ft igh i Laehtoat ynude ia nee aden agaite delta (i i, aie us ie ahs puna he wig ahs fe Dra hia abe HAN musta oh ea ee Wi tye iegenansnel * janie f sedan tei diarat santa ie rt Seeeraehy ie ao za i we H me Da a Tate Whe AY " tae fae UE DR ent } Rte Sah i fit tr py rt AN thy) Arai Ate fangs met My binteg rene ae .) aa 1 ae eh pie a ch iat “ Shelia: Mat 1 ih 4 ua i coe pis at 11 eave ie iNet nt aid taal hep ids is sii 4a ‘ Luthier! som gente 08 Pe Hie 98 tit at (Aa use Wy ete Se inane eat nite bag ns dla AN ‘aie sis ier Lit Ay Heb oh nt ates it Utriehh a by Me es oe yi ad y th vn ae z nea Screed fete hae peta ay milk pay byt ‘as it if une peels f vali Hae sr itsA inquil 2aibavied ine Me i Ce byl) i ste eal Lat sir ir aus fr ae taalany fev ia fiaeicrineld LAME ane gt if rar by cpt ae i isagngbainntan yy i i v) pels Na i it ie cit aos det matte iA ‘ ae att pera A ce Hite et ac wh CORNELL UNIVERSITY LIBRARY BOUGHT WITH THE INCOME OF THE SAGE ENDOWMENT FUND GIVEN IN 1891 BY HENRY WILLIAMS SAGE Gmina? UREA Keacary ie maon The tailless batrachi 7 | i i Cornell University The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924000414775 THE AY SOO 1 ET Y. INSTITUTED MDCCCXLIV. This volume is issued to the Subscribers to the Ray Society for the Year 1896. LONDON: MDCCCXCTIL, THE TAILLESS BATRACHIANS or KUROPE. BY G. A. BOULENGER, F.R.S. PART £, LONDON: PRINTED FOR THE RAY SOCIETY. MDCCCXCVII. YRINTED BY ADLARD AND SON, BARTHOLOMEW CLOSE, EC., AND 20 HANOVER SQUARE, W. PREFACE. NotwirHstanvineG the great progress attained in our knowledge of Kuropean Batrachians within the last twenty years with regard to the distinction of species and varieties, the study of their anatomy, life-histories, and distribution, there exists at the present day no work dealing with them as a whole from these different aspects. For twenty-five years a close student and collector of these animals, which have always exercised an extraordinary fascination on my mind, I have often wished for an opportunity of supplying such a de- sideratum, by utilising the enormous material which had gradually accumulated in the literature, my own notes, and the unrivalled collection in the British Museum. Thanks to the Council of the Ray Society, who, on the special recommendation of Sir W. H. Flower, acceded to my request, I have at last the satisfaction of seeing my hopes realised, and, owing to the talent of my artistic collaborators, Messrs. P. J. Smit and J. Green, in a manner which fully satisfies my aspira- tions. May those, few as they still are, who share my fondness for this group of animals endorse my opinion of the beautiful and accurate illustrations which are one of the principal features of this work, il PREFACE. and derive from the perusal of these pages one-tenth of the pleasure it has given me to write them; I shall feel amply rewarded for the trouble I have taken. I would also express a hope that a little book of this kind, embodying in a concise form the result of much study, at present scattered in a multitude of publications in some ten different languages, may have the effect of stimulating interest to a subject that has been too much neglected, and in the cultiva- tion of which new workers will find much to repay their efforts, especially if applied in other regions of the globe, which, though much richer in Batrachians, have as yet yielded little or nothing to our knowledge of the life-histories. This work has been planned with the view of assisting the beginner as well as of affording new information to the advanced student, and the synoptic treatment has therefore been frequently resorted to in the Introduction; whilst all descriptions in the sys- tematic part are strictly comparative. Technical terms have not been avoided, but the numerous figures in the text should render them easily intelligible to the beginner. At the desire of the Council of the Ray Society this volume is issued in two parts. The second part, continuously paged and with bibliographical and alpha- betical indexes, is to follow in a few months. The plates, twenty-four in number, have been exe- cuted by Mr. P. J. Smit und printed by Messrs. Mintern Brothers. With two exceptions, all the coloured figures have been taken from living speci- mens. Five of the plates (I, Il, III, XVI, XVII) have already appeared in the ‘ Proceedings of the PREFACE. ii Zoological Society’ (1884, 1885, 1891), to illustrate papers of mine, and have been copied with the sanc- tion of the Committee of Publication of that Society. The figures in the text have been drawn by Mr. J. Green, and reproduced by the Typographic-Etching Company. Most of them are original ; if copied from other works, the source has in every case been referred to. A few cuts illustrating a paper by me on Tadpoles in the ‘ Proceedings of the Zoological Society ’ for 1891, and another by Dr. Ridewood on the development of the hyoid, published in 1897, have been reproduced by permission, for which I beg to tender my acknow- ledgments to the Committee of Publication of the Zoological Society. I also wish to publicly express my thanks to Sir W. H. Flower, K.C.B., and to the Rev. Professor Wiltshire for the interest they have shown in the publication of this work; and to the various friends and correspondents who have assisted me with notes or material, especially Dr. W.G. Ridewood, to whom I am indebted for many hints and preparations; M. F. Lataste, who bas placed his private collection at my disposal; Dr. J. de Bedriaga, Count M. Peracca, Mr. F. A. Moller, Dr. F. Werner, and Herr W. Wolter- storff, who have supplied me with an abundance of living specimens from France, Italy, Portugal, Austria, and Germany ; and Messrs. Chevreux and Doumergue, through whose kindness I have been able to make observations on a large number of living specimens of _ Discoglossus from Algeria. G. A. B. British Museum (NATURAL HIsToRY) ; October 28th, 1897, THE TAILLESS BATRACHIANS OF HUROPE. INTRODUCTION. PAGE I. CLASSIFICATION F , é ss 2 II, ExTERNAL CHARACTERS . ; : ; ; 9 III. IntEGUMENT ‘ : : , , 2b IV. DERMAL SECRETION . . . 30 V. SKELETON. ‘ ‘ : : . 33 VI. VIscERA , ; . 49 VII. Hasits . . F : : : ; . 87 VIII. Voicze . : , : ‘ . 61 IX. Pairing AND OVIPOSITION : . 64 X. SPERMATOZOA. ; ; : 75 XI. Eces. : : 4 ; 78 XII. DEVELOPMENT AND MeETAMORPHOSIS : : : 85 XIII. TapPpoLes. : : . ; 99 XIV. Hysrips : 112 XV. GEOGRAPHICAL DISTRIBUTION 115 A 2 INTRODUCTION. I. CiasstricaTion. For many years the Tailless Batrachians (Hcaupata s. AnunA) were classified in a very unsatisfactory manner. The genera Bufo and Hyla were, it is true, always regarded as the types of distinct groups ; but Pelodytes, Pelobates, Discoglossus, Bombinator, and Alytes were placed with Rana, or in separate families in most unnatural associations. Domerit and Biproy, in the eighth volume of their standard work, ‘ Erpétologie générale,’ published 1841, after dividing the order into two sub-orders, Phanéro- glosses and Phrynaglosses (= Aglossa, the two exotic genera Pipa and Xenopus), a primary division first introduced by Wagler in 1830 and which has stood the test of time, combined the Phaneroglossal genera into three families—Raniformes, Hyleformes, Bufoni- formes ; the two latter containing each a single Euro- pean genus, whilst under the former all others were arranged in a series without further subdivision. So unnatural an arrangement evoked criticism from all who were acquainted with the life-histories of the European forms; and an excellent French observer of these animals, A. T'Homas, in a paper published in the ‘Annales des Sciences naturelles,’ 1854, dwelt upon the correlation which exists between the shape of the pupiland the mode of amplexation during parturition, and proposed to group together ou the one hand those genera that have a horizontal pupil and an axillary embrace, on the other those that combine a vertical or triangular pupil with a lumbar embrace. These divisions were later provided with names by Bruc# in 1863 (Plagioglena, Orthoglena), and by pe wIste in 1877 (Alamplewes, Inguinampleces). How exaggerated the importance attached to this correlation, which, besides, holds good only for the Huropean forms, is now apparent to all. Yet the ar- rangement proposed by Thomas was a decided advance OLASSIFIOATION, 3. on that of the ‘ Erpétologie générale,’ and on the views of J. Mizner (1832), who overrated the taxonomic importance of the structure of the ear, and inspired the classification followed by Srannivus in the ‘ Zoo- tomie der Amphibien,’ published in 1854. The faults of the latter classification were further developed by GinwrHsr in 1858 (‘Cat. Batr. Sal.’), whose avowedly artificial scheme has been followed by Favio in 1872 and by De Burra in 1874. Cors’s epoch-making classification in 1865 (‘ Nat. Hist. Review’), based on a correct appreciation of the osteological characters, placed the matter on a sound footing, and his groupings, slightly modified in 1867, received confirmation, so far as European genera are concerned, from Larastu’s study of the larval characters in 1878 and 1879. In revising the classification in 1882 (‘Cat. Batr. Ecaud.’) I could introduce but slight improvements to Cope’s scheme, as may be seen from the arrangement followed in this work, in which the various groups are placed in ascending order. Principal Schemes of Classification. 1. Dumeérin and Bisron, 1841. Tongue absent: PHRYNAGLOSSES. Rana. Discoglossus. Tongue present: (toothed; a ie SHENSE OCHO digits ee pper jaw with disks: Hyleformes . Hyla. toothless . ‘ . Bufoniformes Bufo. 2. Sranntus, 1854. % (absent: AGLOSSA. a bp . elobates. ES present: cae oe re . : . Pelobatoidea | pom ninator, i not dilated; : : Bufo. ee present; man sent; Bufonina { Alytes 7 SLDERE. digits sterni present: Ranina . Rana si eacline dilated at theend . Hyloidea . Hyla. INTRODUCTION. 3. GunruER, 1858. ‘epiuofng oT} [IIA pojtan ate aepiuhsydoumyy ayy pus ‘pouopuvge mou st ‘auhsydourya SUES UVOLXOTY TY} UO posed ‘WOISIATP SITE, % ‘wh + epi ‘ofng_ ° + aapmofng pmudgng yuesqe tO RD UOT } aeproqwwurgquieg ; DULLOPVULQULO | SRY ** i " : padopaaap a en e Apgoasaeduat eS Pie Iv ide soe ee vaqewoa — | eatasoad sancangboonad jmprssop boas TIME e "pagel ease ee radar : : puunay DUD BPIWOT PoFBTIP FOU + padojaaap Apyooyazad y VSSOTNOUMLOU + 4WO1 Ul va.ty ‘punjhp ‘wpigowphan7g + poyettp oe sy81q' “pphgoophegQ *PeIelip 20U “YSSOTNOHLSIAO : purjad vary “YSSOTDYW : yuesqe ensuoy, OLASSIFIOATION. 5 4. Brucu, 1868. Gon vertical: ORTHOGLENA . Pelobatides ea Pupil Alytes. Hyloides . Hyla. round or horizontal : Pracroarexa | anodes . Bana. Bufonides. Bufo. This arrangement is merely the application of the views expressed by Thomas in 1854. 5. Corr, 1865. AGLOSSA. BUFONIFORMIA . Bufonide: . Bufo. Discoglossus. Discoglosside + Alytes. Bombinator. es "| Scaphiopodide oe Hylide . Hyla. RaNIFORMIA : : ‘ . Ranidz . Rana. The Bufoniformia correspond to the Bufoniformes of Duméril and Bibron; the Arcifera are separated from the Raniformia, which agree in the presence of teeth, by the structure of the pectoral arch; and the Discoglosside are defined for the first time by the opisthoccelous vertebree provided with autogenous ribs. Some further improvements were introduced in 1867, when the Bufonifurmia were split into two groups, based on the structure of the pectoral arch, that group corresponding to the Raniformia among the toothed forms, being named Mirmisternia, a term the signification of which I have extended to embrace all non-arciferous forms, whether toothed or not—an improvement since accepted by Cope himself. 6. Mivart, 1869, in one of his classical papers pub- lished by the Zoological Society of London, endea- voured to combine Giinther’s and Cope’s classifications, but his arrangement does not differ materially from 6 INTRODUCTION. the former author’s so far as the grouping of the European genera is concerned. 7. Lavastre, 1879. Ranidee . Rana. LEVOGYRINIDE 8. PROCELIDE : Hylide . Ayla. Vertebre procelous; tadpole with) Bufonide . Bufo. sinistral spiraculum. : Pelobates. Pelobatide 4 pet, dytes. . : Discoglossus. Discoglossidee egies Vertebre opisthocelous; tadpole Alytide . Alytes, MEDIOGYRINIDH s. OPISTHOC@LIDE : { with medial spiraculum. The Alytide are separated from the Discoglosside on account of the shape of the pupil and the aberrant mode of parturition and development. The classification adopted in this work is shown in the following synopsis, which also includes a list of all the species represented in Europe. Order ECAUDATA. Sub-order PHANEROGLOSSA. Eustachian tubes separated ; tongue present. Series A.—ARCIFERA. Coracoids and preecoracoids connected by an arched cartilage (the epicoracoid), that of the one side over- lapping that of the other. Family 1.—Discoauossipa. Vertebree opisthoccelous; short autogenous ribs attached to the anterior diapophyses; diapophyses of sacral vertebra dilated ; upper jaw toothed. CLASSIFICATION, 1. Discoglossus, Otth, 1837. 1. pictus, Otth, 1837. 2. Bombinator, Merr., 1820. 2. igneus, Laur., 1768. 3. pachypus, Fitz., 1838. 3. Alytes, Wagl., 1830. 4. obstetricans, Laur., 1768. 5. cisternasit, Bosca, 1879. Family 2.—PELOBATIDA. * Vertebree proceelous;* no ribs; diapophyses of sacral vertebra strongly dilated; upper jaw toothed. 4, Pelodytes, Fitz., 1838. 6. punctatus, Daud., 1802. 5. Pelobates, Wagl., 1830. 7. fuscus, Laur., 1768. 8. cultripes, Cuv., 1829. Family 3.—Bouronipa. Vertebree proccelous; no ribs; diapophyses of sacral vertebra dilated ; jaws toothless. 6. Bufo, Laur., 1768. 9. vulgaris, Laur., 1768. 10. viridis, Laur., 1768. 11. calamita, Laur., 1768. Family 4.—Hy.ipa. Vertebree procclous; no ribs; diapophyses of sacral vertebra dilated; upper jaw toothed ; terminal phalanges claw-shaped. 7. Hyla, Laur., 1768. 12. arborea, L., 1766. * Tn the European genera. 8 _ INTRODUCTION. Series B.—FIRMISTERNIA. Coracoids firmly united by a simple epicoracoid cartilage; precoracoids resting with their distal extremity upon the coracoids, or connected with the latter by the epicoracoid cartilage. Family 5.—Raniva. Vertebre proccelous; no ribs; diapophyses of sacral vertebra cylindrical ; upper Jaw toothed. 8. Runa, L., 1766. 138. esculenta, L., 1766. 14. arvalis, Nilss., 1842. 15. cameran’, Blgr., 1886. 16. temporaria, L., 1766. 17. greca, Blgr., 1891. 18. iberica, Blgr., 1879. 19. latastii, Blgr., 1879. 20. agilis, Thom., 1855. We are thus now acquainted with twenty well- established species, as against twelve and eighteen enumerated in the two preceding general works on the Batrachians of Europe, viz. ScHreiper’s ‘ Herpe- tologia Europea,’ 1875, and ve Brpriaca’s ‘ Frosch- lurche Europas,’ 1889. Most of the additions made during the last twenty years fall to the genus Rana, and are in fact dismem- berments of the Linnean species R. temporaria. That the forms here admitted as species are fully entitled to that rank, no one acquainted with their histories will contest fora moment. Nor, among recent acces- sions to our knowledge, is the distinction of the two species of Bombinator or the two of Alytes likely to meet with any opposition. But the rank to be assigned to the principal forms of Hyla arborea and Rana esculenta, regarded by me as only varieties, perhaps remains an open question in these days of extreme multiplication of species. EXTERNAL CHARACTERS. 9 TI. Exrernat CHARACTERS. As these Batrachians have no neck, the head passes directly into the body, and, except in Pelobates, where the skin is closely adherent to the rugose skull, the demarcation is consequently difficult to trace. By drawing a line connecting the articular extremities of the mandible we obtain approximately the exact leneth of the head, the greatest width being at the commissures of the jaws. The head is more or less flattened, with prominent eyeballs covered above by the upper eyelid, which is continuous with the rest of the cephalic integument. The upper surface of the head may be limited on each side by an angular line extending from the tip of the snout to the upper Fig. 1. Upper views of heads of (A) Ranu temporaria and (B) Bombinator igneus. eyelid, the canthus rostralis, continued behind as a ridge or fold, the supra-temporal ridge (Fig. 1, a). In some forms, such as Bombinator igneus (Fig. 1, 8B), the snout is simply arched from one lip to the other, and there is no trace of a canthus rostralis. The side of the snout below the canthus rostralis is called the loreal region. The length of the snout is taken by applying the points of the compasses to the anterior border of the (bony) orbit and the very tip of the snout, on the median line. The nostrils are small valvular openings, situated between the end of the snout and the eyes, alter- 10 INTRODUCTION. nately opening and closing as the throat is drawn in and swollen out during respiration. The eye is large, lateral (Fig. 1, a), or directed upwards and outwards (Fig. 1, B), and protected by three lids: the upper thick and coloured, the lower rudimentary and immoveable, and the nictitating membrane, transparent or partially pigmented, which ascends over the eyeball to meet the upper lid. The eye can be covered without the vision being much impaired by elevation of the nictitating lid, as when the frog is under water, or completely closed when the eyeball is drawn in and the upper eyelid descends to meet the lower. The iris is brilliantly coloured with metallic pigment, golden, silvery, or bronzy, with or without an admixture of red or black. The pupil when fully distended is round and large, but when contracted affects various shapes, which afford important characters for the distinction of the genera. Thus the contracted pupil of Rana, Hyla, and Bufo is horizontal, oval with an angle in the lower border Fig, 2. 29 ® ® poe>S Different forms of pupils. a. Bombinator pachypus. B. Alytes obstetricans. c. Pelobatesfuscus. p. Ranaarvalis. E. Bufo vulgaris. in Rana (Fig. 2, D), and Bufo viridis and calamita, with an upper and lower angle in Hyla and Bufo vulgaris (Fig. 2,8). That of Discoglossus and Bombinator may be described as roundish or subtriangular according to specimens, the lower angle being always distinct, but the upper border sometimes convex, when the EXTERNAL CHARAOTERS. 