth on the sides of the lower jaw and in the number of scales in a transverse series. 58. PELMATOCHROMIS auRITUS Blgr. (Plate XXXIX. fig. 4.) Two adult specimens from the Kazzaga market, native name Aibombobombo; three young from Mshale, native name Kanyamitenga. The coloration of the adult when fresh is described as light olive above, silvery below, with a good deal of yellow towards the tail; three or four longitudinal rows of iridescent bluish spots in the middle third of the body; dorsal and anal fins with white spots, the former tipped yellow and black; pectorals yellowish, ventrals white; caudal with white spots and yellow bars, and tipped blackish. Young specimens, measuring 65-70 millim. (Plate XX XIX. fig. 4), have very distinct silvery oblique stripes on the body ; the dorsal and anal fins are unitorm brown. The caudal fin of these young specimens is truncate, not emarginate. I agree with Pellegrin in referring this species, originally described as a Paratilapia, to Pelmatochromis. 59. EcroDuUs DESCAMPs! Bley. Three specimens from Sumbu. Native name: Kasha. Total length 80 millim. In his notes Dr. Cunnington mentions that these are OF FISHES FROM LAKE TANGANYIKA. 569 apparently full-sized females, with eggs in the mouth. A few eggs, measuring 4 millim. in diameter, are still in the pharynx. Dorsal XII15; Anal III 10-11; Scales 37-38 * Lateral lines 7; 17 gill-rakers on lower part of anterior arch. Vertebre 36 (17+19). This genus was established for two species, E. descampsi and EL. melanogenys. ‘The second differs, however, in having the inner rays of the ventral fin longest and the caudal vertebree much more numerous than the preecaudals, characters which it shares with Xenotilapia ; whilst in the first the outer soft rays of the ventral are, as usual, the longest, and the caudal vertebra do not differ much in number from the precaudals. I therefore propose a new genus, Enantiopus, for H. melanogenys and its near ally E. longianalis. 60. ENANTIOPUS MELANOGENYS Bler. Two specimens, 70-75 millim. long, from Kigoma. Native name: Mulunda. 22-97 | Dorsal XIV 15; Anal III 13; Scales 40 .°.; Lateral lines - 92103 jaa) 11 or 12 gill-rakers on lower part of anterior arch. 61. ENANTIOPUS LONGIANALIS Bler. Three specimens from Mbete, native name Aveta; two from Moliro, native name Karonda ; one from Rumonge and one from Sumbua. The last two, females, carrying eggs in the pharynx. ‘These specimens measure 100 to 135 millim. Dorsal XIV 16-17; Anal III 16-17; Scales 40-42 2; Lateral lines gill-rakers on lower part of anterior arch. The number of rays in the anal fin is quite sufficient for separating this species from the preceding. 25-29 , 72 op: 12 or 13 62. XENOTILAPIA sima Blev. Two specimens, measuring 130 millim., from Niamkolo. Native name: Mtwnunu. re 25-26 Dorsal XIV 14; Anal III 11; Scales 37-38 5; Lateral lines ESE 10 or 11 gill- 11-16 rakers on lower part of anterior arch. Dr. Cunnington describes the colour as delicate brownish aboye, silvery below; head dark above. 63. GRAMMATOTRIA LEMAIRII Blgv. Two specimens from Niamkolo, native name Mendi, and one from Kasawa, native name Lungi. ‘These specimens measure from 130 to 210 millim. 41-43 Dorsal XV 14-15; Anal III-IV 10-11; Scales 52-55 at} Lateral lines Ze 10 or e =6 11 gill-rakers on lower part of anterior arch. Dr. Cunnington describes the colour as uniform grey or delicate brownish above, silvery below; opalescent blues below VoL. xvil.—Part vi. No. 5.—October, 1906. 46 570 MR. G. A. BOULENGER ON A COLLECTION the eye and on the gill-cover; dorsal fin fringed with black. The vertebra number 86 (14422). 64. TREMATOCARA NIGRIFRONS, sp. n. (Plate XLI. fig. 1.) Depth of body 34 times in total length, length of head 3 times. Snout with curved upper profile, a little shorter than the eye, which is twice and ? in length of head and a little exceeds interorbital width ; mouth extending to below anterior fourth of eye; head naked, with large cavities as in the other species of the genus. Gill- rakers long and slender, closely set, 20 or 21 on lower part of anterior arch. Dorsal X 11-12; spines feeble, increasing in length to the last, which measures } length of head; last soft rays a little shorter. Anal III 10; third spine as long as longest dorsal. Pectoral acutely pointed, a little longer than the head. Ventral not reaching origin of anal. Caudal with deep crescentic notch. Caudal peduncle once and 3 as long as deep. Scales 29-30; upper lateral line 7-8 ; lower lateral line absent. Yellow; a blackish bar across the interorbital region; dorsal and anal fins with a blackish streak near the edge, which is white. Total length 75 millim. Two specimens from Sumbu. Native name: Atlomo. This species is very closely allied to 7. marginatum Blgy., from which it differs in the longer and much more numerous gill-rakers. Dr. Pellegrin, in his Monograph of the Cichlide, p. 331, has expressed his doubts as to the distinctness of 7. unimaculatum from T. marginatum. A renewed examination of the specimens in the British Museum has convinced me that these doubts are not justified. 7. wnimaculatwm differs from T. marginatum in having a smaller eye, a narrower interorbital space, a strikingly longer caudal peduncle, much stronger dorsal and anal spines, more numerous gill-rakers (17 or 18 instead of 10 to 16), and fewer soft rays in the anal fin (7 or 8 instead of 10). Considering that I have before me three specimens of 7’. marginatum and ten of TL. unimaculatum, I feel confident that such differences, occurring in combination, cannot be merely individual; and the fact that the largest specimen of the former species measures 63 millimetres (without caudal fin) and the smallest of the latter only 70, proves that they cannot be ascribed to age, as suggested by Dr. Pellegrin. The specimens are sufficiently well preserved to show that the second or lower lateral line is totally absent; its representation on Plate XVIII. fig. 3 of my third Report in these ‘ Transactions’ is an error of the artist, which I unfortunately overlooked. 65. TROPHEUS MooRI! Bler. A single specimen from Sumbu. Native name: Kifuta. Dr. Cunnington describes its colour when fresh as: “ Dark reddish-brown, with OF FISHES FROM LAKE TANGANYIKA. 571 some suggestion of green, especially posteriorly; scales in dorsal region outlined with darker; dorsal and anal fins dull red, tipped with black; ventrals dark but reddish.” 66. SmocHRomis DIAGRAMMA Gthr. Numerous specimens, measuring from 75 to 185 millim., from Niamkolo, Sumbu, Vua, and Pembe. Native names: Impuma, Mbarama, Modi, Tempendi, Kikalakata. A female, measuring 92 millim., carries a few embryos in the mouth, these measuring 10 millim. and the vitelline sac 34, Dr. Cunnington has taken the following notes on fresh specimens :—* Olive above, lighter below, but with the scales in the dorsal region darker on the overlap; some 9 or 10 rather faint vertical darker stripes on the body; dorsal with brownish spots, tipped black ; anal with dark red border and two red spots posteriorly ; caudal darker in the posterior third.” Other specimens, of the male sex, are “‘ Greyish above, shading to lighter, almost white below; all scales with olive outlines; head dark above; pectoral and ventral fins with red upper margins; dorsal tipped red and black ; anterior and lower margin of anal bright red, tipped with black.” Females have the * Paired fins and the anal olive, the dorsal tipped yellow and black.” 