TESTUDINIDAE 373 



tortoise), and the islands in the Western Indian Ocean, namely 

 the Mascarenes (Bourbon, Mauritius, and Eodriguez), the Comoros, 

 Aldabra, the Amirantes, and the Seychelles. When they became 

 extinct in Madagascar is not known, but T. grandidicri was a 

 very large species of apparently very recent date. Of the other 

 islands the Comoros only were inhabited by man, the others were 

 devoid of any but small and harmless Mammals. It was on these 

 peaceful islands that large tortoises lived in incredible numbers, 

 and, like the Dodo of Mauritius and the Solitaire of Eodriguez, 

 grew to a size far beyond that of their less favoural)ly placed 

 continental relations. The same applies to the tortoises of the Gala- 

 pagos Islands. Plenty of food, a congenial equable climate, and 

 absence of enemies enabled them to enjoy existence to the fullest 

 extent. There was nothing for them to do but to thrive, to feed, 

 to propagate, to grow, and to vary. At least there w^as nothing 

 to check variation within reasonable limits. Scattered over the 

 many islands, they were prevented from inter-breeding, and thus 

 it has come to pass that not only every group of islands, but in 

 the case of tlie Galapagos almost every island, has or had its own 

 particular kind, be these called varieties, races, forms, or species. 



There are four features of special interest. First, these tortoises 

 grow to a large size, and there are no small species on any of 

 these islands. Secondly, they vary much individually. Thirdly, 

 each island or group of islands has developed its own kind. 

 Lastly, there is the widely spread tendency to reduce the thickness 

 of the bony plates of the carapace, in spite of its size. In some 

 cases, notably T. vosmaeri of Eodriguez, the bony shell is reduced 

 to apparently the utmost limit compatible with mechanical safety. 

 The horny shields are, or were, however, well developed, sometimes 

 much more so than in other recent land-tortoises. Whatever were 

 the original reasons for the development of a strong shell in 

 tortoises, they cannot have prevailed in these islands. 



Where did all these tortoises come from, and how did they 

 get to these oceanic islands ? Accidental transport or migration 

 are out of the question. Land-tortoises are drowned within a 

 few hours. Moreover, there are none of their kind on the con- 

 tinents of Africa, Asia, and South America, although they had a 

 much wider distribution in past geological ages. Consequently 

 we have to assume that they are descendants of tortoises once 

 populating the land which, except the islands, lies now below 