11 pupil appears nearly round, or straight when the subtriangular form obtains (Fig. 2, a). In some specimens of Bombinator the upper border is even emarginate, so as to produce the shape of a heart; and when extremely contracted the pupil divides into three branches, affecting the shape of a y (Fig. 2, a) This type of pupil approaches the vertical, and in some specimens of Discoglossus the much-contracted pupil is in fact a little deeper than broad. In Alytes, Pelodytes, and Pelobates (Fig. 2, 8,0) we find a regular cat’s pupil, vertically elliptical or club-shaped, linear when fully contracted. . The tympanum, or drum of the ear, is absent in Bombinator and Pelobates. When present it may be concealed under the skin, as in some specimens of Discoglossus, Pelodytes, and Bufo vulgaris, or appear on the temple behind the eye as a round or oval disk covered with thin skin (Fig. 1, a, p. 9). The mouth is large, and cleft to beyond the eyes. The jaws are edentulous in Bufo; the upper is armed with numerous closely-set teeth in the other genera— ten to fifteen in each premaxillary, forty to fifty in each maxillary. ‘The teeth have sharply-pointed, slightly-hooked crowns with long shafts applied to the inner side of the premaxillary and maxillary bones. The palate (Fig. 3, p. 12) is pierced by two ° pairs of orifices, the choane, or inner openings of the nostrils anteriorly, and the openings of the Eusta- chian or auditory tubes, situated near the commis- sures of the jaws; the Hustachian tubes are extremely fine, or sometimes even indistinct, in Bombinator. In all the genera except Bufo there are two groups or series of teeth, each implanted on an eminence of the vomerine bones; these vomerine teeth are situated between or behind the choanez, as shown on p. 12, Fig. 8, A, ©, D. The tongue, thick and papillose, flat, is attached in front and in the middle, and the posterior portion, by being bent over and rapidly thrust out of the mouth, 12 INTRODUCTION. acts as an organ of prehension in most tailless Batra- chians; in the Dviscoglosside, however, as in the newts and salamanders, the tongue is entirely or nearly entirely adherent to the floor of the mouth, and the prey is seized by the jaws. The tongue serves also as an organ of taste, for although frogs seize almost any moving object, they will reject before deglutition anything that is noxious to them, as may be witnessed on offering a frog a brandling or manure-worm (Allolobophora fetida), a ladybird beetle, or a young Bombinator. Before being swal- lowed, the food, if it be a small mollusc, crustacean, or hard beetle, is crushed between the tongue, the depressed eyeballs, and the vomerine teeth; if the latter be absent, as in Bufo, the sharp, sometimes Fig. 3. Open mouths of—a. Discoglossus pictus. B. Bufo calamita. oc. Hyla arborea. D. Rana temporaria. Showing the shape of the tongue oe disposition of the choane, Eustachian tubes, and vomerine teeth. serrated edge of the palatine bones supplies their function. The tongue is circular and entire in the EXTERNAL CHARACTERS. 13 Discoglosside (Fig. 8, a); circular and entire or feebly nicked behind in the Pelobatide and Hyla (Fig. 3, c), resembling a mushroom when thrust out; elliptical or pyriform and entire in Bufo (Fig. 3, 8); oval and forked behind in Rana (Fig. 38, p). In the two latter genera the extensibility of the tongue is so great that a large specimen is able to seize a prey two inches distant from it. Owing to the nature of the skin and the absence of true ribs, the shape of the body varies enormously according to the degree of moisture of the sub- cutaneous sacs, the inflation of the lungs, or the condition of the ovaries, and all measurements that are not based on the bony framework are useless for systematic purposes. The frogs of the genus Rana appear hump-backed in a squatting posture, an angle being produced at the articulation of the cylindrical sacral diapophyses with the ilia; the other genera, with flattened and dilated sacral diapophyses, have a rounded back when at rest, as may be seen on com- paring a true frog with a toad or tree-frog. The vent, or cloacal opening, is a small rounded orifice at the very posterior extremity of the body, above and between the thighs. The fore limb is divided into a brachiwm or arm, an antebrachium or forearm, and a manus or hand with four functional fingers, of which the third is the longest. The length of the first finger as compared to the second is often used as a specific character. In order to preclude misunderstandings, it is well to remark that when the first finger is stated to extend as far as or beyond the second, the two are taken to meet halfway. A radimentary pollex is often indicated externally by a tubercle at the base of the inner finger (Fig. 4, 4). The hind limb is longer, usually very much longer, than the fore limb, and divided into four distinct segments—the femur or thigh, the crus or tibia, or leg, the tarsus, and the pes or foot. Anatomically, 14 INTRODUCTION. the tarsus is, of course, part of the foot proper; but as it here forms a distinct segment, the term foot is used in a restricted sense. The foot is measured from the base of the metatarsal tubercle; it has five elon- gate toes, gradually increasing in length to the fourth, the fifth being again shorter. ‘There is in addition a rudimentary sixth toe, the so-called pre- hallue, conspicuous externally in the form of a tubercle or spur at the base of the inner toe. In Pelobates (Fig. 4, 8) this preehallux or inner metatarsal tubercle acquires a very great development, is covered by a horny sheath with sharp cutting edge, and serves as a shovel for digging in the soil. ‘here is often another tubercle on the sole, at the base of the fourth toe; this so-called outer metatarsal tubercle (Fig. 4, c) is merely a thickening of the integument. Fig. 4. A. Hand of Alytes obstetricans. B. Foot of Pelobates fuscus. c. Foot of Bufo calamita (lower views). Other dermal tubercles are usually present, more or less developed, single (Fig. 6, a, p. 16) or paired (Fig. 4, 0), under the digits at the articulations between the phalanges; they are called subarticular tubercles. The fingers are free except in Hyla arborea, in which they are provided with a rudimentary web. The toes are more or less webbed or bordered by membranes in all our Batrachians, but the extent of the web varies greatly according to the species, and is usually more developed in males than in females, especially during the breeding season. The web is usually smooth and transparent; it is thicker, and often warty, especially towards the margin, in Bufo. EXTERNAL CHARACTERS. 15 In our single representative of the arboreal type, the tree-froe—Hyla arborea—the extremities of the fingers and toes are expanded into adhesive disks (Fig. 6, 8), which assist the animal in climbing on vertical smooth surfaces. These disks do not act as suckers, as was once believed, but adhere by rapid and intense pressure of the distal phalanx and special muscles upon the lower surface, which is at the same time provided with numerous glands producing a viscous secretion. The upward rotation of the terminal phalanx, the swollen base of which is hinged upon an interarticular cartilage situated under the extremity of the penulti- mate phalanx, and can be raised or lowered like the claw of a cat, is easily to be observed on a living specimen. When the disk does not adhere (Fig. 5), its upper surface shows a short ridge produced by the claw- like terminal phalanx, and the lower surface is convex with some longitudinal grooves. During adhesion, on the other hand, the claw-like phalanx is no longer visible on the upper surface, but is indicated by a groove, whilst the lower surface is flat and expanded. Section through the extremity of the third toe of Hyla arborea, the bones being indicated by oblique bars. ph?, ph3. Second and third phalanges. ia. Interarticular cartilage. d. Ad- hesive disk. sé, Subarticular tubercle. The adhesion of tree-frogs to smooth vertical sur- faces is, however, not effected entirely through the above-mentioned organs, their function being supple- mented by the suctorial action of the belly, such as is developed, though to a lesser degree, in the young of most Batrachians which, devoid of digital expan- 16 INTRODUCTION. sions, are nevertheless able to slowly ascend a pane of glass. In most of our genera the fourth and fifth toes are cleft almost to the base, and the metatarsal portions of them are separated by the web. In two genera, Bufo and Hyla, this is not the case; the metatarsals of the two outer toes are bound together by the common integument, the angular divergence of the fifth toe proceeding only from the basal phalanx, as may be seen from the following figures (Fig. 6) representing the foot of the common frog and that of the tree- frog. Fia. 6. Lower surface of foot of Rana temporaria (a) and Hyla arborea (B). The tree-frog differs from other Batrachians in having the articulations of the limbs, especially the wrist, marked above by a strong fold, reminding one of the arms and legs of a jointed papier-maché doll. The proportions of the hind limbs to the body are best appreciated by bending the limb forwards against the sige, and ascertaining the position of the tibio- tarsal or tarso-metatarsal articulation ; a certain allow- ance must, however, be made in cases when the abdo- men of a female is much distended with ripe ova. A convenient method of appreciating the length of the tibia as compared to the femur is to fold the limb so as to place the former at a right angle to the axis of the body, as shown in the accompanying figure (Fig. 7) ; in which position the heels (tibio-tarsal articulation) meet when the tibia is nearly equal to the thigh, fail to meet when shorter, or overlap when longer. EXTERNAL CHARACTERS. av Fria. 7. Limbs of Hyla arborea, showing variations in the proportion of the tibia. The dotted line indicates the middle line of the body. The great development of the hind limbs enables these Batrachians to proceed by leaps ; there is, how- ever, one exception among the Huropean forms, viz. Bufo calamitu, in which the limbs are so short as to render hopping impossible, and this toad progresses at a running pace when pursued. The other extreme is furnished by Rana agilis, which owing to its extremely long hind limbs is able to cover a distance of six feet at one leap. The aquatic habits of a Batrachian can, as arule, be gauged by the extent of the web between the toes: thus Bufo calamita and Alytes, which have only rudimentary webs, are very bad swimmers, and do not venture far from the shore; whilst Ieana esculenta and Bombinator, with their broadly-webbed toes, spend nearly the whole of the active period in the water. But there are exceptions: Pelobates, the toes of which are completely webbed, sojourns but a very short time in the water; it is true that in this case the web is probably of use for pushing aside the soil in the process of digging. The following synopsis of the species is based entively on external characters, and applies to both SEXES. I. Tongue circular, entire, adherent, or slightly free behind ; vomerine teeth behind the level of the choane; first finger shorter than second (DiscocLossip#). r B 18 INTRODUCTION. A. Pupil round, subtriangular, or cordiform. — 1. Vomerine teeth in long transverse series ; tympanum distinct or concealed under the skin . . . . 1. Discoglossus pictus. 2. Vomerine teeth in two groups; no tym- panum. Tibia shorter than the foot 2. Bombinator iqneus. Tibia as long as or a little longer than the foot . . . . 38. Boimbinator pachypus. 8. Pupil vertical; vomerine teeth in short trans- verse or slightly oblique series; tympanum distinct. Three palmar tubercles ; fourth finger as long as or slightly shorter than second 4. Alytes obstetricans. Two palmar tubercles; fourth finger much shorter than second 5. Alytes cisternasit. II. Tongue circular, entire or slightly notched, and free behind; vomerine teeth between the choane ; pupil vertical (PeLopatips). A. Tympanum distinct or concealed under the skin; toes webbed at the base and bordered ; inner metatarsal tubercle small, soft 6. Pelodytes punctatus. n. No tympanum; toes nearly entirely webbed ; inner metatarsal tubercle large, hard, com- pressed, sharp-edged. Interorbital space and occiput convex ; meta- tarsal tubercle yellowish or pale brown 7. Pelobates fuscus. Interorbital space and ccciput flat; metatarsal tubercle black =. 8. Pelobates cultripes. IIL. Tongue elliptical or pyriform, entire, free behind; no teeth ; pupil horizontal (Buronipm). Toes at least nearly half webbed, with paired subarticular tubercles; no tarsal fold; inter- EXTERNAL CHARACTERS. 19 orbital space at least as broad as the upper eyelid . . . . . 9. Bufo vulgaris. Toes at least nearly half webbed, with single subarticular tubercles; a tarsal fold; fingers very obtuse ; interorbital space narrower than the upper eyelid . 10. Bufo viridis. Toes webbed at the base, with paired subarticular tubercles ; a tarsal fold usually present; inter- orbital space narrower than the upper eyelid; hind limb not, or but slightly, longer than head and body . . . . 11. Bufo calamita. IV. Tongue cordiform, free behind; fingers and toes dilated at the end into distinct disks; fingers webbed at the base; vomerine teeth between the choane ; pupil horizontal (Hyripz). 12. Hyla arborea. V. Tongue forked and free behind; pupil horizontal (Rantpa). A. Vomerine teeth between or extending slightly behind the level of the choanz; interorbital space not more than half as broad as the upper eyelid . . 13. Rana esculenta. B. Vomerine teeth extending behind the level of the choane ; interorbital space at least one- half the width of the upper eyelid. 1. Tibio-tarsal articulation rarely reaching the tip of the snout, and never beyond; distance between the dorso-lateral folds five to seven times in the length from snout to vent. Inner metatarsal tubercle large, hard, compressed, measuring one-half to two-thirds its distance from the tip of the inner toe; first finger extending beyond second; tibia shorter than the forelimb. . . 14. Rana arvalis. Inner metatarsal tubercle small, soft, oval; first and second fingers equal, or first extending 20 INTRODUCTION. slightly beyond second; tibia as long as or shehtly shorter than the fore limb 15. Rana camerani. Inner metatarsal tubercle small, soft, oval; first finger extending beyond second; tibia con- sider ably shorter than the fore limb 16. Rana temporaria. 2. Tibio-tarsal articulation reaching the tip of the snout or beyond ; tibia nearly as long as the fore limb; distance between the dorso-lateral folds contained four to five and a half times in the length from snout to vent. a. Tympanum not more than two-thirds the diameter of the eye, distant from it. Distance between the nostrils greater than the interorbital width; first and second fingers equal, or first extending slightly beyond second; inner metatarsal tubercle nearly half the length of the inner toe, as long as the diameter of the tympanum . . . 17. Rana greea. Distance between the nostrils greater than the interorbital width; first and second fingers equal, or first extending slightly beyond second ; inner metatarsal tubercle one-third the length of the imner toe, shorter than the diameter of the tympanum . . 18. Rana aberica. Distance between the nostrils not greater than the interorbital width ; first finger extending beyond second; inner metatarsal tubercle one- third the length of the mner toe, shorter than the diameter of the tympanum 19. Rana latastii. bh, Tympanum two-thirds to five-sixths the diameter of the eye, and close to it ; first finger extending beyond second ; inner metatarsal tubercle very promi- nent, but moderately large. 20. Rana agilis. INTEGUMEN?. 21 III. Inrecumenr, The integument of the Ecaudata differs from that of other Batrachians in the slight attachment to the Fia. 8. Rana temporaria, showing the dermal attachments and the dis- position of the lymph-sacs, upper and lower views. The position of the lymph-hearts is indicated by asterisks. a, Abdominal sac. | g. Gular sac. | sb. Subbrachial — sac. b. Brachial » | Y Interfemoral ,, sbp. Subplantar 5 ce. Crural » |b Lateral a sf. Supra-femoral ,, d. Dorsal » |p. Pectoral ss sp. Supra-plantar_,, f. Femoral 55 muscles, thus rendering the skinning of a frog a most easy operation. But the body is not enclosed in a loose sac, as is often stated in popular descriptions ; there are definite lines of attachment (Fig. 8) by means of connective tissue, which form the demarca- 22 INTRODUCTION. tions between the lymph-sacs that exist on both the dorsal and ventral surfaces immediately beneath the integument. The lymph, fed by the moisture absorbed through the skin, is pumped into the veins by two pairs of lymph-hearts, contractile muscular sacs, the anterior of which are situated in the muscular tissue close to the transverse processes of the third vertebra, the posterior on the sides of the urostyle near its extre- mity; their pulsations may be observed on the back of the living frog. The number and extent of the lymph-sacs varies in the different genera. In Rana and Pelodytes there are 22, 4 unpaired and 9 pairs. ‘The dorsal sac extends from the tip of the snout to above the vent, and is limited by the line of attachment which corre- sponds to the canthus rostralis, the supra-orbital border, the supra-temporal fold, and the glandular dorso-lateral fold. The lateral sac is limited below by a band on each side of the belly, which may be marked externally by a slight groove; the abdominal surface of the head and body is divided into three sacs, gular, pectoral, and abdominal, with partitions across the preecoracoid and coracoid regions. The further sacs (brachial, subbrachial, femoral, supra-femoral, inter- femoral, crural, supra- and sub-plantar) belong to the limbs. In Iyla there is besides a close areolar attach- ment of the abdominal integument. The Discoglos- side differ in the absence of coracoid attachment, or if such exists, as in Discoglossus, it is widely interrupted on the middle line; there is, therefore, no distinet pectoral lymph-sac. In Pelobates the skin of the belly is attached in its posterior half. In Bufo the ventral integument is more or less broadly attached, together with the whole of the dorsal integument. One of the most remarkable peculiarities connected with the integument of Batrachians is the cutaneous respiration, or the power which the surface of the INTEGUMENT, 23 skin possesses of effecting those changes in the blood which are usually performed by the lungs or gills, and by which frogs may remain active for long periods under water. Results of various experiments have proved that pulmonary respiration alone is not sufficient to support life without the aid of that of the cutaneous surface; whilst some of the tailed Batra- chians have lately been shown to be deprived of both lungs and gills, so that the contrary does not hold good for them. But in order to carry out this im- portant function it is of course necessary that the surface be kept in a moist state, and this is effected by a secretion of fluid from the skin itself. An immense number of close-set glandular ceca open upon the surface of the skin, the slime-cells, in addition to which a greater or lesser number of larger poison-secreting glands are scattered or disposed in very conspicuous prominent aggregations, such as the dorso-lateral fold of Rana, the large dorsal warts of Bufo and Bombinator, and especially the so-called parotoids, situated above the ear in Bufo and Alytes, and a similar gland on the calf of Bufo calaimita (Fig. 9, p. 24). These glands are pierced with large pores, visible to the naked eye or with the aid of a low magnifier. A round gland, the frontal or pineal gland, the homologue of the so-called parietal eye of reptiles, is more or less discernible under the skin of the forehead in front of and between the eyes; this gland, in the early stages, was connected with the brain. ‘The presence of lime concretions in the skin of old specimens of Bufo vulgaris has been first pointed out by Leydig. The same have been found in B. calamita by O. Seeck. A character which does not appear to have been noticed before is the presence of a filiform line or raphe, on which the skin is much thinner, extending along the middle line of the back from the snout to the extremity of the coccygeal region in all specimens of Bufo. This raphe, which may “pest be observed on 9A, INTRODUCTION. a stripped skin held against the light, gives rise to the yellow vertebral line which is normal in Bufo calamita, and only exceptional in B. viridis and vulgaris, inde- pendently of another light vertebral streak which is sometimes also present. The independence of the two is most conspicuous in cases of deviation of the former, as often happens in Bufo calamita, which appears to be caused by the presence of large glands in the course of the raphe necessitating a winding (Fig. 9). Bufo calamita, showing yellow raphe, independent of light vertebral streak, and its deviation from the median line in the middle of the back. The epidermis is remarkable for the formation of horny cells, as the spines on the warts of some toads and Bombinator (Fig. 10, p. 29), or the sheaths on the tips of the digits in Bufo and Pelobates, or on the metatarsal tubercles when these are much developed. There are also, in some Batrachians, special small, pearl-like warts, mostly pigmentless, the “ tact-spots” INTEGUMENT. 