67. Tivapia NiLotica L. A perfectly typical specimen from Lofu, measuring 310 millim., with dorsal XVII 12, anal III 10, and 23 gill-rakers on lower part of anterior arch. According to Dr. Cunnington, the fish grows to 390 millim. and occurs not only in Tanganyika itself but also in a small lake and in overflowed rivers near Kombe, on the east coast. Native name: Aiumbwi. T. nilotica, which has a wide distribution in Africa and Syria, had not previously been recorded from further south than Lake Kivu, where specimens were collected by Mr. Moore. 68. Truapra BuRTONI Gthr. Several specimens from Kituta and Kakonde River, Mtondwe Bay. Native names: Fwili and Sidia. 69. TILAPIA LATIFRONS, sp. n. (Plate XL.) Depth of body barely twice in total length, length of head 35 times. Profile of snout descending in a steep, straight line; eye twice in length of snout, 5 times in length of head, twice and + in interorbital width ; mouth extending to between nostril and eye; outer teeth strong, bicuspid, followed, after an interspace, by small tricuspid teeth forming 5 series in the upper jaw and 4 in the lower; 3 series of scales on the cheek; large scales on the opercle. Géill-rakers short, 9 on lower part of anterior arch. Dorsal XVI 13; spines strong, increasing in length to the last, which measures + 4@2 572 MR. G. A. BOULENGER ON A COLLECTION length of head; median soft rays produced, longer than head. Anal IIT 11; spines stronger but shorter than dorsals, third 3 length of head; median soft rays produced, like the dorsals. Pectoral acutely pointed, as long as head, not reaching origin of anal. Ventral reaching vent. Caudal peduncle deeper than long. Caudal truncate, oe lateral lines = Olive-green above, the scales outlined with black, bluish green to white beneath; cheek and opercle mottled green and blue, with some black; a black opercular spot; dorsal black, with regular large round blue spots, towards the edge turning to yellow with white spots; anal and caudal similarly coloured, but the spots less distinct; pectorals brown, ventrals grey. Total length 260 millim. A single specimen from Kituta. Native name: Ntunka. This fine Tilapia is allied to T. melanopleura A. Dum. (lata Gthr.), but the inter- orbital space is considerably broader, there are 11 soft rays in the anal instead of 9 or 10, and the markings on the fins are different. slightly emarginate. Scales finely denticulate, 32 70. Trnapta Hori Gthr. Two adult specimens, measuring 140 millim., from Niamkolo, native name Songa Matete, and one young, measuring 75 millim., from Sumbu, native name Kilomo. Two of these specimens have 8 soft rays in the anal fin. I no longer regard 7. rubro- punctata Bley. as separable from 7 horit. 71. TILaPia DARDENNII Bler. Five specimens, measuring 175 to 240 millim., from Niamkolo, Native names: Sangant, Nkungula, Masangano. Females have the mouth and pharynx filled with eggs, 5 or 6 millim. in diameter. 72. TILAPIA LABIATA Bler. Three specimens, measuring 330, 190, and 95 millim. respectively; two from Niamkolo and one from the Kakonde River, running into Mtondwe Bay. Native name: Kawbo or Dafwa. All three specimens have the produced lips characteristic of the species. Dorsal XVII 10-11; Anal IIL 7; 16 to 20 gill-rakers on lower part of anterior arch. Dorsal and anal with round reddish-yellow spots between the soft rays; a yellow spot on the middle of the ventral. 73. TILAPIA MICROLEPIS Bler. ‘wo specimens, measuring 285 and 265 millim. respectively, from Niamkolo, Native name: Aupi or Kuhi. OF FISHES FROM LAKE TANGANYIKA. 573 Dr. Cunnington measured a specimen 610 millim. long, and notes that the fish is said to grow still larger, 74, TILAPIA GRANDOCULIS Bley. Three specimens from Niamkolo, native name Weswa, and one from Sumbu, native name Kisangi. Dorsal XIII-XIV 13-14; Anal III 9; Scales 58-65 7"; Lateral lines"; 14 to 16 gill-rakers on lower part of anterior arch. The ventrals do not extend beyond the origin of the anal. Dr. Cunnington describes the coloration as dark olive-brown above, shading to lighter, with some yellow, below; a series of about five irregular blackish blotches on the side of the body; dorsal with yellow spots, tipped black; pectoral slightly, ventral strongly tinged with yellow; anal with yellow markings; proximal half of caudal darker than the rest. 75, PETROCHROMIS POLYODON Bley. Two specimens from Niamkolo and one from Pembe. Native names: IJnfutw and Llumbu. In the largest specimen, 170 millim. long, the depth of the body equals the length of the head, the diameter of the eye is contained nearly 5 times in the length of the head and once and 2 in the interorbital width. ‘There are 6 series of scales on the cheek, and 19 spines in the dorsal fin. Its colour when fresh is noted as ‘‘ Greenish brown over most of the body, but a little lighter below, with a patch of rather darker colour at the edge of each scale.” A smaller specimen is described as “ Olive above, shading off to lighter and greyish below, scales outlined by rather darker shade; dorsal fin with reddish tinge in anterior region and tipped black; anal with several prominent spots of bright yellow; distal half of ventrals slightly reddish.” 76. PETROCHROMIS TANGANICEZ Gthr. One specimen from Niamkolo and one from Kituta. Native name: Atlongo. CUNNINGTONIA, g. n. Jaws with broad bands of minute, bristle-like teeth with incurved tricuspid crowns. Maxillary largely exposed at the end. Body moderately elongate, covered with small, feebly ctenoid scales; two lateral lines, the upper complete. Dorsal with 13 spines, anal with 5. Vertebre 36 (18+18). Agrees with Petrochromis in the dentition, differs in the maxillary bone being largely exposed when the mouth is closed, 574 Mk. G. A. BOULENGER ON A COLLECTION 77. CUNNINGTONIA LONGIVENTRALIS, sp. n. (Plate XLI. fig. 2.) Depth of body equal to length of head, 3 times in total length. Snout broad, very convex, shorter than the eye, which is 25 times in length of head and equals inter- ocular width; maxillary extending to below anterior border of eye; 2 or 3 series of small scales on the cheek; large scales on the opercle. Gill-rakers rather short, 13 on lower part of anterior arch. Dorsal XIII 13; spines weak, subequal from the 7th, nearly $ length of head; posterior soft rays produced, ? or $ length of head. Anal III 8-9; spines weak, 5rd 4 length of head; soft rays produced like the dorsals. Pectoral acutely pointed, as long as head, reaching beyond origin of anal. Ventral produced into a veiy long filament, nearly twice as long as head. Caudal peduncle once and } as long as deep. Caudal with deep crescentic notch. Scales 38 a lateral lines ae Total length 110 millim. Two specimens from Niamkolo. Native name: Ka/lilolilo. This fish bears a striking resemblance to Tilapia boops Blgr., in which, however, the dentition is very different and the maxillary bone is almost concealed when the mouth is closed, and the vertebre are fewer (16 + 16). Bluish black above, bright silvery beneath; fins blackish. 78. Erermopus cyanostictus Bler. Two specimens from Niamkolo Bay and one from Sumbu. Native name: Kamfoto. Dr. Cunnington describes the colour as dark brownish grey with a number of scattered spots of a brilliant blue. The specimen from Sumbu, a female measuring 75 miilim., contained developing embryos in the mouth; some of them dropped out before the specimen could be preserved. 79. XENOCHROMIS HECQUI Bler. A single specimen from Moliro. Confounded by the natives with the following species under the name of Komangingi. 