194 
This gives to understand that even when the cocoon is sur- 
rounded with saturated water, the diffusion through the wall — if 
any at all — is not sufficient for the animal, a tension difference 
of about 21 °/o cannot bring about an O 2 -diffusion sufficient for 
the animal’s metabolism. 
Under natural conditions the cocoon is not surrounded with 
water saturated with atmosplieric air but with the water in the 
mud. And here is (according to prelirainary tension determinations) 
the oxygen tension low. 
Mostly (summer as well as winter) the oxygen tension is still 
lower than this in the cocoon; an oxygen renewal by means of 
the oxygen in the surrounding water is therefore excluded. 
As it now must be regarded as proved that the oxygen in 
the intercellular spaces is the only source for the renewal of the 
oxygen in the cocoon, it may seem a little capricious that the 
difference between the oxygen percentage in the intercellular spaces 
and that in the cocoon can vary so much (at the same temperature) 
as my analyses have shown. 
It would not be difficult to give a series of explanations of 
the varying tensions difference, but as I cannot say which of these 
are of real importance I should prefer only to give the faet 
without explaining it. 
In all my analyses, however, the difference between the 0 2 - 
percentage in the cocoon and in the intercellular spaces is rather 
considerable, the mean difference (iu July, on Nuphar) being 
3,6 °/o 0 2 . 
The diffusion of the oxygen from the intercellular spaces into 
the cocoon is a very slow process, so slow even that through the 
large hole (or holes) through which the cocoon is in communication 
with the intercellular spaces, a tensions difference of about 3—4 °/o 
is needed in order to bring about a diffusion sufficient of the 
animal. 
From this we can conclude that a diffusion through the very 
diminutive membrane by which the canals in the hooks according 
