On the Veined Structure of Glaciers; with Observations upon White Ice-Seams, Air-Bubbles and Dirt-Bands, and Remarks upon Glacier Theories

XV. On the Veined Structure of Glaciers; with observations upon White Ice-seams, Air-bubbles and Dirt-bands, and remarks upon Glacier Theories. By John Tyndall, F.R.S., Professor of Natural Philosophy, Royal Institution. Received February 24,—Read February 24, 1859. § 1. Introduction. On the 20th of May, 1858, I communicated a paper to the Royal Society, containing an account of observations made upon the Mer de Glace of Chamouni. In addition to the questions there discussed, another of great importance occupied my attention during my sojourn at the Montanvert, and that was the veined structure of the ice. To obtain information on this head, I visited almost every portion of the Mer de Glace and its tributaries; I examined the Talèfre and Léchaud glaciers, and spent several days amid the seracs of the Glacier du Géant. To investigate the connexion, if any, between the structure of the glacier and the stratification of its névé I ascended the Col du Géant, and afterwards inspected the magnificent ice-sections exhibited in the dislocations of the Grand Plateau and other portions of Mont Blanc. During this investigation my convictions were by no means fixed; cases strongly suggestive of the influence of pressure, in producing the structure, came before me, and again other cases appeared which suggested, with almost equal force, the influence of stratification. The result, however, of the observations on the Mer de Glace was a strong opinion that pressure was the true cause of the phenomenon. But I could not help feeling that the facts and arguments which I was in a position to bring forward would still leave the question an open one. They might influence the opinions of others, as they had influenced mine; but I had nothing to advance on which the mind could rest with perfect certainty. In short, neither the Mer de Glace nor its tributaries furnished facts capable of completely deciding the question. The subject being one on which a great deal had been written and retracted, I was unwilling to swell the bulk of the literature connected with it, while a possibility remained that what I had to say upon the subject might also require withdrawal. I therefore thought it better to wait another year; to extend the range of my observations, to visit glaciers in which the mechanical conditions of strain and pressure were different from those of the Mer de Glace. Thus by varying the circumstances, and observing Nature at work under different conditions, I hoped to confer upon the investigation the character and precision of an experimental inquiry. The course of the inquiry in 1858 was as follows:—I first examined the glaciers of Grindelwald; crossing the Strahleck, I ascended the lower glacier of the Aar to the Grimsel, thence to the glacier of the Rhone, thence to the great Aletsch glacier, in the neighbourhood of which I remained eight days. I afterwards spent eleven days at the Riffelberg, and explored the entire system of glaciers between the Monte Rosa and Mont Cervin. I thence proceeded to the Matmark Alp, and remained for five days in the vicinity of the Allalein glacier; I afterwards visited the Fée glacier, and completed the expedition by a visit to the Mer de Glace and its tributaries, and a second ascent to the summit of Mont Blanc. The present paper contains the evidence derived from the sources thus opened to me; and I shall take these sources in the order in which they come before me; the evidence is therefore necessarily of a varied character, and it will I think be found conclusive. Besides those sections which are immediately devoted to the subject of structure, the paper contains others on the cause of the flattening of the air-bubbles in glacier ice, on the problem of glacier motion, and on the origin and cause of the Dirt-bands of the Mer de Glace. § 2. General Aspect of the Veined Structure. The general appearance of the veined structure is well known. The ice of glaciers, especially midway between their mountain sources and their extremities, is of a whitish hue, owing to the number of small air-bubbles enclosed in the mass—the residue, doubtless, of that air which was originally entangled in the snow of which the ice is composed. Through the general whitish mass, however, at some places, blue veins are drawn, so numerous indeed, in some cases, as to cause the blue ice to predominate over the white. A laminated appearance is thus conferred upon the ice, the cause of the blueness being, that for some reason or other, the bubbles distributed throughout the general mass do not exist in the veins, or exist there in much smaller numbers. In different glaciers, and in different portions of the same glacier, these veins exist in different stages of perfection. On the clean walls of some crevasses, and in the channels worked in the ice by glacial streams, they present a most beautiful appearance. They are not to be regarded as a partial phenomenon, or as affecting the constitution of glaciers to a small extent only. Vast masses of some glaciers are thus affected: by far the greater part of the Mer de Glace, and its tributaries, is composed of this laminated ice. The lower portion of the glacier of the Rhone, from the base of the ice cascade downwards, is entirely composed of it, and numerous similar cases might be cited. To observe the structure of a glacier it is not even necessary to see the blue veins. Those who have ascended Snowdon, or wandered among the hills of Cumberland, or even walked in the environs of Leeds or other towns in Yorkshire and Lancashire, where the stratified sandstone of the district is used for architectural purposes, will have observed the exposed edges of the slate rocks, and of the stratified sandstone, to be grooved and furrowed by the action of the weather. In fact some portions of such rocks withstand the action of the atmosphere better than others, and these more resisting portions stand out in ridges while the softer portions between them are worn away. An effect exactly similar is observed upon the surface of the glacier. The laminated ice, exposed to the sun, and to wasting atmospheric influences, melts in a manner similar to the wasting of the rocks; little grooves and little ridges are formed upon the surface of the glacier, the latter being due to the more resisting ice, while the grooves are produced by the melting of the less resisting mass between them. The consequence of this is, that the light dirt scattered by winds and avalanches over the surface of the glacier is gradually washed into the little grooves, thus forming fine lines, which to the practised eye are an infallible indication of the structure of the ice underneath. Visitors to the Jardin have ample occasion to observe these striæ, for they are finely shown upon the surface of the Mer de Glace between the Augle and Trelaporte. When they are followed until they are intersected by a fissure or a stream, it is seen that the superficial groovings always mark the direction of the veined structure within the glacier. § 3. Structure and Stratification:—Marginal Structure. Opinions at present are very diverse as to the origin of these veins. Professor Forbes first regarded them as being caused by the freezing of water which filled fissures in the ice, but he now discards the notion of freezing, and supposes the "incipient fissures" to be closed by "time and cohesion." M. Agassiz despairs of rendering an account of them, but calls them "bands of infiltration." The Messrs. Schlagintweit have also treated the question, but with no greater success. In the paper published by Mr. Huxley and myself, pressure is referred to as the probable cause of the phenomenon, but we were unable at the time to furnish proofs of this. Apart from those who give public expression to their views upon the subject, I know that there are many who reject the pressure theory, and adopt instead of it the explanation that the blue veins of the glacier are merely the continuation of the strata of the névé; a view which has recently been upheld by Mr. John Ball in the Philosophical Magazine. The matter indeed so stands, that in a recent résumé of glacier investigations, Professor Mousson of Zürich omitted the subject of structure altogether, for the express reason that the question is still in complete obscurity. I will not take up the time of the Society in discussing the vague, involved, and often absurd explanations which have been given of the blue veins of glaciers, but state broadly that the question now rests between the pressure theory, and that of stratification. Taking the parallel geological phenomena, the question then is, Does the veined structure of glaciers correspond to the stratification or the cleavage of rocks? In reply to this question, I will remark, in the first place, that the veins are not always, nor even generally, such as we should expect from stratification. The latter ought to furnish us with distinct planes extending parallel to each other for great distances through the glacier; this is by no means the general character of the veins. We observe blue streaks, some a few inches, some a foot, and some several feet in length upon the walls of the same crevasse, and varying from a fraction of an inch to several inches in thickness. In many cases the blue spaces are definitely bounded, giving rise to the lenticular structure described by Mr. Huxley and myself, but more usually they lose themselves as pale washy streaks in the general mass of the white ice. In fig. 1, I have endeavoured to give an idea of a very common aspect of the veined structure. Such