spread over all the tropical regions ; but each species of one region 

 exhibits a certain number of characters common to species of the 

 same region, and different from those of any second, so that we can 

 at once refer them to their native country. From this circumstance 

 the naturalist, in my opinion, is justified in attributing a generic value 

 to such characters, although they may be subject to variation in 

 another genus, or even in one and the same species. A variation in 

 the number of the rows of scales in the genera Zamenis, Tropidono- 

 tus, &c, is of trifling value only, whilst in other genera the relative 

 number of scales is a constant character (e.g. Calamaria, Oxyrhopus). 

 The number of the ocular shields in species of Zamenis, Pituopkis, 

 Tropidonotus, varies much ; but in the species of Simotes, Liophis, 

 Dromicus, &c. the number of these same shields is part of the generic 

 character. Thus we see that in one group of animals a part of the 

 organization is constant, and forms a positive character, whilst in 

 another group the same part is liable to very considerable modifica- 

 tion : and this peculiarity (liability to modification) itself constitutes 

 a character of the group. In every instance where naturalists neglect 

 to make this distinction, and try either to allow modification of a 

 naturally fixed character, or to fix modifications of a naturally variable 

 character as positive ones, artificial groups of animals are established. 

 For instance, the East Indian Simotes is a most natural genus, distin- 

 guished by the form of the rostral shield, by the peculiar system of 

 coloration, and by a single anterior ocular — three naturally fixed 

 characters. Now, if we attempt to force the African Snake (Hete- 

 rodon diadema) into this genus, we must actually modify two of those 

 characters, and the result of this combination is the establishment of 

 an artificial group * . The same is the case if we try to combine this 

 type of a separate genus with Heterodonf. On the other hand, two 

 East Indian genera, Oligodon and Simotes, form together a most 

 natural group, having the same arrangement and shape of the shields 

 of the head, the same form of the body, and even the same system 

 of coloration — three naturally fixed characters ; but the dentition in 

 these two genera of snakes is liable to variation. Now, if we con- 

 sider the latter character a fixed one, so as to become that of fami- 

 lies, the result is the artificial distant separation of those closely- 

 allied genera from one another. 



I have made these remarks to show, that to obtain a natural 

 system of genera, we cannot rigidly adhere to certain preconceived 

 constant characters, but we must apply them as they are made 

 obvious by nature itself. Let us apply these views to the illustration 

 of the genus Flaps of Wagler. Here we see that all the East Indian 

 species are distinguished from those of the other parts of the world 

 by a more slender, vermiform body, and by two longitudinal series 

 of scales fewer in number than in the others. As Wagler rightly ob- 

 served, the ornamental colours % on the upper parts of the body have 

 a tendency to form longitudinal bands {corpore vittatd) ; on the 



* Catal. Colubrine Snakes, p. 26. f Dum. et Bibr. vii. p. 26. 



% In contradistinction to the ground-colour. 



