322 DE. HANS GADOW ON EEMAINS OF SOME GIGANTIC 



more in general the Aldabran species. I feel inclined to associate this specimen with 

 the other remains of T. sumeirei. 



It has to be borne in mind, however, that the bones of the shoulder-girdle of all 

 these extinct Tortoises are subject to a very great amount of variation in size and 

 shape ^. It would not be difficult to select out of the extensive material at our disposal 

 at least half a dozen different types, provided the intermediate forms were neglected 

 or suppressed. 



Phalanges. — Three large terminal and seven middle and proximal phalanges. 



Skulls (PI. XLIII. figs. 6 ff-8). — Mr. Sauzier's collection contains 19 skulls and two 

 mandibles. Two of these skulls and one pair of underjaws belong to T. triserrata 

 according to Dr. Giinther's definition. Six skulls and the other pair of underjaws 

 agree with those of T. inejyta. The remaining eleven skulls differ from those of 

 T. triserrata and T. inepta chiefly in the shape of the ventral surface of the long supra- 

 occipital crest. This surface is broad, triangular, and concave, while it is narrow and 

 ridge-like in T. inepta, narrow and doubly ridged (or, in other words, with a narrow 

 longitudinal groove) in T. triserrata. 



However, all these skulls exhibit a considerable amount of individual variation in 

 their general aspect, slope and size of the crest, relative strength of the various 

 parts of the skull, naso-frontal profile, &c. Most of them approach to a slight 

 extent the Aldabran skulls by the convexity of their frontal region, and, according to 

 Dr. Glinther, the " posterior margin of the paroccipital crest is deeply excised " in 

 T. daudini. Whether these variations are due to age or sex, or are of specific or 

 subspecific value, cannot be determined. It has to be borne in mind that we have no 

 criterion whatever by which we can associate any of these numerous skulls with any 

 particular form of carapace, plastron, or limb-girdle. It is quite possible that the 

 typical skulls of T. triserrata belong to the carapaces which have been distinguished 

 as those of T. inepta, or to T. sumeirei, or vice versd, and the same remark applies with 

 equal force to the various sorts of pelves. 



Unfortunately, this uncertainty is inevitable, because, owing to the circumstance that 

 many of the bones from the Mare aux Songes had to be fished out of a morass just as 

 the labourers happened to come across them, no record of the juxtaposition of the 

 various bones could be, or at any rate has been, made. Until, by happy chance or by 

 a much more careful and extensive mode of research, all the principal parts of one 

 unquestionable individual are found, the association of these bones will be a matter 

 of speculation without any valid basis. 



" Habt alle die Theile in der Hand, 

 Fehlt leider nur das geistise Band." 



' See A. C. Haddon, Trans. Linn. Soc. ser. ii., Zoology, vol. ii. (1879) pp. 156-158. 



