UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN BIOLOGY JUL 1 i bov I FIELDIANA Zoology Published by Field Museum of Natural History Volume 58, No. 10 March 10. 1972 A New Limbless Skink (Reptilia: Scincidae) From Thailand With Comments On the Generic Status of the Limbless Skinks of Southeast Asia W. Ronald Hbyer* Biology Department, Pacific Lutheran University In the course of an ecological study at the Sakaerat Experimental Station, located 250 km. NE of Bangkok, 17 specimens of a limbless skink were collected. The purpose of this paper is to compare this new series with previously known limbless skinks of the region. METHODS AND MATERIALS External characteristics were compared on 24, 10, and 5 alcoholic specimens of each of the three previously known species of limbless skinks from Thailand, the 17 specimens collected from Sakaerat, and on a single specimen of a limbless skink from Australia to which cer- tain of the Thai species have been related at the generic level. Cleared and stained individuals were compared for skeletal anal- ysis of the Thai forms. A single individual was prepared from the Sakaerat series, a single individual was prepared for one previously known species, and two individuals were prepared for the other two species. ACKNOWLEDGMENTS Noel Kobayashi, Prasert Lohavanijaya, Frank Nicholls, and Sukhum Pongsapipatana, all currently or formerly of the Applied Scientific Research Corporation of Thailand, greatly aided the field work. A special acknowledgment belongs to Sukhum, whose efforts at Khao Saton produced the new series of specimens. 1 Formerly Field Research Associate, Division of Amphibians and Reptiles, Field Museum of Natural History. Library of Congress Catalog Card Number: 79-179169 Publication 1142 109 THE UBRARy of THE my 2 1972 i BIOLOGY LIBRARY ^ V r -lft-l/> , n, m IRRU L HMB VC^'y.ERSrTY OF ILI IMTIIO 110 FIELDIANA: ZOOLOGY, VOLUME 58 I thank the following curators and their institutions who allowed the material under their care to be examined (initials refer to insti- tutional abbreviations used in the text) : Harold Cogger, The Austra- lian Museum, Sydney, Australia; Prasert Lohavanijaya, Thai Na- tional Reference Collection, Applied Scientific Research Corporation of Thailand, Bangkok, Thailand (TNRC) ; and Hymen Marx, Field Museum of Natural History, Chicago, U.S.A. (FMNH) Allen E. Greer, Museum of Comparative Zoology; Robert F- Inger, Field Museum of Natural History; and Jay M. Savage, Uni- versity of Southern California, critically reviewed the manuscript, but the responsibility for accurate presentation and interpretation of the data is solely mine. Without the co-operation of my colleagues in the Biology De- partment, Pacific Lutheran University, this report would be consid- erably delayed. Especially helpful were Irene Creso and Harold J. Leraas, who by taking heavier teaching loads allowed me time for re- search. Pacific Lutheran University provided some support moneys for the completion of the manuscript. Field work in Thailand was supported by National Science Foun- dation grant GB 7845X to Dr. Inger. HISTORICAL REVIEW OF THE LIMBLESS SKINKS OF SOUTHEAST ASIA Boulenger (1914) described the first limbless skink collected from Thailand in the widespread genus Lygosoma as Lygosoma angui- noides. Lonnberg (1916) described a new genus and species for a large Thai limbless skink, Isopachys gyldenstolpei. The third species of Thai limbless skink was described by Angel (1920) also as a new genus and species, Typhloseps roulei. Smith (1935) placed all three known Thai limbless skinks in the Australian genus Ophioscincus Peters (1873), the type of which was the limbless species australis. Smith later (1937) placed the limbless species he had previously assigned to Ophioscincus in the Australian limbed genus Rhodona. Mittleman (1952) placed all four limbless species in the genus Ophio- scincus again. Taylor (1963) regarded Isopachys gyldenstolpei as ge- nerically distinct from the remaining Thai species. Taylor further indicated that he was certain of the congenerity of anguinoides and roulei and hesitatingly associated them with the genus Ophioscincus. Doubtless the cause for some of this generic confusion has been a paucity of material so that the skeletal features could not be corre- ¥ HEYER: LIMBLESS SKINK 111 lated with external features. Material is now adequate so that this may be done for the Thai forms. Unfortunately, X have been able to locate only a single specimen definitely referable to Ophioscincus austrcUis Peters (FMNH 97689). Dr. Cogger of the Australian Mu- seum sent two alcoholic and two cleared and stained specimens of an Australian limbless skink referred to as Lygosoma ophioscincus (=0. australis Peters). The species Cogger sent is at least specifically distinct from the species 