25 of Merkel, formed by the grouping together of terminal ganglion-cells. These sense-orgaus are found principally on the limbs and flanks of Rana esculenta, more developed in males than in females; the small warts on the fore limbs of Pelobates, which have by some been compared with the nuptial horny excres- cences, seem to be of the same nature. Very similar productions appear during the breeding season in the females of Rana temporaria and It. arvalis, but are of a merely temporary character, and will be noticed in the chapter on Pairing and Oviposition, together with the horny excrescences which arm the digits or other parts in the males at the same period, and assist them In maintaining their hold whilst pairing. Pigment, usually confined to the cutis, sometimes occurs in small quantity in the epidermis. As Leydig has shown, five kinds of special cells may be pre- sent in the cutis—black, yellow, red, white, and metallic or iridescent, the last being the so-called guanin-cells or iridocytes. The yellow and black pig- ments combine to produce the bright green colour, as in normal specimens of Rana esculenta typica and Hyla arborea : if the yellow pigment be absent, as in the German specimens of Lana esculenta, var. ridibunda, the black and gold produce the dull green or olive hue; if both yellow and gold pigments are absent, the black below the cloudy medium of connective tissue and epidermis produces blue, as in the sky-blue specimens of Rana esculenta and Hyla arborea. In the play of the special pigment-cells or chromatophores, which contract or expand and radiate, we find the explana- tion of the changes of colour which some specimens undergo with so great rapidity. » arvalis , » temporaria greeca latastit . agilis 2 29 1. Length from snout to vent. 1 2. 3. 4. 59.5 8 9 46 4 4 18 44 4° °4 10 37 64 0C«dS Od, 54.5 6 34 38 3 4 14 88 9 8 28 70 9 7 25 66 9 7 26 3.5 5 18 72 8 7 26 61 6 6 20 75 6 6 3 50 55 55 15 bl $5 45 15 62 55 55 22 5. 19 18 12 10 15 6. 7. 8. 9. 10. 18 4 15 7 9 1 3 56 2 #15 ll 38 55 25 14 8 25 3 2 5 155 8 6&6 7 93 #36562 9 21 4 10 5 7 v5 10 4 #18 15 7 3 #19 12 35 5 1511 15 6 6 5 16 17 6 5 3 8 19 6 12 6 17 12 4 5 25 10 12035 4 25 8 = a or > or eo for] 2. Length of ventricle of heart. 3. Breadth of ventricle of heart. 4. Length of inflated lung. 5. Length 6. Length of kidney. 7. Breadth of kidney. 8. Length of testis (in case of difference between the two testes, the of longest lobe of liver. longer is measured). adiposa. 9. Breadth of testis. 10. Length of corpora HABITS. 57 VIL. Hasris. _ Frogs and toads were classified by the earlier authors into terrestrial, aquatic, and arboreal. It would be difficult now, nay impossible, to draw any sharp demar- cation line between the three types, and all we can do is to emphasise the differences which exist between the extreme forms. Thus by opposing the extremes we can show examples among the European representa- tives of terrestrial and aquatic (Alytes obstetricans— ftana esculenta), fossorial and scansorial (Pelobates— Hyla), saltatorial and cursorial (Rana agilis—Bufo calamita), diurnal and nocturnal (Rana esculenta— Alytes obstetricans) types, which are, however, con- nected by insensible gradations. These adaptations are revealed by certain structural characters, such as the development of the web between the toes to facili- tate natation, the great development of the inner metatarsal tubercle to assist in burrowing, the adhesive digital disks for climbing, the relative length of the hind limbs for jumping, and the shape of the pupil, whether round or horizontal or vertical, as indicative of more or less diurnal or essentially nocturnal habits. But here and there we meet with exceptions; such are, for instance, the fully-webbed toes of Pelobates, which spends but a short time of the year in the water; the shovel-shaped metatarsal tubercle of Rana esculenta, var. lesson#, which does not burrow; or the similarity in shape of the pupil in Rana and Bufo, which yet differ so much in their habits. Still, on the whole we may roughly incorporate the species outside the breeding season into the following categories : Diurnal and aquatic: Rana esculenta, Discoglossus, Bombinator. Diurnal and terrestrial: Rana temporaria and allies. Nocturnal and terrestrial: Bufo, Pelodytes, Alytes. Nocturnal and fossorial: Pelobates. Nocturnal and arboreal: Hyla. That most, though not all, are aquatic during the 08 INTRODUCTION. breeding season is mentioned in the chapter on pairing and oviposition, All our species also spend a certain period of their existence as gill-breathing, fish-like larvee: there are exceptional forms which entirely dispense with the larval life, and hop out of the ege in the perfect condition, but we have not to deal with them here, for they are only found between the tropics. Even our most aquatic species leave the water after metamorphosis, and young specimens of Runa esculenta and Bombinator are terrestrial, or live more on the border than actually in the water until able to breed. Sexual maturity is not attained before the third or fourth year in males, a year or two later in females—of the larger species, at any rate. The growth is slow, and so prolonged that of certain species, such as Léna and Bufo, one may say the older the specimen the larger; the gigantic size attained by some in secluded localities appears to be mainly due to the immunity they have enjoyed for a great number of years. Individuals of the common toad have been observed for very long periods, and it is probably no exaggeration to assess the possible duration of life in that species at a century. Among the smaller forms a tree-frog has been kept in con- finement for twenty-two years. The food consists exclusively of live prey; no insect or worm will be taken that has not moved in the presence of the consumer. But it is easy to deceive a Batrachian by agitating any object ; the edible frog is thus captured in France by means of a red rag put on to a line. In fact, any moving object of commensu- rate size will be taken, to be afterwards rejected if dis- tasteful. Discoglossus and Bombinator alone among our Batrachians are able to seize their prey under water. The thoroughly nocturnal forms cannot be induced to feed in the daytime. Frogs and toads occasionally display cannibal instincts, swallowing young of their own species. ‘The prey is secured by throwing out the tongue, except in the Discoylosside, which, not HABITS. 59 bemg provided with such a mechanism, simply seize it with their jaws, as does a newt; the hands often assist in pushing it into the mouth. A certain quantity of sand or fine gravel is usually found in the stomach and the large intestine, together with beetle elytra and other indigestible matter. As in many birds, the feces are enveloped in a muci- laginous coating formed in the large intestine. Although sensitive to cold and retiring at the approach of winter, Batrachians do not fall into complete lethargy. Hibernating specimens found in holes, under heaps of manure or dead leaves, or even buried in the mud under the water, are only sluggish, not dormant. Salt is fatal to most Batrachians and to their eggs. A frog soon perishes when thrown into sea water. Yet Bufo calamita is in this respect exceptional, often breeding in brackish pools or burrowing in sand strongly impregnated with salt. The same toad is also remarkable in being less partial to moisture than -any other of our Batrachians. Among the species that remain concealed during the daytime or in inauspicious weather, Bufo vulgaris and Pelobates lead a solitary existence; whilst Bufo viridis and calamita and Alytes are of gregarious habits, two individuals or more being usually found in the same hole or under the same shelter. Brief allusion should be made here to three popular beliefs, which, although often refuted, still crop up now and then, and, curiously, occasionally find cham- pions in educated men. The first, that toads squirt poison at their enemies, is explained by the fact that these Batrachians, when frightened and trying to escape, shoot out, to a con- siderable distance, liquid from the vent. But this liquid, tapped from the bladder, is as innocuous as pure water, and has nothing to do with the poisonous secretion of which toads are really possessed, but unable to squirt out spontaneously. 60 INTRODUCTION. The second is the supposed showers of frogs and toads, believed to fall with the rain, and to have been carried by a waterspout, some narrators of the phenomenon even stating to have received them on their hats or open umbrellas. But the fact that these so-called showers consist of young frogs, never of larve, and only occur at the time of the metamorphosis of the common species, when myriads leave the water and conceal themselves in holes and fissures in the soil, whence they suddenly emerge when the rain falls, sufficiently accounts for the phenomenon. Half blinded by the rain, people, startled at the sudden appearance of these legions of tiny Batrachians on the ground around them, actually fancy they feel them falling, the delusion being further enhanced in the case of baby tree-frogs, which, under the circumstances, climb up their clothes. | The third belief, in live frogs and toads enclosed in stones, hardly deserves refutation, when we know that air, moisture, and food are indispensable to these creatures. If every case could be properly investi- gated, it would be found that the quarry workman, from whom such tales invariably originate, has been the victim of a delusion, and that the Batrachian he fancied to have hopped out of the stone he was break- ing lay concealed close to it unobserved, and jumped from its retreat at the blow of the hammer. And, finally, it remains to observe that our Batra- chians do not bite, although some exotic forms do. However, Lclobates, when irritated, assumes a very ageressive attitude, screaming and jumping with open mouth towards its tormentor, as if to snap at him. No doubt these antics, accompanied by a repulsive odour produced at the same time, must act as a protection against snakes and carnivores. The Boi- hinator are also believed to startle pursuers by turning over and suddenly exhibiting the brilliant coloration of their lower surfaces. VOIOR. 61 VIII. Voices. All out male Batrachians are endowed with a voice, which they produce at least during the pairing season. Females are mute, or respond to the male by a mere grunt. The larynx is provided with vocal cords, which are set vibrating as the air is rapidly shifted from the lungs into the buccal cavity. In many species the sound is intensified by resonance in special vocal sacs situated in the gular region, or at the sides of the head behind the commissures of the jaws. The vocal sacs are called internal when Fig. 23. A K af lig A B A. Rana esculenta, §, with inflated external vocal sacs (lower view). B. Rana temporaria, $, with inflated internal vocal sacs (lower view). covered by the unmodified gular integument, how- ever much this may be distended ; external when their membrane, a diverticulum of the mylohyoid muscle, projects through slits at the sides of the throat, as in Rana esculenta, or when the skin is thinned and converted into a bladder-like pouch, as in Ayla arborea. In some forms there are two distinct sacs,in others but one. ‘The air penetrates by one or two openings in the floor of the mouth, small, rounded, and situated near the commissures of the jaws, as 62 INTRODUCTION. in Rana, or slit-like, and situated at the side of the tongue, as in most other genera. Pelodytes and Hyla have two openings, whilst Bufo has commonly but one, either on the right side or on the left. In Bombinator iqgnens internal vocal sacs are well developed, but are formed by the skin of the floor of the mouth, which is loose and plicate, and projects through a slit dividing the submaxillary muscle into an anterior and a posterior portion. Discoglossus has ‘rudimentary vocal sacs of the same type, but they are confined to the sides of the throat, near the mandible. Bombinator pachypus, Alytes, Pelobates, Bufo vulgaris, and Rane greed, lutastit, tberica, and agilis lack the vocal sacs. The mode of inflation of the vocal sac is best observed in Hyla arborea, the bladder-like appendage when blown being of enormous size, nearly as large as the body, which, when the animal croaks, is much Fig, 24. Hyla arborea, &, with the vocal sac in the collapsed and inflated conditions. thinned by the emptying of the lungs, as shown in the accompanying figure (Fig. 24). This mechanism of the shifting of the air from the lungs into the throat, and vice versd, explains the fact, paradoxical in appear- ance, of Batrachians being able to croak under water. The voice varies very considerably according to the species, and the names bombina, sonaus, campanisona, vidibunda, eachinnaus, &c., convey the impressions made on the ears of the earlier observers. It affords the trained collector one of the surest means of ascer- VOICE. 63 taining the Batrachian fauna of a district during the spring and early summer months; for although it may not lead at once to the discovery of the performer, owing to his most deceptive ventriloquial accomplish- ments, the whistlng argentine note of Aljytes obstet- ricans (often attributed by country people to the salamander), the hoo-hoo of Bombinator pachypus, the oonk-oonk of Bombinator iqneus, the creck-crech-crech of Pelodytes, the clock-clock-clock of Pelobates fuscus, are so characteristic as to be surely recognised by whom- soever has heard them before. ‘he sound uttered by the breeding Bufo vulgaris may be compared to a feeble oft-repeated bark ; that of Rana temporaria may be rendered by grook, grook, and that of Rana agilis and arvalis by co-co-co. But these are all comparatively feeble monotonous sounds. If we turn to the loud croakers, Rana esculenta, Hyla arborea, and Bufo calamita, all three of which are provided with external vocal sacs, we find a more varied performance, which when produced by a large number of individuals, as is usually the case, is simply deafening, and audible miles away. ‘The croaking of Rana esculenta has been admirably rendered by Aristophanes in his chorus of frogs, Boecexexe& xoag xoa—. Hyla and Dufo calamita are more difficult to imitate, their song consisting of an extremely rapid succession of highly sonorous vibrat- ing notes, bra or cra. These loud vocalists are in the habit of joining in choruses, ceasing suddenly at the approach of the listener. But they can easily be induced to resume the concert by an imitation of their croak either by the human voice or by strokes on a metal object. In confinement they can also be enticed to croak by being rubbed on the back. Most frogs when seized by a snake or other enemy utter a shrill ery. That of Pelobates fuscus is most startling; and specimens of that species, when per- sistently teased, can be aroused to a fit of anger, accompanied by loud cries, lasting for several minutes. 64 INTRODUCTION. TX. Parrina AND OVIPOSITION. It is often stated in books that, in temperate climates at least, tailless Batrachians have a fixed annual period of reproduction, taking place in the end of winter or in spring. This statement, based on a generalisation of the familiar phenomenon presented by our two common Northern species, Rana tempo- revia and Bufo vulgaris, is erroneous, individuals of some species—those of the family Discoglossidex, for instance—breeding several times a year, at distant intervals; and between these two extreme types we have almost every possible gradation. In a first category, to which Bufo vulgaris, Rana femporuria, and Rana arcalis belong, the pairing season is of short duration, and although, with us, dependent on atmospheric conditions and regulated by the rise of the thermometer rather than by the calendar, coincides almost to a day for all the in- dividuals under equal climatic conditions. ‘The males are endowed with genesic fury to the extent of letting themselves be mutilated or even immersed in spirit without letting go their mate, to which they cling tightly with their powerful fore limbs, awaiting the expulsion of the ova to discharge on- them their seminal fluid. In their blind frenzy they will clasp individuals of other kinds, even fishes, putrefied corpses of females that have succumbed to their embrace, or all sorts of floating objects. Common toads may be fished by holding out to them a thick stick, to which they cling with their arms. Ag males of that species are always more numerous than females, great fights ensue, in which the latter sometimes perish under the pressure of several of their would-be possessors. PAIRING AND OVIPOSITION. 65 Kven in this category there are occasional excep- tions. Such are the cases recorded by Martin and Rollinat and W. Evans, of single couples of the common toad found breeding near Argenton, in France, on the 18th June, and near Edinburgh on the 18th June, two or three months after the other individuals in the same localities had deposited their spawn and left the water. These exceptional cases are probably due to some accident having prevented the female from resorting to the breeding-place at the proper time. Males, we know, may long retain their genital ardour, so many of them, owing to their excess In numbers, being unable to find a mate. We next find species which may also be said to have a short breeding season, but which do not con- gregate with the same ensemble, the condition of the genital products not being the same at a given time in all the individuals. In these species the female does not, as a rule, enter the water until ready to spawn, and the embrace is, in consequence, of short duration, often taking place only at night. To this category belong Pelobates, Pelodytes, Hyla, and Rana agilis. The males are not, as a rule, animated with the same frenzy as we observe in the first type, and will often let go the females when disturbed or handled. A third category is represented by Bufo calamita, Bufo viridis, and Rana esculenta, which spawn but once a year, but the breeding season of which extends over a longer period, viz. two or three months, within the same district, with an ultimum mostly depending on the weather. Lastly, all the species of the family Discoglossidex, and apparently also some individuals of Pelodytes, are able to breed twice or more every year. Bombinator spawns first in spring and again in summer, some- times as late as August or the beginning of September ; . Discoglossus and Alytes as often as three or four times at distant intervals during the warmer months, the E 66 INTRODUCTION. former (inhabiting the extreme south) from January to October, the latter from the end of February or beginning of March to the beginning of September. © The time at which each species commences to pair varies, as we have seen, with the temperature, for the early breeders at least. Thus in the south of Europe these would start from January, whilst in the north or at corresponding altitudes in the mountains represen- tatives of the same species would be compelled to wait until the thaw has set in. Again, exceptionally pro- longed winters may cause the individuals of a species to be retarded in this operation, so as to bring them together with others which, under normal circum- stances, do not breed until some weeks later. But, on the whole, our Batrachians may be arranged in the following order, beginning with the most precocious: 1. Discoglossus, Rana temporaria, R. arvalis. 