CHELONIA. 723 



genus Melanochelps, and tlie third by the genus Bellia. The first two of these, 

 according to Gray, contained forms with imperfect, or perfect zygomatic arches. 

 The first, he stated, had never webbed feet, the second had webbed feet, and the 

 third tribe, BelUana, resembled the Emydina. The only one genus, however, 

 among the 15 genera referred to those tribes that possesses an imperfect zygomatic 

 arch is the genus Geoemyda. Although Gray stated that Melanochelys had an im- 

 perfect and weak zygomatic arch, this was an error, as the species E. trijuga, on 

 which the genus is founded, has been satisfactorily proved by my personal observ- 

 ation of many skulls from Ceylon, Southern India, Madras, Central India, and 

 Burma, to have a perfect zygomatic arch, so that the genera of the second and 

 third groups are all distinguished by a perfect arch. It is doubtful whether 

 several of the nine genera, referred to the Bmydina, may not have ultimately to 

 be united with the genus Emys, which I temporarily accept in the sense adopted 

 by Gray, with the above correction. 



Emys trijuga, D. & B., var. hurmana. Plate LVII & LVIII. 



Emys trijuga, Theobald^, Journ. Linn. Soc, vol. x, p. 13, 1868 (pars). 

 Emi/s edeniana, Theobald, Descr. Cat. Rept., Brit. Ind._, 1876^ p. 12. 



Shell not so elongately oval as in Indian examples of the species, moderately high 

 and broad, and broader in females than males, anterior margin slightly posteriorly 

 concave. Lateral margin acutely revolute in the young, reduced to a ridge in 

 nearly adult specimens. In the young slightly expanded at the eighth marginal, 

 the posterior marginal shields being involute, more so in some than in others, 

 and feebly in the adult. The posterior margin is slightly serrated in some 

 young specimens and undulating in adults ; caudal notch hardly developed. Three 

 dorsal ridges, the vertebral ridge the most prominent ; the costal areolar ridges 

 very feebly developed, passing through the areolae which have sometimes a tend- 

 ency to nodosity, and not extending on to the last costal and scarcely on to 

 the first, in which, however, the areolae are well marked. The costal ridges 

 nearly disappear with age, but the vertebral is always distinct, and most prominent 

 on the three last, and on the first vertebral. The nuchal is narrow and almost linear 

 broader behind than in front. The vertebrals are generally broader than long, the 

 second and third being as long as broad. There are frequent exceptions to this, 

 especially in the case of the first vertebral wliicli is often much longer than broad, 

 and in such cases the other vertebrals partake of its elongated character. All the verte- 

 brals, with the exception of the first and last, are almost as broad in front as behind, 

 and the costal margins of the second, third and fourth are all about one length, the 

 anterior costal margin of each being more or less convex, and the posterior concave. 

 The first vertebral is pentagonal, either broader than long, as long as broad, or con- 

 siderably longer than broad. Its lateral margins are more or less concave in the 

 middle, but divergent anteriorly. The posterior margin is transverse, notched for 

 the vertebral ridge. The second vertebral is generally a little broader in front than 



