erty as FOE PEs ’ 5 SO fod :. NS : ¥ e ce Stn! Pipe ie Fs SER SR . TE Shoe Pr 4 Reese Ne yy AAAS VN SOAS, ee: Yer ey 5 Oe eae PREF we ey *: SES ee oe Sse page ioe ; ie ae ; ee S55 Paribas ee oe Te gE et ae a tO rea EP eet Sia and a eet in ye Ne ag HE FIELD MUSEUM LIBRARY WO NNN N 3 5711 00015 4402 OF NATURAL HISTORY ee C , renee Digitized by the Internet Archive in 2012 with funding from Field Museum of Natural History Library http://archive.org/details/fieldianazoology49chic “y FIELDIANA: ZOOLOGY A Continuation of the ZOOLOGICAL SERIES of FIELD MUSEUM OF NATURAL HISTORY VOLUME 49 NATURAL HISTORY ae FOU Sy M ees ¢\ ee ARSHALL CHICAGO NATURAL HISTORY MUSEUM CHICAGO, U.S.A. 1965 THE SYSTEMATICS AND EVOLUTION OF THE ORIENTAL COLUBRID SNAKES OF THE GENUS CALAMARIA ROBERT F. INGER Curator AND HYMEN MARX Associate Curator Division of Amphibians and Reptiles FIELDIANA: ZOOLOGY VOLUME 49 Published by CHICAGO NATURAL HISTORY MUSEUM NOVEMBER 29, 1965 Edited by EDWARD G. NASH Library of Congress Catalog Card Number: 65-2467 1 PRINTED IN THE UNITED STATES OF AMERICA BY CHICAGO NATURAL HISTORY MUSEUM PRESS CONTENTS PAGE INTRODUCTION « «.% 2 % 0 ¢-s 4 3 4 jee & oH eG ee Hee eS 5 GENERIC DESCRIPTION. . . . . . 2. 1 1 ee ee TAXONOMIC PROCEDURE ... 2... 60 8 ee ee ee ee ee ee ew) 8 RECOGNITION OF SPECIES .......... 2... 2. 2... 2.2... 8 VARIATION OF CHARACTERS ............... 2... ... 14 SPECIES:ACCOUNTS «4: 4 @ 4.6 4 6 6 WE 2B bw eR ae RRR Ee ~ OO EVOLUTION AND SPECIATION. ........... 2... 2... 2... 241 Phylogenetic relations of the genus Calamaria ............ 241 Evolutionary trends within the genus Calamaria ........ . . 252 Patterns of evolution. ......... uo Oe ee A ee me DOYSIDUTION. «. con « 2 oe Gd ee ee Ee Ree wee ee oe BE Centers of evolution and dispersal. .............. . . 275 Time of dispersal and speciation. ............... . . 279 DUNMAR Ys 4s ae va Bo wea ee eee ee ee ee ole wee See APPENDIX A—Formation of Dendrogram ............... . 285 APPENDIX B—Faunal List ............... 2... 2... . 298 REPWRENCES 5 4 « 4&6 24 % © & KR w Re & Sd og ewe eS a & » ZOD TINDIES ¢ ie. boos we Oe me Be & Bw ee eo ee ea a we 1. we Be: OE INTRODUCTION The genus Calamaria, distributed from southern China and the Riu Kiu Islands southward through the East Indies to the Philip- pines and Celebes, is one of the two or three most successful genera of colubrids in the eastern Oriental tropics. These highly specialized, burrowing, forest-dwelling snakes are easily recognized as members of a single genus. Very few species of colubrid snakes have as few scale rows (13) as Calamarza; none have more extensive fusion of head scales. In this highly restricted morphological framework, no amount of speciation can disguise the generic identity of a form. One result of this homogeneity is that only four generic names have been pro- posed for the 148 forms named since 1826. Definition of species limits, however, has been a difficult problem in the past because of the homogeneity of the group and because so few specimens have been available to most workers. Since the last revision (de Rooij, 1917) of a large section of the genus, two enormous collections of a few species have been made at several localities in Java (Kopstein, 1941; de Haas, 1941). These collections housed in the Rijksmuseum van Natuurlijke Historie, Leiden, provide a critical means of evaluating variation. Without an opportunity to examine these Javanese snakes in our own laboratories, our study would have been seriously hampered. We are, therefore, especially grateful to Drs. L. G. Brongersma and M. Boeseman of the Rijksmuseum for having loaned us these very large collections. The curators of other museums have also been generous and in many cases took the most liberal view of adminis- trative rules in lending us specimens. As a result we were able to see more than 2600 specimens. We wish to express our thanks to the following colleagues and institutions for permission to study specimens in their collections: Gordon Conway and Chin Phui Kong, Agriculture Department, Tuaran, Sabah; Charles M. Bogert, American Museum of Natural History (AMNH); James E. Bohlke, Academy of Natural Sciences of Philadelphia (ANSP); J. C. Battersby (retired) and Alice G. C. Grandison, British Museum (Natural History) (BMNH); Alan E. 5 6 FIELDIANA: ZOOLOGY, VOLUME 49 Leviton, California Academy of Sciences (CAS); Neil D. Richmond, Carnegie Museum (CM); Chicago Natural History Museum (CNHM); Edward H. Taylor (EHT-HMS); Bengt Hubendick, Naturhistoriska Museet Goteborg (GM); A. Capart, Institut Royal des Sciences Naturelles de Belgique (IRSNB); Enrico Tortonese and Lilia Capocaccia, Museo Civico Genova (MCG); Ernest E. Williams, Museum of Comparative Zoology (MCZ); Villy Aellen, Museum d’Histoire Naturelle, Genéve (MHNG); Jean Guibé, Museum Na- tional d’Histoire Naturelle, Paris (MHNP); Robert C. Stebbins, Museum of Vertebrate Zoology (MVZ); F. Gouin, Musée Zoologique de l’Université, Strasbourg (MZUS); Josef Eiselt, Naturhistorisches Museum Wien (NHMW); Lothar Forcart, Naturhistorisches Mu- seum Basel (NMB); Eric R. Alfred, National Museum of Singapore (NMS); Greta Casteegley, Naturhistoriska Rijksmuseum Stockholm (NRS); L. D. Brongersma and Mart Boeseman, Rijksmuseum van Natuurlijke Historie (RMNH); Tom Harrisson, Sarawak Museum (SM); A. Kleinschmidt, Staatliches Museum fiir Naturkunde in Stuttgart (SMNS); Robert Mertens and Konrad Klemmer, Sencken- berg Museum (SNG); George S. Myers and Walter C. Brown, Nat- ural History Museum of Stanford University (SU); Norman Hartweg and Charles F. Walker, University of Michigan Museum of Zoology (UMMZ); Doris M. Cochran, United States National Museum (USNM); Leif R. Natvig, Universitetets Zoologiske Museum, Oslo (UZMO); Dirk Hillenius, Zoologisch Museum Amsterdam (ZMA); Heinz Wermuth (formerly) and Giinther Peters, Zoologisches Museum Berlin (ZMB); Werner Ladiges, Zoologisches Museum Hamburg (ZMH); Walter Hellmich, Zoologische Sammlung des Bayerischen Staates, Munchen (ZSBS); M. L. Roonwal, Zoological Survey of India (ZSI). We are also grateful to Dr. G. B. Rabb, and Dr. T. M. Uzzell for critical comments on various parts of the manuscript. The drawings, charts, and maps were prepared by Miss Marion Pahl, Miss Phyllis Wade, and Miss Janet Wright. The photographs were taken by Mr. H. V. Holdren. The several versions of the manu- script were typed by Mrs. Marion Anderson and Miss Janet Wright. The research was partially supported by National Science Foun- dation Grant G-6257. CALAMARIA H. Boie Calamaria H. Boie in F. Boie, 1827, Isis, 20, col. 519 and 539—type species Calamaria lumbricoidea Boie by subsequent designation of Brongersma, Inger, and Marx, ms. Changulia Gray, 1834, Ill. Ind. Zool., 2, pl. 86, figs. 6-9—type species Chan- gulia albiventer Gray by original designation. Typhlocalamus Gunther, 1872, Proc. Zool. Soc. London, 1872, p. 589—type species T'yphlocalamus gracillimus Gunther by original designation. Keiometopon Taylor, 1962, Univ. Kansas Sci. Bull., 43, p. 257—type species Keimetopon booliatt Taylor by original designation. Application for conservation of the generic name has been made to the International Commission on Zoological Nomenclature (Bron- gersma, Inger, and Marx, MS). Habitus vermiform; head not distinct from neck; tail short. Eye distinct, diameter greater than, equal to, or smaller than eye- mouth distance; rostral rounded; paired prefrontals fused with inter- nasals and in contact with supralabials; frontal 114 to 8 times width of supraocular; parietal meeting last supralabial; an enlarged para- parietal (Fig. 1); 4 or 5 supralabials, second and third, third, or third and fourth touching eye; nasal single and surrounding nostril; no loreal; a single preocular present or absent; one postocular; no an- terior temporal; mental touching or separated from anterior chin shields; 5 infralabials, first two or three touching anterior chin shield; two pairs of chin shields, posterior ones meeting or separated in mid- line; 3 or 4 gular scales in midline between chin shields and first ventral. Dorsal scales in 13 longitudinal rows throughout body length; scales smooth; no apical pits; anal single; subcaudals paired. Posterior vertebrae without hypapophyses; 6 to 12 maxillary teeth, modified (Fig. 3) in most species; palatal and mandibular teeth conical. This genus can be distinguished from all other snakes on the basis of the combination of the following characters: dorsal scales in 18 rows throughout body; internasals and prefrontals fused; parietal broadly in contact with supralabials. 7 8 FIELDIANA: ZOOLOGY, VOLUME 49 The genus Keiometopon Taylor was based on a single specimen collected in Malaya. Through the courtesy of Dr. E. H. Taylor we have been able to examine the type specimen. It is clearly an anom- alous specimen of Calamaria schlegeli. The frontal shield is fused with the prefrontals and a median suture runs the length of the dorsal surface of the head. The scales on the side and ventral surface of the head are normal. Several fusions of the longitudinal scale rows occur opposite the first fifteen ventrals. Between the twentieth and the twenty-sixth ventrals the second scale row on each side splits returning the dorsal scale count to thirteen, which number continues to the vent. The maxillary teeth have the characteristic modifica- tion of Calamaria. TAXONOMIC PROCEDURE In establishing taxa and their interrelations, we have adopted the following procedure: 1) examine all characters practically accessible; 2) analyze variation of characters within sympatric samples; 3) correlate all characters within individuals within samples; 4) establish sympatric taxa on basis of (2) and (8); 5) extend analysis of variation to allopatric samples; 6) on basis of (4) and (5) establish limits of species; 7) compare Calamaria with other genera of Colubridae to deter- mine distribution of certain characters in the family; 8) on basis of (7) decide which characters of Calamaria are primitive (within the genus) and which derived (or ad- vanced) ; 9) on basis of (5) and (6) establish relations between species (or species groups) ; 10) on basis of (8), (9), and zoogeographic possibilities, work out phylogeny within the genus. The validity of the procedure can be tested through examination of the logic and with fresh material. Perhaps the latter is the best test, for in effect by our taxonomy we are making predictions con- cerning associations of characters in snakes not yet examined. RECOGNITION OF SPECIES Starting with large sympatric samples, such as one from Tjikad- jang, Java (Table 1), we have examined as many characters as pos- INGER AND MARX: THE SNAKE GENUS CALAMARIA 9 sible and then analyzed the distribution of the total sample with respect to these characters. As shown in Table 1, the sample from Tjikdajang may be treated as two isolated populations (linnaez and lumbricoidea) that differ in ventral counts (both sexes), subcaudal counts (females), tail ratio (males), supralabials, three aspects of coloration, and ornamentation of hemipenes. Treated as one population the Tjikadjang sample consists of two clusters of individuals in a multidimensional space generated by the 14 characters. Fragmentation of a species into two such disparate groups at one locality can occur only under very limited circum- stances. The separation may be based on age differences. In this case juveniles (less than 200 mm.) and adults (females with large ova) occur within each group, thus eliminating this explanation. The separation could reflect sexual dimorphism, but the table shows both sexes are well represented in each group. Seasonal changes affect some animals profoundly and produce cyclomorphosis, as in Daphnia (Brooks, 1946). Seasonal variation does not account for the division in this sample as half of the ‘“lumbricoidea”’ group and all of the ‘“‘lin- naev’ group were collected in months January—May; the range of variation (Table 1) within the two groups is too narrow for one-half of the ‘“‘lwmbricoidea’’ group to cause the wide gap between the groups. The only possible explanation remaining for the split in the sam- ple under the single-species hypothesis is polymorphism. Because of the characters used in the analysis (see below), such polymorphism would involve large segments of the genotypes, which in turn in- creases the possibility of recombinations, crossovers, and hence inter- mediate phenotypes. As no intermediate phenotypes appear if we consider coloration, ornamentation of hemipenis, number of supra- labials, and number of ventrals (which almost certainly have a multi- ple factor mode of inheritance), the polymorphism explanation can be rejected. Thus the split single-species hypothesis can be rejected for lack of a sound biological explanation. This hypothesis also runs con- trary to all experience with species of snakes. Clearly the only rational interpretation of the correlated distribution of characters in the Tjikadjang sample is that two reproductively isolated popu- lations are involved. Another large sample, this one from Wonosobo, Java, was ana- lyzed in the same manner (Table 1). In this case three clusters of individuals develop in the multidimensional space. Again the most “ayNOS [VUIUVIIE} WO] STEPNVIQNS JO SUIIE} UL pd}VoO] SMO B[BoS [VSIOP INO} 0} UOLJonpey ; "y{3ug] [8107 JO SY}pURSNOY} jo sua} UT , -) c) FI PPI Me 2 cr geo: ypuULyos L-g 6 CC-€¢G 16 SIl=éL 61 Po-CST OFT S 1 cee pysapou, II-6 €1-Ol 99-87 V9-6¢ 9T-€1 LI-91T 66-81 6L-991 ¢ yoo eee ypjomsib VI-€I VI-OL P8-LIO VOT-6L OZ-LI 0Z-8I1 LY-V¥l 9€-I§I € G ae SIsuanpns val 0Z-8T Z8-GL 86-86 VC-12 62-92 06-08T T8-9LT ¢ 9: 3, a uiysnoqgn.b 8-P 8-P T8-OL ITT-88 6I-LI VG-&6 O8-OLT c9-PVST L L oo * Dapvooisqun) c=[ 8-P 9€-8G VL-OL IIT-6 SI C0d-S6I1 €8-8LI G G -» Se awe DjDjNH.11a val 9T VL 16 9T era OST 9TI I c Soe oes $2)D410}D) eT ote E1Z side TP e's Ce-Fel oe 7 Geo Ceo ol Oo dbs yabayyos [jooulog YON ‘neg euly If] L-€ 8-§ 69-LV 00T—-08 S1-Tt Ic-81 8L-S9T c¢c-9FI 9T Li. =? a ar ae 149]]}9NU 9-G V G8—-LI cor LI-GT ES OL-9ST 9CT G [EP oS ete oe peers DIANI V-G + at | 99-LG TTT-98 LI-ST VE—-06 VL-LOT G9-S8PL 6 TT (00070 0% **82478047)N9D [saqe[eD “Ss ‘uleyqUuog “I]A] C-% 9-T C9-ZGS 611I-L6 Sissel 8S-VG T6-S9T 99-CCI CG 8ST “°° °° * * * *ppsapou ¢ ie CF ae 91 hee C1Z a I Ze ee ee Daprooiuquin} ‘Sea! c-§ €S-LP 68-68 Liss SI-LI T9-I¢T CY-5-,., three had 5-5—. 5(3-4)’ (8-4)’ and two had 5(8-4). and one Supralabials, because of the limited intraspecific variation, are useful in distinguishing species. Infralabials show less interspecific variation than supralabials. Five is the usual number, the first three touching the anterior chin shield. Calamaria borneensis invariably has four infralabials of which only the first two touch the anterior chin shield. In C. lowz the infra- labial formulas are 4(2) or 5(8). Paraparietals—All species of Calamaria have an enlarged scale or shield along the posterolateral margin of the parietal and behind and slightly above the level of the last supralabial (Fig. 1). This shield, which we call the paraparietal, is usually surrounded by five or six scales counting the parietal and supralabial. The functional significance of the difference between these counts is obscure, if indeed there is any. Certain specimens of lumbricoidea have only four scales bordering the paraparietal; in these specimens the paraparietal is short, squarish, and does not reach the end of the parietal. In other lumbricoidea, the paraparietal is oblong as in Fig. 1 and reaches the end of the parietal or slightly beyond. Although the paraparietals of the two snakes shown in Fig. 1 differ in length, spe- cies having five scales bordering the paraparietal do not have con- sistently longer or shorter paraparietals than those with six. The difference in counts reflects the number of scale rows meet- ing the paraparietal. The fourth, fifth, and sixth rows always touch the paraparietal. Species in which the third row also meets the para- parietal have six scales bordering that shield. INGER AND MARX: THE SNAKE GENUS CALAMARIA Li Fic. 1. Position of paraparietal (Ppa) and number of scales surrounding it in Calamaria buchi (top) and C. schmidti (bottom). Pa=parietal. SL=last supra- labial. Unfortunately, we discovered this variation in scutellation after we had returned part of the material loaned to us. As a result, we have almost no information on this character in some species (e.g., pavimentata, septentrionalis) otherwise examined in large numbers. Despite this handicap, the data in Table 4 which lists the larger samples indicate the patterns of variation in this character. Individual variation within a local population is relatively insig- nificant. For example, all of 25 linnaei from Bogor have counts of 18 FIELDIANA: ZOOLOGY, VOLUME 49 TABLE 4.—Frequency Distribution with Respect to Number of Scales Bordering the Paraparietal in Large Samples of Calamaria. Species Seales bordering paraparietal 4/4 4/s 5 /s 5/s 6/s 6/4 7 /, COUNT OSI TSes Aa odode GS woriad 1 3 24 1 VCO 2.0 neta pee ie See aes 1 9 DHOTOUES 66.6652 % 0nd bees es iw! it DOTNCCNSIS: ca sc ce SL Se as 14 gervaist (Manila)........... 1 94 5 gervaist (remainder)........ 2 155 12 3 GPA0OWS CDE ons % 6 x Das sesteaccac 1 1 18 2 OQUTSWOWD 25:8 ose es He Dae a plat leucogaster................ 15 1 1 TOC es aan as a ae A at 2 61 i WOW oso ex oon Sho ee ae 34 lumbricoidea (Philippines)... 9 4 5 lumbricoidea (remainder)... . 2 151 9 Z modesta (Borneo).......... 4 modesta (remainder)........ 3 151 i MUCHET. vcs Shee ba when ss 1 62 1 MUCHOS <6 onc Hie eee Y wee 1 7 schlegeli (Sumatra)......... 1 6 schlegeli (remainder)...... ; 12 1 SIUNCISTS cg once ck weal | 13 1 i! sumatrana............005. 17 virgulata (Celebes)......... 14 virgulata (Java)............ a virgulata (Sumatra)........ 1 i virgulata (Borneo).......... 4 virgulaia (Sulu). .... 6.0.0.5 5 virgulata (Mindanao)....... 4 nt virgulata (Palawan)........ 3 6/6, all 11 griswoldi from Mount Kina Balu have 5/5, and all but 6 per cent of gervaist from Manila have 5/5. The restriction of local variation gives reliability to the geographic variation observed (Table 4). The extent and sharpness of geo- graphic variation is not the same in all species. Calamaria gervaisi, which ranges throughout the Philippine Islands except for Palawan, shows no geographic variation. Although lumbricozdea from the Phil- ippine Islands differ in this character from the rest of the species’ sample, the frequency distributions overlap considerably. Roughly the same amount of variation occurs in vrgulata, but differences between populations are sharper. Our observations indicate that, despite some intraspecific varia- tion and despite limited interspecific variation, this character must be used in any taxonomic work on the genus Calamaria. Relation of mental to anterior chin shields.—The mental shield may touch the anterior chin shields or be separated from them by the first infralabials. We found no intraspecific variation in 31 species; 13 INGER AND MARX: THE SNAKE GENUS CALAMARIA 19 species varied to some extent; nothing can be said about the six spe- cies of which we saw only one specimen each. Intraspecific variation (Table 5) may take the form of a relatively rare anomaly as in lin- naer or schlegelt. In pavimentata, lowi, and virgulata the variation is distinctly geographic. In alidae and modesta variation, though more frequent than in some species, is not geographic. TABLE 5.—Frequency Distribution with Respect to the Relation Between the Mental and Anterior Chin Shields in Variable Species of Calamaria. Asym- Species Touching Separated metrical! oO ~ isr) i) ~~ j~) ~ —_ iw) CUCLON oakea Syne bum oe He sd Be Re 0 GTQDOWSIYE on cc ceed dea ew ba bes 21 leucogaster é UOC eS os estas ws x vi xcs bahen ade We eS Dae lowt (Borneo)................0.. 57 low. (Malaya). ..004¢56040 e004 ones WOCHENL «eee eb ered = we oe wes we MOdESIA... 2. ee ee eee ees palavanensis......... 00.0 eee pavimentata (Laos)............... pavimentata (remainder).......... SCWICGOU < okk-da seeks Bae ee hes ak virgulata (S. Celebes)............. virgulata (N. Celebes)............ virgulata (remainder)............. oe) —_ Co e we) SOO HALPODwWeOAOCcrh bo oD AD — SCoOonrnronNnrKo RPOONNORNOCCOrFOrFH Oo Qe ono 1 One of first infralabials reaching mid-line and thus separating mental from chin shield of that side. The snakes listed as asymmetrical in Table 5 have been instruc- tive. In these specimens the first infralabial of one side reaches the midline and thereby separates the mental from the chin shield of that side. This variant thus forms an intermediate between two extreme types. Nominate species have been described that differ from others only in the relation of the mental to the chin shields. Calamaria simalurensis de Rooij, for example, differs from mo- desta only in having the mental touching the anterior chin shields. Yet in the publication (de Rooij, 1917) describing simalurensis, the illustration of the holotype of another new species, elegans (also from Simalur Island), clearly shows the asymmetrical arrangement. Three other specimens from Simalur we have examined have the mental separated from the chin shields as in Javan modesta and agree with modesta in all other characters. The large Javan sample (167 speci- mens) includes six snakes that have the mental touching the chin 20 FIELDIANA: ZOOLOGY, VOLUME 49 shields and six that have the asymmetrical pattern. This intermedi- ate arrangement helped us to interpret the variation in this character. The holotype of Calamaria schlegeli differs from the types of leuco- cephala and agamensis only in this character. This difference was apparently the reason why Boulenger (1894) separated the last two from schlegelt. The holotype is the only specimen of the more than 200 schlegeli we have seen having the mental touching the chin shield. The discovery of two snakes having the asymmetrical arrangements supports our decision that this sole difference between schlegeli and the two other nominate forms reflects individual rather than inter- specific variation. Table 5 shows that species do differ in this character. The lack of intraspecific variation in 81 species and the low frequency of vari- ants in additional species enable one to use the relation of mental and chin shields as a diagnostic or key character. However, when two samples differ only in this character, probably intraspecific rather than interspecific variation is involved. Chin shields and gulars.—Every species of Calamaria has two well-developed pairs of chin shields. The anterior chin shields are usually larger than the posterior ones and always reach the mid- ventral line. The posterior chin shields also meet in the mid-ventral line in most species, but are separated by a small gular scale in some. Three or four gulars precede the first ventral in the mid-ventral line. If the posterior chin shields do not meet in the mid-line, invari- ably there are four gulars in the mid-line. Only one species (apraeoc- ularis) having the posterior chin shields meeting in the mid-line has four gulars in the median row. The relationship of the posterior chin shields to each other and, consequently, the number of median gulars vary geographically in lumbricoidea (‘Table 21). They also vary in schlegeli, but the varia- tion is not geographic. Mazxillary teeth.—In our earlier paper on Calamaria, we described the distinctive, basally expanded maxillary teeth of many species of Calamaria. We have extended our survey to every species in the genus. Although we examined enough specimens of most species to determine the scope of major intraspecific variation, we did not measure enough teeth to permit statistically valid analysis of minor variations in size and shape. This survey was adequate, however, to confirm our earlier finding that there are two sharply different types of maxillary teeth. INGER AND MARX: THE SNAKE GENUS CALAMARIA 21 Fic. 2. Maxilla and maxillary teeth of Calamaria leucogaster (top) and C. sep- tentrionalis (bottom). Length of maxilla in leuwcogaster 5 mm., in septentrionalis 3.5 mm. The peculiarly modified teeth shown in the lower halves of Fig- ures 2 and 8 have bilobed pulp cavities. The pulp cavity in the un- modified, conical type of tooth (upper halves of Figs. 2 and 8) is conical and slightly curved, paralleling the shape of the tooth. Serial variation is limited to minor changes in size and, in the unmodified dentition, to slight basal thickening of the most poste- rior teeth. This thickening does not result in a tooth resembling the modified ones. Thus, a snake has but one type of maxillary tooth (Figs. 2 and 8). Apparently local, individual, and ontogenetic variation do not affect the form of the maxillary teeth. We examined nine luwmbri- coidea from Tjikadjang, Java, and although two were shorter than 200 mm. and two longer than 475 mm., tooth shape in this series was essentially uniform. The maxillary teeth of lumbricoidea from five land masses, of leu- cogaster from three land masses, of gervaisi from three, of virgulata from four, and of modesta from four were examined. We found no intraspecific variation in tooth shape. Fic. 3. Unmodified maxillary teeth of Calamaria acutirostris (top) and modi- fied ones of C. linnaei (bottom). 22 INGER AND MARX: THE SNAKE GENUS CALAMARIA 23 The number of teeth in a maxilla varies from 6 to 12. In most species the observed range of variation is 2 or 3 (Table 6). Local individual variation in tooth number in certain species may be as great as that over the entire geographic range of most species. For example, acutirostris from southern Celebes has 10 to 12 maxillary teeth; lowz lowt from western Borneo has 8 to 10. Other species, such as lumbricoidea, show almost no local variation. Geographic variation appears in lowi and virgulata. This form of intraspecific variation may be more common than our limited data indicate. Sympatric species differ in tooth counts in varying degrees. Bor- nean specimens of modesta have lower counts (6-8) than do Bornean lumbricoidea (10). The counts of Bornean sulwensis and grabowskyt overlap but do not coincide. Of the species from southern Celebes, curta has distinct counts whereas those of the other three species overlap (Table 6). The tooth counts are not related to the form of the teeth. The species having unmodified maxillary dentition have 6 to 12 teeth, which is almost identical to but slightly exceeds the range (6 to 11) in species having modified teeth. All but six species of Calamaria have modified maxillary teeth. Because only two types of maxillary teeth occur in this genus, the TABLE 6.—Variation in Maxillary Dental Counts in Species of Calamaria. No. of teeth 7 8 9 10 11 12 Species Region No. of specimens curta southern Celebes 3 acutirostris southern Celebes 3 Z 3 muelleri southern Celebes 2 4 2 nuchalis southern Celebes i! 2 lumbricoidea Philippine Islands 3 lumbricoidea Borneo 6 lumbricoidea Sumatra 1 7 lumbricoidea Java Z 2 grabowskyi Borneo Z Z 1 suluensis Borneo 2 2 virgulata Borneo 2 virgulata Java 1 virgulata Philippine Islands a 1 virgulata Celebes Z 3 schlegeli Borneo 4 schlegeli Sumatra y| schlegeli Malaya 3 schlegeli Bangka 1d. schlegeli Java 9 4 lowi lowt Borneo 1 2 4 lowi gimletti Malaya 2 3 prakkei Borneo 5 1 eiselti Sumatra 3 crassa Sumatra (i 24 FIELDIANA: ZOOLOGY, VOLUME 49 usefulness of tooth shape in distinguishing between species is limited despite the absence of intraspecific variation. The number of maxil- lary teeth is potentially of greater value in the definition of species because of the wider intrageneric variation. The maxillae of species having modified teeth are deeper relative both to the teeth and to the size of the snake. The snakes shown in Figure 2 have teeth of the same length relative to total length. Yet the maximum depth of the maxilla (again relative to total length) of the snake having modified teeth is 1.7 times that of the one having unmodified teeth. The unmodified teeth illustrated are clearly longer relative to the maxilla than are the modified ones. Similar differences are shown in other specimens. In one of acutz- rostris (447 mm.), which has unmodified teeth, the maximum depth of the maxilla was 42 micrometer units, the length of the longest tooth 45 units. In a linnaet (292 mm.), which has modified teeth, the maxilla was 45 units and the tooth 22. In one lumbricoidea (393 mm.), which also has modified teeth, the maxilla was 52 units and the tooth 32. The depth of the maxilla relative to total length of these three specimens stand in the relationship of 1.0 to 1.6 to 1.4. Ventrals—The generic range in ventral counts is 119 (joloen- sis, &') to 804 (gracillima, 2). The maximum counts of most spe- cies fall between 146 and 185. The first ventral is preceded by three longitudinal rows of small gular scales and is about as wide as these three rows combined. Sexual dimorphism in ventral counts is characteristic of Cala- maria. Females of most species have higher counts than males. Unless intraspecific comparison of the sexes is restricted to speci- mens collected at one locality, geographic variation complicates the picture. For example, the counts of males of luwmbricoidea from Tjikadjang, Java completely span the counts of females from Mount Kina Balu although sexual dimorphism is sharp at each locality (Table 7). Some of the complexities of variation in ventral counts are shown in Table 7 which lists those species represented by at least 17 specimens. In general, ranges of variation within local populations are rela- tively narrow, all but three of the 30 examples in Table 7 being less than 16. Neither sample size nor absolute number of ventrals exerts much influence on local ranges of variation (Figs. 4 and 5). Intraspecific ranges, considering entire species samples, are highly variable. Part of the differences between species in the ranges of range of variation in ventrals 140 150 160 170 180 190 200 210 220 230 maximum number of ventrals within local samples Fic. 4. Relation between maximum number of ventrals and range of variation in ventral counts within local samples of Calamaria. 26 FIELDIANA: ZOOLOGY, VOLUME 49 local sample size 4 8 10 12 4 16 18 20 “22 24 (26° 28° -30° 432 34 local variation in ventrals Fic. 5. Relation between local intraspecific variation in ventral counts and local sample size in species of Calamaria. ventral counts can be explained by the maximum individual ventral counts (Fig. 6). It is self-evident that a species having a maximum count of 230 could have, for example, a range of almost 100 ventrals (as in lumbricoidea), whereas one having a maximum count of 150 could not. Sample size has little effect on total intraspecific variation (Fig. 7). The ranges of linnaez are about half as extensive as those of grabow- skyt which is represented by less than a fifth as many specimens (Table 7). Species of terrestrial snakes distributed in several islands might be expected to show greater ranges of variation than those occurring in only one island because of the geographic isolation of the popula- tions of the first group. However, the ranges of variation in species such as grabowskyi and borneensis, which are known only from Bor- neo, may be as great as the ranges of species such as leucogaster and schlegeli, which occur on three or four land masses including Borneo. INGER AND MARX: THE SNAKE GENUS CALAMARIA 27 yh (eo) NO je) oO range of variation 1n ventrals Ww (o) 140 150 160 170 180 190 200 210 220 230 240 250 maximum number of ventrals within’ species Fic. 6. Relations between maximum number of ventrals and intraspecific ranges of variation in ventral counts in Calamaria. The range of ventral counts in gervaisz, which occurs on nine of the Philippine Islands, is no greater than that of schlegeli (known from four land masses) and is less than those of low? and modesta (each known from three land masses). The range of ventral counts in females of lumbricoidea from Borneo is 64, or greater than the range of variation in all other species except pavimentata. Thus the fragmen- tation of the geographic distribution into separate land masses does not affect the range of ventral counts significantly. Narrow local variation makes it possible to rely on ventral counts for interspecific comparisons, provided the samples being compared are from the same locality (first four lines of Table 7). Interspecific comparisons of ventral counts based on allopatric samples are not reliable because of the wide intraspecific ranges of variation in some species; such interspecific comparisons must be bolstered by evidence from other characters. Subcaudals.—The observed generic range in these counts is six (septentrionalis, 2) to 44 (schlegelt, ~), the maximum within spe- 28 FIELDIANA: ZOOLOGY, VOLUME 49 TABLE 7.— Intraspecific Ranges of Variation in Ventral Counts of Calamaria in Local and Species-wide Samples. Species lumbricoidea grabowskyi suluensis griswoldi lumbricoidea linnaer linnaei schlegeli schlegeli pavimentata borneensis lowr gervaist modesta modesta acutirostris mueller bicolor sumatrana melanota leucogaster septentrionalis cies falling between 15 and 30 in 89 species. variation are variable (Table 8). Single locality Males Females , TS No. Range No. Range [Mount Kina Balu, Borneo] 7 154-165 7 170-180 6 176-181 5 180-190 2 131-136 3 144-147 4 166-179 5 183-192 [Tjikadjang, Java] 26 179-187 30 197-209 25 182-142 25 149-159 [Wonosobo, Java] 12 184-146 19 150-164 44 145-155 54 162-173 [Penang Island] 4 144-151 4 160-167 [Suisharyo, Formosa] 4 148-155 10 166-176 [Long Mujan, Borneo] 9 155-169 5 170-182 3 198-202 5 221-226 [Manila, Luzon| 25 147-158 26 157-190 [Bandung, Java] 28 165-179 19 191-202 [Tjibodas, Java] 18 155-166 25 165-194 ilar to those of ventral counts. Sexual dimorphism in subcaudal counts (see Table 8) follows the pattern common in the Colubridae. Entire geographic range Males Range 144-196 150-186 129-138 155-179 144-196 130-149 130-149 129-161 129-161 125-168 126-169 163-213 132-164 131-179 131-179 148-161 129-155 139-169 126-157 121-142 126-146 148-166 Females io No. 101 Range 137-229 164-190 142-168 183-192 34-229 148-166 148-166 136-180 136-180 137-206 159-192 215-256 142-190 158-202 158-202 163-174 155-178 151-160 164-175 131-154 129 157 168-188 Intraspecific ranges of The patterns of variation are sim- Males have more subcaudals INGER AND MARX: THE SNAKE GENUS CALAMARIA 29 130] 120 110 100 90 80 70 60 sample _ size 50 40 30 species - wide 20 10 10 20 30 40 50 60 70 80 90 species- wide variation 1n_ ventrals Fic. 7. Relation between intraspecific variation in ventral counts and species- wide sample size in Calamaria. than females in all species of Calamaria except possibly suluensis (Table 36, p. 128) and everetti (p. 183). As in the case of ventral counts comparison of the sexes should be based on specimens from a single locality because of geographic variation. Females of lwmbri- 30 FIELDIANA: ZOOLOGY, VOLUME 49 TABLE 8.