Memoirs MCZ 4703

/IDemotrs of tbe /iDuseum of Gomparattve Zoology AT HARVARD COLLEGE Vol XLVII No SPHAERODACTYLUS BY THOMAS BARBOUR WITH T\A^NTY-SIX PLATES CAIVIBRIDGE, U IPrtnteo for tbe December, S A.: /»useum 1921 iry^S SPHAERODACTYLUS INTRODUCTION There is scarcely a single Neotropical genus \Yhich has proved The to students than the one chosen for revision now more baffling species are often ex- tremely small, adequate series of specimens are not frequently available, and as common in generalized, forms, yet when he comes is homogeneous groups, the characters available for specific Thus while with the fresh definition prove to be variable in the extreme material in hand the reviewer has no great difficulty in separating the various to justify his conclusions conditions observed, the task becomes one of by actually describing the most acute difficulty; preparation of satisfactory keys for the determination of the species hopeless task I was to the long interest in led to attempt West Indian this revision solely by the is while the an almost fact that exploration, which has centred in the owing M C Z a series of several hundred sphaerodactyls has gradual^ accumulated This series represents no less than thkty-one species of the thirty-five which are recognized as valid including typical material of twenty species and fifteen types of valid species acknowledge with deep gratitude the coiu-tesy which I ha\e received from Drs Leonhard Stejneger, G A Boulenger, Lars G Andersson, A G I Ruthven, Mr G K Noble, and Mr H W Fowler, all of whom have aided my The quite satisfactory drawings are by Mr E N Fischer, whose skill studies is already well recognized The sphaerodactyls form a compact group of Gekkonidae highly characAll the species, teristic of the Antillean subregion and the adjacent mainland except one, are found in countries bordered by the Caribbean Sea This one the pecuhar species on Cocos Island off the Pacific coast of Costa Rica is I noticed that three others from South America differed from all the rest in that they lacked the supraorbital spine,' a character of great taxonomic importance Therefore, though they have been called Sphaerodactylus there seems good ' which It is is interesting that the possession of this unique character and not distantly related also Antillean is shared by the genus AristeUiger, SPHAERODACTYLUS 218 viz meridionalis reason to believe that these three species, nat hist., 1888, ser 6, mag nat hist., 2, 1902, ser p 40 Eublepharidae Iguarasse, Brazil), scapularis Boulenger, (Ann 7, 9, p 54 (Ark zool., 1917, 11, no 16, p Ecuador) amazonicus Andersson, St Javier, Manaos, that family be recognized if Boulenger, (Ann mag , Brazil) should be allocated to the A very large series of small lizards from northwestern Peru agree closely with Boulenger's description of scapularis and in the field Mr Noble considered them sphaerodactyls I have, however, examined the series and ascertained their eublepharid affinities and think they should be placed in Lathrogecko or Lepidoblepharis (syn Pseudogonatodes).' Sphaerodactylus Small, generalized, gckkoniform lizards having "Digits narrow, slender, free, with transverse lamellae inferiorly, the apex dilated into a disk, with a circular undivided plate inferiorly; all digits with a sheathed retractile claw, the sheath opening laterally and inwards Scales granular or imbricate Pupil round or subelliptical Eyelid nearly circular No preanal or femoral pores." (Boulenger, Cat lizards Brit, mus., 1885, The rather 1, p 217) primitive character of these lizards paratively simple digital structure, by emphasized by the com- of the digits being similarly disked, all by the eye with certainly almost always, is not always, a round pupil, in spite of if the decidedly nocturnal or crepuscular as well as sometimes diurnal habits, and in lacking preanal or femoral pores Within the genus the species show many diverse types which we may of squamation of assume, and probably correctly, that the simple granules on head and body, quite undifferentiated, represent the most primitive condition This squamation seems to be characteristic of species in which the males and females show the same gamut of colour-variations as e.g in cinereus or elegans These species are small and show great power of adaptation to environment for often they occur commonly in open clearings, in the dense forests and commonly in human habitations as well AristeUiger This also and Phelsuma The granules from being rounded granules (Uneolatus) though at the type of squamation observed in is first ; or tubercular give rise naturally to keeled very small (nigropunctatus) we see , conditions typified and become imbricate, and them increase in size through these again begin to enlarge by such species as notatus, macrolepsis, or pictus, finally leading to such giant species, comparatively speaking, as richardsonii, anthracinus, ' and copei These largest species have the largest dorsal scales not only Since described as a Lepidoblepharis by G K Noble (.