Mr, J. Ball on the Structure of Glaciers. 483 
evaporation and melting in diminishing the mass of snow, or 
on an under-estimate of the amount that annually falls in the 
upper regions. 
Comparing the observations made at the Great St. Bernard, 
and those of M. Dolfuss at his Pavilion on the Aar Glacier, we 
cannot estimate the average annual fall of snow, in the region of 
the Swiss Alps from 8000 to 10,000 feet above the sea, at less 
than 60 English feet. From the experiments of Agassiz on the 
relative density of snow, water, and different varieties of glacier 
ice, it appears that this thickness of snow would be equivalent 
to a bed of névé ice rather more than 8 feet in thickness. 
But in the lower part of the sections of névé, where the snow is 
converted into névé ice, the thickness of the distinct beds is 
seldom more than 6 or 8 inches, and often much less. 
If each layer corresponds to the produce of one year’s fall of 
snow, we should have to believe that the effect of melting and 
evaporation is to reduce the annual deposits of snow to one- 
twelfth or a still smaller proportion of their original bulk. This 
estimate, I feel persuaded, would be a very great mistake if 
applied to the region where the great accumulations take place, 
10,000 feet, and more, above the sea. It is my strong impres- 
sion, derived from frequent observations in the upper regions of 
the Alps, that the visible beds of névé represent each considerable 
fall of snow that has taken place, without any absolute limit 
between those of one year and another. Where a slight fall 
has occurred, or where a more considerable one has been followed 
by a sufficient interval of hot weather, it leaves no trace among 
the strata of the névé, save that the water which it produced has 
helped by percolation to consolidate the icy seams in the beds 
immediately below it. Again, when one fall of snow follows 
another with but a short interval, and before a strong crust of 
granular ice has formed over the first, the two become so nearly 
confounded together in the stratification that subsequently ensues 
as not to be distinguishable to the eye at a short distance. But 
it happens many times in each year, that after a heavy fall of 
snow dry and clear weather succeeds. I see no reason to doubt, 
that in winter as well as in summer, the joint effect of the sun and 
air is to produce a crust of granular ice, whereon the wind often 
deposits dust and other light bodies. At the next fall of snow, 
the main source of waste, evaporation from the surface, is put 
a stop to, and the future diminution in the thickness of the new 
stratum is owing to the gradual consolidation which arises from 
internal melting, from the percolation of water coming from 
above, and the pressure of the fresh strata of snow which con- 
tinually accumulate overhead. 
This explanation accounts for the great irregularity in the 
212 
