Correspondence—Professor J. Milne. 191 
annually. But the annual precipitation amounts to one foot. If we 
suppose this to be formed during the winter, then the heat given off 
in winter by radiation is as much as would evaporate one foot. 
Radiation goes on in summer as well as in winter, indeed more 
actively, as then temperatures are higher. This has to be compensated 
out of the heat then being received. ‘Thus, out of the heat received 
during summer, as much as will evaporate more than one foot of 
ice is spent simply in replacing the heat being radiated away. The 
heat ultimately left will scarcely be sufficient to evaporate the one foot 
of ice which we had to account for. Thus even neglecting the effects 
of the fall of temperature far below freezing-point in winter, the in- 
terception of solar heat by fogs in spring, and other possible causes, 
It is still conceivable for the calculated heat to fall on Greenland, and 
yet not dissipate so much as the observed amount of ice. 
The above calculation gives some idea of the diverse results which 
may be obtained when we reason on uncertain assumptions. ‘lhe heat 
given off in forming a foot of snow or ice out of vapour, if applied to 
changing snow or ice back merely into water, would liquefy nearly 
eight times as much. Whether solar heat would melt or evaporate the 
ice into which it entered probably depends on the dryness or dampness 
of the air. This suggests, what I have often thought probable, that to 
know better the laws of winds might be very helpful in the study of 
Glacial Periods. 
The suggestion of Dr. Roberts’s lucid letter, that the heat disengaged 
in the formation of snow, being disengaged in the upper regions of the 
air, produces little effect at the ground, is well worthy of consideration. 
The same is probably true of rain, and over not merely Arctic regions 
but the whole surface of the globe. But to make this action available 
in support of the theory under discussion, we must show that its effect 
can be increased by increase of eccentricity. It is not sufficient to 
prove that it is intensified by rising temperatures, unless it also be 
shown not to be correspondingly eufeebled when they fall. This re- 
mark applies to many suggested actions. Eccentricity when it seems 
to throw a sword into one scale, often places in the other scale a shield. 
E. Hu. 
THE KURILE ISLANDS, 
Sir,—The following notes upon the Kurile Islands were obtained 
from Mr. J. Snow, a gentleman who has spent several summers cruis- 
ing amongst this interesting group of volcanos, whilst engaged in otter 
hunting. I offer them to you as supplementary to what I wrote upon 
these islands myself in 1878 (see GxoLogicaL Macazine, Decade II, 
Vol. VI. pp. 337—348). 
Chirnot Islands.—On May 29th, 1879, smoke or steam was seen to 
be issuing from the northern of these two islands. It is possible that 
the eruption may have commenced before this date. On the 30th, 
at intervals of from 3} to }.an hour, loud explosions were heard. During 
the night the mountain was seen to be covered with fire. ‘The eruption 
seems to have: formed a new point of land. On the island there are 
four cones, all of which give off steam. Of these four, the most northern 
