Ixxxvl PROCEEDINGS OF THE GEOLOGICAL SOCIETY. 
the mass descending by the increasing weight, part takes the ground, 
and the other portions run out to sea over deeper water. 
The layers are proofs of successive accumulations, and as they vary 
in their texture, they would point to modifications in the conditions 
of the deposit at different times. We may assume that similar ac- 
cumulations are effected on the land, disturbed only by volcanic out- 
bursts and the showermg of ashes and lapilli from volcanic vents, 
for the time im a sufficient state of activity, such showers extending 
as well over any portions of the icy barrier within reach, as over the 
snows of the land. The uniform, or nearly uniform height of the 
barrier, when covering various depths of water, and not aground, 
would apparently indicate some counteracting cause, preventing such 
an accumulation of layers of ice, as by giving a total increased thick- 
ness should enable a greater height to rise above water in the deeper 
situations. It is here assumed that the thickness of the barrier is 
not increasing, and remains generally the same, a fact which may per- 
haps be thought not as yet sufficiently proved. Nevertheless, when 
we consider that the sea-water beneath the icy barrier is not exposed 
to that great depression of temperature, to which when directly in 
contact with the atmosphere, it is subject, and that beneath the lower 
part of the barrier, deep as that lower part is, it would be only water 
of greater specific gravity than that above, which could there find 
its way, the experiments of Sir James Ross lead us to suppose that 
after a certain depth the ice would cease, the temperature being 
too high for its contmuance. Upon this hypothesis the general 
thickness of the barrier would be the same as long as there was 
a sufficient depth of water to secure the needful temperature, the 
accumulation of snow and ice above being met by the melting of the 
ice beneath. 
We should expect the rise and fall of the tide to act upon the 
barrier, tending to break off portions at its outer edge, where, if © 
the lower part plunge into water of sufficiently high temperature, a 
melting beneath would assist in detaching fragments. The needful 
support by the proper amount of submersion being thus to a certain 
extent withdrawn, the masses would strive to rend themselves off and 
adjust themselves in the water relatively to the floatation lme now 
become proper to them. It is evident by the upsetting of the ice- 
bergs, so often observed, that their centres of gravity become changed, 
so that the masses take a new floating position relatively to them. 
In regions where the cold of the atmosphere is so great that little 
general change is produced upon the upper part of the icebergs 
compared with that which is experienced beneath by plunging into 
water above the freezing pomt, we should expect such changes in po- 
sition frequently to happen, any load of mud or stones of the remain- 
ing portion beneath being comparatively of little importance. 
Great tabular masses varying Im size, some even several cubic miles 
in volume, float away from the parent barrier, the tidal streams and 
ocean currents sweeping them onwards. And it should be borne 
in mind that the solid barrier presents a submarine cliff of ice, by the 
side of which a large volume of water would readily pass without the 
- 