—Intraspecific Ranges of Variation in Subcaudal Counts of Calamaria in Local and Species Wide Samples. Single locality Entire geographic range Males Females Males Females Go, eS SE See Species No. Range No. Range No. Range No. Range [Mount Kina Balu, Borneo] lumbricoidea 7 238-24 7 17-19 184 17-27 118 = 13-21 grabowskyt 6 26-29 5 21-24 15 23-29 18 20-28 suluensis Z 18-20 2 17-20 6 18-20 13 14-26 griswoldi 4 16-17 4 138-16 6 16-18 4 18-16 [Tjikadjang, Java] lumbricoidea 26 17-21 30 =. 18-16 184 17-27 118 13-21 linnaer 25 15-19 yas 8-12 74 15-22 95 7-13 |Wonosobo, Java] linnaet 11 17-22 19 9-13 74 15-22 95 7-13 schlegeli 42 28-338 52 23-28 99 25-44 113 19-87 [Penang Island] schlegeli 4 29-32 4 24-25 99 25-44 113 19-87 [Suisharyo, Formosa] pavimentata 4 20-23 10 +=16-18 38 18-33 50 8-20 [Long Mujan, Borneo] borneensis 8 20-22 5 13-19 24 20-26 28 13-21 low2 3 19-22 5 12-14 26 14-26 37 ~=610-18 [Manila, Luzon] gervaist 41 15-19 34 10-14 110 15-21 129 #£10-18 [Bandung, Java] modesta 31 25-29 19 15-19 100 += 19-81 19 W222i [Tjibodas, Java] modesta 14 24-28 18 13-18 100 +=19-81 19) W220 acutirostris 18 20-24 16 13-17 bicolor 12 21-28 7 Le=Zil leucogaster 14 17-26 22. 7 W219 melanota 9 238-26 9 16-20 muelleri 29 16-21 38 9-15 septentrionalis 28 15-19 30 6-11 sumatrana 6 14-20 tal 10-14 coidea from Mount Kina Balu have almost as many subcaudals as males of lwmbricoidea from Tjikadjang, Java (Table 8), though sexual dimorphism is clear at each locality. The evidence for lack of dimorphism in sulwensis and everettz is equivocal. At Mount Kina Balu the few males and females seen of INGER AND MARX: THE SNAKE GENUS CALAMARIA 31 suluensis had almost identical subecaudal counts, whereas in south- eastern Borneo males seem to have slightly higher counts (Table 8). In the case of everettt, none of the males came from the localities of the females. Within local populations the intraspecific range of variation is usually less than six subcaudals. Neither sample size nor maximum number of subcaudals affects the local range of variation (Figs. 8 and 9). When entire species samples are analyzed, intraspecific ranges in- crease primarily because of geographic variation, which is relatively common (see Tables 26, 30, 45). Intraspecific ranges at this sam- pling level vary according to the maximum individual counts within species (Fig. 8); a species such as schlegeli (maximum subcaudal count 44) can have a total range of variation of 20 subcaudals whereas a species such as septentrionalis (maximum subcaudals 19) cannot. The size of the total species sample had little effect on range of vari- ation (Fig. 9). Because of narrow local variation, subcaudal counts are helpful in differentiating between sympatric species. Their usefulness in in- terspecific comparisons when allopatric samples are involved is not as great, though some species are decidedly short-tailed and others long-tailed. Reduction of the number of dorsal scale rows on the tail.—The num- ber of dorsal scale rows on the tail is reduced to four in most species of Calamaria. As in our previous study, we have located the posi- tion of the reduction to four rows by means of the subcaudal oppo- site which it occurs. In Table 9 we have listed the position of the reduction in terms of subcaudals between the level of the reduction and the vent and between the reduction and the terminal scute. Individual variation is extensive in most species and in both sexes, yet not so extensive as to obscure sex dimorphism and interspecific variation (Table 9). The position of the reduction to four rows is affected by at least two factors. Whether measured from the vent or from the terminal scute, the position of the reduction is correlated with the total num- ber of subcaudals. It is self-evident that, as the total subcaudal count decreases, the maximum number of subcaudals separating the level of the reduction from either end of the tail must decrease. Yet the total number of subcaudals accounts for only slightly more than half of the variation in position of the reduction (or between 0.39 and 0.82; see Table 10). 32 FIELDIANA: ZOOLOGY, VOLUME 49 The other factor influencing the position of the reduction is tail shape. This influence is reflected in sex dimorphism and in inter- specific variation. Males of Calamaria are like those of most other snakes in having thicker tails basally than females. This difference is evident in the consistently greater mean distance of the level of the reduction from the vent in males (Table 9). Only four of the 21 means of males are less than 11 subcaudals from the vent and hence in the section of the tail sheathing the hemipenes. By contrast 13 of 21 means of females lie within 11 subcaudals from the vent. Sim- variation range of 14. 16 18 20 22 24 26 28 30 32 maximum number of subcaudals within local samples x 20 x O 18}— 16 14 oO xO 612 Xx os xX Lavi “S211 oO X = 6%) co 8 re) x x ey fe) (@) Oo x X » rr i eRe re nk eo 6 9 x o = “0 0. +O KC xO X x= o 4 fe) fe) x :0=Q I2514 16" VBS 20" 92204245 826728) 305482 34 SG) S8ae40l e425 eas maximum number of subcaudals within species Fic. 8. Relation between intraspecific variation in subcaudal counts and maxi- mum number of subcaudals in local (top) and species-wide samples (bottom) of Calamaria. INGER AND MARX: THE SNAKE GENUS CALAMARIA 33 130 120 re) — 110 x = 100 x x a a 70 sample size oO ro) i 20 |= Q cone 30 x“ ® ° pees Species - wide my ro) © { | local sample size 10 4 6 8 10 12 14 16 18 20 species-wide variation in subcaudals local variation in subcaudals Fic. 9. Relation between intraspecific variation in subcaudal counts and size of species-wide (left) and local (right) samples of Calamaria. ilarly the minimum distance from the vent in the 21 species listed in Table 9 is less in females than in males; the minimum distance in females of all 21 species falls within 10 subeaudals (11 of them within 5 subcaudals) whereas only 14 of the minima of males lie in that range (only three within five subcaudals). Sex dimorphism in the number of subcaudals between the level of the reduction and the terminal scute is, with few exceptions, not significant statistically or otherwise (Table 9). The importance of shape in determining the position of the re- duction is also shown by interspecific differences. In species having slender tails the reduction takes place much farther from the terminal scute than in thick-tailed species. Males of septentrionalis, linnaer, and griswoldi have approximately the same number of subcaudals G’g €=0 Del Eig LLP “ors 6'§& c-0 ES. oe O°cI FVI-OL 6° cI 61-6 GL 91-36 0°6L 9c-FI LOT LI-8 I'v 8t-ol 6°06 86-6 LS Gor, L°OT T6-G G°96 LE-6I 0 0 T°6 =TT-9 6 9 8'§ 8-0 Vil ol-7 GST 06-8 87 L-ZG 6L TI-¥ 8°cI SI-6 6 & =. OL] Sol si-s 8°9T Té-0.10 +0.82 <0.001 total). The tail of griswoldi (Fig. 10B) is one of the most slender in the genus and the reduction is 11.2 subcaudals (67 per cent) from the terminal scute. Males of modesta and grabowskyi have equally long tails (means of subcaudals 26.3 in both), but that of grabowskyi (Fig. 10D) is tapered near the base and that of modesta (Fig. 10C) only near the tip. The mean position of the reduction in grabowskyz males is 18.5 subcaudals (70 per cent of the total) from the terminal scute and 3.7 (14 per cent of the total) in males of modesta. Thus despite individual variation and sexual dimorphism, the position of the reduction to four rows can be used as a valid taxo- nomic character and in some cases helps in distinguishing between species. Because sexual dimorphism in the number of subcaudals between the reduction and the terminal scute is less pronounced than in the number between the vent and reduction, we have given the position of the reduction relative to the end of the tail in the descrip- tions of species. Hemipenis and cloaca.—The retracted hemipenes of slightiy more than half of the species were examined (Table 11). The material available was not sufficient for adequate sampling of each species. We dissected the hemipenis of a single male in nine species, of two males in 11 species, of three males in four species, of four males in two species, of five males in one, of 20 males in one, and of 46 males in one. Our analysis of variation is, therefore, of limited value. TABLE 11.—Frequency Distribution of Species of Calamaria with Respect to Form of Hemipenis. Body of hemipenis Forked or Forked Non-forked non-forked Unknown Calyces SMO OU vc hee 17 ik Papillate:::. 0. ere eee. 6 1 Smooth or papillate...... i 1 Witknowine- eee ree: 2 21 INGER AND MARX: THE SNAKE GENUS CALAMARIA 37 The most common forms of individual variation involve the posi- tion of the fork and the length of the hemipenis. The fork of the retracted hemipenis usually lies opposite the fourth to sixth sub- caudal. In 24 males of lwmbricoidea, the fork was opposite the second subcaudal in two specimens, opposite the third in three, the fourth in 11, the fifth in seven, and the sixth in one. In 12 males of schlegela the fork lay opposite the fourth subcaudal in six, the fifth in five, and the sixth in one. Similar variation occurs in 14 of the 16 other species which have forked hemipenes and of which we examined two or more males. TABLE 12.—Relations of Length of Hemipenis to Total Number of Subcaudals in Two Species of Calamaria. C. lumbricoidea Total number of subcaudals End of hemipenis! 19 20 21 22 28 24 25 26 27 Number of specimens 6 1 1 7 1 8 1 3 1 1 9 3 2 2 2 10 1 3 1 i 1 li 1 C. schlegela Total number of subcaudals 27 28 29 30 31 34 39 40 42 Number of specimens 7 1 8 1 1 9 2° J 1 iL 10 1 1 1 11 1 1 Determined by subeaudal opposite which retractor muscle inserts on hemi- penis. We have measured the length of the hemipenis in terms of sub- caudals, letting the insertion of the retractor muscle mark the end of the hemipenis. The retractor may insert as close to the vent as the sixth subeaudal and as far as the fourteenth; in most males ex- amined the insertion is opposite the eighth to tenth. In the two large series examined, lwmbricoidea and schlegeli, the length of the hemipenis varies. But that variation bears little rela- tion to tail length as measured in terms of subcaudals (Table 12). Similar minor variation occurs in other species. 38 FIELDIANA: ZOOLOGY, VOLUME 49 What may be called major, in contrast to the preceding minor, variation appears in sulwensis and lumbricoidea. The calyces of one suluensis male are smooth-edged, those in a second papillate. The hemipenes of lwmbricoidea are simple (non-forked) with smooth-edged calyces, forked with smooth calyces, or forked with papillate calyces. These variants are geographically localized (Inger and Marx, 1962; see also p. 85 of this paper). As we examined at most two males in 20 of the 29 species dis- sected, the true extent of major intraspecific variation in hemipenes cannot be evaluated from our data. Intraspecific variation as strik- ing as that of lwmbricoidea should not be expected in every species, for only minor variation was found in 20 males of schlegeli from all parts of that species’ wide range. Most species of Calamaria have deeply forked hemipenes having smooth-edged calyces (Table 11). Although the characteristics of many of the species in the table are based on the dissection of one or two males, the previous generalization seems valid. Even if major intraspecific variation is much more common than our data indicate, it is unlikely that we would have found such a high proportion of snakes falling in one cell of Table 11—provided that our selection of specimens for dissection was random with respect to hemipenial form. As we found no ontogenetic variation in hemipenes and as we could not know the form of the hemipenis prior to dissection, randomness in our selection was assured. We dissected two males of each of the two species characterized by non-forked hemipenes and two or three males of four species char- acterized by papillate hemipenes. Despite the small samples in- volved, we believe that much of the interspecific variation indicated in Table 11 will be substantiated by future work. The cloacas of females have three general forms (Inger and Marx, 1962). Cloacas called bulbous have oviducts opening into the cham- ber posterior to the anterior wall of the cloaca. If the anterior end of the cloaca has an indentation and the oviducts open posteriorly, the cloaca is called cardioid. A bilobed cloaca has the oviducts open- ing into the anterior ends of short cloacal lobes. Females of 20 species were dissected. We examined one female of each of seven species, two of each of 10 species, three of one, four of one, and 27 of one (lwmbricoidea). Only four of the two-specimen samples show no variation. The cloacas of the other multi-speci- men samples fall in at least two categories. INGER AND MARX: THE SNAKE GENUS CALAMARIA 39 Variation in cloacal form is apparently not related to stage of the reproductive cycle (Inger and Marx, 1962). Possibly the shape of the cloaca is transitory and modified by differential contraction of muscles. In view of the wide intraspecific variation, we have not utilized cloacal form in our taxonomic decisions. Thickness.—Species of Calamaria clearly differ in body thickness. Great variation in quality of preservation made it impractical to measure the thickness of the body directly. The only satisfactory method of determining bulk was by water displacement which gave us the volume of each animal. Obviously shrunken and broken ani- mals were not used. Unfortunately these measurements were not made until relatively late in our study after a good part of the bor- rowed material had been returned. A graduated cylinder was partly filled so that the meniscus was at a given line. A snake was submerged and the increase in volume (to the nearest 0.5 ml.) gave us the volume of the snake. The volume was divided by the total length of the individual giving the volume per millimeter of length. This measure is an index of thickness. Despite the limitations of small samples, these measurements give clear indication of ontogenetic changes in body thickness, weaker in- dications of sexual dimorphism, and good indications of interspecific variation (Fig. 11). We have made volume measurements for 43 species (Table 13); 21 species are represented by enough measurements to indicate at least the pattern of individual variation. Eighteen of the 21 show ontogenetic variation clearly. Each of these eighteen species shows increase in the volume per millimeter of total length as total length increases. The three species that do not show clear ontogenetic change in this character are lewcogaster, lowi, and sumatrana; the last two species are slender and the first is moderate in thickness. In lumbricoidea, gervaist, schlegelt, and virgulata the males seem to be somewhat bulkier than females of corresponding lengths. Even in these species, however, the data are too few to substantiate this relationship. Differences between species in thickness show up with even quick visual examination. For example, evselti is several times the diameter of lowz; these two species represent extremes of thick- ness in this genus (Fig. 11). The results of visual inspection are con- firmed by our measurements of volume (Table 13). Even relatively subtle differences between species (e.g., evselta vs. crassa or lumbri- coidea) can be detected by these measurements. 00s 40) i) 80 Ol ‘DUDUWDIDI Jo Setoeds ve1y} Ul (“WUI/*;U) Apoq Jo sseuyoIy, (ww) HIONTT TWLOL “TT SI 007 iy[est2 IMO] voplooliqun | SZl tO £0 vO SO 90 ZO 80 60 Ol AdOd XFQNI SSINMOIWL 40 INGER AND MARX: THE SNAKE GENUS CALAMARIA 4] TABLE 13.—Body Volume and Ratio of Volume to Total Length and Numbers of Ventrals and Subcaudals in Calamaria. Museum number lowt SM ‘14’ CNHM 109974 CNHM 72373 CNHM 129001 CNHM 109975 CNHM 129002 CNHM 109976 CNHM 129003 alidae NHMW 16996 apraeocularis MCZ 25265 MCZ 25300 forcarti N MB 8958 ZMA 10073 NHMW 16710 buchi CNHM 71697 pavimentata CNHM 100869 CNHM 11528 virgulata ZMA 10240 RMNH — RMNH -— SU 22405 RMNH —— RMNH —— RMNH -— NHMW 16695 NMS 4 NMS 4 CNHM 63571 NHMW 16716:2 RMNH —— RMNH 39 NHMW 16716:1 MHNP 5780 SNG 19441 joloensis CAS 60901 boesemani GM 3230 Locality Borneo oF Sumatra Celebes Nias 9: Sumatra Indo-China Tonkin Annam Sumatra Celebes Mindanao Celebes +9 Java Celebes Borneo Celebes Java Celebes no data Celebes Jolo Celebes Sex 40 40 10 QV 40 Q, 40 40 40 Qy Qy 40 A Qy 40 40 40 Qy Ay 40 40 10 10 10 Qt OQ A110, Qy Total Length Vol. mm. 238 238 239 268 276 289 291 304 254 148 237 216 235 239 389 268 244 100 150 173 193 195 219 225 237 254 257 258 271 273 283 286 307 336 144 116 ml. DO bo bo ovc WPF EN WHh DD BDEUAnNIEP EE PWWWNNR © OVdtAN On OUOr Vol./TL ml./mm. .008 .008 .018 .009 .014 .014 .014 .010 cocococococo°o 0.012 0.007 0.013 .009 .009 .008 ooo 0.021 .O11 .012 oo .008 .010 .014 .010 .015 .014 .018 O17 .016 .016 .016 .018 .023 .021 .019 .015 .018 cooocoooooooooocochwc 0.010 0.009 Ven- trals 226 225 202 238 196 246 233 228 203 178 220 176 200 177 221 186 160 195 180 182 227 160 182 192 197 208 208 205 216 194 189 192 236 209 119 170 Sub- caudals TABLE 13 (cont.) Museum number Locality brongersmar GM 3229 everettz RMNH -— CNHM 63572 BM 1902 a CNHM 109971 os javanica CNHM 109791 _ Billiton Celebes Borneo be ] lateralis MCZ 43582 Borneo sumatrana MCG 30382 Sumatra ZMA 10238 ie ZMH 3996 : AMNH 2873 ” ZMH 3996 uP RMNH 4860 fs ZMH 2466 m2 ZMA 10237 oe ZMH 2307 22 abstrusa ZMB 5986 ZMB 5986 BM 64.4.7.11 no data 99 Sumatra? borneensis CNHM 109973 Borneo SMNS 4584 u CNHM 76295 a SMNS 4584 = CNHM 72372 ” CNHM 109972 sd suluensis SMNS 34.98 RMNH — RMNH — ‘ MHNP 57-812 2. NHMW 16708:2 ZMH 4212 is NMS 146.17 oe NHMW 16708:1 NHMW 16708:3 Ss CNHM 76294 ae Borneo ’)9 palavanensis BM 94.6.30.51 CAS 62151 BM 94.6.30.50 +4 Palawan 9) gervaist SU 17931 Negros SU 18229 Cebu 10 70 QA +O +O 40 40 Q)40 1040 Q AQ, 40 Ay A 40 40 40 40 QQ Q, Q, 40 40 40 40 QV 40 10 410A, 40 QV 40 Qy 40 Total Length, Vol. Tol: mm. 225 144 168 187 208 169 190 168 169 174 184 200 208 222 227 234 162 162 209 209 210 228 244 275 353 109 165 190 203 221 227 252 259 264 287 134 280 323 ee 120 42 COD re St Loe et SS) WPWNNMRND W Dee OOnOnOFPEPWWHNHE CIN OVE Pah) .75 .20 .75 Vol./TL ml./mm. 0.018 o°ococo cococooooococc ooo ooococococococoe ocooocooo ooo .014 .009 .O11 .010 .012 .O11 .018 .015 .014 .010 .013 .010 .014 .018 .015 O11 Od .012 .014 .014 .013 .016 .022 .031 .009 .012 .016 .015 .019 .020 .020 .026 .023 .028 .007 .021 .025 .009 .008 Ven- trals 155 157 151 152 153 168 150 126 130 149 165 167 173 157 164 174 130 129 152 138 175 171 176 159 177 147 142 158 129 138 133 147 154 161 164 178 174 180 170 149 Sub- caudals 18 TABLE 13 (cont.) Museum number CM 2530 SU 19874 CAS 60471 CNHM 53377 CM 2470 CM 2540 MCZ 37699 MCZ 25775 CM 2485 SU 17929 CNHM 96620 SU 18910 CM 2537 CNHM 533880 MCZ 25752 CM 2538 SU 15965 SU 15963 SU 15947 nuchalis USNM 61201 RMNH -— RMNH 83882 MCZ 45486 NHMW 16715 RMNH -— USNM 61193 leucogaster RMNH -— RMNH -— RMNH -— CNHM 71598 RMNH -— SMNS 4585 RMNH 3994 curta MCZ 25302 MCZ 25301 septentrionalis CNHM 7141 ZSBS 169/47 CNHM 7140 CNHM 7139 déderleini MZUS —— schlegeli ZSBS 1933/0 NHMW 16712:6 CNHM 121033 SMNS 4586 NHMW 16712:3 NHMW 16696:1 NHMW 16712:2 CNHM 121039 MHNP 4417 Locality Luzon Negros Basilan Mindanao Luzon %99 Mindoro Polillo Luzon Negros Mindanao Negros Luzon Mindanao Tablas Luzon Panay Negros Celebes Sumatra Borneo Sumatra Borneo Sumatra Borneo Sumatra Celebes China 99 >? 99 Sumatra Borneo Sumatra Java Borneo Sumatra Java Sumatra Java Billiton TM i) ~ 40 40 40 40 40 40 10 QV QV QQ 4H WO QAAAAQAQ+O AAW AAQ\MAQQA, 40 40 Qh 40 40 40 40 40 40 409, 40 40 % Ay 40 QV A 40 AA 40 40 Total Length, mm. 135 149 170 179 186 186 190 191 196 199 204 210 245 258 262 263 277 281 321 123 184 195 200 249 280 294 151 162 177 195 198 205 223 195 280 168 263 277 342 288 163 189 198 255 275 283 287 317 326 43 Vol. SOy,e fF owH _ BoB Roo Do o one) CONBCNNMDWWP DHARAAHAIWEHPWNNMNNMNMNNH 5 ° . — Vol./TL ml./mm. 007 .013 12 O11 O11 .013 O11 .010 .015 .020 .020 .014 .020 .023 .023 .023 .022 .030 .025 coooooooococoooqoooqocoocno oooococoo oo oooocooo oooco So oooooooooa .008 .016 .026 .020 .016 .029 .034 .015 .006 .028 .026 .020 .020 .018 .015 .029 .021 .024 .029 .047 .038 .012 .016 .018 .024 .025 .027 .028 022 .026 Ven- trals 162 159 133 145 153 146 172 148 155 154 147 148 160 150 163 169 168 177 177 146 142 133 145 156 143 145 126 147 127 148 146 151 143 170 156 160 178 180 172 163 148 152 146 129 152 159 138 169 135 Sub- caudals TABLE 13 (cont.) Total Museum Length, Vol. Vol./TL Ven- Sub- number Locality Sex mm. ml. ml./mm. trals caudals griswoldi NMS -— Borneo ot 236 4 0.017 166 16 CNHM 72434 be of 334 i Bs 0.039 170 17 USNM 134114 és Q 376 9 0.024 190 14 NMS 18949 - ot 379 14 0.037 val 16 MCZ 438580 ” Q 488 18 0.037 186 13 grabowskyr CNHM 76293 — Borneo o 182 2 0.011 186 28 NHMW 16998 % oh 209 6 0.023 150 23 RMNH -— ” of 287 8 0.028 160 25 RMNH -— e ‘eh 307 i) 0.029 162 24 MCZ 43573 e ot 316 8 0.025 176 2 ZMA 10230 ‘ Q 340 10 0.029 180 Oe, NHMW 16693 ‘is Q S00 14 0.039 168 21 MHNP 89-192 ” of 362 11 0.030 181 28 MHNP 89-191 i Q 388 10 0.026 189 VAIL RMNH 568 i Q 451 28 0.062 190 20 ceramensis CNHM 83464 Ceram 2 266 ij 0.026 154 19 ZSBS 947/20 - Q 301 11 0.037 165 20 lautensis RMNH 4716B _ Kokos Ids. ee 215 6 0.028 123 24 modesta MHNP 3299 Java Q io 2 0.013 192 18 Z5BS 1935/0 Pe Q 201 3 0.015 174 21 ZSBS 2635/0 fe Q 224 33) 0.016 199 19 ZSBS 2646/0 is of 269 8 0.030 154 31 ZSBS 2646/0 « Q 269 7 0.026 166 il ZSBS 1934/0 Z ot 286 rh 0.024 iL 26 ZSBS 2646/0 a of 304 bt 0.036 154 29 ZSBS 2675/0 ‘ Q 319 9 0.028 173 18 CNHM 83164 i o 343 12 0.035 162 25 ZSBS 2675/0 is fot 352 12 0.034 171 7A bicolor SM “164” Borneo oy 164 2 0.012 167 23 NHMW 16997 :2 i of 178 3 0.017 139 Oe, MCG 748.36 4 Q 181 4 0.022 159 PAu RMNH 68 ‘ en 281 i 0.025 152, 24 RMNH 1678 a Gh 284 9 0.032 145 21 RMNH -— iJ oh Slat 10 0.032 142 26 RMNH 10542 is 2 3515) 14 0.042 156 20 SM Cd.5.26.7a ‘a Q 391 21 0.054 159 20 muelleri MCZ 25332 Celebes Q 146 il 0.007 fall 13 NHMW 16692:7 té Q 208 3 0.014 165 13 USNM 1208138 a ot 214 A.) 408021 152 18 ZMB 156385 a Q 218 4 0.018 170 18} MCZ 25322 x of 235 6.5 0.028 148 18 SNG 19443 ‘s Q 242 8 0.033 158 13 SNG 19442 fo Ct 254 8 0.031 141 23 MCZ 25317 i Q 298 8.5 02029 169 14 MCZ 25315 te 2 320 9 0.028 2 14 ZMB 13946 ce Q 355 15 0.042 167 2 TABLE 13 (cont.) Museum number hilleniusi ZMA 10078 SM 66 1 99 acutirostris NHMW 187138:11 NHMW 1387138:10 NHMW 1387138:7 NHMW 1387138:9 NHMW 138718:5 NHMW 138713:3 NHMW 1387138:6 NHMW 187138:1 NHMW 1387138:4 NHMW 187138:2 albiventer CAS 14940 crassa NHMW 16994 RMNH —— RMNH 34 RMNH 34 RMNH 34 ZMB 5229 RMNH 34 RMNH 34 melanota ZSBS 2039/0 ZSBS 2039/0 SMNS 4529 RMNH 1152 CNHM 109964 lumbricoidea NHMW 16992 :2-3 NHMW 16992 :2-3 NMS —— RMNH 932 SMNS 3497 RMNH — RMNH 4865 NMS -—— RMNH 1264 CAS 15289 ZMA 10234 RMNH 5806 NMS 1932 MCZ 20971 RMNH 5806 MCZ 43568 ZMH 2466 RMNH 5806 NMS —— NHMW 16700:1 ZMB 13110 Locality Borneo ’) Celebes Singapore Sumatra 99 Borneo bE Malaya Java Borneo Sumatra Nias Malaya Java Mindanao Nias Sumatra Malaya Sumatra Borneo Sumatra de) Malaya Sumatra Borneo Sex 40 40 FQ, 40 AQ, 40 QVQ,40Q, one e +O 40 Q9 40 40 10 40 QQ, Ay +0 40 QA, WI AAAAAWOAAYO OW AAAAAIO A Total Length, Vol. mm. ml. SLT 13 368 19 138 2 150 2, 173 5 193 4 291 14 296 14 330 14 341 24 370 20 447 18 200 6 131 2 168 vA 227 5 330 13 300 14 Bo4 1s B42 16 ota 19 182 4, 189 4 202 3 218 6 230 6 170 2 176 3 177 3 186 3 190 5 193 o 246 6 249 4 254 7 254 5 259 6 277 fa 294 ‘- 301 11 306 14 347 15 351 9 353 1} 367 14 379 bh 382 17 45 Vol./TL ml./mm. oooocococoocoo oo So ooocococooco ooococoe ooococoooococooooococnhwmc So 041 .052 .013 .018 .029 021 .048 .047 .042 .070 054 .040 .026 .015 s08Z .045 Ven- trals Sub- caudals 18 16 TABLE 13 (cont.) Museum number RMNH 618 NHMW 16701:2 RMNH 588 MHNP 89-193 NHMW 16702 margaritophora ZSI 149386 MHNP 39186 NHMW 16707 RMNH 4689 RMNH — ZMB 8462 USNM 70952 bitorques MCZ 25776 ZMB 7444 ZMB 3778 MHNP 00-364 CAS 15295 SNG 19394 MHNP 00-3865 linnaet RMNH -—— RMNH — RMNH — RMNE =— RMNH — RMNH —— RMNH — RMNH —— RMNH 52 RMNH —— prakkei ZSI 13306 RMNH 4360 erseltz NHMW 16701:1 NHMW 16703:4 NHMW 16703:1 NHMW 16703:3 CNHM 134723 NHMW 16703:2 Locality Java Sumatra Malaya Java Malaya Java Borneo Sumatra Java +2 4 9:3 Sumatra 99 +9 99 Phil. Ids. Luzon b He no data Java Singapore Borneo Sumatra ” NM f¢°) ~ 40 40 40 40 40 40 40 40 Q 40 4 AAA Ay 40 Ay AQ QY A QA 40 Ay Q, 40 10 10 10 AQ +0 Qy % 40 40 40 QV Qa 40 Total Length, Vol. mm. ml. 409 21 410 10 413 16 451 18 454 22, 460 34 472 ZZ 490 23 200 4 234 8 255 6. 256 10 259 6 261 9 358 16 250 4 329 10 329 16 328 12 398 Zi. 412 26 420 24 122 Bt 171 2 Zoo 6 242 5 245 6 281 9 292 12 307 20. 336 20 396 38 204 4 243 6. 244 q 333 22 343 2A 365 24 381 28 424 42 Vol./TL ml./mm. .051 .024 .039 .040 .048 .074 .047 .047 oocooocoooc]oa .020 .034 .025 .039 .023 .034 .045 ooococeoo .016 .030 .049 .037 .053 .063 .057 cococococeoe]o .008 O17 .026 .021 .024 .032 .041 067 .060 . 100 ocooocococoooo .020 .027 oo .029 .066 .061 .066 .073 .099 oo°ococoo Ven- trals 182 191 162 187 160 181 179 169 152 150 157 152 160 159 161 172 175 157 170 174 167 168 159 140 143 157 154 154 159 137 150 154 126 130 153 137 137 151 151 151 Sub- caudals As with many other characters, the many known species of Cala- maria can be arranged in order of increasing thickness and form a continuous gradient from one extreme to the other. Coloration.—Taken as a whole, the genus Calamaria is highly vari- able. 46 The dorsal pattern may consist of broad or narrow stripes, INGER AND MARX: THE SNAKE GENUS CALAMARIA 47 narrow or broad transverse bands, and spots of varying sizes. Or the dorsum may be unmarked or divided into dark upper and light lower portions. The individual scales, where they are not involved in the spots, stripes, and bands of the pattern, may be a solid color (e.g., Fig. 40), or have a fine network or speckling on a contrasting background (e.g., Fig. 33). In many species the scales of the first row have light centers and dark edges and form a continuous light stripe or a row of light dots. Dorsally the head usually has the ground color of the body, often with obscure, irregular spots. The upper lip and the underside of the head are usually light and often have dark spots on scale sutures. In a few species the head and trunk colors contrast sharply. The ventrals of most species are yellow or whitish. The common- est pattern consists of a dark spot at the lateral corners of each ven- tral. In some species (e.g., ezseltz) the anterior ventrals are yellow whereas those farther back have a black area that becomes progres- sively larger on successive plates until it covers the entire width of the belly. In some species each ventral has a dark anterior band and a light posterior one (Fig. 56). A pattern consisting of wider dark bands on a light background occurs in two species. A checkered pattern (Fig. 13) occurs in several species. The ventral surface of the tail usually has a continuation of the belly coloration, often with the addition of a dark mid-ventral stripe. Some variation occurs in almost all species. But the type and extent of intraspecific variation differ from species to species. In some species, for example linnaez, both dorsal and ventral colorations vary widely (Figs. 12 and 13). Others, e.g., griswoldi (Fig. 23) do not vary at all. Although we have seen many more linnaez (659) than griswoldi (12), the differences between these two species in ex- tent of variation is not attributable to sample size because rarely do any two specimens of linnaez resemble one another as much as do all 12 of the griswoldz. The patterns of the different parts of the body may vary in differ- ent ways. Generally, the coloration of the head varies less than the trunk. In alidae (p. 236), for example, the trunk may be striped, spotted, or banded, whereas the head varies only in the number of dark spots. By contrast, the trunk of schlegeli (835 examined) is always dark brown or black above the unspotted, yellow first two scale rows, whereas the head may be completely yellow or completely dark brown or any intermediate bicolored condition. 7 from linnae dorsal pattern of Calamaria 1n . lation il var Specimen at bottom of middle row from Bandung, Java. 12. Individue ang, Java. ‘ Ie Fi Tjikad) 48 Fic. 18. Individual variation in ventral pattern of Calamaria linnaei from Tjikadjang, Java. 49 50 FIELDIANA: ZOOLOGY, VOLUME 49 Ontogenetic variation characterizes lumbricoidea (p. 82) and prob- ably bicolor (p. 153). In luwmbricoidea both head and trunk are in- volved but though darkening of the color is the age trend, the change is not exactly parallel in the two areas (Figs. 21 and 22). In bzcolor the entire dorsal surface, except for the tail, becomes darker with age. Intraspecific variation may be common in a local population. All of the snakes (linnaet) in Figs. 12 and 13 except one are from Tji- kadjang, Java. At Bukit Tinggi, Sumatra, seven schlegeli have three of the types of head coloration listed in Table 44. Other species may show little individual or local but conspicuous geographic variation. All specimens of lumbricoidea from Borneo are uniformly dark brown above the first two scale rows; those from the Philippine Islands are striped; those from eastern Sumatra are non-striped and those from western Sumatra striped (Table 21). Despite local and geographic intraspecific variation, coloration frequently is one of the characters differentiating similar species. Striped griswoldi, for example, is immediately distinguishable from the sympatric, non-striped population of lwmbricoidea. Even striped lumbricoidea differ from griswoldi as the light lines of the former are confined to the very edges of the mid-dorsal scales whereas in gris- woldi they cover at least the lateral fourths of each scale (Fig. 28). Furthermore, the belly of griswoldi is immaculate yellow and that of lumbricoidea is almost always banded. Another example of interspe- cific variation involves gervaist, which is non-banded, and bitorques which is. Calamaria crassa and eiseltt form another closely related pair differing in coloration (cf. Figs. 46 and 47). Interspecific differences are not universal, however, and some spe- cies (e.g., grabowskyi, suluensis, and palavanensis) cannot be distin- guished on the basis of coloration. Because of such interspecific similarity and because of wide intraspecific variation in other cases, color differences have to be carefully analyzed in conjunction with other characters before interpretations are made. But this caveat applies to most of the characters used in this study. Correlation of head characters.—Although a certain amount of in- dividual variation occurs in most characters, we consider a species to be variable (e.g., preocular present or absent; mental touching or not touching chin shields) for the purposes of this section only if the variation is geographic or if, as in the case of lumholtzi, the two to four specimens examined were variable. Association of pairs of cephalic characters was tested by means of contingency tables. Chi-square values reached significant levels INGER AND MARX: THE SNAKE GENUS CALAMARIA 51 TABLE 14.—Association of Presence of Preocular with Size of Eye in Species of Calamaria (Numbers in body of table are numbers of species). Preocular Observed Calculated Present Present or or Eye size! Present absent Absent Total Present absent Absent Greater 16 0 0 16 12.5 0.6 2.8 Equal 19 0 3 22 17.2 0.9 4.0 Less 4 ve 6 12 9.4 0.5 Zee chi-square=20.1; n=4; P <0.001 1 “Greater,” ‘“‘equal,’’ and “‘less’’ refer to eye diameter relative to distance of eye from mouth. Species varying from “less” to “equal’’ or from ‘‘equal’’ to “sreater’’ are listed as “equal.” (P< 0.05) in only five tests: preocular and eye size (Table 14) ; frontal ratio and eye size (Table 15); preocular and frontal ratio (Table 16) ; preocular and relation of mental to chin shields (Table 17); number of scales surrounding paraparietal and number of supralabials (Ta- ble 18). Results of tests between all pairs of the cephalic characters are summarized in Table 19. The association of frontal ratio, eye size, and preocular based earlier on data taken from literature (Marx and Inger, 1955), is con- firmed by the material we have since examined (Tables 14-16). These three topographically related characters tend to change as a unit and are what Cain and Harrison (1960) call ‘‘necessary corre- TABLE 15.—Association of Frontal Ratio with Size of Eye in Species of Calamaria (Numbers in body of table are numbers of species). Eye size Observed Calculated —_ os = Frontal — ratio! Greater Equal Less Total Greater Equal Less —2 3 0 0 3 1.0 1.3 0.7 2-214 10 5 i) 16 5.1 7.0 3.8 216-21/s 3 5 3 Ly. 3.5 4.8 2.6 3-34 0 10 t 14 4.5 6.2 3.4 4— 0 2 4 6 1.9 2.6 1.4 Chi-square=27.2; n=8; P <0.001 1 Ratio of frontal width to supraocular width. Species categorized by intra- specific maxima. 52 FIELDIANA: ZOOLOGY, VOLUME 49 TABLE 16.—Association of Presence of Preocular with Frontal Ratio in Species of Calamaria (Numbers in body of table are numbers of species). Preocular Observed Calculated Present Present Frontal or or ratio Present absent Absent Total Present absent Absent —2 3 0 0 3 Ape: 0.1 0.5 2-214 16 0 0 16 12.5 0.6 2.9 214-21/s 9 1 i) i 8.6 0.4 21.0 3-3 14 0 D 14 10.9 0.6 2.5 4— 2 1 3 6 Anat 0.2 at Total 39 2 9 39.1 ies) 9.0 Chi-square= 18.2; n=8; P=0.05 lates.” Their association prohibits the use of them as three inde- pendent estimates of similarity or divergence. Yet it would be an oversimplification to say that we are dealing with a single character, the ocular region, in taxonomic evaluation of a comparison. For as the association is not perfect, the parts of this unit have changed at different rates. A small ocular region may be one in which the pre- ocular has disappeared while the eye has remained moderate in size (e.g., javanica), one in which the eye has become small while the pre- ocular has not disappeared (e.g., helleniusz), or one in which both eye and preocular have become reduced (e.g., schlegeli cuviert). Thus despite their association, each of these characters has to be consid- ered in evaluating a comparison of two samples. TABLE 17.