a structure is not that which we should expect from bedding. Again, taking the Glacier du Géant as a representative case, we have first of all the slopes of the Col du Géant, the collectors of the snow by which the glacier is formed. The fissures on these slopes exhibit beautifully, to a certain depth, the horizontal stratification. The lines of bedding may be seen as far down as the summit of the great ice-fall between the Rognon and the Aiguille Noire; and on the castellated masses at the summit of this fall, to which the name seracs has been applied, the lines of stratification may be distinctly seen. Escaped from the confusion of the fall, the glacier flows gently through a long valley towards its junction with the Léchaud and Talèfre at the Tacul. Throughout the entire length of this glacier the planes of the structure are vertical or nearly so; sometimes they dip a little forward, but at other places they dip an equal quantity backward. Now let the mind figure, if it can, an agency which, as the mass descends the fall, shall turn up the horizontal strata of the Col du Géant and set them vertical, without a single break, throughout the entire length of the Glacier du Géant, and I imagine the effort to conceive of such an agency will be followed by the conviction that the change indicated is inconceivable. Further, we often find, in the central portions of a glacier, the structure feeble, or scarcely developed at all, while at the sides it is well developed. This is often the case where the glacier moves through a valley of tolerably uniform inclination, and where no medial moraines occur to complicate the phenomenon. But if the veins mark the bedding, there seems to be no reason why we should not find them as clearly defined at the centre as at the sides; the fact, however, certainly is that we do not so find them. Let me here show the true significance of this fact. If a plastic substance, such as mud, flow down a sloping canal, the central portions will flow more quickly than the lateral ones which are held back by friction. Now the flow may be so regulated that a circle stamped upon the central portion of such a mud-stream shall move downwards without sensible distortion, thus proving that the central mud is neither compressed nor stretched longitudinally; for if the former, the circle would be squeezed to an ellipse with its major axis transverse to the axis of the stream; and if the latter, it would be drawn out to an ellipse with its major axis parallel to the line of flow. A similar absence of longitudinal compression exists in many glaciers, and in such ice-streams there is no transverse central structure developed. But let a circle be stamped upon the mud-stream near to its side; owing to the speedier flow of the centre, this circle must be distorted to an ellipse, because the part of the circle furthest from the side moves more quickly than the part nearest the side. Hence we shall have an ellipse formed with its major axis inclined downwards, indicating that the mud is compressed in one direction and expanded in another. An exactly similar state of things occurs in many glaciers; the ice near the sides is subjected to a pressure and tension like that here indicated, and we have marginal crevasses as the result of the tension, while the veined structure is, at all events, found associated with the pressure. Fig. 2 will perhaps render my meaning more intelligible, in which \(cb\), \(cb\) represents the sides of a glacier moving in the direction of the arrow. Here, while the central circle retains its shape, the side ones are squeezed and drawn out to ellipses. Marginal crevasses occur parallel to the lines \(mn\), or perpendicular to the tension, while the dotted lines mark the direction of the blue veins which are at right angles, or nearly so, to the crevasses. I have dotted the line marking the direction of the structure along the margins \(ab\), \(ab\). In connexion with this point, I would refer to the instructive papers of Mr. Hopkins*, who has shown that in glaciers which move through valleys of uniform width, the directions of maximum pressure and tension are at right angles to each other, each of them enclosing an angle of 45 degrees with the side of the glacier. I have simply said that the structure in the case described is "associated with the pressure," thus confining myself within the strict limits of the facts. But what has been said shows that the pressure theory affords, at all events, a possible solution of a difficulty, which, without violence to fact, is inexplicable upon the hypothesis of stratification; the difficulty, namely, that a finely developed structure often exists along the margin of a glacier, while it is excessively feeble, or entirely absent, in the central portions. § 4. Transverse Structure.—Glaciers of Grindelwald, the Rhone, &c. In many cases, however, the structure is not thus limited to the margins, but sweeps across the glacier from side to side, without interruption, being as well developed at the centre as at the margins. The stratification theory is wholly incompetent to account for this; the pressure theory requires that to produce this transverse structure the glacier must, at some portion of its route, have been forcibly compressed longitudinally. It was not till after my return from the Mer de Glace in 1857, that the full mechanical significance of a change of inclination in the glacier occurred to me. Bend a prism of glass, we have compression on one side and extension on the other, * Philosophical Magazine, 1845, xxvi. p. 148. See also Proceedings of the Royal Institution, vol. ii. p. 324. with a neutral axis between, the mechanical conditions of the mass being shown by its action on polarized light. The same is true of any other substance,—the concave surface of the bent prism is compressed. Now at the bases of steep glacier slopes, where the inclination suddenly changes, we have a case of this bending, and along with it a thrust of the mass behind. The concave surface is turned towards us, and that surface is thrown into a state of compression corresponding to the thrust, and to the change of inclination. Hence it occurred to me, that the bases of the ice-falls, where the requisite change of inclination occurs, were likely to be the manufactories of the transverse structure. The experience of 1858 completely verified this idea. In illustration of my position I will take a representative case; and to render my observations capable of being easily checked, I will choose one of the most accessible glaciers in the Alps,—the lower glacier of Grindelwald. One portion of this glacier descends from the Viescherhörner; but there is another branch which descends from the Schreckhorn, Finsteraarhorn and Strahleck, and it is to this latter branch that I now wish to direct attention. Walking up this glacier from its place of junction with the tributary from the Viescherhörner, we come at length to the base of an ice-fall which forbids further advance upon the ice. Let the glacier be here forsaken, and let the flanking mountain side, either right or left, be ascended, until a position is attained which affords a complete view of the fall and of the glacier stretching downwards from the base of the fall. The view from such a position will furnish a key to the development of the transverse structure. It is, in point of fact, a grand experiment which Nature here submits to our inspection. The glacier, descending from its névé, reaches the summit of the fall and is broken transversely as it crosses the brow. It descends the fall as a succession of broken cliffy ridges, with transverse hollows between them. In these latter the ice débris and the dirt collect, partially choking up the fissures formed in the first instance. Carrying the eye downwards along the fall, we see, as we approach the base, these sharp ridges toned down, and a little below the base they dwindle into rounded protuberances which sweep, in curves, across the glacier. At the centre of the fall there is not a trace of the true structure to be observed. At the base of the fall it begins to appear,—at first feebly, but soon becomes more pronounced; until finally, at a short distance below the fall, the eye can follow the structural groovings right across the surface of the glacier, while the mass underneath has become correspondingly laminated in the most beautiful manner. It is difficult to convey, by writing, the force of the evidence which the actual observation of this great experiment places before the mind. The ice at the base of the fall has to bear the powerful thrust of the descending mass; but more than this, the sudden change of inclination which it suffers throws its upper portion into a state of violent longitudinal compression. The protuberances are squeezed more closely together, the hollows between them wrinkle up in submission to the pressure; the whole aspect of the glacier here gives evidence of the powerful exertion of the latter force; and exactly at the place where it is exerted, the structure makes its appearance, and being once manufactured, is sent onwards, giving a character to other portions of the ice-stream which have no share in its production. An illustration, perhaps equally good and equally accessible, is furnished by the glacier of the Rhone. Above the great icefall which the traveller descending from the Furca has to his right, the horizontal bedding is exhibited in a more or less perfect manner, to a certain depth, upon the walls of the huge and numerous crevasses here existing. I have also examined this fall from both sides, and an ordinary mountaineer will find no difficulty in reaching a spot nearly opposite the centre of