0. australis. With the lack of a series of specimens clearly referable to 0. australis, the definitive relationships among the Australian and Thai species and between the genera Ophioscincus and Lygosoma remain open to speculation. Finally, Greer (1970), who had only two species, anguinoides and roulei, available as skeletons, placed both species in the genus Ophio- scincus, which he previously (1967) regarded as suspect, but lacked the material to elucidate the generic relationships. Greer (1970) placed Ophioscincus in the subfamily Lygosominae. CHARACTER ANALYSIS The first objective of the study is to determine whether or not the new series of specimens is specifically distinct from the other limbless skinks of Thailand and from the single Australian limbless skink species with which the Thai forms have been included at the genus level. Twenty- three external characters were evaluated for the five samples; 19 skeletal characters were evaluated for the four Thai forms (tables 1, 2; figs. 1-5). The Sakaerat series is unique in two external characteristics (character numbers 19 and 21) with respect to the previously known Thai forms and O. australis. In addition, the Sakaerat series differs from 0. australis in five character states of external morphology (2, 5, 15, 20, 22) ; from the three previously known Thai species in six skeletal states (25, 27, 29, 30, 34, 39) ; from anguinoides in two ex- ternal states (10, 18) and two more skeletal states (24, 26); from gyldenstolpei in ten external states (1, 2, 5, 7, 8, 9, 12, 17, 22, 23) and eight skeletal states (28, 31, 32, 35, 36, 37, 41, 42) ; and from roulei in nine external states (1, 5, 7, 11, 12, 18, 20, 22, 23) and seven skeletal states (26, 28, 32, 35, 36, 38, 41). The Sakaerat series is clearly distinctive from any of the previ- ously known Thai species as well as from 0. australis. These differ- ences are at least indicative of species differentiation. The problem now is to determine the relationships among the five species, and v. c T-H 1— I i-H 1— ( 00 2^ .ia I I I I M O "3 ^ CO O O. Q, • • iH rH CX) Tjt (3 US (I ^ tH rH U3 Tf CO I ^ CO ^ O lO ■if "5 -*" CO* Oi iH fl S i-H t- O O (M O) ^ 1-H i-( 1-1 rH (M W .5 I I I II -H O O ub CO (M o O, . . OJ ;D (M ggl ScO«5 CS >-l >H O (M CO (M»-ItH(M«D'0-<*(M to O ID S 5 in o t> -^ ^ ^ '2 •^ 00 o tH aj oo + 5 ^ o o ' fl c3 g '3 . . -rl< iH o S a> rt 2 "^ ftp. Cco'^cococotHi— ic^ic-^coco co ^oiiu^oic^ £ £ >.co^cooj C3 O, . . Oi i£> Oi a, o, I I I I CO .-I igTco U3 ■>* CO 0) I m ^ s «« 2 S..S--^ "^ « -^ ^ — ' " O to '5 CO ra -2-S 2 S r ^r f« C3 tc CO 3 .2 o o w :;3 o j2 o o ^ ;3 rt Cj ^ OT J-) Jh M +J 3 13 »H O gag b| b J5 01 a> a} M Qj t> Oj S =« ^ s^ =^ -213--;'^ J3-2i2^ "t; rt -t: c3 O) o a CO .rt M o o o o oooooooo j3 CO n £ CO O O O o o Iz; :z; :z; ^z; Pm 112 I 11 5? S 3^ 0) 0) X M « ^ >> >» £ 73 73 8°' 00 rt M ra W V ^ l_l V V, « o _, ^ CO iO U3 (3 CO t- « §11 II! b M ►h fa tJ ® — e* TO e8 o sfj •^ OJ »< Q, »< 3 35 T I ,co 00 T3 o a fl fl o>-ieieoTi»u3«ot«^o6osc> ej NooeoeoeocoeoeoeoeoeO'^ •^i" Tj« 118 114 FIELDIANA: ZOOLOGY, VOLUME 58 Fig. 1. Dorsal, lateral, and ventral views of the skull of /. anguinoides. then determine whether generic rank is warranted. The data for deducing phologenetic relationships are based upon estimation of the direction of change among or between the character states. Phylog- enies are then based upon shared patterns of the advanced charac- ter states. Marx and Rabb (1970) discuss ten criteria of directional change in character states. To apply their criteria meaningfully, one must have data available for the central group from which the smaller study group was derived. In the case of the limbless skinks of Thai- land, one must have a data set on representative members of the family Scincidae. Because of the limited nature of this study, cer- tain of their criteria, such as ecological specialization, geographic restriction, closely related taxa, correlation of applied critera, genetic structure, and fossil record, do not apply. Other criteria, such as uniqueness, relative abundance, and morphological specialization, are applicable in this study. Lack of knowledge of the character states exhibited by most skinks prevents analysis of the following numbers of characters listed in the tables at this time: 3, 4, 5, 6, 7, 10, 11, 12, 13, 14, 15, 16, 18, 19, 20, 21, 22, 26, 33, 34, 40, 41. Discussions of the evolutionary directions of each of the remaining characters follow. Characters 1 and 32 are considered together under 32. For each character, the primitive state is given a lower case letter, the derived state a capital letter, a secondary derivation a capital letter starred, a tertiary de- rivation a capital letter double starred, independently derived char- acter states are indicated by numerical subscripts. HEYER: LIMBLESS