—Association of Presence of Preocular with Relation of Mental to Chin Shields in Species of Calamaria (Numbers in body of table are numbers of species). Preocular Observed Calculated Present Present Mental and or or chin shield Present absent Absent ‘Total Present absent Absent Separated 12 2 5 19 14.8 0.8 3.4 Separated ortouching 2 0 2 4 Ghai Ot 0.7 Touching 25 0 2 PAE Dileewll sal 4.9 Chi-square=9.5; n=4; P=0.05 INGER AND MARX: THE SNAKE GENUS CALAMARIA 53 TABLE 18.—Association of Number of Scales Surrounding Paraparietal with Number of Supralabials in Species of Calamaria. (Numbers in body of table are numbers of species) Seales surrounding paraparietal Observed Calculated Supra- és se labials 5 S5or6e 6 Total 5 5 or 6 6 4 1 0 10 Ld a 0.4 G0 5 22 2 15 39 17.9 1.6 19.5 Chi-square=9.4; n=2; P=0.01 The association of number of scales surrounding the paraparietal with number of supralabials does not have an obvious explanation. The reduction from five to four supralabials occurs in the preorbital region (Marx and Inger, 1955, p. 169) and has no apparent functional relation to an increase in number of scales at the rear of the head. Correlations of this sort, which does not involve necessary correlates, are valuable in attempts to evaluate the relations of populations as they provide additional evidence of common genotypes. Association of the preocular character with the relation between mental and chin shields has no obvious explanation. Separation of mental and chin shields occurs proportionately oftener with the absence of a preocular than one would expect on the basis of chance alone. Contact between mental and chin shields is probably the primitive condition in this genus (p. 254) and absence of preocular advanced (p. 254). Thus the association shown in Table 17 does not result from progressive modification of both characters. Loss of the TABLE 19.—Summary of Chi-Square Tests of Association Between Pairs of Cephalic Characters in Species of Calamaria. (Value of chi-square in upper right half of table, with degrees of freedom as superscripts. Corresponding values of P in lower left hand. Character states as in Tables 14-18) Supra- Para- labials Preocular Frontal Eyesize Mental parietal Supralabials 2, o.0- 14 3.92 2.1? 9.42 Preoculars S04 ; 18 .28 20.1! oe 5.64 Frontal 1022 0.05 ae 27 ..2° 13.08 13.98 Eye size >0.1 <0.001 <0.001 ai 5.94 2.64 Mental > 0.3 0.05 > 0.1 0.2 i 6.64 Paraparietal 0.01 >0.2 0.09 >0.5 a O 54 FIELDIANA: ZOOLOGY, VOLUME 49 preocular is correlated with general reduction in the ocular region. The other characters associated with the preocular reduction of the ocular region, namely eye size and frontal ratio, are not associated with the chin character (Table 19). Consequently the association of preocular and mental appears not to have a functional basis. As these characters are not necessary correlates, both may be used as independent estimates of similarity or divergence of populations. KEY TO FORMS OF CALAMARIA 1. One sipralabial enter? Or0b ts on «ac 2s Oead sdah ers wees lowi lowi (p. 222) Two supralabials entering orbit.....0.0. «620964 s 64o08es44.ndae nine aes 2 2. Second and: third supralabials entering orbit) 3.252444. eas Se eee 3 Third and fourth supralabials entering orbit........ 20.4.6. .5-s0.@9s400- ee 16 3: dP reocular absent...«cctie. DURE By lees ek aes 1 es a eee 4 Preocular PVESENS 6s: 65.40 dices desu Woe ih iste wd oe ase ee a oe ae 9 4. Mental not touching anterior chin shields... ... .¢..24.. 0.0. sheuuinls oe oe 5 Mental touching anterior chin shields... 23.0 .22.50. 3. 42) nae oe eee 7 5. Supraocular and postocular fused into a single shield...... gracillima (p. 229) Supraocular distinct from postocular .. 4.....<4.0hed«..4 06008 2s ee eee 6 6. Frontal 5 to 6 times width of supraocular; nasal larger than postocular; maxil- lary teeth unmodified (fig. 2 top)..................0.0.. schmidti (p. 74) Frontal 214 to 4 times width of supraocular; nasal equal to, or smaller than postocular; maxillary teeth modified (fig. 2 bottom). . .lowi gimletti (p. 226) 7. Eye much smaller than eye-mouth distance; posterior chin shields meeting in MIGHNE Sc 6 ic ied pow dere oa ae iS RSG OE ES BE ORE OO Eee 8 Eye equal to eye-mouth distance; posterior chin shields not meeting in mid- DDN Oh eae gentss dato. See acd a: ani o Oe Sep imes: eas date tea javanica (p. 209) S.. Posterior ventrals dark browite...% BIyeuINg UOISatY 67 68 FIELDIANA: ZOOLOGY, VOLUME 49 Geographic variation.—The striped pattern (Fig. 16) is character- istic of most of the specimens from Borneo, although in 8 the stripes are faint and in 4 they are absent. All Sumatran specimens are un- striped. The 2 presumed to be from Java (see below) have faint stripes. No other geographic variation (Table 20) was found. Distribution.—Sumatra and Borneo (Fig. 17). The only specimens reported from Java (MHNP 39-188 and 39- 189) were obtained by Boudart. Other specimens Boudart reportedly obtained at Batavia (see p. 170) probably were not collected there. This, in turn, casts doubt on the authenticity of the Javanese local- ity for leuwcogaster. SUMATRA (RMNH 8 unnumbered; ZMB 6298, 31408, 33610). Sumatera Utara: Deli (SNG 19410; ZMA 10077). Sumatera Sela- tan: Ampat Lawang (BM 63.12.11.141-syntype; RMNH 3994-syn- type). SARAWAK (BM 1908.5.28.69; MCG 30456-holotype of beccarz). Kindi District (BM 1911.1.80.25). First Division: Matang (BM 95.2.28.30-holotype of brookiz; SM unnumbered), Kuching (SM un- numbered), Santubong (CNHM 71598-99), Bidi, near Bau (SM unnumbered). Second Division: Lubok Antu (SM unnumbered). Third Division: Long Mujan (BM 1922.11.24.5-holotype of smith). Fourth Division: Niah (CNHM 129004, 131601-04). Fifth Divi- sion: Lawas (CNHM 67279). NORTH BORNEO. Kota Belud District: Kiau, 915 meters (NMS 320), Mount Kina Balu (BM 94.6.30.58; MCZ 43575-76). Labuan District: Labuan (BM 94.6.30.52). INDONESIAN BORNEO: Long Petah (ZMA 10110); Upper Maha- kam (RMNH unnumbered). BORNEO (BM 68.1.27.22; SM 3 unnumbered; ZMB 8012); central Borneo (SMNS 4585). ? Java. Jakarta (MNHP 39-188, 39-189). Specimens examined.—A42. Calamaria ulmeri Sackett. Figure 18. Calamaria ulmeri Sackett, 1940, Not. Nat., no. 41, p. 2—northwest of Blang- kedjeren, Atjeh Province, Sumatra; Marx and Inger, 1955, Fieldiana, Zool., 37, p. 206. Holotype.—Academy of Natural Sciences of Philadelphia 21535. INGER AND MARX: THE SNAKE GENUS CALAMARIA 69 Diagnosis.—Maxillary teeth unmodified; third and fourth supra- labials entering orbit; mental not touching anterior chin shields; paraparietal surrounded by 5 shields and scales; lateral corners of ventrals with dark pigment. Fic. 18. Holotype of Calamaria ulmert. Description.—Rostral wider than high, portion visible from above 28 length of prefrontal suture; prefrontal shorter than frontal, touch- ing first 2 supralabials; frontal hexagonal, 214 times width of supra- ocular, about 34 length of parietal; parietal 114 times length of prefrontal; paraparietal surrounded by 5 shields and scales; nasal smaller than postocular; preocular present; neither ocular as high as eye; eye twice the eye-mouth distance; 5 supralabials, third and fourth entering orbit, fifth the largest, first 3 subequal and slightly larger than fourth; mental triangular, not touching anterior chin shields; 5 infralabials, first 8 touching anterior chin shields; both 70 FIELDIANA: ZOOLOGY, VOLUME 49 pairs of chin shields meeting in midline; 3 gulars in midline between posterior chin shields and first ventral. Tail incomplete but tapering gradually from base; reduction to 4 dorsal scale rows, if it occurs at all, must take place near the end of the tail. Six scale rows are present opposite the twenty-third sub- caudal of this incomplete tail. Ten unmodified maxillary teeth (1 specimen). Ventrals 186 in female; subcaudals 23+ in female. Snout-vent length, 284 mm.; incomplete tail, 28 mm. Body and tail brownish above, each scale with a dark network and a dark central spot; these spots unite to form dark stripes run- ning the length of body and tail; head brown above with small dark spots; upper halves of supralabials dark, lower halves yellowish with dark sutures; underside of head yellowish with small dark spots; lat- eral corners of ventrals and subcaudals dark brown, otherwise entire ventral surface immaculate yellow. Distribution.—Northern Sumatra (Fig. 17). SUMATRA. Sumatera Utara: Atjeh Province, about 40 km. north- west of Blangkedjeren, 2080 meters altitude (ANSP 21535-holotype). Calamaria lautensis de Rooij Calamaria lautensis (part) de Rooij, 1917, Rept. Indo-Austr. Arch., 2, p. 1638, fig. 66—Pulau Si Laut, Cocos Islands; Werner, 1929, Zool. Jahrb., (Syst)., 57, p. 172; de Haas, 1950, Treubia, 20, p. 569; Marx and Inger, 1955, Fieldiana, Zool., 37, p. 203. Lectotype.-—Rijksmuseum van Natuurlijke Historie 4716A, here designated. Taxonomic notes.—As the Amsterdam Museum syntype (ZMA 10111) from Cocos Islands is damaged, we are selecting one of the Rijksmuseum syntypes as the lectotype. See page 139. Diagnosis.—Maxillary teeth unmodified; third and fourth supra- labials entering orbit; preocular present; mental not touching ante- rior chin shields; paraparietal surrounded by 6 shields and scales. Description.—Rostral slightly wider than high, portion visible from above ¥% length of prefrontal suture; prefrontal slightly shorter than frontal, touching first 2 supralabials; frontal hexagonal, 214 to 21% times width of supraocular, about 3/; length of parietal; parietal 1% to 1% times length of prefrontal; paraparietal surrounded by 6 shields and scales; nasal smaller than postocular; preocular present; neither ocular as high as eye; eye less than eye-mouth distance; INGER AND MARX: THE SNAKE GENUS CALAMARIA 71 5 supralabials, third and fourth entering orbit, fifth the largest, first, third, and fourth subequal, second slightly larger; mental triangular, not touching anterior chin shields; 5 infralabials, first 3 touching an- terior chin shields; both pairs of chin shields meeting in midline; 3 gulars in midline between posterior chin shields and first ventral. Body thickness index 0.028 (1 specimen); tail thick, tapering near end to a sharp point; dorsal scales reduce to 4 rows on tail opposite third to fifth subcaudal anterior to terminal scute. Hemipenis forked opposite fifth subcaudal, calyces smooth (1 specimen). Eight to 9 unmodified maxillary teeth (8 specimens). Ventrals: males, 123-1380 (N=2); female, 146. Subcaudals: males, 29-32 (N=2); female, 14+ (incomplete). Total length: males, 215-244 mm.; female, incomplete. Ratio of tail to total length, 0.181-0.184 (N=2). Body and tail dark brown above, each scale with a light network; most scales with a dark central spot; spots frequently unite to form short longitudinal stripes; anteriorly 2 or 3 light spots or vertical bars on side of body; scales of first row whitish, the light area becom- ing progressively reduced toward rear of body; head dark brown above with obscure lighter spots; dark pigment reaching lip on at least the first 2 supralabials; posterior supralabials with varying amounts of dark pigment; underside of head yellowish with dark brown spots; ventrals and subcaudals with dark lateral margins; tail with a dark mid-ventral streak; otherwise ventral surface immacu- late yellow. Distribution.—Cocos Islands (Fig. 17). PULO SI LAUT (RMNH 4716A-lectotype, 4716B-syntype; ZMA 10111-syntype). Calamaria curta Boulenger Calamaria curta Boulenger, 1896, Ann. Mag. Nat. Hist., (6), 18, p. 62—South Celebes; 1897, Proc. Zool. Soc. London, 1897, p. 224, pl. 14, fig. 2; de Rooij, 1917, Rept. Indo-Austr. Arch., 2, p. 160; Smith, 1927, Proc. Zool. Soc. London, 1927, p. 224; Werner, 1929, Zool. Jahrb., (Syst)., 57, p. 171; de Haas, 1950, Treubia, 20, p. 566; Marx and Inger, 1955, Fieldiana, Zool., 37, p. 201. Holotype.—British Museum (Natural History) 96.4.29.35. Diagnosis.—Maxillary teeth unmodified; third and fourth supra- labials entering orbit; mental touching anterior chin shields; second pair of chin shields meeting in midline. FIELDIANA: ZOOLOGY, VOLUME 49 —~l bo Description.—Rostral higher than wide, portion visible from above 14 to 24 length of prefrontal suture; prefrontal shorter than frontal, touching first 2 supralabials; frontal hexagonal, 1°/s to 21% times width of supraocular, about ‘/; length of parietal; parietal 114 to 124 times length of prefrontal; paraparietal surrounded by 6 shields and scales; nasal smaller than postocular; preocular present; neither ocular as high as eye; eye equal to or greater than eye-mouth dis- tance; 5 supralabials, third and fourth entering orbit, fifth the largest, first 4 subequal or second slightly larger than others; mental triangu- lar, touching anterior chin shields; 5 infralabials, first 3 touching an- terior chin shields; both pairs of chin shields meeting in midline; 3 gulars in midline between posterior chin shields and first ventral. Body thickness index 0.015-0.029 (2 specimens); tail tapering gradually from base to an extremely long terminal scute; dorsal scales reduce to 4 rows on tail opposite third to sixth subcaudal an- terior to terminal scute. Cloaca of female cardioid (1 specimen). Seven unmodified maxillary teeth (8 specimens). Ventrals: male, 156; females, 153-170 (mean 162; N=4). Sub- caudals: male, 28; females, 14-17 (mean 15.3; N=4). Total length: male, 1538 mm.; females, 195-807 mm. Ratio of tail to total length: male, 0.105; females, 0.067—-0.072 (mean 0.073; N=4). Color dark brown above, most of dorsal scales with a fine dark net- work; dorsal scales darker in anterior corners; head dark brown above and on sides with obscure darker spots; supralabials with dark sutures and yellowish ventral borders; head yellow below with dark brown spots; ventrals dark, usually brown in anterior 24 and yellowish on posterior edges; small yellowish spots usually present in dark portion of each ventral; subcaudals yellow and brown, usually yellow con- fined to posterior halves. Ecological notes.—Smith (1927) found four specimens in and under rotting logs. He stated that their food consisted of earthworms. Distribution.—Southern Celebes (Fig. 17). CELEBES: peak of Mount Bonthain, 1500-1800 meters (BM 1926. 8.20.167-168; MCZ 25301-02) ; south Celebes, 610 meters (BM 96.4. 29.35—holotype). Specimens examined—d. INGER AND MARX: THE SNAKE GENUS CALAMARIA 73 Calamaria acutirostris Boulenger. Figure 19. Calamaria acutirostris Boulenger, 1896, Ann. Mag. Nat. Hist., (6), 17, p. 394— Loka, Mount Bonthain, Celebes; 1897, Proc. Zool. Soc. London, 1897, p. 223, pl. 13, fig. 2; Boettger, 1898, Kat. Rept. Samm. Senck. Naturf. Ges., pt. 2, p. 86; de Rooij, 1917, Rept. Indo-Austr. Arch., 2, p. 161; Smith, 1927, Proc. Zool. Soc. London, 1927, p. 224; Werner, 1929, Zool. Jahrb., (Syst.), 57, p. 171; de Haas, 1950, Treubia, 20, p. 563; Marx and Inger, 1955, Fieldiana, Zool., 37, p. 199. Lectotype.—Naturhistorisches Museum Basel 1686, here desig- nated. Fic. 19. Paratype of Calamaria acutirostris (NMB 1688). Diagnosis.—Maxillary teeth unmodified; third and fourth supra- labials entering orbit; mental touching anterior chin shields; second pair of chin shields not meeting in midline. Description.—Rostral higher than wide or as wide as high, scarcely visible from above; prefrontal subequal to frontal, touching first 3 supralabials; frontal hexagonal, 224 to 31% times width of supra- ocular, about 24 length of parietal; parietal 11% times length of pre- frontal; paraparietal surrounded by 6 shields and scales; nasal slightly smaller than postocular and larger than preocular; preocular present, neither ocular as high as eye; eye less than or equal to eye-mouth distance; 5 supralabials, third and fourth entering orbit, fifth the largest, first smallest, next 3 subequal; mental triangular, touching anterior chin shields; 5 infralabials, first 3 touching anterior chin shields; posterior pair of chin shields not meeting in midline; 4 gulars in midline between anterior chin shields and first ventral. Body thickness index 0.013-0.070 (10 specimens) ; tail thick, either tapering gradually or abruptly near end to a point; dorsal scales re- 74 FIELDIANA: ZOOLOGY, VOLUME 49 duce to 4 rows on tail opposite first to sixth subcaudal anterior to terminal scute. Hemipenis forked opposite fifth to seventh subcaudal, retractor muscle beginning opposite ninth to twelfth subcaudal, sulcus bifur- cate, calyces smooth (5 specimens). Cloaca of female bilobed (2) or bulbous (1). Ten to 12 unmodified maxillary teeth (9 specimens). Ventrals: males, 148-162 (mean 154.6; N=18); females, 163-174 (mean 167.6; N=16). Subcaudals: males, 20-24 (mean 21.5; N=18); females, 18-17 (mean 15.2; N=16). Total length: males, 138-415 mm.; females, 150-447 mm. Ratio of tail to total length: males, 0.086-0.111 (mean 0.094; N=17); females, 0.057-0.069 (mean 0.068; N=16). Color reddish brown above without markings; dorsal scales with- out network; head dark brown above; dark pigment ending on a more or less oblique line on supralabials, usually reaching lower edge of an- terior supralabials; underside of head yellow with dark spots on ante- rior infralabials and chin shields; ventrals yellow, usually with a brownish speckling across their anterior edges; subcaudals yellow, edged with brown. Ecological notes—Smith (1927) found ten specimens under stones on Mount Bonthain. Distribution.—Southwestern Celebes (Fig. 17). CELEBES (NMS 8 [2]; ZMB 15636): Mount Bonthain, Djikoro (CNHM 88161; MCZ 25292-96, 25298; USNM 120810), Loka (BM 96.12.9.61-63—paratypes; NHMW 16714; NMB 1681—lectotype, 1684, 1688-89, 5228-29—all paratypes); Bua Praeng (NHMW 16713: 1-11; SNG 19440). ? Java (ZMB 138947). This specimen was collected by the same collector (Fruhstorfer) and on the same date as SNG 19440. The snake probably came from Celebes. Specimens examined.—s4. Calamaria schmidti Marx and Inger. Figure 1. Calamaria schmidti Marx and Inger, 1955, Fieldiana, Zool., 37, p. 197, fig. 27 — Bundu Tuhan, Mount Kina Balu, North Borneo. Holotype.—United States National Museum 130240. Diagnosis.—Maxillary teeth unmodified; four supralabials, sec- ond and third supralabials entering orbit. INGER AND MARX: THE SNAKE GENUS CALAMARIA 75 Description.—Rostral broader than high, portion visible from above about 14 length of prefrontal suture; prefrontal squarish, sub- equal to length of frontal, touching first 2 supralabials; frontal pentag- onal, 5 to 6 times width of supraocular, about 24 length of parietal; parietal 124 times length of prefrontal; paraparietal surrounded by 6 shields and scales; nasal larger than eye or postocular; no pre- ocular; postocular not as deep as eye; eye small, diameter 14 eye- mouth distance; 4 supralabials, second and third entering eye, fourth longest, third 34 or more length of second, first and third equal; mental triangular, not touching anterior chin shields; 5 infralabials, first 83 touching anterior chin shields; both pairs of chin shields meet- ing in midline; 3 gulars on midline between second pair of chin shields and first ventral. Tail ending in a blunt point; dorsal scales reduce to 4 rows on tail opposite the seventh to eighth subcaudal anterior to terminal scute. Six or seven unmodified maxillary teeth (2 specimens). Ventrals: males unknown; female, 144 (1 specimen). Subcaudals: females, 14 (2 specimens). Total length of complete female 253 mm.; ratio of tail length to total length: female, 0.075 (1 specimen). The paratype is badly damaged and cannot be measured. Color purplish gray above, uniform, scales without network; head without markings; supralabials same color as back; anterior infra- labials and first pair of chin shields purplish gray, remainder of under side of head yellowish; ventral surface without markings; anterior ventrals yellowish, belly becoming increasingly more purple posteri- orly, but lighter than dorsal color; under side of tail darker than belly but slightly lighter than dorsal surface. Food.—The holotype contained an earthworm. Distribution.—Mount Kina Balu, North Borneo. NorTH BoRNEO: Mount Kina Balu, Bundu Tuhan, 1370 meters (USNM 130240—holotype), Mount Kina Balu (SU 8568—paratype) Calamaria lumbricoidea H. Boie. Figure 20. Calamaria lumbricoidea H. Boie in F. Boie, 1827, Isis, 20, p. 540—Java; Schlegel, 1837, Phys. Serp., pt. 2, p. 27, pl. 1, figs. 14-16; Duméril and Bibron, 1854, Erp. Gén., 7, p. 89; Giinther, 1858, Cat. Colubrine Snakes Brit. Mus., p. 5; Jan, 1862, Arch. Zool. Anat. Phys., 2, p. 8; 1865, Icon. Ophid., 10, pl. 2, fig. 2; Lidth de Jeude, 1890, Notes Leyden Mus., 12, p. 254; Boulenger, 1894, Cat. Snakes Brit. Mus., 2, p. 333; Boettger, 1886, Ber. Senck. Naturf. Ges., 2, p. 105; 1894, 72 Semon, Zool. Forsch. Austr., p. 125; 1898, Kat. Rept. Samm. Senck. Naturf. Ges., pt. 2, p. 83; de Rooij, 1917, Rept. Indo-Austr. Arch., 2, p. 153; Werner, 1929, Zool. Jahrb., FIELDIANA: ZOOLOGY, VOLUME 49 (Syst.), 57, p. 170; Kopstein, 1930, Treubia, 12, p. 274; Marx and Inger, 1955, Fieldiana, Zool., 37, p. 183; Inger and Marx, 1962, Syst. Zool., 11, Di oc. Calamaria lumbricoidea lumbricoidea, de Haas, 1950, Treubia, 20, p. 570. Calamaria vermiformis Duméril and Bibron, 1854, Erp. Gén., 7, p. 85—“‘Java”’ [in error]; Jan, 1862, loc. cit., p. 8; 1865, loc cit., pl. 2, fig. 3; Lidth de Jeude, 1890, in Weber, Zool. Ergebn., p. 182; 1922, Zool. Meded., 6, p. 247; Boulenger, 1894, op. cit., p. 333; 1896, Cat. Snakes Brit. Mus., 3, p. 646; 1912, Fauna Malay Penin., p. 155; Bartlett, 1895, Sarawak Note Book, no. 1, p. 83; Werner, 1896, Verh. Zool. Bot. Ges. Wien, 46, p. 17; 1929, loc. cit., p. 170; Boettger, 1898, loc. cit., p. 838; Brown, 1902, Proc. Acad. Nat. Sci. Philadelphia, 1902, p. 180; Cohn, 1905, Zool. Anz., 29, p. 545; Baumann, 1913, Zool. Jahrb., (Syst.), 34, p. 270; Smith, 1916, Jour. Nat. Hist. Soc. Siam, 2, p. 162; 1925, Sarawak Mus. Jour., 3, p. 4; 1930, Bull. Raffles Mus., no. 3, p. 58; 1931, zbid., no. 5, p. 27; de Rooij, 1917, op. cit., p. 153; Holtzinger-Tenever, 1920, Arch. Naturg., 85, p. 86; Mertens, 1924, Zool. Anz., 60, p. 158; Lonnberg and Rendahl, 1925, Ark. Zool., 17A, p. 2; de Haas, 1950, loc. cit., p. 575; Tweedie, 1953, Snakes Malaya, p. 49; Marx and Inger, loc. cit., p. 185; Taylor and Elbel, 1958, Univ. Kansas Sci. Bull., 38, p. 1044. Calamaria vermiformis vermiformis, Marx and Inger, 1955, loc. cit., p. 187. Calamaria temmincki Duméril and Bibron, 1854, Erp. Gén., 7, p. 87—‘‘Su- matra or Borneo’; Giinther, 1858, loc. cit., p. 5. Calamaria. grayi Giinther, 1858, Cat. Colubrine Snakes Brit. Mus., p. 5— Philippine Islands; Boettger, 1886, loc. cit., p. 105; Boulenger, 1894, loc. cit., p. 08; Taylor, 1922, Sndkes' Philip. Ids:, p. 184% 1922," Philipwviour: Scl.;,21,/p; 204; Werner, 1929, loc.cctjap 170. Calamaria dimidiata Bleeker, 1860, Nat. Tijds. Ned. Indie, 21, p. 295—Java. Calamaria melanorhynchus Bleeker, 1bid.—Ampat Lawang, Sumatra. Calamaria alkeni Bleeker, 7zbid.—Ampat Lawang, Sumtra. Calamaria flaviceps Giinther, 1865, Ann. Mag. Nat. Hist., (8), 15, p. 90— Sarawak and Borneo. Calamaria philippinica Steindachner, 1867, Verh. Zool. Bot. Ges. Wien, 17, p. 514, pl. 8, figs. 4-6—Philippine Islands; Boettger, 1886, loc. cit., p. 105. Calamaria stahlknechtii Stoliczka, 1873, Jour. As. Soc. Bengal, 42, p. 119, pl. 11, fig. 2—near Deli, Sumatra; Boulenger, 1885, Ann. Mag. Nat. Hist., (5), 16, p. 388; 1894, loc. cit., p. 885; Modigliani, 1889, Ann. Mus. Stor. Nat. Genova, (2), 7, p. 119; de Rooij, 1917, loc. cit.,.p. 154; Werner, 1929; loc. cit., p. 170; de Haas, 1950, loc. cit., p. 574; Marx and Inger, 1955, loc. cit., p. 206. Calamaria variabilis Lidth de Jeude, 1890, in Weber, Zool. Ergebn., p. 1838, pl. 16, fig. 8—Buitenzorg, Java. Calamaria vermiformis var. sumatranus Lidth de Jeude, 1890, Notes Leyden Mus., 12, p. 18—Deli, Sumatra. Calamaria bungaroides Werner, 1901, Zool. Anz., 24, p. 300—type locality unknown; 1929, loc. cit., p. 178; de Rooij, 1917, loc. cit., p. 173; de Haas, 1950, loc. cit., p. 565; Marx and Inger, 1955, loc. cit., p. 200. INGER AND MARX: THE SNAKE GENUS CALAMARIA 77 Calamaria bruegeli Mertens, 1924, Zool. Anz., 60, p. 158—central Borneo; de Haas, 1950, loc. cit., p. 565. Calamaria géringt Vogt, 1925, Zool. Anz., 62, p. 64—Java; de Haas, 1950, loc. cit., p. 567; Marx and Inger, 1955, loc. cit., p. 202. Calamaria goeringi, Werner, 1929, loc. cit., p. 171; Brongersma, 1930, Treubia, 12, p. 302. Changulia lumbricoidea, de Haas, 1941, Treubia, 18, p. 369. Calamaria gracilis (non Boulenger), Angel, 1941, Bull. Mus. Nat. Hist. Nat., (2), 13,, p. 411. Calamaria vermiformis grayi, Marx and Inger, 1955, loc. cit., p. 188; Leviton, 1968, Proc. Calif. Acad. Sci., 31, pp. 379, 389, 393, 404. Lectotype.-—Rijksmuseum van Natuurlijke Historie 105438 (ex 42). Taxonomic notes.—Dumeéril and Bibron did not compare verm- formis with the very similar species lwmbricoidea. Boulenger (1894) differentiated these two forms solely on the basis of eye size, which he said was smaller than the eye-mouth distance in lwmbricoidea and equal to that distance in vermiformis. De Rooij (1917) merely copied Boulenger’s descriptions. In his descriptions Boulenger stated that the ventral coloration of lumbricoidea was whitish whereas that of vermiformis was highly variable, ranging from uniform yellow through increasing amount of black transverse bands to uniform black. Our examination of 375 specimens from all parts of the combined ranges of lumbricoidea and vermiformis convinces us of their conspeci- ficity. We confirm Boulenger’s observation of the range of ventral coloration. We cannot, however, confirm Boulenger’s observations on eye size; Javanese snakes agreeing in all details with the type series of lumbricoidea have eye diameters ranging from slightly smaller to larger than the eye-mouth distance. Snakes from Borneo, indistin- guishable from the type series of vermiformis, show the same varia- tion in eye size. Both sets of-specimens have the same distinctive round snout and a relatively thick body. Especially significant is the fact that only these two nominal forms out of the entire genus show ontogenetic variation in coloration of the head (see below). The type locality of both vermiformis and lumbricoidea is Java. All of Boie’s material was from Java and subsequent collecting has turned up all of his species of Calamaria on Java. Consequently the type locality of the older name, lumbricoidea, is correctly given. Numerous specimens agreeing with the types of lwmbricoidea in those few characters by which these two type series differ (see Geographic variation) have been collected over a long period of time on Java by various collectors. 78 FIELDIANA: ZOOLOGY, VOLUME 49 Seven specimens with the characteristics of vermiformis (1.e., banded ventrally) have been reported from Java. Of these only one (MHNP 89-187) is reported from a specific locality, Batavia (=Ja- karta). As none of the recent large Javanese collections of Calamaria include vermiformis, which is one of the common forms in other parts of the generic range, we doubt that this color form occurs on Java. The detailed original description and figures of stahlknechti agree in counts and coloration with typical Sumatran specimens of luwm- bricoidea. Snakes from Nias identified by Boulenger (1894) and de Rooij (1917) as stahlknechtt have been examined and are also lumbricoidea. The holotype of géringi does not differ from Javanese lumbricoidea. Our examination of the holotype of bruegelz confirms our previous opinion (Marx and Inger, 1955) that it is synonymous with vermi- formis and therefore with lumbricoidea. The original description of Calamaria bungaroides fits juvenile Bornean and Sumatran lumbricoidea exactly. We have examined the types of temmincki, dimidiata, melano- rhynchus, alkent, flaviceps, vermiformis var. sumatranus, and variabilis and agree with Boulenger (1894) that they are synonymous with lumbricoidea and vermiformis of Boulenger. Diagnosis.—Maxillary teeth modified; third and fourth supra- labials entering orbit; preocular present; mental touching anterior chin shields; paraparietal surrounded by 4 or 5 scales and shields; belly yellow with black crossbars wider than width of one ventral. Populations of lwmbricoidea from Java and parts of Sumatra have immaculate light bellies. Javanese lumbricoidea are distinguished from all other species known from that island by the other characters mentioned in the diagnosis. Those characters also distinguish the light-bellied Sumatran lumbricoidea from all other Sumatran Cala- maria except albiventer. The last has two lateral light stripes which differentiate it from lumbricordea. Description.—Rostral wider than high, or higher than wide, por- tion visible from above about half length of prefrontal suture; pre- frontal equal to or slightly shorter than frontal, touching first 2 supralabials; frontal hexagonal, 14% to 214% times width of supra- ocular, about 34 to °/s length of parietal; parietal 114 times length of prefrontal; paraparietal surrounded by 4 or 5 shields and scales; nasal smaller than postocular; preocular present, higher than post- INGER AND MARX: THE SNAKE GENUS CALAMARIA 79 ocular; postocular usually not as high as eye; eye smaller than, equal to, or greater than eye-mouth distance; 5 supralabials, third and fourth entering orbit, fifth the largest, first 4 subequal or third the smallest; mental triangular, touching anterior chin shields; 5 infra- labials, first 8 touching anterior chin shields; both pairs of chin shields usually meeting in midline; usually 3 gulars in midline be- tween posterior chin shields and first ventral (see Geographic varia- tion below). Fic. 20. Calamaria lumbricoidea (NHMW 16698:3). Body thickness index 0.012-0.074; tail thick, tapering abruptly near tip to sharp point; dorsal scales reduce to 4 rows on tail opposite first to eleventh subcaudal anterior to terminal scute. Hemipenis variable (see Geographic variation). Cloaca of female bilobed, bulbous, or cardioid, not varying geographically (Inger and Marx, 1962). Nine (1 specimen), 10 (19) or 11 (2) modified maxillary teeth. Ventrals: males, 144-196 (mean 172.0; N=182); females, 137-229 (mean 190.8; N=116). Subcaudals: males, 17-27 (mean 21.5; N= 134); females, 18-21 (mean 16.8; N=118). Total length: males, 149-498 mm.; females, 120-642 mm. Ratio of tail to total length: males, 0.068-0.114 (mean 0.080; N=126); females, 0.089-0.0838 (mean 0.056; N=1138). Color dark brown or black above, with or without narrow, light stripes; scales without network; first one or two scale rows yellow; ‘paddiys svore yovlq Jo UMOIG “Dap20I1UQUN) DILDULDID UL UOIZBIO[OO pRdY UI UOTVIIRA dIJeUeS0JUQ “TZ “DI oss 00S OS OOr OSE 0 CZ 002 OSsl OO (ww) yyBua; |OyoY f OO€ 0 ; rine al fff [fr Xxxla g W W rf S| @a Sass] Soa%scSs s} saSS8wSagwx SS S|INX S x S cs WwW [WS Saws N|N DEE geet es rf he Ge r f f f rf ve ry rk Xr) eee reece f W Ww |trrw f xX W WF q TY «OK SDIN=N puppy =) DIJDWNS=S ayAa =] pAD|DW=W ODUDpPUIW=X oauslog=q DAD/( =[ 80 =Borneo Leyte Nias B L N Mindanao Sumatra Thailand 81 SM S'S BSB|ISB B x M SSoMgc B 150 200 250 300 350 400 450 500 550 100 Total length(mm) Yellow areas not stippled. . Ontogenetic and geographic variation in body pattern of Calamaria lumbricoidea. 22 FIG. 82 FIELDIANA: ZOOLOGY, VOLUME 49 ventral surface yellow, with or without black crossbands (see Onto- genetic variation and Geographic variation). Ontogenetic variation.—The heads of juveniles (up to about 200 mm., total length) are yellow in preservative whereas those of adults are dark brown or black. Darkening of the head begins on the snout and gradually spreads to cover the entire head (Fig. 21). Juveniles of most populations have yellow bands across the dor- sum. These bands are one-half to two scales wide (Fig. 22). Geographic variation.—Figure 21 suggests that Bornean juveniles tend to have lighter heads than do those from Java. Adults from Java often retain a narrow light band or pair of light spots on the rear of the parietals. Such spots or bands rarely occur in adults in other parts of the range (Fig. 21). The yellow dorsal crossbands are usually less than one scale wide and usually interrupted in juveniles from western Sumatra and west- ern Malaya. In juveniles from Borneo and the Philippine Islands these bands are usually continuous and one to two scales wide. Yel- low crossbands are absent in juveniles from Java, eastern Sumatra, and central Malaya (Fig. 22). Tweedie (1953) states that longitudinally striped lwmbricoidea are confined to high altitudes and uniform brown snakes to low altitudes. The striped Malayan snakes examined by us come from 1,800 to 1,400 meters above sea level and non-striped ones from 500 to 1,350 meters. Limiting observations to Malaya might lead to the conclu- sion that the variation is altitudinal. However, all lumbricozdea from Java (350 to 1,400 meters) and Borneo (50 to 1,480 meters) are non- striped though they span the altitudinal range of the Malayan speci- mens. All lumbricoidea from the Philippine Islands (sea level to 850 meters) are longitudinally striped. Clearly this aspect of dorsal pattern varies geographically rather than altitudinally (Table 21). The non-striped Malayan snakes occur along the western edge of the peninsula and the striped ones farther inland. In Sumatra non-striped snakes occur along the east- ern coast and striped ones on the west. The ventrum in lwmbricoidea varies from uniform yellow to almost solid black. Black on the belly usually takes the form of black cross- bands that occupy 2 to 4 adjacent ventrals separated by varying numbers of yellow scales. In three specimens the belly is entirely black. Some specimens without black crossbands have black speck- ling that varies from a few tiny spots on scattered ventrals to a more or less continuous dark stripe midventrally. Geographic variation INGER AND MARX: THE SNAKE GENUS CALAMARIA 83 TABLE 21.—Frequency Distribution of Dorsal Patterns and Number of Gulars in Samples of Calamaria lumbricoidea. Dorsal pattern No. of gulars Localities Striped Uniform 3 4 TRAUONC < «, 2:446-428 da-wee ewe ore we 1 I Malaya: Panbane< sed seen ds8 aes 8 7 1 Pele, ced OY VR Ss 4 4 SClANGOr ecu cease sewn 4 3 JONOLEs «5.5 bok Same owecks 2 2 Sumatra: Padang area........... 20 3S 25 WG es: Oe dono ee ose de 8 1 cs Palembang..... nes 1 1 S! Rambé.uscsaccenadus ii 1 Bindjay <4 62cc ccd deces 2 2 LONGED Gs nao as ea eew eed 2 2 Ophir District.......... 