106 ANIMAL INTELLIGENCE 
never fail to notice them at considerable intervals, 
say once in five to nine minutes, even when hibernating 
most profoundly. The respirations were peculiar. 
Sometimes one deep inspiration, preceded by a sudden 
relaxation of the enlarged chest, would be succeeded by 
a long pause; again there would be a series of very 
slight respiratory movements. It was always possible 
by the respirations alone to predict when the animal 
was approaching the waking state. 
The awakening was never sudden but gradual, often 
extending over hours. I have seen something like this, 
though less remarkable, in the bat. This is no doubt 
protective to the vital machinery, for when Hall main- 
tained that bats, suddenly awakened from the hiber- 
nating condition, died speedily, he was, in the main, if 
not entirely, correct. 
A brief consideration of natural daily sleep will throw 
light on winter sleep, etc. 
Sleep is favoured by moderate exhaustion, a good 
condition of nutrition, and the absence of all sorts of 
peripheral stimuli. 
The case of the youth having but one good eye and 
one hearing ear, who could be put to sleep by closing 
these up, is very instructive. This lad did not, however, 
continue to sleep indefinitely, but awoke after a number 
of hours, showing that though there are certain 
conditions in the environment that favour sleep, the 
latter is essentially a condition of the central nervous 
system, and dependent on laws governing the latter. 
This view makes it clear that sleep is naturally a 
nocturnal condition for most animals, owing, no doubt, 
to the evolution of life in relation always to the 
environment. The fact is, we cannot conceive of life 
except in, and by reason of, in a sense, some environ- 
ment. The change of the seasons, day and night, and 
all the periodicities of the inorganic world have, as a 
