Temperature of the Terrestrial Globe. 19 
and the polar water. Whence it follows that these two differences 
nearly compensate each other, and that to produce equilibrium, the 
sea ought to be, under the equator, a little lower than under the poles, 
and that if this difference of level establishes a current, it should take 
place at the surface, from the poles to the equator, and in the inte- 
rior in a contrary direction. Let us now see the effect of a thaw of 
the polar ice. 
We have two kinds of polar ice, that which is formed upon the 
continent, the continent being land covered with snow and ice, or 
simply ice and that which is formed by the sea itself. ‘The first are 
evidently only glaciers, like those which are formed in the alpine re- 
gions of every climate. They contain absolutely no salt, this is con- 
firmed by the observations of MM. Egédé, Sabie and Wrangle. 
The second are the waters of the sea frozen, and M. Wrangle in- 
forms us that between the 70° and 71° of N. Latitude this congela- 
tion does not exceed the depth of nirfe or ten feet. Thus these enor- 
mous masses of floating ice, which rise even four hundred and five 
hundred feet above the sea, and have at least eight to nine times more 
thickness under the sea, are glaciers of the former kind, formed up- 
on a base of frozen sea water, which cannot be twenty feet thick ; 
and this ‘base itself contains so little salt, that it was believed for a 
long time that it did not contain any.* Thus we may consider the 
polar ices and the water which runs from them as containing only a 
minimum of salt, perhaps less than the water of most rivers. As 
moreover it is detached from only a few sides of the icy platform in 
comparison with the mass of waters which melt every summer from the 
surface of these great platforms, and which contain no salt, we may 
without sensible error, consider the entire mass of the waters which 
flow each summer from the polar regions as a water without salt, 
and we can pronounce, without uncertainty, upon the direction of 
these waters. 
The sea water of these latitudes being like the water of the gla- 
ciers, at the temperature of 0, these two waters act, with regard to 
each other, as their specific gravity impels them, that is to say, the 
water from the glaciers will glide upon the surface of the sea towards 
the equator, without sinking at all; for although, during this move- 
* [have proved that salt water in freezing retains a part of its salt. See my 
Grundiss der theoretischen Physik, T. U1. and the 4nnalen der Paysik, T. LVIL. p 
144, It follows from my experiments that the inferior parts of the ice of sea water 
must contain a little more salt than the superior. . 
