162 
transmutation by means of Natural Selection which belong 
alike to colours which conceal, to colours which act as 
warnings, and to insects which mimic the coloration of 
insects of different genera and species. 
In the first place, after what has been said, it becomes a 
mere truism to assert that the efficacy of Natural Selection 
will be in inverse ratio to the amount of accidental death. 
We have seen that it is a logical demand of the theory, 
when dealt with from an abstract point of view', that we 
must get rid of the idea of accidental death. Here, then, 
is a good test to apply. The theory of defensive colouring, 
if it be so admirable an illustration of the action of Natural 
Selection, ought to introduce us to a sphere where acci 
dental death is reduced to a minimum ; or, at the very 
least, is not so common as in other spheres. 
We shall be justified in applying this test especially to 
the insect w'orld, for Mr. Wallace gives several conclusive 
reasons for the fact that the phenomena of mimicry 
abound among insects more than in other departments of 
the animal kingdom.* What evidence, then, can be 
adduced from the observation of naturalists to show that 
the insect world is less free than other spheres of nature 
from accidental death ? I have not been able to find any 
such evidence adduced by the scientific theorist, but a very 
different testimony is given by the practical naturalist. 
Mr. Samuel H. Scudder says:— 
“ The great sources of destruction here, as in all groups of animals, 
are in early life. How large a proportion of the eggs that are laid 
by butterflies ever finally produce imagines ? Let those answer 
who have attempted to follow their history in their native haunts. 
My experience leads me to believe that, at the very least, nine-tenths 
—perhaps ninety-nine hundredths—never reach maturity. Hymen- 
opterous and dipterous parasites beset them at every step ; the eggs, 
although so small and often heavily ridged, cannot escape the 
Contributions, pp. gg-ioo.