1842.] ticular, is a sort of standing evidence of its mechanism, and, rightly understood, must lead to the most important confirmation of any mechanical theory. This, you may believe, I have made an object of very particular attention. I have now examined so many glaciers as to have a very clear idea of the empirical laws which that structure follows. Lately, I begin to perceive a connection between that structure and the facts of motion already cited. If these two classes of facts can be well brought into harmony with one another, we should have a very good chance of consolidating them into something like a theory. my next letter, I will give you some account, at all events, of my observations on the subject, which are sufficiently definite, and probably also (without considering it as proved), of what seems likely enough to be its true explanation. I go to-morrow to the Great St. Bernard, to meet M. Studer.-Believe me, very sincerely yours. THEORY TO INCLUDE BOTH STRUCTURE AND MOTION. 17 In IV. THIRD LETTER on GLACIERS, addressed to The Laws of Structure of Glaciers generally detailed-The "Dirt-Bands " described. Mechanical Theory of the Structure-Dirt-Bands correspond to Annual Intervals-Glaciers probably move in Winter. ZERMATT, North Side of Monte Rosa, My Dear Sir-I arrived here two days ago by a very interesting and unfrequented route. I mentioned in my last, that M. Studer and I had agreed to visit together the valleys eastward of the great St. Bernard. The Convent was our place of rendezvous, and we afterwards descended to Orsières, and turned into the valley of Bagnes. Crossing the Alpine chain at the head of the valley, by the Col de Fenêtres, we went down to Valpelline on the Italian side, and ascended that *Edinburgh New Philosophical Journal, October 1842. C valley quite to its origin. We then crossed to the western branch of the valley of Erin, by the Col de Collon or Arolla, a very striking glacier pass. Thence M. Studer went to the Val d'Anniviers, and rejoins me here by the way of Visp, whilst I ascended the other branch of the Eringer Thal from Evolena by way of the Ferpecle glacier, and crossed over the mountains to this place, by a pass higher and much longer than the Col du Géant, which presents, certainly, the grandest views I have hitherto met with in the Alps. I must not, however, stop to describe, as my present object is to fulfil the promise in my last respecting the structure of glacier ice. The internal veined or ribboned structure presented by all glaciers in a greater or less degree, appears to be the only true essential structure which they possess, and which, you will recollect, I described in a paper printed in your Journal for January last.* The existence of granules divided by capillary fissures, as well as of large crevasses, are equally unessential to glacier structure, and subordinate to the other. Whatever other result may flow from the examination of glaciers this summer, by the many persons who are probably at this moment directing their attention to them, this, I am sure, will be admitted, that the veined structure is not peculiar to some glaciers, as some would maintain, nor to some years, as has been alleged by others; but that it is perfectly general and systematic, having one general type or form, which is varied according to external mechanical circumstances. Being then the most essential and intimate part of the glacier formation, as well as one of its most obvious and universal features (especially on those glaciers which are most commonly visited), it is equally singular that it should not have been sooner noticed, or if noticed, never once alluded to by the eminent and ingenious authors who have treated of existing glaciers and their effects. With respect to the general type or form of this structure, I am happy to say that I have found not the slightest reason to modify the description which I have given in the paper above * [See page 1 of this volume.] 1842.] FORMS OF SURFACES OF STRUCTURE. 19 alluded to, of the conformation of the glacier of the Rhone. The description is characteristic, not of that glacier only, but of every other, with certain modifications similar to the variation of the parameter of a curve; variations, therefore, not in kind but in degree. The most beautiful structure I have ever met with is in the glacier of La Brenva, in the Allée Blanche, which was one of the earliest I examined this season, and in which I found all that I had seen, though imperfectly, on the glacier of the Rhone (which it resembles in the circumstances of being derived from an icy cascade, and in having a considerable breadth in proportion to its length), developed in a manner so clear and so geometrically precise, as gave me the most lively satisfaction. I refer to my former paper for the figure and description