112 MR. G. A. BOULENOER ON THE OSTEOLOGY OF [Jan. 20, 



more or less developed, in the Ophidians. As ohserved by Shufeldt, 

 the bony dermal tubercles of the head are so intimately adherent to 

 the bones that it is extremely difficult to remove them in adult 

 specimens: but in the half-grown skull oi H. horridum, which is 

 quite as large as that of the adult H. suspecfum, they could be de- 

 tached very readily, and the skull shows all the sutures perfectly 

 distinct, as correctly figured by Gervais and by Bocourt^ I may 

 add that bony tubercles are also present on the dorsal and lateral 



Fig. 3. 



Skin of tlie iiiidtlle part of the bodj^ of H. ^iiHpectnm, epiderinis removed, 

 showing the bony tubercles; nat. size. 



surfaces of the bodv, but wide apart and gradually diminishing in 

 size towards the ventral surface, which is entirely devoid of dermal 

 ossifications or shows mere ring-like traces of them, except on the 

 prseanal region, which, like the i)ack, is studded with bony tubercles. 



Vertebral Cohnirn. 



This consists of 8 cervical, 26 dorsal, 2 sacral, and 40 caudal 

 (z='/Q^ vertebrae in H. horiidum, of 8 cervicals, 26 or 28 dorsals, 2 

 sacrals, and 27 caudals (=63 or 65) in H. suspectum. In the (2 ) 

 specimen of the latter species, with 28 dorsal vertebrse, both centrum 

 and arch of the 21st vertebra are anchylosed with the 22nd. 



I was anxious to examine the atlas, as I had been much struck 

 by Dr. Shufeldi's statement (P. Z. S. 1890, p. 214) that it is " com- 

 posed of five separate pieces ; three of these are devoted to the for- 

 mation of its anterior cup for the cranial condyle. Of these three 

 pieces, one is a mid-ventral one, while either of the others are ventro- 

 laterally situated. Each side of the neural arch is formed by one of 

 the two of the remaining pieces of the five of the component ele- 

 ments of this vertebra ; and in a large specimen of this lizard none 



1 Miss. Sc. Mex., Rept. pi. xx. e. 



