LAND-TOETOISES AND AN EXTINCT LIZAED FEOM MAUEITIUS, 321 



Adult specimens of Chelone midas have a similar, although much smaller, caudal 

 claw ; in a large male specimen in the Cambridge Museum the horny claw or spur is 

 about 2 cm. in length, and covers about three vertebrse, two anchylosed and one free. 



None of the textbooks of zoology, comparative anatomy, and herpetology written 

 since 1877 have as yet condescended to mention this important instance of an anchy- 

 losed terminal caudal vertebral complex in the class of Reptiles. 



Cervical Vertehrce (PL XLIV. figs. 20-25). — Three atlas vertebrae, representing two 

 different types, one with slightly joined neural arches and with a ring-shaped perforated 

 body, the others with a completely solid unperforated body — differences which cannot 

 be explained away by age. Although the atlas of the various families of Tortoises 

 exhibits many modifications, the specimen A, with the solid body, is peculiar; its 

 body does not contain the odontoid process, because the latter has left the three typical 

 articulating facets or impressions upon the body of the atlas. Specimens B and C 

 agree more with the atlas vertebrae figured by Dr. Giinther. 



Pelves. — Five nearly complete specimens, numbered 1. to V. (PI. XLIII. figs. 4 & 5j. 



Pelvis No. I. has to be assigned to T. triserrata according to Dr. Giinther's defini- 

 tion. It is the largest known, measuring 23 cm. in height atid 23 cm. in width. 

 The bridge between the obturator foramina is very broad, namely 3-8 cm. Unfor- 

 tunately, the tuberosity of the ischiadic symphysis, resting upon the plastron, is lost. 

 The ventral ridge of the ischiadic symphysis is very prominent. 



Pelvis No. II. belongs to a smaller specimen, its measurements being 17 and 

 15-5 cm. The obturator foramina are wide, the bridge consequently narrow, resem- 

 bling that of Aldabran Tortoises. The lateral ridge or crest of the shaft of the ilium 

 is very prominent, more so than in T. triserrata. This pelvis differs markedly in two 

 peculiarities from those which have been determined as belonging to T. triserrata and 

 T. inepta. First, the longitudinal ridge on the ventral side of the ischiadic symphysis 

 is very low, instead of being very prominent. Secondly, the ischiadic tuberosity has a 

 deep cavity on its ventral surface, and is rough instead of being smooth, suggesting 

 that it fitted upon and was partly fused with a corresponding tuberosity of the posterior 

 end of the plastron. Such a plastron is that which I have distinguished as belonging 

 to T. sumeirei. 



Pelves III., IV., and V. resemble each other, and those of T. inepta more than other 

 species ; but it has to be noted that they exhibit a certain amount of variation in the 

 extent of the lateral iliac ridge — in fact, that they are intermediate between the typical 

 specimens of T. triserrata and T. inepta. They may belong to T. leptocnemis, the 

 pelvis of which is diagnosed as resembling that of T. triserrata, but with a narrower 

 ilium. 



Scapulm and Coracoids. — The four specimens are all different. Two are more like 

 those of T. inepta; one more like that of T. triserrata; the fourth has a very flat 

 scapular shaft, resembling in its transverse configuration that of T. triserrata and still 



VOL. XIII. — PAET VIII. No. 3. — April, 1894. 3 a 



