968 THE AMERICAN NATURALIST. [Vor. XXXV. 
the segmentation nucleus passes through fewer divisions than 
in the whole egg also points in the same direction. 
The regeneration of pieces of Hydra, Tubularia, and Planaria 
shows that at every level of the body the cells have the power 
of producing any part of the organism,! and we must regard 
them as totipotent, or nearly so. The same conclusion is 
arrived at from a study of the development of isolated blasto- 
meres. These facts, and others that might be given, furnish, 
I think, a strong argument against Weismann's hypothesis, 
and I believe we must look in a different direction for a solu- 
tion of the problem of development and of regeneration. 
We should be careful to distinguish between the idea of pre- 
determination and that of pra formation. The former includes 
the latter, but also much more, and is, in fact, an almost neces- 
sary idea in connection with the phenomenon of development, 
at least in one sense. Every egg of the sea-urchin is prede- 
termined to give rise to a sea-urchin so long as the conditions 
remain as we find them at present, but it by no means follows 
that the hypothesis of preformation is a necessary consequence 
of this view. It is true that since each egg passes through the 
same stages to reach the same goal we must regard the series 
of changes as predetermined in the structure or composition of 
the egg, and in the external conditions surrounding the egg, but 
the way in which the development takes place may be in no 
way connected with the presence of a praformed germ of the 
embryo or with the presence of praeformed germs of any sort 
in the nucleus. In fact, a process of pure epigenetic develop- 
ment, as generally understood nowadays, may also be prede- 
termined in the egg. It is well, I think, to be clear on this 
point, since otherwise we grant too readily the necessity of 
assuming some sort of praeformation hypothesis. 
In my Woods Hole lecture of last year I considered the 
hypothesis of formative stuffs and reached the conclusion that 
it is insufficient to explain many of the phenomena of regener- 
ation. Let us see if this view will apply to the development of 
parts of the egg and embryo. In the first place, it should be 
noticed that the hypothesis was primarily invented to account 
! At least so far as the different layers are concerned. 