80. PLEcopuS PaRADoxuS Bler. Two specimens from Kasakalawe (native name NVkanka), two from Niamkolo (n. n. Nswepa), one from Sumbu and one from Moliro (n. n. Komangingi). As this remarkable fish was only known from young and defective specimens, a new description from the material now at hand seems desirable. Depth of body nearly equal to length of head, 34 to 4 times in total length. Snout short and convex, as long as the eye in the adult, shorter in the young; diameter of eye 3 to 33 times in length of head, once to once and + in interorbital width; mouth OF FISHES FROM LAKE TANGANYIKA. 575 extending to below anterior fourth of eye; jaws strong, lips thick; 18 or 20 teeth in the upper jaw. 14 or 16 in the lower; interocular region scaly, convex; 3 or 4 series of scales on the cheek; larger scales on the opercle. 18 to 20 gill-rakers on lower part of anterior arch, the posterior rather elongate. Dorsal XIX-XX 11-13; spines rather strong, subequal from the 8th or 9th, 2 to $ length of head ; posterior soft rays longest, $ length of head. Anal III 12-13; spines very strong, third longest, 2 to 4 length of head. Pectoral acutely pointed, ? to # length of head. Ventral not reaching anal. Caudal peduncle once and 3 to twice as long as deep. Caudal with deep 58-66 , s ld a 5-6 ° ing : crescentic notch. Scales 75-78 93 lateral lines [-,; 9 series of scales between the lateral lines anteriorly. Olive-grey above, in the fresh condition, shading to whitish below, with 7 or 8 somewhat irregular, yellowish-brown longitudinal lines; two or three ill-defined dark brown bars across the back and a round spot at the root of the caudal fin sometimes present; a black opercular spot; dorsal and caudal fins brown, with round yellow spots. Total length 290 millim. The vertebree number 38 (18+20). The occipital crest is very strong. MASTACEMBELIDA. S1. MastacEMBELUS Moort Bler. Five young specimens, 40 to 70 millim. long, from Utinta, dredged among shells in about 15 fathoms. Native name: Molombo. The dorsal spines number 26 or 27. 82. MASTACEMBELUS CUNNINGTONI, sp. n. (Plate XLI. fig. 3.) Depth of body 10 to 11 times in total length, length of head 7 to 74 times. Vent equally distant from end of snout and from root of caudal fin, its distance from the head equal to 24 to 23 times length of latter. Snout 3 to 4 times as long as eye, ending in a trifid appendage which is a little longer than the latter; cleft of mouth extending to below nostril or not quite so far; a strong, erectile spine below the nostril ; two strong preopercular spines. Dorsal X XIX 95-105, spines increasing in length; anal II 115-125, second spine equal to penultimate dorsal; both fins completely embracing the rounded caudal; the distance between the first dorsal spine and the head 3 to 34 times in length of latter. Pectoral twice and } to 3 times in length of head, Scales very small, 26 to 28 between origin of soft dorsal and lateral line. Uniform brown aboye, or olive-brown mottled with darker, white beneath; caudal blackish, edged with orange. Total length 580 millim. Two specimens from Kituta, native name Morobe, and one from Kazzaga, native name Molombo. 576 ON A COLLECTION OF FISHES FROM LAKE TANGANYIKA. Closely allied to M. congicus Blgr.; distinguished by the fin-formula and the coloration. 83. MASTACEMBELUS T&NIATUS Bler. Two young specimens, 75 and 38 millim. long, one from Niamkolo Bay (native name Morodi), the other from Kasawa (native name 1e Molombo). The dorsal spines number 33 or 35. 84. MASTACEMBELUS OPHIDIUM Gthr. A single specimen, measuring 330 millim., from Burton Gulf. Native name: Molombo. Length of head 8} times in total length. 33 dorsal spines. Pale olive-brown above, with roundish spots of a dark brown; white beneath; dorsal and pectorals spotted with brown; anal and caudal of a light orange colour, unspotted. We are now acquainted with 7 species of Mastacembelus from Lake Tanganyika, They may be distinguished as follows :— I. Vent equally distant from end of snout and from root of caudal fin. A. No preopercular spine. Dorsal XVIII 85; AnalJI90... . ee en cay ene enaLusH Bor! Dorsal XX V-XXVITI 70-80; Anal II 70- 80 So 8 6 go op np a AUG OeRD Isis, B. Two preeopercular spines. Dorsal XOXeEXs 95-105), 7 Amal N15 1125 at een en 7 cunningtoni Blgr. II. Vent nearer caudal fin than end of snout. A. One or two preopercular spines. Dorsal XXXI 70; AnalIIT70 . . . oc ete Minellipstteribion: Dorsal XXXVI [-XXXIX 50-60; Anal Ul 50- 60 . see ew ee 4 «6M. tanganice Gthr. B. No preopercular spine. Dorsal XXX MIX OXNV 85); Anal 11 85) = ee M. teniatus Blegr. III. Vent nearer end of snout than caudal. Dorsal XXX-XXXIII 100-120; AnalI 115-180 ... . - « + « WM, ophidium, Gthr. PAU LOG a xvi—parr vi. No. 6.— October, 1906. FISHES FROM LAKE TANGANYIKA. PLATE XXX. Fig. 1. Mareusenius tanganicanus, p. 545, . Pellonula miodon, p. 546. © bo = nat. size. . Alestes rhodopleura, p. 547. 4 nat. size. ©, XEN Gi Framscool, ob VC XVIL Bale & Danielseon, Lt? imp. ith Green del. et J 3. ALESTES RHODOPLEURA. 2.PELLONULA MIODON. 1. MARCUSENIUS TANGANICANUS. PLATE XXX. 580 FISHES FROM LAKE TANGANYIKA. PLATE XXXI. Dinotopterus cunningtoni, p. 550. a. Adult, $ nat. size; 6. Upper view of head, } nat. size; ¢. Dentition, natural size ; d. Young, natural size. Tramiel Soc Ul XV F¢-XXKL Bale & Danielsson, iia imp CUNNINGTONI. DINOTOPTERUS DIGI NOOTTG oO bo FISHES FROM LAKE TANGANYIKA. PLATE XXXII. Fig. 1. Neobola minuta, p. 549. Enlarged. . Chrysichthys sianenna, p. 551. . Phyllonemus typus, p. 552. . Haplochilus pumilus, p. 554. bo eo ‘SNTINAd SNIIHOOIdVH } SNdAL SNNHNOTIAHA € VNNANVIS SAHLHSISAMHD 2 VLONIUN VIOROUN T | Qa ‘jap wae 32 “Jap Weer “+f ‘durr w ak aoss]arue (| Y ajeq N 584 FISHES FROM TAKE TANGANYIKA. PLATE XXXIII. Auchenoglanis occidentalis, var. tanganicanus, p. 558, with upper view of head and open mouth. 'SONVOINVONVL YA ‘STTIVLNHCIOD0 SINVTDONEHONY = =| a) ay 4 masin ¢ dan ry OSS eeeChy ae aa SE repre irtuctrrmeenaeny newer 9522s avi ee SIA SN ef Seopine Ph DOOM GE =e No. 7.—October, 1906. 586 FISHES FROM LAKE TANGANYIKA. PLATE XXXIV. Synodontis melanostictus, p. 553, with upper and lower views of head. % natural size. ‘SOLOILSONVTIHIN SILNOGCONAS ‘det ps Siete ® aeq Wall 34 Yep wees) AIXXX Ob? TAK 72 °G GOES FISHES FROM LAKE TANGANYIKA. PLATE XXXYV. Fig. 1. Lates angustifrons, p. 555. ia? #4 5 upper view of head. 2 » microlepis, p. 555, upper view of head. 3. Lamprologus multifasciatus, p. 558. "3 reticulatus, p. 580. F calliurus, p. 960. SHaMiilivo LS SOIVTN LG wa ‘SALVIOSVEILINN SNOOIOMUNVI € “SIGHTOUDIN 12 SNOUAILSNONVY SHLVT 1 Titeaa pal TosstsTaed R ae Es ANIX AG HAN 720° 0? oOo PLATE XXXVL_ FISHES FROM LAKE TANGANYIKA. PLATE XXXVI. Fig. 1. Lamprologus brevianalis, p. 555. = oo SO cunningtoni, p. 557. mondabu, p. 057. callipterus, p. 559. SMa ici. Az “NEVGNOW ‘T€& INOLONINNOO Te SITVNVIAGHE SNOOIOXdNVT T dot py) mosayerme(y RwaAeq - Ti] 3 ysp waeig-r MANE AB MAX 720 YJ ?CO OP" 2 or (tap) bo FISHES FROM LAKE TANGANYIKA, PLATE XXXVII. Fig. 1. Paratilapia frontosa, p. 561. 1a. Dentition. 2. Pelmatochromis polylepis, p. 566, young. SIGdHTIATOd SINOMHIOLVNIEd ¢ VSOLNOUA VIdVIILVUdVd 1 dan 2a] wosspmeq 9 eg pz Ll 4! tata gah? h S,ae D HAXEX tb HAX 720° 7°S6 "Lo PA Maya * i ' pant vi. No. 8- FISHES FROM LAKE TANGANYIKA, PLATE XXXVIII. Fig. 1. Bathybates minor, p. 564. 2. Haplotaxodon microlepis, p. 566. WO XVI PY XXXVUL Dd ans.Lovl, Fo iy F; p. t Bale & Danielsson, L*? im 3 2. HAPLOTAXODON MICROLEPIS. MINOR. BATHYBATES ] PLA TE 596 FISHES FROM LAKE TANGANYIKA. PLATE XXXIX. Fig. 1. Pelmatochromis melanostigma, p. 567. Mp ie pleurospilus, p. 567. oe - rhodostigma, p. 568. 