of that structure; I have found the same conoidal surfaces, and the same false appearance of horizontal stratification on the terminal face of the glacier, arising from the veins dipping inwards at first at an angle of only 5°, rising to 10°, 20°, up to 60° and 70°, if we follow the medial line of the glacier, or axis parallel to its length. The sides of the glacier, in like manner, have their cleavage planes or veins dipping inwards towards the centre at an angle determined by the declivity of the rock or moraine which supports them, gradually becoming more vertical as the centre of the glacier is approached, where they twist round by degrees, so as to become transverse to its length, and to form part of the system of planes dipping inwards first described. Fig. 1 exhibits a section parallel to the length. Fig. 2, a transverse section. You are already aware that this structure consists in the alternation of more or less perfectly crystallized ice in parallel layers, often thinning out altogether like veins in marble, not unfrequently parallel and uniform like a ribboned calcedony or jasper. I will, for brevity, merely state the modifications which this fundamental type undergoes, bringing together glaciers of all classes, but reserving the detail of examples and proofs, of which my experience has already furnished me with a great number, to another occasion. If a glacier be long and narrow, as the Lower Aar, or the Mer de Glace of Chamouni, the frontal dip is the least conspicuous part of the phenomenon; and if it terminate in an icy cascade, as in the second case, it might escape observation altogether. The vertical planes parallel to the length, or nearly so, usurp nearly all the breadth of the glacier, and only in the centre is a narrow space, where not unfrequently the structure appears quite undefined. I have satisfactorily made out, however, in every glacier which I have had the means of examining with that view, that the conoidal structure, however obscured, exists in all parts of the true glacier, modified, according to its length and breadth, in the manner which figs. 3 and 4 indicate. I need not add, that these rude sketches are not intended to be considered as rigorously exact, but only to explain generally my meaning. There is yet another modification, but only a modification, of the above, namely, in the case of extremely steep glaciers, but which are coherent, and not crevassed into pyramids. There are numberless examples of these in all the higher valleys of the Alps, which do not descend into the hollows, but festoon 1842.] GLACIERS OF SECOND ORDER-DIRT-BANDS. 21 the steep sides of snowy mountains. They are, I believe, what Saussure called glaciers of the second order, and have no relation to névés, so far as I can attach a meaning to that term. They are of hard ice, and almost invariably present an appearance of stratification parallel to the soil on which they rest. This stratification is only apparent; the cleavage planes dip forwards and outwards, instead of dipping inwards, as in the terminal portion of glaciers of less inclination. The surfaces of crystallization have, in this case, absolutely the form of a scallopshell, the lip or front being always inclined below the horizon. I attach importance to the community of feature in glaciers of every form and inclination, because it indicates that the origin of the structure cannot be unimportant, considering its generality; and in this particular case of small steep glaciers, it appears, I think, that M. de Charpentier, who has justly denied the stratification of glaciers in general, has wrongly admitted the existence of strata in the case in question, which he regards as formed by the intercalation of mud from the soil in some manner, which, if I recollect rightly, he does not very clearly describe. Now, these seeming strata of mud I have examined in a multitude of cases, and found invariably to result merely from the percolation of dirt from the moraine, sometimes even accompanied by small fragments of rock, into the more spongy and less crystalline veins of the glacier mass which already existed; and the proof is that, by cutting with a hatchet, we gradually gain the pure ice, equally veined with the exterior, but not discoloured. I may observe, in passing, that the fissures which, in the lower part and near the sides of glaciers, form the granules, about which so much has been written, are stopped by the independent formation of the veins in the ice, which thus demonstrate their prior origin. One afternoon I happened to ascend higher than usual above the level of the Mer de Glace, and was struck by the appearance of discoloured bands traversing its surface nearly in the form indicated in fig. 4. These shades, too indistinct to be noticed when near or upon the surface, except upon very |