2. Rann agilis, Alytes. 3. Bufo vulgaris, Pelobates, Pelodytes. 4. Bufo viridis, Bufo calamita, Hyla, Bombinator, Tana esculenta. Taking the three British species as an example, Rana temporaria in normal years with us breeds in the beginning of March, Bufo vulgaris in the beginning of April, and Bufo calamita from the middle of May ; their nuptial periods do not even overlap. Leaving out the genus Alytes, which is altogether exceptional in its mode of parturition, all our Batra- chians are compelled to resort to the water for the purpose of depositing their spawn. In some species, such as our common frog, the males precede the females, and may be found in winter in a semi-torpid condition, but already attired in their full breeding costume, at the bottom of pools thickly coated with ice. The common toad travels long distances on land, or following the course of brooks, in quest of suitable brveding-places ; and if the sexes meet on the way the male often secures immediate possession of the female, accomplishing the rest of the journey on her back. PAIRING AND OVIPOSITION. 67 The Natterjack toad, B. calamita, also often pairs on land, and being a bad swimmer, and only spawning at night, numbers in embrace may be found in the day- time in holes on the banks, their presence being revealed by their loud croak. The tree-frogs usually pair only at night or in the evening. ‘Thoroughly aquatic species like Bombinator and Rana esculenta resort, of course, to the water before mating. Alytes, the most terrestrial of European Batrachians, pairs and oviposits on land. The female never goes into the water, and the male contents himself with a hip-bath for the purpose of releasing his progeny from the egg-capsules entrusted to his care. Species differ greatly in the choice of a site for their nursery. Alytes exercises the greatest judg- ment, avoiding ponds or pools already largely stocked with tadpoles of other species, or any water which is not permanent. ‘I'he common toad and the tree- frog, also, are judicious in their choice, and their offspring are never doomed through drying up of the site selected, an eventuality which the common frog and the Natterjack toad do not appear to be able to foresee. The latter species especially often spawns in roadside ditches and puddles of a most temporary nature, although suitable places may be near by, and we often find masses of their young tadpoles accumu- lated in a small hole with scarcely any water left, exposed to the sun-rays, where they must perish unless saved by a timely rainfall. Frost often destroys the brood of the common frog. It has been observed by Héron-Royer, and I have more than once been able to verify the fact, that different species which breed in the same water some- how manage to keep clear of each other. In a pond near Paris the above-mentioned author found Rana temporaria occupying the north and west, Rana agilis and Bufo vulgaris the north-east, Pelodytes punctatus the south-east, and Hyla arborea the south, east, and west. In this country, where Rana temporaria and 68 INTRODUCTION. Bufo vulgaris often breed in the same place, we always find the tadpoles forming distinct colonies. A dip of the net will secure specimens of one species only. Courtship does not take place in any of the tailless Fie. 25. Pairs in embrace. a. Discoglossus pictus. B. Pelodytes punctatus. Batrachians. The female is seized by the first comer, and when he has once secured his hold he will remain for a period varying between a few minutes and several days in possession until the evacuation of the ova, unless he be dislodged by a stronger competitor. PAIRING AND OVIPOSITION. 69 Impregnation accompanies or immediately follows, in one or more emissions, the discharge of the eggs. The mode of embrace varies according to the species ; it is axillary in all our genera with horizontal pupil, and lumbar in the others, as shown in the following Synopsis : Fie. 26. Pairs in embrace. A. Bufo vulgaris. B. Rana arvalis. I. Male hoiding the female round the waist. a. The hands joining on the pubic region (Fig. .25, A). Discoglossus, Bombinator, Alytes, Pelobates. 70 INTRODUCTION, s. The forearms meeting on the pubic region (Fig. 25,8). ; Pelodytes. II. Male holding the female behind or above the arms. a. The hands dug into the axils or just above them (Fig. 26, a). Bufo vulgaris, B. calamita, Hyla arborea. B. The hands pressed against the breast (Fig. 26,3) . « ‘ . Bufo viridis, Rana. Owing to the mode of embrace the fore limbs of the males are always stronger than those of the other sex, and in some species acquire about the breeding- time a great muscular development. The inner side of the inner finger may develop a thick pad, supported hy a corresponding development of the bones of the rudimentary pollex and first functional finger, which, pressed against the breast or pubic regicn of the female, assists in maintaining the hold, often leaving a deep scar if the pairing has been of prolonged duration. There are developed in addition temporary horny excrescences, in the form of small granules or spines, which are present in most species, and afford useful characters for their distinction during the breeding season. When this is over the excrescences are shed in most species. They, however, often persist through- out the year in Bufo, Discoglossus, and Bombinator. These nuptial asperities are absent in Alytes and Pelobates, and scarcely distinguishable in our Hyla. When present, their shape and coloration, as well as their distribution on various parts of the limbs or even of the body, vary according to the species. With regard to distribution, we find them located in fully developed breeding individuals— On the inner side of the two inner fingers, the ante- brachium, the brachium, the breast at the base of the arm, the chin, ue belly, and along the toes (Fig. 27, B) : : Pelodytes. On the inner and upper side of the three inner PAIRING AND OVIPOSITION. 71 Fria. 27. Cc Cc Nuptial excrescences of males. A. Discoglossus pictus. B. Pelodytes punctatus. c. Bombinator pachypus (hand and foot). oD. Bufo vulgaris (hand). E. Rana temporaria (hand). 72 INTRODUCTION. fingers, on the chin, the belly, and on the free edge of the web between the toes (Fig. 27, a). Discoglossus. On the inner side of the three inner fingers and the antebrachium, and on the third, the second and third, or the second, third, and fourth toes (Fig. 27,0) . é Bombinator pachypus. On the inner side of the two inner fingers and the antebrachium ' . Bombinator igneus. On the inner side of the three inner fingers (Fig. 27,D) : Bufo. On the inner and upper side of the inner finger (Hie. 27,8). ; . Rana. At the base of the inner finger ‘ , Hyla. It sometimes happens that old females present at least traces of these organs, the appanage of the males. I have recorded such a case in a Rana temporaria with eges in the oviducts; Boscd has found the same anomaly in a Pelodytes punctatus, and Méhely in five specimens of Boibinator pachypus. These horny productions are usually dark brown or black; they are greyish in Rana esculenta and agilis, colourless in Hyla arborea. Their aspect under the microscope, and the differences they present according to the species, have been studied by Lataste, who has shown that they are low and obtuse in Rana esculentu and agilis, erect and pointed in Rana temporaria and arvalis and Discoglossus, pointed and oblique or hooked in Bombinator and Pelodytes, pointed, erect, and crenu- lated in Bufo vulgaris and viridis, and obtuse and crenulated in Dufo calamita (Fig. 28). As in all Batrachians in which no courtship takes place, the males are not distinguished by any orna- mental appendages or more vivid colours, but in Rana temporaria and arvalis they acquire during the pairing season a peculiar swelling of the skin, by which the back, and especially the gular region, take on a bluish tinge and a flabby texture, strongly contrasting with the aspect of the same parts later in the year. The - PAIRING AND OVIPOSITION. 73 females of the same species, on the other hand, develop on the back and limbs, during the breeding season, peculiar pearl-like structures of the epidermis, which are regarded as sensory, being provided with a special innervation, same as we find developed during the spawning season in the males of many Cyprinoid and Salmonoid fishes. These organs have been dealt with at great length by Huber and Maurer. Fie. 28. > sae é eealo se hes, 2; an Be3 0 o es 3 Ps 2s LAUT Sections through the nuptial asperities of males, much en- larged (after F. Lataste). a. Rana temporaria. B. Rana esculenta. c. Bufo vulgaris. v. Bufo calamita. &. Disco- glossus pictus. ¥. Bombinator pachypus. The males of most of our Batrachians make them- selves conspicuous during the pairing time by their voice, whether they be provided with vocal sacs or not. Although there are some, Rana esculenta and Hyla, in which the clamour is not restricted to the breeding season, most others are comparatively mute outside that period, and the vocal sacs undergo a sort of regression, as is especially noticeable in Rana temporaria and Bufo calamita. 74 INTRODUOTION. Our larger species of toads and frogs do not appear to be able to propagate until four or five years old. But it has been ascertained on specimens kept in confinement that Alytes, Bombinator, Pelobates, and Hyla—which are of proportionally larger size immediately after transformation—attain much sooner their full development, and may breed in the third year of their existence. A curious fact, for which we can at present find no explanation, is the very considerable excess of males over females in the genera Pelodytes, Pelobates, and Bufo, when adult at least, as has been observed not only during the breeding season, but also at other times, and especially during the winter, when numbers congregate in holes. In other genera the proportion of both sexes appears about equal; Born and Yung have even ascertained that in the young of the genus Rana the females are in the proportion of 54 to 61 per cent. It is believed that in some lower animals abundance of food is an important factor determining sex. The two above-mentioned physiologists have conducted a series of experiments on tadpoles of frogs (Rana temporaria), and by giving them a more abundant or more artificial food have obtained a much higher ratio of females, viz. 70 to 95 per cent. A. von Griesheim, after examining the sexes of 440 specimens of the common frog captured immediately after the metamorphosis, found only 36 per cent. of males. It would be highly interesting to make a similar investi- gation on the young of some species, such as Bufo rilgavis or Pelobates fuscus, in which the number of adult males far exceeds that of females. SPERMATOZOA. 75 X. SPERMATOZOA. The seminal elements of tailless Batrachians have received a great share of attention, not only on the part of anatomists and physiologists, but even of systematic workers. This is due to the great amount of difference in structure and size between these elements in the various genera; and also to the fact that otherwise closely allied species of the genus Rana may differ in their spermatozoa more considerably from each other than from others morphologically more remote. The failure to obtain hybrids between such near allies as the male Rana temporaria and the female R. arvalis has been attributed principally to these differences; for other species of the genus Bufo, which stand very far apart, but which have almost identical spermatozoa—as, for instance, B. vulgaris and B. calamita, and B. vulgaris and LB. viridis—cross with comparative facility under artificial fecundation. One great source of errors in describing and com- paring the spermatozoa has been due to the changes which these elements undergo in their development, and care should be taken to ascertain whether or not they are in aripe condition. The safest way to avoid mistakes is not to extract the spermatozoa from the testicles, but to collect them in the seminal ducts from specimens in embrace. This is, however, difficult to do in species which are destitute of a seminal bladder. Thanks to the investigations of Leydig, La Valette St. George, Spengel, and Pfliiger, which I have been able to supplement for some of the less known species of Rana, we are now well acquainted with the sperma- tozoa of nearly all the European species. Those of Discoglossus are remarkable for their great size, measuring three millimetres, whilst those of other species do not exceed one-tenth of a milhmetre. The so-called head is spirally wound like a corkscrew; the 76 INTRODUCTION, tail is about equally long, and provided with an ex- tremely delicate, wavy, vibratile, hyaline membrane or crest. As in the newts, they are collected together in bundles, although devoid of a spermatophore. Fia. 29. A B Cc D E F Spermatozoa of Batrachians. a. Bombinator pachypus. B. Pelo- bates fuscus. c. Bufo vulgaris. Dp. Hyla arborea. EB. Rana esculenta. F. Rana temporaria. (After La Valette St. George and Leydig.) In Bombinator a crest is present along the side of the spindle-shaped head, and the tail is moderately long, filiform, and devoid of a crest. The crest may be partially detached from the head, and has been taken to belong to the tail. In Alytes and Bufo the head is staff-shaped, pointed in front; the tail is about twice as long, with a deli- cate crest. In Pelobates the head is long and thin, with strong spiral windings, corkscrew-shaped, passing gradually into the extremely long, delicate, filiform tail. The spermatozoa of Pelodytes bear much resem- blance to the preceding, but the head is less twisted ; the tail is also extremely long and filiform. SPERMATOZOA. 770 In Hyla arborea the head is spindle-shaped, acutely pointed in front, obtusely behind; the tail filiform, and about twice as long. In Rana esculenta and arvalis the head is staff- shaped and obtuse in front, and about half as long as the filiform tail; and in Rana temporaria and agilis the head is very long and thin, not or but scarcely longer than the tail, and attenuate into a fine point at both ends. The spermatozoa of R. greca and latastii bear a greater resemblance to those of #. arvalis, but have a more elongate head, the diameter ten to twelve times in the length; the head is cylindrical, feebly attenuate, and quite obtuse at both ends. The spermatozoa of R. camerant and R. iberica are still unknown. Attention has been drawn above to the importance of securing for comparison the seminal elements in their fully ripe condition. La Valette St. George has shown that at certain stages of development these are almost identical in R. esculenta and ft. temporaria ; and that accumulations of protoplasm at certain points of the unripe element or spermatocyte, as he calls it, produce a shape which may be very unlike that of the true spermatozoon. So careful an observer as Leydig has unfortunately fallen into the error of figuring spermatocytes as spermatozoa in Bombinator, Bufo, and Hyla. Pfliiger has observed that spermatozoa with pointed heads possess a greater facility for penetrating the envelops of the ova of other species to reach the nucleus, a fact which explains the different results obtained in attempts at crossing in both directions, to which highly interesting experiments allusion will again be made in the chapter on hybrids. 