the fall, from which both the fall itself and the glacier below it are distinctly visible. Here a similar state of things to that already presented to his view reveals itself. The fall is structureless; the cliffy ridges are separated from each other by transverse hollows, following each other in succession down the slope; those ridges are toned down to protuberances at the base of the fall, becoming more and more subdued, until low down the glacier the transverse swellings disappear. As in the case of the Grindelwald glacier, the squeezing of the protuberances and of the spaces between them is visibly manifested. Where this squeezing commences the transverse structure also commences, and in a very short distance reaches perfection. All the ice that forms the lower portion of the glacier has to pass through this structure mill at the base of the fall, and the consequence is that it is all laminated. The case will be better appreciated by reference to figs. 3 and 4, the former being a sketch, in plan, and the latter a sketch in section of a part of the ice-fall and of the lower portion of the glacier of the Rhone. \(aebf\) is the gorge of the fall, and \(fb\) its base. The transverse ridges are shown crossing the fall, being subdued at the base to protuberances, which gradually disappear further down the glacier. The "structure" sweeps across the glacier in the direction of the fine curved lines. On the plan I have also endeavoured to show the radial crevasses of the glacier; they are at right angles, or nearly so, to the structure. As would be inferred by those acquainted with what I have already written upon the influence of curvature, the side \(bcd\) of the glacier is much more violently crevassed than the side \(fgh\). Fig. 4 shows the cliffy ridges of the fall, and of the rounded protuberances below it, in section. The shading lines below denote the structure. The protuberances are so powerfully squeezed in some cases that they scale off at their surfaces. Fig. 5 is a representation of such scaling off which I have observed at the bases of several cascades, including those of Grindelwald, the Rhone, the Rognon, and the Talèfre, each of which has also its "structure mill" at the base of its cascade. Fig. 5a is an example of scaling off which I have produced by artificial pressure. It is to be borne in mind that the structure once formed prolongs itself into places which have no part in its formation; it would therefore be hasty to infer the relationship of structure and pressure from an observation of them at a particular portion of the glacier. I have sometimes seen the veined structure parallel to the crevasses for a short distance: there are some transverse crevasses on the Glacier du Géant a little above Trelaporte which illustrate this; but it would be altogether erroneous to infer from this that the law which makes the structure perpendicular to the pressure, and hence, as a general rule, transverse to the crevasses, finds an exception here. It is perfectly manifest that the structure which is brought into this unusual relationship to the crevasses has been developed far higher up; that the change of conditions from longitudinal pressure to longitudinal strain is too weak and transitory to obliterate it. To effect obliteration, a force commensurate with that which produced the structure must be brought into play, and at the place now referred to no such force exists. § 5. The Aletsch Glacier. Having made the foregoing observations upon the glacier of the Rhone, I proceeded to the Aletsch glacier, and during a residence of eight days at the hotel upon the slope of the Eggischhorn, made frequent excursions upon the ice. I had never previously seen this grand ice-stream, and my interest in it, at the time of my visit, was greatly augmented by the arguments which Mr. John Ball had founded upon its deportment against the pressure theory of the veined structure. I shall here limit myself to a few brief remarks upon this subject. I have already stated, and this must be particularly remembered, that the veined structure often appears in places which have no share in its production. The longitudinal structure in the centre of the stream of the Aletsch, for four miles above the base of the Eggischhorn, is not due to the lateral pressure endured by the glacier during these four miles. It is due, as Mr. Ball himself suggests, to the mutual thrust of the branch glaciers, which unite to form the trunk stream; and, once formed by this thrust, it perpetuates itself throughout a great portion of the trunk stream. But it is urged against this view, that pressure exerted in new directions—the longitudinal pressure, for example, endured by the stream in its descent, and acting through long periods, ought,—if pressure has the power ascribed to it, to obliterate the first structure. Now here, again, it must be remembered that it is the portions of the ice near the bed of the glacier that yield, and that the upper portions of the ice, in many cases, are simply floated upon the moving under portions. Were the uniform "long reach" referred to by Mr. Ball strictly examined, it would, in all probability, be found that the ice near the surface is no more compressed than a log of timber would be if placed upon the glacier, and permitted to share its motion downwards. I may sum up by saying that a close examination of the glacier satisfied me, not only that it presented no phenomena which were at variance with the pressure theory, but also exhibited some which, as far as I could see, were perfectly fatal to the theory of stratification. The state of the ice at the base of the Eggischhorn, as shown in fig. 6, is certainly quite in harmony with the pressure theory; another fact observed upon the glacier shall be referred to at a future page. § 6. Glaciers of Monte Rosa. I will next endeavour to describe the phenomena of structure exhibited in the system of glaciers in the neighbourhood of Monte Rosa. The general mechanical conditions of these glaciers will be evident to an observer stationed upon the Görner Grat, a point of view well known to travellers, and famous for the magnificence of the panorama which it commands. As the observer stands here, facing Monte Rosa, the great Görner glacier, coming down from the heights of the old Weissthor at his left, flows beneath him. It is joined, in its course, by a series of glaciers from the sides of the opposite range of mountains. First of all comes the western glacier of Monte Rosa, which really ought to give its name to the trunk stream, as it is the most considerable of its tributaries. Into the glacier of Monte Rosa, and before the latter reaches the trunk valley, a glacier from the Twins, Castor and Pollux, pours its contents. Afterwards we have the Schwarze glacier, which lies between the Twins and the Breithorn; then the Trifti glacier, which lies upon the flank of the Breithorn, and afterwards the glaciers of the little Mont Cervin and of St. Theodule. The accompanying sketch (fig. 7) will render intelligible what I have to say regarding these glaciers. The small Görnerhorn glacier, which comes down the sides of Monte Rosa, is a very singular one. In comparison with the western glacier of Monte Rosa its mass is insignificant, and it is abruptly cut off by the latter along the line $a b$, a moraine occurring here, which may be regarded as forming, at once, the terminal moraine of the one glacier and the lateral moraine of the other. Thus the smaller glacier coming down the mountain side abuts against its more powerful neighbour, and we should infer from the inspection of the glacier that its terminus is subjected to great pressure. Let the observer now suppose himself transported to the Görnerhorn glacier, at some distance above the terminal moraine \(a b\); he will find there the transverse structure, if at all developed, excessively feeble and defective; let him now walk downwards towards the moraine \(a b\): every step he takes brings him to a place where the ice is subjected to a greater pressure, and every step also brings him to a better structure: both phenomena go hand in hand. At the end of the glacier, alongside the terminal moraine and under it, the structure is finely developed. If the observer now cross the glacier and ascend the rocks called *Auf der Platte*, from which he can command a near view of the Görnerhorn glacier, and embrace a large portion of it, he will be able to observe the gradual perfecting of the structure as the region of pressure is approached. Towards the extremity of the glacier the surface becomes wrinkled, the groovings denoting the structure become more and more pronounced, the dirt striæ being more closely squeezed together; and from these external aspects he may infer, with certainty, the gradual perfecting of structure within the glacier. The western glacier of Monte Rosa next commands our attention. This great stream occupies the valley between Monte Rosa and the Lyskamm, receiving the snows of the opposite sides of both. The branch of the Görner glacier coming down from the Weissthor throws itself across the flow of its powerful neighbour, and deflects the latter, both of them afterwards moving together down the trunk valley, with a moraine, as usual, between them. Before quitting the "Platte," we will suppose that the observer has endeavoured to form some idea of the mechanical conditions of the Monte Rosa glacier. He would see the mass arrested in its descent by the Görner glacier, and compelled to accompany the latter. A certain component of the weight of the glacier is borne by the ice where it comes into contact with the Görner glacier. The observer would infer, from mere inspection, that if the