FRYER— FORMATION OF ALDABRA, ETC. 421 



the eggs were found sub-fossil in the phosphatic rock underneath the guano. At 

 Assumption bones of several specimens were found in the pits in the rock, and they also 

 seem of great age. The bones in the rock of Aldabra are deeply imbedded and cannot 

 have been fossilised very lately, and consequently if human transportation is to account 

 for the distribution of the tortoises it must have taken place in very early times. 



The difficulty then arises that we know of no nation which could have sailed the 

 Southern Indian Ocean in such early times and have acclimatised land tortoises on the 

 majority of its islands. There is therefore this extraordinary problem to be solved, viz. 

 that most of the islands of the South Indian Ocean, whether of coral or not, were 

 inhabited by giant land tortoises which, to quote Dr Gadow (Comb. Nat. Hist. Reptiles), 

 " are so absolutely terrestrial that wherever we find them, unless they have been brought 

 artificially by man, these tortoises must have arrived by land." It has been shown that 

 neither of Dr Gadow's alternatives will satisfactorily account for the facts. Tortoises are 

 not the kind of animal to be distributed by drifting on logs, unless they belong to a small 

 species. If the "giants" are derived from a small species distributed in this way then 

 they show a wonderful case of parallel evolution, and further, if this is so, small tortoises 

 must have been drifting about the Indian Ocean in considerable numbers to account for 

 the colonisation of so many islands. Finally, it is generally held as impossible that land 

 tortoises could have had aquatic ancestors*. All means of escape from the problem are 

 therefore withdrawn and any explanation must be liable to great objections. Nothing 

 further can therefore be said, and it remains to conclude with a short account of the 

 natural history of the tortoise. 



The day is spent in the midst of the densest jungle or in a large clump of Pandanus, 

 a very favourite haunt being a clump of this tree near a brackish pool. At night they 

 become active and move about in search of food, which in the wet season consists of grass 

 and leaves, and in the dry of bark, if nothing else can be obtained : as a rule their move- 

 ments are very slow and deliberate, but at times, perhaps when moving to fresh feeding 

 grounds, they crash straight through the jungle at a pace which is astonishing for so 

 clumsy an animal. 



The breeding of the tortoises takes place about March, and the hard-shelled eggs are 

 buried in small guano-filled concavities in the rock. As the atoll was left at the end of 

 January only one nesting-place was found, and this contained no eggs, though the guano 

 had been obviously disturbed. Tortoises are never found far within the Pemphis bush 

 and as the latter appears to be increasing at the expense of the more open country it may 

 in time seriously restrict them. Possibly the present greatest enemy of the tortoise is the 

 rat, which is reputed to destroy the eggs. Frigate birds and herons would undoubtedly 

 devour the young ones if accessible, but the thick bush affords abundant protection. 

 There has recently been a suggestion put forward to remove the tortoises to some 

 Seychelles island for safety: it is, however, quite unnecessary, for tortoises are already 

 plentiful in the Seychelles and there does not appear to be any danger of their extinction 

 on Aldabra unless hunted by man. 



Of other chelonians the green turtle (Clielone my das), the hawksbill {Chclone 



* Vide Geographical Journal, Sept. 1910, Discussion on the "South- West Indian Ocean." 

 SECOND SERIES— ZOOLOGY, VOL. XIV. 54 



