NATURAL SCIENCES OF PHILADELPHIA. 249 



among nearest allies. Therefore emlnyologic indications are through out use- 

 less in classification, and it is necessary to lieep carefully separate the state- 

 ments of observations on development of a given species, and not transfer 

 such facts to the history of another species for the purpose of completing it. 

 We cannot even range these histories in conformity with family groups." For 

 us this statement, though no doubt largely true, is an indication of imperfec- 

 tion, first, of knowledge of true affinities of recent, but especially of extinct 

 adults, and second, of imperfection of knowledge of development. The posi- 

 tion appears to be based on negative evidence, while the opposing can and 

 does stand on nothing but positive. 



/?. Examples of (he inexact parallelism.. 

 1. The genera of the Batrachian family Scaphiopodidte form a series of steps 

 differing a little more than as repressions or permanent primary conditions 

 in the development of the highest.* Thus two of the genera, which are North 

 American, maintain their tul)ie eustachii and tympanum through life, while 

 three European lose them at an early period. The three European genera also 

 advance beyond the larval character of the American in the ossification of the 

 basis of the xiphisternum into-a broad style. Thus we have two series estab- 

 lished, which differ only in the two characters named. Each shows its devel- 

 opmental steps iu a similar manner, the European series extending further; 

 thus, — 



European. North American. 



1. Temporal fossa over arched. 



Cultripes. * * 



Temporal roof not ossified. 



2. Fronto-parietal bones ossified, involving 



derm. 



Pelobates. Scaphiopus. 



3. Fronto-parietals ossified, distinct from 



derm. 



* * (Unknown.) * * 



4. Fronto parietals not -ossified, distinct 



from derm. 



Didocus. Spea. 



In this case Didocus cannot be said to be identical as a genus witli an un- 

 developed stage of Cultripes, since while the cranium of the latter is iu tlie 

 condition of Didocus it bears a long tail, and the limbs are but little developed. 

 Nor is Didocus identical with the undeveloped condition of Pelobates, since 

 both cranium and limbs of the latter are developed before the tail is absorbed. 

 Nor is Pelobates identical with the undeveleped condition of Cultripes, since 

 while the cranium of the latter is that of the former, the limbs and tail are still 

 larval. The same relations exist between the other members of the family. 

 The genus Scaphiopus is not an undeveloped form of Pelobates as to its audi- 

 tory organs, for when the latter is identical with the former in this respect, it 

 bears otherwise entirely larval characters. Nor is Spea an arrested Scajihio- 

 pus, the relation being here precisely that between Didocus and Pelobates. 

 Spea approaches more closely an arrested Didocus in all respects, but that 

 when the latter posesses the auditory apparatusf of the former, it is a larva in 

 limbs and tail, and that it loses this apparatus before reaching the other 

 characters of Spea. Tne relations of these genera, as compared with those of 

 the Trachycephalus, Cystignathid;e and Bufu series, may be represented as fol- 

 lows : the lines represent the developmental scale of each. 



* See Journal Academy, 1800, on Arcifera. 



tThe possession of caVum tympani and tuba Eustachii in tlie undeveloped condition of 

 this genus is only assumed from its close relation to Pelobates. 



J According to"Brucli and Tschudi in Pelobates. I have found traces of the eustachian 

 diverticula in a tailed Pelobates f us cus, whoseT&ody measured 1 in. 4 liu., from Mus. 

 Peabody Institute, Salem, Mass. 



1868.] 17 



