NATURAL SCIENCES OF PHILADELPHIA. 273 



specific* and that one and the sam^ species (if origin be the definition,) has, in the 

 natural succpssion, existed in more than one genus. 



Apart from any question of origin, so soon as a species should assume a 

 new generic cliaracter it ceases, of course, to be specifically the same as other 

 individuals which have not assumed it. If supposed distinctuess of origin be, 

 howevei-, a test of specific difference, we shall then have to contend with the 

 paradox of the same species belonging to two dift'crent genera at one and the 

 Same time. 



It follows, therefore, in our interpretation of nature, tliat groups defined by 

 coloration alone are not to be regarded as genera, as is done by some orni- 

 thologists and entomologists. They are simply groups of species in which 

 distinctive generic characters had not appeared up to the period of repro- 

 duction. Inasmuch as in development certain specific characters appear first, 

 among them part or all of the coloration ])attern, it is obvious that the latter 

 do not belong to the generic category. The employment of such characters 

 then, in this sense, is only to commence reversing the terms generic and 

 specific, and to inaugurate the process of regarding each species as type of a 

 separate genus. 



(i. Of probable cases of transition. 



Thus the transition between the toothed and edentulous conditions in Ceta- 

 cea takes place in the ordinar}- growth of the individuals of the genus Glo- 

 biocephalus, and the transition between the ossified and non-ossified types of 

 Clielonia occurs during the life of the individuals of the genus Dermatemys. 



But in attempting to demonstrate this proposition we must bring forward 

 facts of another kind. The anti-developmentalists are accustomed to put 

 such changes aside, as part of the necessary history of established types ; hence 

 ■we will not appeal to such. 



1. The frog Ranula affinis, of South America, was described by Peters 

 as probably a climatal variety of European Rana temporaria. In this 

 he is supported by the fact that the specific characters do not differ more 

 than would characterize it as a local variety, were it an inhabitant of Europe. 

 But I have found that it differs generically in the non-ossification of the 

 ethmoid boue, as has been confirmed by Steindachner, and represents an 

 embryonic condition of the same bone in Rana. It is in fact an undeveloped 

 Rana. That this is a true genus is confirmed by many specimens, by an addi- 

 tional species (R. palmipes), and by the fact that the allied genus Trypher- 

 opsis, embracing three species in the same region, differs in the same way from 

 the otherwise identical genus of the Old World, Hylorana. 



2. The South African Saurians Chama?saura a n g u i n a, and Mancus m a- 

 c r o 1 e p i s, are very closely allied in specific characters in all respects, though 

 distinct. They have one important ground of generic distinction; the latter 

 has one pair of limbs less than the former. They are rudimental in CiiamiC- 

 saura, and the disappearance in Mancus is but another step in the same di- 

 rection. The difference in specific characters is of much less degree. 



3. In the genus Celestus there are numerous species, which range from a 

 slender, snake-like form with weak limbs, to stouter, strong-limbod forms with 

 a more saurian build. Among these the Haytian C. ph oxin us is well dis- 

 guished by form and coloration. An allied genus from the same region is 

 Panolopus, which in specific characters approaches the 0. p h o x i n u s very 

 closely, much more so than any Celestas (one sijecies possibly excepted). But 

 in generic characters it is distinguished by the loss of all its toes and the non- 

 separation of nine plates on the end of the muzzle. The genus Diploglossus, 

 on the other hand, occupying a superior place on account of the division of 

 the frontonasal into three, is, in specific characters (of D. m o n o trop i s) much 



I * See Proceedings Academy, 1807, p. 86, where I observe that generic characters are 

 probably less inherent than specitis. 



1868.] 