SKINK 115 Character 2 — Tail tip shape. Most skink tails taper and end in a point. Two patterns are noted in the Thai limbless forms. The first is a tapered tail ending in a point composed of a single enlarged scale, which is interpreted as one type of specialization. Living specimens of the new species would often drive the tip into one's hand, feeling much like being pricked with a probe. The second pattern is a bluntly rounded tail. a=tail tapered, ending in a point; Ai=tail tapered, ending in a sharp scale; A2=tail cylindrical, ending bluntly. Character 8 — Prefrontal scale. The common head scale pat- tern among non-burrowing skinks includes paired prefrontal scales. In gyldenstolpei, the only Thai limbless species that lacks prefrontal scales, the prefrontals are probably fused with the frontal scale. Fusion of dorsal head scales is a well documented phenomenon among burrowing reptiles. b= prefrontal scale present, B= prefrontal scale absent. Character 9 — Frontonasal scale. As character 8. nasal scale present, C= frontonasal scale absent. c=fronto- Character 17 — Tertiary temporal scale. The common skink pattern includes distinct, paired, tertiary temporal scales. In the single species, gyldenstolpei, within this study that does not have distinct tertiary temporals, paired parietal shields are present. The parietal shields result from the fusion of a secondary temporal with a tertiary temporal on each side of the head. d= tertiary temporal scales present, D= tertiary temporal scales absent. Fig. 2. Dorsal lateral, and ventral views of the skull of /. gyldenstolpei. 116 FIELDIANA: ZOOLOGY, VOLUME 58 Fig. 3. Dorsal, lateral, and ventral views of the skull of /. rouiei. Character 23 — Postmental scale. The common skink pattern is the presence of one or more postmental scales. The homologues of postmental scales may well be present in the forms not having clearly enlarged azygous scales posterior to the mental, but they are indistinguishable from the other ventral body scales. Reduction of scale size in the chin region would allow more flexibility. Combined with the specialized state of exaggerated retroarticular processes, as found in gyldenstolpei and rouiei, a change of the feeding mechanism associated with the fossorial habitat is suggested. In the specimens which had food in the gut and were prepared for clearing and stain- ing, the predominant food item was termites, which the lizards prob- ably ate underground. e= postmental scale present, E= postmental scale absent. Character 24 — Nasal bone contact. In the common skink pat- tern paired nasal bones contact each other medially. Rarely, here seen in anguinoides, an anterior process of the frontal bone separates the nasal bones. Functionally interpreted, the anterior projection of the frontal is more efficient in handling burrowing stress applied to the dorsonasal skull region in two ways: 1) Stress applied to the nasal region spreads more posteriorly throughout the entire frontal bone, and 2) a maximum of three articulations of the nasal and frontal bones rather than four allows a stronger nasal region of the skull. f= nasal bones in contact, F= nasal bones separated. Character 25— Frontal bone. Greer (1970) indicates that a paired condition of the frontal bone is a primitive condition. g= paired frontal bones, G= single (fused) frontal bone. HEYER: LIMBLESS SKINK 117 Fig. 4. Dorsal, lateral, and ventral views of the skull of the Sakaerat series. Character 27 — Lateral process of parietal. Greer (1970) indi- cates that a finger-shaped lateral process of the parietal is found only within the subfamily Scincinae. As the subfamily Scincinae is the primitive subfamily of skinks, it is logical to assume that a finger- shaped process is a primitive character state. h= lateral process of parietal finger shaped, H= lateral process of parietal broad and shallow. Character 28 — Jugal. A complete postorbital arch is the com- monest pattern among skink genera, with burrowing forms showing a reduction of the arch. The functional diminution of the eye com- bined with a streamlining of the skull for more efficient burrowing apparently provides the selective pressure to reduce the postorbital arch. All the Thai species have incomplete arches. In addition, two species have greatly reduced jugals. i= complete postorbital arch, 1= incomplete postorbital arch, moderately developed jugal, I*= in- complete postorbital arch, reduced jugal. Character 29 — Palatine bones. Greer (1970) indicates that widely separated palatine bones represent a primitive condition among skinks. He is uncertain whether the condition of narrowly separated palatines, as seen in anguinoides, is primitive with respect to palatines meeting on the midline or is a secondary derivation. j = palatines widely separated, J= palatine bones in contact, or