120 THE EVOLUTION OF INSTINCT 
of them are killed and others paralyzed.” In Ammophila 
“the prey may be stung so slightly that it can rear and strug- 
gle violently or so severely that it dies almost at once, and 
in neither case is a break made in the generation of Ammo- 
philes, since in the former, the egg or larva is so firmly fas- 
tened as to keep its hold, while in the latter the dead and 
decomposing caterpillar is eaten without dissatisfaction or 
injury.” 
An important fact which Eimer has apparently failed to 
consider is that in most of the solitary wasps the nest with 
its provision and egg is abandoned as soon as completed and 
never seen again, the wasp dying before its progeny emerges, 
The same is true among the more primitive bees, such as 
Osmia, where elaborate provision is made for the larva 
which only emerges the following year. The typical con- 
dition among the solitary species of both bees and wasps is 
one in which the parent never sees its own offspring, never 
has an opportunity of watching the results of its own experi- 
ments. Unless gifted with a truly preternatural intelli- 
gence what means has a solitary wasp or bee which never 
sees the larva or pupa of its own species of knowing what 
conditions of food and habitation are the most suitable for 
its progeny? Far from being the only possible explanation 
of the origin of the remarkable instinct he has cited, the 
doctrine of Eimer is improbable to the point of absurdity. 
Romanes who has treated the evolution of instinct at 
length in his Mental Evolution in Animals regards both 
natural selection and use inheritance as important factors. 
Instincts due to the first factor he calls primary, while 
those due to the latter are called secondary. Several in- 
stances are cited in which it is claimed there is strong proof 
that certain instincts have been produced through the ac- 
cumulated effects of experience. Other instances are given 