structure be due to pressure, it ought to be most fully developed near the moraine which separates the Monte Rosa from the Görner glacier. If he now pass from the "Platte" to the ice, and cross to the centre of the Monte Rosa glacier to \(A\), he will find the structure there excessively feeble, if at all developed. Let him now walk straight down the glacier towards \(B\), where the pressure is most intense. Every step he takes downwards brings him to more perfectly veined ice; and I am not acquainted with a more splendid example of laminated structure than that exhibited by this glacier along the moraine, and for some distance from it, at its southern side. The system of glaciers which next come under review are exceedingly instructive. In no place in the whole range of the Alps are the effects of pressure and the phenomena of structure more strikingly exhibited. I have endeavoured, in the sketch, to render the aspect of these glaciers intelligible. The Schwarze glacier moves down a steep mountain slope, and welds itself to the Monte Rosa glacier at the bottom. But the great mass of this latter enables it to pursue its way without being compelled to swerve sensibly by its feebler neighbour. The latter is forced to bend abruptly, and from a wide irregularly-shaped field of névé, it is squeezed between the Trifti and the Monte Rosa glaciers to the narrow band represented in the figure, and moves thus downwards. The Trifti glacier itself is perhaps a still more striking illustration of the power of ice to yield when subjected to pressure for a long period. The aspect of the real glacier is much the same as that shown upon the sketch. It also is compelled to change its direction, and to flow as a narrow stripe along the trunk valley, being hemmed in between the strip of the Schwarze glacier and that of the glacier of the little Mont Cervin. A beautiful system of bands is to be seen at the lower portion of this glacier. The inspection of the sketch will show better than words the modifications of shape which the lower portion of the glaciers undergo by the pressure of their higher portions, and the resistance of the trunk stream. They are turned aside, firmly welded together, and form a series of parallel narrow bands, separated from each other by moraines. They are all well seen either from the Görnergrat or the summit of the Riffelhorn. I have examined each of these glaciers, and find the same to be true of all of them. High up the structure is feeble; as we descend it becomes more pronounced, and at the places where the tributaries join the trunk, and the ice has to bear the full thrust of the mass behind it, we have a finely developed structure. § 7. Coexistence of Structure and Stratification.—The Furgge Glacier. The evidence of the association of pressure and glacier lamination which I have thus far laid before the Society, will, I think, be admitted to be very strong. I have no hesitation in saying that the stratification theory has nothing to urge at all to be compared with it in point of cogency. Still I cannot help feeling how a critical and well-informed mind might weaken the force of what I have adduced. Difficult as the conception is, it might be urged that the structure, so fully developed near the margins of glaciers, may be due to a turning up of the strata edgeways, in consequence of a wide névé being squeezed into a narrow channel,—just as a sheet of paper, if forced through a groove less than itself in width, would turn up at its edges. It might also be urged that the structure developed alongside and under the medial moraines, is due to the placing side by side of these folded-up strata; the perfect welding of both and the clearer development of the structure being conceded as possible consequences of the mutual pressure. This indeed is Mr. Ball's view of the subject; and M. Agassiz assumes such a folding-up of the strata of the Unteraar glacier. With regard to the transverse structure also, it might be said that we do not know how the interior of the mass is affected in descending the ice-falls. The mind, it is true, finds great difficulty in conceiving of any agency which could set the strata which were horizontal above the fall, vertical below it; still this difficulty may be due to our ignorance of the mechanical conditions of the mass during its descent. In this way it would be quite possible to fritter away conviction to a mere opinion, and hence arose a strong desire on my part, either to confirm these surmises, or to place the pressure theory, once for all, beyond the power of such attacks. One conclusive observation is still wanting to establish the analogy between glacier lamination and the cleavage of slate rocks. In the latter case the arrangement of the strata has been traced by their organic remains; and, indeed, stratification has often been visibly exhibited coexistent with cleavage, both crossing each other at a high angle. If a similar state of things could be detected upon a glacier, it would at once lay the axe to the root of all the scruples above referred to, and place the pressure theory upon an unassailable basis. The consciousness of this was sufficient to stimulate me in the search of such evidence. I had visited all the glaciers hitherto mentioned, and others not mentioned, without obtaining more than one clear case of the kind: this case I observed upon the Aletsch glacier on the 6th of August. Not far from the junction of the Middle Aletsch glacier with the trunk stream, a crevasse exposed a wall of ice 50 or 60 feet in height, upon which the stratification was exposed, and cutting the stratification at a high angle were the groovings which marked the true veined structure. The association was distinct; my friend Professor Ramsay was with me at the time; I drew his attention to the fact, and to him the case was perfectly conclusive. Thus the Aletsch glacier, which had been referred to by Mr. Ball as furnishing evidence against the pressure theory, gave us a fact, which, as far as I could see, was perfectly fatal to the theory of stratification. But the case was solitary, and although inspiriting at the moment, its effect upon the mind became feeble as time passed, and no repetition of the observation occurred. I had remained at the Riffel from the 9th to the 18th of August, exploring all the adjacent glaciers, and adding each day to my stock of knowledge; but I met no case in which the structure and the bedding were so clearly and independently exhibited, as to leave an adherent of the stratification theory no room for doubt. Wednesday the 18th of August was to be my last day at the Riffel, and it was devoted to the examination of the Furgge glacier, which occupies the space between the pass of St. Theodule and the Matterhorn. Crossing the valley of the Görner glacier, I climbed the opposite mountain slope, and passing the Schwarze See, soon came upon the glacier referred to. I walked up it until I found myself in a kind of cul de sac, flanked by precipitous ice-slopes, and opposed in front by a cascade composed of four high terraces of ice. The highest terrace was composed principally of broken cliffs and peaks of ice, and it had let some of its frozen boulders fall upon the platform of the second terrace, where they stood like rocking-stones on the point of falling. The whole space at the foot of the fall was covered with quantities of crushed ice, while some coherent masses, upwards of 200 cubic feet in volume, were cast to a considerable distance down the glacier. Upon the face of the terraces the stratification of the névé was beautifully shown. Above the fall the névé extends as a frozen plain, quite undisturbed, so that the bedding took place with great regularity; and being broken through for the first time at the summit of the fall, the lines of stratification were peculiarly well defined and beautiful. Towards the right of the fall, looking upwards, this was particularly the case; for here no pressure had been exerted upon the beds sufficient to contort them or to rupture their continuity. The figure of a vast lake pouring its waters over a rocky barrier, which curves convexly upwards, thus causing the water to rush down it, not only longitudinally over the vertex of the curve, but also laterally over its two arms, will convey to the mind a tolerably correct conception of the appearance of the fall. Towards the centre the ice was powerfully squeezed; the beds were bent, and their continuity often ruptured, so as to exhibit faults; but they were as plain, and as easily traced, as in any other portion of the fall. I thought I saw structural groovings running at a high angle to the stratification. Had the question been an undisputed one I should have felt sure of this, for the groovings were such as always mark the structure. The place being dangerous, I first observed it from a little distance through my opera-glass; but at length, resigning the instrument to my guide, and leaving him to watch the tottering blocks overhead, and to give me warning in case of their giving way, I went forward to the base of the fall, peeled the grooved surface away with my axe, and found the true veined structure underneath, running, in this case, nearly at right angles to the stratification. The superficial groovings were not uniformly distributed over the whole face of the terrace, but occurred here and there where the ice had yielded most to the pressure. I examined several of these places, and in each instance found the superficial grooving to be the exponent of the true veined structure underneath, the structure being in general nearly vertical, while the lines of bedding were horizontal. The coarse bands which marked the division of the