just separated. Character 30 — Pterygoid bones. As pointed out by Greer (1970), the evolutionary trend within the skinks is toward a greater development of a secondary palate. The pterygoids, when in medial 116 FIELDIANA: ZOOLOGY, VOLUME 58 Fig. 3. Dorsal, lateral, and ventral views of the skull of /. roulei. Character 23 — Postmental scale. The common skink pattern is the presence of one or more postmental scales. The homologues of postmental scales may well be present in the forms not having clearly enlarged azygous scales posterior to the mental, but they are indistinguishable from the other ventral body scales. Reduction of scale size in the chin region would allow more flexibility. Combined with the specialized state of exaggerated retroarticular processes, as found in gyldenstolpei and roulei, a change of the feeding mechanism associated with the fossorial habitat is suggested. In the specimens which had food in the gut and were prepared for clearing and stain- ing, the predominant food item was termites, which the lizards prob- ably ate underground. e= postmental scale present, E= postmental scale absent. Character 24 — Nasal bone contact. In the common skink pat- tern paired nasal bones contact each other medially. Rarely, here seen in anguinoides, an anterior process of the frontal bone separates the nasal bones. Functionally interpreted, the anterior projection of the frontal is more efficient in handling burrowing stress applied to the dorsonasal skull region in two ways: 1) Stress applied to the nasal region spreads more posteriorly throughout the entire frontal bone, and 2) a maximum of three articulations of the nasal and frontal bones rather than four allows a stronger nasal region of the skull, f = nasal bones in contact, F= nasal bones separated. Character 25— Frontal bone. Greer (1970) indicates that a paired condition of the frontal bone is a primitive condition. g= paired frontal bones, G= single (fused) frontal bone. HEYER: LIMBLESS SKINK 117 Fig. 4. Dorsal, lateral, and ventral views of the skull of the Sakaerat series. Character 27 — Lateral process of parietal. Greer (1970) indi- cates that a finger-shaped lateral process of the parietal is found only within the subfamily Scincinae. As the subfamily Scincinae is the primitive subfamily of skinks, it is logical to assume that a finger- shaped process is a primitive character state. h= lateral process of parietal finger shaped, H= lateral process of parietal broad and shallow. Character 28 — Jugal. A complete postorbital arch is the com- monest pattern among skink genera, with burrowing forms showing a reduction of the arch. The functional diminution of the eye com- bined with a streamlining of the skull for more efficient burrowing apparently provides the selective pressure to reduce the postorbital arch. All the Thai species have incomplete arches. In addition, two species have greatly reduced jugals. i= complete postorbital arch, 1= incomplete postorbital arch, moderately developed jugal, I*= in- complete postorbital arch, reduced jugal. Character 29 — Palatine bones. Greer (1970) indicates that widely separated palatine bones represent a primitive condition among skinks. He is uncertain whether the condition of narrowly separated palatines, as seen in anguinoides, is primitive with respect to palatines meeting on the midline or is a secondary derivation. j = palatines widely separated, J= palatine bones in contact, or just separated. Character 30 — Pterygoid bones. As pointed out by Greer (1970), the evolutionary trend within the skinks is toward a greater development of a secondary palate. The pterygoids, when in medial 120 FIELDIANA: ZOOLOGY, VOLUME 58 . - 15 rou I e I - 12 gy Idenstolpe? a7l A^BCD0^*P angulnoides - I I A.FI*LM* EKNO*' new species - 7 A.I*LM* abcdefghi jklmnop Fig. 6. Proposed phylogenetic relationships among the four Thai limbless skinks. See text for details. ary trends seem apparent in the Thai forms in addition to the basic reduction displayed by all. o= functional pectoral girdle, 0= reduced pectoral girdle, with at least one pair of ribs attaching to the sternum, 0*= reduced pectoral girdle with sternum well developed, no rib attachment, 0**= extremely reduced pectoral girdle, interclavicle and sternum almost gone, no rib attachment to sternum. Character 42 — Osteoderms. Most skinks have osteoderms, but they are rarely fused together or co-ossified with the skull. This latter condition correlates with a more rigid skull for burrowing. p= osteoderms present, P= osteoderms fused over head and co-ossi- fied with the skull. The four taxa for which both external and skeletal data are avail- able may be compared with respect to evolutionary specialization by assigning the following numbers to the character states: 0= lower case letter, 1= capital letter