204 The American Naturalist. [March, 
One reason why I have endeavered to emphasize the unanimity 
of opinion upon this point among those who deny Lamarck’s 
principle in this: If there are definite lines in blastogenic variation 
which cannot be explained by Selection, or by Environment 
acting upon germ-cells, we must find some other causes or laws 
governing them. Therefore the Lamarckians must first establish 
their claim that there are definite lines of variation; second, that 
these lines have not been directed by Selection (see Query 6). 
The opinions of Lamarckians on this point is that “there are 
variations which follow from their incipient stages a certain 
definite direction towards adaptation, independent of Selection in 
their origin.” * This, it will be observed, does not exclude the 
existence of variations of the class accounted for by Weismann, 
but it constitutes substantially a distinct class of variations which 
Weismann, Lankester, and others do not account for, because, 
upon their hypothesis, we have no evidence that there is such a 
class” 3 
This opinion has frequently been asserted without adequate 
support from observation, otherwise we should not find such can- 
did writers as those quoted above dismissing it so summarily. 
The fact is, it is very difficult, if not impossible, to prove that 
there are definite lines of variation (which cannot be explained 
by Selection) from the examination of zoological and botanical 
collections, for we are, from the nature of the material, princi- 
pally examining variations by divergence in space. In such 
complete fossil series as are now available paleontologists enjoy 
the distinct advantage of following divergence both in space and 
time. They are thus in a better position to study lines of varia- 
tion than ever before, because they are in at the birth, so to speak, 
of many useful and adaptive characters, and can follow the gradual 
rise from the minute infinitesimal stages to the advanced condition 
in which are constituted what we call specific and generic char- 
acters. Not only so, but it is possible to observe pedigrees, since 
the condition of surrounding parts prior to their appearance is 
known 
The history of the teeth of the Mammalia affords the most 
* That is, no class of variations which conform to direct individual adaptations. 