beds were also seen underneath, when the surface of the ice was removed. Having perfectly, and with deliberation, satisfied myself of these facts, I made a speedy retreat; for the ice blocks were most threatening, and the time of day that at which they fall most frequently. We now resolved to try the ascent of the glacier to the right; it was much riven, but perfectly practicable to a good iceman. To me it was also perfectly delightful; in fact, as regards the relationship of structure and stratification, this glacier taught me more than all the others I had visited taken together. Our way lay through fissures which exposed magnificent sections, and every step forward added further demonstration to what I had already observed at the base of the fall. The bedding was perfectly distinct, and the structure equally so, the one being at a high angle—sometimes at a right angle—to the other. Among these crevasses the pressure was in some cases greater than on the fall, and the structure proportionally more pronounced. The crumpling of the beds demonstrated the exercise of the pressure, and the structure went straight through such crumplings, thus furnishing me with numerous parallels to the case observed by Professor Sedgwick, Mr. Sorby, and others, of the passage of slaty cleavage through contorted beds. Indeed I question whether the phenomena of cleavage and bedding, in the case of slate rocks, were ever exhibited, side by side, with a distinctness equal to that of the stratification and "structure" of ice in the present instance. Fig. 9 represents a crumpled portion of the ice, with the lines of lamination passing through those of bedding at a high angle. Fig. 10 represents a case where a fault occurred, the veins at both sides of the line of dislocation \( ab \) being inclined towards each other. The lines \( mn, mn \) represent of course the lines of bedding, and the lines crossing them the structure. These observations are conclusive as regard the claims of the rival theories of structure and stratification*. § 8. On the White Ice-seams of the Glacier du Géant, and their relation to the Veined Structure. From an elevated point at Trelaporte I observed a remarkable system of white bands sweeping across the Glacier du Géant in the direction of the structure. From one of the moraines near the junction of the three tributary glaciers, the same system of bands present a very striking appearance. They consist of a hard white ice, more resistant than the general mass of the glacier, and in some cases rising to a height of three or four feet above the surface. On close examination I found that they penetrated the glacier only to a limited depth. In fig. 11 I have given the sections of two of these * While correcting this proof, I find a case of this kind figured by M. Agassiz in the atlas to his 'Système Glaciaire,' pl. 8. fig. 3. veins, about 15 feet deep, which were exposed on the walls of a crevasse high up the glacier. They constituted a kind of *inverted glacier trap*, and I was led to a knowledge of their origin in the following way. In one of my earliest visits to the base of the ice-fall of the Talèfre, I observed a curious disposition of the veined structure on the walls of some of the crevasses: fig. 12 represents one case of the kind, and fig. 13 another, and numerous similar ones find a place in my note-book. In the former case the veins fell *backward* as well as forward, being vertical through the central portion of the curve. In fig. 13 the position of the veins varies in a very short distance from the vertical to the horizontal. I found that the portions of ice which showed the phenomena, formed, when seen from a point of view sufficiently commanding, a part of a system of crumples or protuberances which swept round the base of the fall, between the moraine which descends along it from the Jardin, and its highest lateral moraine. I have already referred to the protuberances which sweep across the Strahleck branch of the Lower Grindelwald glacier, and of those of the glacier of the Rhone; those to which I now refer were of the same character. Right and left from the position where the crumples were most pronounced they gradually became subdued, shading off to a mere undulating surface; the sides of a crevasse intersecting this surface longitudinally presented the structural arrangement shown in fig. 14. It will be observed that the directions of the veins change in accordance with the undulations of the surface. Supposing the squeezing of the mass to become so violent that the gentle undulations shall become steep crumples, the deviation of the structure from parallelism with itself would, of course, be augmented. This prepares us to understand the exact phenomena observed at the base of the Talèfre cascade. Fig. 15 represents a series of crumples following each other in succession at the place referred to; at the base of each I found a vein of *white* ice, $a$, wedged into the mass. This interrupted the continuity of the structure; the abrupt change in its direction at opposite sides of the white band being, as shown in the figure, in every case observed. I found that the width of the seams was exceedingly irregular, varying, at different portions of the same seam, between 6 inches and 3 or 4 feet. I also found that a seam sometimes became forked so as to form two branches, which thinned gradually off until they finally vanished. Fig. 16 is an example of this kind: the seam was divided at the point \(a\); one of its branches ran up the face of the crumple, thinned off and disappeared at \(b\); the other widened considerably, but finally thinned off and also vanished. Along the bases of the crumples the fillets of water which poured down their faces were collected and flowed. The streams thus formed ran in many cases alongside the existing veins of white ice, and had worn for themselves deep channels in the glacier. The thought soon suggested itself, that the seams themselves were formed by the gorging up of those channels by snow in winter, and the subsequent consolidation of this snow during the descent of the glacier. Indeed the channels of the streams seemed the exact matrices of the seams of white ice*. * The fact of one branch of a vein running up the face of a crumple, seems to prove that the ice, which at one time constitutes the base of a crumple, does not always remain so; the bases of the crumples are sometimes lifted up by the squeezing. The horizontal structure at the fronts of many of the crumples seems due to a local forcing forward of one protuberance over that next below it. Were the matter tested by strict measurement, I think it would be found that different portions of the crumples move downwards with different velocities. According to this view, upon the general motion of the glacier there are local motions superposed. I afterwards traced the seams of white ice of the Glacier du Géant to their origin amid the ridges and hollows at the base of the great ice-fall of Le Rognon. In some cases the seams opened out into two branches, which, after remaining for some distance separate, would unite again so as to enclose a little glacial island; at other places lateral branches were thrown off from the principal seam, presenting the form of a glacier stream which had been fed by tributary branches. Fig. 17 is the plan of an actual stream observed at the base of the ice-fall; fig. 18 is the plan of a seam of white ice observed the same day lower down the glacier; their relationship is evident. I may remark that I have observed other seams produced by the gorging of crevasses with snow, and the subsequent closure of the fissures. Considering the place where they are formed, these channels cannot escape compression; but let me remove all uncertainty on this point, by proving that not only at the base of the seracs, but throughout almost its entire length, the Glacier du Géant is in a state of longitudinal compression. The first proof I have to offer is that the transverse undulations of the glacier, to which reference has been so often made, become gradually shorter as they descend. A series of three of them, measured along the axis of the glacier on the 6th of August 1857, gave the following respective lengths—955, 855, and 770 links, the shortest undulation being the furthest down the glacier. Now these undulations, as I shall subsequently show, are due to a regularly recurrent action, and are doubtless originally of the same length; that the lower ones are shorter than the higher must therefore be due to compression. The following observation is, however, more conclusive. About three-quarters of a mile above the Tacul, and to the left as we ascend, there is a green patch upon the craggy mountain side. From this spot, as a station, I set out with a theodolite a line (No. 1) transverse to the axis of the glacier. From a station lower down, chosen in a couloir along which the stones are discharged from the end of a secondary glacier which hangs upon the slope of Mont Tacul, I set out a second line (No. 2) transverse to the axis of the glacier. A third line (No. 3) was set out across the glacier about a quarter of a mile still lower*. The mean daily motion of the centres of these three lines is given in the annexed Table, and also their distances apart. | Mean daily motion of three points upon the axis of the Glacier du Géant. | |-----------------------------|-----------------| | inches. | Distances apart.