436 TOWER, DARWINISM. 
of chance favorable position. No evidence of adaptive variations 
was discovered. It appears that the major elimination is the one we 
have been considering, so that whatever eliminative processes occur 
subsequent to this one, they are proportionately of less moment, 
in fact, are not unlike the statistical results that come from the 
examination of the cases of mimicry and protective resemblance. 
It must be remembered that these data are for germination at 
the first opportunity presented, and that seeds might germinate later. 
It is not easy to evaluate this aspect of the problem; it introduces 
a complication but does not alter the character of the general result. 
Moreover, seeds delayed in germinating are constantly accumulating 
chances against their growth. With Solanum rostratum seeds, tests 
were made of these delayed germinations under laboratory con- 
ditions, with the result that few delayed germinations took place, 
and these did not differ in any discoverable manner from the normal. 
In the normal plots of 1905 at Apam, the growing plants were 
pulled up and destroyed at the census, so that no seeds were 
matured. In following years, delayed germinations appeared as 
folows; — 1906, favorable, 19; normal, 36; 1907, favorable, 21; 
normal, 7; 1908, favorable, 2; normal, 2; 1909, favorable, 0; 
normal, 0; with no plants forming seeds on the plots in any year. 
It does not appear that delayed germinations alter at all the cor- 
rectness of the general conclusion regarding elimination. 
WITH ANIMALS. 
The examination of the fundamental problem in elimination and 
survival with animals is much more difficult than with plants, which, 
in many respects are ideal for this kind of investigation. With the 
resources that I had available, the most satisfactory subjects and 
locations were in the tropics, where restricted habitudinal relations, 
as well as greater freedom from human interference provided the 
best experimental arrangements. For these experiments I used 
isolated habitudinal groups of Chrysomelid beetles of the genera 
Calligrapha, Leptinotarsa, Labidomera and Doryphora, but for the 
purposes of this paper, I shall utilize examples from the genus 
Leptinotarsa, only. In principle, the results with all of the animal 


