INFILTRATION OF SALT WATER. 877 
ridian as in the torrid zone ! The immediate surface 
to-day was often entire, though we plunged almost 
knee-deep in water helow it. This you will easily un- 
derstand when I tell you that the thermometer in the 
sun gave, for four successive hours to-day, a mean of 
nearly 80°. The surface thaw percolates through the 
loosely-compacted snow, and, forming a pasty sub- 
stratum, is protected from re-freezing by the very snow 
through which it has descended. Our mean temper- 
ature of late has varied but little between 25° and 
27° for any twenty-four hours. 
" The infiltration of saline water through the ice as- 
sists the process of disintegration. The water formed 
by surface or sun thaw is, by the peculiar endosmitic 
action which I believe I have mentioned elsewhere, 
at once rendered salt, as was evident from Baume's 
hydrometers and the test of the nitrate of silver. The 
surface crust bore me readily this evening at a tem- 
perature of 21° and 19°, giving no evidences of thaw. 
Beneath, for two inches, it was crisp and fresh. As 
I tried it lower, cutting carefully with my bear-knife, 
it became spongy and brackish ; at eight inches mark- 
edly so ; and at and below twelve, salt-water paste. 
On the other hand, all my observations, and I have 
made a great many, prove to me that cold, if intense 
enough, will, by its unaided action, independent of 
percolation, solar heat, depending position, or even 
depth of ice, produce from salt water a fresh, pure, and 
drinkable element. 
^^May 31, Saturday. Walked to-night to the south- 
ward in search of seal : found the ice in motion, and 
had some difficulty in getting back. Wind from south- 
ward, and freshening, after a day of nearly perfect 
calm. The drift is somewhat to the eastward. The 
