REMARKABLE TORTOISE, TESTUDO LOVERIDGII. 515 



orthogenetic, brought about either gradually or as a mutation, and 

 that the furtive habit of hiding beneatli stones was the natural 

 result, since the tortoise no longer possessed adequate protection 

 from enemies. 



Possibly both principles come into play, the reduced armour 

 and loss of ribs being oi-thogenetic, and the depression and relative 

 condition of the vertebrae being svibsequently induced by the rock- 

 dwelling habit. 



Again, the metabolism of bone formation may be to some 

 extent influenced by climatic conditions, lack of lime in the 

 soil or water etc., or by their food. In this connection it is 

 interesting to note that Mr. Loveridge has recently found two 

 specimens of Cinixys helliana in the T. loveridgii locality which 

 have remarkably depressed carapaces (31 °/^ of total length as 

 against over 40%); and, further, Tornier describes one from 

 Bussisi, the T. tomieri locality, which has its bony plates 

 excessively thin, and actually fenestrated. Possibly this indi- 

 vidual, if let alone, might have founded a race of soft-shelled 

 tortoises in this genus also. 



The actual process of development of the bony plates of both 

 carapace and plastron in T. loveridgii is really quite normal ; in 

 spite of the astounding appearance which they present in the 

 adult, they are similar in general form to those of the young of 

 other species. In some, of course, one set of elements may 

 develop more quickly in proportion, altering the balance of the 

 fenestrse or their outlines, but the same process is gone through 

 by them all. 



On dissecting a series of the young of Testudo Jwrsjieldii, I find 

 that the very young (50 mm.) presents a bony network of 

 costal and neural plates, modified ribs, and undeveloped plastron, 

 similar to the young T, loveridgii (60 mm.). The two are 

 compared in text-fig. 2 ] . Series in other species, chosen at 

 random, show the same thing, with, of course, variations. In 

 some the dermal plates are more equalized, and in the early 

 stages present a less net-like form, each costal being kite- 

 shaped, rather as in the Turtles. In the Giant Tortoise 

 {T. nigrita) from the Galapagos, a very young specimen (95 mm.) 

 has the bony carapace still extensively fenestrated ; the costal 

 plates are more or less equally developed, each being dilated 

 beneath the vertebro-costal sutures, and from thence narrowing 

 to the marginal region, the fenestrte between are therefore 

 equal. The neural plates are broadly developed ; second, fourth, 

 and sixth are not in contact with the corresponding costals. 

 The general structure is therefore in essentials the same as in 

 T. loveridgii. In Freshwater Tortoises, such as Emys, development 

 is px'oportionately quicker beneath the vertebral shields, a speci- 

 men of 50 mm. having this part of the carapace entire, whilst 

 the costal region is only half completed. 



In the following section the development of the dostal and neural 

 plates will be dealt with further, in connection with their origin. 

 Proc. Zool. See— 1922, No. XXXYI. 36 