78 INTRODUCTION. XI. Eaas. The eggs are spherical vitelline bodies, surrounded by a thin, elastic, cortical membrane or chorion and one or two gelatinous envelops, formed during their passage down the oviducts: the outer capsule swells out in the water after oviposition, and varies accord- ing to the families or genera; the inner is absent in the Pelobatide. With the exception of Alytes, the upper pole of the vitelline sphere is always pigmented, varying from pale brown to black, and this coloration may extend on the whole sphere,—as, for instance, in Budo vulgaris, or on the whole save a small whitish spot at the lower pole, as in Rana temporaria. The inner mucilaginous envelop is formed in the upper portion of the oviduct, and surrounds each ovum; the outer, which has been termed “adhesive envelop,” is formed lower down the oviduct, and either surrounds each single ovum, whether free or agglutinated in masses, or forms a common investment in which the ova are disposed irregularly or in files. All eggs, except those of Alytes, are deposited in the water. If laid singly, each is fixed by its adhesive envelop to submerged bodies; if in strings or bands, these are twined round plants. Egg-masses are usually also fixed to plants, except those “of Rana temporaria, which are endowed with a peculiar buoyancy, and simply float on the surface. The eggs of some Batrachians have a strong odour of fish, this being most noticeable, among European forms, in Pelobates fuscus. The amount of protection which the mucilaginous envelops afford the embryo varies considerably. In Discoglossus, Pelodytes, Pelobates, and Bufo they soon partly dissolve, so as to release the embryos almost before these are able to execute any spontaneous move- ments; the embryos, so to say, drop out and become fixed by their adhesive subbuccal apparatus to the EGGS. 79 outer surface of their remains. In Bombinator, Hyla, and Rana the embryo develops much further within the egg, and becomes liberated by its own action. In Alytes, which is exceptional among European Batrachians in the eggs being laid on land in strings, taken charge of by the male parent, who carries them about twisted round the legs, all the embryonic stages are passed through within the egg, and the young is born as a true tadpole. Contrary to the other Batra- chians, the food-yolk of which is in insufficient quantity to form an external appendage of the embryo, we find here a large yolk-sac. These eggs, which measure three and a half or four millimetres in diameter, are of a uniform yellow colour, and are protected and connected in rosary-like fashion by a comparatively tough but highly elastic transparent investment, which represents the adhesive envelop of the typical Batrachian ovum, and contracts thread- or hair-like between every two eggs. The following synopsis shows the principal dif- ferences between the eggs of the European species, and will serve to their identification, the dimensions given being those of recently deposited ones. I. With large unpigmented vitelline sphere, 33—4 millimetres in diameter, laid on land in rosary- like strings, carried by the male twisted round his legs, the larva leaving the envelops in an advanced condition after the loss of the ex- ~ ternal gills and the formation of the spiraculum. Alytes, II. With small vitelline sphere (1—3 millimetres in diameter), entirely or partially brown or black, deposited in the water, the embryo leaving the envelops with external gills or before their appearance. ; a. Deposited singly or in small groups of two to twelve; vitellus brown or black above, white or yellowish beneath, the embryo leaving the 80 INTRODUCTION. envelops with external gills or just before their appearance, with the tail more or less developed. Vitellus 1—13 millimetres in diameter, black above . . . » « Discoglossus. Vitellus 2 millimetres in diameter, brown above. Bombinator. b. Irregularly disposed in a thick mucilaginous band (formed by fusion in the cloaca of the contents of the two oviducts) ; vitellus dark brown or black, with the lower pole white, 13—2% millimetres in diameter, the embryo leaving the envelops before the appearance of the external gills and tailless. Pelodytes, Pelobates. c. In regular files in two long mucilaginous strings; vitellus entirely dark brown or f black, or with a whitish lower pole, the embryo leaving the envelops before the appearance of the external gills and with a rudimentary tal. . . . . . . Bufo. Vitellus 13—2 millimetres in diameter; eggs, in each string, in three or four files when floating, in two when stretched. . . . .B. vulgaris. Vitellus 1—14 millimetres in diameter; eggs in three or four files when floating, in two when stratehed. ... «4 6 «© « « » A aradis, Vitellus 1—14 millimetres in diameter, with whitish lower pole; eggs in two files when floating, in one when stretched. D. calaiita. pb. In large lumps; upper pole brown or black, lower whitish or yellowish; the embryo leaving the envelops with external gills and a well-developed tail. 1. Envelop, when fully swollen out, measuring 3—4 millimetres in diameter; vitellus yellowish white with brown upper pole, 13 millimetres in diameter; embryo yellowish. Hyla. Fia. 30, iy / Eggs of—a. Bombinator pachypus. B. Discoglossus, pictus. c. Alytes obstetricans. Dp. Pelodytes punctatus. EB. Bufo vulgaris. ¥. Hyla arborea. G. Rana temporaric. fa 82 INTRODUCTION. 2. Envelop 7—10 miilimetres in diameter ; embryo brown or blackish. . . Rani. («) Vitellus miparti, brown or blackish above, yellowish or white below, or the light spot covering at least the lower third; eggs submerged. Diameter of vitellus 1}—2 millimetres. R. esculenta, R. arvalis. Diameter of vitellus 2—3 millimetres . LR. aqilis. (b) Vitellus nearly entirely black, with a small whitish lower pole. . Diameter of vitellus 2—3 millimetres ; eggs float- np + 4 ; . . Lt. temporaria. Diameter of vitellus ‘12 millimetres ; eggs submerged . . . . . B, latastis. It may be noticed that, so far as species are con- cerned, the size of the egg stands in no relation to that of the parent, a fact which is also conspicuous in the tadpole. Thus the largest Batrachian, Bufo vul- garis, has eggs no larger than the smallest, Pelo- dytes punctatus; and in Rana temporaria they are considerably larger than in J. esculenta, which reaches both in the larval and perfect states to a much greater size. Héron-Royer has observed that young females produce eggs of a somewhat smaller size than those of full-grown specimens of the same species, and I have been able to verify his observation on tana temporaria. ‘he number of eggs also varies according to the size of the female. Héron-Royer, who has counted them in most of the Huropean species, gives the following numbers as the result of his computation : Bufo viridis . . . . . ~=10,000—12,000 Teanga esewlenta .. 2. . ~=-10,000 Bufo vulgaris... 1. A,972— 6,840 Bufo calamita. . . . . 8,000— 4,000 tana temporaria, . . . ~ 2,856— 4,005 Pelobates fuseus . 2... -:1,200— 2, 236 Rana arvalis . 5. 1,000— 2000 EGGS. 83 Pelodytes punctatus . . . 1,000—1,630 Rana agilis Bose, 18h oe 669—1,200 Hyla arborea... 800—1,000 These results, however, do not always agree with my own, as will be seen further on under the various specific headings. In the Discoglosside, which spawn two or three times a year, each brood consists of 300 to 1000 eggs in Discoglossus pictus, 80 to 100 in Bombinator pachypus, and 20 to 90 in Alytes obstetricans. This makes a probable maximum total of 3000 for the first species, 300 for the second, and 270 for the third. These numbers are, however, only approximative, for no observer has yet been able to follow the spawning of a given specimen throughout the year, and to record the total number of eggs produced. Manipulating frogs’ eggs is not easy, on account of the sticky mucilage which adheres to the fingers or instruments, and the counting is a very trying opera- tion. Approximate computations are more easily arrived at with the egg-strings of toads, which can be divided into small portions, in which the eggs are counted and the result multiplied by measuring the length of the whole strings. Fia. 31. OOO @ & © © © © Cleavage of egg of Rana temporaria (after Ecker). After fecundation the vitelline sphere undergoes a process of division, or segmentation, which can be readily observed under the microscope or an ordinary hand lens, and of which the above diagrammatic 84, INTRODUCTION. figure from Ecker’s ‘Icones Physiologice’ will give an idea. This type of segmentation, in which the first furrow completely cleaves the whole egg into two, each suc- cessive one again subdividing that cell with which it is related, the first two furrows vertical and the third horizontal, is termed complete or holoblastic. The first plane of division passes through the plane of symmetry of the future embryo, the tissues of which are formed from these cells. The egg of Alytes contains a much greater quantity of nutritive matter, and the segmentation is less regular and more limited ; it belongs to the meroblastic type. The time required for the eggs to hatch varies con- siderably, according to the species. Thirty hours may suffice for those of Discoglossus pictus, whilst those of Alytes obstetricans take at least, under the best conditions, nearly three weeks. The eggs of other species fall between the two extremes, the evolution being of course accelerated or retarded by the variable temperature with which early breeders have to contend, or even temporarily arrested when the thermometer descends to freezing-point or below. DEVELOPMENT AND METAMORPHOSIS. 85 XII. Devetopment anp Meramorpuosts. In the first condition, when the embryo has become distinct from the vitelline sac, the head is large and distinct from the elongate body, the tail absent or rudimentary. The head is usually cleft below by a median longitudinal groove, in the middle of which a transverse or rhomboidal depression represents the first rudiments of the mouth; on each side, in front of the mouth, a pit indicates the nostril; behind it is a grooved fold, the cephalic crescent, which develops into a single or paired prominence, the holder— “crochets” of Rusconi—acting as an adhesive apparatus by means of which the helpless embryo fixes itself at first to the outer surface of the mucilaginous envelop of the egg, and later to weeds or submerged objects. Eyes are absent. A small bud-like tubercle on each side of the posterior border of the head is the rudi- ment of the external gills, and vertical grooves in front and behind the bud represent the visceral clefts, the intervals between which will later become converted into the four branchial arches. The adhesive apparatus mentioned above varies con- siderably, and its conformation affords the means of distinguishing genera or even species at a period when the tadpole characters are not yet developed. It is mainly through the observations of Héron-Royer and the more scientific researches of Thiele that we have become acquainted with its modifications in most of the European forms. Thiele has shown that the term * sucker,” which has been bestowed on this organ by various authors,is a misnomer. There is no muscular suctorial apparatus developed in connection with it ; it is glandular and secretory, producing a sticky mucus or slime, which serves to fasten the larva to its resting-place.* The mode of its development and * From this apparatus the ventral adhesive disk of certain tadpoles living in mountain streams of the Himalayas, Burma, and the Malay Archipelago is no doubt developed. 86 INTRODUCTION. retrogression is shown in the following series of figures of Bufo vulgaris (after Thiele). Fie. 32. Development and retrogression of the subbuccal apparatus in Bufo vulgaris. It appears as a crescentic groove very early, before any other organ, and disappears, after having under- gone various changes, in the beginning of the tadpole period. Its fullest development coincides with that of the external gills. It is single and discoid, with a crescentic groove, in Discoglossus; single and Y-shaped in Pelodites and Pelobates ; single and V-shaped or crescentic in Bufo; and paired, forming two small disks, in Bowbhinator, anu, and Hyla. These disks are close together in Bombinator, widely separated in Myla and Ranw tenporaria and agilis, whilst they are connected by a fine transverse ridge in Rana escu- _lenta, which thus affords a link between the single and paired types of holders. Hyla differs from all others in the disks shifting forwards so as to be, in the later stages, situated on a level with the mouth. The condition in Discoglossus appears to be the most primitive, from which those shown by Bombinator, fiana, and Hyla on the one hand, Pelobates and Bufo on the other hand, may be derived. The following DEVELOPMENT AND METAMORPHOSIS. 87 diagram, copied from Thiele’s memoir, shows roughly the shape of this organ, and its position with respect ce the mouth at the close of the first larval period. Nw . E F G H A. Discoglossus pictus. B. Pelobates fuscus. c. Bufo vulgaris. D. Bufo viridis. ». Rana esculenta. ¥F. Rana temporaria. G. Hyla arborea. u. Bombinator pachypus. As the embryo grows the tail elongates and shows a muscular portion with chevron-shaped myotomes, bordered above and below by a membranous crest. The gills become digitate or branched; the olfactory pits shif. more forward, and become converted into functional nostrils communicating with the mouth; the eye may be detected at the side of the head, appearing first as a pigmented ring under the trans- parent epidermis; the mouth becomes bordered by fleshy lips; the anus is perforated; and the larva is able to feed, having thus far subsisted on, the vitellus contained in the abdomen. The external gills, two or three in number on each side, the second and third often rudimentary and con- cealed under the first, are largest and persist longest in Rana temporaria. In this species and in the toads they are strongly pigmented, like the body; in most others they are not, or but feebly, pigmented. A little smaller but likewise branched in Rana esculenta, Bufo vulgaris, and Discoglossus, they remain very short and 88 INTRODUCTION. unbranched or bifid in Bufo viridis and calamita and in Hyla arbored. Z In Alytes, within the egg, the external gills are extremely developed, but single; and the very long and slender brauches are confined to the ventral side, as in other tailless Batrachians. On entering the second period, or true tadpole stage, the external characters of which will be more fully described in the following chapter, an opercular fold covers the external gills (the right some time before the left in the forms with lateral spiraculum), which atrophy and are replaced by in- ternal ones, small branched fila- ments disposed along four cartila- einous arches. The anal tube is developed ; the mouth acquires horny mandibles and the lps horny teeth; the nostrils assume a more dorsal position; the sub- buccal adhesive organ disappears ; and the opercular fold having fused with the skin above the gill-arches, Youngtadpoleof Hyla the head becomes confluent with ee ce the globular swollen belly, in which branchialarches(ba), the extremely elongate intestine the heart (c),and the shows through the transparent coiled-up intestine : 5 : (i). x 3. abdominal membrane, coiled up like a watch-spring (Fig. 34). In the third period the hind limbs appear as buds at the base of the tail, and gradually attain their full development during the tadpole life. The fore limbs grow simultaneously, and even more rapidly, but remain concealed within a diverticulum of the bran- chial chambers until fully formed, when they burst through the skin (unless the left spiraculum be utilised for the egress of the corresponding limb), leaving in DEVELOPMENT AND METAMORPHOSIS. 89 front of them a small cleft, out of which the gill- filaments of the branchial arches in process of dis- appearance may often be seen projecting (Fig. 35). Alytes obstetricans, towards the end of the larval stage, showing the gills (br) protruding through the cleft in front of the arm, and the median spiraculum (sp). Then only the caudal crests become reduced and the tail gradually absorbed; the gill-arches entirely dis- appear; the lungs, which had co-existed as accessory respiratory and hydrostatic organs, assume alone (or, rather, together with the skin) the respiratory func- Fie. 36. ry Development of Discoglossus pictus, x 13. tions; the horny armature of the mouth and lips is shed in pieces; the lips are absorbed and the buccal cleft extends; the eyes become free and acquire move- able lids; the lachrymal canal is shifted towards the 90 INTRODUCTION. eye and perforates the lower eyelid; the intestine shortens; the anal tube ceases to function, and dis- appears with the last vestiges of the caudal crest, which had become detached from the vent; and the _ young frog, usually still bearing a stumpy tail, leaves the water. The metamorphosis is completed. The skull and its appendages also undergo very important changes. The cranium in the tadpole is a large undivided cartilage with narial openings and large suborbital fenestra. The suspensorium of the lower jaw is extremely elongate, and sends up a strong process, connected with the cranium by a fw) bridge, in front of the eye. The premaxillaries are Fia, 37. ip PPO, ee eco f, ) pe Fete a t ua chy | me Su A. Skull of full-grown tadpole of Pelobates fuscus. B. The same at the end of the metamorphosis, after the loss of the horny beak. chy. Ceratohyal cartilage. eo. Exoccipital. fp. Fronto- parietal. di. Lower labial cartilage. me. Mandibular (Mecke- lian) cartilage. ma. Nasal processes of chondrocranium. po. Preorbital process. pro. Prootic. su. Suspensorium (palato- quadrate), wl. Upper labial cartilage. represented by a single or paired cartilage, the upper labial, loosely attached to the diverging anterior pro- cesses of the cranium, which supports the upper horny beak; and to this corresponds a pair of cartilages, the lower labials, ultimately the mento-Meckelian or symphysial bones, supporting the lower beak and attached to the short mandibular or Meckelian cartilage. In the transformation of the mouth this mandibular cartilage acquires a greater length, and the suspensorium becomes reduced in proportion and shifted backwards. DEVELOPMENT AND METAMORPHOSIS. 91 The ceratohyal cartilage at first articulates by a condyle with the suspensorium below its praorbital process, and extends across the gular region towards its fellow, from which it is separated by a narrow space, with one or two small cartilaginous pieces (basihyals, Parker ; copule, Gaupp), behind which is a paired plate (hyobranchials, Parker ; branchial plate, Gaupp). To each branchial plate the four cartilaginous arches bearing the tufted internal gills are attached by con- nective tissue, and ultimately fuse with it (in the third period). The homology of these branchial arches has been the subject of much discussion. According to Parker, they are to be regarded as subcutaneous cartilages, and do not belong to the category of true visceral arches; they are called accordingly eatra- branchials—the true branchial arches of fishes and tailed Batrachians being the ceratobranchials, repre- sented by small styliform cartilages first attached to the outer border of the hyobranchial cartilage, with which they fuse, the fourth to persist throughout life, and ossify as the thyrohyal. This view is strongly refuted by the latest worker on the subject, Gaupp, whose conclusions are as follows : The hyobranchial apparatus of the larva consists of two hyalia [= ceratohyals] (Fig. 38, ch), con- nected mesially by a pars reuniens ; continuous with it follows the copula [basihyal, bh], which connects the hyal with the branchial skeleton. The latter is made up of two branchial plates [hyobranchials, hbr] and four branchialia [ceratobranchials, clr] on each side, which pass into the branchial plates with their inner extremities and into each other with their dorso-lateral extremities. The whole hyobranchial skeleton forms a cartilaginous continuum. For a time a cartilaginous piece is developed in front of the pars reuniens, which may be regarded as a first copula. During metamorphosis the larval apparatus is for the greater part lost; the pars reuniens and the anterior copula are absorbed ; the branchialia disappear with- 92 INTRODUCTION. out a trace. ‘I'he body of the hyoid cartilage of the transformed animal is composed of the two united branchial plates, the posterior copula, and of newly formed cartilage, constituting the manubria of the principal cornua, probably derived from the copula. The alar and postero-lateral processes are new formations which only appear comparatively late. The so-called ‘processus thyroidet”’ [thyrohyals, ¢] have nothing to do with the branchial processes (nor with the ceratobranchials of Parker, which Gaupp terms spiciula), but are remains of the posterior portion of the larval branchial plates. Thus, of the larval apparatus nothing is preserved but the hyal arch, and a portion of the copula and of the two branchial plates. Dr. W. G. Ridewood has recently investigated the development of the hyobranchial skeleton of Pelodytes, and the careful figures he has drawn are reproduced here with his kind permission. He agrees with Gaupp in the main results, but shows the postero-lateral process to be the persistent proximal portion of the first ceratobranchial. In Pelodytes as well as in Pelobates the anterior processes are expanded and bent inwards, and the cornua become broken, the anterior portion fusing with the alar processes, whilst the posterior becomes free from the hyoid plate, as may be seen on Fig. 38. In order to complete this rapid sketch of the evolution of the hyobranchial apparatus, which is of erveat importance for the correct understanding of the hyoid apparatus in the perfect form, it remains to notice that in the Discoglosside the first copula or basihyal is of much larger size than.in other frogs, and persists longer; whilst the second extends farther back, and completely separates the hyobranchial plates from each other. The ossification of the thyroid processes takes place only after transformation. The only cranial ossifications to develop during the DEVELOPMENT AND METAMORPHOSIS. 