588 The American Naturalist. [June, 
itself many of the same objections which Prof. Weismann brings against 
the gemmule and epigenetic hypotheses. The inconceivability as to 
number and complexity of gemmules is quite matched by the incon- 
ceivable complexity of a germ-plasm, which in the sixth generation is 
composed of 32 ancestral germ-plasms, in the tenth generation of 1024, 
etc., even if the reduction by one-half be accomplished for each gen- 
eration as explained pp. 357, ff. Moreover, must not each of these 
germ-plasms be divided by the vast number of ova or spermatozoa 
which appear in any individual’s lifetime, and which are all potential 
for a possible conception ? I am not aware that Prof. Weismann treats 
this point; but at any rate it is difficult to see how, upon his hypothe- 
Sis, we can escape getting down within a few generations to the ulti- 
mate unit which ‘cannot be divided without the loss of its essential 
nature.” (P. 357.) 
The fundamental difficulty, the causing from a remote part of the 
organism such a peculiar effect upon the germ that the disease re- 
appears in the developed organism, this problem is unsolved by the 
bacillus hvpothesis, as also by the gemmule and epigenetic hypotheses. 
But do we need these hypotheses? This must be granted, that the 
germ-plasm, whether continuous or not, is a living being within the 
body and also of it (cf., pp. 103, 170, 267, e¢ passim); that the law of 
interdependence of function—disorder in one function exciting sym- 
pathetic affections in others—applies to reproduction ; that farther, the 
ference of abnormal motion rather than matter? It seems probable 
that epilepsy, for example, could cause specific changes in the germ, 
but this may not be heredity, which is a specific effect of such a nature 
Spermatozoon and ovum, and how this effect is carried through conju- 
gation and development. 

