600 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1911. 
significance. One goes to sleep before the moment of complete 
exhaustion and in order to prevent it. Sleep is a function for the 
defense of the organism, and we have the desire to sleep before feeling 
an overpowering need of it, just as we are hungry and thirsty before 
there is an imperative necessity for eating and drinking, just as 
the swallow migrates before it is overcome by the cold, as a bird 
builds its nest before the time for laying its eggs. All these instinctive 
acts, all these instincts, begin before there is an absolute necessity. 
This conception of sleep as an active instinct preceding exhaustion 
changes the question entirely. Sleep is no longer a categorical 
positive necessity, but it becomes a very pliable, modifiable act. 
Like all instincts, it is regulated by the law of momentary interest. 
We sleep if our interest in slumber is the greatest at a given moment; 
but we could ward off sleep if some other instinct should predominate. 
This pliability of instinct enables us to understand the variations of 
sleep, the various causes of drowsiness and awaking. It enables us 
to understand dreams. The theory of Claparéde has therefore a 
great advantage over all the others, since it alone can be applied to 
all the varied forms of sleep. Further, it does not exclude physio- 
logical theories, since it can accept them as stimuli of the sleep 
instinct. Inhibition, the fatigue-producing substances, the sensa- 
tions of fatigue, and the like, become the causes of the interest that 
we take, at a given moment, in going to sleep. 
But Claparéde’s theory reaches beyond the sphere of pure physi- 
ology into that of psychology. Nevertheless, it offers a new field of 
activity for the physiologist, who, relieved from the anxiety of search- 
ing for a complete explanation for sleep, need no longer be deterred 
by the inadequacy of his own theories, but may seek to complete and 
state precisely those parts of the biological theory of Claparéde that 
pertains to his own peculiar sphere as a physiologist and are not 
yet solved. Sleep is an instinct of defense, but defense against what ? 
This instinct is brought into play by various stimuli, but which of 
these are physiological? We shall not inquire further as to why we 
sleep, nor as to what it is that makes us sleep, but ask what does 
sleep protect against? What is it that gives us a desire to sleep? 
My friend Piéron, lecturer at ’Ecole des Hautes-Etudes, has 
asked himself these questions, and has wished very much that I 
would help him to answer them. For six years we have been making 
numerous experiments to try to solve the problem, and it is the 
actual result of our researches that I would like to explain to you 
in concluding this lecture. 
Sleep is an instinct that obeys a Tose of momentary interest, and 
there is very little hope of finding a physiological cause for its being 
brought into play. This explains why the physiological phenomena 
which accompany sleep are often not constant, why we can not 