93 Fig. 38. ” elearalee ‘ Gy v bh { Transformation of the mandible and hyobranchial apparatus of Pelodytes punctatus. (After Ridewood, ‘ P. Z.S.,’ 1897.) Nine different stages, No. 9 representing the adult. Figs. 1—7 dorsal view. Figs. 8 and 9 ventral view. as, bh. ebr. ch. . Dentary bone. . Lateral foramen. . Hyoidean cornu. . Convex surface by which Angulo-splenial bone. Basihyal. Ceratobranchial. Ceratohyal. the ceratohyal articulates with the palato-quadrate cartilage. . Hyobranchial plate. . Hyoglossal sinus. . Internal or mesial part of the mandibular cartilage. . Lateral part of same. . Mento-Meckelian (symphy- sial) bone. pa. Anterior process, pal, Antero-lateral process. . Postero-lateral process. . Space enclosed between the hyobranchial plate and the proximal ends of the first and second cerato- branchials. . Spicules of branchial arches. t. Thyrohyal. tf. Thyroid foramen. . Ventral splint-bone. 94 INTRODUCTION. tadpole condition are the exoccipitals and prootics (cartilage bones) and the fronto-parietals, or frontals and pavietals if these ossify from distinct centres, and parasphenoid (membrane boues). ‘The preemaxillaries, maxillaries, and squamosals, next in order, appear only some time after the loss of the suctorial or larval mouth. Whilst the fore limbs are enclosed, each moiety of the shoulder-girdle is widely separated from the other ; the scapula is directed upwards, and the coracoid and precoracoid form a loop turned downwards and inwards. At the same period the ilium is_per- pendicular to the vertebral column. The tail proper of the tadpole remains in a noto- chordal condition, no cartilage being ever formed in that region. But both dorsal and ventral cartilages are developed in the basal portion of the caudal region, and ossify as two dorsal elements or arches and a single ventral strip behind the ninth or sacral vertebra ; they gradually fuse to a continuous cylinder, the urostyle or coccyx, from which both chorda and spinal cord ultimately disappear, to form in the adult a solid bony rod. The vertebra of the trunk are formed, as first discovered by Dugés, on two different plans. In the Discoglosside, Pelobatide, and Hylide the chorda remains for a long time exposed along the ventral surface, and, owing to the absence of cartilaginous formation around it, disappears without ever becoming invested otherwise than by a thin elastic membrane ; it can easily be stripped off below the vertebrae in specimens on the point of metamorphosing. ‘This has been termed by Gegenbaur the epichordal type. In the Bufonide and Ranide, which represent the peri- chordal type, the greater share of the formation of the whole vertebra falls to the (paired) dorsal cartilage, but there is in addition a narrow ventral or hypo- chordal cartilage which fuses with the dorsal or becomes connected with it by calcified tissue; the DEVELOPMENT AND METAMORPHOSIS. 95 notochord is thus completely surrounded by a thick sheath in tadpoles with imperfectly developed limbs. This mode of formation of both the arch and the greater part or whole of the so-called centrum from the same cartilage explains why there is never a neuro-central suture in these Batrachians. Fra. 39. Side view of cartilaginous skeleton of tadpole of Bombinator pachypus, with outlines of the external parts, (After Goette.) A, Anus. cor, Coracoid and ov. Occipital vertebra. E. Eye. preecoracoid. p. Pes. M. Mouth. f. Femur. pe. Pelvis. N. Nostril. fi. Fibula. ru. Radius-ulna. S. Spiraculum. h. Humerus. se. Scapula. as. Astragalus. il. Lower labial. su. Suspensorium. ea. Caleaneum. m. Manus. t. Tibia. ch. Chorda. me. Meckelian. ul. Upper labial. chy. Ceratobyal. v. Vertebra. During segmentation of the dorsal cartilages men- tioned above, which send out the transverse pro- cesses or diapophyses, there appears between each two centra an intervertebral cartilage, out of which the articulating knob of the centrum is formed, and be- comes attached either to the vertebra anterior (pro- coelous type) or posterior (opisthoccelous type) to it, if not remaining as an independent, intervertebral, ossified sphere, as we sometimes find in specimens of Pelobatide. Ossification appears first in the neural arch, next descends to the centrum, and sets in last in the articulating condyle. 96 INTRODUOTION. The annexed figure (40) of the vertebral column (ventral aspect) of a recently transformed Pelobates cultripes shows well the independence of the inter- vertebral spheres and the three elements which enter into the formation of the urostyle. A glance at this figure will also show to what extent the importance of the mode of articulation of the vertebre has been overrated by those systematists who have proposed to divide the tailless Batrachians pri- marily into proccelian and opistho- ceelian types. As a matter of fact, both occur in the family Pelobatide, some of the exotic genera of which are normally opisthoccelian. During larval life the habits vary considerably according to the species. Thus Hyla is always on the move, swimming about in every direction like a fish; Bom- binator, Alytes, Pelobates, and Rana keep more to the bottom, approaching the borders of the ponds or pools during the hotter part of the day to sun themselves in shallow water; Discoglossus, Pelodytes, and Bufo are in the habit of coming frequently to take air on the surface, where they swim about with upturned bellies. The notion that tadpoles are exclusively herbivorous has long been exploded ; they feed on both vegetable and animal substances, and usually show a preference for the latter, in addition to the microscopical organ- isms contained in the ooze, which they swallow in large quantities. Their carnivorous or even cannibal habits are only too noticeable in confinement, and they have been utilised for making small and delicate skeletons. The tadpoles of Bombinator and Alytes are almost exclusively carnivorous, the real difference between them and the perfect animal being that the former feed readily on dead or even decomposing matter, DEVELOPMENT AND METAMORPHOSIS. 97 whilst the latter takes only living prey. Nevertheless the intestine in all tadpoles is exceedingly elongate, being several times longer and of a greater calibre than after the metamorphosis. Bataillon has ascer- tained by direct experiment that the abbreviation of the gut is not localised to any special region. Tad- poles of Alytes obstetricans having been anzsthetised a few days before the egress of the fore limbs, the abdominal cavity was cut open, and the intestine, after having been slightly uncoiled, marked at regular dis- tances by silk threads tied round it. The abdomen was then sewn up, and the tadpoles, having recovered, continued their evolution. At the close of the meta- morphosis the gut was again examined, and it was found that the threads, which were tied at intervals of twenty millimetres, were then only seven or eight millimetres apart. I have measured the uncoiled digestive canal in fully developed tadpoles and in recently transformed young of the three following species. It will be seen from the measurements, in millimetres, how consider- able the range of variation between different forms may be, both in the larval and perfect conditions. Thus in Rana esculenta the length of the intestine of the tadpole is nearly ten times that of the perfect form, whilst in Alytes obstetricans its length is only as 4:1; Pelodytes punctatus, with 6: 1, is inter- mediate. This difference is correlated with the diet, Rana esculenta being the most herbivorous, and Alytes obstetricans, as noticed above, almost ex- clusively carnivorous in the tadpole condition. Length of intestine in tadpole (1) and young (11) of— i: i, Alytes obstetricans (25 mm.body-length) 235 60 Pelodytes punctatus (20 ,, re ) 194 32 Rana esculenta (23° 4, * ) 472 45 Before closing this chapter on Development, the highly interesting fact remains to be alluded to that G 98 INTRODUCTION. whereas the transformed tailless Batrachians, unlike their tailed precursors, occupy too high a position in the zoological scale to be able to regenerate ampu- tated extremities, the hind limbs reproduce without difficulty during the larval existence, a fact first observed by Spallanzani in the last century, and since verified by Gimther and Barfurth. In order to be regenerated it is, however, necessary that the limb be not completely removed. A basal fragment of the femur must remain to give off the cartilage- cells which, most curiously, will become differentiated into the various segments, proceeding from the distal to the proximal; and the mutilation must have taken place some time before the close of the third larval period or true tadpole stage. It has been stated that reproduction of the limbs can only be obtained in young larvee, long before they have reached their full size. Thisis, however, contrary to my own experience on large, full-grown tadpoles of Alytes obstetricans, in which the well-developed hind limb, after having been amputated below the knee-joint, was regenerated with the greatest facility. Fie. 41. Series of embryos of Bombinator pachypus. (Atter Goette.) Showing the formation of the larval mouth, together with the atrophy of tbe adhesive disks; the gradual disappearance of the external gills, and the formation of the median spiraculum by approximation of the two opercular slits; and the develop- ment of the gut (the abdominal membranes being removed). wha ao Soon sam a ans PAA sr eee peo Span ") ats 2 a SRS. OAS State RN att ae at ag a $F at 5 nos , Ses Se Ms . le , RF win * Us = BR, ee oT RU ye 6. TE OE om, a Seti, SD csevapla BIOS, + Ee oieee ty edhe a TADPOLES. 99 XII. Taprougs. The term tadpole is restricted to the developmental stages between the formation of the spiraculum and the egress of the fore limbs, and corresponds to the second and third periods of Dugés. Descriptions are preferably drawn up from specimens in the third period, which begins with the budding of the hind limbs. form.—The head and body are so fused that it is Fria. 42. Open mouth of—a. Pelodytes punctatus. B. Alytes obstetricans. b.o. Buccal orifice. Up. Labial papille. m. Mandibles. ¢. Series of labial teeth. (From ‘ P. Z.58.,’ 1891.) 100 INTRODUCTION. extremely difficult to discern the limit between the two. The term ‘ body” is, therefore, used in the description as meaning both head and body. Its lon- gitudinal measurement is taken to the origin of the hind limbs. The tail consists of a fleshy muscular portion bordered above and below by membranous expansions, termed respectively the upper and lower crest. By depth of the tail is meant its greatest depth, crests included; and the length is measured from the posterior extremity of the body. _f- Mouth (Fig. 42).—This term is used in its wider sense, z, ¢. to include the much-developed lip surround- ing, like a funuel directed downwards, the horny beak, not unlike that of a cuttle-fish, which forms the entrance tothe mouth proper. The characters offered by this circular lip are among the most important for the distinction of species. ‘The lip may be entirely bordered by fleshy papilla (/.y.), or these may be re- stricted to the sides, or to the sides and the lower border. Its inner surface is furnished with ridges armed with series of minute bristle-like erect horny teeth (¢), each of which, when — strongly magnified, is seen to be formed of a column of superposed cones, — hol- lowed out at the base x B and capping each other Horny teeth of Rana uyilis (4) (Fig. 43); the summit or and Pelobates fuscus (B). (After erown of each of these Van Bambeke.) ‘ cones is expanded, spatu- late, hooked backwards, and usually multicuspid. The denticles are absent in Pelobates and Peludijtes, and present in all the other genera. The number of these columns is very great. F. E. Schulze has counted as many as 1100 in the lip of Pelobates fuscus. Fria. 43. TADPOLES. 101 By drawing an imaginary line across between the mandibles the lip may be divided into an upper and a lower portion, the series of teeth above the upper mandible being termed wpper lubial, those below the lower mandible being lower labial. These are described as first, second, third, &c., proceeding from the outer border towards the beak in both the upper and lower sections of the lip, as shown in the accompanying figures (p. 99). Hach series is reckoned as one, whether continuous or more or less interrupted in the middle; this method being far more simple, and at the same time more correct, considering the great amount of individual variation, than those used by some authors who distinguish between ‘‘median”’ and “lateral” series, according as to whether or not the series is broken up in the middle, The first series, either in the upper or lower division of the hp, may be marginal (Fig. 42, a, t.1), or it may be within the border, which is then occupied by fleshy papille (Fig. 42, Bs). The arrangement of the series is expressed by a formula—%, for instance, indicating the number in the upper and lower divisions, the figures being separated by a transverse line corresponding to the position of the horny beak. The labial teeth are usually arranged in a single row on each ridge (Fig. 42, a); in the Discoglosside, however, each ridge, or at any rate the second, bears two or even three rows of teeth (Fig. 42, 8). The beak itself is made up of horny elements like the labial teeth; its edge, when . sufficiently magnified, is seen to be denticulate, each i \ denticle representing the cusp of a single tooth. The inside of the mouth and the pharynx (Fig. 44) are beset with long pointed mucous papille, ‘ taste- organs” of F. E. Schulze, the longest of which border the choane (ch), one in front and one behind, and are followed by a pair with a denticulate fold between them ; a cushion-like, hemispherical swelling (¢) repre- sents the tongue, on the anterior border of which a pair of papille are inserted. A broad cleft with 102 INTRODUCTION. denticulate anterior fold leads into the branchial chambers (b7), which contain the four branchial arches. Fic, 44. Transverse section through the head of larval Pelobates fuscus. (After Schulze.) a. Upper part. 3B. Lower part. 6r. Bran- chialchambers. ch. Choane. l. Lip. m. Mouth. ph. Pharynx. t. Tongue. These arches bear on the convex outer side the delicate arborescent gills, and on the concave inner side de- velop a membranous septum with vermicular perfora- tions, a special sifting or filtering contrivance through which the water absorbed by the mouth has to pass before reaching the respiratory organs of the branchial apparatus. Spiraculwn.—The water is expelled from the bran- chial chambers by one or two tubes opening by one orifice in all European Batrachians. This orifice is the spiraculum. In the Discoglosside two tubes are present, which gradually converge towards each other at the end of the first period of development, as shown on Fig. 41, p. 98, and in the tadpole proper discharge through one transverse, slit-like or crescentic opening, situated in the middle of the breast (Fig. 45, a, sp.). In all other tailless Batrachians the tube is single, and TADPOLES. 1038 opens on the left side, straight backwards in Bufo, backwards and upwards in the other genera (Fig. 45, B, sp.; Fig. 46, a, B). Fia. 45. A B Lower view of tadpoles of—a. Alytes obstetricans. n. Hyla arborea. (From ‘P. Z.8.,’ 1891.) The Lines of Muciferous Crypts or Sensory Canals.— All tadpoles are provided with these organs, the homo- logues of the lateral line in fishes ; they stand in imme- diate relation to the nerve branches, and are regarded as organs of a special sense possessed by aquatic animals, feeling, in its broadest sense, having been admitted as their possible use, and the function of determining waves of vibration in the aqueous medium having been suggested. The surface openings of these tubules or canals form lines which may differ greatly 104 INTRODUCTION. as regards their degree of development in individuals of the same species, and their arrangement also varies, within certain limits, irrespective of the species. I Fie. 46. = SS, So meee We z Tadpoles showing sensory canals. a. Rana agilis. B. Pelodytes punctatus. C. Alytes obstetricans. (From ‘P, Z, 8.,’ 1891.) have found them usually most distinct in Pelodytes, Peluhutes cultripes, Rana agilis, R. latastii, and Bom- binator igneus ; least in Bufo. I append figures (Fig. 46) of tadpoles of Rana agilis, Pelodytes punctatus, a" Aaitobsitiny, ances + . our Soa une J indy ease Le. yb 4igs /, Z, Caines ‘P30 TADPOLES. 105 and- Alytes obstetricans, in which these organs are represented by dotted lines. In addition to these lines, all tadpoles show more or less distinctly a small whitish gland in the middle of the head between the eyes, the so-called frontal gland or pineal gland, which in early stages is connected with the brain. A glan- dular streak extending from the nostril towards the eye is the lachrymal canal. The presence of a number of pulsating lymph-sacs has lately been discovered by Weliky on each side of the dorsal caudal crest. Pigmentary Network.—In addition to the ordinary pigment-cells, the tadpoles of afew genera (Discoglossus, Bombinator, Pelodytes) present a system of fine black lines, most apparent on the caudal crests and the more feebly pigmented parts of the body, which afford excellent characters for their identification. It is a fact worthy of notice that when the end of the tail has been injured these pigmentary lines are not reproduced on the regenerated portion. Their aspect is shown in _ the following figures (from ‘P. Z. 8.,’ 1891), repre- senting much-enlarged pieces of the upper caudal crest of Pelodytes (A) and Discoglossus (8). “ Fie, 47. A B The following key has been drawn up to facilitate the identification of the genera and species in the tad- pole condition. I. Spiraculum median; anus median; tail rounded or obtuse at the end; a papillose edge all round 106 INTRODUCTION. the lip, sometimes narrowly interr upted above ; labial teeth in 2 series, disposed in two or three rows, at least in the second series. A. Spiraculum equally distant from the anterior and the posterior extremity of the body ; tail at least once and a half the length of the body, and three and a half or four times as long as deep ; caudal crests with a polygonal network of fine black lines . A . DiscoGLossvs. B. Spiraculum nearer the posterior than the ante- rior extremity of the body; tail not more than once and a half the length of the body, twice to twice and a half as long as deep; caudal crests with fine black decussating lines. BoMBINATOR. Mouth trigonal : : B. igneus. Mouth elliptical . ; . B. pachypus. c. Spiraculum nearer the anterior than the poste- rior extremity of the body ; tail at least once and a half the length of the body, twice and two-thirds to thrice and a half as long as deep ; no black network : . ALYTES. II. Spiraculum sinistral; labial teeth disposed in a single row in each series or ridge. A. Anus median. 