Zoological Club of the Linnean Society/* 603 



possessed by each in common with the families most nearly con- 

 nected with it. The circular succession of affinities among them 

 he had hitherto been unable to trace with precision; although he 

 had ascertained the existence of osculant species or genera be- 

 tween some of the families. The genus Terrapene^ Merrem, or ra- 

 ther the sub-family Sternoihcerina, Bell, appeared to be inter- 

 mediate between the Testudinidce and Emydidw^ exhibiting the 

 gibbous form of the land tortoises, but approximating to the fresh- 

 water ones by its shorter sterno-costal suture and by its more pal- 

 mate feet. In like manner the Emydidce are connected with the 

 TrionychidcB by means of the Matamata, Chelysjimbriata^ Dura., 

 in which not only does the sternum become shorter than in the 

 more prominent types of the former family, but the beak also 

 ceases io be horny, and is replaced by the soft projecting snout 

 of the genus Trionyx. Between the remaining families he was 

 yet unable to point out the connecting links. In the course of his 

 observations he adverted to the importance of some characters 

 hitherto little attended to in the discrimination of species, and par- 

 ticularly to that furnished by the area, or scabrous central portion, 

 of each plate. This is always fully developed when the young 

 tortoise quits the egg, and never increases in size, the subsequent 

 growth of the plate being produced by additional lateral layers. 

 Hence it furnishes a constant clue to the determination of the age 

 of the individual. In the Testudo elegans^ Schaepff, the area is 

 large with scarcely any lateral striae. This species is consequently 

 founded on a very young shell, and an examination of the area in 

 shells of various sizes shows it to be the young of T. radiata. 

 Other instances were also adduced in confirmation of the remark ; 

 and the same character was shown to be essential in distinguishing 

 species when the shells had attained their full growth. In T. 

 geometrica for instance, the very small area affords almost the only 

 point of discrimination between it and the closely allied species, 

 T. actinodesy in which the area is considerably larger. Mr. Bell 

 further adverted to a novel kind of Box-tortoise, in which the 

 moveable portion of the shell forming the box existed in the cos- 

 tal plates instead of in the sternal as in Terrapene. The Land- 

 Tortoises exhibiting this striking peculiarity he had recently de- 



