EVOLUTION IN A DETERMINATE LINE. 1 09 



the opercular flap, i. e., the opening valve, is prepared progres- 

 sively, so that the embryo can escape in due time. 



X. /;/ its Provision for the Latest Embryonic Characters in 

 Different Species. The egg capsules of four species of Chimara 

 known to the writer, present distinctive characters, and among 

 these are some which are prophetic of the structures of the adult 

 fish. Thus the tail sheath of the capsule of C. colliei is thick and 

 relatively short, fitting the tail of the embryo of a species in 

 which no adult opisthure is noticeable : in C. mitsukitrii, on the 

 other hand, the opisthure of the adult is of extreme length, and 

 the tail sheath of its egg capsule is correspondingly long and 

 delicate. The egg cases of C. phantasma and C. monstrosa sug- 

 gest correctly intermediate conditions in the adult opisthures. 



To return to the theme of the present paper : A point of gen- 

 eral interest presented by this complicated capsule is its bearing 

 upon the factors of evolution. For, considering always that the 

 substance of the capsule is only indirectly connected with the 

 egg, i. e., as a secretion formed by the parent after the mechanism 

 of heredity has already been established in the egg, it neverthe- 

 less (i) "foresees" with startling exactness the size and shape 

 of the young fish when many months hence it comes to hatch 

 out, and (2) it provides a series of progressive modifications 

 adapted to the developing physiological needs of the young. It 

 is evident, accordingly, that if natural selection be adduced to 

 explain the present phenomenon it encounters difficulties more 

 numerous and complex than in usual instances. In the latter 

 cases selection concerns itself with variations which affect the 

 progeny directly ; but in the present case variations must have 

 occurred in the lines both of the progeny and, indirectly, of its far 

 less individual capsule-forming capabilities with the result that 

 a succession of closely correlated steps in variation must have 

 coincided in both distinct directions. 



To make more apparent the complicated nature of this proce- 

 dure an example might be devised in which numbers appear. 

 Thus : At each point in the development of the young Chiintzra 

 let the number of important fortuitous variations be represented 

 by 1,000 (which I think all will agree is a number of possibilities 