| | No. 1 . . . . . | 20·55 | 2477 links. | | No. 2 . . . . . | 15·43 | | | No. 3 . . . . . | 12·75 | 2215 links. | The advance of the hinder lines upon these in front is most strikingly shown by these measurements; and the proof that the Glacier du Géant is in a state of longitudinal compression is thus complete. Here then we have a vast ice press, and here we have the pure snow filling the transverse channels of the streams. We are thus furnished with an experimental test on a grand scale of the pressure theory of the veined structure. In 1857 I examined a great number of these seams of white ice, and found in many of them a finely developed lenticular structure. In 1858 I also examined the seams, and found some of them "ribboned" in the most exquisite manner by the blue veins; indeed I had never seen the veins more sharply and beautifully developed. This structure was observed in portions of the seams at and near the centre of the glacier, where the differential motion observed at the sides does not exist. This fact, I think, throws grave difficulties in the way of any theory which makes the veined structure dependent on differential motion, and more especially a theory which requires "a very considerable amount of this differential motion to produce any sensible degree of stratification in the vesicles." § 9. On the flattening of Air-bubbles in Glacier Ice, and its relation to the Veined Structure. Those who have given their attention to the subject, know that the bubbles contained in glacier ice are, in general, not spherical, but flattened; and that from their shape conclusions of the greatest import have been drawn regarding the internal pressures of glaciers. M. Agassiz draws attention to this subject in the following words:—"The air-bubbles undergo no less curious modifications; in the neighbourhood of the névé, where they are most numerous, those which one sees on the surface are all spherical or ovoid, but by degrees they begin to be flattened, and near the end of the glacier there are some that are so flat that they might be taken for fissures when seen in profile. The drawing, * These three lines are drawn upon the sketch map at page 268 of Part I. of these researches (FF', GG', HH'). fig. 10, represents a piece of ice detached from the gallery of infiltration; all the bubbles are greatly flattened. But what is most extraordinary is, that far from being uniform, the flattening is different in each fragment; so that the bubbles, according to the face which they offer, appear either very broad or very thin. I know of no more significant fact than this, since it demonstrates that each fragment of ice is capable of undergoing in the interior of the glacier a proper displacement independently of the movement of the whole. "The same flattening of the bubbles," continues M. Agassiz, "is found at a greater depth. While engaged in my boring experiments, I observed attentively the fragments of ice brought up by the borer. I found in them almost flat bubbles, perfectly similar to those of the fragment figured above, at all depths from 10 to 65 metres. It follows hence that a strong pressure is exercised on the interior of the glacier." The description of the "flattening" here given is correct: all observers agree in corroborating it, and every observer with whom I am acquainted draws substantially the same conclusion from the phenomenon that M. Agassiz does. Professor Thomson's speculation upon the subject is particularly refined and ingenious. Mr. John Ball converts the flattening of the bubbles into evidence against the pressure theory of the structure in the following way:—"As Agassiz has pointed out," writes Mr. Ball, "and I have frequently verified his observations upon this point, though the air-cavities show traces of compression reducing them to the form of flattened lenses, the directions in which they are flattened are most various, and show no constant relationship to the planes of the veined structure. Here then we have direct evidence that separate portions of the ice have been acted upon by pressure sufficient in amount to modify their internal arrangement, but that these pressures have not acted in the same, or nearly the same direction." Granting the inference that the observed flattening "furnishes direct evidence" of pressure, the foregoing argument would, I confess, be a very formidable one. If the bubbles are thus flattened by pressure, and if the veined structure, as I contend, be the result of pressure, and approximately at right angles to the direction of the force, we ought to have the bubbles squeezed out in planes parallel to the structure. The fact that the bubbles are not so squeezed out, would then afford a strong presumption that the structure is not produced by pressure. I expect, however, to be able to prove that the shape of the bubbles is not a "direct evidence" of pressure, as hitherto assumed; and I think, as I do so, it will be seen how necessary it is to associate experiment with an inquiry of this kind, if we would read aright our observations. In a paper in the Philosophical Transactions on the Physical Properties of Ice, I have shown that when a sunbeam traverses a mass of ice, the latter melts at innumerable points in the track of the beam, and that each portion melted assumes the form, not of a globule, but of a flower of six petals. The planes in which these flowers are formed are independent of the shape of the mass and of the direction of the beam through it; they are always formed parallel to the surface of freezing. This is a natural consequence of the manner in which the particles of ice are set together by the crystallizing force. By the slow abstraction of heat from water its particles build themselves into these little stars, and by the introduction of heat into a mass so built the architecture is taken down in a reverse order. In watching the formation of artificial ice, by the machine of Mr. Harrison referred to in my paper, I have seen little solid stars formed, by freezing, which were the exact counterparts of the little liquid stars formed by melting. So far as I can see, the complementary character of the phenomena is perfectly natural, and presents no difficulty to the mind in conceiving of it. When the beam is intense, and its action continued for some time, the flowers expand, so as to form liquid plates within the mass. Looked at edgeways, these liquid spaces appear like fine lines; which proves that the melting is not symmetrical laterally and vertically, but that the ice melts in the planes of freezing much more readily than at right angles to these planes. If an air-bubble exists within ice, and if the ice melts at the concave surface of this bubble, as might be expected from the foregoing facts, the ice will so yield that the composite cell of air and water will not be spherical, even though the bubble of air may originally have been so. In the planes of freezing the mass yields most readily, and the cavity containing the air and water will appear as if flattened by a force acting perpendicular to these planes. This is not a deduction merely, but an observation which I have made in a hundred different cases. What I have here said applies to ordinary lake ice; but glacier ice has no definite "planes of freezing." The substance is first snow, which sometimes, it is true, falls regularly in six-rayed crystals, as observed by myself on the summit of Monte Rosa; but it is usually disturbed by winds, while falling, and whirled and tossed by the same agency after it has fallen; the mountain snow is often melted, mixed with water and refrozen. Even after it has become consolidated it is often shattered in descending precipitous slopes. In such ice definite planes of crystallization are, of course, not to be expected. If we suppose a mass of lake ice to be broken up into fragments, and these fragments thrown together confusedly and regelated in their new positions to a continuous mass, we have an exact image of the character of the glacier ice in which this flattening of the bubbles in different directions has been observed. In the paper already referred to, I have given a sketch of a piece of ice composed of such segments, and have described the effects obtained with it. That ice was sold to me as Norway lake ice. I am not aware whether glacier ice is ever imported into this country from Norway; but if it be, the piece in question must, I think, have belonged to it. It is so like all the glacier ice that I have examined since that time, and so unlike all the lake ice, that I feel little hesitation in saying that it belonged to the former*. No matter how coherent and optically continuous a mass of ice may be, a condensed sunbeam would at once tell us whether it belonged to a lake or to a glacier. * Perhaps formed from the connecting together of confused fragments. I have given in fig. 19 a sketch of a piece of ice taken from the end of the great Allalein glacier, on the Swiss side of the Monte Moro. On reference to M. Agassiz's figure, it will be quite manifest that we are both dealing with the same phenomenon; we have the division of the ice into "angular fragments," the flattening of the "bubbles," and the non-parallelism of their directions in the different fragments. Fig. 20 is a sketch of a piece of ice which showed the veined structure. The line AB was parallel to the veins, and it will be seen that the "bubbles" are inclined to this line at different angles, and in different azimuths. The circles indicate, of course, that the "bubbles" were there parallel to the horizontal face of the slab, while the lines indicate that they were perpendicular. In one case the bubbles are seen in plan, in the other case in section. The ellipses show the bubbles foreshortened where their planes are oblique to the surface of the slab. Associated with the air-bubbles, and usually beyond comparison more numerous in ice taken from the "ends" of glaciers, were the round liquid disks which I