\nn N Y acad sci., 1921, 29, p 133) INTRODUCTION actually but proportionately and are posed that these species are in all 219 It comparatively rare any wise ancestral the one must not be sup- to the other; but are simply examples chosen as typifying arrested stages pointing out the path by which such a form as anthracinus, like cinerexis at first sight similarity is happen tobe for instance, may have been derived from one The apparent apparently wholly unrelated dis- bridged over by a series of hypothetical ancestral intergrades which by various scattered species In anthracinus (syn asper) and scaber there is a distinct sexual dichromatism,the males illustrated copei fsyn picturatus) , being monochrome, the females fantastically banded and decorated with ocellate How^ever, a similar dichromatism exists in decoratus and torrei spots among the species having granular scales A condition similar to this has been independently arrived at in some species of Gonatodes, far from closely Jamaican richardsonii for number both I This apparently have traced eight specimens, sexes were probably represented closely related to the three others but and allied in lacking the is It is found that species with very large scales, there running down the middorsal region is (St pidus of this is in colouration, probably, in several, indeed, in most of the a narrow zone of fine granular scales This probably has been at least three is rather wide and not very Vincent) and macrolepis (Martinique) shows no sign of a zone Antilles, banded and all This curious character has not times independently acquired in that the zone sharply defined in lincenti not true of the Structurally this species more primitive middorsal zone of granules been mentioned previously is (St Kitts), sputator (St The species (St Lucia) ; while festus from the Upper Lesser Eustatius) and elegantulus (Antigua) have a very narrow zone, sharply defined although the body-scales are not so very large In Porto Rico, Mona, and the Virgin Islands (macrolepis), the local species although large scaled, has no middorsal zone On the other hand, in Cuba, Haiti, and Andros Island there are very large species with very large and anthracinus respectively which have the middorsal defined Thus it may be observed that the character appears scales, viz scaber, copei, granular zone well in three separated groups of species No explanation has been offered suggesting why this character should be acquired It seems reasonable, however, that it may have to with the flexibility of the body The large overlapping scales tend to form a sort of cumbrous armour and the expansion and contraction of the body connected with breathing may be facilitated by this zone of small, nonimbricating scales lying upon a flexible integument Analogous, but more obviously necessary, are the lateral grooves of soft skin or fine scales seen in SPHAERODACTYLUS 220 certain anguids, without which respiration would be impossible owing to the integumentary armament memoir was prepared for publication some time ago it behooves comment supplementally on Noble's valuable paper The bonj'- structure Since this me to and phyletic relations of Sphaerodactylus description of a March, new genus published and in the allied lacertilian genera, with the American museum novitates, 10, 1921, no 4, p 1-16, figs In this paper Noble gives the results of his studies of the skeletal features number of a of gekkoid genera made with the aid of cleared preparations With most heartily agree, namely that the Eublepharidae constitute an unnatural assemblage and it is evident that the family serves no one of his conclusion I an omnium gatherum combining the end results of descent not closely related Thus the family "instead of being useful purpose, being simply several lines of a very ancient group as hitherto believed, even if be a very recent assemblage, may a conservative one." Noble derives Sphaerodactylus from a genetic series of which Gonatodes, Lathrogecko, and Lepidoblepharis represent successive stages of modification Various internal characters are cited and illustrated to support the evidence derived from a comparative study of the digital apices; this persuasively and while stress is it may be correct it is by no means not placed upon fortuitous similarity modifications are very easily brought about many through very I is presented very certain that too much believe that often skeletal by changes in feeding habits or other causes and often even \^hat are apparently such may be very conservative In other an organism may offer, within reasonable limits, ephemeral features as colour-characters words the external facies of better criteria for postulating relationships than certain skeletal features Noble's work on the claw-sheaths is very valuable, and while here again it possible that similarity maj' not indicate actual phyletic relationship neverthe- less and is very considerable that some such descent has taken place this evidence is taken in connection with the internal characters of hyoids the probabiUty when girdles It would be very interesting to know Mr Noble's views on the connection Aristelligcr \^ith this series for it is is is not close I sphac redact yls surprising in of almost inconceivable that this relationship strongly suspected this genus to be the closest ally of the That no fossils remain to cast upon the subject light is not view of the fragility of their bony skeleton The sphaerodactyls are preeminently island gekkos and they probably occur INTRODUCTION 221 upon nrarly every island in the Caribbean area Central America and the species are few Of South America we know little They This, considering the collecting which or nothing if hardly an island, so recent has been its The two not absent.