THE EVOLUTION OF INSTINCT 125 
the results of experiment? To anyone who has made any 
first hand study of lepidopteran psychology the supposition 
_ that the ingenious mechanical devices shown in the cocoon 
were. hit upon as the result of a series of experiments is 
about as probable as that a cat would be able to comprehend 
the differential calculus. Yet we read such phrases as “such 
calculation,” “what wonderful contrivance,” applied to such 
performances by some of the foremost students of animal 
psychology a couple of decades ago. How absurd they all 
‘seem now! To be able to improve a habit there must be an 
opportunity for repeating the action. Caterpillars might 
be supposed, by the mere act of once constructing a cocoon 
to have adapted their organization to this operation, and 
we might suppose that this modification affects the germ 
cells so that the next generation spins with somewhat 
greater facility. But caterpillars which constructed their 
cocoons badly would have no opportunity to improve, and 
their bad methods would be handed on, and confirmed 
more and more in their badness. 
After the moth emerges from the cocoon she soon deposits 
her eggs upon a species of plant which affords suitable food 
for the larve. How did the moth come to have this instinct? 
Can any one believe that the moth watched the results of 
laying eggs on different kinds of plants and formed the 
habit of ovipositing on those upon which the larvee happened 
to thrive? 
Many of the most complex instincts of insects are in re- 
lation to constructing some sort of protective dwelling for 
the winter and in making provision for their progeny, and 
in neither of these cases is there, in most forms, room for im- 
provement through profiting by failures. Protective dwel- 
lings are usually made but once, and in providing for the 
young there is usually no opportunity for the parent to 
