276 FOSSIL REPTILES. 



genera, and of the sutures which unite their bones, fully 

 convinced him that the remains in question belonged to the 

 subgenus on which we are now writing. The description of a 

 carapace found in the quarries of the village of Melsbroek will 

 prove this. Its contour is oval, a little narrowed behind, but 

 not more so than in the emys centrata. The ribs unite unin- 

 terruptedly with the marginal pieces, as they do in all the 

 emydes and land-tortoises. The curve of these ribs is pretty 

 nearly the same as in the emys centrata. The vertebral plates 

 are singularly narrow, more so than in any living emys, and 

 even than in the fossil emys of Sheppey, which we have just 

 spoken of. This peculiarity may be remarked in this speci- 

 men, that the seventh and eighth ribs are each of them united 

 to their opposites, between the eighth and ninth vertebral 

 plate. This is a circumstance which also takes place in 

 relation to the seventh pair in the emys of Sheppey, and which 

 in the eighth pair, but only in the internal face, is found in the 

 emys centrata, but is just the same in the emys expansa, as in 

 the specimen of which we now speak. 



Comparing this carapace with that of any sea-tortoise of the 

 same size, a specific character, very strongly marked, is in- 

 stantly discovered. The fossil tortoise has the intervals of its 

 ribs completely ossified, and no vacancy remains between them 

 and the pieces of the edge, which are also much more broad 

 in proportion than those of a sea- tortoise. In the testudo 

 mydas, for instance, at the age when its carapace is no more 

 than thirteen or fourteen inches in length, a vacancy remains 

 between the ribs, not ossified, which almost equals one-half of 

 the length of the rib. A part of this vacancy has been found 

 remaining, even in an individual whose carapace was nearly 

 four feet in length ; and it has been verified upon many of a 

 size intermediate between the last two mentioned. All this 

 serves to prove that the fossil tortoises of Melsbroek cannot 

 belong to testudo mydas ; nor in fact to any marine tortoise, 

 for the ossification does not take place more rapidly in any of 



