YASUKAWA ET AL.— I'ln I OCil'NY Ol GEOEMYDINE l UR I LES 125 



because llic cliaiaclci is liit'lilv vanahk' 

 williiu ilic I csi luliiioidca ((iatliic\ aiul 

 Mcylaii, I^ASS; \asuka\\a, uupiiblislicel 

 data), aiul is also seen in Mcso/oic prim- 

 itive tesiiuiinoids sueli as Alo/ii^o/cinvs 

 (Hirayama, unpublished data). 



There are no synaptMnoi phs elcaii\ 

 supporting llic nionophvK ol ( icoeinN diuac. 

 The charaelcr siales sliaicd b> iiiosi 

 mcinbcis ot iliis siihtamiK largely repiesenl 

 liiiiinii\c coiuliiions in ilie liataguridae 

 (llira>ania, 19X4; Table 1). This may 

 suggest the paraphyletie naluie ol ilie 

 Gcoemydinae as an assemblage ol primitive 

 stocks of the Hataguridae. Nevertheless, 

 because uc could also dcicci no s\iiapo- 

 inoi phs siippoi iiiiL' sisici group r^lat lousliip 

 Willi llic Baiagunnac ol aii\ pailiculai 

 clade wilhin the ( ici)cm\clinac, ihis |iroblcm 

 remains lo be icaddrcssed on the basis of 

 additional daia. In cousitleration of ia\(v 

 nomic stabilii\ , wc propose lo iciupoi ai il\ 

 rclain ihc iccoiMiiiion ol liic sublainiU 

 ( icocnu dniac as a "uiclalaxou", lor uhicli 

 no characler evidence sujiiioris or negates 

 ihc inonophyly (Ciaulhiei cl al., 19SS), 



()ui analyses re\ealed dial ihc dciicnud- 

 inae consists ol ihe A/(////('//ms and ilic 

 Gcocniydu groups. In ilic lollouiii!' sub 

 sections, we discuss ihc pli\ loyciu wuliiii 

 each o\ ihese groupN. W c also jiroposc 

 se\cial modi I icai ious lo ihc classiricaiion 

 by Das id (1994) ( 1 able 4) on the basis ol 

 evolutionary relationships revealed in this 



si lld\ . 



Main cm\ s ■^roiif 



I he icsulis ol ihis study may support the 

 inonophyly ol the Mutin'/nys grouji as 

 consisting of Mauremys and Saailia with 

 one synapoiuorph (»l |1|, sniootli skin ol 

 the posterior head). 1 \cn so, however, tliis 

 character state was also observed in Stem J 

 except f or Gcocmyda silviiiu u and Pyxulca. 

 Ihis suggests that 'S\ (I) has evolved more 

 than one time in the Gcoemydinae. and 

 reduces it value as e\ idence for monophyly 

 of the Matircmys group Other character 



slates c\liiliiicd In \liiiiiciiiv\ and SiUiiliii 

 seem lo nu)slly re|Mescnl primitive condi- 

 tions ol Ihe (icoemydinae (Table 1), or ol 

 ihe whole lamily Balaguridae (Hirayama, 

 iyS4). riuis, the Maureiuys group may be 

 icreired to as the primitive stock of this 

 I a 111 I ly. 



Monophyly of Ihe genus Maurcnivs was 

 su|-ipoited b\ two character states, 14 (1) 

 and 15 (1), while Sacalia exhibited no 

 auiajiomorphs in our parsimony analyses. 

 McDowell (1964) suggested that Sacalia is 

 disci iiiimaled from Maureinys in lour char- 

 acter states, or these, however, only one 

 (a lew \ery large scutes without intervening 

 granules on forearms) is a possible synapo- 

 inoi ph of the Sacalia species. Although ue 

 did not employ scutellation of the forearm 

 in ihc phylogenetic analyses due to its 

 extensisc intrageneric \ariaiion in the 

 Gcocnivdii group, we confirmed ihe unic|ue 

 state mcniioiicd abo\c in ihis character in 

 Siiciilia. ( oiisidci ing it as an autapomorph 

 ol Sacalia, we regard both Maureinys and 

 Sacalia as \ alid. 



(icociincia '^roi//) 



In out results, the monoplnly of the 

 ( n'Dcinyda group is supported by three 

 cli iiacier slates, absence of the squamosal- 

 i|u idratojugal contact (7 |1|), absence of 

 ihe jugal-quadratojugal contact (8 (!]), 

 and anterior neurals with shorter posterior 

 sides (:> |1|). Of these, 25 (1) was constant 

 within the Cn'ocniyda group exclusive of 

 .\<)i<)c/n'lys, in whicli the anterior neurals 

 IkkI shorter anterior sides hkc those in the 

 Maiiivtnys group and most batagurines. 

 Hirayama (19S4) depicted a sister group 

 relationship between Saaiiia and SoUKlwIys 

 as supported by jugal parietal contact, and 

 distinctly widened first vertebral which 

 sometimes reaches the second marginal. 

 Ihe present study revealed that the jugal is 

 close to the parietal but separated from il 

 in Noiodu'lys, while Ihc two make weak 

 contact in Sacalia. The first vertebral was 

 widened iiu! usu.ilK ickIiciI the second 



