200 The American Naturalist. [March, 
-definite variability. This is the only sense in which the terms 
“ definite variations ” and “indefinite variations ” can be fairly used 
in this discussion. 
In two of his most recent essays Yaon says : 
“My theory might be disproved in two seke e by 
actually proving that acquired characters are transmitted, or by 
showing that certain classes of phenomena admit of absolutely no 
explanation unless such characters can be transmitted. Only 
if it could be shown that we cannot now or ever dispense 
with the Lamarckian principle would we be justified in accept- 
mob 
We may gather evidence from the data of Embryogeny, or of 
Ontogeny and Phylogeny. Itis neither possible nor desirable to 
separate these data; but as previous writers have dealt extensively 
upon the evidence of embryogeny, I will emphasize the ontological 
and paleontological evidence, with which I am, in fact, much more 
familiar. I shall endeavor principally to concentrate attention 
upon the phenomena to which future observation must be espe- 
cially directed. We already have. a number of valuable essays 
and criticisms in this line,” but none, so far as I have seen, 
examine the question in view of all the difficulties which the 
adoption of either principle involves us. 
I believe we are far from understanding all the phenomena of 
variation, and put the question, therefore, in the following form: 
Does our present knowledge of variation in living and fossil forms 
lend greater support to Lamarck’s or to Weismann’s principle? 
1. What is the Origin of Variability ?—According to Weis- 
mann, the ultimate or primordial origin of variability is somato- 
genic, *—that is, we must trace variability back to the unicellular 
organisms in which the environment acts pay upon the whole 
20 Biol. Mem., p. 388. 
21 Nature, Feb. 6th, 1890, p. 322. 
22 Especially those of Ryder, Cope, Eimer, and iter A very valuable review 
of the whole ag is found in C. V. T paper, “ On the Causes of Variation in 
Organic Forms.” Proc. Am. Assoc. Adv. Sc 
23 Biol, iaoi p. 277. “ The origin of cae individual variability cannot be 
found in the pa — but it must be sought for in the lowest,—the unicellular 
organisms. uch organisms reproduce by division, individual acquired characters 
will be se oa wo the offspring.” 
