372 



FOSSIL TURTLES OF NORTH AMKRICA. 



more than a few common species. Along the western coast of Central and South America, in 

 situations where they are liable to be carried down by rivers to the Pacific Ocean, there live 

 about 50 genera of lizards. How can it be explained that such a fragment ot these has reacht 

 the Galapagos Islands ? 



An instructive chart, publisht by Dr. Alexander Agassiz (Bull. Mus. Comp. Zool., xxm, 

 pp. 56-74, plate iii) exhibits the floor of the ocean in that region, and demonstrates the presence 

 of a tongue of uplift which extends from the Galapagos Islands to the Central American coast 

 just north of the Gulf of Panama. It seems therefore probable that at some time in the past, 

 evidently after the Oligocene, more probably during the Miocene, there existed a mass ot land 

 which included the Galapagos Islands and joined narrowly the Central American coast. This 

 land mass might be compared to the island of Madagascar, with the long axis at right angles 

 with the neighboring continent, instead of parallel with it. Over the narrow bridge there 

 past the ancestors of the present turtles of those islands and a few genera of lizards, together 

 with some other animals and sundry plants. The descendants of these invaders have differen- 

 tiated and formed a part of the present fauna and flora of the islands. It is not unlikely that 

 some genera which invaded that land perisht afterwards. 



When we try to account for the presence of the gigantic tortoises found at one time, but now 

 nearly extinct, on the islands of the Indian Ocean, we are met with the same difficulties that 

 we have been considering. If the conformation of the land and sea was the same as now when 

 the ancestors of these tortoises reacht their island homes, they must have been transported over 

 the seas for distances varying from 250 to 750 miles. 



The solution of the various problems connected with the distribution of animals and plants 

 in southern Africa and southern Asia and the islands of the Indian Ocean have driven many 

 geologists, zoologists, and botanists to the conclusion that at one time there was a land connect- 

 tion between India and South Africa, across the Indian Ocean. For a masterly discussion 

 of this question the reader is referred to the presidential address of Mr. W. T. Blantord before 

 the Geological Society of London in 1896 (Quart. Jour. Geol. Soc. Proc, p. 98), and also to 

 Giinther's address already cited. Blanford concludes that there was such a land connection 

 as has been mentioned above, that this continued thruout the Upper Cretaceous, and was 

 broken up into islands at an early Tertiary date. Dr. Gunther adopts this conclusion and adds 

 that "the slow evolution of this chelonian type (Testudo) which has scarcely changed since 

 the Eocene, and its wide distribution over the Northern Hemisphere, justify the supposition 

 that it was in existence already before the Tertiary, before the bridge was broken through 

 which allowed ot its passage southwards or northwards." 



That primitive Testudinidae were in existence toward the close ot the Upper Cretaceous is 

 very probable, but that the genus Testudo had at that time made its appearance there is no 

 proof, no probability. We do not know of its existence with certainty before the Upper Eocene 

 in Africa and the lowest Oligocene in America. The less advanct forms of the family, as 

 Hadrianus and related genera, may have reacht India or South Africa by the end of the 

 Cretaceous, but we have no evidence of this. Hence, unless we are willing to betake ourselves 

 to what ought to be our last refuge and hold that Testudo had a polyphyletic origin (using this 

 term in its original sense), that the Testudos of America and those of the eastern continents 

 sprang independently from Hadrianus or from the latter genus and another, we must place at 

 a later date than the earlv Eocene the arrival of the ancestors of the gigantic tortoises on Mada- 

 gascar and the islands of the western portion of the Indian Ocean. 



The conclusion reacht by Mr. Richard Lydekker regarding the time of separation of Mad- 

 agascar from the mainland was that this occurred during the Miocene (Geog. Hist. Mammals, 

 p. 223). Mr. Blanford had previously concluded that the depression of 1,000 fathoms or more 

 which led to the formation of the Mozambique Channel had taken place during Pliocene or 

 Postphocene times. However, the recent discovery of species that can hardly be separated 

 from Testudo, in the Fayum in Egypt, makes it necessary to recognize the fact that members of 

 the genus had reacht that continent as early at least as the Upper Eocene. 



Testudinidae at their various stages of development are known at earlier dates in North 

 America than elsewhere. It seems therefore probable that North America was their center of 

 distribution. From western North America thev may be supposed to have reacht Central 

 America and to have migrated thence to the present Galapagos Islands. From western North 



