Mr. J, Ball on the Structure of Glaciers. 497 
structure which is found to be an almost universal characteristic 
of glaciers. 
In favour of this view it may be urged that we have, in the 
rst place, the positive fact that the névé is a mass made up of 
beds alternately more and less compact and crystalline,—that, 
80 far as we are able to trace them downwards, these beds con- 
tinually diminish in thickness,—and finally, that the glacier ice 
formed out of these beds has a proper structure, which differs 
only in degree from that of the névé. The irregularity in the 
thickness and appearance of the veins, and the recurrence at 
regular intervals of larger and more strongly-marked bands, are 
facts exactly in accordance with this explanation of the origin 
of the veined structure. At first sight it seems, however, im- 
possible to reconcile it with the position which is generally 
assumed by the surfaces of the veined structure, but perhaps a 
little consideration may lessen, though I do not pretend to say 
that it will completely remove the difficulty. All great glaciers 
are made up from the confluence of separate ice-streams. We 
Suppose each of these to consist in its origin of beds of névé 
transformed into glacier ice, the strata of which must gradually 
have conformed their slope to the form of the channel down 
which they have moved. Hence at the point of junction of two 
glaciers, we have brought side by side two great masses of ice 
made up of concave shells lying one within the other. In the 
gradual advance of the glacier, two, or it may be several, such 
masses are pressed together within a channel much narrower 
than the aggregate of those which they separately occupied, 
Those who have followed Professor Tyndall’s experiments, will 
readily understand that under such conditions concave shells of 
ice would be gradually brought to the shape of deep troughs, 
of which the transverse section would have the form of an 
elongated letter U. It is obvious that the appearance pre- 
sented by a horizontal section through a mass made up of such 
surfaces would be that of lines parallel to its length. But a 
horizontal section does not correctly represent the condition of 
things scen in descending over the surface of a glacier. Each 
successive portion of the glacier has undergone in an increased 
degree the process of ablation, by which it is estimated that a 
thickness of from 10 to 12 feet is in many cases annually re- 
moved from the surface of the glacier. We should therefore 
obtain a nearer approximation to the truth by supposing an 
elongated solid, composed of deep trough-shaped surfaces lying 
one within the other, to be cut through by a plane inclined at 
a small angle to its longer axis, so as to reduce its thickness at 
the lower end toa mere tongue. The likeness to a glacier would 
be further increased if the edges were rounded off so as to 
Phil. Mag. 8.4. No. 96, Suppl. Vol. 14. 2K 