3a. ae ie lower surface of head and anterior part of body. A, 3 auritus, p. 668, young. “SOLINAV dy VNDILSOUOHY 2e SNTIASOUNATd d2 VNDILSONVTAN SINOUHIOLVINTSd UT oOuno.my 1) [?P uses) “(a ae oo 7 NIU IE UN 900? PLEO 598 FISHES FROM LAKE TANGANYIKA. PLATE XL. Tilapia latifrons, p. 571. SNO@AILVI VidvVTiL TRE UNO HR 7 OL PLATE XL 600 FISHES FROM LAKE TANGANYIKA. PLATE XLL. Fig. 1. Trematocara nigrifr ons, p. 570. 2. Cunningtonia longiventralis, p. 574. 3. Mastacembelus cunningtoni, p. 575, 3 nat. size. 3 a. Ditto: upper view of head. 3 nat. size. INOLONINNOAD SOTHAENADVISVN € SITIVHULNSAISNOT VINOLONINNND ¢@ SNOWMIGOIN VEVOOLVNGaL 7 dor 7 ‘wossyarme (| R seg ay 33 Tap sp VS al $4 < &) [TE AE MAN G26 7D 7° OL, ussac 601 LIST OF THE PAPERS CONTAINED IN VOL. XVII. Page | Page Beravye-Baxrr, Groree T., F.LS., F.Z.8., &e. Fawcert, Lieut.-Colonel J. Matcoum. A Revision of the Amblypodia Group of Notes on the Transformations of some Butterflies of the Family Lycenidee 1 South-African Lepidoptera .......... 165 Bovutencer, Guorce Arsert, F.R.S., V.P.Z.S. A Contribution to our Knowledge of the Mircnenr, P. Caatmers, M.A., D.Se., Secretary Varieties of the Wall-Lizard (Lacerta to the Society. muralis) in Western Europe and North On the Intestinal Tract of Mammals .... 437 INGA ox on Gatoce Olde Dob DOCS sa ell | Fourth Contribution to the Ichthyology of | Broan, C. Tare, B.A., F.Z.S. Lake Tanganyika.—Report on the Col- | A Monograph of the Fishes of the Family lection of Fishes made by Dr. W. A. Teor iccrti die eax au aietee cts ea Keone 191 Cunnington during the Third Tanganyika Expedition, 1904-1905 ............ 537 VOL. XV1I.— PART VI. No. 9.— October, 19 06. 4L INDEX OF SPECIES, ETC., IN VOL. XVII. Acanthicus, 261. genibarbis, 219. hystrix, 262. Acesina, 25. aberrans, 144. ariel, 145. arisba, 144. —— paraganesa, 145. -— zephyretta, 145. Aecestra, 302. acus, 304. —— amazonum, 305. — gladiolus, 305. ——- gladius, 303. — knerii, 304. oxyrhynchus, 304. Africa, on transformations of Lepidoptera of South, 165. , on the varieties of the Wall-Lizard of Northern, 351, Alactaga indica, 482. Alestes, 543. batesti, 548. brevis, 548. grandisquamis, 548. macrolepidotus, 548. macrophthalmus, 540, 547. rhodoplewra, 540, 547, 578. Amblypodia group of the Lycenide, revision of the, 1. Amblypodia, 1, 15, 25. abseus, 112. acetes, 45. —— achelous, 82. —— uacron, 140. adatha, 69. admete, 141. Amblypodia adonias, 102. — edias, 124. @xone, 3d. agaba, 83. agelastus, 136. —— agesias, 95. agesilaus, 93. —— agluis, 44. —— agnis, 125. — alaconia, 101. — albopunctata, 60. — alce, 75, 76. -— alea, 133. —— aliteus, 61. —— allata, 128. —— ainantes, 42. —— ameria, 22. —— amisena, 6. anmon, 143. amphimuta, 90. anarte, 81, 125. andersonti, 17. anita, 18. annetta, 19. anna, 19. anmella, 114. annulata, 52. anthelus, 77, 78. anthore, 58. antimuta, 98. anunda, 76. apidanus, 115. aracves, 43. areste, 119. aricia, 87. —— «roa, 85. saturatior, 76, 77. 604 Amblypodia aronya, 57. arracana, 18. arvina, 147. asopid, 138. ate, 29. atkinsoni, 21. —— atosia, 87. — atraz, 180. aurea, 106, —— auxesia, 80. —— awiothea, 153. —— azenia, 140. —— bazaloides, 55. bazalus, 119. bieolora, 53. bupola, 104. —— buxtoni, 60. — ceca, 149. —— camdeo, 46. ——- canula, 149. —— capeta, 113. —— centaurus, 39. —— chinensis, 118. ——- corinda, 121. — courvoisiert, 152. —— critala, 152. —— cryonthe, 34. —— eupido, 151. — darana, 18. —— detrita, 123. -—— diardi, 113. —— disparalis, 152. —— dodonea, 184, —— elfeta, 48. —— epimete, 141. —— epimuta, 88, 89. — erebina, 52. —— eridanus, 49. dilutior, 49. —— eucolpis, 141. —— eumolphus, 104. eupolis, 41. fuisina, 19. farquhari, 104. —-— fruhstorferi, 80. fulgida, 114. fulla, 150, INDEX. Amblypodia ganesa, 146. grynea, 102. hainana, 20, — helius, 58. anthelius, 58. hellenore, 104. hercules, 28. hesba, 95. horsfieldi, 107. —— hypomuta, 99. inornata, 91. kuhni, 113. — lycenaria, 60. —— meander, 32. —— metamuta, 100. —— micale, 36. —— minnetta, 60. —— muta, 101. — myrtale, 63. —— myrtha, 61. —— myrzala, 143. — nakula, 39. —— narada, 16, 162. —— -— erichsonii, 18. —— naradoides, 18. —— nobilis, 74, 75. —— oberthuri, 101. —-— cenea, 130. —— enotria, 124, — olinda, 60. —— ovomaculata, 95. —— padus, 46. palowna, 6. paraganesa, 144. perimuta, 139. — phrywus, 58. polita, 49. pryeri, 86. pseudo-centaurus, 39. pseudomuta, 87. querceti, 130. quercetorum, 4, 5. —— quercoides, 68. —— rana, 130. —— ribbei, 34. rochana, 12. selta, 84. Amblypodia silhetensis, 72. sophax, 58. —— sophrosyne, 36. strophe, 153. superba, 56. —— tamyrus, 58. —— taoouna, 16, 17. tephlis, 53, 5d. —— theha, 56. —— timoleon, 9. —— tounguva, 138. —— tristis, 52. viardi, 113. vihara, 66. —— viola, 46. —— violacea, 130. vivarna, 5, 6. vivianda, 61. —— yendava, 125. —— zeta, 135. (Narathura) fulla andamanica, 151. Ambulyx postica, 165, 174, 188. Amphilius platychir, 540. Ancistrus, 221, 251. aliernans, 229. angulicauda, 241. —— annectens, 225, 330. —— aspidolepis, 230. —— aurantiacus, 226. bachi, 233. barbatus, 240. —— brachyurus, 231. —— calamita, 258. —— chagresi, 26. — cirrhosus, 256. —— depressus, 240, —— dolichopterus, 254. —- duodecimalis, 229. —— etentaculatus, 226. — gibbiceps, 227. gigas, 230. guacharote, 238. —— quentheri, 241, 332. gymnorhynchus, 254. heteracanthus, 237. hoplogenys, 255. leucostictus, 255. INDEX. 605 Ancistrus lituratus, 227. longimanus, 226. medians, 229. megacephalus, 234. multiradiatus, 228. —— mypstacinus, 238. —— niveatus, 235. —— oligospilus, 232, 332. parahybe, 241. pictus, 231, 237. platycephalus, 235, 332. —— punetatissimus, 236. punctatus, 229. scaphirhynchus, 231. schomburghii, 233. setosus, 239, 332. —— stigmaticus, 259. —— temminckii, 254. — undecimalis, 226. —— vittatus, 232. Antherea belina, 169. Anthropopithecus niger, 518. Apporasa, 20. atkinsoni, 21. Arctitis binturong, 500, 501. Arges, 307. boulengeri, 310. —— brachycephalus, 313, 350. chote, 313, 350. cyclopum, 311. —— eigenmanni, 312, 350. —— fest, 316, 350. Jfissidens, 316. —— guentheri, 310, 350. homodon, 309, 350. — longifilis, 314. marmoratus, 314, 350. orientalis, 313, 350. peruanus, 317. prenadilla, 316. sabalo, 315. simonsti, 317, 350. taczanowskii, B15. vaillanti, 312. whymperi, 311, 316. Arhopala, 1, 25. aberrans, 144. 606 Arhopala aboe, 147. abseus, 112. ace, 123. acestes, 121. acetes, 45, 160. —— achelous, 82. — acron, 140, 160. —— adala, 148. —— adatha, 30, 36, 69. —- adherbal, 32. —— admete, 141. waigoensis, 141, 160. — adonias, 102, 158. —— adorea, 67. —— adriana, 117, 164. —— adulans, 148. —— cdias, 124, 160. —— wwxone, 35, 160, —— agaba, 83. ——- agelastus, 89, 136. —— agesias, 95, 160, 164. —— ayesilaus, 64, 93, 158, 164. —— aglais, 44. —— agnis, 82, 125. —— agrata, 123. —— aida, 64. —— alaconia, 101. —- albopunctata, 60. — alee, 76, 164. —— alcestes, 75. —- alea, 133. —— alemon, 133, 164. —— alesia, 102. —— aliteus, 61, 162. —— allata, 128, 160. alpha, 69. amnantes, 42, 162. anatriv, 42. amazona, 40. ammon, 143. ammonides, 143, 160. amphea, 112. amphimuta, 90, 158. amytis, 33, 34, 162. anamuta, 128, 169. anarte, 81, andamanica, 151, 162. INDEX. Arhopala anella, 110, anila, 94. —— anniella, 114. annulata, 52, 162. — antharita, 78. anthelus, 78. —— anthore, 58, 162. —— antimuta, 88, 98. —— anunda, 76, 160, 164. apidanus, 115, 164. palawanus, 116. —— appianus, 32, —— arama, 72. — araxes, 43. —— arca, 116. — areste, 119. —— argentea, 57. — aricia, 88. —— ariel, 145, 160. —— aroa, 85, 86, 164. —— aronya, 57. — artegal, 1165. —— arvina, 147. —— arzenius, 66. — asia, 92. — asoka, 118. —— asopia, 138, 160. — aie, 29. — athara, 76. — atosia, 