1844.] SURFACES OF DISCONTINUITY IN A VERTICAL PLANE.

59

in applying the same reasoning to the resistance of the frontal dip, exhibited in the second figure of this letter. When a fluid, or semi-fluid, is very viscous, there is a great resistance to its onward motion in the direction which gravity and the fall of the bed prescribe. Let L M be the surface, N O the bed of a glacier; then the resolved force is usually considered as acting on the particles m n, in the directions m m', n n', parallel to the bed. But if we reflect that, owing to the length of the glacier, and the toughness or consistency in its mass, the resistance of the line of particles nv is enormous, the plane of complete resistance N O will virtually be twisted in the direction N' O', and the particle tends to be thrust forwards and upwards, which will evidently produce the frontal dip.

(3.) But there is a peculiarity in the vertical plane which did not exist in the horizontal one. In the case we first considered, the veined structure exists almost entirely in the neighbourhood of the sides of the glacier, and is lost towards its centre, being due to the influence of friction, which varies with the distance from the side; the central part, e f g h (fig. 11), moving nearly uniformly, would cease to exhibit a linear arrangement. The completion of the curve is due to the influence of the curvilinear bottom, combined with the opposing mass of the glacier in front; and this

L

move onward in the direction in which the effective pressure is greatest; and it is plain, that, owing to the diminishing relation between the weight of the superincumbent particles and

the frontal resistance, the direction in which the particles will tend to slide over one another, or to produce rents, will approach verticality at the surface, and on the whole will, therefore, tend to produce lines of discontinuity, such as N M.

(4.) Considering the glacier at different points of its length,

Fig. 16.

it is evident, by similar reasoning, that near the region of the névé a the frontal dip will be all but vertical, because there the horizontal resistance is enormous; whilst at the lower end b, where it tends to vanish, the shells will tend to parallelism with the bed. It is needless to add, that the relative movements of the particles over one another, producing discontinuity, are not to be confounded with their absolute motions in the glacier, exactly as under head (1.) I must, however, observe, that as the tendency of any particle due to the hydrostatic pressure will be to describe ultimately the whole curve N m, M within the glacier, this may account for some of the facts, or supposed facts, which indicate a tendency in the ice to expel bodies engaged in it, as well as the convexity of the glacier at all times, and its remarkable rise of surface during winter.

Lastly, The ablation of the surface of the glacier during its descent from a to b (fig. 16) will tend continually to give the observed elongated forms of the superficial bands, by cutting the shells of structure obliquely.

1844.]

GREAT GLACIER OF ALETSCH.

61

IX. EIGHTH LETTER on GLACIERS,* addressed to PROFESSOR JAMESON by PROFESSOR FORBES.

Observations on the Motion of the Glacier of Aletsch-Experiments on the Plasticity of the Ice of the Mer de Glace, by Observing the Distortion of a Limited Space of Ice devoid of Crevasses-Movement of a Glacier of the Second Order 8000 Feet above the Sea.

GENEVA, 30th August 1844.

My Dear Sir-The theory of glaciers has now reached that point when it can only receive some material addition by the multiplication of accurate measurements; and these measurements must be conducted in the manner which will best discriminate between rival hypotheses, and, if possible, yield direct instead of indirect proofs of each fundamental fact assumed. In my former letters I have insisted sufficiently upon the importance of the results which a system of nice measurement has introduced into this branch of science, and their value to the theorist who afterwards wishes to put numerical for unknown quantities in his investigations; I also showed that there is a continuity and approximate constancy in the motions of glaciers, which permits us to obtain, with certain precautions, in a few days, better results than any one had previously acquired during the lapse of months or years. I have now to announce to you that I have pushed these measurements to a still greater degree of minuteness, and with results which show that the methods I have employed are trustworthy, and are able to afford the direct solution of questions which at first appeared to admit of only indirect or inductive proof.