^ dadian species, molei and buergeri, are very scarce is Mexico and their distribution in northern has been done, indicates that thej" are very rare dad are rare in Trini- Indeed, zoologically Trini- No continental connection sphaerodactyls have been fomid in the Leeward group composed of Aruba, Buen and Curasao; nor on Margarita nor yet on Old Providence, St Andrews, or the Corn Islands, but many of these locations have hardly been touched Aire, herpetologically In the Greater Antilles and in the Bahamas as well as in many of the Lesser which we can speak with authority, they fairly swarm Sphaero-dactylus argus in Jamaica occurs in every native hut on the Luguanea plain about Kingston, under every stone wall, under fallen leaves, and dry vegetable trash Antilles of numbers and other in great species elsewhere are very common In general, the species are confined each to a single island; the exceptions occur mostly in the Bahamas or Here many human migration may serves special mention but recentlj' separated from each other be playing some part This species, first Sphaerodadylus notatus de- made known from Key West, Bahamas and on Cuba, the very have been given a name (difficilis) which on most viduals of the islands are of the occurs shghtlj^ ^'arying Haitian indiis, I think, probably justified had been usually believed that notatus was of purely fortuitous Whereas occurrence in the city of Key West, further investigation shows that the species it occurs widespread in the Florida Kej^s and upon the mainland southern tip of the peninsula of Florida as well case of chance transportal, but whether wind and wave cannot how be told in the This beyond doubt, is extreme a definite by human agency or by the action Nor is it of surprising, for zoogeographers have been inclined to believe that no weight could be given to evidence based on the existing dispersal of small gekkonids and scincids dence among There is very clear e\Ti- the islands of the Pacific that small lizards of these famiUes have been carried hither and yon beyond doubt, by the wide-sailing Polynesians the Pacific Islands, however, one finds the isolated islands distributed far same species widely scattered and wide throughout an immense area In upon The ' Boulenger (Cat lizards Brit mus 1885, 1, p 224) records a specimen of what lie calls S.fantaslicus from Caracas The locality record has never been verified and although the description is in rather vague terms and there is no indication to show whether Boulenger's description was drawn from this specimen or not, it nevertheless reads like fantasticus from Guadeloupe Boulenger's other specimen listed is from Antigua and is probably eleganlvlus SPHAERODACTYLUS 222 exact reverse lay West Indies and while we the case in the is upon the evidence stress any very great of Sphaerodactylus as bearing a relation to the past geographic changes in the region, that the distribution defined not at aU haphazard The it is only fair to point out species are generally well and bear evidence that no frequent interbreeding by fresh stock through flotsam and jetsam has taken place This is as isolation by long the arrival of evident on is are not constrained to little Navassa, for example, where the species population must be very small, as upon Cuba where it is numberless It is, therefore, reasonable and conservative to state that with the exception of notatus (which has obviously been carried about and by the condition in the Bahamas as yet very incompletely known), the distribution of the Antillean sphaerodactyls bears no evidence within itself of being the result of fortuitous dispersal recorded from certain islands, Archipelagos, it Saba, Redonda, Nevis, the Turks and Caicos e.g should be strongly emphasized that these islands are as yet very known and imperfectly While sphaerodactyls are not there found to occur, especially The majority is no reason to suppose that the genus will in the northeastern Lesser Antilles of the species are found more abundantly upon the coastal plain of the various islands than in the interior highlands Dense to offer a less suitable environment than open, cultivated, and inhabited Nevertheless, while collectmg Peripatus near the Mountains of Jamaica, I human far more crannies of walls often, certainly probably with the number of individuals in now was ever Thus in other situations abundance behind pictures on the and Blue in the seen habitations offer so safe and favorable an environment that now they are rarely seen furniture, in chinks lands found Sphaerodactylus goniorhynchus abundantly under species have found that cinereus occur in great forest appears Cuna Cuna Pass rotten logs to the very summit, but neither argus nor richardsonii Some not be wall, and wpinscoting and elegans and behind and in in similar places the ratio of one hundred to one as compared living out of doors The fact that so many species prefer lowland to highland habitat and some indeed even frequent the windrows able that of pebbles Little of interest such colourless hves movement makes it the more remark- can be written regarding the habits of creatures which They are, of course, Indian natives that this is very great beautifully silky quality in tiny jewels at high-tide mark, the species are so definitely fixed in their distribution all many West and seaweed is harmless though one cannot persuade so Their beauty of form, of skin, and of The integument life live and the eyes They move about with funny of many of these little species has a peculiar little and creatures gUsten like gliding jerks and pass readily INTRODUCTION 223 over smooth surfaces both perpendicular and overhanging Cuba they in often come to one's reading table creeping daintily to within the sphere of lamplight to catch httle moths and flying ants the name Wood-slave is upon Porto Rico; and all In coimtry houses generally applied in In the British Islands while in Spanish ; it is Santa Lucia Cuba, Salamandra, Salamanquita, or Salamanquesa, obviously based upon the assumption that there is some relationship with the The tiny elegans is suggested that this name is due