87, 88, 164. —— atrav, 130, 160. ~ —— aurea, 106. —— auvesia, 80, 160. —— auzea, 80. —— avatha, 96. —— axiothea, 153. azata, 48. —— azenia, 140. azinis, 122. barami, 126. basiviridis, 108, 158, 164. bazaloides, 55, 156, 162. —— bazalus, 55, 119. bella, 122. belphecbe, 142. birmana, 144. borneensis, 107. Arhopala brahma, 99, 158. brooket, 84, 156. buddha, 148, 160. buruensis, 122. —— buwxtoni, 60. — ceca, 149. camdeo, 46. canaraica, 133. canulia, 149, 160. carolina, 50. catori, 92, 93, 158, 164. centaurus, 39, 162. coruscans, 39. chinensis, 118, 160, 164. clavissa, 57. comica, 134. constancee, 33. corinda, 102, 121, 160. coruscans, 40. critala, 152. eryonthe, 34. dajagaka, 127. davaona, 96, 158. davisoni, 88, 98. detrita, 123. deva, 98, 164. diardi, 113, 164. disparalis, 152. dodonea, 134, 164. dohertyi, 64, 156, 162. drucei, 69. duessa, 143. elfeta, 48. elopura, 86, 164. epimete, 141, 160. epimuta, 88, 158, 164. eumolphus, 104, 158, 164. ewpolis, 41, 156. eurisus, dl. farquhari, 104, 105. fruhstorfert, 80. — fulgida, 114. —— fulla, 150. ganesa, 146, 162. loomisi, 146. LUCA ee seminigra, 145. eridanus, 47, 49, 156, 162. INDEX. Arhopala gunongensis, 97, 158. havilandi, 82. helianthes, 58. helius, 58, 156. anthelius, 58, 156. — hellenore, 104, 105. hercules, 28, 162. herculina, 28, 156. leo, 28. tyrannus, 29. —— hesba, 95, 160. — hewitsoni, 132, 160, 164. —— hylander, 31. —— hypomuta, 91, 98, 99, 158. ijauensis, 80, 156. —— inornata, 91, 158. —— irregularis, 111, 158. —— japonica, 134, 160, 162. khamti, 129. — kiriwinii, 38, 156. —— kounga, 122. —— kuhni, 118, 158. kurz, 92. — labuana, 65. leo, 28. loomisi, 146. —— lycenaria, 60. malayica, 71, 158. meander, 31, 32, 162. —— metamuta, 97, 100, 164. micale, 36. mindanensis, 62, 155. minnetia, 60. mirabella, 61. moeller7, 118. moolaiana, 89, 1538. moorei, 97, 164. morphina, 117, 158. muta, 101, 164. myrtale, 63, 156. myrtha, G1, 156. myrzala, 143. nakula, 39. nicévillei, 73, 158. nobilis, 74, 158. oberthuri, 101, 158. OAS horsfieldi, 106, 107, 108, 153. 607 608 Arhopala ocrida, 102. eeta, 139. —— ened, 130, 164. —— enotria, 124, 160. —— olinda, 60. —— opalina, 51. ——- ovomaculata, 95. padus, 46, 156, 162. paraganesa, 144, 162. paramuta, 1385, 162. —— pastorella, 88, 89. periander, 30. perimuta, 139. perissa, 137, 160. phenops, 68, 156. philander, 30, 162. phryeus, 58. pirama, 40, pirithous, 40. pryert, 86. pseudocentaurus, 40, pseudomuta, 86, 87, 156. —— quercoides, 68, 156. —— rafflesii, 86. —— rama, 130, 164. —— roona, 186. —— sandakani, 68. —— sangira, 57, 156, —— sceva, 85. -—— sella, 84. —— semper, 126. —— silhetensis, 72. —— similis, 94. —— singla, 120, 164. —— sophrosyne, 36. —— staudingeri, 109, 158. subfasciata, 80. —— superba, 36. tameanga, 126. teesta, 119. —— tephlis, 53, 156, 162. — theba, 56, 162. tounguva, 138, 160. —— triangularis, 114. —— trogon, 107. turbata, 120. —— tyrannus, 28, INDEX. Arhopala vihara, 66, 158. viola, 51, 156, 160. v-major, 93. waterstradti, 97. —— wildei, 151, 156. wimberleyi, 102. yendava, 125. zephyretta, 145. zeta, 135. (Mahathala) ameria, 22. Asprotilapia leptura, 542, 565. Astroblepus, 318. griwalve, 318. Ateles variegatus, 509, 510. Atherura africana, 484. Auchenoglanis biscutatus, 555. occidentalis, 540, 553. tanganicanus, 553, 584. Babirussa babirussa, 469, 471, 519, 520. Baker, G. T. Bethune-, see Bethune-Baker. Barbus platyrhinus, 540. serrifer, 540, 549, tropidolepis, 540, 549. Barilius moorit, 540, 549. tanganice, 540. Basiana postica, 174. Bassaris astuta, 496. Bathybates fasciatus, 541, 564, 565. ferox, 541, 564, 565. minor, 541, 564, 565, 594. Bethune-Baker, G. T., Revision of the Amblypodra Group of the Lycenide, 1. Bettongia cuniculus, 449, 451. penicillata, 449, 450, 451. Bombyx astreas, 180. bauri, 180. — bellatrix, 182. delegorguei, 172. meteus, 178. patens, 173. vaillantina, 182. Boulenger, G. A., on the Varieties of the Wall- Lizard (Lacerta muralis) in Western Europe and North Africa, 351; on the Fishes collected by Dr. W. A. Cunnington during the Third Tanganyika Expedition, 537. Bradypus infuscatus, 455, 456, 517. Brontes, 307. prenadilla, 313, 316. Bunea tyrrhena, 171, 186. Butterflies, revision of the Amblypodit group of, 1. Byblia ilithyia, 166, 186. Callioratis bellatrix, 182, 190. Callithrix cuprea, 509, 520, 521. Camelus dromedarius, 473. Canis familiaris, 498. vulpes, 497, 498. Capoéta tanganice, 549. Capromys pilorides, 487. Castor fiber, 482, 517. Cavia porcellus, 484, 485. Cebus fatuellus, 509, 510, 520, 521, Centetes ecaudatus, 491, 492. Cercoleptes caudivolvulus, 495, 496, Cercopithecus niger, 511. stairsi, 511. werneri, 511, 512. Cervus axis, 474. Cherocampa eson, 175, 190. orpheus, 165, 176, 188. osiris, 176, Chetostoma, 244. Cheetostomus, 222, 244, 251, 259. aculeatus, 230. —_ alga, 255. —— anomalus, 250, 334. —— aspidolepis, 230. —— aurantiacus, 236. — bachi, 233. brachyurus, 231. branickit, 250.' brevis, 247, 334. calamita, 258. cirrhosus, 256, 257. cochliodon, 243. —— dentex, 243. —— depressus, 240. dermorhynchus, 248, dolichopterus, 254, Jischeri, 248, fordii, 261. gibbiceps, 227. —— gibbosus, 243. — gigas, 230, INDEX. Chetostomus guacharote, 238. —-~ guairensis, 249, gymnorhynchus, 254. heteracanthus, 237. hoplogenys, 255, —— hystrix, 261. itacua, 234, jelskti, 258. —— latifrons, 253. leucostictus, 255. —— lituratus, 227. loborhynchus, 246. —— macrops, 234. maculatus, 246, 336, malacops, 255. marcapate, 246, 336. marginatus, 249, 334. medians, 229, megacephalus, 234. microps, 247, 249, 251, 336. mystacinus, 238. nigricans, 236. nigrolineatus, 244, Regan niveatus, 235. —— nudiceps, 254. nudirostris, 251. oligospilus, 232. pictus, 231. platycephalus, 235. punctatissimus, 236. —— punctatus, 229. scaphirhynchus, 231. sericeus, 252, serratus, 260. setosus, 239. spinosus, 260. stannit, 248. taczanowskti, 247. tectirostris, 255. thomsoni, 250, 336. — trinitatis, 238. —— undecimalis, 226, —— variolus, 256. vittatus, 232, Chalethiops, 550. Charaxes vuranes, 168, 186. Chirogaleus cocquereli, 503, 505, VOL. XV.—PartT vi. No. 10.—Octoder, 1906. 4M 609 610 INDEX. Ohiromys, 536. Chromis paterfamilias, 538. (Tilapia) philander, 538. Chrysichthys brachynema, 540, 551. cranchit, 540. —— furcatus, 552. myriodon, 540. sianena, 540, 551, 582. Cirina forda, 170, 186. Oitharinus gibbosus, 540, 548. Olarias liocephalus, 540. robecchii, 540, 550. Cochliodon, 242. cochliodon, 242. Copaxa flavinata, 170, 172, 186. Crenis boisduvalt, 167, 186. natalensis, 167, 186. Cricetus frumentarius, 482, 483, Cryptoprocta ferow, 500, 501. Cunnington, W. A., on the Fishes collected during the Third Tanganyika Expedition by, 537. Cunningtonia, 573. longiventralis, 542, 574, 600. Cyclopium, 307. cyclopum, 312. guentheri, 310. humboldtii, 311. Cyclothurus didactylus, 516. Cynictis penicillata, 500, 501. Cynonyeteris collaris, 492. Daman, 5365. Darapsa marginata, 177. Darasana, 25. neward, 135. perimuta, 139. Dasypus minutus, 458, 459, 516, 518, 519. seacinctus, 516. -— villosus, 458, 459, 516. Dasyurus viverrinus, 446, 447. Deltwrus, 222. angulicauda, 241, parahybe, 241. Deudorix mecenas, 11. tumoleon, 11, 12. Didelphys virginiana, 445. Dinotopterus, 550. cunningtoni, 540, 550, 580. Dipus egyptius, 482. Dolichotis patachonica, 484, 487. Dromicia nana, 449. Echidna, 442, 525. hystriv, 44-4. Ectodus cdescampsi, 541, 568. melanogenys, 569. Egybolis vaillantina, 182, 190. ilephas africanus, 466. Enantiopus, 569. longianalis, 537, 538, 541, 665, 569. melanogenys, 541, 569. Equus granti, 479, 480. Eretmodus cyanostictus, 537, 542, 574. Erinaceus algirus, 491. Estigmene atropunctata, 179, 188. —— dissimilis, 165, 179, 188. Euchloris devocata, 184, 190. Europe, on the varieties of the Wall-Lizard of Western, 351. Eurytela hiarbas, 166, 186. Farlowella, 302. acus, 304, amazonum, 305. carinata, 305, gladiola, 305, 348. gladius, 303. gracilis, 303, 848, knerit, 304. ovyrhynchus, 304. Faweett, Lt.