1863.] CRASPEDOCEPHALI OF BAHIA. 53 



any better characters, and still persist in considering them individuals 

 belonging to two species, having no other basis for their separation 

 than the frequent occurrence of C. atrox in Guiana, whilst the other 

 species is never found there. 



Comparing my three specimens, which differ from those of C. atrox 

 in the last-mentioned respects, and first as regards the size of the 

 superciliary shields, I cannot come to any very precise decision, as 

 they are all not full-grown. Comparing with one another old and 

 young specimens of C. atrox, I find that not only the superciliary, 

 but all other head-shields are proportionately larger in young indi- 

 viduals, so is the pit in the cheek ; and the whole head is flatter, 

 especially the occiput, and more elongate in adult specimens. I 

 compared the three specimens with those of corresponding size of 

 C. atrox, but I could not arrive at any decided opinion ; and, con- 

 sidering the difference in size of the figures in Schlegel's plate 19, 

 they also do not allow me to draw any safe inference from the rela- 

 tive size of the superciliary shields in each species. Besides, I 

 am not acquainted with the absolute size each sj)ecies may attain. 

 As regards the size and number of the mental shields, I cannot 

 find any very striking difference ; in some specimens of C. atrox I 

 have found one, in others two, and even three pairs of chin- shields ; 

 in the three specimens which differ in other respects from them, 

 I always found only one pair. The labial shields are certainly 

 smaller in my three specimens which do not agree with C. atrox. 

 But more striking still is the shape of the scales and their keel. 

 The three specimens I am inclined to regard as referable to C. brasi- . 

 liensis have narrower scales, their keel lower, narrower, longer, and 

 reaching to their tip. At first glance these specimens have a less 

 hirsute appearance than those of C. atrox. In accordance with the 

 narrowness and the smaller number of their scales, their body ap- 

 pears more slender. 



I am well aware that the coloration does not afford safe specific 

 characters, except in comparatively few instances ; but as all the 

 specimens I referred to C. atrox agree so well in this respect, dif- 

 fering from my three supposed C. brasilienses, which again agree 

 among themselves, I may be allowed to state in what one and the 

 other differ. The specimens I refer to C. atrox are all greyish yel- 

 low or olive, and have along the body irregular brown, black-edged 

 spots with sinuated margins, which occupy about as much space as 

 the ground-colour. In yoimg specimens the colours are generally 

 brighter, and the spots more distinct. Underneath they are all, with- 

 out exception, chequered with dark grey or black. 



The three specimens of supposed C. brasiliensis are olive-green ; 

 similar brown, black-edged spots, with sinuated margins, occupy 

 their back, but occur at much wider intervals, so that they occupy 

 much less space than the ground-colour ; underneath, all three are 

 dirty-yellow, punctulated with black, but not at all chequered. 



These differences appear very striking, but I refrain from attach- 

 ing undue weight to them. Schlegel describes some specimens of 

 C. brasiliensis with "larges taches carrees " (Essai, ii. p. 533). 



