LAND-TOETOISES AND AN EXTINCT LIZARD PEOM JIAFEITIFS. 319 



fore legs are 45, the hind legs 30 cm. long; neck and head 39-5 cm., tail 30-5 cm. in 

 length. 



The back view shows a very large, broad, and sharply-curved caudal shield, which 

 strongly resembles that of T. triserrata, and differs from that of any Aldabran or 

 Galapagos specimens. Front and side views show that there is no nuchal shield 

 whatever. 



The under view shows a slightly forked projection of the anterior end of the plastron, 

 with two gular shields, indicating two gular bony plates as in the forked plastra 

 D, E, F, G. 



The first marginal shield is very large and in broad contact with the first costal, 

 agi'eeing in this respect with Aldabran specimens. 



The profile of the carapace, the scarcely serrated marginals, and the markings of the 

 vertebral and costal shields most resemble the corresponding parts of T. ponderosa = 

 female of T. elephantina. 



The anterior portion of the plastron, which is well shown in the photograph, agrees 

 in length, narrowness, and forked termination with T. daudini. 



According to Boulenger's Key, pp. 153-154, Cat. Chelonians, British Museum, this 

 specimen would come nearest to T. nigrita and T. nigra s. elephantojms (nuchal absent, 

 gulars distinct, shields of carapace concentrically striated in the adult, profile of cara- 

 pace declivous in front). But T. nigrita differs considerably in the shape of the 

 anterior end of the plastron, the profile of the distinctly humped vertebral shields, and 

 the sen-ated marginals. T. elephantopus differs likewise in the shape of its plastron. 

 Both T. elephantopus and T. nigrita, moreover, differ in the shape of thek much 

 smaller caudal shield, and above all in the shape of the head. The head of the 

 photographed specimen, T. sumeirei, agrees much more with that of the Aldabran 

 type, while T. triserrata and T. inepta appear to have resembled the Galapagos 

 types. 



There remains the question whether the forked plastra D, E, F, G belong to the 

 same race of Tortoises as T. sumeirei. 



This question is difficult to settle ; we do not know the carapaces which belonged to 

 the plastra D, E, F, G. 



However, this much is certain — (1) that the plastra D, E, F, G cannot have 

 belonged to Galapagos Tortoises, because of the double gulars and because of their 

 locality ; (2) that the type of T. sumeirei cannot be a specimen introduced from the 

 Galapagos Islands, because of the shape of its head, plastron, and double gulars; 

 (3) that T. sumeirei cannot be one of the true Aldabran species, because it has no 

 nuchal shield and because of its different caudal shield; (4) T, sumeirei exhibits 

 quite a new combination of characters, namely double gulars, without nuchal, and 

 is indigenous in Mauritius. At any rate, we have here a Mauritian Tortoise which is 

 fundamentally of the Aldabran type, but combines with Aldabran features several 