-Col. J. M., on Transformations of South- African Lepidoptera, 165, Felis caracal, 501, 502. serval, 501. Fishes, monograph of the Loricariide family of, ik, of the Third Tanganyika Expedition, 537. Flos, 25, ahamus, 115, artegal, 115, Galago garnetti, 506, 507. Galeopithecus, 53d. Galictis barbara, 495. Gazella dorcas, 474. marica, 518. Geometra devocata, 184. Gephyrochromis moorit, 542. Giraffa camelopardalis, 474, 475. Gorilla savagei, 513, 514. Grammatotria lemairiz, 541, 565, 569. Gynanisa maia, 171. Hapale penicillata, 508. Haplochilus pumilus, 540, 548, 554, 582. tanganicanus, 540, 547, 554. Haplochromis desfontainesti, 539. ishmaeli, 539. Haplotaxodon, 565. microlepis, 541, 566, 594. Harttia, 297. loricariiformis, 298. Hemiancistrus, 221, 259. aspidolepis, 230. —— brachyurus, 231. fordii, 261. guacharote, 239. heteracanthus, 237. —— hystrix, 261. itacua, 234. longipinnis, 235. medians, 229. megacephalus, 234. mystacinus, 238. oligospilus, 232. —— pictus, 231. scaphirhynchus, 231. schomburgku, 233. serratus, 260. —— spinosus, 260. trinitatis, 239. vittatus, 232. Hemiloricaria, 270. caracasensis, 270. Hemiodon, 296. acipenserinus, 296. depressus, 296. platycephalus, 295. Hemiodontichthys, 296. acipenserinus, 296. depressus, 296. Hemnipsilichthys, 220. gobio, 221. Henucha delegorquei, 172, 186. smilax, 171, 186. Herpestes albicauda, 500, 501, INDEX. Herpestes griseus, 500, 501. ichneumon, 500, 501. Hesperia apidanus, 115. centaurus, 39. meecenas, 11, Heterobranchus, 550. Hisonotus, 265. levior, 268. leptochilus, 268. notatus, 268. Hyena crocuta, 498, 499. Hydrocherus capybara, 484, 486, 535. Hydrocyon lineatus, 540, 547. Hyloicus juniperi, 177, 188. Hypoptopoma, 263. bilobatum, 263, 265. carinatum, 264. —— guentheri, 264, 338. —— gulare, 268. joberti, 265. steindachnert, 265. thoracatum, 263, 265. Hypostomus, 202, 221, 244, 251, 259. alatus, 211. —— asperatus, 217. —— aurantiacus, 236. —— auroguitatus, 217. barbatus, 240, —— bufonius, 258. —— calamita, 258. — cirrhosus, 256. cochliodon, 242. commersonit, 206. — duodecimalis, 226. emarginatus, 210. erinaceus, 251, 252. etentaculatum, 226. granosus, 306. guacari, 205. guacharote, 238. —— guttatus, 240. ——- horridus, 210. itacua, 234, —— nigricans, 236. niveatus, 235. —— nuiliceps, 254, pantherinus, 209. 4m 2 611 612 Hypostomus pardalis, 228, pictus, 237. —— plecostomus, 205, 209. punctatus, 207. robinit, 205, 215. serratus, 260. spinosus, 260. squalinum, 210. subcarinatus, 207. temminckii, 254, vicinus, 236. Hyrax capensis, 460, 461, 516. Hystria cristata, 484. Ichthyology of Lake Tanganyika, 537. Lctonyx zorilla, 495. Indris brevicaudatus, 503. Intestinal Tract of Mammals, 437. Tots, 25, Traota, 1, 8, 16. boswelliana, 12. lazarena, 13, 156. —— mecenas, 10, 11. — nila, 14. — rochana, 12. timoleon, 9, 10, 11, 12, 162. Julidochromis ornatus, 541. Labeo cylindricus, 540, 543, 548. forskaliz, 548, Lacerta acrolampra, 403. agilis, 373, 389, 392. bedriage, 369, 410, 411, 413, 420. —— bruggemanni, 384. | depressa, 354, 359, 368, 369. | derjuginti, 880. faraglionensis, 394. | —— filfolensis, 401. genei, 377, 405, 413. horvathi, 358, 365, 366, 367, 378. lilfordi, 372, 373, 376. mosoriensis, 366. murdalis, 351-436, balearica, 369, 3870, 372, 373. pityusensis, 370, bedriage, 384, 404, 409, 411, 413, 414, 415, 424, 434, 436, bocaget, 361, 369, 415, 416, 420, 426, 428, 436. INDEX. Lacerta muralis breviceps, 376, 378, 379, 420, 428. brueggemanni, 376, 377, 380, 381, 383, 384, 386, 387, 388, 389, 394, 401, 404, 405, 420, 422, 424, campestris, 380, 388, 389, 390, 391, 392, 394, 395, 396, 397, 399, 404, 405, 420, 432, 434. — — cettii, 370. -—— —— chalybdea, 365, —— —— caerulea, 394, 396. —— —— ceruleo-cerulescens, 394, 396. cerulescens gallensis, 394, 396. monaconensis, 394, 396. —— —— concolor, 394. as RE, ALOT. depressa, 366. —— — elegans, 394. — — faraglionensis, 394, 396. —— —- filfolensis, 386, 388, 401, 404, 420, 432, 434, flaviundata, 384, 385, 386. fusca, 361, 368, 372, 373, 376, 398, 405, 406, 415. corsica, 406. nigriventris, 384, gener, 406. gigliolii, 369, 372, 373, 374. hispanica, 363, 368, 369, 420, 426, 428. latastiz, 394. lilfordi, 369, 372, 373, 420, 424, 430. liolepis, 363, 368, 369, 416, 420, 426 livornensis, 392. maculata, 894, -reticulata, 410. — —— — -tigris, 410. maculato-striata, 394, —— —— milensis, 408. modesta, 394, 396. monticola, 365, 366, 869, 420, 426, 428. multifasciata, 390, 392. naxensis, 408. neapolitana, 377, 380, 381, 384, 387, 388, 394, 406. campestris, 390. lineata, 406. livornensis, 390. INDEX, 615 Lacerta muralis neapolitana nigriventris, 384. ventromaculata, 384. —— — nigriventris, 376, 377, 380, 383, 384, 387, 388, 389, 394, 398, 399, 401, 403, 413, 420, 422, 434. pityusensis, 361, 363, 369, 370, 420, 430. —— —— punctato-fasciata, 394, -striata, 394. —— —— quadrilineata, 383, 384, 399, 404, 405, 407, 419, 420, 424, 434, 436. rasquinetit, 360. reticulata, 394, 399. sardoa, 404, 413, 420, 434. —— -— sawicola, 365, 380. serpa, 376, 387, 388, 389, 390, 394, 396, 398, 399, 400, 401, 404, 420, 424, 432, 434, sicula, 396. steindachneri, 368. striata, 390, 394, 396. striato-maculata, 394. — — tiliguerta, 370, 376, 377, 388, 408, 415, 419, 420, 434, 436. —— — typica, 369, 376, 377, 383, 387, 389, 391, 394, 399, 404, 420, 426, 428. —— -— vaucheri, 365, 415, 418, 420, 486. viridiocellata, 394. olivacea, 394, 396. -— ouycephala, 368, 411. ——— bedriage, 411. — _— hispanica, 368. reticulata, 411. podarcis cettii, 408. genei, 405. puccina, 394. sardoa, 413, 414, -— saxicola, 366. —— schreiberi, 361. —— serpa, 377, 390, 394, 396, 405, 409. —— sicula, 394, 396. —— taurica, 389. —— gener, 406, — tiliquerta, 390, 404, 408, 419, — viridis, 389, 391, 394. —— gadovii, 361. vivipara, 358, 373, 379, 380, 419. Lagomys ogotona, 518. Lagostomus trichodactylus, 487. Lama huanaco, 473. Lamprologus brevianalis, 541, 555, 562, 590. brevis, 541, 561. —— callipterus, 541, 559, 560, 562, 590. —— calliurus, 541, 560, 562, 588. —— compressiceps, 541, 561, 562. —— cunningtoni, 541, 557, 562, 590. —— elongatus, 541, 556, 557, 562. fasciatus, 541, 562. —— furcifer, 541, 562. hecqui, 541, 559, 560, 562. —— lemairii, 541, 556, 558, 560, 562. modestus, 541, 558, 562, mondabu, 541, 557, 562, 590. moor, 541, 562. —— multifasciatus, 541, 558, 562, 588. —— reticulatus, 541, 560, 562, 588. tetracanthus, 541, 556, 562. tretocephalus, 541, 557, 562. Lasianeistrus, 224. Lates angustifrons, 541, 555, 588, microlepis, 541, 555, 588. Lemur mongoz nigrifrons, 503, 504. rufifrons, 503. Lepidoptera, on transformations of South-African, 165. Lepus cuniculus, 488, 489. timidus, 517. Liposarcus, 222. altipinnis, 227. jeanesianus, 228, —— multiradiatus, 228. pardalis, 228. scrophus, 227. varius, 228, Lizard (Wall-), on the West-European and North- African varieties of, 351. Lophuron marginatum, 177, 188. Loricaria, 270, 273. acipenserina, 296. acuta, 287. —— amazonica, 286. —— anus, 289, —— apeltogaster, 292, —— aurea, 301. -— barbata, 301. —— bransfordi, 270. 