have described in my paper on the Physical Properties of Ice. Associated with each liquid disk was a vacuous spot, which shone with exceeding lustre when the sunbeams fell upon it. That the spots were vacuous, and not bubbles of air, I proved by permitting them to collapse under warm water; the collapse was complete, and no trace of air arose from them. These, I doubt not, are the "bubbles" observed by M. Agassiz "near the end of the glacier," and which were "so flat that they might be taken for fissures when seen in profile." These "vacuum disks," as I have usually called them, were invaluable as indicators of the planes of crystallization. When a condensed sunbeam was sent through the mass, the six-petalled flowers, which always indicate the planes referred to, started into existence parallel to the disks. Consequently, as the beam passed through different fragments, flowers were formed, in different planes, along the track of the beam. True air-bubbles, associated with water, also occurred in these masses of ice, and such composite cells were always flattened out in the planes of the vacuum disks. The fact then is that many of the so-called air-bubbles are not air-bubbles at all, and that the so-called "flattening" is in reality no flattening at all; and that pressure, in the sense hitherto conceived, has had nothing whatever to do with the shape of these bubbles. In glacier ice, as in lake ice, their shape is determined by the crystalline architecture. The conclusion that they were *squeezed* flat seems to have been drawn by M. Agassiz, and reproduced by subsequent writers, without due regard to the difficulties associated with it. That the pressures of a glacier are so parcelled out as to squeeze contiguous fragments of ice, not exceeding a cubic inch in size, in all possible directions, is so improbable, that reflection alone must throw great difficulties in the way of its acceptance. It is with some diffidence that I here venture to express an opinion upon a question that I have not specially examined; but it appears to me probable that the decomposition of glacier ice into large granules, regarding which so much has been written, may be connected with the foregoing facts. The ice of glaciers is sometimes disintegrated to a great depth; causing it to resemble an aggregate of jointed polyhedra more than a coherent solid. I was very near losing my life in 1857 on the Col du Géant by trusting to such ice; and last summer I found vast masses of it at the end of the Allalein glacier. Blocks a cubic yard and upwards in volume, fell to pieces to their very centres on being overturned; they were an aggregate of granules, whose average volume scarcely exceeded a cubic inch. From the constitution which the foregoing observations assign to glacier ice, this disintegration seems natural. The substance is composed of fragments which are virtually crystallized in different planes; and it is not to be expected that the union along the surfaces, though they may be invisible when the ice is sound, is as intimate as that among the different parts of a mass homogeneously crystallized. Besides, ice no doubt, and all uniaxal crystals, expands by an augmentation of temperature, differently in different directions, and hence a differential motion of the particles on both sides of one of the above surfaces when the volume of the substance is changed by heat or cold is unavoidable. Such surfaces then would become surfaces of discontinuity, and perhaps produce that granular condition which has occupied so much of the attention of observers. § 10. Physical Analysis of the Veined Structure. The relation of pressure and structure has been shown in the foregoing pages, but the mode in which the pressure acts remains yet to be considered. As regards their causes, slaty cleavage and slaty structure have been reduced to one and the same; but as regards the operation of that cause, no two things can, I imagine, in some respects at least, be more different. In a note at page 336 of the 'Proceedings of the Royal Society' for January 1857, I refer to an experiment in which a clear mass of ice was caused by pressure to resemble a piece of fissured gypsum, and I there promised the full details of the experiment in due time. In my paper on the Physical Properties of Ice this promise is fulfilled; I have shown how a mass of compact ice may be liquefied by pressure, in parallel planes perpendicular to the direction of the force, and explained the effect by reference to the ingenious deductions of Mr. James Thomson from Carnot's maxim. Let the attention now be fixed on the state of a glacier at the base of one of the MDCCCLIX. ice-falls where it is bent so as to throw its surface into a state of longitudinal compression. According to the above experiments, the glacier whose temperature is 32° Fahr. must here be liquefied in flats, perpendicular to its axis. A liquid connexion is thus established between all the air-bubbles which are intersected by any one flat, and a means for the escape of this air from the glacier is thus furnished. The water produced is also partially expressed, partially absorbed by capillary attraction of the adjacent bubble ice, and partially refrozen when the pressure is relaxed. It is, I think, perfectly manifest that such a process, each step of which may be illustrated by experiment, must result in the formation of the blue veins*. All the experiments and observations recorded in the paper on the Physical Properties of Ice were made with reference to the glaciers, and one experiment there recorded illustrates the present point more forcibly than any words could do,—it is that in which I have thrown a prism of ice into a state of compression, which brings one side of it into the exact condition of the glacier at the base of an ice-fall. Fig. 21 is precisely the same as that given at page 226 of my paper, the prism being placed horizontal instead of vertical, so as to show its bearing upon the present point more distinctly. The original shape of the piece of ice is given in fig. 22, which by compression between the surfaces AB and CD is reduced to the shape and condition of fig. 21. The vertical lines represent the planes of liquefaction, and they correspond exactly to the planes of the blue veins in the glacier. The application of the same principles to all cases where pressure comes into play is sufficiently obvious. In the experiments with the hydraulic press, the portion of ice between each two liquefied flats transmits the pressure without sensible yielding. We have no difficulty in conceiving that the same holds true of glacier ice, and that pressure may be transmitted through one portion of a bubbled mass of ice and produce the liquefaction of another portion, without sensible distortion of the bubbles contained in the former. One of the objections which have been urged against the pressure theory is thus, I conceive, completely answered,—the objection, namely, that the white ice which transmits the pressure ought to have its bubbles flattened. Indeed this objection continued to be of weight only so long as it was imagined that the observed flat bubbles had been squeezed to this shape; a notion, which I think will no longer be entertained. * The mechanical actions which accompany the development of ordinary slaty cleavage, must, I think, also manifest themselves to some extent in the glacier. § 11. Remarks on Glacier Motion. It is only by slow degrees that we master from actual observation, a problem so large as that presented by the glaciers; the muscular labour alone being such as to render the expenditure of a considerable amount of time unavoidable. The examination of the various questions connected with glaciers, has been therefore, in my case, distributed over some years, and not until last summer was I able to devote the requisite attention to the subject of the present section, which, however, is essential to a right comprehension of the physics of glacier motion. It would be a problem eminently worthy of any geologist, to lay down upon a trustworthy map of Switzerland the directions of the striæ on the rocks over which ancient glaciers have moved; and to one who sees its importance and desires exact information upon this subject, it must be a matter of surprise that nothing of the kind, in a systematic way, has yet been attempted. A suitable map furnished with such lines of direction, carefully and conscientiously drawn, would impart more satisfactory information than all the volumes that ever have been, or ever will be written upon the subject. Here is a piece of work loudly calling for accomplishment, and one on which any young geologist may base an honourable reputation. Mr. Hopkins, I believe, was the first to urge the existence of roches polies at the ends of existing glaciers and along the continuations of existing glacier valleys as an evidence in support of the sliding theory. That such facts exist is known to every body, and that the rocks are thus polished and rounded by the glaciers sliding over them is incontrovertible. Let a traveller, if he wish to obtain a wealth of information upon this subject, transport himself to the terminus of the Unteraar glacier, and walk thence down the valley through which the river Aar now flows. On all sides he will obtain the most striking evidence that the base of the valley was once the bed of the glacier. The rocks are polished and striated, and present at some places the appearance of huge rounded mounds, which, at first sight, would appear to offer an insuperable barrier to the motion of the glacier, but which show by their aspect that the ice actually moved over them, grinding off their angles and furrowing their summits