I894-] 391 [Poulton. 
of the species, and which have been due to the operation of 
natural selection upon the nervous system. Certain cases which 
are most strongly held to be the outcome of the transmission of 
gained experience and the acquired results of practice certainly 
cannot be explained in this way. 
For instance, how upon any such hypothesis can you explain 
tlie wonderful structure of the cocoon spun by the larva of an 
insect? The view would be, [suppose, that the ancestral larva 
spun a cocoon which was not much of a success and was in con- 
sequence attacked by enemies ; that the larva observed these 
attacks, and accordingly improved its cocoon. But tliat is not 
the way in which the struggle for existence is waged with insects. 
If the larva failed, it failed, and that would be the end of the 
matter. It has no chance of improvement ; it has no opportunity 
of leai-ning by experience. Its only chance of survival is to avoid 
experience of foes altogether ; experience is the most dangerous 
thing in the world to an edible insect. This becomes still more 
oJ)vious when we remember that failure or success is almost always 
determined long after the cocoon is made. The caterpillar [)erhaps 
spins the cocoon in autumn, but the real stress of competition will 
come in winter, when insect-eating animals are pressed hard with 
hunger and search high and low for food. But the caterpillar by 
this time is a chrysalis and of course has no opportunity of improv- 
ing the cocoon. The selective testis applied long after the opera- 
tion has been performed, and Avhen there is no possibility of gaining 
by experience. We are thrown back, then, solely upon natural 
selection, which acts on the nervous system of the caterpillar, 
and thus compels it to make the cocoon in a certain way. In 
other words, those caterpillars which are impelled by their 
nervous system to make ill-formed, conspicuous cocoons have no 
chance of living, and in future stages producing offspring. 
Hence, the selection caused by the keen sight of foes first raises 
and then maintains at a high level the standard of cocoon- 
making. 
This contention as to the uselessness and danger of experience 
applies to the whole of those smaller defenceless anitnals which 
liave no chance of fighting with their enemies or of escaping 
when once they have been detected. 
Another special kind of instinct has been greatly relied on 
by Romanes as evidence for tlie Lamarckian theory of transmitted 
