NOVITATES ZOOLOGICAE XXII, 1915. 429 



2. Variation with Sex 



In the smooth-scute J T. yigantea the ? is alwa3's much more strongI_y dome- 

 shaped than the S, and never shows any signs of the scutes being embossed. 

 In the striate-scuted T. elephantina the ? is not conspicuously more dome-shaped 

 than the higher shelled cJc?, but it often exhibits scutes as highly embossed as 

 any i . In daudinii only one considerably dome-shaped and in qouffci no ? ? are 

 known ; while in sumeirei, if the specimen I have of a half-grown ? is reallv 

 sumeirei and not a hybrid, this sex is proportionally even more dome-shaped 

 than in gigantea, thus indicating a representative development in this group to 

 chathamensis among the Galapagos races. Like all the " Giant Tortoises," the 33 

 attain a much larger size than the ? ?. 



4. Individual Variation 



Here we are faced by a very awkward situation, due entirely to the inter- 

 breeding of the various races on the Seychelles and Mauritius, thus making it 

 necessary to include under one name a number of individuals which show differences 

 as great as, and often greater than, those between such Galapagau races as T. nigrita 

 and T. abingdonii. If we took the extremes, such as the very old S in the Tring 

 Museum which was brought back from the Seychelles by Dr. Brauer (see Plates 

 Liii., Liv.), and the large S from Cerf Island in the British Museum (see Plates 

 XLix,, L.), no one would believe them other than very distinct species ; for 

 Dr. Braner's 3 is long and comparatively narrow, has a considerably depressed 

 carapace, and the scutes show little or no signs of a boss — i.e., the edges of 

 the scutes are very little lower than the centre. The Cerf Island 3i on the 

 other hand, is short, very broad and very high, the carapace not depressed at 

 all, and the scutes enormously embossed, the centre of the scutes being 

 often 4-5 inches higher than the edges. Unfortunately we do not know the 

 exact habitat of these extremes, but we do know that a large number of 

 intermediate specimens between the extremes represent interbreeding in all 

 sorts of degrees between at least four species or races. The colour is also very 

 variable, running from pale-horn brown to deep black. The ? ? and young are 

 mostly dark slate-grey or brownish black, very seldom indeed being pale horn- 

 brown. Although, as proved by the above-mentioned large 3 from Cerf Island, 

 the adult 3 3 oi eleplmntina do occasionally become very pale horn-brown, we 

 generally find that the large cj cf of a decided pale brown colour are nearly alwaj's 

 entirely smooth-scnted, and therefore must be included inxAev gigantea. The brown 

 colour appear.':, however, to be found in long and narrow and short and broad 

 specimens promiscuously, but seems less frequent among those with embossed than 

 those with flat scutes, although the Cerf Island 3 in the British Museum is 

 one of, if not the most embossed specimens known. I have a much older 3 at 

 Tring, also brought from Cerf Island ; this individual is entirely black, deeply 

 striated, and has strongly embossed scutes (see Plates li. and lii.) ; but, 

 although being very aged, while the British Museum specimen was about 

 eighty years old only, it is much smaller. As examples of the short and long 

 forms of T. gigantea, see Plates lix., LX.'and lxi., lxii. In none of the races 

 of the group under discussion do we find the yellow jaws and throat, as found 

 in certain Galapagos races. 



