670 BASHFORD DEAK, 



rapidly than in Lepidosteus^ yet not nearly as rapidly as in the 

 Teleost. The shortening in the duration of the larval period, of which 

 the transitional Ämia provides so interesting an example, is clearly 

 a developmental result which could hardly have been suggested on 

 a priori grounds. For it certainly seems far more consistent with 

 our preconceived ideas that the development of the older and simpler 

 forms ^) should be briefer than that of the forms derived from them : 

 but in this case (Ämia) we are given a well marked instance that 

 the later forms have become specialized in the acquisition not merely 

 of adult characters, but of developmental processes as well ; that they 

 have tended in their evolution to acquire their adult characters at an 

 earlier period, to elaborate their conditions of yolk nutriraent so that 

 the growth processes might be shortened, and that the larva should 

 more speedily be rendered seif maiutaining and seif protecting. The 

 shortening of the developmental, especially the larval period in the 

 more recent forms, appears to bring with it a further and most 

 interesting feature: for the shortening of the growth processes has 

 produced a need for masking the stages of transformation. It would 

 thus seem that a directive energy were making itself feit within 

 tissues apparently undifferentiated, and giving them their definitive 

 uses, long before they could have had the time to have performed even 

 a part of the changes through which in phylogeny they have passed, 

 A comparison of the larvae of the Ganoids and Teleosts, Siluroids 

 especially, briugs to light with surprising clearness how perfect a 

 ränge they present in transitional characters. Taking into accouut the 

 forms of the early larvae, pigmentation, mouth and auus, gill characters, 

 opercula, fin diÖ'erentiation, yolk absorption, brain and seusory struc- 

 tures, all of these diöerent types may readily be arranged in a 

 gradatioual scries. At the base of this series Stands Lepidosteus, near 

 it and in some ways even below it is Äcipenser, next is Amia, uext, 

 and very closely related, is Ämiurus, and finally are the many remain- 

 ing forms of Teleosts. 



Zoological Laboratory of Columbia College, 

 New York City, April 21, 1896. 



1) In spite of the fact that among fishes it has generally been re- 

 cognized that the shark's period of immaturity is longer than the lung 

 fish's, the lung fish's than the Ganoid's, the Ganoid's than the Teleost's. 

 This time relation might readily piovo a caenogenetic one, dependent, 

 for example, on richer yolk-supply. 



I 



