54 
Proceedings of the Royal Society of Edinburgh. [Sess. 
batch of Ligise in distilled water virtually frees them from such soluble 
salts as adhere to the surface of the body. Before immersion in the 
Erlenmeyer flasks the animals were so washed that a sojourn of two or 
three minutes in a final small quantity of distilled water permitted the 
escape of no chloride sufficient to be detected with silver nitrate. After a 
number of hours in the much larger quantity of distilled water contained 
in the Erlenmeyer flask, a marked white cloud was invariably produced 
on the addition of a drop of silver nitrate. It is quite probable that 
death from immersion in distilled water is primarily due to the loss of 
essential salts. 
It has already been mentioned that an animal rendered cedematous by 
immersion in one-quarter sea-water may recover on prompt transference to 
full-strength sea-water. One had abundant opportunity of repeating this 
experiment — with similar findings — in the case of animals, whether 
oedematous or not, brought near the point of death by immersion in 
distilled water. With practice one came to estimate with considerable 
accuracy the chances of restoring the animal by this means. After high 
degree of inco-ordination of the gill-movements restoration is not possible, 
at any rate with immersion in simple sea- water ; provided the inco-ordina- 
tion is slight, recovery may be confidently expected. 
We may now discuss the cause of variability in the period of survival 
of animals immersed in distilled water. As possible influencing conditions 
the following were experimentally taken into account: (1) sex; (2) colour, 
whether permanent (some animals, for example, are reddish, some blotched 
with white) or temporary (depending on retraction or expansion of 
chromatophores according to the prevailing background) ; (3) size ; (4) the 
condition as regards moult. It was found that sex could be discounted ; 
so also colour in so far as it did not depend simply on moult. Size proved 
to be of influence, a small animal having on the whole less chance than a 
large of surviving beyond a certain point. At the same time size is not 
the onty factor, for a small animal may survive twice as long as another 
three or four times its size. A decisive influence is exercised by the state 
of the animal as regards moult. 
As the question of moulting is to be discussed in the next communica- 
tion, all that need be said at present is that a freshly moulted animal 
is easily distinguished from one that has moulted long since. Records were 
taken of the survival period of a large series of animals selected at random ; 
after death each animal was examined to determine the state as regards 
moult. Without exception, those that long survived were found to have 
recently moulted ; on the other hand, every animal that died unusually 
