constitute one of the most beautiful features of the higher crevasses.

258. How are they produced? Evidently by the thawing of the snow. But why, when once thawed, should the water freeze again to solid spears? You have seen icicles pendent from a house-eave, which have been manifestly produced by the thawing of the snow upon the roof. If we understand these, we shall also understand the vaster stalactites of the Alpine

crevasses.

254. Gathering up such knowledge as we possess, and reflecting upon it patiently, let us found on it, if we can, a theory of icicles.

255. First, then, you are to know that the air of our atmosphere is hardly heated at all by the rays of the sun, whether visible or invisible. The air is highly transparent to all kinds of rays, and it is only the scanty fraction to which it is not transparent that expend their force in warming it.

256. Not so, however, with the snow on which the sunbeams fall. It absorbs the solar heat, and on a sunny day you may see the summits of the high Alps glistening with the water of liquefaction. The air above and around the mountains may at the same time be many degrees below the freezing point in temperature.

257. You have only to pass from sunshine into shade to prove this. A single step suffices to carry you from a place where the thermometer stands high to one

where it stands low; the change being due, not to any difference in the temperature of the air, but simply to the withdrawal of the thermometer from the direct action of the solar rays. Nay, without shifting the thermometer at all, by interposing a suitable screen, which cuts off the sun's rays, the coldness of the air may be demonstrated.

258. Look now to the snow upon your house roof. The sun plays upon it, and melts it; the water trickles to the eave and then drops down. If the eave face the sun the water remains water; but if the eave do not face the sun, the drop, before its quits its parent snow, is already in shadow. Now the shaded space, as we have learnt, may be below the freezing temperature. If so, the drop, instead of falling, congeals, and the rudiment of an icicle is formed. Other drops and driblets succeed, which trickle over the rudiment, congeal upon it in part and thicken it at the root. But a portion of the water reaches the free end of the icicle, hangs from it, and is there congealed before it escapes. The icicle is thus lengthened. In the Alps, where the liquefaction is copious and the cold of the shaded crevasse intense, the icicles, though produced in the same way, naturally grow to a greater size. The drainage of the snow after the sun's power is withdrawn also produces icicles.

259. It is interesting and important that you should be able to explain the formation of an icicle; but it is

important that you should realise the way in ... the various threads of what we call Nature are

together. You cannot fully understand an icicle gst knowing that solar beams powerful enough ave the snows and blister the human skin, nay, it added, powerful enough, when concentrated, up the human body itself, may pass through and still leave it at an icy temperature.

§ 38. The Bergschrund.

Having cleared away this difficulty, let us turn more to the crevasses, taking them in the order formation. First then above the névé we have al Alpine peaks and crests, against which the

often reared as a steep buttress. We have learned that both névés and glaciers are moving downwards; but it usually happens that the ment of the highest portion of the buttress to the is great enough to enable it to hold on while wer portion breaks away. A very characteristic e is thus formed, called in the German-speaking n of the Alps a Bergschrund. It often surrounds k like a fosse, as if to defend it against the asof climbers.

. Look more closely into its formation. Imagine snow as yet unbroken. Its higher portions cling to the rocks, and move downwards with extreme slowness. But its lower portions, whether from their

greater depth and weight, or their less perfect attachment, are compelled to move more quickly. A pull is therefore exerted, tending to separate the lower from the upper snow. For a time this pull is resisted by the cohesion of the névé; but this at length gives way, and a crack is formed exactly across the line in which the pull is exerted. In other words, a crevasse is formed. at right angles to the line of tension.

§ 39. Transverse Crevasses.

262. Both on the névé and on the glacier the origin of the crevasses is the same. Through some cause or other the ice is thrown into a state of strain, and as it cannot stretch it breaks across the line of tension. Take, for example, the ice-fall of the Géant, or of the Talèfre, above which you know the crevasses yawn terribly. Imagine the névé and the glacier entirely peeled away, so as to expose the surface over which they move. From the Col du Géant we should see this surface falling gently to the place now occupied by the brow of the cascade. Here the surface would fall steeply down to the bed of the present Glacier du Géant, where the slope would become gentle once more.

263. Think of the névé moving over such a surface. It descends from the Col till it reaches the brow just referred to. It crosses the brow, and must bend down to keep upon its bed. Realise clearly what must occur. The surface of the névé is evidently thrown into a

state of strain: I breaks and forms a crevasse. Each fresh portion of the neré as it passes the brow is similarly broken, and thus a succession of crevasses is sent down the fall. Between every two chasms is a great mansverse ridge. Through local strains upon the fall those ridges are also frequently broken across, towers of ice—series-being the result. Down the fall both ridges and séracs are borne, the dislocation being angmented during the descent.

Here

264. What must occur at the foot of the fall? the slope suddenly lessens in steepness. It is plain that the crevasses must not only cease to open here, but that they must in whole or in part close up. At the summit of the fall, the bending was such as to make the surface convex; at the bottom of the fall the bending renders the surface concave. In the one case we have strain, in the other pressure. In the one case, therefore, we have the opening, and in the other the closing of crevasses. This reasoning corresponds exactly with the facts of observation.

and stretch it straight. inch or an inch apart, Bend your arm, the dots finally brought together.

265. Lay bare your arm Make two ink dots half an exactly opposite the elbow. approach each other, and are Let the two dots represent the two sides of a crevasse at the bottom of an ice-fall; the bending of the arm resembles the bending of the ice, and the closing up of the dots resembles the closing of the fissures.