PRIMARY GROUPS OF THE SERPENTS. 14? 



deadly of all serpents, namely, the Crotalidce or rattle- 

 snakes^ a small but very remarkable gi'Oup_, celebrated 

 for the subtlety of their poison. The rattle-snakes of 

 America are probably typical of this division^ aad are 

 the only ones^ in fact, which possess the instrument so 

 called, and to which the name can be properly applied. 

 As these reptiles are frequently brought to this country 

 alive, they need not be particularly described. We shall 

 merely remark^ in this place, that these and such others 

 as belong to this sub-family, are weU distinguished from 

 the preceding by having a round hollow hole^ or per- 

 foration^ behind each nostril : the rattles increase with 

 age, and an additional one is gained each time the rep- 

 tile casts its skin. All the species are natives of Ame- 

 rica. They move slowly, and seldom bite unless provoked, 

 or when seizing their prey : the scales on the head are 

 similar to those of the body, but in the sub-genus Cro- 

 tolophorus these plates are large. The Trigonocephali, 

 or angular-headed serpents, are equally venomous with 

 the rattle- snakes^ but are destitute of any caudal appa- 

 ratus for occasioning noise. Some of these, as the 

 genus Tisiphone, have the tail terminated by a sting, 

 and the genus Craspidocephalus is known by the sub- 

 caudal scales being double_, and those of the head and 

 back being similar : other peculiarities belong to some 

 of the species, but they are of minor importance. We 

 pass, therefore, to the Platurin(E, or flat-tailed Vv^ater 

 serpents, placed by M. Cuvier in this family, on account 

 of their teeth j by which, as he justly concludes, they 

 are rendered essentially different from the HydropliidcE, 

 or true aquatic serpents already mentioned : in these the 

 head is enveloped with plates, and those of the tail are 

 double, this part being compressed in the form of an oar : 

 hence it is probable that this is the point of union be- 

 tween the HydrophidcB and the CrotalidtB. After these 

 are placed certain genera (whose true affinities, however, 

 are very ambiguous), which agree with the vipers, ex- 

 cept that their subcaudal plates are more or less simple ; 

 these genera are Trimesurus, Ophicephalus, and Acan- 



L 2 