Of this class, by far the most important appeared to be the manner in which the glacier alters its form in such a way, and to such a degree, as to suffer its central portion to descend towards the valley with double or treble the velocity of its lateral parts. Such, for instance, I have found to be the case in the middle region of the great glacier of Aletsch, where its inclination is small (about 4°), and where the continuity of the

* Edinburgh New Philosophical Journal, October 1844.

ice with the side wall is preserved without the interference of large fissures. I there found that, whilst the velocity of the ice at 1300 feet, or about a quarter of a mile, from the side, is 14 inches in 24 hours; at 300 feet distant from the side it was but 3 inches in the same time; and, close to the side, it had nearly, if not entirely, vanished. Facts like this seem to show, with evidence, what intelligent men, such as Bishop Rendu, had only supposed, previously to the first exact measures in 1842, that the ice of glaciers, rigid as it appears, has in fact a certain ductility" or "viscosity," which permits it to model itself to the ground over which it is forced by gravity—and that, retaining its compact and apparently solid texture, unless the inequalities be so abrupt as to force a separation of the mass into dislocated fragments, such as it is well known that every glacier presents, when the strain upon its parts reaches a certain amount as when it has to turn a sharp angle, or to descend upon a rapid or convex slope.

The mutual action of the parts of the glacier, the drag which the centre exerts upon the sides (and, by an exact parity of reasoning, the top upon the bottom), seemed to me so obvious, after measurement had proved their variable velocity, and observation had shown that this was not necessarily accompanied by a general dislocation of the mass-that I should scarcely have thought of attempting a direct proof of the yielding and ductile nature of glacier ice, had I not been favoured by Mr. Hopkins with copies of his two ingenious papers on the subject of glaciers, read to the Cambridge Philosophical Society on the 1st May and 11th December 1843, which were put into my hands here less than a month ago, by his friend Mr. Williamson. I there found it stated that there is "a necessity of proving, by independent experimental evidence, that glacier ice does possess this property of semi-fluidity or viscosity, if we would attribute to that property the effectiveness of gravity, in setting a glacier in motion."-First Memoir, p. 3.

Since Mr. Hopkins admits the fact of the swifter central motion of the glacier, he must have recourse to some mechani

1844.]

MR. HOPKINS' DIFFICULTIES CONSIDERED.

63

cal explanation of the fact. This he does by assuming the existence of vertical fissures, parallel to the sides of the glacier, dividing it into a series of longitudinal stripes, whose adjacent surfaces, according to him, slide over one another, and, in the case of a glacier forcing its way through a gorge, the lateral portions are altogether arrested, whilst the central parts slip down between them.*

These parallel stripes of ice are supposed by Mr. Hopkins to be of considerable breadth, and to have no sort of analogy with the ribboned structure, to which the readers of my earlier letters will recollect that I have ascribed a similar origin, being lines of discontinuity arising from the crushing of one portion of the semirigid glacier past another. This Mr. Hopkins regards as "no more possible than that a mass should permanently maintain a position of unstable equilibrium." The veined structure of glaciers he considers to be unexplained, and, in the present state of science, inexplicable.

Although the general absence of such a system of longitudinal fissures as Mr. Hopkins has figured in page 14 of his First Memoir, and the regularity and continuity of motion of the glacier and of its parts, wholly inconsistent with the jostling of huge masses of dislocated ice, might be considered as a sufficient answer to this modification of the theory of De Saussure, the consideration of this demand for a direct proof of the flexibility of glacier ice led me to think of its practicability; and I shall now state what I have succeeded in doing, towards the solution of this practical question in the only way in which it admits of being treated, namely, by the assiduous observation of the motion and change of form of a small compact space of ice on a glacier. The Mer de Glace of Chamouni offers fewer fit points for such an experiment than many other glaciers, since in all its middle and lower portions the ice is excessively crevassed near the sides. There is one spot, however, between the Angle" and Trelaporte, below the little glacier of Charmoz,

* [Mr. Hopkins' figures are reproduced in Plate II., figs. 1, 2, and are referred to again, later in this volume.]