1 j 1 ii INAS doe Se ha ee ih pe eRe ee 6 1 5 Borneo: Kina Balt. wi: sis eedees Le 13 LaWas. .dvecedeiatonens 1 I Pa Brayou es ic.0< 64 awregrs 1 1 Baram District......... 4 2 2 TN rs od, ants cca haceanee A 2 1 1 M AVON os «vd aeensane ds 1 1 Batu Song............. 1 1 Tandon? ic ccddssnencex Z 2 Mahakam............. 2 Z Sandakan..... cea s 1 a PO ee olny odd sow Seta Ged Gow, eee 125 125 NSUUNG.4 92 Sedo awed iene thes poet 1 1 Wid aNAGso &hxedeinee ons oe 8a nes 14 13 1 TOOL 4 & &-esnennint & a ine ae eee oa 1 1 BAST ov x. sites Se Si dn ae dae 1 1 LF: eee a or 1 1 of ventral coloration is shown in Table 22. The belly pattern shows not only inter-island variation, but also variation within Sumatra as a comparison of the Deli and Padang specimens indicates. Populations from the Philippine Islands, Borneo, Java, western Sumatra, and central Malaya have 3 mid-ventral gulars whereas 4 is the common number elsewhere (Table 21). Within islands the number of ventrals varies to a remarkable ex- tent (Tables 28 and 24). Yet at any given locality, lumbricoidea follows the common pattern of this genus by having relatively little individual variation (see p. 24). For example, in the Kina Balu sam- ple the range of ventrals is 170-180 in seven females and 154-165 in seven males or only a fraction of the range for the entire Bornean sample. Despite much overlap, the populations from Java and the Philippine Islands have higher ventral counts than do those from other areas. rm me poeyeredas ‘s[eIjUdA ‘s[eIqQUdA yey Moy uvypiseg Le 22s Dy. Ue legs > ae ee oyog bb Ses 2c Ae ee eee -ovurpulyyy ORT or a VAL baie BE aes 3 Oe apes oe oaul0g Q rte ees pes cou ee aia StIN he te hee aha ae [210 J, ss © eve > is}. e slo, ea, Satie suequis[e dg ee th malate- fe) cat fe tena’ io) cet ce teuhiel comme jeysue'y eS ee neyal . e ON Ma MC ONO. Cut Oe De tt OR suvped “3 e sche? eipetisaieireneetce B1ep Iayyo ou -B1ZBUUNS sivel is: i era(eialh vind lsepye} enish (lyst Celie Mte Reis Ste aeMce late BARR L pkanalln ska er puryivyy, MOTJOA AylyRoo'T “DAPLOIQUN] DILDUWDYIDD Jo sjeryUdA UO Yortg Jo JUNowyY 0} yoodsey YYW uoNqiysiq Aouenbely— ZZ ATAV L, 84 INGER AND MARX: THE SNAKE GENUS CALAMARIA 85 Similar geographic variation—slight differences between areas with extensive overlap—is shown by subcaudal counts, the position of the reduction to four dorsal scale rows, and the ratio of tail length to total length (Tables 25 and 26). The hemipenis of lumbricoidea shows variation not seen in other species of Calamaria (Inger and Marx, 1962). Males from the main- land, Sumatra, Nias, and Java are alike in having the hemipenis forked beginning between the second and sixth subcaudals (as is characteristic of the genus) and in having both rami covered with papillate calyces; the sulcus forks at the base and a branch runs to the apex of each fork (as in Fig. 4F of Dowling and Savage, 1960). The same type of hemipenis was found in males from North Borneo and Sarawak. However, males from widely separated local- ities in Borneo have undivided hemipenes that have smooth calyces; the sulcus is bifurcate in these simple hemipenes. Aside from the range of variation among these Bornean specimens, the striking aspect is the lack of variation within a local population, such as that of Mount Kina Balu. Characters that vary geographically in lumbricoidea are not asso- ciated in any fixed way with the result that geographic variation in one character does not coincide with that of another (Table 27). If subspecies were named they would have to be based on a single character and the choice of that character would be arbitrary. Distribution.—From southern Thailand to Java, Borneo, and Leyte (Fig. 72). THAILAND: Bangkok (UMMZ 65331). MALAYA: Perak: Gunong Kledong (NMS unnumbered), Larut Hills (BM 1900.6.14.22), 1870 meters (NMS unnumbered), Taiping (ZMB 26457). Pahang: Fraser’s Hill (NMS 30229, 4 unnumbered), Cameron Highlands, Kuala Terla (CNHM 723874), Cameron High- lands, Tanah Rata, 1400 meters (NMS 2 unnumbered). Selangor: Kepong Forest Reserve (NMS 18880, 30462), 640 meters (NMS 9474), Kepong, Bukit Lagong (NMS 8827). Johore: Gunong Pulai (NMS 2 unnumbered). Nias (BM 84.12.81.7, 84.12.31.8/2]; RMNH 4865). Gunong Sitoli (MCG 30881); west Nias (ZMA 10284). SUMATRA (AMNH 2879; BM 62.12.4.123—syntype of alkenz; NHMW 16699[2]; NMB 10769; RMNH 33/2], 48, 78, 5806/8], 1 un- numbered; USNM 56489). East coast of Sumatra (BM 983.6.5.8); N 961 E61. G O6T oF SOL “0g. oe T a G ! { G T & T G & T g 6 T I I I 480 8 T8t 84F Skt GLE 69T 99T S1I9}U90 SSBIT) G € i CB iC ve G T IT G G Z i 8 € : 691 O9f L&I TST [ST STE S01 Re aah Seeeh e 3 spur|s] sulddliyg Pa eS ett ILO Cee suel(peyity, sii) (ey <0) Te) yehleiatey (oie = tableMe neg BUuly 3 fefte. Ratio in thousandths of total length. SARAWAK (BM 45.10.2.45—syntype of flaviceps): First Division: Bidi (SM Cd.5.26.1¢; Cd.5.26.15b), Kuching (SM Cd.5.26), Matang (BM 73.3.4.23). Second Division: Saribas (SU 8575). Fourth Divi- sion: Niah, Rumah Sigi (CNHM 129005), Niah, Kuala Sakoloh (CNHM 129006), Niah (CNHM 181620-27), Mount Batu Song, 300 meters (BM 92.10.7.8), Baram district (BM 97.3.4.5, 1902.2.11. 25.16-17; NHMW 16992:1). Fifth Division: Lawas (CNHM 67278), Lawas, Sungei Usop (SM Cd.5.26.1d), Pa Brayong (CNHM 71646). NORTH BORNEO: Ranau District: Mount Kina Balu (MHNP 89- 193), Bundu Tuhan, 1370 meters (USNM 180252). Kota Belud District: Mount Kina Balu, Tenompok, 1480 meters (MCZ 43568), Kiau, 915 meters (BM 1929.12.22.110-111; MCZ 48564-65; NMS 16817), Kenokok River, 1000 meters (MCZ 43566-69; UMMZ 82829[2]). Sandakan District: Sandakan (CNHM 15008). YJOOWS pay1o] yQoows osfduiis -ayeided -pex10j ayeyided poaxysoj ayelided pays0j ayerided poxyJs0j ayelpided poy10j ayeyjided poxy10j (GL)8I-L (61)IZ-9T (LI)OZ-FI (#2) LZ-2Z (861T)6ZS-88T (TLI)9LI-G9T og opm (ED8I-L (6T)GZ-1T (LT)IZ-#T (€2)92-6L (€LT)00Z-LET (O9T)I8SI-FFI F-E apIMm GT LZ LST i (ZL)9T-8 (ST)8I-ZT (ST)8I-ST (0Z)TZ-LI (€0Z)0ZZ-LET (@8T)96I-ZLT & Juasqe Zi (LD LI-e1 GT (1Z)Z2-02 Pol (OST)ESI-LII > (FI)SI-IT (9T)6I-Z1 (LI)6I-9T (22) PZ-6T (E61) LIZ-9LT (OLT)T6I-8GI & ModeU (ZE)9T-L (LI)IZ-&T (9T)LTI-F1 (12)€2-02 (88T)LOZ-LLT (OLTI)E6I-F9T fF Quasqe (FI)8I-Z1 (61)0Z-9T (8T)TZ-ST (72)LZ-02 (O8T)FP8I-ELT (69T)TZT-G9T yp MosreU 6 aye[ided poexsoy (ZT)ST-OL (LT)0Z-ST (8T)8I-LT (22) 72-12 (ZLI)PLI-69T (8¢1)Z91-FSI juasqe poy1oj ST &% 9LT v saodjeg adeys & 2 5 52) 5 2) siemny) ;s8ul ; : = a -purq Stusdituwia YY eUOlJONpYyY eST@pneoqns eS[V1IQU9 A alu -oane “DAPOIUQUN) DILDUWY)IDY UL UOLVBLIVA d1ydeis0ar Jo AreWUING—')7 AIAV ‘Sosoyqjueied Ul SUBATT ¢ “OpIM o[BOS 9UO UBY} SSdT=MOLIVU SaPIM Sa[vIS OM} 0} UO Spuegq=aPI A ; “6G PBL Jo suunjoo yAYSIe 07 YIFY=Aavoy $ZZ a[qey, Jo suuN[oOd YANO} 0} puooss=yySrq ; AAvay AABay yuasqe qyuasqe yuasqe yuasqe “V4 8] AAvay AABOY AAvOY AABBY (Bur “pueq [e1qUa A pedis waojlun waojlun waojlun wi4ojlun pedis wiojiun wsojiun pedis waojlun uorly -B10]09 [Vsiog Spuv[sT oulddriyd oaulog vunjeN BALL SBIN UId}SOM usaysea eiyeunsg U1a}SOM [ei}Ua— “BARR IN PaEreaT 90 INGER AND MARX: THE SNAKE GENUS CALAMARIA 91 INDONESIAN BORNEO: Boven Mahakam (RMNH 2 unnumbered), Tandjong (BM 96.2.17.11-12), southeast Borneo (SMNS 3498, 3498a, 3497; ZMH 2421). BORNEO (BM 56.9.27.9—syntype of flaviceps, 72.2.19.55, 1 un- numbered; USNM 49802); central Borneo (SNG 19889—paratype of bruegeli); north Borneo (NHMW 16992:2-3; ZMB 18110). BASILAN: near Isabela (MCZ 43469). MINDANAO: Zamboanga Province: Zamboanga City (CAS 15291), Zamboanga (City?) (CAS 620380). Cotabato Province: Mount Malin- dang, Gumay (CNHM 96529), Cotabato (NMB 8518), Saub (MCZ 25778-80; USNM 120811), Cotabato coast (ZMA 10236). Misamis Province: Lake Lanao (CAS 15286-89), Lanao, Lumbatan (CNHM 15010). Davao Province: Mount Apo, Todaya, 855 meters (CNHM 53013). NEGROS: Negros Oriental: Dumaguete (SU 15952). BOHOL (NHMW 16815). LEYTE: Cabalian (MCZ 25781). SAMAR: Osmera (USNM 122215). PHILIPPINE ISLANDS (BM 1946.1.2.17, 1946.1.2.21—-syntypes of grayi; ZMB 5446). JAVA (BM 44.2.32.65, 66.8.14.344, 70.6.7.46, 73.5.8.26; CNHM 83163; MHNP 165; NHMW 16698(6], 16816, 2 unnumbered; NMS 258; RMNH 10419[4], 10548—lectotype of lumbricoidea, 42[2]—para- types of lwmbricoidea; SNG 193888, 19393; ZMA 10233[2]; ZMB 1552, 4069, 4117, 6936, 8545—holotype of géringi, 8758, 10261; ZMH 4651- 4652, 5880). M. Pangherango (MCG 30457); ?Tjikorai (NMS 5). Djawa Barat: Benkang near Serang (AMNH 71513), Djakarta (NMS 544), Bogor (RMNH 4827—holotype of varzabilis; SNG 193886, Med- amendoeng Peak, between Bogor and Sindanglaja (MHNP 95-52[2}), Sindanglaja (MCZ 7523), Bandung, 800 meters (RMNH 7463/3], 1 unnumbered), Mount Gede, Tjibodas, 1450 meters (RMNH 8604), Rarahan, near Tjibodas (MCG 6780, 30878), Mount Garut (NHMW 16697:1), Mount Garut, plantation Dajeuhmanggoeng (RMNH un- numbered), Tjikadjang, 900 meters (RMNH 6833/4], 126 unnum- bered), Tjusurupan (SNG 19887, 19391), Tasikmalaja, 350 meters (NHMW 16697:2). Djawa Tengah: Wonosobo, Kepil, 550 meters (RMNH 8555), Wonosobo, Selomojo, 600 meters (RMNH unnum- bered), Wonosobo, Tlogodjati, 1000 meters (RMNH unnumbered). 92 FIELDIANA: ZOOLOGY, VOLUME 49 Questionable localities: Java (BM 61.8.12.44, 63.12.11.171—holo- type of dimidiata; MHNP 174 and 3300[2]—syntypes of vermiformis; ZMB 4118—syntype of vermiformis). Java: Batavia (MHNP 39- 187; RMNH 80—holotype of temmincki). Java: Bandung (RMNH co.118). Dutch Indies (MNHP 12-51, 12-52, 12-54; RMNH 3975). Malayan Archipelago (NHMW 16694). Ternate (BM 78.1.381.8-9). No data (RMNH unnumbered). Specimens examined.—s886 Calamaria griswoldi Loveridge. Figure 23. Calamaria lumbricoidea griswoldi Loveridge, 1938, Proc. Biol. Soc. Washing- ton, 51, p. 43—Luidan River, Bundu Tuhan, Mount Kina Balu, North Borneo; de Haas, 1950, Treubia, 20, p. 570. Calamaria griswoldi, Marx and Inger, 1955, Fieldiana, Zool., 37, p. 1838. Holotype.-—Museum of Comparative Zoology 48580. Diagnosis.—Maxillary teeth modified; third and fourth supra- labials entering orbit; preocular present; mental touching anterior chin shields; color blackish brown above with a narrow white line between successive scale rows; immaculate yellowish white below. Description.—Rostral wider than high, portion visible from above 24 length of prefrontal suture; prefrontal */¢ length of frontal, touch- ing first 2 supralabials; frontal hexagonal, 114 to 2 times width of supraocular, about 24 to 34 length of parietal; parietal 124 to 2 times length of prefrontal; paraparietal surrounded by 5 shields and scales; nasal smaller than postocular; preocular present; neither ocu- lar as high as eye; eye equal to or slightly greater than eye-mouth dis- tance; 5 supralabials, third and fourth entering orbit, fifth the largest, first 4 subequal; mental triangular, touching anterior chin shields; 5 infralabials, first 8 touching anterior chin shields; both pairs of chin shields meeting in midline; 3 gulars in midline between posterior chin shields and first ventral. Body thickness index 0.017-0.089 (5 specimens); tail thick, tap- ering from base to a sharp point (Fig. 10B); dorsal scales reduce to four rows on tail opposite ninth to thirteenth subcaudal anterior to terminal scute. Hemipenis forked opposite fourth subcaudal; suleus spermaticus forked; calyces papillate (2 specimens). Nine to 10 modified maxillary teeth (4 specimens). Ventrals: males, 155-179 (mean 168.5; N=6); females, 188-192 (mean 187.2; N=5). Subcaudals: males, 16-18 (mean 16.7; N=6); females, 18-16 (mean 14.5; N=4). INGER AND MARX: THE SNAKE GENUS CALAMARIA 93 Fic. 28. Calamaria griswoldi (NMS R18949). Total length: males, 192-425 mm.; females, 875-488 mm. Ratio of tail to total length: males, 0.059-0.068; (mean 0.063; N=6); fe- males, 0.048-0.056 (mean 0.052; N=4). Color dark brown above; dark portion of scales without network; blackish brown stripes occupying central 24 of each scale row above the first, yellowish stripes on edges on adjacent scale rows; scales of first row yellow, immaculate in anterior part of body, usually each scale with a small dark spot in posterior half of body; head dark brown above; supralabials yellow in lower 24; head below immaculate yellow; an oblique light bar running forward from gular region onto rear of parietals; ventrals immaculate yellow; subcaudals yellow, usually a faint zig-zag dark line mid-ventrally. Distribution.—Highlands of northwestern Borneo (Fig. 72). NORTH BORNEO: Ranau District: Ranau (USNM 134114), Mount Kina Balu (BM 5.11.7-25), Lumu Lumu, 1525 meters (BM 1929.12.22.112), Bundu Tuhan, 1370 meters (CNHM 72434; USNM 180288), Luidan River, near Bundu Tuhan (MCZ 48580—holotype, 94 FIELDIANA: ZOOLOGY, VOLUME 49 43581—-paratype). Kota Belud District: Tenompok, 1430 meters (NMS 18949, 1 unnumbered). Western North Borneo (SM Cd. 5.26.c-d). SARAWAK. Fourth Division: Kelabit Plateau (SM unnumbered). Specimens examined.—12. Calamaria albiventer Gray. Figure 24. Changulia albiventer Gray, 1834, Ill. Ind. Zool., 2, pl. 86, figs. 6-9—Penang Island. Calamaria albiventer, Giinther, 1858, Cat. Colubrine Snakes, p. 4; 1864, Rept. Brit. Ind., p. 197; Boulenger, 1894, Cat. Snakes Brit. Mus., 2, p. 336; 1912, Fauna Malay Penin., Rept. and Batr., p. 156; Flower, 1896, Proc. Zool. Soc. London, 1896, p. 886; 1899, zbid., p. 674; Werner, 1929, Zool. Jahrb., (Syst.), 57, p. 170; Smith, 1930, Bull. Raffles Mus., no. 3, p. 58; Tweedie, 1953, Snakes Malaya, p. 50; Marx and Inger, 1955, Fieldiana, Zool., 37, p. 199; Batchelor, 1958, Malaya Nat. Jour., 12, p. 107. Calamaria indragirica Schenkel, 1901, Verh. Ges. Basel, 13, p. 164, fig. 3— Indragiri, Sumatra; de Rooij, 1917, Rept. Indo-Austr. Arch., 2, p. 154; Werner, loc. cit., p. 170; de Haas, 1950, Treubia, 20, p. 568; Marx and Inger, loc. cit. p. 202. Calamaria ornata Werner, 1909, Mitt. Naturh. Mus. Hamburg, 26, p. 229, fig. 6—Sungei Lalak, Indragiri, Sumatra; de Rooij, loc. cit., p. 178; de Haas, loc. cit., p. 572; Marx and Inger, loc. cit., p. 204. Syntypes.— British Museum (Natural History) 1946.1.2.10, 1946. eZee. Taxonomic notes—The striking color patterns of indragirica and ornata are identical to that of albiventer. As the counts and scale proportions also agree, we are putting the two Sumatran names in the synonymy of albiventer. Diagnosis.—Maxillary teeth modified; third and fourth supra- labials entering orbit; preocular present; mental touching anterior chin shields; body with 4 narrow light stripes. Description.—Rostral broader than high, portion visible from above about 24 length of prefrontal suture; prefrontal 34 length of frontal, touching first 2 supralabials; frontal hexagonal, 124 to 2 times width of supraocular, about ’/s length of parietal; parietal 114 times length of prefrontal; paraparietal usually surrounded by 5 shields or scales; nasal smaller than or equal to postocular; preocular present, neither ocular as high as eye; greater than eye-mouth dis- tance; 5 supralabials, third and fourth entering orbit, fifth the largest, first 4 subequal; mental triangular, touching anterior chin shields; 5 infralabials, first 3 touching anterior chin shields; both pairs of INGER AND MARX: THE SNAKE GENUS CALAMARIA 95 chin shields meeting in midline; 3 gulars in midline between posterior chin shields and first ventral. Body thickness index 0.026 (1 specimen); tail tapering gradually from base, abruptly tapering at tip to a sharp point; dorsal scales reduce to 4 rows on tail opposite fifth to eighth subcaudal anterior to terminal scute. Cloaca of female bilobed (1 specimen). Nine modified maxillary teeth (8 specimens). Ventrals: males, 1483-144 (N=2); females, 147-162 (mean 156.4; N=10). Subcaudals: males, 21-22 (N=2); females, 15-19 (mean 17.0; N=10). Fic. 24. Calamaria albiventer (CAS 14940). Total length: males, 205 mm. (N=2); females, 170-361 mm. Ratio of tail to total length: males, 0.088-0.098 (N=2); females, 0.047-0.088 (mean 0.066; N=9). Color of body brown above; two pairs of light longitudinal stripes; lower half of first scale row and adjacent edges of ventrals brown; adjacent halves of first and second scale rows yellowish; a broad dark stripe on upper half of second scale row, all of third and fourth 96 FIELDIANA: ZOOLOGY, VOLUME 49 rows, and lower half of fifth row; this dark band sometimes split by a lighter brown stripe formed by light speckling on adjacent halves of third and fourth scale rows; yellowish stripe on adja- cent halves of fifth and sixth scale rows; a dark stripe on vertebral scale row and adjacent halves of sixth rows; top of head light brown with a dark network or dark spots; upper third or half of supralabials brown; remainder of supraliabals and underside of head yellowish, usually with dark spots on infralabial sutures; ventrals yellow except for dark lateral edges; underside of tail yellowish with a dark median stripe. In life reddish brown above; upper light stripe red, lower light stripe bluish white; both light stripes edged with black; underside of head lemon yellow shading into red on neck; underside of body and tail bright red (Flower, 1899, p. 674). Geographic variation.—Two Sumatran specimens (syntypes of indragirica) have 147 and 148 ventrals and 19 subeaudals. One of these snakes is a female; the second was too desiccated to determine the sex. The third Sumatran animal (holotype of ornata) has 148 ventrals and 17 subcaudals (Werner, 1909b, p. 229); sex was not given. If these 3 snakes are females, Sumatran females have lower ventral and higher subcaudal counts than females from Malaya (156-160; 15-17). Distribution.— Western Malaya and east-central Sumatra (Fig. fO2\ MALAYA: Penang Island (BM 1946.1.2.10 and 1946.1.2.18—syn- types, 60.8.19.1269[2], 98.9.22.39). Perak: Larut Hills, 1870 meters (NMS unnumbered). Province Wellesley (BM 96.6.25.28). SINGAPORE (CAS 14940). SUMATRA: Sumatera Tengah: Indragiri (NMB 1697-98—syn- types of zndragirica), Indragiri, Sungei Lalak (holotype of ornata; not examined). ?India (BM 2 unnumbered). Specumens examined.—12. Calamaria hilleniusi new species. Figure 25. Holotype.—Zoologisch Museum Amsterdam 10078, a male from Samarinda, Indonesian Borneo. Paratypes.—Sarawak Museum unnumbered, a female from north- western Borneo; Agriculture Department, Sabah, No. 72, (male) and, INGER AND MARX: THE SNAKE GENUS CALAMARIA of Chicago Natural History Museum 142088, (female) both from Tua- ran, Tuaran District, Sabah (=North Borneo). Diagnosis.—Maxillary teeth modified; third and fourth supra- labials entering orbit; preocular present; mental touching anterior chin shields; paraparietal surrounded by 5 shields and scales; eye about 24 eye-mouth distance; dorsal scales without network. Description.—Rostral wider than high, portion visible from above equals half length of prefrontal suture; prefrontal equals length of frontal, touching first 2 supralabials; frontal hexagonal, 21% to 3 times width of supraocular, about 34 length of parietal; parietal 114 to 1% times length of prefrontal; paraparietal sur- rounded by 5 scales and shields; nasal oriented laterally, equal to postocular; preocular present; both oculars as high as eye; eye about 24 eye-mouth distance; 5 supralabials, third and fourth entering orbit, fifth the largest, fourth almost as long as fifth along ventral border, dorsal margin of fourth labial half of its ventral margin; first 4 supralabials subequal along labial border; mental triangular, touching anterior chin shields; 5 infralabials, first 8 touching anterior chin shields; both pairs of chin shields meeting in midline; 3 gulars in midline between posterior chin shields and first ventral. Body thickness index 0.041-0.052 (2 specimens); tail tapering gradually from base to a point; dorsal scales reduced to 4 rows on tail opposite third to ninth subcaudal anterior to terminal scute. Hemipenis in holotype too small for accurate description, sex determined by examination of gonads. Ten modified maxillary teeth (8 specimens). Ventrals 147-155 in males (holotype 155), 154-161 in females; subcaudals 18-21 in males (holotype 18), 14-16 in females; total length males 317-830 mm. (holotype 317), females 368-370 mm.; ratio of tail length to total length 0.072-0.106 in males (holotype 0.072), 0.078-0.076 in females. Color of holotype brown above, yellow below, the dark pigment ending abruptly on center of third scale row; dorsal scales without network; head brown above, the dark pigment ending in an oblique line from upper edge of second supralabial to lower third of fifth; supralabials yellow, the first dark brown dorsally and anteriorly; a small yellow spot near lateral margin of parietal; head immaculate yellow below; ventrals immaculate yellow; dorsal brown color ex- tending ventrally on tail to cover most of subcaudals, leaving a short yellow stripe along central part of each row of subcaudals. 98 FIELDIANA: ZOOLOGY, VOLUME 49 Fic. 25. Holotype of Calamaria hilleniusi. Color of paratypes dark brown above, one with 21 narrow (2-8 scales wide) crossbands of lighter brown, towards rear of body cross- bands become lighter; other paratypes uniformly dark brown above; dark pigment ending abruptly along second scale row; in anterior half of body first two scale rows immaculate yellow; in posterior half scales of first two rows with dark anterior edges; head dark brown above; supralabials immaculate yellow except for their dorsal edges; head immaculate yellow below; ventrals yellow, first 10 to 25 immaculate; other ventrals with dark spots at antero-lateral cor- ners, the dark area becoming wider in posterior part of body and occupying entire width of ventrals just before vent; subcaudals yellow, edged with dark brown. Comparisons.—Calamaria hilleniusi is similar to lumbricoidea in the relation of mental to chin shields, paraparietal and supralabial counts, and the absence of a network in the dorsal seales. It differs from lwmbricoidea in the presence of light crossbands in some adults INGER AND MARX: THE SNAKE GENUS CALAMARIA 99 and the reduction of the ocular region, which is reflected in a smaller eye and wider frontal. The reduction to four dorsal scale rows takes place closer to the tip of the tail and the number of subcaudals is lower in male hilleniusi than in males of Borneo lumbricoidea (com- pare above description with Table 27). Calamaria hilleniusi differs from other species related to lum- bricoidea, e. g., griswoldi and albiventer, in coloration and in the size of the ocular region. The patterns of hilleniusi are duplicated in Borneo in C. bicolor. Calamaria hilleniusit also resembles this species in its small eye and relatively wide frontal, but differs in having a laterally oriented nasal, the prefrontal separated from the third supralabial (compare Figs. 25 and 40), the mental touching the anterior chin shields, lower subeaudal counts (18-21 versus 21-28 in males; 14-16 versus 18-21 in females), and shorter tail ratios (0.072-0.106 versus 0.106-148 in males; 0.073-0.076 versus 0.079—-0.093 in females). Calamaria hilleniusi resembles schlegeli closely in body form, wide frontal, and small eye. The pattern of the holotype of hilleniusi is identical to that of schlegelt. Calamaria hillencusi differs from schleg- eli in having the mental touching the anterior chin shields, lower subcaudal counts (18-21 versus 27-42 in males; 14-16 versus 28 in females), smaller tail ratios (0.072-0.106 versus 0.120-0.2138 in males, 0.073-0.076 versus 0.099 in females), and thicker body diameter. The holotype of hilleniusi is known to be sympatric with schlegeli at Samarinda. The specimen of schlegeli from Samarinda ZMA 10079) is a male with the following characteristics: 186 ventrals, 34 subeaudals, and tail ratio 0.156. The difference in ventral counts between this specimen of schlegeli and the holotype of hilleniusi suggests that the latter probably has higher ventral counts. Distribution.— Borneo (Fig. 72). Calamaria muelleri Boulenger. Figure 26. Calamaria muelleri Boulenger, 1896, Ann. Mag. Nat. Hist., (6), 17, p. 894— Loka, Mount Bonthain, Celebes; 1897, Proc. Zool. Soc. London, 1897, p. 223, pl. 14, fig. 1; Boettger, 1898, Kat. Rept. Samm. Senck. Naturf. Ges., pt. 2, p. 86; de Rooij, 1917, Rept. Indo-Austr. Arch., 2, p. 161; Smith, 1927, Proc. Zool. Soe. London, 1927, p. 224; Werner, 1929, Zool. Jahrb., (Syst.), 57, p. 171; de Haas, 1950, Treubia, 20, p. 272; Marx and Inger, 1955, Fieldiana, Zool., 37, p. 204. Lectotype.—Naturhistorisches Museum Basel 1690, here desig- nated. 100 FIELDIANA: ZOOLOGY, VOLUME 49 Taxonomic notes.—Species of Calamaria on Celebes fall into three morphological groups, one of which is characterized by modified maxillary teeth and relatively low ventral counts. This group of four species shows no particular specialization though two of the species, nuchalis and muellert (Fig. 26), have narrowed snouts. Calamaria muellert and C. nuchalis are sympatric in southwestern Celebes in the vicinity of Mount Bonthain (Fig. 29). In that region the two differ in ventral coloration (immaculate in nuchalis, speckled in muellerz,) in lateral coloration (white stripe on first scale row in nuchalis, absent in muellerz), and in the number of scales and shields surrounding the paraparietal (6 in nuchalis, 5 in mueller). These differences also distinguish these species in central Celebes where they occur though they have not been collected at the same localities. Calamaria muelleri is apparently sympatric with brongersmaz, the third Celebesian species having low ventral counts, in the Lake Posso area. At that place the two differ in the relative lengths of frontal and parietal shields (frontal shorter than parietal in bron- gersmar, usually longer in muellerz), relation of the mental and an- terior chin shields (separated in brongersmaz, touching in muelleri,) shape of snout (broadly rounded in brongersmai, narrow and with an enlarged rostral in muellerz), and ventral coloration (bold black squares in brongersmaz, speckled in muellerz). The fourth of these species, boesemant, is not sympatric with muellert. The latter differs from boesemani in having the mental touching the chin shields and the frontal longer than the parietal and in lacking a light stripe on the first scale row. Diagnosis.—Maxillary teeth modified; third and fourth supra- labials entering orbit; preocular present; mental touching anterior chin shields; 5 shields and scales surrounding paraparietal; dorsal scales with a light or dark network; subcaudals 21 or less in males and 15 or less in females; portion of rostral visible from above at least equal to length of prefrontal suture; head distinctly tapered in front of eyes. Description.—Head distinetly narrowed anteriorly; rostral as high as wide, portion visible from above equal to or longer than prefrontal suture; prefrontal 24 length of frontal, touching first 2 sup- ralabials; frontal hexagonal, 134 to 3 times width of supraocular, equal to or longer than length of parietal; parietal 11% times length of prefrontal; paraparietal surrounded by 5 shields and scales; nasal smaller than postocular; preocular present; neither ocular as high as eye; eye equal to or slightly greater than eye-mouth distance; 5 INGER AND MARX: THE SNAKE GENUS CALAMARIA 101 supralabials, third and fourth entering orbit, fifth the largest, first 4 subequal or second slightly larger; mental triangular, touching anterior chin shields; 5 infralabials, first 3 touching anterior chin shields; both pairs of chin shields meeting in midline; 3 gulars in midline between posterior chin shields and first ventral. Body thickness index 0.007-0.042 (10 specimens); tail gradually tapering to a moderate point; dorsal scales reduce to 4 rows on tail opposite second to eighth subcaudal anterior to terminal scute. Hemipenis forked opposite fifth or sixth subcaudal; retractor beginning opposite seventh to ninth subcaudal, calyces smooth (4 specimens). Cloaca cardioid (1) or bulbous (1). Nine to 11 modified maxillary teeth (8 specimens). Ventrals: males, 129-155 (mean 146.7; N=80); females, 155-178 (mean 167.7; N=388). Subcaudals: males, 16-21 (mean 18.5; N=29); females, 9-15 (mean 12.8; N=88). Total length: males, 116-262 mm.; females, 186-855 mm. Ratio of tail to total length: males, 0.072-0.101 (mean 0.089; N=29); females, 0.041-0.062 (mean 0.050; N=87). Color brown above, each scale with a fine dark network; scattered scales with dark central spot; spots not forming lines; first scale row usually lighter than those above but without light longitudinal stripe; head brown, densely spotted with blackish brown; a dark stripe usually running from nasal back through eye along lower por- tion of prefrontal and upper margins of supralabials, descending slightly to include upper half or two-thirds of last supralabial and lower edge of parietal; remainder of supralabials yellow with a dark stripe along lower edge or with dark sutures; underside of head yellow with dark spots; ventrals highly variable but always with dark brown pigment, in the form of small spots in longitudinal rows, in the form of speckling, or as transverse bands along anterior half of each ven- tral; underside of tail with a broad median brown stripe and a nar- rower brown stripe along outer edges of subcaudals, the two stripes separated by a yellow stripe; in some specimens underside of tail dark brown with a few isolated spots. Distribution.—Central and southwestern Celebes (Fig. 29). CELEBES (NMS 2 unnumbered; ZMB 15688), south Celebes (BM 96.4.29.82-33): Mount Bonthain (NHMW 16691; ZSBS 545/20), Djikoro (AMNH 68381-82; BM 1926.8.20.159-166; CNHM 83162; MCZ 25815-24, 25326-33; USNM 120818), Loka (CNHM 104619; NMB 1690—lectotype, 1691-93—-paratypes, 5227), Idrulaman(BM 102 FIELDIANA: ZOOLOGY, VOLUME 49 96.4.29.27[2]), Bua Praeng (NHMW 16692: 1-10; SNG 19942-43) ; Luhu (BM 96.12.9.67), between Lake Posso and Gulf of Tomini (NMB 1694). Fic. 26. Paratype of Calamaria muelleri (NMB 1692). No data (BM 1946.1.1.82-84; MZUS 2 unnumbered). ?Java (ZMB 13946, 13948/3]). As these snakes were collected by the same person (Fruhstorfer) and on the same date as SNG 19942-438, they certainly must have come from Celebes. Specimens examined.—b68 Calamaria joloensis Taylor Calamaria joloensis Taylor, 1922, Phil. Jour. Sci., 21, p. 203—central Jolo Island, Philippine Islands; Werner, 1929, Zool. Jahrb., (Syst.), 57, p. 170; Marx and Inger, 1955, Fieldiana, Zool., 37, p. 203; Leviton, 19638, Proc. Calif..Acad. Sci., 31, p. 387. Holotype.—California Academy of Sciences 60901. Taxonomic notes.—The holotype has the lowest ventral count we have observed in the genus. The number of supralabials, eye size, position of nasal, and relation of mental and chin shields recall hilleniusi of Borneo. But joloensis differs from that species in having relatively longer parietals (equal to half of the head length), a blunter tail, fewer ventrals, and light network on each dorsal scale. Except for the much smaller eye and lower ventral count, 7oloenszs is similar to gervaist of the Philippine Islands. As the unique specimen of joloensis is not sympatric with either hilleniust or gervaisi, a certain decision as to the specific differentia- tion of joloensis is not possible now. However, because we have not INGER AND MARX: THE SNAKE GENUS CALAMARIA 103 observed intraspecific variation in eye size that could account for the difference between gervaisi and joloensis, we recognize the latter as a full species. For a similar reason, the shape of the head suggests that joloensis is specifically distinct from halleniusze. Diagnosis.—Maxillary teeth modified; third and fourth supra- labial entering orbit; preocular present; mental touching anterior chin shields; eye about two-thirds eye-mouth distance; dorsal scales with a light network. Description.—Rostral broader than high, portion visible from above about half length of prefrontal suture; prefrontal shorter than frontal, touching first 2 supralabials; frontal hexagonal, 3 times width of supraocular, about 24 length of parietal; parietal 114 times length of prefrontal; paraparietal surrounded by 5 shields and scales; nasal larger than postocular; preocular present; neither ocular as high as eye; eye ’/1) of eye-mouth distance; 5 supralabials, third and fourth entering orbit, fifth the largest, first, third, and fourth sub- equal and smaller than second; mental triangular, touching anterior chin shields; 5 infralabials, first 3 touching anterior chin shields; both pairs of chin shields meeting in midline; 8 gulars in midline between posterior chin shields and first ventral. Body thickness index 0.010; tail short, thick, tapering near end to a blunt tip; dorsal scales reduce to 4 rows on tail opposite sixth subcaudal anterior to terminal scute. Hemipenis bifurcate. Eight modified maxillary teeth. Ventrals 119 in male; subcaudals 138; total length 144 mm.; ratio of tail length to total length 0.069. Color faded in preservative but dark brown above, each scale with a light network; head dark brown above and on sides; head below evidently with brown spots on labials; ventrals yellowish, those in anterior part of body with dark pigment in antero-lateral corners, increasing amounts of dark pigment on ventrals farther back on body; in posterior two-thirds of body most ventrals with a dark band across anterior halves; underside of tail with dark pigment but pattern too faded for accurate observation. Distribution.—Jolo Island, Philippine Islands (Fig. 68). CENTRAL JOLO ISLAND (SU 60901—holotype). 104 FIELDIANA: ZOOLOGY, VOLUME 49 Calamaria bitorques Peters. Figure 27. Calamaria bitorques Peters, 1872, Monatsber. Akad. Wiss. Berlin, 1872, p. 585 —Philippine Islands; Boulenger, 1894, Cat. Snakes Brit. Mus., 2, p. 338; 1896, Cat. Snakes Brit. Mus., 3, p. 646; Boettger, 1886, Ber. Senck. Naturf. Ges., 1886, p. 105; Taylor, 1922, Snakes Phil. Ids., p. 185; Werner, 1929, Zool. Jahrb., (Syst.), 57, p. 170; Marx and Inger, 1955, Fieldiana, Zool., 37, p. 200; Leviton, 1968, Proc. Calif. Acad. Sci., 31, p. 390. Calamaria gervaisi (non Duméril and Bibron), Boettger, 1898, Kat. Rept. Samm. Senck. Naturf. Ges., pt. 2, p. 83 (part). Holotype.—Zoologisches Museum Berlin 7444. Taxonomic notes.