with or without subscript, 2= capital letter with one star, 3= capital letter with two stars. The maximum index of specialization is 20. The actual totals for the taxa are: new species — 7, anguinoides — 11, roulei — 12, gyldenstolpei — 15. PHYLOGENETIC RELATIONSHIPS In constructing a hypothetical phylogeny for the limbless skinks, the character states were listed in rows by species for those charac- ters which were analyzed with respect to evolutionary direction. The rows of character states were then examined for shared advanced HEYER: LIMBLESS SKINK 121 states. Each step in the phylogeny (fig. 6) represents the maximum number of shared advanced character states for the maximum num- ber of species. For example, all four Thai species share three ad- vanced character states. An assumption made for purposes of the analysis is that once a specialization has occurred, it does not re- verse itself. The available data on australis is too limited to include in the phylogenetic analysis. Of the characters analyzed, it has two ad- vanced character states; a rounded tail-tip which it shares with gyldenstolpei and no indication of an auricular crease which it shares with anguinodes and the new species. Discussion of the relationships of australis to the other species is deferred to the next section. The hypothetical phylogeny, once constructed, should then be examined for common character state clusters and independently derived characteristics. All four species share three advanced character states: an incom- plete postorbital arch, loss of the external ear, a reduced pectoral girdle. Because the study is based on looking for characteristics that differentiate the taxa, two characters have not been mentioned pre- viously of which all specimens examined have advanced states: a lower eyelid composed of a thickened scale and a snout covered by a thickened cuticle. The combination of the five characteristics apparently defines the minimum adaptations necessary in order to lead a burrowing life. This appears true at least for the forms stud- ied herein. It would be interesting to see whether the character states are of value in predicting basic adaptations to burrowing in other skinks. The maximum number of derived character states shared by any combination of three species is three: a fused frontal bone, broad and shallow lateral process of the parietal, and palatine bones overlapping or just separated. The maximum number of derived character states shared by any two species of the characters not previously used is four: postmental absent, pterygoids in contact, well-developed retro- articular process, reduced pectoral girdle lacking rib articulations. CHARACTERISTICS DERIVED IN PARALLEL AND THE GENUS CONCEPT Basically, I am considering the shift to a completely fossorial habitat, roughly indicated by complete loss of limbs, to be an evolu- tionary shift worthy of generic recognition if no closely related forms 122 FIELDIANA: ZOOLOGY, VOLUME 58 show a gradation of limbs to limblessness such as is demonstrated by the genus Brachymeles in the Philippines. The only likely candidates in Southeast Asia are the genera Lygosoma and Saiphos. As mentioned previously in the character analysis, the new spe- cies shares certain character states which Greer (1970) uses to define the subfamily Scincinae. The advanced character states of these same characters are used to define the subfamily Lygosominae, which appear as the step containing GHJ in the phylogeny. Put in accordance with Greer's (1970) findings, the new species belongs in the subfamily Scincinae, and anguinodes, gyldenstolpei, and roulei belong in the subfamily Lygosominae, as do the genera Lygosoma and Saiphos. The only scincine previously known in Southeast Asia is the genus Eumeces, composed of species with relatively well-developed limbs. Thus I consider the new species to be representative of a distinct genus. Because two subfamilies are represented by the Thai limbless skinks, the following character states have been derived independ- ently in the two subf amilial burrowing stocks : AT *LM * . This matter of independent character state derivations has bearing on the generic status of the three Thai lygosomine limbless species. The problem with the other three Thai species, then, is to deter- mine 1) whether they represent a single invasion of the fossorial habitat by a common ancestor which then speciated (all three spe- cies belonging to a single genus), or 2) whether the species represent two or three independent entries into the fossorial habitat (two or three genera) and 3) what the relationships are to the genera Lygo- soma, Ophioscincus, and Saiphos. In order to determine whether the three species could be ac- counted for by a single invasion, it is necessary to ignore any ad- vanced characteristics dealing