108 



Current Herpetol. 20(2) 2001 



signed Heosemys silvatica to Geoemyda. 

 We followed this generic arrangement like 

 a few other recent revisions (David, 1994; 

 McCord et al., 1995; Yasukawa and Ota, 

 1999). 



McCord et al. (1995) also assigned their 

 new species, Geoemyda yuwonoi, and the 

 two other species, Heosemys depressa and 

 H. leytensis, to Geoemyda. However, 

 generic status of the three species remains 

 controversial (Fritz and Obst, 1996; Iverson 

 and McCord, 1997b; Yasukawa and Ota, 

 1999). Recently, McCord et al. (2000) 

 assigned G. yuwonoi to a new genus, 

 Leucocephalon, and reassigned G. depressa 

 to Heosemys on the basis of analysis of 

 mitochondrial DNA variation. Since we 

 did not examine those three species, we 

 tentatively followed McCord et al. (2000). 

 For the genus Geoemyda as defined above, 



G. japonica (13), G. silvatica (6), and G. 

 spengleri (12) were examined. 



Heosemys. — McDowell (1964) assigned 

 five species, Heosemys grandis, H. spinosa, 



H. depressa, G. leytensis and G. silvatica, 

 to Heosemys, probably on the basis of 

 absence of the quadratojugal (squamosal in 

 his terminology). He, however, did not 

 directly examine the latter three species, of 

 which H. leytensis and H. silvatica were 

 recently reassigned to Geoemyda (see 

 above). We examined H. grandis (3) and 

 H. spinosa (4). 



Leucocephalon. — This genus contains 

 only one species, Leucocephalon yuwonoi, 

 which was formerly assigned to Geoemyda 

 or Heosemys (McCord et al., 2000). We 

 have had no chance to examine this spe- 

 cies. 



Mauremys. — This genus currently con- 

 sists of eight species, M. annamensis, M. 

 caspica, M. iversoni, M. japonica, M. 

 leprosa, M. mutica, M. pritchardi, and 

 M. rivulata (Fritz and Wischuf, 1997; 

 McCord, 1997), of which six, M. annamen- 

 sis (6), M. caspica (1), M. japonica (10), 

 M. leprosa (2), M. mutica (11), and M. riv- 

 ulata (6), were examined. There were no 



essential differences in any character among 

 M. caspica, M. leprosa, and M. rivulata. 

 They were formerly considered as three 

 subspecies of M. caspica, and are most 

 likely monophyletic (Busack and Ernst, 

 1980; Ernst and Barbour, 1989; Iverson 

 and McCord, 1994). We thus treat them as 

 constituting a single OTU, the Mauremys 

 leprosa species-group. 



Melanochelys. — This genus contains two 

 species, M. tricarinata and M. trijuga, of 

 which only M. trijuga (10) was examined. 



Notochelys. — We examined six specimens 

 of A^. platynota, the only representative of 

 this monotypic genus. 



Pyxidea. — This is a monotypic genus 

 consisting solely of P. mouhotii. This 

 species, formerly assigned to Geoemyda 

 (e.g., McDowell, 1964) or Cyclemys 

 (e.g., Wermuth and Mertens, 1961; Prit- 

 chard, 1979), is treated as Pyxidea by 

 most recent authors (e.g., Hirayama, 

 1984; Ernst and Barbour, 1989; Iverson, 

 1992; David, 1994). Five specimens were 

 examined. 



Rhinoclemmys. — This is the only bat- 

 agurid genus occurring in central and 

 South America, and includes nine spe- 

 cies: annulata, areolata, diademata, 

 funerea, melanosterna, nasuta, pulcher- 

 rima, punctularia, and rubida (David 

 1994). Hirayama (1984) suggested the 

 polyphyly of the genus, but all subsequent 

 authors have apparently ignored this 

 account (e.g., Ernst and Barbour, 1989; 

 Iverson, 1992; David, 1994). Therefore, we 

 tentatively treated them as composing a sin- 

 gle genus, and examined R. annulata (2), R. 

 areolata (2), R. funerea (3), R. pulcherrima, 

 (5), R. punctularia (4), and R. rubida (1). 



Sacalia. — Iverson and McCord (1992b) 

 recognized three species, 5. bealei, S. quad- 

 riocellata, and 5. pseudocellata, for this 

 genus. They showed some differences in 

 coloration and shell shape, but none in 

 internal morphology. We examined S. 

 bealei (2), S. quadriocellata (2), and Sacalia 

 sp. (2: skeletal specimens not identified at 