—C. bitorques is very similar to C. gervaisz in counts and shape of head scales. The two differ in coloration and size. The former has black edged crossbands on the anterior part of the body (Fig. 27). C. gervaist may have narrow yellow rings behind the head forming a wide cross band, but the band thus formed does not have dark edges. The difference in color pattern is not an effect of ontogenetic change for the characteristic pattern of betorques is present in specimens in the size range 150-260 mm., which is well within the size range of gervazst. The largest gervaist we have seen measured 821 mm. None of the more than 80 gervaist from Luzon exceeded 290 mm. By con- trast 11 of 20 b¢torques (known from Luzon only) were longer than 300 mm., with a maximum of 420 mm. We have examined specimens of both species from Manila, Mount Isarog (Camarines Sur), and Bulusan (Sorsogon). The size and coloration differences just noted distinguish these two species at each of these localities. These species also differ in average number of subcaudals. Fe- male gervaisi from Luzon have 10-15 subeaudals (mean 12.67, mode 13, N=78); female bitorques have 12-17 (mean 15.3, mode 15, N=14). Male gervaisi from Luzon have 15-20 subeaudals (mean 16.68, mode 16, N=65); male brtorques have 17-20 (mean 18.83, N=6). Female gervaist from Manila have 10-14 subcaudals (80 specimens) and fe- male bitorques 15-17 (2). Female gervaisi from Mount Isarog have 11-18 subcaudals (4 specimens) and female bitorques 14-15 (2). A single female of gervaisi from Bulusan has 16 subcaudals and a single female of betorques 14. Males of both species were not available from these localities. Diagnosis.—Maxillary teeth modified; third and fourth supra- labials entering orbit; preocular present; mental touching anterior chin shields; paraparietal surrounded by 5 shields and scales; 2 to 6 dark, black-edged, crossbands behind the head. INGER AND MARX: THE SNAKE GENUS CALAMARIA 105 Description.—Rostral as wide as high, portion visible from above about 24 length of prefrontal suture; prefrontal shorter than frontal, touching first 2 supralabials; frontal hexagonal, 114 to 2 times width of supraocular, about °/; length of parietal; parietal 11% times length of prefrontal; paraparietal surrounded by 5 scales and shields; nasal much smaller than postocular; preocular present; postocular not as high as eye; eye equal to or slightly greater than eye-mouth distance; Fic. 27. Holotype of Calamaria bitorques. 5 supralabials, third and fourth entering orbit, fifth the largest, first 4 subequal; mental triangular, touching anterior chin shields; 5 infralabials, first 3 touching anterior chin shields; both pairs of chin shields meeting in midline; 3 gulars in midline between posterior chin shields and first ventral. Body thickness index 0.016—-0.063 (7 specimens); tail thick, ta- pered near end to blunt point; dorsal scales reduce to 4 rows on tail opposite sixth to eleventh subcaudal anterior to terminal scute. Cloaca of females bilobed (1) or cardioid (1). Nine modified maxillary teeth (2 specimens). 106 FIELDIANA: ZOOLOGY, VOLUME 49 Ventrals: males, 150-157 (mean 154.0; N=5); females, 157-197 (mean 173.4; N=14). Subcaudals: males, 17-20 (mean 18.8; N=6); females, 12-17 (mean 15.3; N=14). Total length: males, 150-833 mm.; females, 222-420 mm. Ratio of tail to total length: males, 0.066-0.087 (mean 0.076; N=5); fe- males, 0.039-0.086 (mean 0.057; N=14). Color light to dark brown above; dorsal scales with fine dark network; 2 to 6 dark-edged saddles on anterior part of body reaching the second or third scale row; behind area of saddles dorsum usually with scattered small, dark spots; scales of first row usually immac- ulate yellow at least in anterior half of body; head brown above with obscure dark markings; supralabials usually cream-colored without dark spots; a dark streak extending from nasal across lower edges of prefrontal and parietal and upper edge of last labial; underside of head yellowish, usually immaculate; ventrals immaculate yellow or with dark lateral edges; underside of tail immaculate yellow or with narrow dark mid-ventral streak. Distribution.—Luzon (Fig. 68). LUZON (BM 72.8.20.49; ZMB 3778, 7444-holotype, listed in Ber- lin Museum catalogue as from “‘Luzon’’), central Luzon (SNG 19394). Cagayan Province: Cape Engano (BM 95.9.21.4). Isabela Province (CAS 15295). Rizal Province: Manila (MHNP 00-864- 65), area between Subic Bay south to Clark Air Force Base (MVZ 73672). Camarines Sur Province: Mt. Isarog, Curry, Pili (CNHM 142515-16). Sorsogon Province: Bulusan (CNHM 142510-14). PHILIPPINE ISLANDS (BM 72.10.11.16, 72.10.11.20, 1 unnum- bered; MCZ 25776). Specimens examined.—20. Calamaria gervaisi Duméril and Bibron. Figure 28. Calamaria gervaisii Duméril and Bibron, 1854, Erp. Gén., 7, p. 76—‘‘Java’”’ {in error]; Gtinther, 1858, Cat. Colubrine Snakes Brit. Mus., p. 4; Jan, 1862, Arch. Zool. Anat. Phys., 2, p. 8; 1865, Icon. Ophid., 10, pl. 2, fig. 1; Fischer, 1885, Jahrb. Wiss. Anst. Hamburg, 2, p. 80; Boettger, 1886, Ber. Senck. Naturf. Ges., 2, p. 105; 1898, Kat. Rept. Samm. Senck. Naturf. Ges., pt. 2, p. 83; Boulenger, 1894, Cat. Snakes Brit. Mus., 2, p. 338; Taylor, 1922, Snakes Philippine Ids., p. 186; Werner, 1929, Zool. Jahrb., (Syst.), 57, p. 171; Marx and Inger, 1955, Fieldiana, Zool., 37, p. 180. Calamaria gervaisii gervaisti, Taylor, op, cit., p. 187; Leviton, 1963, Proc. Calif. Acad. Sci., 31, pp. 390, 394, 407. INGER AND MARX: THE SNAKE GENUS CALAMARIA 107 Calamaria mindorensis Boulenger, 1895, Ann. Mag. Nat. Hist., (6), 16, p. 481 Mindoro; 1896, Cat. Snakes Brit. Mus., 3, p. 646; Taylor, op. cit., p. 190; Werner, loc. cit., p. 171; Marx and Inger, loc. cit., p. 207. Calamaria gervaisii iridescens Taylor, 1917, Phil. Jour. Sci., 12, p. 360— Can- laon Volcano, Negros; 1922, op. cit., p. 188; Brown and Alcala, 1961, Ecology, 42, p. 632; Leviton, loc. cit., pp. 384, 398, 402. Calamaria tropica Taylor, 1922, Snakes Philippines Ids., p. 194—-near Naujan, Mindoro; Marx and Inger, loc. cit., p. 206. Calamaria polillensis Taylor, 1923, Phil. Jour. Sci., 22, p. 549—Polillo, Polillo Island; Werner, loc. cit., p. 171; Marx and Inger, loc. cit., p. 205. Calamaria hollandi Taylor, 1923, Phil. Jour. Sci., 22, p. 550—Port Holland, Basilan; Werner, loc. cit., p. 171; Marx and Inger, loc. cit., p. 202. Calamaria gervaisi hollandi, Leviton, loc. cit., pp. 379, 393. Calamaria gervaisit polillensis, Leviton, loc. cit., p. 403. Syntypes.—Museum National d’Histoire Naturelle ¢2314~-7202, two specimens labeled “‘Java’’. No author subsequent to Duméril and Bibron (1854) has reported this species from outside the Philip- pine Islands. We now believe, contrary to our previous opinion (Marx and Inger, 1955), that the type locality cannot be restricted to a particular island within the Philippines except arbitrarily. Taxonomic notes.—We have examined the holotype of mindorensis Boulenger and cannot distinguish it from gervazst. Boulenger (1895) did not compare mindorensis with any other Calamaria. Although our counts (ventrals 177, subcaudals 17) on the holotype differ from Boulenger’s (ventrals 193, subcaudals 15) we are certain that the specimen examined is the holotype of mindorensis because of the agreement in coloration, size (total length 232 mm., tail length 13 mm., according to our measurements; 240 and 13 mm., according to Boulenger), and shield arrangement. Furthermore it is the only Cala- maria from Mindoro in the collection of the British Museum. Calamaria tropica Taylor was distinguished from all other species on the basis of the presence of loreals. These “‘loreals’ are tiny, triangular scales much smaller than even the nasal and are wedged between prefrontal, preocular, and second supralabial. Though one of these scales appears on each side, they differ in size and we can only conclude that they represent fragments of the prefrontals. An additional specimen of gervaist (SU 21804) has a similar fragment on one side of the head. The holotype of tropica is certainly conspecific with the other Mindoro specimens of gervaist we have examined. Calamaria polillensis was distinguished by Taylor (19238) from gervaist on the basis of its lower ventral and subcaudal counts and the presence of a neck ring. As may be seen in Tables 28-31, the gel I I I §° SPT i aa T i! z I I ; I I I < : ¢ g i ZO TEL PSI ¢ @ : : €°9CT I Ll : Zel : G*6FT : 1 I I I I L I I 7 I z I 9 8 OL 19°06 IST ¢ g 6 we nae TS — uray e9r o9f £Sf ‘Sr It str Sr eYr 68% 981 881 S19]U99 SSBID "as~pasah DILDULD]DD JO Soe UL syUNOD [eIjUeA Jo UOINqIIysIq AoUeNbeIy—'8Z ATAVL ail») fo, (8) (6) @) Oe) ellen = marie, ue [ise q “-"* Zuepule yy “IA Ont, ti Ova dug O80 0 UCBMEBSE IL aD OmOND UND Odo. ody IW Pc. DeOND OfDpOeOLD G10 0.006 nqeg pLicuretien autare “u0B jue a) :. Al W avnelesl eS Vendon. OTe ajonseundg "* “$0188 Nep souseanD J cate vel enters : (18302) SOIB9N eS ailsike fe) ice yee O[LO[] :Avued o) Helis! ©. sesusice) edteWelre mein ils OT[Od oo 0 23 SOUBEE SOG +2 Fa ape TAT ey or 01 Votre sulinbe jy IVA 2 0S ee OSSoul Pee MORAL Ol Ooo BllUB IL PREC OM OT OKC " ({@0}) uozn'T 108 L8tl r8°0F0'OLT L° cbt O9T SFI L°OLT 98° 0FR' COL aSFursyy mr Nar i! 68T G82 941 i.) T i T oT I I I ¢ ¢€ & ra Or ¢ G CG OG I I qT T I if I tT oT z € I v g tL -Sf 9 7 T SI rome GLI OLE L9T TOT i I 9 & I9T 8ST S19}U9) SSBIT) T CST 6ST T 671 971 ETI Ce ee ae ee ee uvARYL J, oe we ee we ‘AITULY IJ IA ence ar eR ee Bote er Me ee em wes wns J, nid, a See de SS ( [e3 04) oVUuepuUly[yy BUuRl[aysedg BT a es “Gh Sr tai Santee sd ete onns eseVlLInzn'y ES: Gh cae ajonseuNng "* * “SO1B9N ap sousong ee es) SO189N Lee es thee s OO SABER ee ae ee a ee er ee ee Svlqr J, a oe a ee Ta tee pas ee are ae ere cee O[od Sie el ACE Ge a ae de Wel O1OPUI[ : Bo eb “eeie sey steey ee on feels ueleq(eg etoie): Gos Selig cans aMep ons emtetne IOPIsa11I09 si, Glare Vey SI SanOMio eu eretons uene ues Ss Ws ‘outanee mee oak emea utente UBOTIUIS iat Semen eee ne ere soueg so'T™ a telren open Leas sullinbe fy IA a, Qeuecd ee eatemsem subereercete omnseg ee ee Ce a ee ee eo youq go aS are de” ey las Ta ee, fee ellue ee eC eC net ee ({@104) uozn'T, "281MA19B DILDUDIDD JO SayeUla UL SJUNOD [eUaA Jo UOIyNqIIysIq AoUenbely—'6Z ATAVL 109 110 FIELDIANA: ZOOLOGY, VOLUME 49 counts of the holotype and paratype we have seen fall within the ranges of gervaisi from other islands; the two paratypes not examined lie within the same range of variation. Taylor (ibid.) compared C. hollandi with polillensis, which we have just shown to be a synonym of gervazsz, but not to gervaisi. The character—extent of the rostral visible from above—by which Taylor distingiushed holland: from polillensis shows considerable variation in gervais?. The direct comparison of the holotype of hollandi to various populations of gervaisz does not reveal any differences of diagnostic value. Taylor (1917) distinguished his Negros form, C. gervaisz iridescens, from gervaist from other islands in the Philippines by ventral and subeaudal counts and by size. Marx and Inger (1955) substantiated the count differences on the basis of 17 specimens from Luzon and Mindanao. However, with larger series available these differences do not appear to be taxonomically significant (Tables 28-31). Only the holotype of zridescens is unusually long (306 mm.). The largest of the 5 paratypes measured 240 mm. Of the gervaisi we measured, 18 of 97 from Luzon, 1 of 2 from Tablas, 4 of 8 from Panay, 1 of 9 from Mindanao, and 10 of 49 from Negros exceeded 250 mm. Fic. 28. Calamaria gervaisi (CNHM 150381). INGER AND MARX: THE SNAKE GENUS CALAMARIA 111 As approximately the same proportion of the Luzon and Negros samples fall in this size range, the supposed difference between 77r7- descens and other gervaisz has no taxonomie significance. Diagnosis.—Maxillary teeth modified; third and fourth supra- labials entering orbit; preocular present; mental touching anterior chin shields; paraparietal surrounded by 5 shields and scales; sub- caudals of males under 28, of females under 20; no light stripes above second scale row; a dark-edged, interrupted, light stripe on first scale row usually present; no dark-edged dorsal saddles; head not conspicuously tapered before eyes. Description.—Rostral broader than high or higher than broad, portion visible from above 14 to equal to prefrontal suture; pre- frontal about */; length of frontal, touching first 2 or 3 supralabials: frontal hexagonal, 114 to 21% times width of supraocular, about 34 to */; length of parietal; parietal 114 to 124 times length of prefrontal; paraparietal surrounded by 5 shields and scales (rarely 3, 4, or 6); nasal smaller than postocular; preocular present; neither ocular as high as eye; eye slightly less than or slightly greater than eye-mouth distance; 5 supralabials, third and fourth entering orbit, fifth the largest, first 4 subequal, or fourth narrower than others; mental triangular, touching anterior chin shields; 5 infralabials, first 3 touch- ing anterior chin shields; both pairs of chin shields meeting in mid- line; 3 gulars in midline between posterior chin shields and first ventral. Body thickness index 0.007-0.030 (21 specimens); tail thick, tapering gradually to a point; dorsal scales reduce to 4 rows on tail opposite third to twelfth subcaudal anterior to terminal scute. Hemipenis forked opposite fourth to sixth subcaudal; retractor muscle beginning opposite ninth or tenth subcaudal; sulcus bifurcate; calyces smooth (2 specimens). Cloaca of female shallowly bilobed (3) or cardioid (1). Eight or 9 modified maxillary teeth (9 specimens). Ventrals: males, 182-164 (mean 151.2; N=81); females, 142-190 (mean 166.1; N=101) (Tables 28 and 29). Subcaudals: males, 15- 21 (mean 17.2; N=110); females, 10-18 (mean 13.2; N=129) (Tables 30 and 81). Total length: males, 79-251 mm.; females, 104-321 mm. Ratio of tail to total length: males, 0.063-0.096 (mean 0.075; N=81); females, 0.041-0.070 (mean 0.051; N=98). Color light or dark brown above; each scale above first row yellowish with a dark network; some scales with a dark spot forming 112 FIELDIANA: ZOOLOGY, VOLUME 49 longer or shorter dark stripes; scales of first row dark or with clear yellow centers forming a row of yellow dots; head above same color as body with irregular dark spots; lower half of supralabials yellowish, occasionally sutures dark brown; 1 or 2 complete or incomplete yellow rings on head and neck, present or absent (Table 32); under- side of head yellowish, usually with small dark spots; ventrals yel- lowish with dark lateral edges, or yellowish with dark transverse bands across anterior halves or any condition between these two extremes (see Geographic variation and Table 32); underside of tail similar to belly, with or without a dark mid-ventral streak. TABLE 30.—Frequency Distribution with Respect to Certain Characters of Males of Calamaria gervaisi. Subcaudals 15 16 17 18 19 20 Gi Mean + SE Lwzony 2-3: 4 30 19 8 3 1 16.68+0.10 Mindoro.... 2 al 1 170 Pohllo........ il 19 Panay: 23.0% 2 1 18.3 Negros. . 7 9 9 1 18.15+0.17 Cebu... 2 18 Mindanao.. . 1 1 1 19.3 Basilan..... at Ly Position of reduction to four dorsal scale rows! DAO. STS. COS SO IIT ICTS 1, Luzon...... ts 10. 22> 1410) 2) 2 i 9.5+0.2 Mindodo.... 1 2 1 9.3 Polillo.:. 2. 1 12 Panay ..2.005 Zak 1253 Negros , 5 & 6 6 3, 5 11.5+0.3 Cebu.......: 1 1 9.5 Mindanao... af 1 if 8 Basilan..... 1 5 Tail ratio? Class centers 64-67 70.73 76 79-82 85 88 V91- 9h. 97 PUZONe ss. -) 1 4 + a a | lyf 6 1 1 73+0.7 Mindoro.... 1 1 i i 76 Polillo:...4 il 96 Panayve.. a. a Al 71 Negros... 2% oto: “WSO Rs eA es 2: sal: 76+0.9 Cebure 235. i 83 Mindanao... 1 1 1 85 Basilan..... 1 88 ! Position given in terms of subcaudals counted from vent. * Ratio given in thousandths of total length. INGER AND MARX: THE SNAKE GENUS CALAMARIA 113 Geographic variation.—C. gervaist shows an extremely wide range of ventral counts if the entire species sample is considered, but rel- atively narrow ranges of variation within local samples (Tables 28 and 29). For example, the total range of ventral counts in males from Negros Island extends from 144 to 164 whereas males from Cuernos de Negros have exactly half that amount of variation. In general, the populations from the southern Philippine Islands have lower ventral counts than those from the central and northern is- lands. The differences between the 2 large samples, Luzon and Ne- gros, are statistically significant though we attach no taxonomic importance to a difference of only 5 ventrals. Subcaudal counts (Tables 30 and 31) of the northern populations are lower than those from the central and southern populations. TABLE 31.—Frequency Distribution with Respect to Certain Characters in Females of Calamaria gervaist. Subcaudals 10 ii 12 13 if 15 16 i7 18 Mean + SE TOON Se gees ek es 1 10 19 34 12 2 12.67+0.11 Mindoro: «.< 66. acs s's i 1 1 16 POWIG 6 664.555 ¢o% 40% 1 15 Tablas ee ae 1 1 1325 PONGY 6-5 baleen 3. U8 13.5 NG@G@TOS: 4 besa eos 2.2 6 TL: 3 1 13.60+0.25 Mindanao......... 4 Zz 14.7 Position of reduction to four dorsal scale rows! 3 4 5 6 ff 8 9 10 11 Luzon............. 1 5 8 23 16 8 38 4 6.38+0.2 Mindoro........... 1. 1 8 PONG tae. 8k. cS wes ed 1 ff Tablas............ 1 1 7,5 Panay ..c.¢seaghbas 1 2 Ok 11.8 Neeros: ..6.60.0.45 2 4 8 4 9 2 7.7+40.3 Mindanao be ne 2 1 1 6.8 Tail ratio? Class centers 42 45 48 51 5, 57 6O 63 66 69 WGUZON GS os cess hoe ee on LO To TZ v4 1 2 49+0.6 Mindoro 2 ee = 1 i 1 60 Politlo.... 2.20.45... 1 68 Tablas. . rf 1 53 Panay.. 2 ie Ce 48 INegros. ....-...... 1 ot Sebew6e. A 8 1 51+1.2 Mindanao......... 1 1 1 1 1 1 61 1 Position given in terms of subcaudals counted from vent. * Ratio given as thousandths of total length. we od) [ntl GU I OlOsONO- DEO OPOO O00 0l0 0 GClneG 6 ooo. 0 DO 000 UBVMBSBL Aor OO nN 0 G.0fO Om GOO mm aoa onan 0D 5 uveAey}e J, mote Ono tH On GO Ordo Ofo OD OOO OOO 0 6.0 0 ody IW 6 op do Oo 0 D6" 6 obo OD 660 GO.0-0 AgUL yoy IIA Oe OLOMONO. OP Cet? Oe8tn DO 'ONdaD Oo OOo ON 6.020 070 D0 0 wns J, mm OO OO rr Hono do CbecGc Em dun CED UO OLOkO. Do) 040.0 Ol DNdl 0, 00.0 nqag arirel carcoWventen sew cnven ts Meuse) tolecuget toe ci ccmtemtel fonkooks BuB][eyse 4) eT Solna micro. Dadecy ou koto *ONON Oso nocor G, 3 ‘sc esevlinzny eo, fei soy se) fey 8) (00 6: eo) ie, iu 0) ie: etueuici els site ue .elisita uov[uey IN @ 16: jot Velve: Gey ierielce) 1s Ye) Oo q o ue) = e FT 2 0S 9B tithe Bete aii eatin ee aaa uefnen :OLOPULIAT eer Joptsa1109 T = | > + @ «dete sigeep Set Rane panei varia uene ues mle 6 RO RR TE stapes ST ace Oana BIa]][IPIOD p.® ohbig hore tee oe UBOTIUIS r Zi. PEERS Rae OR acai og sourg Sola | Z Ce ere ee ee eT eee ae PROMO: oO yeARly VIN Oh Wiikne ae maeriea eS Surinbe yj “IN laghevsei je) SueieD ae Jalgs\ weylsae! GP Tanis) tase Mee eae me miemeeee ens qassa'T Or nN g = bo 3 = T PN oe sedancoie ahi OS aaa cain ee ee joRVq , IQ ct tte tsetse reese pple yy :uOZzn'T] & | 9 S vi [ejotued pulyeq seleog jejelued puryaq salvos [eqoleg IB[JOo ,AIB][OD puodas JO UOTIISOG Ie][Od a[BuIs JO UOT}ISOg ON SIv]joo ajqnog "2StMAlah DIADUWDYID UL SIVTJOD MOT[AA JO uolyNqiystq Aouenbely—'zZE ATAVL 114 INGER AND MARX: THE SNAKE GENUS CALAMARIA 115 The differences between the Luzon and Negros samples are statis- tically but not taxonomically significant. Position of the reduction to 4 dorsal scale rows is farther removed from the vent in the central populations than in the northern and southern ones. Again differences between Negros and Luzon samples are statistically but not taxonomically significant. The ratio of tail length to total length shows no significant geo- graphic variation (Tables 30 and 31). As noted above, the position and number of yellow collars varies (Table 32). Almost all Mindanao specimens have a yellow ring crossing the posterior ends of the parietal scales. Most specimens from Negros have no yellow collars, but if they are present, a single one occurs 4 or 5 scales behind the parietals. Few Luzon specimens have yellow rings; those that do have either a single ring crossing the ends of the parietals or two collars, one crossing the parietal and the second 4 or 5 seales behind the parietals. As Table 32 shows, variation in this character is extremely local. For example, most of the specimens from Los Bafios, Luzon, have a single yellow collar, which is found in only 2 of 99 from Manila, less than 30 miles away. Similarly, most of the specimens from Dumaguete, Negros, have yellow collars, whereas only !/; of those from Cuernos de Negros, 10 miles away, do. Coloration of the ventrals varies from immaculate yellow (except for lateral edges) through increasing amounts of dark spots, ending in broad dark bands crossing the anterior 4% or 24 of each ventral (Table 33). All gradations between these extremes may appear in a single local population, as for example, at Manila. In other sam- ples, for instance, the small one from Baguio, Luzon, or the larger one from Cuernos de Negros, most individuals may be concentrated at one end of the scale. The characters discussed above vary locally and independently of one another. Under these circumstances subspecies should not be recognized. Distribution.—Philippine Islands (Fig. 68). LUZON (BM 72.8.20.50; NHM W 16705[2]; NMB 1704-06; USNM 06345; ZMB 3746[2], 4021[3], 6272[3],7441[2]). Central Luzon (SNG 19395-98) ; north Luzon, Cordillera Range (CNHM 109988); north Luzon, Lesseb (NHMW 16993). Mountain Province: Balbalan (CAS 61555-56), Baguio (CNHM 100872; MCZ 25758-56; NMB 7905-06, 7908). Pampanga Province: Mount Arayat (CAS 61812- 116 FIELDIANA: ZOOLOGY, VOLUME 49 TABLE 33.—-Frequency Distribution of Ventral Patterns in Calamaria gervaisi. Heavy spotting Few dark or narrow Wide dark Immaculate spots dark bands bands Luzon (total) nivcsssacces, BZ 44 44 13 Manta? =... fn0% eee: 20 39 30 4 Da6Gy oa « nk ow sh hele 1 Montalban. . ccaars 4 Z 6 Basi0 res od: wae os il 7 Mt. Maquiling........ 2 i 1 Mt, Arayat....514.<6<%¢ 2 LOS, Ban0S dic te sa ac 3 uf 2, SiMMlOal, 24 os cain ee eens 3 Cordillera. cages ant: i! 13). Rizal Provinee: Manila (AMNH 683880; BM 19388.8.7.49; CAS 61746-57; CM 2468-2507, 2509-86; MCZ 20086, 25757-67, 25770; MNHP 5506, 003859-60; NHMW 16704; SU 15969; ZMH 497), San Juan (CAS 15292). Montalban (CM 2587-88; CNHM 15025- 31, 15083-86; MCZ 20087). Cavite Province: Corregidor Island (USNM 71700). Laguna Province: Los Bafios (CAS 61156-57; SU 7353, 7362-64), Mount Maquiling (CM 2540[2], 2542; CAS 15815- 16), Siniloan (SU 15958-61). Camarines Norte Province: Daet ( ZMH 4062). Camarines Sur Province: Mount Isarog, Curry, Pili CNHM 142517-21). Sorsogon Province: Bulusan (CNHM 142522). POLILLO (CAS 62455—holotype of polillensis; MCZ 25775—para- type of polillensis). INGER AND MARX: THE SNAKE GENUS CALAMARIA 117 MINDORO (BM 95.11.7.27—holotype of mindorensis; CAS 73766— 69). Abra de Ilos, near Calapan (MCZ 37699), Naujan (CAS 62069 —holotype of tropica). TABLAS (MCZ 25751-52). PANAY. Iloilo (SU 15958-57, 15962-65). NEGROS (NMB 9514-15). South Negros (BM 77.12.138.29). Neg- ros Occidental Province: Victorias (USNM 78118), Silay (USNM 78134-85, 80582), Mount Canlaon (CM 2626, 2633, 8807, 8809- paratypes of zridescens), Cabagna-an Barrio, about 16 km. E. of La Castellana (SU 19873-76). Negros Oriental Province (MCZ 17577, 46681—-paratype of zridescens): Luzuriaga (SU 15949), Dum- aguete (CNHM 573808; SU 15947-48, 15951), Batinquil Barrio, 5-6 km. W. of Dumaguete (SU 18774), Bonghong Sitio, about 10 km. W. of Dumaguete (SU 18889), 1.5 km. W. of Valencia (SU 17981), 4-5 km. W. of Valencia (SU 17929-380), 5-6 km. W. of Valencia (SU 18238, 18908), 6 km. W. of Valencia (SU 18241-42), north side of upper Maite River ravine (SU 31866), south side of Maite River, east side of Cuernos de Negros (SU 18230-37, 18239, 18770), north peak area of Cuernos de Negros (SU 18240, 21804-05, 21900), east side of Cuernos de Negros (SU 18788), Banika River area, about 5-6 kms. W. of Dumaguete (SU 18443), Bais (CNHM 61625), Maya- posi environs, about 20 km. W. of Bais (SU 18760, 18910), Ocoy River valley, about 3 km. W. of Palimpinon (SU 18776). CEBU (SNG 19399). Antuwanga area, about 7 km. SW. of Cebu City (SU 18229). MINDANAO. Davao Province: Tagum (CNHM 58378-80), Mount Apo, 850 meters (CNHM 583877), Mount McKinley, 915 meters (CNHM 58874-76). Cotabato Province: Tatayan (MCZ 25771). Zambaonga del Norte Province: Malindang (SU 15968), Masawan (CNHM 96620), Dapitan Peak (SU 23172), 6-10 km. SE of Buena Suerte, west side of Dapitan Peak, about 760 meters (SU 28145, 23159-60, 238162-66). Misamis Occidental Province: 10-16 km. SE of Buena Suerte, west side of Dapitan Peak, 1070-1825 meters (SU 23134-40, 238142-48, 23146, 23148-58, 23188). BASILAN. Port Holland (CAS 60471—holotype of hollandz). ?Jolo (SU 15950). PHILIPPINE ISLANDS (BM 72.10.11.17, 5 unnumbered; NHMW 16706[6], 16814[4]; UMMZ 65483[2]; ZMB 1551/2], 5461[2]). 118 FIELDIANA: ZOOLOGY, VOLUME 49 ?Java (MHNP c2314-7202[2]—syntypes. No data (CNHM 100871; SMNS 8844). Specimens examined.—s817. Calamaria brongersmai new species. Calamaria virgulata (non H. Boie), Boulenger, 1897, Proc. Zool. Soc. London, 1897, p. 225 (part); Schenkel, 1901, Verh. Ges. Basel, 13, p. 164; de Rooij, 1917, Rept. Indo-Austr. Arch., 2, p. 162 (part). Holotype.—Naturhistoriska Museet Goteborg 3229, a male from Penapuan (?Pinapolan) near Luwuk and Biak, near the tip of the eastern peninsula, Celebes. Collected by Walter Kaudern in De- cember, 1919. Paratype.—Naturhistorisches Museum Basel 1707, a female from near Lake Posso, central Celebes. Diagnosis.—Maxillary teeth modified; third and fourth supra- labials entering orbit; preocular present; mental not touching anterior chin shields; paraparietal surrounded by 5 shields and scales; ven- trals with dark squarish spots; no light lateral stripe on any scale row; an oblique dark stripe behind eye. Description.—Rostral wider than high, portion visible from above equals % length of prefrontal suture; prefrontal shorter than frontal, touching first 2 or 8 supralabials; frontal hexagonal, 2 to 214 times width of supraocular, about 34 length of parietal; parietal 1144 times length of prefrontal; paraparietal surrounded by 5 shields and scales; nasal smaller than postocular and oriented obliquely forward; pre- ocular present; neither ocular as high as eye; eye equal to or slightly smaller than eye-mouth distance; 5 supralabials, third and fourth entering orbit, fifth the largest, first, third, and fourth subequal and slightly shorter than second; mental triangular, not touching anterior chin shields; 5 infralabials, first 8 touching anterior chin shields; both pairs of chin shields meeting in midline; 38 gulars in midline between posterior chin shields and first ventral. Body thickness index 0.009 (1 specimen); tail thick, tapering at tip to a compressed point; dorsal scales reduce to 4 rows on tail opposite third to fourth subcaudal anterior to terminal scute. ight modified maxillary teeth (2 specimens). Ventrals 155 in male, 157 in female; subcaudals 18 in male, 11 in female. Total length of male 225 mm., of female 189 mm.; ratio of tail length to total length 0.089 in male, 0.048 in female. INGER AND MARX: THE SNAKE GENUS CALAMARIA 119 Body and tail brown above, each scale with a light network; dark brown spots in center or on edges of scattered scales; some scales of first 2 rows with yellow spots, not forming continuous lateral stripe; head brown above with darker spots; an oblique dark brown stripe from eye to posterior corner of fifth supralabial; other supra- labials with dark sutures; head below yellowish with brown spots on labials and chin shields; ventrals and subcaudals yellowish with scattered dark squarish spots. Comparisons.—All Celebesian species of Calamaria except some virgulata and boesemani have the mental touching the chin shields and differ from brongersmai in this regard. In addition, acutirostris and curta are distinguished from brongersmai by having unmodified maxillary teeth. Calamaria apraeocularis has no preocular. The latter, and virgulata have higher ventral counts (females in excess of 179) than brongersmar. Besides the difference in the mental shield, nuchalis and muellert are distinguished from brongersmai in having the frontal equal to or longer than the parietal and in lacking bold black squares ventrally. Calamaria brongersmaz is sympatric with C. boesemani in eastern Celebes (Fig. 29) where the two differ in ventral coloration (checkered in brongersmaz, immaculate in boesemanz), lateral coloration (white stripe absent on first scale row in brongersmaz, present in boesemanz), eye size (much greater than eye-mouth distance in boesemani, smaller in brongersmaz), orientation of the nostrils (pointed obliquely forward in brongersmaz, laterally in boesemanz), and number of scales sur- rounding paraparietal (5 in brongersmai, 6 in boesemant). One of the specimens (NMB 1707) of brongersmai was identified by Boulenger (1897) and Schenkel (1901) as virgulata (non H. Boie, equals modesta of this paper). It differs from modesta in having lower subcaudal counts and an oblique black stripe behind the eye. The species agreeing with brongersmaz in having the mental separ- ated from the chin shields, 5 supralabials, a preocular, modified maxillary teeth, and the paraparietal surrounded by 5 scales are: bicolor, lateralis, lumholtzi, schlegeli, everettt, modesta, virgulata, and palavanensis. Comparisons with virgulata, modesta, and boesemani are made in the preceding paragraphs. None of these species has a dark, checkered belly or a dark, oblique stripe behind the eye. In addition, bicolor differs from bron- gersmaz in having higher subcaudal counts and a relatively longer tail. Calamaria lateralis and everettt have the reduction to four dorsal scale rows closer to the vent than does brongersmar. Calamaria 120 FIELDIANA: ZOOLOGY, VOLUME 49 e muellert C nuchalis O brongersmat * boeseman! Fic. 29. Distributions of certain species of Calamaria from Celebes. schlegeli and palavanensis have higher subeaudal counts than bron- gersma. Calamaria lumholtz: differs from brongersmai in having a light stripe low on the side, usually no preocular, and the prefrontal as long as the frontal. Distribution.—Central and eastern Celebes (Fig. 29). Calamaria prakkei Lidth de Jeude. Calamaria sumatrana (non Edeling), Sclater, 1891, Jour. Asiatic Soc. Bengal, 60, p. 233; 1891, List Snakes Indian Mus., p. 11. Calamaria prakkei Lidth de Jeude, 1893, Notes Leyden Mus., 15, p. 252— Sandakan Bay, North Borneo; de Rooij, 1917, Rept. Indo-Austr. Arch., INGER AND MARX: THE SNAKE GENUS CALAMARIA 121 p. 156; Werner, 1929, Zool. Jahrb., (Syst.), 57, p. 170; de Haas, 1950, Treubia, 20, p. 573; Marx and Inger, 1955, Fieldiana, Zool., 37, p. 205. Calamaria prakkii, Boulenger, 1894, Cat. Snakes Brit. Mus., 2, p. 337; Bart- lett, 1895, Sarawak Note Book, no. 1, p. 83. Lectotype.—Rijksmuseum van Natuurlijke Historie 4860, a male here selected from the type series of six specimens from Sandakan Bay. Taxonomic notes.—Calamaria prakket is sympatric with sulwensis and probably with everettz in eastern North Borneo. All three of these species have relatively low ventral and relatively high subcaudal counts, though prakkei appears to have more subcaudals. All three are alike in having a light stripe on the first scale row and a network on the dorsal scales. The mental touches the chin shields in prakkez and suluensis but not in everettz. The tails of suluensis and everetti taper gradually from the base whereas that of prakkez is thick and tapers only near the tip. Diagnosis.—Maxillary teeth modified; third and fourth supra- labials entering orbit; preocular present; mental touching anterior chin shields; paraparietal surrounded by 5 shields and scales; sub- caudals of males 30 or more, of females 24 or more; ventrals fewer than 145. Description.—Rostral as wide as high, portion visible from above 14 to 24 length of prefrontal suture; prefrontal °4 length of frontal, touching first 2 supralabials; frontal hexagonal, 1144 to 214 times width of supraocular, about */; length of parietal; parietal 124 times length of prefrontal; paraparietal surrounded by 5 scales and shields; nasal subequal to postocular; preocular present; neither ocu- lar as high as eye; eye greater than eye-mouth distance; 5 supra- labials, third and fourth entering orbit, fifth the largest, other 4 subequal; mental triangular, touching anterior chin shields; 5 infra- labials, first 8 touching anterior chin shields; both pairs of chin shields meeting in midline; 8 gulars in midline between posterior chin shields and first ventral. Body thickness index 0.020-0.027 (2 specimens); tail long and thick, tapering near end to sharp point; dorsal scales reduce to 4 rows on tail opposite seventh to fifteenth subcaudal anterior to ter- minal scute. Hemipenis forked opposite the twelfth subcaudal, retractor be- ginning opposite fourteenth subcaudal; calyces papillate distally (1 specimen). Seven to eight modified maxillary teeth (6 specimens). 122 FIELDIANA: ZOOLOGY, VOLUME 49 Ventrals: males, 126-132 (mean 128.6; N=5); females, 142-144 (N=2). Subcaudals: males, 31-82 (mean 31.4; N=5); females, 24—- 25 (N=Z). Total length: males, 172-245 mm.; females, 230-256 mm. Ratio of tail to total length: males, 0.165-0.167 (mean 0.166; N=8); fe- Wales, 0.096-0.105 (N=2). Color brown above, each dorsal scale above first row with a dark network; scattered mid-dorsal scales with a dark central spot, spots not forming longitudinal stripes; scales of first row yellow in centers forming a distinct longitudinal light stripe running length of body; adjacent edges of ventrals and first scale row dark brown forming a dark stripe; a similar stripe formed along adjacent halves of first and PALAWAN @ everetti x prakke1 @ palavanensis BORNEO Fic. 30. Distributions of Calamaria everetti, C. palavanensis, and C. prakket. INGER AND MARX: THE SNAKE GENUS CALAMARIA 123 second scale rows; in anterior sixth of body a dark stripe formed on adjacent edges of second and third rows; light brown collar about 6 scales behind parietal, collar complete or interrupted mid-dorsally; head brown above with darker spots; lower half or third of supra- labials yellow, sutures brown; head yellow below with brown spots on sutures; ventrals yellow, immaculate except for brown lateral edges; underside of tail yellow, usually a midventral dark stripe. Distribution.—Northeastern Borneo and Singapore (Fig. 30). NORTH BORNEO. Sandakan District: Sandakan Bay (RMNH 4360—lectotype, 10866[5|—paratypes). SINGAPORE (ZSI 13306). Specimens examined.—. Calamaria suluensis Taylor. Figure 31. Calamaria suluensis Taylor, 1922, Snakes Phil. Ids., p. 189—Cagayan Sulu, Philippine Islands; Marx and Inger, 1955, Fieldiana, Zool., 37, p. 190; Leviton, 1963, Proc. Calif. Acad. Sci., 31, p. 388. Calamaria pendleburyi Smith, 1931, Bull. Raffles Mus., no. 5, p. 27—Kiau, Mount Kina Balu, North Borneo; de Haas, 1950, Treubia, 20, p. 572; Marx and Inger, loc. cit., p. 191. Calamaria sumatrana (non Edeling ),Werner, 1909, Mitt. Naturh. Mus. Ham- burg, 26, p. 228. Calamaria grabowski (non Fischer), Marx and Inger, loc. cit., p. 189 (part). Neotype.