614 Loricaria brevirostris, 299. —— brunnea, 270. —— cadew, 279. —— carinata, 292. castanea, 287, 288. catamarcensis, 280. cataphracta, 291. —— etrrhosa, 286, 291. depressa, 296. dura, 291. evansii, 290, —— filamentosa, 274, 292. flava, 205. —— frenata, 302. —— jubata, 278, 338. konopickyi, 281. labialis, 288. leviuscula, 284, 285. — lamina, 294, 295. lanceolata, 277. laticeps, 295, 348. latirostris, 283, 340. —— lima, 281, 282. —— macrodon, 293. —— macromystax, 294, —— macrops, 290, 342. —— muculata, 286. —— magdalene, 279. —— microlepidogaster, 283, 338. —— nigricauda, 275, 340. —— nudirostris, 288. —— nudiventris, 290. —— panamensis, 301. parahemiodon, 286. parva, 276. — paulina, 283. —— phowocephala, 276. —— platycephala, 295. platystoma, 298. —— platyura, 270. plecostomus, 208. punctata, 285, 342, rostrata, 290, 300, 301. setifera, 291. simillima, 292, 342. spiaii, 289, = stetndachneri, 281. INDEX. Loricaria strigilata, 283. stubelii, 287. teffeana, 280. —— typus, 286, 287. —— uracantha, 278. ——- valenciennesi, 286. —— variegata, 293, vetula, 293. Loricartichthys, 270, 273. maculatus, 286. Loricariide, revision of the Fishes of the family, 191. Ludia smilax, 170. Lutra vulgaris, 495. Lyceena japonica, 134, Lycenide, revision of the Amblypodia group of the ils Macacus arctoides, 511. rhesus, 511, 520, 521. Macropus bennettii, 449, 452, 517. Mahathala, 1, 20. —— ameria, 22, 162. atkinsoni, 21, gone, 23. hainani, 23, 156. (Apporasa) atkinsont, 21. Malopterurus electricus, 540, 554. Mammals, on the Intestinal Tract of, 437. Manatus inunguis, 464, 516. Manis tricuspis, 454. Marcusenius discorhynchus, 545. petherict, 545. tanganicanus, 540, 545, 578. Mastacembelus congicus, 576. cunningtoni, 542, 575, 576, 600. —— ellipsifer, 542, 576. —— frenatus, 542, 576. —— moorit, 542, 575, 576, —— ophidium, 542, 576. teniatus, 542, 576. tanganice, 542, 576. Megaderma, 492. Metanastria aculeata, 180, 188. Metarctia meteus, 165, 178, 188. Microlepidogaster, 265. perforatus, 269. Microthrissa, 547. Midas rufimanus, 508. INDEX. 615 Mitchell, P. C., on the Intestinal Tract of Mammals, Nilasera, 25. 437. adriana, 117. Mormyrus caschive, 546. amantes, 42. longirostris, 540, 546. | —— apella, 42. Mus arboricola, 482, 483. | ——areste, 119. decumanus, 482, 483. |} — asoka, 118. centaurus, 39. Mustela subpalmata, 495. fulgida, 114. Mycetes beelzebul, 509, 519, 520. | moelleri, 118. | opalina, 51. Muscardinus avellanarius, 482. ——- seniculus, 509. Myrina lazarena, 13. pirama, 39. Myrmecophaga didactyla, 516. | — pirithous, 39. Nandinia binotata, 500, 501. | —— subfasciata, 80. Narathura, 25. wimberleyi, 102. achelous, 82. Nudaurelia belina, 169, 186. —— agelastus, 89, 136. Ochotona pusilla, 517. ——- agnis, 125. . rufescens, 440, 488, 518. — alhopunctata, 60. Odaxothrissa, 547. —- amaria, 22. ' | Ophiusa demta, 183. indeterminata, 183, 190. Ornithorhynchus anatinus, 442, 443, 525. —— ammon, 143. — amphinuta, 86. —— anniella, 114. | Orycteropus, 453, 460. — anthelus, 76, 78, Otaria californiana, 493, 494. —— aroua, 85. Otocinclus, 265. atosia, 87. affinis, 267. buatoni, 60. Jjimbriatus, 267. centaurus, 39. — flexilis, 267. farquhari, 104. joberti, 265. inornata, 91. lophophanes, 269. — kurz, 92. maculicauda, 268. —— lyceenaria, 60. / —— nigricauda, 268. —— mazwelli, 104. | notatus, 268. —— metamuta, 100. | —— perforatus, 269. moolaiuna, 89. | —— vestitus, 267. roona, 136. vittatus, 267, 338. — subfasciata, 151. Ouacaria rubicunda, 509, 511. vihara, 66. | Ovis tragelaphus, 474. Nasalis larvatus, 511. Oxyloricaria, 297. Nasua vittata, 496. barbata, 301. Neobola, 543, 550. —— brevirostris, 299. argentea, 500. —— frenata, 302, 344. bottegi, 550. | — guentheri, 299, 344. loricarviformis, 298. lyra, 300, 346. panamensis, 301. minut, 540, 549, 550, 582. Neoplecostomus, 306. granosus, 306. —— microps, 306. platystoma, 298. Nephele variegata, 177, 188 | —— robusta, 300, 346. 616 Oxyloricaria rostrata, 300. Oxyropsis, 263. —— wrightiana, 264. Panaque, 242. cochliodon, 242, 243. dentex, 243, —— gibbosus, 243. -—— nigrolineatus, 244. Panchala, 25. apidanus, 115. — birmana, 144. — diardi, 113. —— dodonea, 184. —— ganesa, 146. —— morphina, 117. paraganesa, 144, paramuta, 135. —— rama, 130. singapura, 114. Papilio apidanus, 115. centaurus, 39. —— dorimond, 115. —— eumolphus, 104. — helius, 58. hiarbas, 166. ilithyia, 166. timoleon, 9. varanes, 168. Parahemiodon, 270. typus, 286, Parancistrus, 221, 222. aurantiacus, 236. —— mgricans, 236. punctatissimus, 236. Paratilapia calliura, 541. dewindti, 541, 563. frontosa, 541, 561, 592. furcifer, 541, 563. leptosoma, 541. longiceps, 565. macrops, 541, multicolor, 538. —— nigripinnis, 541. pfeffert, 537, 541, 563. robusta, 565. serranus, 589. stenosoma, 541, 564, 566. INDEX. Paratilapia ventralis, 541, 563. Parotocinclus, 265. maculicauda, 268. ~ Pellonula acutirostris, 547. miodon, 540, 546, 578. vorax, 547. Pelmatochromis auritus, 541, 568, 596. lateralis, 538. melanostigma, 541, 567, 596. —— pleurospilus, 537, 541, 567, 596. polylepis, 537, 541, 566, 592. rhodostigma, 541, 568, 596, Peragale lagotis, 446, 448. Perissodus microlepis, 542. Perodicticus potto, 507, 508. Petaurus sciureus, 449, 517, 518, 519. Petovia dichroaria, 184, 190. Petrochromis polyodon, 542, 573. tanganice, 542, 573. Petrogale penicillata, 449, 461. Phacocherus ethiopica, 469, 470. Phalangista vulpina, 449, 450. Phascogale penicillata, 446. Phascolomys, 449. Phoca vitulina, 493. Phocena communis, 467, 468. Phyllalia flavicostata, 173, 186. patens, 165, 173, 186. Phyllonemus, 552. —— typus, 540, 552, 582. Pieris abyssinica, 168. gidica, 168, 186. Pimelodus cyclopum, 311. Plecodus paradoxus, 5642, 574. Plecostomus, 202, 203, 221. —— affinis, 207. — alatus, 211. —— angulicauda, 241. —— anne, 210. —— asper, 220. —— aspilogaster, 206. —— atropinnis, 215. —— auroguttatus, 217. — barbatus, 240. bicirrhosus, 205, 209. biseriatus, 210. borellit, 212, 326. Plecostomus boulengert, 206. brasiliensis, 205. —— brevicauda, 216. carinatus, 209. —— cochliodon, 242. —— cordove, 212, 326. —— emarginatus, 210. feste, 208, 326. Jlagellaris, 291. flavus, 205. Francisci, 211. gurman, 214, 328. genibarbis, 219, granosus, 306. guttatus, 241. heylandii, 221, horridus, 210, johnit, 217. latirostris, 213, 330. —— lima, 214, 216, 217. limosus, 206. luetkent, 217. macrops, 214. maculatus, 286. microps, 306. pantherinus, 213. parahybe, 219. pardalis, 228. pellegrini, 218. plecostomus, 205, 206. —— punctatus, 207. —— robini, 215. —— scabriceps, 207. —— scopularius, 210. —— seminudus, 206. -—— spiniger, 206. —— spinosissimus, 208. —— tenuicauda, 210. —— ternetzi, 213. — une, 215. — vaillanti, 210. —— vermicularis, 216. verres, 209. —— villarsi, 210. —— virescens, 210. —— wertheimeri, 218. VOL. XVII.—PART YI. commersonii, 206, 207. INDEX. Plecostomus wuchereri, 216, 328. Plotus, 516. Podarcis muralis albiventris, 394. campestris, 390. doderleini, 394. —— filfolensis, 402. —— —— faraglionensis, 394. — — maculata, 394. rosewentris, 380. -— sicula, 394. —— — tiliquerta, 409. —— oaycephala, 411. tiliquerta, 405. Pogonopoma, 205. Polyommatus apidanus, 115. centaurus, 39. eumolphus, 104. helius, 58. helus, 39, Polypterus, 543. -—— congicus, 540, 544. Polyptychus grayi, 174, 188. Porcula salvania, 469, 471. Potamogale, 491. Procyon lotor, 496, Protopterus, 551. ethiopicus, 540, 544, Pseudacanthicus, 259. fordii, 261. —— hystrix, 261. —— serratus, 260. spinosus, 260. Pseudancistrus, 221, 222. barbatus, 240. depressus, 240. —— guttatus, 240. —— setosus, 239. wertheimeri, 218. Pseudergolis wedah, 167. Pseudohemiodon, 270. Pseudoloricaria, 270, 272. Pteropus gouldi, 492. medius, 492. Plerygoplichthys, 221. anisiti, 221. No. 11.—October, 1906. lineata, 379, 394, 405. viridiocellata, 394. 