700 
In a visit to the glacier of Chamouni in the summer of 1821, I 
was struck with the very remarkable positions of several large blocks 
‘of granite resting on the glacier in various parts. ‘They were perched 
on stools of ice ok less diameter than the blocks themselves, which 
overhang their supports on all sides, as a mushroom does its stalk. 
The position of these large masses was rendered the more striking 
when contrasted with that of small fragments of stone, equally (to 
appearance) exposed to all the local heating and cocling influences, 
but which were uniformly found te have sunk mto the ice, and 
that the deeper, (within certain limits) the less their size. On con- 
sideration, the cause became apparent, and, as it affords a very 
pretty illustration of the laws of the propagation of heat through 
bad conductors, and the steps by which an average temperature is 
attained in large masses from a varying source, [ will here state it 
as it occurred to me at the time. 
With regard to the sinking of small masses into the ice when 
heated by the sun, it is the natural effect of the greater power of 
absorbing heat which stone possesses beyond ice. Whenever the 
sun shines, the stone will detain more of its heat than an equal sur- 
face of ice would do ; and as it gives this out to the ice below nearly 
as fast as it receives it, a greater depth of ice is melted in a given 
time beneath the stone than in the parts around. On the other hand, 
at night, ice radiates terrestrial heat nearly or quite as copiously as 
stone, and thus they are on a par in frigorific power. 
The elevation of great masses above the general level, which at 
first sight would appear to contradict this explanation, is however 
equally a consequence of the laws of the propagation of heat. To 
conceive this, let us imagine a very large block of stone, at the com- 
mencenmient of the summer, to lie on a level surface of ice, in a si- 
tuation exposed to the direct rays of the sun, where the mean tem- 
perature of day and night is (even in summer) but little above the 
freezing-point, but where, however, no fresh snow falls during the 
whole summer. In the day time then, while receiving the sun’s 
rays, the upper surface of the stone will be strongly heated, and a 
wave of heat will be propagated slowly downwards through the 
stone towards the ice, diminishing in intensity rapidly, however, as 
it travels, since each superior stratum only divides its excess of 
_temperature with that below. Long before this can reach fhe ice, 
however, night comes on. The surface cools below the mean or 
even below the actual temperature of the air by tadiation, and a 
mave of cold is propagated (or, which comes to the same thing, heat 
is abstracted from stratum to stratum) by the same laws. This fol- 
lows close on the wave of heat below and travels with equal velo- 
city. In consequence, the heated stratum parts with its heat, now, 
both upwards and downwards, and thus the intensity of the wave of 
heat diminishes with much greater rapidity as it proceeds down- 
wards. It is manifest, that were the thickness of the stone infinite, 
the wave of heat being always followed close up by the wave of 
cold, and a perpetual tendency to an equilibrium of temperature 