—Chicago Natural History Museum 76294, a female from Deramakot, North Borneo; here designated. Taylor’s type of swluensis was destroyed in the Bureau of Science, Manila, during World War II. We know of no other specimens from Cagayan Sulu, the type locality. The neotype, collected 180 km. from the type locality is approximately the same size (neotype total TABLE 34.—Comparison of Females of Calamaria suluensis and three topotypic C. pendleburyt. Tail Species Locality Ventrals Subcaudals ratio! suluensis Cagayan Sulu? 154 25 102 sf Sandakan, North Borneo 168 23 92 es Deramakot, North Borneo? 164 26 98 a Bongon, North Borneo 148 25 109 pendleburyt Mount Kina Balu 144-146 17-20 67-84 1JIn terms of thousands of total length. 2 Holotype. Data from Taylor (1922). 3 Neotype CNHM 76294. 124 FIELDIANA: ZOOLOGY, VOLUME 49 237 mm., tail 28 mm.; holotype total 266, tail 27), has very similar counts (Table 34), and is the same sex. Taxonomic notes—We have seen two snakes from northeastern Borneo whose counts (Table 34) agree with those of the type of Cala- maria suluensis (type locality Cagayan Sulu, Philippine Islands) ; they also agree with the original description of swlwenszs in other details. We are assigning these two snakes to sulwensis, which differs from topotypic pendleburyz only in ventral and subcaudal counts and in relative tail length (Table 34). A female from Bongon agrees with topotypic pendleburyz in ventral counts and with sulwensis in subcaudal counts and tail length. Because of this intermediate speci- men and because the only differences between pendleburyi and sulu- ensis (Table 34) involve characters known to undergo remarkable infraspecific geographic variation, we are forced to consider these two forms conspecific. The snakes known as sulwensis (or pendleburyt) and grabowskyi (or baluensis) are very similar in coloration, size of eye, shape of head shields, shape of tail, and arrangement of scales on the chin and throat. In all, the head is dark brown speckled with obscure black spots; the scales of the first row have white centers forming a con- tinuous stripe; the ventrals vary from immaculate yellow to yellow heavily banded with dark brown. In all the frontal is relatively nar- row and long, the eye is large, the nasal is almost as large as the oculars, and the mental touches the anterior chin shields. The sub- caudal counts are moderate to high and the tail begins to taper almost at its base. In the Mount Kina Balu area of North Borneo snakes were col- lected at Saiap, Kiau (915 meters), Kenokok (1000 meters), Tenom- pok (1430 meters), and Bundu Tuhan (1370 meters). The distances between localities are not great: Saiap—Kiau, 20 km.; Kiau-Kenokok, ea. 138 km., Kiau-Bundu Tuhan, 8 km.; Bundu Tuhan-Tenompok, 4km. These distances are small enough to warrant the assumption that gene flow among populations in these localities has been likely at least up to very recent times. The distribution in counts shown in Table 35 can be interpreted in two ways: (1) The snakes of this small area belong to one species having a remarkably wide range of variation in counts. As the two sets of counts do not overlap, one is almost foreed to conclude that polymorphism is involved. As this wide, local range runs against the evidence accumulated for all the other species in the genus, this alter- native lacks support. INGER AND MARX: THE SNAKE GENUS CALAMARIA 125 TABLE 35.—Comparison of Counts and Tail Ratios of Calamaria grabowskyi and C. suluensis From Two Areas of Borneo. Map Ventrals Subcaudals Tail ratio? Localities no.! G S2 G S) G ) MALES Mt. Kina Balt... < Pattee eR ne neh en Hue a10de3ulg a ee BALTRIN WIdIION ‘ON ‘Yabalyos DIUYUYIDD JO OLVkY [ley pur ‘sjepnvoqns ‘sjeUaA Ul UOIZeIIVA dIYdeisoexy)— CH ATAV, 160 1 eg T I Q g 9 6 II 7 a vA I T Pee eer ee ee a ee BAC T se Se OE wR wR wee ew ooulog ) uo ig I Ee ee ae Se ee ee By surg G T T T e G ¢ c aay Sa Tat i ws capa See ice) ee cee eth eryeuns I af T I ve wi Wich Se Pla) Sy ST ae ays a10desuls af T C G ince. (8) 81 L’8t 0c-LI SHTIVN Uva sur yy sjepneoqns ‘8NOS [VUIUIA} WOIJ S[BpNBoqns jo SUlJe} UL UOI}ISO ; "yj Sua] [2307 JO syypuRsNoY} ul ONLY ; 9° S16 O0FG-80d 0° 98T V6T-O8T 9° LVG 09G-LEG I¥G LVC—GES a (2) 102 661 10c-L6I 961 COC-P8T 6 106 G0c-V6l - GIG : 06T Sol OLT-O9T G° 166 LES- L0G (1) L3G LLY G*L8T 681-981 6 O8T S8I-8LI Vol L6OT-68T ued asUuey s[@ijua A mw OD OD NN CO1d bk @: ke) ep @: te: 0” “e: “eco. (6! (0, (0) “ey ome, 6,0) .6, other ene (sdaoun) TN N [eqquag DON OEE RORDEO."OMECrICN C.-C.) G.0) aro (824D]]029) UIOY}10 NY :seqalag @, 18) Tey (6, 16) Oh te! coy jo! fe, 0: uel 0) ei vet (ate). 0 (1suLDaU) 048qey10D ' * (sisuabunoqupz) esueoquiez uleyyNoS :OBVURPUlI ng UBMEIeI er CM ORC NOU COM I Of OTA "Ad (ninhiyov.q) osulog Biyeung [eqquey oi edceY (er ing N O z ™m zzzut 4NOZ 1VIYOSSO4 | VISAV IVAW VIQNI VOINSV |VISAVIVW VIGNI VOIYSV | VISAV IVAW VIGNI VOIYsV] VISAVIVW ~~ VWIQGNI VOINAV 246 INGER AND MARX: THE SNAKE GENUS CALAMARIA 247 no one has suggested that every opisthoglyphous genus represents an independent evolution of grooved maxillary teeth; in fact, such a suggestion would be unreasonable. Probably the African opistho- glyphs represent very few lines of evolution. Bogert’s tentative clas- sification has three clusters of opisthoglyphous genera. The ecological distribution of the African opisthoglyphs suggests adaptive radiation within a continental area similar to that under- gone by other vertebrates (e.g., marsupials in Australia). Five of the genera are arboreal, nine terrestrial, fifteen fossorial, and three of unknown habits. One of Bogert’s clusters of opisthoglyphs (Groups XIII to XVII) includes five arboreal genera, seven terrestrial, and five fossorial. Opisthoglyph genera form only 27% (15 of 55) of the total colu- brid genera of India and Southeast Asia (Smith, 1943) and only 26% (14 of 54) of the total in the Indo-Malayan area (Smith, 19380; de Haas, 1950). The two halves of the Oriental Region together have only 18 opisthoglyph genera. Ten of the 18 belong to the sub- family Homalopsinae and represent a single line of evolution that probably is not closely related to any other opisthoglyph snakes. The remainder includes two terrestrial genera, Telescopus and Psam- mophis, which are widely distributed in Africa and are at best mar- ginal members of the India fauna. Five other genera are arboreal and one is of unknown habits. We conclude from a comparison of the frequency and ecological distribution of the African and Oriental opisthoglyph colubrids that those of Africa, including the fossorial types, represent the results of the radiation there of one or a very few ancestral stocks. From this conclusion follows our belief that the burrowing genera of Africa and those of the Oriental Region reflect the convergence of at least two colubrid stocks that had diverged prior to their adoption of fossorial habits. Whether two or three stocks are involved depends on the relations between the Indian and Indo-Malayan genera. If we temporarily exclude Oxyrhabdium, the basic difference between these two sets of genera lies in the presence of posterior hypapophyses in those from India and Southeast Asia. Though workers in the past have hesi- tated to recognize subfamilies on the basis of this character, many have nonetheless been sufficiently impressed by it to use it for de- limiting groups of related genera. Dunn (1928) distinguished subfamilies on the basis of presence or absence of the posterior hypapophyses. Bogert (1940) used the posterior hypapophyses as the basis of the primary dichotomy of *poj}IWO BVULYDO Useysve pue ‘eAvleyy ‘BAL “eIZBUUNS ‘QOUIOG Ul DLUDWDIDD JO SAT}[VIO] Jo aWIOG *UOLSEY [eJUENO Ul VIIUES Pliqnjoo [elJOsso} poezl[eloeds jo suOIyNqIysIq “G9 “Sy VINVNVTYO @ WAIGEVHYOTIOI &y WNIGIHdOdsWHA § SITOHAOIGI ¥ NOlagvyoanasd ¥ WAIG8VHYOWVTVO M WNIGEWHYAXO Z VIHLAT9 vandldsv ININIWIS SNDYIDO1d WH SITOHdOIDW 1d SIHdOTAX Om @NVI1SV8 v NYMV l¥d OVNVONIW % MeN a ake Sug AVA ® vONynsne QNVITIVHL> ‘ te roar ° OYOAUNIW ives VNIHSOGNI’ rm X vsOw4sogd) j ) (%/ VMY NINO? | of 248 INGER AND MARX: THE SNAKE GENUS CALAMARIA 249 African colubrid genera, stating that “.. . if allowance for variation is made, they do characterize more or less natural groups.” Smith (1948), after first stating (p. 28) that the posterior hypa- pophyses indicate “‘. . . stages in evolution, but not necessarily phylo- genetic relationship,’’ used the character (p. 138) to relate several genera with Natrix. Smith then listed (p. 139) several genera, in- cluding Aspidura, Blythia, Xylophis, Haplocercus, and Plagiopholis, as forming ‘‘a degenerate assemblage, perhaps derived from the pre- vious group”’ (i.e., Natrix, Pseudoxenodon, etc.). Phylogeny is clearly implied. Brongersma’s discovery (1938) that hypapophyses may be pres- ent or absent in Chrysopelea presents the kind of evidence that has weakened confidence in the character. However, occasional intra- generic or even intraspecific variation does not justify our thinking that each occurrence of the character is an instance of independent evolution. Weare following Bogert and Smith, among recent workers, in recognizing interrelationship of genera based on the presence of posterior hypapophyses. The genera of India and Southeast Asia having this character are aquatic, terrestrial, or fossorial, suggesting adaptive radiation of a stock of colubrids. We, therefore, consider that the fossorial genera, Aspidura, Blythia, Plagiopholis, X ylophis, and Haplocercus, which for the sake of convenience, at least, we will refer to as “‘natricine,”’ represent a stock different from that that gave rise to the Indo-Malayan genera. Oxyrhabdium is the only specialized burrowing Indo-Malayan genus having posterior hypapophyses. Probably it is related to the Indian genera (Leviton, 1957) and is a member of the same stock. Besides the character of the vertebrae, it resembles the Indian genera in having a narrow anterior temporal, a moderately high number of maxillary teeth and in certain characters of the skull. The burrowing genera of India and Southeast Asia have distri- butions that are peripheral geographically or ecologically (Fig. 65). Aspidura, Haplocercus, and Xylophis are confined to Ceylon or the southern tip of the Indian peninsula. Blythia and Plagiopholis are confined to the mountainous northeastern fringe of the Oriental Re- gion. Oxyrhabdiwm, the Indo-Malayan genus that appears to be re- lated to the Indian genera, is known only from the Philippine Islands (Fig. 65). Thus, it, too, has a peripheral distribution in the Oriental Region. If we correctly interpret the relations of these natricine genera, as a group they have a disjunct, peripheral distribution. é as 2. Ae ad & Fic. 66. Skulls of Old World fossorial colubrids showing bones reinforcing side of snout. A. Top, Miodon gabonensis (CNHM 58999, length of skull 11 mm.). Middle, Aspidura trachyprocta (CNHM 98767, 11 mm.), maxilla and premaxilla missing. Bottom, Oxyrhabdium modestum (CNHM 96532, 14 mm.), premaxilla missing. INGER AND MARX: THE SNAKE GENUS CALAMARIA 251 Fic. 66. B. Top, Calamaria gervaisti (CNHM 15025, 7 mm.), maxilla dis- placed. Bottom, Pseudorabdion longiceps (CNHM 71596, 8 mm.). The group comprising the Indo-Malayan burrowing genera (ex- cluding Oxyrhabdium), is probably monophyletic or at least formed of closely related stocks. This group, which we will refer to as “‘cala- marine,” has a central, continuous range (Fig. 65). The distribution of these two groups of Oriental genera suggests that the calamarines form a relatively recently evolved fossorial group that is more suc- cessful and has forced the older, natricine burrowers to the ecological and geographical periphery of the Oriental Region. The anterior portions of the skulls of these burrowing colubrids support the division into three phylogenetic groups. The lateral sur- face of the snout in Calamaria (Fig. 66B) is protected and strength- ened by a bony wall formed by the greatly expanded lateral portion of the septomaxilla (Marx and Inger, 1955, p. 169). The four species of Calamaria examined (gervaisi, leucogaster, schlegeli, and septentrio- nalis) have squarish prefrontals behind the enlarged septomaxillae. A single skull of Idiopholis everettt (CNHM 1203877) and one of Pseu- dorabdion longiceps (CNHM 71596) also have squarish prefrontals and expanded septomaxillae forming bony lateral walls to the snouts (Fig. 66B). FIELDIANA: ZOOLOGY, VOLUME 49 bo Or bo In Aspidura trachyprocta (CNHM 98767) and Oxyrhabdium mo- destum (CNHM 965382) the lateral portion of the septomaxilla is rela- tively small, but the prefrontal has a long anterior projection (Fig. 66A) that tends to form a lateral bony wall of the snout. In the African burrowers, Calamelaps unicolor (CNHM 74822), Chilorhinophis carpentert (CNHM 81022), Miodon gabonensis (CNHM 58999), and Aparallactus wernert (CNHM 81061), the prefrontal lacks an anterior projection and the lateral flange of the septomaxilla is relatively small. The side of the snout has a lateral bony wall of varying extent, the wall formed mainly by the nasal and maxilla (Fig. 66A). The nasal in all of these African snakes has a broad dorsal surface and a ventrally projecting lateral flange at least poste- riorly. The maxilla in all four is short and, except in Aparallactus wernert, relatively deep and thick. These burrowing snakes have used three methods of strengthen- ing the side of the snout. The three groups formed on the basis of the bones involved in this strengthening correspond to the groups established on the basis of other characters. Evolutionary trends within the genus Calamaria.—Analysis of evo- lutionary trends in the genus Calamaria depends on the determination of which characters are primitive and which are advanced or derived. Primitive is used here in the restricted sense of referring to what was typical of the stock from which the genus Calamaria arose; derived is applied here to characters that appeared at or after the time the stem of the genus branched off from the rest of the colubrids. This classification of characters can be made only after their distribution in the family Colubridae is analyzed. A trait common to Calamaria and most colubrid genera almost certainly was characteristic of the stock from which Calamaria evolved; such a character is certainly primitive as we use that term here. A character common to Cala- TABLE 59.—Frequency Distribution of Minimum Number of Supralabials in Indo-Malayan Colubrid Genera. Minimum of labials Area LF Cr 8? @G 10" WT ormore hotel No. of genera Bast Indies! 27 Zito) OF GE ams 6 2 2 46 Malaya? see ee Le ea De SO a Ces if 0 32 1 Data from de Rooij (1917) after allowing for uniting Dendrophis with Den- drelaphis and Simotes with Oligodon. * Data from Tweedie (1953). pbiipd yaoi A eeoneeh iG todos iad nee M logins M09 GuaaT IniWr2sioeg2 ti OG2 eC BBA YEG OF cht Mk Cho Spey Ih Oka @e@ x; ea 7 he, h : ; ®) ta "eo 52 ‘ yy i » A Poh : y ne haw S bem ; re we *. . * Bin en pes pie CEN ee UO er em ed Be Oo ot ed er Se 5 ot ag ms: BE ‘ 8.8 8.8 é.28 a.@r 0.5 0.8 8.88 8.er @.7 q g.81 8.8% 2.8L 8.8% @.0 0.08 1.08 &.@8 S.IE TTS £6 q @.88 6.18 {.€8 O.88 0.28 3.08 8.2 j b.@ @.38 $.8b 0.86 8.78 6.22 T.8h fe 1 S.3L @.08 $88 8.88 S.dr 8.8% 9.86 6.68 8.8 P eer BBE F.@r 8.88 6.76 8.28 0.48 G.1d 8.06. 3.2 4 _. @.0b V.88 €.83 8.88 0.08 BES I.48 6.98 6.08 S.88 9.6 ; 3.01 8.8h &.08 €.08 V.Ab 2.86 S.88 8.88 F.08 I.é8 9,08 TV. BIT r.8E A.BE B.88 VTA 8.88 €8.7S 9.@8 b.16 9.7E S.86 8.88 8. MMi TLAST 8.Sh O. ID €.9b O.Sb 0.08 $8.88 @.06 B.Eb T.8e S88 YF BSD @1YSS TiSS! 8.86 1.08 B.2E @.98 O.8b S.TE 0.68 8.88 2.08 I. > AGG GBS iTS agVS: .88 6.68 8.08 8.88 2.0b 6.Eb S.VE 8.86 ¢ £3 MESS LTE OOg: Seame V.er 8.88 B.SS TV. 8.08 S.be 0.86 3. IS M2SI7 BiGes SHON SpGai O1OS? @.8E 9.18 O.B6 G.Ob 3.SE B.S 0 I PBST Spee MES, EO dy) OBS b.08 CCL 6.38 F.08 S.38 6. 3 eM or ti , By = 8.28 @.86 8.@8 8. : 2i6 tee Ss. 3.88 8.88 &. 2 Bby E.R -b. o ® FIELDIANA: ZOOLOGY, VOLUME 49 bo Or bo In Aspidura trachyprocta (CNHM 98767) and Oxyrhabdiwm mo- destum (CNHM 96582) the lateral portion of the septomaxilla is rela- tively small, but the prefrontal has a long anterior projection (Fig. 66A) that tends to form a lateral bony wall of the snout. In the African burrowers, Calamelaps unicolor (CNHM 74822), Chilorhinophis carpentert (CNHM 81022), Miodon gabonensis (CNHM 58999), and Aparallactus wernert (CNHM 81061), the prefrontal lacks an anterior projection and the lateral flange of the septomaxilla is relatively small. The side of the snout has a lateral bony wall of varying extent, the wall formed mainly by the nasal and maxilla (Fig. 66A). The nasal in all of these African snakes has a broad dorsal surface and a ventrally projecting lateral flange at least poste- riorly. The maxilla in all four is short and, except in Aparallactus wernert, relatively deep and thick. These burrowing snakes have used three methods of strengthen- ing the side of the snout. The three groups formed on the basis of the bones involved in this strengthening correspond to the groups established on the basis of other characters. Hvolutionary trends within the genus Calamaria.—Analysis of evo- lutionary trends in the genus Calamarza depends on the determination of which characters are primitive and which are advanced or derived. Primitive is used here in the restricted sense of referring to what was typical of the stock from which the genus Calamaria arose; derived is applied here to characters that appeared at or after the time the stem of the genus branched off from the rest of the colubrids. This classification of characters can be made only after their distribution in the family Colubridae is analyzed. Aas “oe Is aoe $§ Se nee 222 casss & RESE enn atenicalenloo tare aesasr 8 e2eees shame . etsgs 5 SERKkES a2eeerh= | eeisseds 2 SSSeh8sea Qeaaatads eeoueg 3 8 222858858 enensrnaaa _ Serweenrsoo F S8588a25598 Haamenaanes _ aeeesnesnes 3 BSSasaseaees utenneanaer- , thoaornxanmowe= ° SHSRSSEESS55 noennernnseans asorsoia seas 3 S2nese¢2eseee 7 namenencHncssme eeaecsrosnageias & S2SSRBReSZSARR ereenanrnenannean Bt 1818 vy ea/el C2! c9/e9 BIE co SrSsentaduesans RAASSSESTRSLSRSR nan eneontaantresn FOO COUGH OC CIC hestnassesessana + RASSLSSARBSSARSS ™ Demin arnennanen Seen staeeseaserer 4S SZAsSSS5S8255S5545 Rateancaneeeraterta 5 KeasesrManeworisansean 2 SSeszeeasseerssssess enonnreeanartconenne Ngacsornsgansnwsevsews SASSRZASSSSSSSTSSSSR CONE SHANK ONDA AANAIO easaaseneseeeceaesas F HASSSSSASTSSSSCSSSRE menwarnetasaoenanceda LC Noress+egn8®rnwenenasonntsesag HN S2RSSARRSeSseeeegesene Reenteenetentechadtase 4 Adaseectanisssseseessese A SF RSSASSSSASBSSSESSSRSS RHoenwunaenmennanesonerae _ Sieaeeasangeceseuessera F SZSASSAZSARSSESSRSSSS4 S52 Srnaeneneanarawmanxzennase 6 coyrn RoOraAoh IBOmMONM SOD a SSRSRSSRSASHSSASRSERSASAS meter nrandhadtawnethsoqetas SASAGTMEMAKCHSOSHANHSwBOwIMa GS SYSSSsnsssssleagsss2esssss auancancnnaane east eonn 4 Anedgnrnrovnnroeo’S ereonvsorsts = SRZSSS8RSSeSE5ss SSesassces naaaendnanatens Hnoedtees hesneesssnssrace astsssers 7 RERSSSSSR8SE8ER SSSR88E258 tonntwornneeiaue conan 6 eooowreny awvcoz co omen an hn S2SSSS 5528525585 253855 warnowncannaaated neonnen is Nedgodrassseraguess aseqea 7 ASAASSASAAARARSRSS SASS55 anleerrsencarera4ee waenanwee _ Sneernageesare sess Sseesssas = SHSSSRRSRSSSRSK S489 SSCesssess wnHONaANNtTANBrOANTEMEMHaArHARTA= 3 SSguGoeSnagtesssgaeesesnsesstess F RSRSASASSSSSSSASSSSSESRSGSAAASSE Swmonmomreronammacannrnenrtashaas 4 aNadbaesestanassssoesasenesseregraeas 4 SSUSSSSSSSSSSS4SSSSSSSSSSRnssae Bem NSoemuddhanarennnUnnonweuaerera _ BN WR BFNHEKASSEI GSH Una e sees ANSSSIASSSSAESERSSSSIGSLSERZRAISLS DOD HKEENHNSHATHATNHNeM|ETMAVISTIEMSE 4S esanressssaseseaesserscsasaegesasias F ZAMNSSSSHSSSLASASASHSASITSSSSRSASISS AD OCMAMAN ETOH AL NMCOHANMTENNET HEL SHS AMWmArFSoSwowANroMrnorers poll >t AMON YH tae Mona a RSASSRSARSRSASARSASASSSSSSSSSSGAARSSS RNSKA CHAM NNT AHK HI MOMoeerIAAAnEedK SHUG SUAS OHHSoSHIS MBSA sarSseseESerss @ SARALSAASASAASABRSAAAAASSASSISSSASSAB WER ATA HE AANDEARAE ACH AARNE AANADHATIWWA AnAnowownwenna — Netnre2n wv oon MB NYM RB he oOmt TOS HN = SESSSSRSSASRNARRASSASRAASHRSSSESSSSSAS enh annaneehansooanhnexeqercnanoennaiaan CONCK SU HSK Se Sa SaagerentaussnegerSesuren © NENSHASHSASSASASARSSSSARSSSASSSESSBIAS CAPE MWANTHHAWSHUTAMTHADERLERANROTER ANAS oMUOrMngaINg Qoaz WOM DHE OSOERBANLSAAO HOH ie Sr oh = . SAaSage RSRSSSSASSSRARSZSASSSSSHRISAS wens enen seqaenanwquraretnadsoesnecenn ener mtmonn on ag DOFPODOHOHNHBSHHAONH DMB a SASRASSSSHSSAASRASRASASISSSSRSSRSEFSBSSESS WORARAMANDANHHNHOOUTAANTTAETHHENAATABROSNITIG Arann BABDBNONMFOnMDN DaetyOHoOAMDF HTH SHE DSOTe HH HAASAN KARARSRASARSSARSFARSSSSASTSLSESERSEIAS NOSE HHAOHBOEHNASHRAOAMRABTOOAMHDEOINTOMANKE AHA 9109 Be tie ot elit Rieti iene te 3 Bie /A8 19060 to las er ina one alone aaa SxosnseseosgresSusssdareearansgscesasascsess SSAAZSESSSAAASASASARAASSASSSRSSISSSSHRSTESES ANCOADAAHMATOADVAAMMMHONSTANKHHONNSAMIATDT || moro 2 QAMWnar he ao moenowrstoo TFHMHONOAWDAWMAONOWO ‘es SSRZRASSANANSARRSSRASARSSSSASSSSSRSSSGSSS : 2 2 3 -] 2 as 2 3 Bos ~ st i s: ares 5 g Bis ers 2 3 BOK ahs = : @ 5 28 2 Ss sief Sie lee —+lumholtzi \____________ bicolor \ po llentass NN) 4+ Jumbricoidea | ie +g$riswoldr ——_____+albiventer ——+brongersmai a L_yoloensis | L____1 bhi torques | eres | q | — gervaisi | — / + suluensis / = Lpranker +grabowskyi everetti _i————_ palavanensis | _ opmuelleri — -modesta F _-battersbyi | | aN +forcarti | —tvirgulata | Ly - +sumatrana oS | ~s\ -nuchalis | | ae —+margaritophora ———- doderleini | | 2 +crassa | | a ae -erselti | =a aD StrUsa | — { . L Soares er | _______| borneensis \ a ee ee! [ Iinnaet | . . -javanica | +buchi +pavimentata +septentrionalis | low gs L | i N L 1 i bh w Ww ine) Le) = ame Nn [e) oO oO UN oO On oO Horizontal connections give level of Fuller explanation in Appendix A. Dashed lines indicate alternate connections. Phenetic relations among species of Calamaria having modified maxillary teeth. FIG. 67, mean character differences of species concerned. @ mueller e gervaisi A bitorques M joloensis Fic. 68. Distributions of Calamaria bitorques, C. gervaisi, C. joloensis, and C. muelleri. Two localities of gervaisi and bitorques from extreme southeastern Luzon are not shown. 260 INGER AND MARX: THE SNAKE GENUS CALAMARIA 261 The last three groups in Table 61 have widely disjunct distribu- tions. Although the possibility of extinction of related forms in in- tervening areas cannot be eliminated, parallelism within a genus of so many forms seems to us to be a more probable explanation of these phenetic assocations. Furthermore, as in two of these groups com- parisons were limited to 9 characters instead of the usual 14 we do not have the same degree of confidence in the phenetic resemblance. Between mean character differences of 15 and 20, species clusters formed are: schlegeli—Malaya, Sumatra, Borneo, Java bicolor group—Java, Borneo javanica—Java, Billiton melanota group—Java, Borneo, Bangka pavimentata—Burma, southern China, Indo-China septentrionalis—eastern China, northern Indo-China lumbricoidea group suluensis group muellert group sumatrana—Sumatra virgulata—Sumatra, Borneo, Java, Celebes, southern Philippines alidae—Sumatra apraeocularis—Celebes The first three of these clusters form coherent geographic units. Probably they are phylogenetic as well as phenetic groups. The fourth in this tabulation, a cluster of groups, is also geographically coherent and is probably a phylogenetic grouping. To this enlarged lumbrv- coidea group the grabowskyi group may be added; though the con- nection is shown by Fig. 67 at a mean character difference of 25, the similarity of the grabowskyi and sulwensis groups, considered alone, is at a higher level as indicated by the dashed line in the dendrogram. The alidae-apraeocularis cluster is disjunct geographically and may represent another case of parallelism. Although virgulata and sumatrana are similar and have overlapping ranges, we hesitate to call this a patristic group because the taxonomy of virgulata is itself uncertain (p. 187). The highest mean character difference joining species we believe are closely related is 22.5, the level of similarity of low7z and gracillima. Several species are not closely related phenetically to any group. Calamaria modesta is similar to palavanensis at a mean character difference of 20, and to the grabowskyi group, of which palavanensis is part, at a value of 27. Calamaria battersbyz, for which only 8 char- 262 FIELDIANA: ZOOLOGY, VOLUME 49 acters could be used, ties in to the sumatrana-virgulata group at a value of 29. Calamaria buch joins the melanota group at the level of 25. The muellert group is apparently the most generalized of those listed in Table 61 and in the tabulation above. The average devia- tion, calculated from Table 60, of the muellerz cluster from all other species is 27.5. The average deviation of other clusters are: mar- garitophora—30.5; lumbricordea—81.1; eiselti—31.7; melanota—32.0; boesemani—32.8; grabowskyi—83.4; sumatrana—33.4; suluensis— 34.2; bicolor—385.6; pavimentata—37.9; rebentischi—41.7. The fact that the mueller group has character values closer to the average for these 44 species than any other group does not prove that it is the stem group from which others arose. We think it far more likely that the lumbricoidea group represents the stem population because it occupies the heart of the generic range, it is not specialized (see low average deviation, above), and its variability demonstrates that the genetic possibilities for speciation are present. These topics are expanded below. Mean character differences were calculated for the six species having unmodified maxillary teeth, using the same characters and the same procedures (Table 62). Only C. leuwcogaster and ulmeri are closely similar to one another. None of the species having four supralabials forms part of a clus- ter with species having five supralabials. These more specialized species form three separate groups (shown at the right of Fig. 67) that join one another at high mean character differences (80). The measure of phenetic resemblance thus seems to bear out our state- ment (p. 257) that supralabial counts were reduced to four independ- ently in several lines. There is also a strong tendency for species having the paraparietal surrounded by six scales to form clusters from which those having five scales around the paraparietal are excluded. For example, in Table 61 the sixth, seventh, ninth, and tenth clusters consist of spe- cies having the paraparietal surrounded by six scales; those in the other groups all have paraparietals encircled by five scales. Some of the clusters of species on p. 261 have one state or the other of this character, but several of the species in other clusters (e.g., pavimen- tata, schlegeli, virgulata) are variable in this respect. Sharp dichotomies among species are not based on any other character. Species lacking a preocular (e.g., 7avanica) may form part of a cluster with species having preoculars. Species having the mental INGER AND MARX: THE SNAKE GENUS CALAMARIA 263 shield touching the anterior chin shields (e.g., grabowskyz) may form a cluster with species having the mental and chin shields separated. The clusters of similar species do not show consistent associations of characters that are not necessary correlates (cf. Tables 61, 64). For example, the lumbricoidea, muelleri, and suluensis groups have five supralabials; only the last two have reticulate patterns in the dorsal scales, and only the last has a light, dark-edged stripe on the first scale row (Table 64). The lumbricoidea and melanota groups are alike in the presence of preoculars and in the relation of mental and chin shields but differ in supralabial counts and pattern of dorsal scales. Other examples of this nature are apparent from comparison of Tables 61 and 64. This almost exhaustive recombination of char- acters, to which we have referred above (p. 257), suggests an evolu- tionary sequence. Patterns of evolution.—Dobzhansky (1955) cites numerous studies of Drosophila showing that many populations, phenotypically rather uniform, in fact carry a large store of genotypic variability. As many as 40 per cent of the chromosomes of some species are hetero- zygous for deleterious recessives. Ecologically specialized forms of Drosophila (e.g., D. prosaltans) often have less genotypic variability than their more widely distributed, less specialized relatives (Dob- zhansky and Spassky, 1954; Carson, 1959). On the basis of such evidence Dobzhansky suggested the “balance hypothesis” of the origin and maintenance of adaptive norms. According to this hy- pothesis successful adaptation depends on an array of genotypes heterozygous for gene alleles, gene combinations, and chromosome arrangements. As an alternative Dobzhansky proposed the “‘classical hypothe- sis’’ which involves the gradual substitution and fixation of favorable mutants. This scheme is at least implicit in much taxonomic litera- ture (e.g., p. 256 of this paper; Gans, 1959, pp. 167-173). According TABLE 62.— Matrix of Mean Character Differences Among Pairs of Species of Calamaria Having Unmodified Maxillary Teeth. Species number No. Species it Z 3 4 5 1 leucogaster 2 ulmeri 9.6 3 ~=lautensis 30.1 29.6 4 curta 48.6 30.8 41.0 5 acutirostris 53.6 37.5 46.6 33.9 6 schmidti 47.0 62.5 60.4 52.8 DDO 264 FIELDIANA: ZOOLOGY, VOLUME 49 to this hypothesis, most individuals will be homozygous for most loci and most chromosome arrangements. Carson (1959) suggests this is a common result of homoselection in specialized marginal populations. The success of the ‘“‘balance’’ system involves more than heterotic effects in individuals. Wright’s (1949b) model of adaptive peaks of successful gene combinations to which the gene pools of populations return after temporary disturbance or relaxation of selection pres- sure is predicated on heterozygosity in an organized species system. This heterozygous system has the capacity to withstand temporary ¢ D E F G H J SPECIES CONTEMPORARY A>A’ AA AA Fic. 69. ‘Classical’ (left) and ‘‘balance’’ hypotheses (right) applied to one locus. Each bifurcation represents development of reproductive isolation. See text for fuller explanation. (After Throckmorton, 1962.) changes in the environment without destruction of the basic adapta- tions and thus has genetic homeostasis (Lerner, 1954). Throckmorton (1962) outlines a phylogeny in Drosophila based on the “‘balance hypothesis.’”’ A heterozygous stem population gives rise to populations that express completely all of the possibilities for diversification (Fig. 69). Under this hypothesis it is possible for two species that have identical genetic compositions, at least at this par- ticular locus, to be derived from separate populations unlike either. By extending the diagram (as in Fig. 70) so that it includes another locus that, for the sake of discussion, modifies a second character, one may arrive at a set of species expressing most of the possible combinations of states of these two characters. INGER AND MARX: THE SNAKE GENUS CALAMARIA 265 The two contemporary species, K and N (Fig. 70), having iden- tical combinations of these characters are not derivatives of the same ancestral population. In fact, each is more closely related phylo- genetically to species having different combinations of states of these characters. The similarity of K and N is the result of parallelism, which we believe accounts for many of the similarities between spe- cies of Calamaria in particular characters and occasionally in several characters. This outline, if it does in fact describe the phylogenetic history of any group, depends on the existence for some time of populations K L M N SPECIES A’ AN’ B’B’ AABB A A‘BB’ N A’ B’B’ CONTEMPORARY A A’ BB’ AABB Fic. 70. Balance hypothesis applied to two loci. See text for explanation. having heterozygous genotypes. Have such populations existed in the genus Calamaria? At the very least such species populations exist now. Calamaria lumbricoidea is the best example. It is variable over narrow or wide geographic areas in dorsal and ventral coloration, type of hemipenis, number of gulars, position of reduction on the tail, ventral and subcaudal counts, and in ontogenesis of dorsal pigmenta- tion (pp. 82 ff.). Two related species, griswoldi and albiventer, have modifications of the dorsal striping found in some populations of lumbricoidea. A third related species, hilleniusi, lacks stripes but may have yellow transverse bars, which are found in the young of most populations of lumbricoidea. Subcaudal counts are low in gris- 266 FIELDIANA: ZOOLOGY, VOLUME 49 woldi, moderate in albiventer; ventral counts are low in albiventer, moderate to high in griswold?. Reduction of the dorsal scales on the tail in albiventer occurs in the middle and in griswoldi in the upper part of the range of variation of lwmbricordea. We are far from certain that lumbricoidea (or its antecedent pop- ulation) is in fact the ancestor of albiventer, griswoldi, and hillenius?. Fic. 71. Pleistocene relations of land areas (after Smit-Sibinga in de Beau- fort, 1951). Present land areas stippled. Pleistocene coast lines and inland drain- ages indicated by solid lines. A. First interglacial period. B. Last glacial period. However, such a relationship is a distinet possibility, for lumbricoidea does have a great store of variability and its geographic range spans those of the other three. During the Pleistocene when the area of Sundaland was being fragmented and reunited at intervals (Fig. 71; see also below), a variable population could have split off geographi- cally isolated fragments in which certain variations became fixed either because of inbreeding and random phenomena such as acci- dents of sampling or because of strong selection for homozygosity (Carson, 1959). INGER AND MARX: THE SNAKE GENUS CALAMARIA 267 Wright (1982) has concluded that ‘*. . . evolution depends on a certain balance among its factors. There must be gene mutation, but an excessive rate gives an array of freaks, not evolution; there must be selection but too severe a process destroys the field of vari- ability and thus the basis for further advance; prevalence of local inbreeding within a species has extremely important evolutionary consequences but too close inbreeding leads merely to extinction. A certain amount of crossbreeding is favorable but not too much.’’