( 1«'J ) 



On the whole the several vertebrae resemble much those of T. ahimjdoiui, but 

 the ossification of the neural arches is less extensive, although more solid, with 

 deeper incisions for the intervertebral foramina. The zygapophyses of the 

 atlciH (PI. XX, Fig. a) are very long, styliform to effect articulation with the 

 neural arch of the second vertebra, which is pushed backwards by the intervening 

 unusually large " odontoid process." The dorsal and ventral median profiles of 

 the second vertebra are nearly parallel, its outlines being thus very different from 

 those in T. abingdonii ; its haemal crest is also simple for the greater part of 

 its length, not bifid as in that species. The third certebra (PI. XX, Fig. b) has 

 a condijle behind as well as in front* thus somewhat exceeding in length the 

 fourth (Fig. c) which is concave in front like the other vertebrae ; there is no 

 conspicuous hollow behind the anterior zygapophysis in this vertebra or in the 

 fifth, whilst it is well marked in the sixth.\ The sixth has a rather low haemal 

 crest, which is highest in the middle and not towards the front, as in T. ahinydonii ; 

 the bi]iartition of the posterior condyle is incomplete. The summit of the neural 

 crest of the seventh vertebra (PI. XIX, Fig. u) is deeply excised and situated 

 in advance of the middle of the length of the vertebra, the anterior portion of 

 the roof of the central arch being shortened. The eighth vertebra resembles 

 much more that of T. ephippium (Nov. Zool. 1890, PI. XXII) than of T. 

 ubingdonii {Gig. Land Tort. PI. L.) ; especially the anterior condyles are trans- 

 versely elongate. 



Other differences in form are best described by giving in a comparative table 

 (on next page) the measurements of the vertebrae of T. galapagoensis, ephippium, 

 and abingdonii. The figures are in millimetres. 



Dorsal vertebrae. — From a comparison of the few skeletons which have served 

 for my previous notes on this part of the vertebral column (I T. elephantopus, 

 1 T. ficina, 3 T. ephippitnn), I am inclined to draw the conclusion that it does not 

 afford any reliable characters indicating the distinctness of the various island-races. 

 The longest are generally about the fourth and fifth, the posterior rapidly decreasing 

 in length. In the tnale individual of T. galapagoensis, two, the ninth and tenth, 

 jiarticipate directly in the formation of the protuberance for the articulation of the 

 ilium. This protuberance is continued forward into a ridge, which is reached by the 

 pleurapophyses of the eighth and seventh vertebrae. The eleventh and twelfth 

 vertebrae (PI. XXI) are reduced to the size and shape of the succeeding caudal, and 

 their j>]eurapoph3-ses articulate with the head of the ilium ; I enumerate them as 

 caudals. The thirteenth is the last vertebra, which has a free pleurapophysis. 



In tliti female specimen the nine anterior vertebrae are preserved. Tliey differ 

 considerably in their comparative length from those of the male, the posterior being 

 lengthened at the expense of the anterior. 



"* Me.^sure.ments or Duus.u. Vkktehrae. 



1st 'Jnd 3rd 4th 5th Gth 7th 8th 9th 10th 



(J 54 mm. OO mm. 93 mm. 100 mm. 97 mm. 73 mm. 50 mm. .30 mm. 20 mm. 14 mm. 

 ? 47 „ 61 ,, 67 ,, 09 ,, G:S ,, 45 „ 46 „ 30 „ '20 „ — 



^ .lackfeon already notices tliis (/.*-. p. KU), but without bcinj^ aware of the si.unilieanee of this 

 character. 



f An error has crept into the lettering of the four niiiUlle fitjures of I'l. XLI.K of Giijnnlir Luml 

 Tiirtoisoi. Tiic letters n and u should i)C reversed; the top and bottom ligures (side views of the liftii 

 and si.\th vertebrae) are correctly lettered. 