with specializations to a burrowing way of life. These would be expected to be important as far as the speciation process is concerned, but, as seen above, many of the characteristics associated with burrowing have been derived inde- pendently in the two subfamilies. In other words, a single lygoso- mine ancestor could have made the initial shift to a burrowing life -style, fragmented into three populations, and the three popula- lations could have then diverged, if the only types of characteristics that differentiate the three populations have to do with further adaptations to an underground existence. The following advanced HEYER: LIMBLESS SKINK 123 character states are represented in the three species: A1A2BCDEFI* KLM*NO*0**P. This grouping does not include the character states used in the phylogeny up to this point. Of these advanced states, only one does not correlate with burrowing specializations in my opinion: K. Two species, gyldenstolpei and roulei, share the char- acteristic— the pterygoid bones in medial contact. On the basis of this one characteristic together with the confusion regarding the generic limits of many lygosomines, I think it best at this time to be nomenclaturally conservative and regard the three Thai limbless lygosomine species as representing the same genus. Using this in- terpretation, each of the species is quite specialized, but it is easiest to envision a common ancestor for gyldenstolpei and roulei within the genus. Until the lygosomine genera are redefined, it seems best to call attention to the extreme specializations demonstrated by the three limbless Thai skinks, anguinoides, gyldenstolpei, and roulei, by plac- ing them into a genus, Isopachys Lonnberg, 1916, distinct from Lygosoma, Saiphos, or Ophioscincus. Because the species here included in Isopachys have previously been included in the genus Ophioscincus, further comment on their relation to 0. australis is warranted, although any decision is tenta- tive due to lack of skeletal material at this time. Ophioscincus australis has one unique scale-field difference in comparison with the three Thai lygosomines: three primary temporals. This character state is shared with the other species of Australian limbless skink examined. In addition, the other Australian limbless skink has sev- eral more head scales and a proportionately larger pectoral girdle than any of the other limbless skinks examined, suggesting a rela- tively more recent entry into the fossorial habitat. It is possible that when further material is available, the Australian limbless skinks may prove the endpoints in a degenerative cline paralleling Brachy- meles. These bits of morphological data, combined with the zoo- geographic improbability of congeneric burrowers in Southeast Asia and Australia (unless burrowers are man-carried, they are usually not zoogeographically vagile forms) lead to the conclusion that the three Thai species do not belong in the genus Ophioscincus. The purpose of the paper was to compare the new series of speci- mens from Sakaerat with the other known limbless skink species from Thailand. The taxonomic conclusion has been reached that the Sakaerat specimens represent a new genus and species, which are now described. 124 FIELDIANA: ZOOLOGY, VOLUME 58 Davewakeum, new genus. Figures 4, 5. Type species. — Davewakeum miriamae, new species. Diagnosis. — The only scincine genera in which some or all of the species are limbless are: Barkudia, Brachymeles, Cryptoposcincus, Davewakeum, Grandidierina, Malacontias, Melanoseps, Nessia, Ophio- morus, Paracontias, Pseudoacontias, Scelotes, Scolecoseps, Sepsophis, Typhlacontias, Voeltzkowia. Barkudia and Sepsophis differ from Davewakeum in having an external ear. The remainder of the genera either lack an external ear or the character is variable among species within a genus. Ophiomorus has a transparent disk in the lower eye- lid, Davewakeum has a thickened, opaque lower eyelid. Brachymeles, Melanoseps, Scelotes, Scolecoseps, and Typhlacontias have supranasal scales: Davewakeum lacks supranasal scales. Cryptoposcincus, Grandi- dierina, Malacontias, Nessia, Paracontias, Pseudoacontias, and Voeltz- kowia lack prefrontal scales, Davewakeum has prefrontal scales. Definition. — Nostril pierced in a single nasal scale; upper eyelid absent, lower eyelid composed of one thickened scale; tympanum absent, no indication of external ear; no supranasal scales; single frontonasal scale present; prefrontal scales paired; frontoparietal scales paired; parietal scales paired; interparietal scale present; limbless. Nasal bones in contact medially; four teeth per maxilla; no ptery- goid teeth; palatine and pterygoid bones well separated medially; basipterygoid process of basisphenoid reduced; frontal bone divided; stapes articulating with quadrate; pectoral