

TOWER, DARWINISM. 421 
or in the inaccuracies of observation. In general, the errors that 
might arise from judgment were eliminated by placing the obser- 
vations directly in nature. I shall present in condensed form some 
results derived from the observation of groups of mimetic butterflies, 
some instances of protective resemblance and some instances of 
the eliminating effect of the physical environment. 
MIMETIC GROUPS. 
The cases of so-called mimicry, of which numerous instances 
have been described, are, in general, explained with facility by 
the Hypothesis of Natural Selection. That these resemblances 
exist, is true and it appears that the Hypothesis of Natural Selection 
is probably the most plausible and generally accepted explanation 
for them. The explanation of these cases rests upon the assumption 
that a predacious enemy finds the mimic far more edible than 
the model, which is usually assumed to be inedible. The predacious 
enemy, discriminating between the edible and the inedible, selects 
for consumption those individuals least resembling the non-edible 
form. As a consequence, there is constant elimination from which 
only those individuals of the mimic escape which most closely 
resemble the model. This is a plausible explanation, and my 
question was, wheher evidence of such elimination could be obtained. 
In these instances, especially in the butterflies, birds are generally 
assumed to be the efficient, intelligent, discriminating factor that 
produces the selective elimination in the edible form of those 
least resembling the model. 
Accordingly in 1904 after a preliminary survey in the previous 
year, locations were staked out in the rain-forests of southern 
Vera Cruz, Mexico, where butterflies of the alleged edible and 
non-edible families Danaidae, Heliconidae and Pieridae were 
abundant, presenting instances of so-called mimicry. These locations 
were plots from ten to a hundred metres square and were marked 
by long iron stakes driven deep into the ground. Through oppor- 
tunities made possible by the Carnegie Institution of Washington I 
was enabled to visit these areas two or more times a year, so 
that in all between 1904 and 1911, twenty-two censuses were taken 
upon each of the plots, or 220 censuses for the series. 