! Almost all these requirements are satisfied by lumbricoidea. Al- though we know nothing about its genotype directly, phenotypic variation within demes in ventral and subcaudal counts (Table 23) and in coloration (Tables 21, 22) is strong evidence of genotypic vari- ation. The prevalence of one type of dorsal coloration over wide areas shows that, despite some variation, mutation has not run wild. The narrow range of ventral counts within and the differences be- tween local populations (Tables 23, 24) indicate that local inbreeding has taken place. Yet the wide geographic range of particular dorsal patterns is evidence that some crossbreeding has also occurred. The events of speciation in the lumbricoidea groun (Fig. 72) can be visualized along the following lines. In one fragment (albiventer), isolated in Malaya during a period of high sea levels, a uniformly light venter became fixed, dorsal striping of one sort became constant, ventral counts were restricted to a low range and subcaudal counts to a moderate one. Possibly in the same interval a fragment (gris- wold), isolated in northern Borneo, lost all trace of dark ventral bars, the dorsal pattern of light lines became fixed, ventral counts became restricted to high levels and subcaudal counts to low ones. At the same time the remainder of the species population in Su- matra retained most of its original variability in these characters. Dorsal striping between all adjacent scale rows may have been char- acteristic of the Sumatran population at this time. Then when continuity between land masses was re-established during the next eustatic lowering of sea level, the Malayan isolate, now a reproductively isolated species, spread westward into Sumatra. The northern Bornean isolate apparently had become specifically adapted to moderately high elevations, a type of specialization Car- son (1959) visualizes as one of the consequences of homoselection and and random drift in small marginal populations. Such a specializa- 1 Although the statement quoted is 30 years old, the idea of a balance among forces is still widely held, as can be seen in Wright (1955), and in Mayr’s (1968, p. 2) analysis of the “synthetic theory”’ of evolution. “paproaiqun) *Q pue ‘isnruanry “OQ ‘ypjomsib “Dd ‘1aquaarqyD DILDWDIDD JO SUOTINGIISIG ‘ZL “DIY Ol Oll OOl eee ve - : AS atl Oy) ey, ee VE CS ~_-7 TdJUIATG]e GB IPJOMSI1I9 © ESPNS eae 2 , e Poplooliquiny= @ ( Cae (ae a on AS =) YLVYANS © \ WZ nl? e50 VN VONIW 268 INGER AND MARX: THE SNAKE GENUS CALAMARIA 269 tion could have prevented the dispersal of gr¢swoldi through the low- lands of the newly re-formed Sundaland. The striped but otherwise variable Sumatran population at this time spread eastward into Malaya, into Borneo and Java, and probably even reached the Phil- ippine Islands. When these populations were isolated from one another during the next rise in sea level, a new, more vigorous epigenotype, having as one of its phenotypic manifestations a uniform dark dorsum in adults, gradually replaced the old, ‘‘striped’’ lwmbricoidea genotype in Borneo. Following the next lowering of sea level and the subse- quent reunion of Sundaland, this new vigorous genotype spread into eastern Sumatra, slowly replacing the ‘“‘striped’’ genotype until the latter became restricted to the western fringes of Sumatra. The new genotype also reached the lowlands of western Malaya and similarly replaced the striped variant until the latter came to be restricted to its present range in the mountains of central Malaya. The genotype associated with uniform dorsal coloration may have spread to Java at this time or the phenotype may have developed there independ- ently. Apparently, this genotype did not reach the Philippines. Thus, the new, more vigorous genotype in some areas replaced the older one completely and in others pushed the older one to the geo- graphical or ecological limits of the species range. During the interval of land fragmentation following the Pleisto- cene, differentiation has continued. Yellow dorsal rings and dark ventral bars have been lost in scattered populations (p. 82), and the ranges of variation of ventrals and subcaudals have continued to dif- ferentiate from population to population. On the basis of the few specimens known, speculation concerning C. hilleniusi is perhaps not justified though it is tempting to suggest that its history parallels that of griswoldt. This picture of speciation in the lwmbricoidea group may be over- simplified; it is certainly speculative. Nevertheless, it fits the known Pleistocene history of Sundaland (Molengraaff and Weber, 1921; de Beaufort, 1951). It also resembles closely Wright’s theoretical picture of the population structure most likely to evolve continuously (Wright, 1949a, pp. 381-882): a species population consisting of semi- isolated local populations in which the varying effects of local selec- tion and inbreeding may produce new favorable gene combinations that will spread through adjacent demes as opportunity presents itself only to be replaced by later, still more favorable combinations. In several of the postulated local populations that became geograph- 270 FIELDIANA: ZOOLOGY, VOLUME 49 ically isolated temporarily, local selection or inbreeding carried the process of differentiation too far. These populations became repro- ductively isolated, recognizably distinct species (albiventer, griswoldz, and, probably, hzlleniusz). Their internal uniformity has effectively cut them off from the continuing process of evolutionary advance of which the parent population, lumbricoidea, still seems capable. The evolutionary potential of lwmbricoidea is probably not con- fined to speciation, that is, to the production of new species that are morphologically distinct but essentially in the same adaptive mold. Calamaria lumbricoidea is generalized enough that it could evolve in any number of adaptive directions. It could, for example, become more slender and shorter-tailed and, hence, approach the form of virgulata. It could reduce the number of supralabials, shorten the tail, and thus develop in the direction of the form of Calamaria sep- tentrionalis. Calamaria schlegeli may be another species that, like lwmbricozdea, has been evolving actively and providing the source of other, repro- ductively isolated, populations. Like lwmbricoidea, schlegeli has an extensive geographic range encompassing those of the postulated off- shoots and morphological variability that can be arranged hierarchi- cally from local individual to macro-geographic variation. The preocular scale (p. 155, Tables 42, 43) may be present or ab- sent in a geographically restricted sample (e.g., western Sumatra), and when present in such a sample it may be large or small. The variation of the preocular may also be severely limited over a wide geographic range, asin Java. The same patterns of variation appear in head coloration (Table 44). Ventral and subcaudal counts, how- ever, vary from deme to deme but show no macro-geographic pat- terns (Table 45). Thus observed variation in schlegelt can be interpreted as the re- sult of a combination of moderate local inbreeding, gene flow, selec- tion, and mutation or recombination—the same combination that we have suggested makes the continued evolution of lumbricoidea very likely. The species in the bicolor cluster (Table 61) probably are branches from the schlegeli stock. Calamaria pavimentata is also subject to local differentiation and individual variation (pp. 215 ff.), the latter in elements of coloration and the former in ventral and subcaudal counts (Table 54). How- ever, what this wide-ranging species lacks that lumbricoidea and schlegeli have is variation at the macro-geographic level. Possibly the largely continental range of pavimentata has not provided the INGER AND MARX: THE SNAKE GENUS CALAMARIA 271 same opportunities for temporary geographic isolation enjoyed by populations of Calamaria lumbricoidea and schlegeli in their largely insular ranges. Another factor that may inhibit the evolution of pavimentata is the environment. The range of this species is close to what is prob- ably the ecological limits of the genus (see below). Although Mayr (1968, p. 546) seems to believe that isolated populations at the eco- logical boundary of a species range represent the situation having the greatest opportunity for ‘“‘adaptive shifts and evolutionary novel- ties,” the species frequency distribution of Calamaria (pp. 272 ff.) suggests the opposite, that the greatest opportunities for this group have been in the center of the generic range. The insular species gervaisz shows individual variation, local dif- ferentiation, and macro-geographic variation. Because variation in gervaist is limited to ventral and subecaudal counts and to minor aspects of coloration, it seems to lack the potentiality for evolution seen in lumbricoidea. Nevertheless, an old isolate from gervaist may have developed into the species bitorques. Some Luzon specimens of gervaisi have two yellow rings on the neck enclosing a darker area (p. 112). Multiplication of these rings, which might have happened in a small, isolated population, would result in the pattern of bitorques which has two to six dark saddles separated by yellowish rings. The Philippine archipelago has been unstable geologically, with periods of submergence and fragmentation of land masses alternating with periods of emergence and land connections. Under these circum- stances so conducive to speciation, it is surprising that more species of Calamaria have not evolved in the Philippine Islands. The only other species having the characteristics we have come to expect in continuously evolving populations is low?. Calamaria lowt has a moderately extensive range from Malaya to Java and Bor- neo, and varies over narrow or wide areas in the number of supra- labials bordering the eye, the extent of spotting on the side of the body, the relation between mental and chin shields, and the numbers of infralabials, ventrals, and subcaudals (pp. 221 ff.). Geographic variation is sufficient for the definition of subspecies. The potenti- ality for continued evolution of low? is probably more limited than that of the lumbricoidea stock, for example, because of its extreme specialization. By virtue of reduction in the orbital region, reduction of supralabial counts, extreme slenderness of the body, and shortness of the tail, low2 probably can evolve only in the direction of increasing fossorial specialization. Some of the special adaptations of low7z, such 212 FIELDIANA: ZOOLOGY, VOLUME 49 as reduction of the orbital region, would have to be reorganized in order to permit low7z to evolve along more generalized lines. The limitation of evolutionary directions does not, however, cut lowz off from the usual processes of speciation. Geographic variation in isolated land masses is evident now. If geographic isolation con- tinues long enough, inherent barriers to interbreeding may develop. Probably gracillama owes its origin from the lowz stock to these processes of geographic variation and isolation. Distribution.—Although two species of Calamaria, septentrionalis and pavimentata, occur as far north as 30° N, the genus is essentially tropical with a pronounced center of distribution in Borneo and Sumatra (Fig. 65, Appendix B). Of the 50 species, 21 are known from Borneo and 16 from Sumatra, thus accounting for 64 per cent of the species (five species common to both islands). In all directions from these two islands, the number of species drops off rapidly. The southern Philippine Islands have 5 species, Celebes 8, Java 9, Bali 1, and the Malay Peninsula 6. Four species occur in Indo-China and China, one of them (pavimentata) entering Formosa and the Riu Kiu Islands. A single species, ceramensis, found in Ceram, represents the southeasternmost extension of the generic range. The northern and western limits of the range are apparently formed by ecological barriers. Low temperatures probably limit dis- persal of Calamarza northward and increasing seasonal aridity may inhibit dispersal westward. The ranges of septentrionalis and pavi- mentata (Fig. 73) almost coincide with the areas of southeastern Asia having at least 25 cm. of rain in the period November through April. The southern boundary of the generic range is a geographic bar- rier. The eastern limits are a combination of ecological and geo- graphical barriers. The area between Celebes and New Guinea has probably been occupied largely by seas throughout the Tertiary and Quaternary (Umbgrove, 1949). Thus a geographical barrier has in- hibited the dispersal of Calamaria eastward from Celebes, and has allowed only one species, ceramensis, to get as far as the Moluccas. The rapid decrease in species of Calamaria eastward from Su- matra in the Sunda chain parallels changes in the distribution of other organisms. Mayr (1945) notes that Sumatra has 440 species of birds, Java 340, Bali 166, and Lombok 119. The decrease is con- spicuous in Java itself, for only 245 species of birds occur in the east- ern half of the island and probably only 170 reach the eastern tip. The numbers of Calamaria in these islands are: Sumatra 16, Java 9, Bali 1, and Lombok 0. Definite localities are available for only five of INGER AND MARX: THE SNAKE GENUS CALAMARIA 273 the Javan species. All five—modesta, virgulata, lumbricoidea, linnaez, and schlegeli—occur in western Java, only the last three in central Java, and only the last two in eastern Java. Calamaria schlegeli is the only species to reach Bali. Mayr (1945) attributes the impover- ishment of the bird fauna within Java to the increasing aridity east- CHINA A Pavimentata 4 yunnanensis ® septentrionalis © buchi Fic. 78. Distributions of Calamaria buchi, C. yunnanensis, C. pavimentata, and C. septentrionalis. Dashed line shows northern and western limits of area having at least 25 em. of rain in the interval Nov. 1—Apr. 30 (after Goode’s School Atlas). ward in the island and notes that lowland rain forest is replaced by drier monsoon forest in eastern Java. The same climatic factor could account for the decrease in species of Calamaria eastward in Java and for the presence of only one species in Bali. The Lesser Sunda Islands ott+++totttettete+ttett++tttete eulyy ‘euly -opu] +ototott+++++++tttttt++to++++teote SpueysT auld -diiyd fottttot++++++tttttt++4++4tttt saqalag +H+++++++++++++++4+444+4t++++e 44 +4+4++++++++++tot++++44++tttt+t+eo++ +OFFFFEEFFEFEFEF FFF FF +++++44444 +4FAFEFEFFFEFFFFFE EEF HE EEE ++ +44 BARR BACL eB1jyeuNnsg ovulog (‘aoueSqe O19Z ‘9DUAIINDOO SdyVOIPUI USIS SN[q) aj wuelie) eu quasqe 7 Se ee ee tra oul @ saeiny ieee 2 eae ee eee peylpowun of fe. 0, te) 0; 10, ‘p <0 ‘el dea) ate ‘elie? eh te . peyIpou “499 |, Pe ee hays. 5 oz cg > ee ee RRO le - Cg < YjeaJ, ee quasqe ‘**“queseid “uaa 7B SSULyIeW YY BIT yIOMJOU JNOYIIM id Sa oe hes yIOMJOU YIM sepeag ‘* peyeiedas "* dulyono} ‘spyalys Ulyo pue [eUeW] Sa a fs ""*y) < sfepneoqns ‘wstydiowip xag ° SL< s[eiqueA ‘WISIYdIOWIp xag y's) ae tap ne Volts, Nie ale: “a piettenne: oMietne C8l < S[BIJUOA ergs "**"@'Q < SaTBaS [BSIOP JO UOTJONpeyy ACEC DICE) ACCC Mee O qyuasqe ‘DILDUWDID SNUAL) BY} JO SUOTIVIIVA [BoIsO[OYdIOPL [Vdloullg Jo uoINqiaysIq dIYydevis0en—'Eeg ATAV,L 274 INGER AND MARX: THE SNAKE GENUS CALAMARIA 275 become increasingly arid eastward. The absence of Calamaria east of of Bali may result from this climatic deterioration and from the age and depth of the straits between Bali and Lombok. Centers of evolution and dispersal.—The numbers of species of Calamaria occurring in different land masses (Appendix B) are not closely correlated with the sizes of those areas. The parts of China and Indo-China included in the range of Calamaria far exceed in size any of the other areas but have very few species. The next two largest areas, Borneo (286,969 square miles) and Sumatra (163,557 square miles), have most of the species. Celebes (69,277 square miles) is considerably larger than Java (48,842 square miles) but has fewer species. The southern Philippine Islands—Mindanao, Jolo, and Palawan—have almost the same area (41,432 square miles) as Java but only half as many species. Searcity of species in southern China can be explained on the basis of unfavorable climate (see above). The small number of spe- cies in the Philippine Islands cannot, however, be explained in eco- logical terms as the climate and vegetation are not very different from those of Borneo or Java. For the same reason, the drop in species number in Celebes probably does not have an ecological basis. If one considers the Tertiary and, especially, the Pleistocene his- tory of the Indo-Australian archipelago, the distribution of species becomes understandable provided one assumes that the centers of evolution and dispersal lie in the Borneo—Sumatra area. If Celebes had been the major center of evolution of Calamaria, one would expect to find many species, most of the morphological types, and at least several species in common with adjacent areas. The number (8) of species of Calamaria in Celebes is small compared to that of neighboring Borneo (21). Most of the morphological types are present (Table 63), but the absence of one—species having four supralabials—requires at least a subsidiary center of morphological radiation elsewhere. The only non-endemic Celebesian species is C. virgulata, which may be a composite of reproductively isolated forms (p. 187). Aside from the questionable taxonomy, virgulata does not provide evidence of a faunal source in Celebes, for it occurs in Mindanao, Palawan, Sumatra, and Java as well as Borneo and Cel- ebes. It could not have spread directly from Celebes to all of these areas, whereas it could have reached them directly from Borneo. On geological and zoogeographical grounds, Celebes is a poor site for the major center of evolution of Calamaria. The Makassar Strait, with channel depths exceeding 2000 meters, has separated Borneo 276 FIELDIANA: ZOOLOGY, VOLUME 49 and Celebes at least since the Eocene (Umbgrove, 1938). Given this circumstance fossorial forest snakes would have great difficulty dis- persing between Borneo and Celebes. If Celebes had been the main center of evolution of Calamaria, only the limited possibilities of acci- dental dispersal could have carried stocks from this source to the main area of the generic range. Consequently, it would be necessary to assume a second center of radiation elsewhere, and in view of the many species in Borneo, the second center would be more important than a Celebes center. The distribution of other snakes is also against locating the main center in Celebes. Celebes has 27 genera of non-marine snakes, 25 of them occurring in Borneo which has 47 genera (de Haas, 1950). Except for two endemics, the genera of Celebes snakes represent the results of sweepstakes migration (Simpson, 1940) across an ocean strait from Borneo. Darlington (1957) shows that the vertebrate fauna of Celebes is composed of groups that have moved in from adjacent areas. Celebes nonetheless has been the site for a minor radiation of Calamaria. Seven of the eight species are endemic and, as noted above, most of the morphological trends taken by the genus appeared in these eight species. Probably no more than three stocks reached Celebes, a stock having unmodified teeth and one or two having modified dentition. The Philippine Islands show little evidence of being an important center of evolution of Calamaria. The Tertiary and Pleistocene his- tory of the Philippines is one of alternating periods of land emergence and subsidence (van Bemmelen, 1949), which should have favored fragmentation of populations and radiation. But the fauna is small (only 6 species) and several striking morphological types are absent (Table 63). Furthermore, the entire amphibian and reptilian fauna of the Philippines appears to have come from other areas and gives no evidence of a major faunal source in the Philippines (Inger, 1954). Indo-China and China are not likely places for the center of radi- ation of Calamaria. They have only four species, lack most of the morphological types, and the contemporary climate is unfavorable. The last obstacle may not be a valid one, for the pre-Pleistocene cli- mate was probably warmer and more humid (Dorf, 1959). If this area had been an important center of radiation, one should expect to find some representatives of less specialized Calamaria (e.g., species having five supralabials or unmodified maxillary teeth). This expec- tation is not fulfilled. INGER AND MARX: THE SNAKE GENUS CALAMARIA 2770 What remains of the range of Calamaria is Sundaland, the land masses of Borneo, Sumatra, Java, and Malaya. Throughout most of the Tertiary and during the glacial intervals of the Pleistocene, these areas (with the possible exception of Java) and much of the intervening South China Sea were areas of land emergence (Umb- grove, 1949). This vast region, which until very recent times was covered with tropical rain forest, has been an ideal area for the evo- lution of Calamaria. The presence of the great majority of species and of all the principal morphological types (Table 63) constitute sufficient evidence for that conclusion. Given the intimate connections among the various parts of Sunda- land, it may be superfluous to search for the center of radiation among them. However, as Java was apparently largely submerged during the Eocene, Miocene, and Pliocene (Umbgrove, 1949, figs. 47, 48, and 55), it probably could not provide the necessary environment over as long a period as Sumatra, Borneo, and Malaya. If Java has been an important area of morphological radiation of Calamaria, it has been so only since the Pliocene. As Java has only nine species of Calamaria and lacks several morphological types (Table 63) including the primitive ones having unmodified maxillary teeth, it is not a likely area of major radiation. Malaya has been a relatively stable land mass through the Terti- ary and Quaternary but has few species (6) of Calamaria, no endemic species, and lacks forms having unmodified maxillary teeth. Even if one assumes that the fauna of Malaya is poorly known—an assump- tion not supported by the large number (100) of species of non- marine snakes already reported from there (Tweedie, 1953)—the Calamaria fauna is not rich enough to suggest that Malaya was a major center of radiation. Sumatra and Borneo remain as the contemporary land masses most likely to have been major centers of radiation or parts of a single major center. Both had large areas above sea level during most of the Tertiary and Quaternary (Umbgrove, 1949) and both have large assemblages of Calamaria. Sumatra has 16 species of which 10 are endemic (if Nias is included as part of a Sumatran area), and Borneo 21, of which 12 are endemic (if Cagayan Sulu may be included as part of Borneo). The Bornean species cover every prin- cipal morphological diversification of the genus (Table 63). The Sumatran assemblage lacks only species having four supralabials. All but one (albiventer) of the non-endemic species of Sumatra occur in Borneo and half of the non-endemics of Borneo are known from 278 FIELDIANA: ZOOLOGY, VOLUME 49 Sumatra. The large proportion of endemics in each island indicates that each has probably been a center of radiation. The preceding discussion has been more or less limited to present day land masses when perhaps it should not have been. Geologic evidence for a much more extensive land area in Sundaland during the Pleistocene (Molengraaff and Weber, 1921) and throughout most of the Tertiary (Umbgrove, 1949) is very strong. As the area between Borneo and Sumatra during periods of emergence would have been lowlands and covered with rain forest, it would have made an ideal habitat for Calamaria. Probably it would be more correct to refer to one large central area of evolution comprising Borneo, Sumatra, and the now sub- merged inter-insular portion of Sundaland. This enlarged center, if at all important, had its main significance to radiation as a unit prior to Pleistocene times. During and after the Pleistocene, because of repeated inundation of the inter-insular region, this enlarged unit did not exist. Evidence from species having wide ranges indicates that Sumatra and Borneo are not only centers of evolution but also centers of dis- persal. Six species of Calamaria have distinctly larger ranges than the others. Two of them, pavimentata and septentrionalis, are re- stricted to the mainland of Asia and islands off the coast of China. The other four—lumbricoidea, schlegeli, virgulata and modesta—occur in Borneo and Sumatra, which are more or less centrally located in their ranges. These species could have spread to Malaya and Java from Sumatra or Borneo. One species, albiventer, is known only from Sumatra and Malaya suggesting dispersal between those two areas. The presence of lowz in Malaya and Borneo suggests dispersal between those areas. The distribution of the uniform color variety of lumbricoidea (Table 21) also supports the latter route of exchange, probably during the Pleis- tocene (p. 267). The uniform color variety of lwmbricoidea occurs in Borneo and Great Natuna among other places (Table 21) and has recently been collected on Tioman Island, 20 miles off the east coast of Malaya (Hendrickson, MS). As this color variety is not known from central or eastern Malaya (Table 21), its occurrence on Natuna and Tioman suggests a dispersal from Borneo westward toward Ma- laya or an origin in a pre-Recent land mass in what is now the South China Sea. Probably lumbricoidea and virgulata spread to the Philippines from Borneo. In fact, all of the Philippine stocks of Calamaria came INGER AND MARX: THE SNAKE GENUS CALAMARIA 219 from Borneo. The Pleistocene drops in sea level caused by the formation of northern glaciers expanded land areas not only within Sundaland, but also between Sundaland and fringing areas. The gaps between the southern islands of the Philippines and Borneo were at least greatly narrowed if not temporarily obliterated. At those intervals the chances for successful dispersal from Borneo to the Phil- ippines were increased considerably. Celebes can probably be ig- nored as a source for the Philippine Calamaria because the geologic evidence for land connection with the Philippines is weak (Umbgrove, 1938; van Bemmelen, 1949) and because the Celebes fauna is itself relatively impoverished. The one species common to the Philip- pines and Celebes is the dubious C. virgulata; it cannot be used as unequivocal evidence for Celebes—Philippines exchange since it could as well have reached both of those areas from Borneo. The mainland of Asia is not a likely source of the Philippine stocks because of the great distance and because the forms of Calamaria present in the closest parts of the mainland all have four supralabials, a character absent in the Philippine Calamaria. There is no evidence from other reptiles and amphibians of direct dispersal between southeastern Asia and the Philippines (Inger, 1954). Because of the great age of the Makassar Strait between Borneo and Celebes (p. 275), species of Calamaria (or their ancestral stocks) could only reach Celebes by some means of accidental dispersal. The size of the Bornean fauna and its proximity to Celebes make it the logical source of the Celebes assemblage of Calamaria. Probably two or three stocks of Calamaria invaded Celebes (see above). Time of dispersal and speciation.—Twenty-two species of Cala- maria are endemic to either Borneo or Sumatra; only five species are known from both islands. Even granting that additional collecting will increase the number of species common to both islands, the ratio of endemic to common species probably will not change much. To assume otherwise implies that endemics are easier to find than non- endemics. If the collections available to us are any measure, the reverse is true. The average number of specimens seen of the 27 species found in either Borneo or Sumatra is 12; the average number seen of the five common to Borneo and Sumatra is 51.! The preponderance of endemics in Borneo and Sumatra suggests that many of these species became distinct in the Pleistocene. Al- though we have no evidence for the date, it is reasonable to place the origin of the genus Calamaria at least in the Pliocene. This date 1 This average is based only on specimens collected in Borneo and Sumatra. 280 FIELDIANA: ZOOLOGY, VOLUME 49 makes the genus at least 5-13 million years old, depending on whether the origin is at the beginning or middle of the Pliocene. If the en- demic species of Calamaria in these two islands date from the Plio- cene, they are several million years old. Their failure to disperse between these islands in that length of time implies a slow rate of dispersal. Some evidence concerning rates of dispersal of Calamaria can be obtained from the presence of species on certain islands. Nias is covered with Pliocene marine sediments (Umbgrove, 1949). Hence its three species, C. lumbricoidea, abstrusa, and forcartt, could not have arrived there prior to the Pleistocene. As all three occur in Sumatra, which had a large land surface throughout the Tertiary and Pleistocene, probably these three species invaded Nias from Sumatra during the Pleistocene. At present, relatively shallow water, i.e., less than 200 meters deep, extends from the northern tip of Nias to Sumatra (U.S. Hydrographic Office Chart 3122). This sixty-mile gap was at least very much narrowed if not completely obliterated several times during Pleistocene glacial stages. Though estimates of the length of the separate glacial stages vary, apparently none of these stages exceeded 100,000 years (Charlesworth, 1957; Dunbar, 1949). As some time was required for the ice to reach its maximum extent and, hence, to lower sea level to the maximum and as some time was required for forest to invade the new land, the maximum time available for overland migration of Calamaria from Sumatra to Nias was less than 100,000 years. If three species, including widely distributed as well as restricted ones and representing 19 per cent (8/16) of the Sumatran fauna, could move 60 miles in less than 100,000 years, it is unreasonable to assume that only 5 of 32 species (the total Calamaria fauna of Borneo and Sumatra) could disperse across an area three and one-half times as wide as the Nias channel in more than one million years of pre- Pleistocene time. Consequently, we think that the argument for a very slow rate of dispersal, on which a pre-Pleistocene origin for most of the Sumatran and Bornean endemics depends, is weak. The alter- native suggestion of a Pleistocene origin for many of these endemics seems to be in accord with more of the known geological and zoogeo- graphical facts. The relatively rapid rate of evolution suggested for these snakes is paralleled by the rate attained by the endemic cyprinid fishes of Lake Lanao, Mindanao. Myers (1960) estimates that the 18 species in the lake evolved in little more than 10,000 years from a single INGER AND MARX: THE SNAKE GENUS CALAMARIA 281 species, Puntius binotatus, that reached Mindanao from North Bor- neo during the Pleistocene. That many forms of Calamaria apparently evolved during the Pleistocene in one or two parts of the range does not prove that the same is true in all areas. The Celebesian species may have developed during the same time. At least the Pleistocene was a favorable in- terval for the genus to disperse to Celebes from Borneo. If the three Philippine species, gervaisz, bitorques, and joloensis, or their ancestral stock reached that area during the Pleistocene and differentiated then, their similarity to the Celebes species mueller would be under- standable in terms of a relatively recent origin. The northern limit of warm temperate forest in eastern and south- eastern Asia at the height of the last glaciation was at about 27° N (Charlesworth, 1957, fig. 293), or only a few degrees south of its post- glacial boundary. The continental forms of Calamaria could, there- fore, have occupied their present ranges during the Pleistocene or even earlier, since the climate of the Pliocene was milder than that of the Quarternary (Dorf, 1959). SUMMARY Calamaria is a highly successful member of a group of fossorial colubrid genera that form a characteristic part of the snake fauna of the Indo-Malayan tropical forests. Although all of these genera have lost or consolidated some of the head scales, reduced the number of scale rows, and strengthened the side of the snout in a distinctive way, Calamaria is one of the most specialized within the group. Whether because of its greater specialization or for other reasons, Calamaria appears to be the most successful of these calamarine genera. It has far more species (50) than all of the other genera combined and its geographic range is more extensive than the com- bined ranges of the other genera. Various levels of specialization are evident within the genus Cala- maria. Most of the species have laterally compressed, flask-shaped maxillary teeth; this modification is accompanied by a deepening of the maxilla. Some species have reduced the supralabials to four; some have lost the preocular; others have extremely short tails; others have become extremely slender. The specializations in the various char- acters appear in almost all possible combinations. These combinations are evidence of a remarkable amount of par- allelism. Apparently many stocks have had the capacity to evolve in several directions. Another element in the pattern of evolution of this genus is the importance of variable species, which, though rela- tively generalized, have had a balance among mutation, response to selection, inbreeding, and gene flow that has enabled them to evolve rapidly and proliferate new forms. Both extrinsic and intrinsic factors have played a role in the evo- lution within the genus. The periodic fragmentation of the heart of the range during the Pleistocene provided opportunities for the de- velopment of reproductive isolation. The tendency toward local dif- ferentiation seen in many species and reflecting a rapid rate of genetic change led to the development of genetic isolation and speciation. Most of the evidence now available suggests that many species, per- haps as large a proportion as one-half, differentiated during Pleisto- cene or post-Pleistocene times. 