girdle reduced, all ele- ments present, one pair of ribs attaching to the sternum; no indica- tion of limb bones; osteoderms not noticeably fused or co-ossified with the skull. Range. — Known only from a restricted region on the edge of the Khorat Pleateau, 60 km. S of the provincial capital of Nakhon Ratchasima, Thailand. Etymology. — The genus, masculine in gender, is named for Dr. David B. Wake, to whom I owe much professional and personal gratitude. Davewakeum miriamae, new species. Figure 7. Holotype. — Field Museum of Natural History number 182550. An adult female from Khao Saton, 300 m. above the Kasetsart University Forestry Station, 60 km. S of Nakhon Ratchasima on Highway 304, Amphoe Pak Thong Chai, Changwat Nakhon Ratcha- HEYER: LIMBLESS SKINK 125 sima, Thailand, collected August 27, 1969 by Sukhum Pongsapi- patana. Diagnosis. — Same as for genus. Fig. 7. Dorsal and lateral views of head of Davewakeum miriamae, new species. Description of holotype. — Body slender, head not differentiated from body; scales smooth; rostral large, its apex prolonged backward onto top of head forming an angle; nostril pierced in nasal scale, bordering first supralabial; nasals forming broad median suture, scale slightly broader than long; no supranasals; frontonasal about two- thirds as broad as long, barely contacting frontal; prefrontals well developed, not quite in contact medially; frontal largest scale on head, almost diamond shaped, three-fourths as broad as long; 3 supraoculars, 2 contacting frontal; frontoparietals separate, not in contact on the midline, just shorter than interparietal; interparietal just longer than broad, barely contacting the frontal anteriorly; parietals meet behind interparietal, greatest length less than twice width; 1 pair nuchals. Two loreals; 4 supraciliaries, 2 presuboculars, 4 postsuboculars; 2 primary temporals; upper temporal borders parietal, third supra- ocular, posterior supraciliary, posterior postsubocular, lower primary temporal and secondary temporal; lower primary temporal borders upper primary temporal, posterior postsubocular, fifth and sixth supralabials, and secondary temporal; 1 secondary temporal, about equal in size to lower primary temporal; 1 tertiary temporal, broader than long; 6 supralabials, the first much enlarged, the fourth under the eye, not contacting the eyelid. 126 FIELDIANA: ZOOLOGY, VOLUME 58 Mental large, posterior edge straight; 5 infralabials; azygous post- mental, 3/7 as broad as long; 2 pairs of chinshields, second longest, each pair separated by one scale. Rostral, nasal, first supralabial, and mental with much thickened cuticle, first two infralabials with noticeably thickened cuticle. Twenty- two scale rows at mid-body; dorsal scales smooth, just broader than ventrals; 107 scales between mental and vent; 5 slightly enlarged preanal scales, 4 bordering vent; 101 subcaudal scales, prox- imal two rows slightly smaller; subcaudal and dorsal caudal scales slightly larger than lateral caudals; tail tip a blunt point terminating in a single, pointed scale. Coloration in life and preservative tan with darker brown spots on each scale forming a series of lines. Pattern is uniform over entire body and tail except the belly and sides of the body are lighter, and the dark lines are somewhat indistinct. Measurements (mm.). — Snout- vent, 90.0; total, 160.5; head width 5.0. Paratypes.—FMNn 182535-45, 182546 (cleared and stained), FMNH 182547-49 from the type locality; FMNH 182534 from the Sakaerat Experiment Station, Amphoe Pak Thong Chai, Changwat Nakhon Ratchasima. A sample will be deposited in the Thai Na- tional Reference Collection, Applied Scientific Research Corpora- tion, Thailand. Variation. — The paratypes range in snout-vent length from 48.0 to 114.0 mm., the largest total length is 173.5 mm. The fronto- parietal scales are in contact in two specimens, separated in the rest. Nine specimens have an additional single enlarged nuchal scale, one specimen has two pairs of nuchals. One specimen has 5 supracili- aries, the remainder 4. Two specimens have 3 presuboculars on one side of the head, one specimen has 3 on both sides, the remainder have 2 on both sides. The number of postsuboculars ranges from 3 to 6. The fourth supralabial often contacts the lower eyelid. The number of scalerows around mid-body is either 20 (2 individuals) or 22. The number of ventral scales between the mental scale and the anus ranges from 93 to 113. The number of subcaudal scales ranges from 79 to 101. There are from 0 to 5 slightly enlarged preanal scales. Some individuals are almost uniform brown with tan venters. Ecological notes. — All the Khao Saton specimens were taken in dry evergreen forest. The Sakaerat specimen was taken in partially cleared dry evergreen forest. Specimens were collected on March 9, HEYER: LIMBLESS SKINK 127 lOS Fig. 8. Map of Thailand showing localities for four species of limbless skinks. A=Isopachys anguinoides, G=Isopachys gyldenstolpei, Open circles =/. anguinoides and gyldenstolpei in sympatry, R=Isopachys roulei, Solid circle =/, anguinoides and roulei in sympatry, 'M.