1. Spiraculum directed upwards and_back- wards; lower lip bordered with papillee ; series of labial teeth ¢ or a. Tail obtusely pointed, a fine black de- cussating lines ; an inverted fold on each side of the lip, the upper edge of which bears a long series of teeth; beak white with black edge. . PELODYTES. -b. ‘Tail acutely pointed, without black lines ; first series of upper labial teeth short ; beak entirely black . . Prvosates. ‘l'ail once and a half to twice as long as the body. ‘ 0 gl’. Fuscns, TADPOLES. 107 Tail not more than once and a half as long as the body . : . LP. cultripes. 2. Spiraculum directed straight backwards, tail rounded at the end; both upper and lower lip with toothed edge; series of labial teeth 2. . . Boro. Mouth at least as wide as the interocular space, which is twice as great as the distance between the nostrils; second upper series of labial teeth very narrowly interrupted in the middle . 2B. vulgaris. Mouth nearly as wide as the interocular space, which measures once and a half the dis- tance between the nostrils; second upper series of labial teeth more or less inter- rupted in the middle . ‘ B. viridis. Mouth considerably narrower than the inter- ocular space, which is nearly twice as great as the distance between the nostrils; second upper series of labial teeth very widely interrupted in the middle. B. calamita. z. Anus dextral; spiraculum directed backwards and upwards; lower lip bordered with papille. 1. Anus opening above the lower edge of the tail; upper caudal crest extending far forwards on the back, almost to between the eyes, which are lateral and visible from the ventral as well as from the dorsal aspect of the body ; series of labial teeth 2 : : ; : . Hyta. 2. Anus opening close to the lower edge of the tail; upper caudal crest not extending forwards beyond the vertical of the spira- culum ; eyes on the upper surface. Rana. a. Series of labial teeth 2-33. Interocular space at least twice the distance between the nostrils, and much greater 108 INTRODUCTION. than the width of the mouth; tail acutely pointed, at least nearly twice as long as the body . ; . R. esculenta. Interocular space but little wider than the distance between the nostrils or the width of the mouth; tail once and two-thirds to twice as long as the body . -B. arvalis. b. Series of labial teeth 224. a. Tail obtusely pointed, once and a half to twice as long as the body. Second series of upper labial teeth widely in- terrupted in the middle; first series of lower labial teeth at least two-thirds the length of the second; width of mouth a little less than the interocular width, which equals about once and a half the dis- tance between the nostrils . L. teiporuria. Second series of upper labial teeth continuous or narrowly interrupted; first series of lower labial teeth at least two-thirds the length of the second; width of mouth quite as great as the interocular space, which equals about once and a half the distance between thenostrils . &.greeca. Second series of upper labial teeth widely in- terrupted in the middle; first series of lower labial teeth hardly half as long as the second; width of mouth much less than the interocular space, which equals nearly twice the distance between the nostrils . : : . LR. cberieca. 6. Tail acutely pointed or submucronate, at least nearly twice as long as the body. Interocular space once and a half the width of the mouth or the distance between the nostrils; no tubercle on the upper man- dible. ‘ Lt. latastiv. Interocular space ‘at least twice the width of TADPOLES. 109 the mouth or the distance between the nostrils ; usually a black tubercle on the upper mandible. ; . RL. agilis. With regard to size, the tadpoles rank as follows : Maximum length on record. Millimetres. 1. Pelobates fuscus . is « 175 2. is Culiripes s « = » « +» 120 3. Lana esculenta... » &: & WE 4. Alytes obstetricans . . a 2 OO Dy 5p CUSEOPINSTE «we we a aw oe «| 6D 6. Pelodytes punctatus. . . . . . 65 1s ORG GGUS ee ww OD 8. Bombinator igneus. . . . 2. 50 0. Hyla arborea 2. 1 «© » « « » 49 10, Rana dberiea, 1. «© « a2 es «6 AD Lhe gy OTRO ae & 2 oa oo ws e 48 e.g. HOMIOTONG 6 ok ee es «~~ O 13. Bombinator pachypus. . . . . 48 14. Bafowilis . 2 sw a we we we ee 16. deena taba, 16. ,, arvalis . . . . . 48 17. Discoglossus pictus. . . . 38 18, Bafowulgaris . ie 2 « « » » BB 19. ,, calamita . . . . . . .~) 80 But if we compare the maximum length of the tad- pole with the maximum length of the adult, we obtain the following order : Maximum length of adult from snout to vent. Millimetvres. 1. Pelobates fuscus . 2. se BO 2. Alytes obstetricans . . . . . . 54 Os ga ASIETOIR ys oe 3 Oe we we a AD 4. Pelobates cultripes . . . . . . 88 5. Pelodytes punctatus. . . . . .) 45 6. Bombinator iqneus . . . . . . 50 110 INTRODUCTION. Millimetres. 7. tiple apboree . 4 6 « 2 « « « BD 8. Rana tbericu . . . oe ee 00 9. Bombinator pachypus @ jai. . ch fen aby. SO 10, Ranu esculenta. . . 125 iL. 6 gy. green. : . . 66 Tei oye NOGOIE a, Go. RB eG i Se BO 13. 4, datas. 2 + « « . . 65 Wh cay PE ce og we . « 0 15. temporaria... « UD 16. Bufo i: » ~ 20 17. Discoglossus pictus . . 2 1. 08 1S, Bufo colada. « « « «2 » + 80 ID. gg PUNGORIE Se we we - 180 It may be noticed that the five species which head the list as having proportionally the largest tadpoles are the only ones in which the pupil is vertical. The structural differences which separate the genera aud species in their tadpole condition reflect, on the whole, pretty accurately the system based upon the perfect animals, although here and there the modifi- cations are of unequal importance. We must bear in mind, however, that such a correspondence, if existing in the European Batrachians, is not universal. Larval forms such as the tadpoles are outside the cycle of recapitulation, the ontogeny being broken by the intercalation of the larval phasis. The horny beak, the circular lip with its horny armature, the spi- raculum, the enclosure of the fore limbs in diverticula of the branchial chambers, and such special adapta- tions as the ventrai disk or sucker of certain exotic mountain forms, clearly point to tadpoles having had a developmental history of their own. We need, therefore, not be surprised at occasionally finding, within the same genus, very different types of tad- poles, or even a total suppression of the free larval stages, as is actually the case in the large and widely distributed genus Rana. str Ny, w Metin, ree n sn davies svete al Hananyecn eed hy Oi aM sa aioyunia fe 2 ayant gg? S yyuweaWietine. qo einige + Ms, "ny si ra aT Tee ea ae Ee ' ja / peel Sv aa as ua ga Se ie ses “sid RWG Re aw OMEN SS a ee ci >a ar a ce ring aint Borromeo - any iS ne i ‘te a grass Seentenees Pridbpoler, wet COLLET TADPOLES. lil The tadpoles are represented, of the natural size, on Pls. I, H, and Il]. The mouth (a) is enlarged five diameters in figs. 1 and 2, Pl. II; seven diameters in fig. Pl. I, Fig. 99 bed 4, Pl. 1; fifteen diameters in figs. 3 and 5, and fig. 1, Pl. 1; ten diameters in the rest. PLATE I, 1. Discoglossus pictus. Montecristo. 2. Bombinator igneus. Seeland, Denmark. 3. — pachypus. Mondorf, Luxem- burg. 4, Alytes obstetricans. St. Germain, near Paris. uo = var. bosce. Sierra Hs- trella. 6. — _ cisternasii. Sierra Morena. 7, 8. Pelodytes punctatus. Near St. Malo. PLATE II. . 1. Pelobates fuscus. Prague. ; — cultripes. Near Bordeaux. 3. Bufo vulgaris. Near London. 4. — viridis. Breslau. 5. — calamita. Near St. Malo. 6, 7. Hyla arborea. Near St. Malo. 8. — — var. meridionalis. Near Nice. PLATE III. 1. Rana esculenta. Near St. Malo. 2. — arvalis. Breslau. 3. — temporaria. Near London. 4. — grexca. Parnassos. 5. — «@ & «= BO Rana temporaria. . . . 43 17, 18. one latastttvw . . . . 648 Bombinator iyweus . 6 ee AO 19, 20s lane CeUOW. Gs oe we we eA With regard to the vertical range, the species may be arranged in the following order : Maximum elevation reached. Feet. 1. Rana temporaria. . 2. . . 10000 9. Rana camerani . . . . . . 8000 3. Bufo vulgaris . . . . «© « « 7000 118 AN Oe Wl ae WHO Bee CQ ~TS CUR LT LS 19. ol. . Pelodytes punctatus . Pelobates fuscus . Bufo vulgaris . Hyla arborea sis . Rane esculenta + 5. f Alyles obstetricans 4, * (Bufo viv Ldis 6. Luin yreea 7 . Bombinator pachypus S. Itana iberica 9. dene agilis. . 10. Bufo calamita . 11. Rana esculenta 12. Hylu arborea INTRODUCTION. Feet. 6500 6500 5800 5500 4500 42.00 4,000 3500 3300 The other species do not ee to 3000 feet. * The following table shows the distribution in the different countries of Kurope: 1 2. . Discoglossus pictus oer . Bombinator igneus puchypis ; Alytes obstetricans » clstermesils . a cultripes . » viridis . mo Ge » calanita + 5 evades. camercitt ae 4 temporaria oy Ese < Guiven. aes » tberica. letastil, nn agilis 1. Great Britain and Iveland. 2. Scandinavia and Denmark. 3. France. 4. Belgium und Holland. 5 Germany and Switzerland. 6. Spanish Peninsula. 7. Italy. Di t+++4ti ti: 3. Dt++4++44+4: t4i of! 4. DHs Di ttt ti +) 8. Austria-Hungary and Roumania. 9. Dalmatia, Bosnia, and countries south of the Danube. 10. Russia. Di ttt+ttit¢i i ttt 9 D+: ++ +4: i429 Ds Pt: t+ ttt: Dtti tt +4 +: +2 8. 9. + as + + + + + + + + + + + + ee + + neo at + + 10. + +4++44+4+4++ 4+ * 15,000 feet in the Himalayas. GEOGRAPHICAL DISTRIBUTION. 119 It may also not be without interest, in order to bring out their associations, to enumerate the species occurring in the environs (i. e. a radius of about fifteen miles) of some of the principal towns the Batrachian fauna of which has been thoroughly searched by com- petent workers. It will be seen, then, that no district can boast of possessing more than ten out of the twenty recognised species, and that, in spite of their northern position, Paris and Bonn yield to no other and are richer than many more southern localities. 1. London (three species). Bufo vulgaris, B. calamita, hana temporaria. 2. Copenhagen (nine species). Bombinator igneus, Pelobates fuscus, Bufo vul- garis, B. viridis, B. calumita, Hyla arborea, Rana esculenta, It. arvalis, Rh. temporaria. 3. Paris (ten species). Bombinator pachypus, Alytes obstetricans, Pelo- dytes punctatus, Pelobates fuseus, Bufo vulgaris, B. culamita, Hyla urborea, Rana esculenta, It. temporaria, Rt. agilis. 4, Bordeaux (nine species). Bombinator pachypus, Alytes obstetricans, Pelo- dytes punctatus, Pelobates cultripes, Bufo vulgaris, B. calanita, Hyla arborea, Rana esculenta, Lt. agilis. Brussels (five species). Bufo vulgaris, B. calamita, Hyla arborea, Rana esculenta, Lt. tenvporaria. 6. Berlin (nine species). Bombinator igneus, Pelobates fuscus, Bufo vul- garis, B.vividis, B. calumita, Hyla arborea, Rana esculenta, Le. arvalis, R. temporaria. 7. Bonn (ten species). Bombinator pachypus, Alytes obstetricans, Pelo- bates fuscus, Bufo vulgaris, B. viridis, B. calamitu, Hyla arborea, Rana esculenta, R. arvalis, R. temporaria. rt 120 INTRODUCTION. 8. Geneva (eight species). F Bombinalor pachypus, Alytes obstetricuns, Bufo vulgaris, B. calumita, Lyla arborea, Runa esculenta, Le. temporaria, Le, agilts. 9. Coimbra (ten species). Discoglossus pictus, Alytes obstelricans, 1. cister- nusti, Pelodyles yraetatus, Pelobates cul- tripes, Bufo vulgaris, B. calanita, Hyla arborea, Rana esculentu, Rt. ‘berica. 10. Turin (seven species). Pelobutes fuscus, Bufo culyavis, Be vavidis, Hyla arborea, Lana esculenta, R. latastit, De. ayilis, 11. Palermo (five species). Discoglossus pictus, Bufo vulgaris, B. viridis, Tyla arborea, Runa esculenta. 12. Vienna (ten species). Bombinutor iynens, Bo pachypus, Pelobates fuscus, DBufo culgaris, Bo viridss, Hyla arborea, Lana esculenta, I. arvalis, 2. temporaria, R. agilts. 13. St. Petersburg (three species). Bufo vulyaris, Rane urcalis, R. temporaria. 14. Moscow (eight species). Bombinator iqneus, Pelobates fuscus, Bufo vul- garis, DB. viridis, Hyla arborea, Luana esculenta, Lt. arvalis, Lt. temporaria. And finally, in order to complete the sketch of the Geographical Distribution, a list is appended of all the tailless Batrachians known from other parts of the Palearctic region in its widest sense, viz. North Africa and Asia north of the 30th parallel. Column A stands for North-western Africa, B for North-eastern Africa, C for South-western Asia, D for North- western and Central Asia, E for Hastern Asia with Tibet, and F for Japan. The names of species occur- ring also in Hurope are prefixed with an asterisk. GEOGRAPHICAL DISTRIBUTION. DiscocLossipm&. *Discoglossus pictus, Otth . Bombinator orientulis, Blgy. PELOBATIDA, Pelodytes caucasicus, Blyy. *Pelobates fuscus, Laur. - syriacus, Bttgr. . BuFONID&. *Bufo vulgaris, Laur. » formosus, Bler. » viridis, Laur. » raddii, Strauch » mauritanicus, Schleg. » regularis, Reuss HYLips. *Hyla arborea, L. » stepheni, Blgr. . RANID A. Rana boulengeri, Gthr. » Umnocharis, Boie » rugosa, Schleg. * |, esculenta, L. » plancyi, Lat. * 4, arvalis, Nilss. . * 4, camerant, Bigr. >» macrocnemtis, Blgr. . * ,, temporaria, L. » amurensis, Blgr. » martensit, Blgr. e japonica, Blegr. * 4, ayilis, Thom. . » mascareniensis, D.& B. . schmackert, Bttgr. . Rhacophorus buergeri, Schleg. yi davidi, Sauv. " schlegelti, Gthr. A. + +: B. +++ D+: ++ ++: +4+++ ++) ++ ++ Pi + 121 ++ +++ [Aé page 123. 70 \N NS NY SS \\ WO ey SS is \ DISCOGLOSSUS PICTUS. V/Z7, BOMBINATOR IGNEUS. ” PACHYPUS, ALYTES OBSTETRICANS, [oe CISTERNASII. DISTRIBUTION OF EUROPEAN DISCOGLOSSID&. 50 DISCOGLOSSIDA. 123 Order ECAUDATA, Four limbs and no tail. Radius and ulna, and tibia and fibula confluent; tarsus (astragalus and caleaneum) elongate, forming an additional segment in the hind limb. Frontal bones confluent with parietals. Sub-ordcr PHANHROGLOSSA. Hustachian tubes separated ; tongue present. Series AW—ARCIFERA. Pectoral arch with the opposite halves moveable, the coracoids and precoracoids connected by an arched cartilage (the epicoracoid), that of the one side over- lapping that of the other. Family 1.—DISCOGLOSSIDA. Vertebree opisthoccelous ; short ribs articulated to the anterior diapophyses; diapophyses of sacral vertebra dilated. Upper jaw toothed. The genera combined under this family constitute ‘a most interesting and perfectly natural group, as 1s abundantly evidenced by the bony structure, the larval characters, &c. That they occupy the most lowly position among the Heaudata, and show the nearest approximation to the Caudata, is another point on which there can be no question. The opisthoccelous vertebree with distinct ribs, the increased number of carpal and tarsal ele- ments, the non-extrusible tongue, the presence of the azygos (posterior cardinal) vein discovered by Hoch- stetter in Bombinator, and since shown by Howes to be 124 DISCOGLOSSIDE. fairly distinctive of the whole family, the structure of the urogenital apparatus, together with other cha- racters, give ample foundation to this proposition. This family comprises only four genera. Three are confined to the Palwarctic region and represented in Hurope; the fourth, Liopelma, Fitzinger, closely allied to Alytes, is the only representative of the Batrachians in New Zealand. The range of the species dealt with in this work is shown on the accompanying map. The European genera are distinguished as follows : Tympanum present, distinct or hidden; pupil roundish or triangular; diapophyses of sacral vertebra moderately dilated . 1. Discoglossus. Tympanum absent; pupil roundish or triangular ; diapophyses of sacral vertebra very strongly dilated. . . . . . . . 2. Bombinator. Tympanum distinct ; pupil vertical; diapophyses of sacral vertebra strongly dilated. 3. Alytes. The relationships of these three genera cannot be well expressed in a linear arrangement LDiscoglossus is unquestionably the most generalised, and Bombinator and Alytes are almost equally related to it, the latter being, on the whole, more affine to Discoglossus than to Bonbinator, especially with regard to the osteological characters. 1. DiscoaLossus. Otth, Neue Denkschr. allgem. Schweiz. Naturf. Ges., i, 1837, p. 6. Pupil roundish or triangular. Vomerine teeth in long transverse series behind the choana. Tongue circular, entire, scarcely free behind. Tympanum more or less distinct or concealed under the skin. Fingers free, toes webbed ; outer metatarsals separated by web. Diapophyses of sacral vertebra moderately dilated. Urostyle articulated to two condyles. A single species, confined to the western parts of the Palearctic region. DISCOGLOSSUS. 125 The pupil is described as roundish or triangular ; in its contracted condition it varies considerably according to individuals. It is never absolutely round, for it forms an angle below, whence a subtri- angular form results, I have seen specimens in which the vertical diameter somewhat exceeds the horizontal, so that the fully contracted pupil might be termed vertically pear-shaped, or, the upper border forming an open angle, kite-shaped. 1. Discocossus PIoTus. (Plate IV.) Cetti, Anf. e Pesce. Sard., iii, p. 38 (1777). Rana temporaria (non L.), Rozet, Voy. Alg., i, p. 230 (1833). Discoglossus pictus, Otth, N. Denkschr. allgem. Schweiz. Nat. Ges., i, 1837, p. 6, figs.; Tschudi, Class. Batr., p. 80 (1838); Bonaparte, Icon. Faun. Ital., Rett. Anf. (1858); Duméril & Bibron, Erp. Gén., viii, p. 425 (1841); Ginther, Cat. Batr. Sal., p. 35 (1858); Strauch, Erp. Alg., p. 77 (1862) ; De Betta, Atti Ist. Venet. (3), xiii, 1868, p. 77, and Faun. Ital., Rett. Anf., p. 67 (1874); Schreiber, Herp. Eur., p. 112 (1875); Camerano, Atti Acc. Torin., xiv, 1879, p. 4438, pl. —, figs. 6—8; Lataste, Act. Soc. Linn. Bord., xxxiii, 1879, p. 275, pls. ii—v; Bedriaga, Bull. Soc. Nat. Mosc., 1881, p. 292; Boulenger, Cat. Batr. Ecand., p. 445 (1882); Camerano, Mem. Ace. Torin. (2), xxxv, 1883, p. 204; Bedriaga, Arch. f. Nat., 1883, p. 254; Héron-Royer, Bull. Soc. Zool. France, 1885, p. 565, pl. xiv; F. E. Schulze, Sitzb. Ges. Naturf. Fr. Berlin, 1886, pp. 5 and 31; Héron-Royer & Van Bambeke, Arch. Biol., ix, 1889, p. 280; Bedriaga, Bull. Soc. Nat. Mosc, 1889, p.545, and Amph. Rept. Portug., p. 22 (1889); Héron- Royer, Bull. Soc. Et. Sc. Angers (2), xix, 1889, p. 160, pls. i and ii; Boulenger, Trans. Zool. Soc., xiii, 1891, p. 160, and Proc. Zool. Soc., 1891, p. 620, pl. xlvii, fig. 3; Anderson, Proc. Zool, Soc., 1892, p. 24; Bedriaga, Suppl. Amph. Port., p. 13 (1893); Mina-Palumbo, Nat. Sicil., xui, 1893, p. 262. Discoglossus sardus, Tschudi, in Otth, 1. ¢., and Class. Batr., p. 80; Bonaparte, l. c.; Camerano, Atti Acc. Torin., xiv, 1879, p. 437, pl. —, figs. 4 and 5. Pseudis sardoa, Gené, Syn. Rept. Sard., p. 24, pl. v, figs. 1—3 & 6 (1839). Rana picta, Schlegel, in Wagner, Reisen Alg., iii, p. 134 (1841). Colodactylus czerulescens, Tschudi, Faun. Per., Herp., p. 68, pl. xi, fig. 2 (1845); Peters, Mon. Berl. Ac., 1873, p. 414, pl. iii, fig. 1. Discoglossus scovazzit, Camerano, Atti Acc. Torin., xiii, 1878, p. 548, and xiv, 1879, p. 447, pl. —, figs. 1—8. Discoglossus auritus, Hévon-Royer, Bull. Soc. Zool. France, 1888, pp. 205 and 220, and Bull. Soc. Et. Sc. Angers (2), xix, 1889, p. 177, pls. i and ii. 126 DISCOGLOSSID. Vomerine teeth in a long, straight or slightly curved series, narrowly interrupted in the middle, behind the choanz, and extending outwards to the Fra. 48. vertical of their outer borders or even a little beyond. Tongue moderately large, circular, thick, adherent, scarcely free behind. Head much depressed, a little broader than long; snout rounded or obtusely pointed, projecting considerably be- yond the mouth, as long as or a little longer than the diameter of the orbit ; no canthus rostralis; dorsal region slightly grooved ; nostril a little nearer the tip of the snout than the eye; eye moderate; interorbital space as broad as or a little narrower than the upper eyelid, and equal to the distance between the nostrils; tympanum measur- ing three-fifths to two-thirds the diameter of the eye, sometimes feebly distinct, often completely hidden. Open mouth. Fie. 49. Heads of males, upper views. a. Corunna. B. Oran. c. Tlemsen. p. Sardinia. DISCOGLOSSUS. 