and sides. All along the valley towards Meyringen, similar evidences exist. In fact, the phenomenon is very common, and admitted on all hands. The conclusion which Mr. Hopkins has drawn from these facts is unavoidable; the glaciers must have slidden over the rocks on which such traces are left. To an eye a little practised in those matters, the precise limits reached by the ancient glaciers are perfectly visible. The junction of the rounded and abraded portions of the mountains, with those portions which in ancient times rose over the then existing ice, is perfectly distinct; and I should say in the valley of the Aar reaches to a height of more than a thousand feet above the present bed of the river. The valley of Saas, in the Canton de Valais, furnishes magnificent examples of the same kind. At all places, from the base of the ancient glacier to its surface, sliding must have occurred; the evidence of it is perfectly irresistible. The summit of the Grimsel pass constituted the bed of an ancient névé; and the groovings and polishings, at the very summit of the pass, show that the ancient névés, as well as the ancient glaciers, slid upon their beds. In company with my friend Professor Ramsay, and assisted by his great experience, I visited the sites of other ancient névés, and found the same true of all of them; they all slid more or less over their beds. No investigator of glacier motion can shut his eyes to those facts, nor refuse to give them their proper weight. The sliding theory is beyond doubt to some extent true; and many of the objections raised against it, and still repeated in works intended to instruct the public, are altogether futile. Here, as in other cases, we find that the extreme facts have been dwelt upon principally by rival theorists, and coexistent truths have, by partial treatment, been rendered apparently hostile to each other. It is perfectly certain that a glacier changes its form by pressure like a plastic mass, but it is equally true that it slides over its bed. § 12. On the Dirt-bands of the Mer de Glace. In walking over the Mer de Glace, we soon observe differences in the distribution of the dirt upon its surface; but while standing on the glacier itself, no orderly arrangement of the dirty and clean spaces is observed. From a point, however, which commands a view of a large portion of the glacier, it is seen that the dirty spaces are arranged so as to form a series of broad brown curves, which follow each other in succession down the glacier. They were first observed by Professor Forbes from the heights of Charmoz, on the 24th of July, 1842, and from the same place, on the 16th of July, 1857, I observed them. Last summer I counted eighteen of them from the same position, and this agrees with the number observed by Professor Forbes. This agreement, after an interval of sixteen years, proves the regularity of their occurrence. These bands were different from anything of the kind I had previously seen, and I felt that the explanation given of the "dirt-bands" observed by Mr. Huxley and myself would not completely account for those now before me. They were perfectly detached from each other, and resembled sharp hyperbolas with their vertices pointing downwards. From Charmoz, however, I could only see the existence of the bands, but could not see their origin. To observe this, I climbed a summit to the left of a remarkable cleft in the mountain range between the Aiguille du Charmoz and Trelaporte, which strikes most visitors to the Montanvert. The place is that referred to by Professor Forbes at page 84 of the 'Travels,' and which he was prevented from reaching by the fire of stones which a secondary glacier sent down upon him. From the pile of stones on the summit I, however, infer that he or some of his guides must have reached the place. Fig. 23 represents what I observed from this excellent station. From the base of the Aiguille du Géant and the Periades, a glacier descends which is separated by the Aiguille and promontory of La Noire from the great glacier which descends from the Col du Géant. A small moraine is formed between both, beside which the letter $a$ stands in the diagram. The glacier descending from the Col is bounded on the west by the small moraine $b$, and between $b$ and the side of the valley is another little glacier derived from one of the lateral tributaries. With regard to the "dirt-bands," the following significant fact at once revealed itself. The dirt-bands extended over that portion of the Glacier du Géant only which lay between the moraines $a$ and $b$, or, in other words, were confined to the ice which had descended the great cascade between Le Rognon and La Noire. It was perfectly evident that the cascade was in some way the cause of the bands. The description which I have already given of the ice-fall of the Rhone and of the Strahleck arm of the Lower Grindelwald glacier, applies generally to the fall of the Glacier du Géant. The terraces, however, are here larger, and the protuberances at the base of the fall of grander proportions. These latter are best seen from a point near A upon the Glacier du Géant; they are steepest on that side, in consequence of the oblique thrust of the western tributaries of the glacier. All that I have said regarding the toning down of the ridges to rounded undulations which sweep in curves across the glacier, applies here also. Referring to the section of the glacier of the Rhone in fig. 4, it will be seen that the word "dirt" is written opposite to each hollow. In fact the depressions between the protuberances are, to some extent, the collectors of the fine superficial dirt. This is also the case upon the Glacier du Géant; but here I noticed that the frontal slopes of the protuberances were also covered with a fine brown mud. Lower down the glacier the swellings disappear, but the dirt retains its position upon the ice, and afterwards constitutes the dirt bands of the Mer de Glace. A remarkable change in the form of the bands occurs where the glacier is forced through the neck of the valley at Trelaporte. They sweep across the Glacier du Géant in gentle curves with their convexity downwards; but in passing Trelaporte the arms of the curves are squeezed more closely together, the vertices are pushed sharply forwards, so that on the whole the bands resemble a series of hyperbolas which tend to coincide with their asymptotes. Looking down from the Convercle upon the Glacier du Talèfre, a series of swellings like those upon the Glacier du Géant are observed. Along the intervening hollows streams run, and sand and dirt are collected, forming the rudiments, so to speak, of a series of dirt-bands; but these latter never attain anything like the precision of those upon the Mer de Glace. I saw no such bands upon the Léchaud, for here the necessary ice-fall is absent: if bands at all exist on this glacier, they must, I imagine, be of a very rudimentary and defective character. I will not occupy the time of the Society in describing my various expeditions up the Glacier du Géant in connexion with these bands; but one circumstance, to which the definite printing of the bands is mainly due, must be mentioned. The Glacier du Géant lies nearly north and south, being only 14 degrees east of the true north. Standing with his back to the Col du Géant, an observer looks northward, and consequently the frontal slopes of the protuberances to which I have referred have a northern aspect. They therefore retain the snow upon them long after it has been melted from the general surface of the glacier. The summer of 1857 was unusually warm in the Alps, but its great heat was not sufficient entirely to remove the snow. No doubt, in colder summers, the snow is retained upon the slopes all the year round. Now this snow becomes the collector of a fine brown mud, which is scattered over the surface of the glacier. It catches the substance transported by the little rills and retains it. The edges of the snow still remaining, when I was on the glacier, were exceedingly black and dirty; and in many cases the entire surface of the snow appeared as if fine peat mould had been strewn over it. Lower down the glacier this snow melts, but it leaves its sediment behind it, and to this sediment the distinctness of the dirt-bands of the Mer de Glace is mainly due. The regularity of the bands depends on the regularity with which the glacier is broken, and the ridges or terraces formed as it passes over the brow of the fall. It is the toning down of these ridges which produces the undulations, which are to some extent modi- fied by the squeezing at the base of the fall; and it is the undulations which produce the bands. Thus the latter connect themselves with the transverse fracture of the glacier as it crosses the brow of the fall. In the figure I have given the general aspect of the bands, but not their number. Thirteen of them exist on the Glacier du Géant. I may add that the bearing I have assigned to this glacier differs from that assigned to it on the map which accompanies Professor Forbes's 'Travels on the Alps,' and which I had with me at the Montanvert. The reason is, that on the map the true north is drawn on the wrong side of the magnetic north, thus making the "Declination" easterly instead of westerly. I have since learned that this error is corrected in the smaller work of Professor Forbes. It has been affirmed that the dirt-bands cross some of the medial moraines of the Mer de Glace, and they are thus drawn upon the map of Professor Forbes. Were this correct, my explanation would be untenable; but the fact is, that the bands are confined to the Glacier du Géant from beginning to end. Royal Institution, February 1859.