= Davewakeum miriamae. 1969 (1), March 18, 1969 (1), August 27, 1969 (10), October 3, 1969 (1), November 6, 1969 (1), December 8, 1969 (1), January 1, 1970 (2). There does not appear to be any correlation with the collections and the rainy season. The first specimen was collected at Sakaerat at 21:40 hours, active on the surface of the ground 0.75 m. from a small stream. All the Khao Saton specimens were caught under cover in the daytime. Two specimens were taken from under rocks 30 cm. 128 FIELDIANA: ZOOLOGY, VOLUME 58 in diameter, the rest were collected from 1 to 15 cm. underground. Of the 14 specimens taken underground, six were found between tree buttresses. On August 27, 1969, Mr. Sukhum Pongsapipatana dug up ten individuals in a 4 m.^ area. This area of high skink density (two Riopa bowringi were also dug up) was not obviously differen- tiated from surrounding areas of the forest. Distribution. — The species is only known from the two adjacent localities listed above. This region is at the eastern edge of a low montane region separating the Khorat Plateau from the Central Valley of Thailand (fig. 8). Etymology. — The species is named after my wife Miriam, whom I thank for her unbelievable patience and understanding. She dis- covered the first specimen while helping with field work one evening An ARTiFicLUi Key to the Limbless Skinks of Thailand lA. Second supralabial under middle of eye Isopachys gyldenstolpei IB. Third or fourth supralabial under middle of eye 2. 2 A. Fourth supralabial imder middle of eye Davewakeum miriamae. 2B. Third supralabial imder middle of eye 3 3A. 18 scale rows around midbody laopachys rovlei. 3B. 22-26 scale rows around midbody Isopachys anguinoides. Localities and Specimens Examined of Isopachys Isopachys anguinoides (fig. 8): Thailand — Chon Buri: Bang La Moung, TNRC 522:653; Prachuab: Nong Kae, TNRC 522:2-8, 522:9 (cleared and stained), 522:10-13, 522:161-2, 522:188-192 (1 of this untagged series cleared and stained), 522:651-2; Prachuab Khiri Kahn: Bangtaphan, type locality; Hua Hin, FMNH 177487- 88, 177521; Gulf of Siam: Koh Tao. Isopachys gyldenstolpei (fig. 8): Thailand — Kanjanaburi: Ap- prox. 6 km. from Kanjanaburi, TNRC 522:1, 522:14-15, 522:122- 124, 522:125 (cleared and stained); Prachuab: Nong Kae, TNRC 522:169; Prachuab Khiri Kahn: Hua Hin, FMNH 178323, 178324 (cleared and stained) ; Koh Lak Paa, type locality. Isopachys roulei (fig. 3): Thailand — Chon Buri: Ang Hin, FMNH 177265-66, 177267 (cleared and stained); Bang Lamung, TNRC 522:495, 522:653, Bang Saen, FMNH 177268. HEYER: LIMBLESS SKINK 129 REFERENCES Angel, M. F. 1920. Sur un Saurien nouveau de la famille des ophiopsisepid^s. Bull. Mxis. Hist. Nat., Paris, 1, pp. 4-6. BOULENGER, G. A. 1887. Catalogue of the lizards in the British Museum (Natural History). Vol. Ill, 2nd ed. Taylor and Francis, London, xii+575 pp. 1914. Descriptions of new reptiles from Siam. J. Nat. Hist. Soc. Siara, 1(2), pp. 67-70. Greer, A. E. 1967. A new generic arrangement for some Australian scincid lizards. Brevi- ora, 267, pp. 1-19. 1970. A subfamilial classification of scincid lizards. Bull. Mus. Comp. Zool., 139(3), pp. 151-183. LONNBERG, E. 1916. Zoological results of the Swedish zoological expeditions to Siam 1911- 1912 and 1914. 2. Lizards. Kungl. Svenska Vetenskapsakademiens Hand- lingar, 55(4), pp. 1-12. Marx, H. and G. B. Rabb 1970. Character analysis: an empirical approach applied to advanced snakes. J. Zool. London, 161, pp. 525-548. MiTTLEMAN, M. B. 1952. A generic synopsis of the lizards of the subfamily Lygosominae. Smith- sonian Misc. Coll., 117(17), pp. 1-35. Peters, W. 1873. Mittheilung iiber neue Saurier (Sphaeriodactylus, Anolis, Phrynosoma, Tropidolepisma, Lygosoma, OpMoscincus) aus Centralamerica, Mexico, und Australien. Monatsbr. Akad. Wiss. Berlin, 1873, pp. 738-747. ROMER, A. S. 1956. Osteology of the reptiles. Univ. Chicago Press, Chicago, xxi+772 pp. Sewertzoff, a. N. 1931. Studien iiber die Reduktion der Organe der Wirbeltiere. Zool. Jahrb. Anat., 53, pp. 611-700. Smith, M. A. 1935. The fauna of British India, including Ceylon and Burma. Reptilia and Amphibia. Vol. II. Sauria. Taylor and Francis, London, xii+440 pp. 1937. A review of the genus Lygosoma (Scincidae: Reptilia) and its allies. Rec. Indian Mus., 39(3), pp. 213-234. Taylor, E. H. 1963. The lizards of Thailand. Univ. Kansas Sci. Bull., 44(14), pp. 687-1077. T ^ tv /I Iff n:: ':. : ^ : Mm: ' ' ill 1 UNIVERSITY OF ILLINOIS-URBANA 590. 5FI C001 FIELDIANA, ZOOLOGY$CHGO 55-58,61 1968-72 3 0112 009379840