127 Fingers rather short, obtusely pointed, first shortest, third longest, second and fourth equal; no sub- articular tubercles ; three palmar tubercles, the inner (rudiment of pollex) largest and very prominent, the two others flat and close together, at the base of the third and fourth fingers. Hind limb rather long; the tibio-tarsal articulation reaches the tympanum, the eye, or the nostril; tibia longer than the femur, the heels overlapping when the legs are folded at right angles to the rhachis. Foot a little shorter than the tibia ; toes rather slender, one-fourth or one-third webbed in females and young, three-fourths or nearly entirely in adult males; no subarticular tubercles; no tarsal fold ; asmall rounded inner metatarsal tubercle, the length of which equals one-third to one-half the length of the inner toe. Skin remarkably shiny in life, smooth or with small warts or short glandular folds on the back ; a more or less developed glandular lateral fold from the eye to above the shoulder, often prolonged on the side of the body to the inguinal region; a fold of the skin may be present across the occiput behind the eyes. Lower parts smooth except the thighs, which are granulate near the vent; throat and belly sometimes with isolated small granules. Coloration very variable. Pale brown, grey, greyish olive, yellowish or red above, uniform or with dark, often light-edged spots; some specimens with a broad yellow vertebral stripe, which may be bordered on each side by one or two dark stripes; the lateral glandular fold light, often reddish or golden; a dark streak on each side from the tip of the snout to the eye, and a dark temporal blotch or streak ; limbs with transverse dark spots, sometimes forming regular cross-bands; a more or less distinct light triangular or heart-shaped spot often present on the middle of the back between the fore limbs, and another, more in the form of a streak, on the coccygeal region. The dorsal spots sometimes small and irregular, some- 128 DISCOGLOSSIDA. times forming handsome symmetrical markings; alarge triangular, heart-shaped, or chevron-shaped marking is usually present between the eyes. Lower parts ivory-white, uniform or closely speckled with brown, yellowish or more or less carneous under the limbs. Iris golden in its upper portion, dark bronzy in its lower two-thirds, the dark lateral streak of the head passing through the eye; sometimes entirely bronze- brown, as in Bombinutor, with a fine golden border to . the pupil, interrupted at the lower angle. Fia. 50. Lower view of male with nuptial excrescences. Male distinguished from the female by much stronger muscular fore limbs, more devcloped inner carpal tubercle, a great enlargement and flattening of the inner finger, and more fully webbed toes, the web even extending to the inner metatarsal tubercle. Vocal sacs are ina rudimentary condition, and do not com- DISCOGLOSSUS. 129 municate by any openings with the floor of the mouth. Horny blackish excrescences are largely developed and distributed as isolated minute spines over the ventral surfaces and on the hind limbs; they form large groups on the inner palmar tubercle and the two inner fingers, a band round the chin, and often border the web of the foot. These excrescences usually persist longer than in most other Batrachians, being found, more or less developed, in adult specimens all the year round. GEOGRAPHICAL Variations.—The polymorphism of this species has given rise to the establishment of various species and sub-species, which were believed to be restricted to certain parts of the habitat of the genus. Thus a Discoglossus sardus, Tschudi, was stated to be peculiar to Sardinia and Corsica, and a D. scovazzii, Camerano, was described from Morocco. Lataste, in his monograph of 1879, had, it seemed, disposed once for all of these supposed species by showing the inconstancy of the characters adduced for their separation ; and, for my part, after examin- ing avery large material from almost every part of the habitat, I fully agree with him. But, since 1879, the D. sardus has been maintained, no longer as a species, but as a sub-species, by Camerano, and the D. scovazzii, or a form very closely agree- ing with it, has been revived by Héron-Royer under the name of D. auritus; it may therefore be well to explain in a few words why I take no notice of them even as varieties. According to Camerano’s latest definition, D. sardus, from Sardinia, Corsica, and neighbouring small islands, differs from the typical D. pictus from Sicily, Malta, and the Spanish Peninsula in having a less acuminate snout, more robust limbs, and the length of the tibia contained twice, or a little over twice, in the length of head and body. The two former characters are really too slight to be easily appreciable ; and besides, I find Maltese, Algerian, and Portuguese specimens with the snout I 130 DISCOGLOSSIDE. quite as short and broadly rounded as in the Sardinian form. As already observed by Lataste, this difference is no greater than that between specimens of Rana temporaria, and is certainly not more constant. With regard to the third character, I need simply refer to my tables of measurement and to Camerano’s own table, where Sardinian as well as Sicilian specimens are shown to have the tibia less than half the length of head and body. Bedriaga also is inclined to retain D. sardus as a distinct form, although admitting that the existence of annectant specimens renders a precise definition impossible. On the whole, he finds the body shorter in proportion to the head and limbs in Corsican and Sardinian specimens than in those from Algeria and Portugal. That there is nothing constant in that supposed difference may be seen from the following measurements (in millimetres) of four adult males, the heads of which are figured above (p. 126),—a from Corunna, 6 from Oran, ¢ from Tlemsen, prov. Oran, and d from Luras, Sardinia : a. b c. d. From snout to vent : BS .. 62 .. Tl «x 60 Length of head ‘ i... 19... 20... 18 Width of head at angles of mouth TO: 22: OE ee BB va DL ig 5, below eyes . » 16 .. 17d. 20 ., J8 From end of snout to fore limb OF ses D250 en LO ges “DO Fore limb 5 ‘ 84... 850... 850. 85 Hind limb. : 83 am 95 cs 92 a. 91 Tibia . ‘ : » O28 wn SF ne 32 ae 30 It will be noticed that the Spanish and Algerian specimens a and J have the head a little longer in proportion to the body than in the Algerian ¢ and the Sardinian d; that the width of the head below the eyes is as great in proportion in the Algerian ¢ as in the Sardinian d; and that the hind limb is shorter in pro- portion in the Algerian c than in the Sardinian d. Héron-Royer’s D. awritus is founded on Algerian specimens (which had been previously referred by Camerano to D. scovazaii), supposed to differ from the Europeans in having the tympanum distinct, the tem- DISCOGLOSSUS. 181 poral spot larger, and the inner metatarsal tubercle smaller. Buta male specimen from Algiers, collected by Mr. Sclater, has the tympanum completely con- cealed ; and, on the other hand, the organ is perfectly distinct in a male from Corunna and in another col- lected by M. Boscd on the Sierra Morena; whilst a number of specimens from Europe and Africa are in- termediate between the two extremes. I have now before me a large number of living specimens from Oran. In some the tympanum is very apparent, whilst in others its presence can hardly be detected. The shape and extent of the dark temporal band and the size of the metatarsal tubercle are subject to variation in specimens from the same locality, and I have failed to find any constancy in the other very trivial distinc- tive characters pointed out by Héron-Royer. Dr. Anderson, who has collected specimens in Algeria, some of which show no trace of a tympanum exter- nally, concurs with me in rejecting D. auritus even as a local form. Specimens from Sardinia, Corsica, Montecristo, and Giglio, it has been observed, never show the striped form so frequent in Spain and Africa; but neither do the Maltese specimens, which are referred by Camerano to D. pictus. Measvrements (in millimetres). rah g 1. 2 3. 4 5. 6. 7. 8. From snout to vent 58 ...55 ...60 ... 73 ...55...54...43 ...76 Length of head 18 ...16 ...18 ... 21 ...16...15...14 ...19 Width of head d 19 1.18 ...21 ... 25 ...19....16...15 ...23 Diameter of eye Bsc Be Dee Cay Bis he Blom 6 Interorbital width . Sb Dena A ace AO os Bow Baw BD ve D From eye to nostril 26.4 45. 5b 4 4 4a 8 _ 3 endofsnout . 8 ... 75... 85... 10 0. 7... 7... 6 2.9 Fore limb . ; 84 ...29 2.85 0. 40 ...27...26...24 ...85 Hind limb . 83 ...80 2.91 1.105 ...84...77...64 22.95 Tibia ; 28 ...26 ...80 ... 85 ...26...25...20 ...83 Foot . ; ‘ 26 1.25 ...26 ... 82 ...24...22...18 ...80 1, 5. Corunna: Seoane. 4. Oran: Chevreux. 2,6. Coimbra: Gadow. 7. Giglio: Florence Mus. 8. Luras, Sardinia: Camerano. 8. Algiers: H4ron-Royer. 132 DISCOGLOSSIDZ. SKELETON.—Ethmoid short, not extending poste- riorly beyond the anterior third or two-fifths of the parasphenoid, embracing above a large fontanelle which extends to the line of the anterior borders of the orbits, and which remains exposed between the diverging anterior borders of the fronto-parietals ex- cept in very old specimens ; its upper lamina obtusely pointed and partly covered by the nasals, which are large and form a long suture in the median line. Fronto-parietals large, in contact along their posterior two-thirds or more. Zygomatic branch of the squa- mosal joining by suture an ascending process of the maxillary. Vomers large, their posterior toothed border nearly straight, and covering the feeble pala- tines, narrowly separated from each other in the median line; pterygoids regularly trifurcate, with a small round outer wing-like expansion at the meeting- point of the three branches, joining a corresponding process of the angular bone of the mandible; the inner branch of the pterygoid in contact with the parasphenoid, which is L-shaped, obtuse or truncated, or indentated anteriorly, and not quite reaching the vomers. Mento-Meckelians quite indistinct. Hyoid a large, broad cartilaginous plate with rounded lateral wings, small postero-lateral processes, and proximally slender, mesially broad and lamellar ceratohyal cornua without forward processes ; in addi- tion to the rather large thyrohyals, which meet with their inner angles, there is a V-shaped, slender, ventral ossification, which is in contact with the latter at its apex, and reaches forward to the anterior notch of the cartilaginous plate; this ossification often paired, the two branches being disconnected at the apex. Vertebral column twice to twice and one-third as long as the skull. Vertebre imbricate and completely covering the spinal cord above, with a long postero- median neural process. Second, third, and fourth vertebre with short diapophyses, +0 which short ribs are attached. The first rib is the shortest, horizontal 133 DISCOGLOSSUS. Whole skeleton of female and fore limb of male. 184 DISCOGLOSSIDA, or directed a little forwards ; the second is the longest, nearly horizontal or directed a little backwards, and sends off from the middle of its upper border a short, slender process, directed outwards and backwards ; the third rib, which may be somewhat curved, is directed slightly backwards, and in length is intermediate be- tween the first and second. The following vertebra have no distinct ribs, but the diapophyses are longer and more slender, the last two or three being turned more or less distinctly forwards. The diapophyses of the sacral vertebra, which cover the anterior extre- mities of the ilia, on which they may shift forwards and backwards to some extent, are moderately dilated, their diameter at the free extremity being somewhat less than their length. Two condyles for articulation with the urostyle. The latter is a little longer than the rest of the vertebral column, and bears on each side at its base a short, slender, transverse process directed obliquely backwards. Preecoracoids slender, curved, entering the glenoid cavity; coracoids stronger, feebly curved; supra- scapula ossified ; omosternum cartilaginous; sternum produced into two long, slender, diverging processes. Humerus once and a half to once and two-thirds as long as radius-ulna. Carpus with eight bones, two of which are in contact with radius-ulna ; a single bone in the pollex, very large in the male ; metacarpal of inner digit enlarged in the male, with strong inner crest. Pelvis nearly two-thirds the length of the vertebral column ; pubis cartilaginous ; acetabulum open. Femur and tibia with the epiphyses cartilaginous, the former strongly sinuous and shorter than the latter; astra- galus a little longer than calcaneum, from which it is perfectly distinct throughout, owing to the absence of calcified epiphyses, and one-half or rather more than half the length of tibia-fibula; three tarsalia in the distal row, one or two of which may ossify ; and two bones to the prehallux. Distal phalanges obtuse, slightly expanded at the apex. DISCOGLOSSUS. 185 Measurements or Skeeron (in millimetres), Length of skull. : § — is Width of skull ; ‘ : 24 6 17 Least interorbital width 4 va 3 Dorsal vertebral column 18 ms 17 Urostyle . : 19 ae 19 Humerus . 17 Ge 12 Radius-ulna 11 me 8 Manus : A: ee 11 Pelvis ‘ 7 3h 23 Femur 26 mde aul Tibia 31 an 24. Tarsus 17 she 13 Pes i 25 21 Hasirs.—Discoglossus pictus resembles the true frogs im the quickness of its movements. It is active by day as well as by night, and usually found in or about water, objecting neither to brackish pools nor to cold running mountain streams. Like its relative Bombinator, and unlike all other European Hcaudates, it is able to seize its prey under water, in the same manner as newts do, and is therefore easily fed in confinement with earthworms dropped into the tank with which the terrarium must be provided. Although particularly shy when handled, it does well in captivity, and will usually take food immediately on arriving after a long journey in a-small box. When seized it becomes covered with a slimy secretion, which renders it very difficult to handle, but which has no special odour nor any irritating action on our mucous membranes, so far as I can judge from experi- ments upon myself. Like Lataste, who was the first to study the habits in captivity, and who later had the good fortune to observe this Batrachian in a wild state in Algeria, I have never heard the rutting male produce more than a feeble note, sounding to my ear as ha-a, ha-a-a, or wa, wa, wa-wa-wa, issued in rapid succession, although specimens have repeatedly paired and bred in my terrarium. But according to Héron- Royer, specimens under his observation uttered at night a constantly repeated.ra-a, ra-a, loud enough to disturb him when produced in a yard outside his 186 DISCOGLOSSID. window. Discoglossws cannot be described as mute, but its voice is certainly the feeblest among all our members of the tailless tribe. The pairing season extends from January, in Algeria, to September and October, and the female is able to spawn three times in the year. The embrace, which never lasts long, is lumbar, the Fie. 52. Male and female pairing. male’s hands joining on the pubic region. The re- markably small eges are produced slowly, one at a time, and feebly adhere to the bottom of the pool or tank, where they form a single layer, more or less closely set. Only exceptionally have I found eggs attached to the weeds with which I had furnished the small tank in the terrarium, the pair keeping to the bottom during oviposition. The spermatozoa, which from their large size can be detected with the naked eye, are ejaculated in several bundles. The larve, as a rule, live only from one to two months before transforming, and keep small, being in this respect comparable to those of the common toad ; but tadpoles born late in the summer may remain for nine or ten months in that condition. Specimens which I received from Algeria in January and February, 1897, although in apparently excellent health and readily feeding on worms and insects, did DISCOGLOSSUS. 137 not begin to pair, so far as I could observe, until the beginning of June. The first eggs, about 400 in number, I found on the bottom of the tank on the morning of June 7th, the embryos emerging on the 10th; whilst on the 14th, or exactly a week later, the young had lost their external gills and entered the tadpole stage, measuring 11 mm. total length, buds of the hind limbs being discernible on the 17th. On the 23rd the largest specimens measured 21 mm., and on the 3rd July some had reached their full length, viz. 25 to 30 mm. On the 5th July some had put on the striped livery. The fore limbs first appeared on the 6th, and the metamorphosis was completed by the 8th, when the first specimens left the water with a reduced caudal appendage, and measuring 8 to 10 mm. from snout to vent, the last metamorphosing as late as August 6th. Striped speci- . mens were in the proportion of one to four spotted or immaculate. On the morning of June 14th the tank again con- tained eggs, 895 in number, which, from their con- dition, must have been deposited in the night of the 12th—13th (I was absent on Sunday, 15th), and the embryos of which were liberated in the afternoon of the 14th, thus less than forty-eight hours after the deposition of the eggs—a process of development more rapid than has been observed in any other species of Batrachians. The young from this brood began to leave the water on July 20th, whilst others retained the larval condition until October 5th. None were of the striped form. A third lot of 936 eggs were deposited during the night of July 6th—7th; they hatched in the night of 8th—9th, and the young completed their metamorphosis between August 16th and October, a few being still in the tadpole condi- tion whilst these lines are passing through the press. About two-thirds of them belonged to the striped form. In all these cases I was struck by the ex- tremely small proportion of eggs that did not develop, 138 DISCOGLOSSIDE. and the very slight amount of mortality that took place among the embryos; whilst of the tadpoles all except such as I killed for study went through the metamorphosis. In fact, the success in rearing these broods of Discoglossus has been greater than I have experienced in dealing with any other kind of tailless Batrachians under similar conditions. I had no diffh- culty in feeding the tiny young on aphides. On the 17th July eggs had again been produced during the night, but they numbered only 306. They hatched on the 18th; most of them were destroyed shortly after bya fungus.