617 618 Pterygoplichthys duodccimalis, 226, —- etentaculatum, 226. — -—- gibbiceps, 227. jeanesianus, 228. —— juvens, 221. —— lituratus, 227. —— multiradiatus, 228. pardalis, 228. —- punctatus, 229. undecimalis, 226. Rapala amnisena, 6. Regan, C. T., Monograph of the Fishes of the Family Loricarvide, 191. Rhanidophora cinetigutta, 183, 190. Rhinelepis, 202, 205, 261. acanthicus, 262. —— agassizii, 219. aspera, 220, genibarbis, 219. —— hystrix, 261. lophophanes, 269. strigosa, 220. Rhineloricaria, 270, 271. Rhinoceros unicornis, 478, Rhinopoma, 492. Rhodogastria astreas bauri, 180, 188. Rhytina, 465, 523. Satadra, 25. abseus, 112. ——- agaba, 83. —— alea, 133. —— anthelus, 78. apidanus, 115. creste, 119. atrax, 130. —— bazalus, 119. —— bupola, 104. —— canaraica, 133. chinensis, 118. chola, 118. diardi, 113. —— fulgida, 114. lazula, 118. patuna, 119. selta, 84. —— singla, 120. —— teesta, 119. INDEX. Satadra zeta, 135, Saturnia forda, 170. Simia satyrus, 513. Simochromis diagramma, 537, 542, 571. Smerinthus grayit, 174. Spathodus erythrodon, 542. Sphingurus prehensilis, 484. Sphineg eson, 175. junipert, 177. Sturisoma, 297. Stygogenes, 307. cyclopum, 311. guentheri, 310. humboldtii, 310, 311. Surendra, 1, 3. amisena, 4, 6, 7, 156, 162. biplagiata, 5. discalis, 4, 5. florimel, 4, 7. latimargo, 5. palowna, 4, 6, 7. —— quercetorum, 4. latimargo, 5. stimula, 4, 7, 8. vivarna, 4, 6, 156. Suricata tetradactyla, 500, 501. Sus scrofa, 469, 470. Synodontis granulosus, 540, melanostictus, 540, 553, 586. multipunctatus, 540, 553. —— punetulatus, 554. schall, 554. zambesensis, 554, Talpa europea, 490. Tamandua tetradactyla, 455, 457. Tanganyika, Report on the Fishes of the Third Expedition to, 537. Tapirus americanus, 477. Tarsius, 503, 535. Latusia keppleri, 458, 516. Telmatochromis temporalis, 541, 561, vittatus, 541. Thaduka, 1, 24. multimaculata, 20, 24, 162. Thecla apidanus, 115. narada, 16. nila, 10. Thecla rama, 130. timoleon, 12. Tilapia boops, 542, 574. burtoni, 542, 571. dardennii, 537, 542, 572. galilea, 538. grandoculis, 542, 573. horit, 542, 572. labiata, 542, 572. —_ lata, 572. latifrons, 542, 571, 598. —— martini, 539. melanopleura, 572. —— microlepis, 542, 572. natalensis, 538, 539. —— nilotica, 538, 542, 571. —— philander, 538, 539. —— pleurotenia, 542. —— rubro-punctata, 572. simonis, 538. strigigena, 538. trematocephala, 542. Tinolius eburneigutta, 184. Trabala pallida, 181, 190. Tragqulus meminna, 471, 472. Trematocara marginatum, 541, 570. unimaculatum, 541, 565, 570. Tropheus annectens, 542. nigrifrons, 541, 565, 570, 600. INDEX. Tropheus moortt, 538, 542, 570. Tupaia tana, 490, 492. Ursus malayanus, 496, 498. Usta terpsichore, 169, 186. Varicorhinus tanganice, 540, 549. Vipera aspis, 373. berus, 373. Xenocara, 251. brevipinnis, 257. —— bufonia, 258. chagrest, 256, 336, cirrhosa, 256. dolichoptera, 254. —- gyinnorhynchus, 254, hoplogenys, 2505. latifrons, 253, 338. montana, 258, 336. —— occidentalis, 257, 336. stigmatica, 259. temminckii, 254, Xenochromis hecqui, 542, 574. Xenomystus, 220. gobio, 221. Xenotilapia ornatipinnis, 541, 565. sima, 041, 565, 569. Xenurus unicinctus, 458, 459, 516. Xerus capensis, 480, 481. Zootoca lulfordi, 372, 375, 401, 402. quadrilineata, 405, 408, END OF VOLUME XVII. PRINTED BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET. TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON (continued). To Fellows. To the Public. pasa ede upsets VOLUME XIV. (1896-1898, containing 47 Plates) . Price5 5 O a On O) Partl. (1896, containing 2 Plates)... . . ios Oa) 2-0) 0.82.0 » 2 (1896, containing 6 Plates)... oO O50) Owen O » 8. (1897, containing 9 Plates) . . .. . oO roe ee (0) » 4. (1897, containing 10 Plates) . . . . . er BLAU (0) 220% 20 » 5. (1898, containing 2 Plates) . . . . .- spar Opee Outs 0, Sun ORO: » 6. (1898, containing 7 Plates) . . ... Spe OS aeO ony Aue O » 7. (1898, containing 6 Plates) . . .. . rp OAs ie) enOwlLo en » 8. (1898, containing 5 Plates and TitleandIndex),, 0 9 0 ay Or bevO VOLUME XV. (1898-1901, containing 52 Plates) . Price5 15 6 714 0 Parr ]. (1898, containing 8 Plates) . . . . - 57 O whoo, Ted Lee) » 2. (1899, contaming 5 Plates)... . . 3 20 0) OCT s'O » 93 (1899, containing 4 Plates) . . . : . Bais OO 012 0 » 4. (1899, containing 4 Plates) . . . .. yee 0) 012 0 » 5. (1900, containing 24 Plates) . . .. . eon eoumeO 310 0 » 6. (1901, containing 4 Plates)... .. Pri Oi Orla. » 7. (1901, containing 3 Platesand Title andIndex),, 0 9 O OPT eNO VOLUME XVI. (1901-1903, containing 38 Plates) . Price5 8 O 7 4 0 Parr 1. (1901, containing 6 Plates) . . . =. - FeO MLS aa LF45 30 » 2 (190%, containing 5 Plates). 2. es» OTL 3 COV HN 0) » 9. (1901, containing 9 Plates) . . ..- . jen Oetoe'9 Lode? O » 4. (1902, containing 5 Plates) . . .. . pena Olt Sy U0, 012 0 » 5. (1902, containing 4 Plates) . . . « . pL ONLar 36 018 0 » 6. (1902, containing 3 Plates) . . - . +. » OD DY its ae » 7. (1902, containing 3 Plates) 5. a . ORAS ee OF Loe » 8. (1908, containing 3 Plates andTitle and i Tadex), Oss. 6 018 O VOLUME XVII. Parr 1. (1903, containing 5 Plates) . . . . . Pricel 2 6 110.0 » 2 (1908, containing 3 Plates). . .. . Re Olan 6 018 0 » 3. (1904, containing 13 Plates). . . . - CP in ae 8 LAGE » 4. (1905, containing 8 Plates). . . + . oe ap le eer ae) 1 LOF sO » 5. (1905, containing 50 Text-figures) . . 50s: .“0 ORO , 6. (1906, containing 12 Plates and Title and ENS Lee 110 0O CONTENTS. VI. Fourth Contribution to the Ichthyology of Lake Tanganyika.—Report on the Collection of Fishes made by Dr. W. A. Cunnington during the Third Tanganyika Expedition, 1904-1905. By G. A. Boutencur, L.F.S., V.P.Z.S. (Plates XX X= MUD) 0 hs a ede el Pace Oo dhist of Papers contaimed im Voly XV U tee ee OU Index of Species, Gc, im: Viol: NV ee et ee OS Titlepage and Contents to Vol. XVII. THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Tux scientific publications of the Zoological Society of London are of two kinds—* Proceedings,” published in an octavo form, and “Transactions,” in quarto. According to the present arrangements, the “ Proceedings” contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended to be published in the “ Proceedings” by the Committee of Publication. A large number _of coloured plates and engravings are issued in the “ Proceedings,”. to illustrate the new or otherwise remarkable species of animals described in them. Among such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “Proceedings”’ for each year are issued in four parts, in the months of June, August, October, and April, the part published in April completing the volume for the last half of the pr eet year. Beginning from January 1901 they form two half-yearly volumes. The “ Transactions” contain such of the more important serena cansonars made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. "They are issued at irregular intervals. Fellows and Corresponding Members, upon payment of a Subscription of £1 1s. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1881, if they exceed the value of five pounds. Fellows also have the privilege of subscribing to the Annual Volume of the ‘ Zoological Record? for a sum of £1 10s. (which includes delivery), but this privilege only holds good if the subscription is paid before the First of December in each year. Such of these publications as are in stock may be obtained at the Society’s Office (8 Hanover — Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, 1.C.), or through any bookseller. « P. CHALMERS MITCHELL, October, 1906. ge Secretary. 1944 ae \ ny} 4 { ih Rais Mi, ARIES ii I | Ih i | | i ITHSONIAN INSTI l lll ll 9088 00850