to the habit of hiding behind pictures on the wall, Salamander and in the Sabandija country in is Cuba the called Salamanquita de only picture is It la Vii-gen The name often of the Virgin used in the mountains of Oriente and Central has been Cuba applied to various species Sphaerodactyls seem in general to lay but a single white egg slightly oval, it is covered with a hard, brittle chalky shell Rouiid or The habit of many species of laying in the old termite galleries of rotten logs as well as in other dark moist situations would again seem to enhance the ease with which fortuitous means might act upon the genus of transport Nevertheless, possibly because the eggs are delicate and sensitive to salt water or any other disturbance, they certainly seem most fragile, there is no evidence to be gathered from the dis- from the regular typical West Indian forms Their whether the dispersal seems to conform to certain more or less known rules reason for this homogeneous dispersal be found in the even action of flotsam and tribution which sets jetsam bringing all them off — the same immigrants to all the islands as some postulate or in previous changes in the geographic form and relations of the islands as seems much more plausible A word as is laid methods pursued in the body of this revision Great stress on the number of dorsal scales which counted along one single row, equal to the the distance from the tip of snout to the centre of eye The best method in making identifications is to verify the count given in the diagnosis by comparing and counting the scales in the figure of the dorsal scalation of each species Thus the method of making the count can be checked and then applied to an examina- tion of the particular specimen in hand Where there is no occasion to revise or to change the allocations to synonymy made by Boulenger in the Catalogue of Uzards in the British museum (1885, no allusion is made to them and they are accepted as proposed Differences opinion and synonyms subsequent to 1885 are given in It has been found extremely difficult 1) of full to find stable characters, those given in the specific descriptions, although their inclusion is many of sanctioned by SPHAERODACTYLUS 224 The median groove are really generic and not specific time-honored usage have no diagnogtic value whatsoever in the rostral, the superciliar j' spine for instance characters yet their mention in specific descriptions has always been customary The conditions in the supralabial series are so variable that So placed on characters apparent there though also, no reliance can be in a less nimiber and shape of the scales separating each supranasal from opposite side, subject to is much The shape variation of its degree, the fellow of the head and body is useful to observe, although the tail varies widelj' in proportion to the bodj'-length generally longer and Still it is much more best characters are to be found in the of far more value is The in males form, and arrangement of the scales of Lines be manj^ or few or the whole series may than than has previously been supposed in diagnosis features of pattern are strongly fixed these m females Colour also as becomes daily more evident, both the dorsal and ventral surfaces is size, slender break up into Many may may be absent but even a remnant series of spots ; Ukely to be verj- definitely located with reference to the original complete pattern Thus also markings, such as the spectacles so often seen or sacral regions, haphazard, but may is or may not be present but very distinctly occur in the Greater Antilles and among the two extremes of the fixed The present their location if is not brilliantly cross-banded species Bahamas and genus on scapular curiously enough are found They are either species with tiny granular enormous tectiform dorsals and among these two divergent groups There are probably sex-linked dichromatism has become well established scales or with many more sex-Unked colour-characters than are now Their recognized dis- covery will further reduce the range of apparently fortuitous variability of each species somewhat surprising fact to observe how close is the correspondence between the distance from the tip of snout to the ear and the length of the fore It is a Umb If long Whether may the first distance this fact is longer than usual so also the fore limb is likewise has any special significance does not appear be some relation between the length of the arm and There a corresponding head- makes more convenient the picking of small objects from the ground with the mouth or by using the thick fleshy and only slightly extensible tongue length which While one occasionally sees a sphaerodactyl struggling to subdue a its wings flapping a most inconvenient mon to find them eating ants if ineffectual protest, Little ants swarm far moth, more com- in houses in the tropics as everyone who has lived there knows to his sorrow, and account for the lizards frequenting houses too it is little some degree Ants creeping over smooth this maj' in PLATE 23 PLATE Sphaerodactylus pictus Garman Cf Plate 8, fig C Z 6,071 Sphaerodactylus becki Schmidt Fig 5-8 Fig 1-4 M Amer Mus Nat Hist., 12,595 Mrs William Beutenmiiller, del Navassa 23 Mem Mus Comp Zool Sphaerodactylus HELIOTYPE CO BOSTON Plate 23 PLATE 24, PLATE Fig 1-8 Sphaerodactylus sputator (Sparnnan) Royal Mus Stockholm Cf Plate 8, fig 3, 24 Mem Mus Comp Zool Sphaerodactylus HELIOTYPE 00 BOSTON Plate 24 PLATE 25 PLATE 26 & Liitken Sphaerodactylus microlepis Reinhardt Cf Plate 9, fig C Z 10,787 Sphaerodactylus elegantulus Barbour Fig 5-8 M C Z 12,084 Cf Plate 9, fig Fig 1-4 M Mem Mus Comp