1844.] MOTION OF A GLACIER OF THE SECOND ORDER. 67 have observed coincidences in our views which he appears not to have noticed; and he would probably have hesitated before laying down so broadly as he has done an objection to the Viscous Theory, very easily refuted, and some peculiar views which he considers distinctive of his manner of considering the subject, from De Saussure's and my own. I shall probably, on another occasion, endeavour to show that, by following out his own principles, the results must inevitably merge in mine, when what is inadmissible shall have been subtracted. P.S.-The influence of the Dimension, Slope, and absolute Elevation (or surrounding temperature) of glaciers upon their motion, is a matter of observation in detail which offers no peculiar difficulty, and which deserves to be extended. Having measured the rate of motion of perhaps the largest glacier in Switzerland (the Aletsch), I have also measured one of the smallest, a glacier of the second order, near the Hospice of the Simplon, almost 8000 feet above the sea, and not many hundred feet in length. The velocity was little more than an inch in twenty-four hours, a result corresponding with the extreme dryness of the névé at that elevation, indicated by the very trifling issue of water from beneath, and to the insignificant vertical pressure of so small a mass, notwithstanding its considerable slope. A similar result, it must be owned, might be expected in this case upon almost any theory. 68 NINTH LETTER ON GLACIERS. X. NINTH LETTER on GLACIERS, Remarks on the Recent Observations made on the Glacier of direction of M. Agassiz. New Confirmation of the Plas sation of the Glacier in its downward course in conseque tance-Continuity of Motion-Motion accelerated in Fin Central Velocity of the Glacier-Motion of Glaciers of Snow Beds. EDINBURGH My dear Sir-However satisfied one may clusiveness of their own experiments, it is alwa they are confirmed by others even in their mi especially if the observations have been made at all different. In this respect, I find with communication read at the Institute on the 9t that M. Agassiz's coadjutors on the glacier obtained results so perfectly accordant with have done me the favour of publishing on form they would have satisfactorily established, had tions been awanting, the viscous theory of gla which alone they are reconcileable; the single nism to my own measurements being one whi in favour of that view. I propose to give a brief summary of the show their correspondence with my own. This amounting almost to coincidence-is, of cour factory guarantee for their accuracy, as far others, the goodness of the instruments, and the observers, must, in the mean time, be tak It is hardly necessary to premise that M. friends now admit that all glaciers move fast and slowest at the sides. * Edinburgh New Philosophical Journal An 1845.] PECULIARITIES IN THE MOTION OF THE AAR GLACIER. 69 But a new fact still less reconcileable with the dilatation theory resulted from the first attempt to apply geometrical measurement to the motions of this glacier; viz., I. The glacier of the Aar moves fastest in its middle region, and slower in its upper and lower regions (i. e. towards the origin and termination). The slowness in the upper region does not so plainly follow from the facts at present before us, but the retardation towards the termination of the glacier is undoubted. The following are the motions originally ascertained, in ths of a year, or, more exactly, 289 days. We prefer retaining the original measures in Swiss feet; the stations are in descending order, and a quarter of league (4000 feet) apart* (the second in order indicates the rock called Hôtel des Neufchâtelois.) 169.2 Swiss feet. 177.1 141.3 150.1 133.1 83.7 58.3 This result is very different numerically from that which I obtained on the Mer de Glace of Chamouni, but the difference is of the kind which might have been expected from their great diversity of situation and circumstances. I never expected, or pretended to find in the Mer de Glace the same peculiarities of velocity as in other glaciers; on the contrary, I endeavoured to show what were the local peculiarities as to slope and breadth, which probably produced the law of variation of the motion which I observed, slowest in the middle, and quickest towards either end, precisely the reverse of that observed by M. Agassiz; but I neither depreciate the accuracy of his surveyor, nor contend that one cause of motion sways the glacier of the Aar, and another that of the Montan vert. On the contrary, the difference appears to me entirely conformable to the * Bulletin de la Societé de Neufchâtel, 8th November 1843. viscous theory; and the glacier of the Aar, in this respect, a more instructive example than the Mer de Glace of Chamouni. I have shown in one of the passages of my work just cited, that the velocities of the different portions of the glacier depend among other things, on their inclination or slope; and hence, 1 should have inferred, that in a glacier which did not slope faster and faster towards its lower end, till it becomes almost precipitous, there would be accumulating resistance due to the friction of the ice on the bed of a long, nearly uniform, gently sloping valley, such as that which contains the glacier of the lower Aar, which must magnify the tendency which the ice has to be squeezed forwards and upwards against the mass immediately in advance of it, which produces the frontal dip of the ribboned structure or slaty cleavage of the ice, in the way that I have explained in my Seventh Letter. In a glacier then, whose slope is nearly constant and small, I should expect a condensation of the ice longitudinally, and a swelling of the surface depending upon the motion of the plastic ice in the direction of least resistance. Now this is exactly what we have in the results of measurement. If the annexed figure represent the U s V W W W W W W W R ww plan of the glacier, and the ice be divided into imaginary compartments by vertical sections; since, whilst AB moves 177 feet, CD moves but 141 feet, there is a condensation of the mass of ice ABCD, from back to front, of no less than 36 feet in that time, and so for the successive slices EF, etc. How then, is this shrinking to be accounted for? Not by the mere * [See page 59.] 71 1845.] MOTION OF THE AAR GLACIER ACCOUNTED FOR. internal melting, for that would produce merely a lowering of the surface, and a subsidence of the level of the ice; such as I have shown* actually takes places in other glaciers whose sections move with increasing velocity on the whole. There is only a vis à tergo which can approximate the sections together, and, as we read in the Comptes Rendus, squeeze the moraine longitudinally, giving it a greater breadth,† and condense the entire body of the ice so as to make it more compact in texture.‡ If we take a vertical section instead of a plan (see next page), the slice abcd must be condensed into the higher and shorter solid cdef, and so of the rest, and the surface will be a swelling one, as acen, which might even rise towards valley, but generally need only be less sloped than the bed. The effect of superficial thaw and internal subsidence diminishes this again, and gives it the form of the dotted curve an'. In these diagrams the varying velocity in different parts of the transverse section is, for simplicity, kept out of view. A retardation of the foremost portion of a viscid stream, and consequent heaping of its surface, is exactly imitated in the models formed of plaster of Paris, which I have elsewhere described, and which, though of uniform fluidity from end to end, and therefore not subject to the objection arising from the cooling of lava, where a precisely similar fact is observed, reproduce faithfully the motions of the glacier of the Aar. *Travels, p. 153. + Les "Moraines médianes s'élargissent dans la même proportion que l mouvement se ralentit." Comptes Rendus, 9th Dec. 1844, p. 1301. Ibid, p. 1306, line 29. |