ORGAN-FORMING SUBSTANCES IN EGGS OF ASCIDIANS. 223 



ground for supposing that evolutionary changes first occur in the 

 end stages and only later affect the earlier stages in the life cycle? 

 In spite of the age-long controversy as to the inheritance or non- 

 inheritance of acquired characters there is no satisfactory evidence 

 that particular modifications of any adult part ever produce specific 

 modifications of the germ. All the evidence available seems to 

 show that the soma stands to the germ in the relation of environ- 

 ment and that the only influence exercised by the former upon the 

 latter is of a general character, as Weismann has so ably argued. 1 

 Furthermore the difficulty of conceiving of any method by which 

 adult characters might be transferred to the germ is well known. 

 No hypothesis ever yet proposed for the solution of this problem 

 harmonizes with the established facts of oogenesis and spermato- 

 genesis. If such transfer occurs, of which there is no sufficient 

 evidence, it can only take place by methods of which we are at 

 present wholly ignorant. 



On the other hand there is much to be said in favor of the 

 view that the germ is primary, the adult secondary and that 

 heritable modifications first arise in the germ and only later 

 appear in the adult. Apart from the fact that the germ gives 

 rise to the adult and to other germs, it is known that in 

 certain cases apparently slight modifications of the germ may 

 produce profound modifications of the adult, whereas the reverse 

 is not known to be true. One of the most convincing evidences 

 of the truth of this view is found in cases of cross breeding, par- 

 ticularly in hybridization, where it is certain that hybrid charac- 

 ters of an offspring are directly due to the hybrid character of the 

 germ, since they can have no other possible cause. The evidence 

 drawn from experiments on eggs and embryos on first thought 

 seems to be conflicting ; in some cases fragments of eggs or 

 embryos give rise only to partial larvae and injured eggs produce 

 only embryos showing more or less serious defects ; in other 

 cases entire embryos are produced under these conditions, but 

 these results cannot be regarded as destructive of this argument 

 for, as, has long been maintained by Roux, such cases of entire 



1 Undoubtedly one important cause of germinal variation is to be found in the in- 

 fluence of changing environment upon the germ, but this is far from saying that par- 

 ticular modifications of the adult are transmitted to the germ. 



