ON" THE ORIGIN OF GENERA. 93 



a little before. The identity of Didocus with the undeveloped 

 Pelobates is thereupon lost ! 



So may have been the relations between Pelobates and Cul- 

 tripes. Pelobates was probably once identical with the undevel- 

 oped Cultripes ; but the same acceleration has concentrated the 

 characters more rapidly than the other larval stages, leaving Pelo- 

 bates behind. 



This I conceive to be the explanation of this relation : when 

 the i^arallelism is inexact by two steps, as in Spea to Didocus, by 

 the obliterated ear and ossified xiphisternum. The continued 

 concentration of characters has been carried to earlier stages till 

 the identity exists in the adult state of neither one, but at a pe- 

 riod of larval life of both, shortly preceding the adult period of 

 the lower. The relations between the Amblystomidie and Pletho- 

 dontido?, which I have elsewhere * jDointed out, have probably 

 had their origin in this way. 



If we attempt to prove the identity of the modern Mammalian 

 foetal circulation with that of the modern adult fish, we may find 

 nearly an exact parallel in this respect, as it is the basis of class 

 distinction ; but in other respects the identity will not exist, ren- 

 dering the parallel inexact or remote. The structure of the ori- 

 gins of the aorta is at one time identical with that of the shark, 

 with one exception : in the former but four aorta-bows appear to- 

 gether ; in the latter five. In the former the first disappears as 

 the fifth comes into being. This is simply a continuation of ac- 

 celeration. The first generalized representative of the Mammalia 

 lost the first aorta-bow toward the latter part of its growth, and 

 became the next genus in advance of the selachian. The fact 

 that these bows do not appear exactly simultaneously, but rather 

 successively, renders it necessary that in a regularly shortening 

 period of possession of transitory characters, one such, as the ex- 

 istence of the first aorta-root, should vanish before the appearance 

 of a permanent, the fifth, in the more specialized types, where 

 acceleration reaches its maximum. Tliis is indicated by the fact 

 that in the Batrachia, where the acceleration has not attained so 

 high a degree, the first and fifth aorta-bows co-exist for some time, 

 though the first and second disappear before maturity. 



So also with the splitting of the bulbus arteriosus. As in the 

 Batrachia, the pulmonary ductus communis only is to be sepa- 



* "Jour. Ac. Nat. Sci.," Phil., 1866, p. 100. 