282 INGER AND MARX: THE SNAKE GENUS CALAMARIA 283 Though the generic range extends from eastern China and Burma to Ceram, Celebes, and the Philippine Islands, the majority of spe- cies (64%) of Calamariza occur in Sumatra and Borneo. Because of the large proportion of the species found on these islands and be- cause of the geological stability of these islands during most of the Tertiary and Quaternary, they were probably the principal centers of evolution and dispersal of the genus. APPENDIX A FORMATION OF DENDROGRAM The method used in constructing the dendrogram of phenetic re- lations is based on mean character difference (Cain and Harrison, 1960). Fourteen characters were used: number of supralabials, num- ber of scales surrounding the paraparietal; the relationship between mental and chin shields; presence or absence of preocular; dorsal scales with or without pigment network; light stripe on first or sec- ond scale row bounded above and below by dark pigment; light spots or bars on the sides opposite the vent; number of maxillary teeth; position of the reduction to four scale rows on the tail; number of ventrals; number of subcaudals; sex dimorphism in ventral counts; sex dimorphism in subecaudal counts; and body thickness. The char- acters are described in the introduction (pp. 14-50). For the purposes of comparing species, the first seven characters were scored as follows: CHARACTER STATE SCORE supralabials four 0 i five 100 paraparietal surrounded by five scales 0 a six scales 100 mental and chin shields separated 0 “ touching 100 preocular absent 0 “ present 100 dorsal scale pigment network absent 0 a present 100 light, dark-edged stripe absent 0 . present 100 light spots opposite vent absent 0 nM present 100 285 oooocooocoocoocooooocoscoococoococococooc9sc ns quoA ye ssull ‘sieq ‘syods 14 3VT 0 OOT 98 0 OOT TG OOT 0OT a 00T OOT €¢ 0 0 IG 0¢ 0 O¢ 0 0 "y 0 0 0¢ 0¢ cs Oot OOT 0€ OOT OOT 6G 0OT OOT ae 0OT OOT 0OT 00T 0 OOT US) OOT €6G 0 0OT 67 0 OOT _ 0 0OT 8E 0 0 9¢ 0 0 0 0 LE 0 0 GV MOL y10M40U Ss a[eos yin =| ysiy U0 sa Bas 0 adiiy4s [es1og 1YsBVT IOJOD N re Sex dimorphism subcaudals Sex dimorphism ventrals 91 [op) Subeaudals RS YVAR Ventrals OOT Reduction 0OT 0OT Preocular Mental 00T 00T 0OT 0OT =) So moo Oo O 1d Paraparietals ooo oO OSC OSoCOSCOSCOS Supralabials S uyjasva DSSDLI SUpYyonu nLoydoprsvb..0U wyabayos 101091 COU) $1)D19}D) DISapOUL uhysmoqgnib sisuauvanpod yja1aaa vayyoid sisuanns wwus1abuosg wsvpa1ab sanb10}1Q s7swaojol 149) )ONUL ULL | 44]U9Q1G)D epjomsr416 DapwILLQUin} ‘DILDULD]D) JO Saldedg UIBMJog SedUeleYIq JepeVIVYyD ua, SUIYR[NI[VD Ul pas~y Selodg JeJOVIeYQO—'F9 AIAVL 286 uaa ye SBUlI ‘sieq ‘sqods 1ysIT MOI a[eos 4s UO adtijs WysrT 10JOD 0OT YOM OU yy sa]eos [esioq If Girth Sex dimorphism subcaudals Sex dimorphism ventrals Subcaudals ° Ventrals &6 ion '2 29 Reduction 4 S&S 2 00T S re oooooooqocoo Preocular S OOT OOT Mental © Paraparietals Supralabials UigsiaqqDq Siipjna0avidD Ad aed yayoau SISUIWDLIBI ayasyuagas 0UuL1yj.90.16 uno} $1jpUuorsuajidas DID UaWMWADA Lyong paunanl 1aDUU} DIOUDIaUL SISUAIULOG YysvaL1of py0)nb.110 DUDLJOULNS dunwasa0g DSN.1ISQD Luia}Lepop (*JUOD) DIUDWDIDD JO Saloedg UsEeMJog SedUaIEYIG JoyVIeYO Uva, BUIZB[NI[eD Ul pass Sse1OdG JaJOVIVYQ—'F9 ATAV 287 288 FIELDIANA: ZOOLOGY, VOLUME 49 If a species showed geographic variation in one of these seven characters (e.g., schlegeli in preocular and vzrgulata in the mental shield character) or extensive individual variation (e.g., C. gervaisi in the light stripe), it received a score of 50 for that character. For the remaining seven characters, we used a slight modification of the method outlined by Cain and Harrison (1960), who simply set the highest value of a given character in a suite of species equal to 100 and scored all other values of that character as decimals of the highest. In the case of average number of ventrals, which ranges from 136 to 296, the lowest score for any species is .46.1. Thus the maximum possible interspecific difference in ventral counts would be 54. As the maximum possible interspecific difference is 100 in the first seven characters, this procedure would give greater weight to each of those characters than to the ventrals. But the reason for undertaking this numerical approach was our inability to justify weighting these characters. To avoid introducing bias, we subtracted the smallest average ventral count (136) from all of the averages, thus arriving at a zero value for at least one species. The maximum interspecific difference in ventral counts then reached 100. We made a similar subtraction in the following characters: sub- caudal counts, reduction to four scale rows, percentage of sex dimor- phism in ventral and subcaudal counts, number of maxillary teeth, and body thickness. To obtain the average of ventral counts for a given species, we added the means for males and females and divided by 2. This pro- cedure gives equal weight to both sexes. Subcaudal counts were treated the same way. For sex dimorphism in ventrals, we divided the difference between the means of males and females by the mean of males for the particu- lar species. This procedure gave sex dimorphism in ventrals as a percentage of the average male count (generally the smaller of the two means). Sex dimorphism in subcaudal counts was determined the same way except that the base was the mean female count, which is the lower mean. If only one sex was available for a species, ventral and subeaudal counts and sex dimorphism were not used in comparing that species with others. The value for thickness of each species was derived from the mean of those specimens in the size range 300-399 mm. Because relative _'In actual practice, we multiplied this score by 100 in order to avoid the decimal points. That practice is assumed for the rest of this discussion. INGER AND MARX: THE SNAKE GENUS CALAMARIA 289 thickness increases with increasing size, we felt it necessary to elimi- nate juveniles as well as extremely long specimens (1.e., greater than 400 mm.), which occur in only four species. The character values for the 44 species having modified maxillary teeth are given in Table 64. From this set, we calculated the matrix of mean character differences presented in Table 60. Certain characters used in our “quantitative” analysis of inter- specific similarity have a continuous pattern of variation. Ventrals, subeaudals, reduction, and sex dimorphism in ventrals and subcau- dals are apparently under multigenic control such that a change in any part of the genetic background causes a small recorded change in the value of the character. For example, in the percentage evaluation of characters, the char- acter “‘reduction”’ has values at all levels between 0 and 100. Within a species, variation in this character proceeds in small steps relative to the total range of the character in the genus (see Table 46). The mean value for the species schlegeli is 18.5. A slight change in the mean is possible and would result in only a slight change in the per- centage value in Table 64. Similar comments could be made of the characters mentioned above. The pattern of variation in other characters used in our analysis is radically different. ‘Presence or absence’’ characters vary or their variations are recorded in relatively large steps. The number of scales around the paraparietal is either 5 or 6; the number of supralabials is either 4 or 5. Although one must assume that these characters also are controlled by many genes, a change in any part of their genetic background causes either a large recorded, relative change in the character or none at all. Changes at either end of the scale are thus magnified resulting in many interspecific differences of high or low magnitude. The technique of estimating similarity by use of mean character differences works without bias if the pattern of variation is approx- imately the same for each character. Sneath (1961) uses 14 body proportions to illustrate two methods of analysis. As body dimen- sions vary continuously and as their minor variations are recorded as well as the major ones, the patterns of variation of body propor- tions are similar. If differences between species in these two general kinds of char- acters are lumped together in calculating mean character differ- ences, bias might be introduced. However, by using equal numbers of both kinds of characters, except in seven species, we believe we 290 FIELDIANA: ZOOLOGY, VOLUME 49 have minimized this danger. After removal of the seven species for which we have inadequate data, estimates of mean character differences between lumbricoidea and the remaining species, based upon the two sets of characters, are reasonably close in the species most similar to lumbricoidea. If a joint estimate of 30 is used as the cut-off point—and our dendrogram analysis gives us little confidence in mean character differences beyond that level—the two separate estimates of that mean differ by 5 or less in 9 of 18 comparisons and exceed 15 in only 6 of the 18. Having satisfied ourselves, at least, that the method of scoring characters did not introduce a bias, we made a dendrogram (Fig. 67) of phenetic relations based on the mean character differences (“‘m.c.d.,”’ the abbreviation used by Cain and Harrison, is adopted in the fol- lowing discussion) in Table 60. The method is similar to the un- weighted variable group method described by Sokal and Michener (1958) for use with correlations. Each species was joined to that one with which it had the lowest m.c.d. unless the second one had a lower m.c.d. with a third species. In that event the first species was usually joined to the group formed by the second and third. For example, albiventer and griswoldi have an m.c.d. of 6.2. But griswoldi and lumbricoidea have an m.c.d. of 5.1, the lowest m.c.d. in which either is involved. The nucleus of a group is formed by the last two and albiventer is joined to that group. Two species were always paired as the nucleus of a group if their m.c.d. with one another was lower than the m.c.d.’s of each with all other species. There were eleven such pairs: lumbricoidea-griswoldt, margarttophora-nuchalis, everetti-palavanensis, muelleri-joloensis, su- luensis-prakket, lateralis-lumholtzi, rebentischi-ceramensis, apraeocul- aris-alidae, melanota-borneensis, pavimentata-septentrionalis, and low1- gracilima. One triad of species, crassa, evselti, and doderleinz, had such similar m.e.d.’s (11.6, 11.7, 12.4) that they formed the nucleus of a group. In adding a species to a group the average of its m.c.d.’s with all the members of the group determined the level of its connection. For example, in joining gervaisi: to the group muelleri-joloensis-bi- torques, the average (8.9) of its three m.e.d.’s (8.2, 9.1, 9.5) with those three species gave the level of the connection. In combining groups, the m.c.d.’s of the members of one with all members of the other were used to arrive at a mean m.c.d. Thus to determine the level at which the prakkei-suluensis group joined the muelleri-jolo- ensis-bitorques-gervaist cluster required the average of eight m.c.d.’s. INGER AND MARX: THE SNAKE GENUS CALAMARIA 291 Some situations did not fit the simple logic of this outline. Cala- maria grabowskyi has the same m.ec.d. (12.8) with sulwensis as with everett?, each of which has a lower m.c.d. with another species. The m.c¢.d. (19.8) of sulwensis and everett2 is moderately high. The aver- age m.c.d. of grabowskyi with the nucleus group everetti-palavanensis is lower than that with suluensis-prakkei; grabowskyi was therefore attached to the former group. Some species or groups of species had equivalent average m.c.d.’s to a third group but much higher m.c.d.’s with each other. For example, brongersmai had an average m.c.d. with the muelleri-jolo- ensis-bitorques-gervaist group of 16.6; the mean m.c.d. of prakkei- suluensis with the muellert group was 17.3; the mean m.c.d. of bron- gersmat with prakkei-suluensis was 31.6. If we followed the logic of the method rigorously, brongersmaz should be included in the muellert group and the m.c.d. of the enlarged muellert group to prakkei-suluensis recalculated. This procedure would lower the level at which the last two groups join and would obscure the similarity of prakkei-suluensis to the restricted muellert group. Inclusion of brongersmaz in the muellera group and recalculation of the m.c.d.’s of the latter with all other groups would in every instance lower the level of the connection between groups. One interpretation of this situation is that the similarity of brongersmai to the muellert group results from parallelism; another is that bron- gersmat is an off-shoot of the muellerz stock that developed after the latter split off from the other stocks. Our guess is that the first ex- planation is more nearly correct. At any rate, in order not to obscure the similarity of the mwellerz group to the others, we arbitrarily did not include brongersmaz in it though the dendrogram shows the level of its similarity to the muwellerz group. Calamaria modesta caused a similar problem that was handled in the same way. The solid line indicates the level of its similarity to the grabowskyi-everetti-palavanensis group and the dashed line the level of its similarity to palavanensis. Several alternative ways of connecting groups in the dendrogram were possible. These are indicated by dashed lines. The solid lines indicate the connections that gave the lowest m.c.d.’s over all. AMBON ceramensis AOR low? BALI schlegeli BANGKA schlegeli BASILAN gervaisi lumbricoidea BILLITON javanica schlegeli BOHOL lumbricoidea BORNEO battersbyi* bicolor borneensis* everettr* grabowskyi* gracillima* griswoldi* hilleniusi* lateralis leucogaster lowd lumbricoidea lumholtzi* melanota* modesta prakkei rebentischi* schlegeli schmidti* suluensis virgulata * endemic BURMA pavimentata CAGAYAN SULU suluensis CELEBES acutirostris* apraeocularis* boesemani* brongersmai* curta* muelleri* nuchalis* virgulata CERAM ceramensis CHINA pavimentata septentrionalis COCOS ISLANDS lautensis* FORMOSA pavimentata GREAT NATUNA lumbricoidea HAINAN septentrionalis HONG KONG septentrionalis INDO-CHINA buchi* pavimentata septentrionalis JAVA bicolor javanica lateralis linnaei 293 APPENDIX B: FAUNAL LIST lowt lumbricoidea modesta schlegeli virgulata JOLO 2gervaist joloensis* LEYTE lumbricoidea LUZON bitorques* gervaist MALAYA albiventer low? lumbricoidea pavimentata prakker schlegeli MINDANAO gervaist lumbricoidea virgulata MINDORO gervaist NEGROS gervaisr lumbricoidea NIAS abstrusa forcarti lumbricoidea PALAWAN palavanensis* virgulata 294 FIELDIANA: ZOOLOGY, VOLUME 49 PANAY SIMALUR leucogaster gervaisi modesta lumbricoidea margaritophora* POLILLO SULU ARCHIPELAGO heli* ees mechelt gervaisi ; virgulata modesta RIOUW ARCHIPELAGO . igus SUMATRA schlegelt sumatrana* abstrusa - RIU KIU ISLANDS : ulmeri ca eat albiventer orraulile imenta . P alidae* lumbricoidea doderleini* gervaist SAPARUA eiselti* THAILAND ceramensis forcarti lumbricoidea * endemic REFERENCES BARBOUR, THOMAS 1927. Two new Bornean snakes. Proc. Biol. Soc. Washington, 40, pp. 127-128. BEAUFORT, L. F. DE 1951. Zoogeography of the land and inland waters. viii+208 pp., 10 figs. London. BEMMELEN, R. W. VAN 1949. The geology of Indonesia. Vol. 1A, General geology of Indonesia and adjacent archipelagoes. xxlii+732 pp., 375 figs. The Hague. BOGERT, C. M. 1940. Herpetological results of the Vernay Angola Expedition. Bull. Amer. Mus. Nat. Hist., 77, pp. 1-107, pl. 1, text figs. 1-18. BOULENGER, G. A. 1893. Descriptions of new reptiles and batrachians obtained in Borneo by Mr. A. Everett and Mr. C. Hose. Proc. Zool. Soc. London, 1893, pp. 522-528. 1894. Catalogue of the snakes in the British Museum (Natural History). 2, xi+382 pp., 20 pls. London. 1895. Descriptions of two new snakes of the genus Calamaria. Ann. Mag. Nat. Hist., (6), 16, p. 481. 1897. A catalogue of the reptiles and batrachians of Celebes, with special refer- ence to the collections made by Drs. P. and F. Sarasin in 1893-1896. Proc. Zool. Soc. London, 1897, pp. 193-287, pls. 7-16. 1910. A revised list of the South African reptiles and batrachians, with synoptic tables, special reference to the specimens in the South African Museum, and descriptions of new species. Ann. South African Mus., 5, pp. 455-588. 1915a. A list of the snakes of the Belgian and Portuguese Congo, Northern Rhodesia, and Angola. Proc. Zool. Soc. London, 1915, pp. 1938-2238, figs. 1-2. 1915b. A list of the snakes of East Africa, north of the Zambesi and south of the Soudan and Somaliland, and of Nyassaland. Jbid., pp. 611-640, figs. 1-3. 1915c. A list of the snakes of North-East Africa, from the tropic to the Soudan and Somaliland, including Socotra. Jbid., pp. 641-658. 1920. A list of the snakes of West Africa, from Mauritania to the French Congo. Ibid., 1919, pp. 267-298, figs. 1-2. BRONGERSMA, L. D. 1938. On the presence or absence of hypapophyses under the posterior precaudal vertebrae in some snakes. Zool. Meded., 20, pp. 240-242, fig. 1. BRONGERSMA, L. D., R. F. INGER, and HYMEN MARX —— Proposed use of the plenary powers to conserve the generic name Cala- maria Boie, 1827, and the specific name Calamaria linnaei Schlegel, 1837 (Class Reptilia, Order Serpentes). ms. submitted to Inter. Comm. Zool. Nomen. BROOKS, J. L. 1946. Cyclomorphosis in Daphnia. I. An analysis of D. retrocurva and D. gale- ata. Ecol. Monogr., 16, pp. 409-447. 200 296 FIELDIANA: ZOOLOGY, VOLUME 49 Brown, W. L., JR. and E. O. WILSON 1956. Character displacement. Syst. Zool., 5, pp. 49-64, figs. 1-6. CAIN, A. J. and G. A. HARRISON 1960. Phyletic weighting. Proc. Zool. Soc. London, 135, pp. 1-31. CARSON, H. L. 1959. Genetic conditions which promote or retard the formation of species. Cold Springs Harbor Sym. Quant. Biol., 24, pp. 87-104. CHARLESWORTH, J. K. 1957. The quarternary era. Vol. 2, pp. 595-1700. London. CHERNOV, S. A. 1962. On one still unknown snake of the genus Calamaria. Trudy Zool. Inst. Acad. U.S.S.R., 30, pp. 382-384, 1 fig. [Russian text]. DARLINGTON, P. J. 1957. Zoogeography: the geographical distribution of animals. xi+675 pp., 80 figs. New York. Davis, D. D. 1946. Observations on the burrowing behavior of the hog-nosed snake. Copeia, 1946, pp. 75-78. DOBZHANSKY, THEODOSIUS 1955. A review of some fundamental concepts and problems of population genetics. Cold Springs Harbor Sym. Quant. Biol., 20, pp. 1-15. DOBZHANSKY, THEODOSIUS and BORIS SPASSKY 1954. Genetics of natural populations. XXII. A comparison of the concealed variability in Drosophila prosaltans with that in other species. Genetics, 39, pp. 472-487. DorF, ERLING 1959. Climatic changes of the past and present. Contr. Mus. Paleo. Univ. Michigan, 13, pp. 181-208, 1 pl., 3 figs., 7 maps. DOWLING, H. G. and J. M. SAVAGE 1960. A guide to the snake hemipenis: a survey of basic structure and syste- matic characteristics. Zoologica, 45, pp. 17-28, pls. 1-3, text figs. 1-6. DUELLMAN, W. E. 1958. A monographic study of the colubrid snake genus Leptodeira. Bull. Amer. Mus. Nat. Hist., 114, pp. 1-152, pls. 1-81, text figs. 1-25, maps 1-25. DuUMERIL, A. M. C. and G. BIBRON 1854. Erpétologie général ou histoire naturelle des reptiles. 7, pt. 1, xvi+ 780 pp. Paris. DUNBAR, C. O. 1949. Historical geology. 573 pp., New York. DUNN, E. R. 1923. On a collection of reptiles from Sarawak. Jour. Malayan Branch, Roy. Asiatic Soc., 1, pp. 1-4. 1928. A tentative key and arrangement of the American genera of Colubridae. Bull. Antivenin Inst. America, 2, pp. 18-24. EXDELING, A. C. J. 1870. Recherches sur la faune erpétologique de Sumatra. Nat. Tijds. Ned. Indié, 31, pp. 376-388. INGER AND MARX: THE SNAKE GENUS CALAMARIA 291 FLOWER, S. S. 1899. Notes on a second collection of reptiles made in the Malay Peninsula and Siam, from November 1896 to September 1898, with a list of the species recorded from those countries. Proc. Zool. Soc. London, 1899, pp. 600-697. GANS, CARL 1959. A taxonomic revision of the African snake genus Dasypeltis (Reptilia: Serpentes). Ann. Mus. Royal Congo Belge, 74, pp. 1-237, 13 pls. GUNTHER, ALBERT 1866. Fifth account of new species of snakes in the collection of the British Museum. Ann. Mag. Nat. Hist., (3), 18, pp. 24-29, pls. 6-7. 1896. Descriptions of two new species of snakes from Sarawak. J[bid., (6), 17, pp. 228-229. HAAS, C. P. J. DE 1941. Some notes on the biology of snakes and on their distribution in the dis- tricts of west Java. Treubia, 18, pp. 327-375. 1950. Checklist of the snakes of the Indo-Australian Archipelago. Jbid., 20, pp. 511-625. HIGHTON, RICHARD 1956. Systematics and variation of the endemic Florida snake genus Stilosoma. Bull. Florida State Mus., 1, pp. 73-96, figs. 1-8. INGER, R. F. 1954. Systematics and zoogeography of Philippine Amphibia. Fieldiana, Zool., 33, pp. 181-531, figs. 28-98. INGER, R. F. and HYMEN MARX 1962. Variation of hemipenis and cloaca in the colubrid snake Calamaria lum- bricoidea. Syst. Zool., 11, pp. 32-38, figs. 1-3. JAN, GEORGIO 1862. Enumerazione sistematica delle specie d’ofidi del gruppo Calamaridae. Arch. Zool. Anat. Phys., 2, pp. 1-76, 7 pls. JAN, GEORGIO and FERNANDO SORDELLI 1865. Iconographie générale des Ophidiens. Livr. 10, 6 pls. KOPSTEIN, FELIX 1926. Reptilien von den Molukken und den benachbarten Inseln. Zool. Meded., 9, pp. 71-112. 1941. Uber Sexualdimorphismus bei malaiischen Schlangen. Temminckia, 6, pp. 109-185. LERNER, I. M. 1954. Genetic homeostasis. 134 pp. New York. LEVITON, A. E. 1957. A review of the Philippine snakes of the genus Ovyrhabdium. Wasmann Jour. Biol., 15, pp. 285-808, figs. 1-2. LEVITON, A. E. and W. W. TANNER 1960. The generic allocation of Hypsiglena slevint Tanner (Serpentes: Colubri- dae). Occ. Pap. California Acad. Sci., no. 27, pp. 1-7, figs. 1-2. LOVERIDGE, ARTHUR 1944, Further revisions of African snake genera. Bull. Mus. Comp. Zool., 95, pp. 119-247. 298 FIELDIANA: ZOOLOGY, VOLUME 49 Marx, HYMEN and R. F. INGER 1955. Notes on snakes of the genus Calamaria. Fieldiana, Zool., 37, pp. 167- 209, figs. 21-27. MAYR, ERNST 1945. Wallace’s Line in the light of recent zoogeographic studies. Quart. Rev. Biol., 19, pp. 1-14. 1963. Animal species and evolution. xiii+796 pp., Cambridge. MERTENS, ROBERT 1929. Herpetologische Mitteilungen, XXIII-X XV. Zool. Anz., 86, pp. 57-68. MOLENGRAAFF, G. A. F. and MAX WEBER 1921. On the relationship between the Pleistocene glacial period and the origin of the Sunda Sea (Java- and South China-Sea), and its influence on the distri- bution of coral reefs and on the land- and freshwater fauna. Proc. Akad. Wetens. Amsterdam, 23, pp. 395-447, 3 figs. MYERS, G. S. 1960. The endemic fish fauna of Lake Lanao, and the evolution of higher taxo- nomic categories. Evolution, 14, pp. 323-333. PETERS, WILHELM 1872. Ubersicht der von den herren Mse. G. Doria und Dr. O. Beccari in Sara- wak auf Borneo von 1865 bis 1868 gesammelten Amphibien. Ann. Mus. Civ. Stor. Nat. Genova, 3, pp. 29-45, pls. 2-6. PopE, C. H. 1935. Reptiles of China. Nat. Hist. Central Asia, 10, li+604 pp., 27 pls., 78 text figs. Roois, NELLY DE 1917. The reptiles of the Indo-Australian Archipelago. II: Ophidia. xiv-+ 334 pp., 117 figs. Leiden. SCHENKEL, E. 1901. Achter Nachtrag zum Katalog der herpetologischen Sammlung des Bas- ler Museums. Verh. Ges. Basel, 13, pp. 142-199, figs. 1-6. SIMPSON, G. G. 1940. Mammals and land bridges. Jour. Washington Acad. Sci., 30, pp. 187- 163. SMITH, M. A. 1925. Contributions to the herpetology of Borneo. Sarawak Mus. Jour., 3, pp. 15-36. 1927. Contributions to the herpetology of the Indo-Australian Region. Proc. Zool. Soc. London, 1927, pp. 199-225. 1930. The Reptilia and Amphibia of the Malay Peninsula. Bull. Raffles Mus., no. 3, pp. 1-149. 1931. The herpetology of Mt. Kinabalu, North Borneo, 18,455 ft. Jbzd., no. 5, pp. 3-82, 2 pls. 1943. Fauna of British India. Reptilia and Amphibia. III Serpentes. xii+ 583, figs. 1-66, London. SNEATH, P. H. A. 1961. Recent developments in theoretical and quantitative taxonomy. Syst. Zool., 10, pp. 118-189, figs. 1-5. INGER AND MARX: THE SNAKE GENUS CALAMARIA 299 SNEATH, P. H. A. and R. R. SOKAL 1962. Numerical taxonomy. Nature, 193, pp. 855-860, figs. 1-5. SOKAL, R. R. and C. D. MICHENER 1958. A statistical method for evaluating systematic relationships. Univ. Kansas Sci. Bull., 38, pp. 1409-1488, figs. 1-6. STEENIS-KRUSEMAN, M. J. VAN 1950. Malaysian plant collectors and collections. Flora Malesiana, 1, clii+ 639 pp. TAYLOR, E. H. 1917. Snakes and lizards known from Negros, with descriptions of new species and subspecies. Philippine Jour. Sci., 12, pp. 353-379, pls. 1—2, text figs. 1-2. 1922. The snakes of the Philippine Islands. Philippine Bur. Sci., Manila, Monogr. 16, 312 pp., 87 pls., 32 text figs. 1923. Additions to the herpetological fauna of the Philippine Islands, III. Philippine Jour. Sci., 22, pp. 515-557, pls. 1-8. 1962. New Oriental reptiles. Univ. Kansas Sci. Bull., 41, pp. 209-268, figs. 1-18. THROCKMORTON, L. H. 1962. The problem of phylogeny in the genus Drosophila. Stud. Genetics, 2, pp. 207-3438, figs. 1-54. TWEEDIE, M. W. F. 1953. The snakes of Malaya. Gov. Print. Office, 139 pp., 12 pls., 27 text figs., Singapore. UMBGROVE, J. H. F. 1938. Geologic history of the East Indies. Bull. Amer. Assoc. Petrol. Geol., 22, pp. 1-70, 1 pl., 24 text figs. 1949. Structural history of the East Indies. Cambridge Univ. Press, 63 pp., 9 pls., 68 text figs. WALL, FRANK 1921. Remarks on a specimen of Calamaria javanica. Rec. Indian Mus., 22, Dp. 729; WERNER, FRANZ 1909a. Neue oder seltenere Reptilien des Musée d’ Histoire naturelle de Belgique in Brussel. Zool. Jahrb., (Syst.), 28, pp. 263-288, 2 figs. 1909b. Uber neue oder seltenere Reptilien des Naturhistorischen Museums in Hamburg. Mitt. Naturh. Mus. Hamburg, 26, pp. 205-247. WRIGHT, SEWALL 1932. The roles of mutation, inbreeding, crossbreeding and selection in evolu- tion. Proc. 6th Int. Congr. Genetics, 1, pp. 356-365. 1949a. Adaptation and selection. Jn Genetics, Paleontology, and Evolution. ed. Jepsen, Mayr, and Simpson. pp. 365-889. Princeton, New Jersey. 1949b. Population structure in evolution. Proc. Amer. Philos. Soc., 93, pp. 471-478. 1955. Classification of the factors of evolution. Cold Springs Harbor Symp. Quant. Biol., 20, pp. 16-24. INDEX Principal taxonomic discussion indicated by bold-face type. asbtrusa 42, 177, 179, 182, 186, 280 Acanthophis antarcticus 207 acutirostris 12, 18, 22, 28, 24, 30, 45, 73, 119, 172 adaptation 241 adaptive radiation 247 agamensis 20, 154, 162, 165 albiventer 7, 45, 94, 99, 265, 278 albopunctatus 187 alidae 19, 41, 47, 177, 235, 257, 261 alkent 76 allopatric populations 13 Amblyodipsas 255 Ambon 293 anceps 186 annamensis 213 Aor 293 Aparallactus wernert 252 apraeocularis 20, 41, 119, 172, 237, 257, 261 arcticeps 63 Asia, key to species 60 Aspidura 241, 249 trachyprocta 250, 252 Bali 293 baluensis 124, 126, 129, 1380 banaensis 213 Bangka 293 barriers, ecological 272 geographic 272 Basilan 293 battersbyz 208, 261 beccari 63 benjaminsi 198 berezowskit 213 bicolor 18, 19, 30, 44, 50, 63, 99, 119, 147, 149, 163, 240, 262 group 259, 270 bilineata 201 Billiton 293 bitorques 18, 46, 50, 104, 260, 271, 281 Blythia 249 body thickness 39 boesemani 41, 100, 119, 171, 262 bogorensis 138, 1389 Bohol 293 booliati 7, 154, 163 popes 16, 18, 26, 30, 42, 194, 201, 257 301 Borneo 293 key to species 58 Brachyorrhus 243 brachyura 186 brevis 201 brongersmai 42, 100, 118, 172 brooki 63 bruegeli 77 bucht 41, 211, 262, 273 bungaroides 76 Burma 293 burrowing snakes 241 Cagayan Sulu 293 Calamelaps 241 unicolor 252 calamarine genera 254 calamarius 201 Calamorhabdium 2538, 254 Cantoria 254 Celebes 293 key to species 61 centers of evolution 275 Ceram 293 ceramensis 44, 280, 232, 257, 272 Changulia 7 albiventer 94 lumbricoidea 77 multipunctata 201 characters, origin 256 primitive 252 Chilorhinophis 255 carpentert 252 China 293 chin shields 20, 51 Chrysopelea 249 classification, phenetic 257 cloaca 36 Cocos Island 293 collaris 186 coloration 46 Coluber calamarius 201 contaminata 201 convergence 241 correlation of characters 50 crassa 23, 39, 45, 50, 169, 173, 177, 180, 186, 257 curta 12, 23,43, 71, 119.172 cuviert 52, 165, 240 302 dendrogram formation 285 dimidiata 76 dispersal time 279 distribution 272 ecological 247 divergence 245 déderleini 43, 169, 173, 177, 239, 257 doerianense 227 dorsal scale rows 31 dumerili 162 ecological distribution 247 barriers 272 egregia 187 erselti 23, 39, 46, 47, 50, 169, 175, 180, 186,257, 262 group 209 electa 198 elegans 19, 138, 140 everetti 19, 29, 31, 42, 119, 121, 126, 133, 136 evolution 241 centers 275 patterns 263 eye size 51 flaviceps 76 forcartt 41, 177, 180, 182, 184, 280 Fordonia 253 Formosa 293 formosana 213 frasert 227 frontal ratio 51 frontal width 14 fusion of head shields 253 gastrogramma 198 gastropicta 201 genotypic variability 263 geographic barrier 272 variation 50, 143, 155, 191, 270 gervaisi 16, 18, 21, 27, 30, 39, 42, 50, 102, 106, 251, 260, 271 281 gervaist 106 hollandi 107 iridescens 107 polillensis 107 gimlett: 28, 221, 226 goeringt 77 goring. T7 grabowski 123 grabowskyi 18, 19, 28, 26, 30, 36, 44, 50, 124, 129, 133, 136, 262 group 259, 261 gracillima 7, 24, 229, 257, 261, 272 gracilis 77, 186 grayt 76 Great Natuna 293 griswoldi 12, 15, 18, 30, 33, 36, 44, 47, 50, 92, 99, 265 grooved maxillary teeth 243 gulars 20, 274 FIELDIANA: ZOOLOGY, VOLUME 49 Hainan 293 Haplocercus 249 head scales, loss 243 hemipenis 36 Heterodon 241 heterozygous genotypes 265 hilleniust 45, 52, 96, 102, 265 hoevenii 166 hollandi 107 Homalopsinae 247 Hong Kong 293 hosei 149 hypapophyses 2438 Idiopholis 254 everett, 251 Indo-China 293 indragirica 94, 130 intestinalis 240 tridescens 107 iris 163 Java 293 key to species 60 javanica 42, 52, 209, 223, 227, 257, 261, 262 jeuder 232 Jolo 293 joloensis 24, 41, 102, 260, 281 Keimetopon 7 booliati 163 klosst 240 labials 15 lateralis 12, 42, 119, 146 lautensis 44, 70, 138, 139 leucocephala 20, 63, 149, 154, 165 leucogaster 18, 19, 21, 26, 30, 39, 43, 63, 126, 186, 251, 262 Leyte 293 lineata 223 linnaet 9, 15-19, 22, 26, 30, 33, 46, 47, 50, 198, 200, 209, 257, 273 local variation 270 lowi 18, 19, 27, 30, 39, 41, 221, 257, 261, 244 gimletti 23, 221, 226 lowt 16, 28, 222 wermuthi 222, 225, 229 lumbricoidea 7, 9, 16, 18, 20, 21, 28, 24, 26, 27, 29, 30, 37-39, 45, 50, 75, 98, 154, 177, 186, 193, 2145 2405 2075 262, 265, 200, 210, 200 griswoldi 92 group 259 lumbricoidea 76 lumholtzi 15, 50, 119, 147, 148, 173 Luzon 293 Lytorhynchus 241 Macrocalamus 2438 INDEX 303 macrurus 149 maculolineata 166 maculosa 201 Malaya 293 margaritifera 166 margaritophora 46, 166, 177, 180, 186, 201, 257, 262 margaritophora 166 Maticora 240 maxilla 252 maxillary teeth 20 mearnsi 187 mecheli 19, 186, 233, 257 melanorhynchus 76 melanota 30, 45, 198, 208, 227, 257, 262 group 259, 262 mental shield 18, 254, 274 Mindanao 293 mindorensis 107 Mindoro 293 Miodon gabonensis 250, 252 mjobergi 138, 140 modesta 16, 18, 19, 21, 27, 30, 36, 44, 119, 126, 187, 138, 173, 177, 186, 207, 261, 2738, 278 bogorensis 138 Moluccas, key to species 61 monochrous 138 moultoni 130 muellert 12, 18, 23, 30, 44, 99, 119, 172, 257, 260, 262, 281 group 259, 262 multilineata 201 multipunctata 201 nasal bone 252 natricine stock 249 Natrix 207, 249 Negros 293 Nias 293 nigroalba 163 nuchalis 18, 23, 43, 190, 119, 170, 172, 177, 180, 186 occipitalis 240 oculars 15 Oligodon 255 ontogenetic variation 39, 50, 82, 153 Oreocalamus 248 origin of characters 256 ornata 94 Oxyrhabdium 244, 249 modestum 250, 252 palavanensis 19, 42, 50, 119, 133, 134, 261 Palawan 293 Panay 294 parallelism 257 paraparietals 16, 51, 262, 274 patterns of evolution 263 pavimentata 17, 19, 27, 30, 41, 181, 201, 212, 227, 239, 261, 262, 270, 272, 273, 278 pendleburyi 123, 126 pfeffert 213 phenetic classification 257 relations 258 phenotypic variation 267 Philippine Islands, key to species 61 philippinica 76 Phyllorhynchus 241 phylogenetic relations 241 groups 251 phylogeny 264 picteti 149 Plagiopholis 249 Pleistocene history 266, 278 Poecilopholis 244 polillensis 107 Polillo 294 prakkei 23, 46, 120 prakkii 121 prefrontal 252 preocular 51, 254, 262, 274 Pseudorabdion longiceps 251 Pseudoxenodon 249 quadrimaculata 213 quinquetaeniata 187 ravent 148 rebentischi 230, 257, 262 reproductive isolation 12 reticulata 201 Rhabdophidium 254 rhomboidea 201 Riouw Archipelago 294 Riu Kiu Islands 294 roelandti 163 Samar 294 Saparua 294 Scaphiophis 241 schlegeli 8, 12, 15-16, 18-20, 23, 26, 27, 30, 31, 37-39, 48, 47, 50, 99, 119, 150, 154, 178, 177, 186, 240, 251, 261, 270, 273, 278 cuviert 52, 165, 240 schlegeli 162 schmidti 74 semiannulata 149 semidoliata 239 septentrionalis 17, 21, 27, 30, 31, 33, 43, 218, 239, 251, 257, 261, 270, 272, Za, 215 septomaxilla 251, 252 sexual dimorphism 24, 28, 31, 274 siamensis 213 Simalur 294 stmalurensis 19, 188, 140 sinkawangensis 163 smithi 63 304 FIELDIANA: ZOOLOGY, VOLUME 49 snout 251 sondaica 201 speciation 269 stahlnechtii 76, 184 Stenorhabdium 244 Stilosoma 241 subcaudals 27, 274 subminiata 207 Sulu Archipelago 294 suluensis 12, 15, 18, 23, 29, 30, 38, 42, 50, 121, 123, 133, 186, 262 group 259 Sumatra 294 key to species 59 sumatrana 18, 30, 39, 42, 120, 1238, 177, 180, 181, 186, 261 © group 262 sumatranus 76 supralabials 51, 252, 253, 274 sympatric samples 8 Tablas 294 taxonomic significance, characters 14 teeth 274 temmincki 76 tessellata 201 Tetralepis 2538, 255 Thailand 294 thickness, body 39 time of dispersal 279 transversalis 201 tropica 107 Typhlocalamus 7, 229 ulmeri 68, 262 uniformis 213 variabilis 76 variation, characters 14 geographic 50, 148, 155, 191, 270 local 270 ontogenetic 39, 50 phenotypic 267 ventralis 223 ventrals 24, 274 ventrimaculata 201 vermiformis 76 grayt 77 versicolor 201 virgulata 12, 18, 19, 21, 28, 39, 41, 118, 119, 186, 138, 172, 177, 186, 261, 270, 275, 278 group 262 vittata 207 wermutht 222, 225 Xenocalamus 241, 255 Xylophis 249 yunnanensis 238, 273 zamboangensis 187 *e “y im cn ¥ ee ae ee PO Stage ae od bs wwe ' a is rn SUN ne oy or >. : RE MS Ry rete « ee Se ey ay: ae a ~~ eh ds F 4 4 \*- a aor « ¢ = Eee ey : J a eR Ve we ; - ‘ [oem ec ab n = . i ~ Sees Go Slew ee ature = 2 ; epee Shoh ee oten AB he heat ow egre grarareers: a ete, 3h HL Te te SS =e ya See ANTS ad Anew ds ee eey cy hy eh i CO ae he SS li yess “rr tn Wey Trae ? were s TF apse ge eR Tt Pew Nag ar Bee Mey ey Ate ew h % “4 Aa ie) h 4 : " “ “ ’ LPI pe ee = > “b . whee ot * Wes y 5 a ; aed o . ‘ . , ~~ Ps eS ae es Fw mee Saree yt Aste ‘ Hite we scene we { : ce tees : At RY A he Se pn wee Rr sD Ye i \ : : - ; aie et Tt eo a aa eee Ae. ; ee er A ee Ce a = > a Y sv = ed Shs ere Pye 9 9 ees ee