in this way, and the railway-engineers experienced curious difficulties in making their artificial cuttings. Some of these had to be roofed over and converted into tunnels, to prevent the constant slipping in of the sides. In the case of rapid streams, the banks are always changing, and short work will be made of the alluvial flat if artificial checks are not resorted to.

Sections in the alluvial infilling of our valley will give us an insight into the work of the river as an accumulator; and it will be well to find our way down one of the steep zig-zag paths to a little bridge, from which we can view both banks. The pebbles are not thrown down without method or arrangement, as they might be in the products of a landslide which has come to rest. On the contrary, a distinct grouping in beds or layers is visible (Plate III). Here a coarse mass of boulders lies heaped together, the rounded blocks sticking out conspicuously on the steep bank above us, like skulls in a giants' graveyard. But below them the pebbles lie with their longer diameters generally horizontal, as they would if one took a heap of them and smoothed it out on a flat table with the hand; the water in this case has acted as the hand, and has, in swaying them this way and that, finally formed a layer of them, which we technically term a stratum. The plural of this word is strata, and when we say a rock is stratified, we mean that natural agents have deposited its constituents in a series of strata one upon the other.

Beneath the stratum of pebbles comes one of sand, with a few small stones scattered through it. When this was formed, a quantity of finer material, perhaps a muddy and sandy flow from some side-valley, was being brought into the higher reaches of the stream. Or the stream itself became checked in its flow by alterations in its course or in the form of the valley-floor above, and could no longer carry along the larger stones. So it merely brought away the finer sand, with a few little pebbles, and for a time made beds out of these, as if there were no coarse materials up the valley. We shall see later, in observations at the river-mouth, how far this process of natural sifting may be carried.

If we examine the coarser strata in this section of the alluvium, we shall find the sand there also; but it is distributed between and over the surfaces of the stones, which are far more conspicuous. Under the bridge some one has

PLATE III.]

Photographed by Mr. R. WELCH.

confidingly left one of the shallow wooden pails which are used for carrying water or milk, or for washing clothes, throughout this mountain district. If we borrow it, and

throw some of the material from a coarse-grained stratum into it, we can imitate the final separating action of the stream. Let us fill up the pail with water and shake it briskly from side to side. If we pour the contents out quickly, the rush of water will carry with it pebbles and sand and mud alike, and we reproduce the action of the river when it formed the original stratum. But if we deal more gently, we find the top layers of the water becoming thick with the mud and sand, among which, if the sun is shining, we can see a number of tiny flakes of mica gleaming; and now we can pour off this finer material and leave the coarser and more unwieldy stones behind in the bottom of the pail. The particles of sand and mud and fine mica have probably quite as high a specific gravity as the stones that remain behind; but they offer so large a surface in proportion to their mass that they are easily washed about in the water and become poured off with it.

We have now imitated the action of the river when its current is more gentle, and when it begins to sift apart the alluvial materials for redistribution in the lower part of its course. We may even go further, and by careful washing, and by allowing the disturbed material to stand for a short time, we may find that the sand will settle, while the finest mud can be poured off. Then we can, as a second stage, wash away the sand, leaving only the coarser pebbles; and thus we may, like the river in some places, produce a triple separation.

The layer-structure, which we call the stratification, is not very regular in these rough materials, and no one stratum extends for any long distance in the section shown us in the bank. The beds of finer sand lie on one another, sloping in various directions, a certain number being deposited by a current swirling one way, and then others following under the influence of a different movement of the stream. Such a structure, where the beds do not all slope one way, is known as current-bedding. This irregular stratification is further disturbed by the pressure of coarse boulders, which become rolled down in exceptional seasons; and the group of blocks sticking out above us proves how very variable is the action of a mountain-stream.

When a great bank of such detritus has been exposed for a long time to rain and weathering, gullies are cut in it by the storm-rills from above, and these vertical grooves become rapidly deepened. But the larger blocks serve to protect the portions immediately below them from denuding action, and conical pillars result, carved out by rain, some exceptionally large stone forming a cap at the summit of each. These tall cones can be seen on any hillside where banks of old detritus have been subsequently excavated into cliffs; Glencullen, near Dublin, contains fair examples of these earth-pyramids, and the soil at the bottom of any hedge, when examined after a shower, will show the same feature pleasantly in miniature.

The more we remain in the bottom of the present stream-cut, beneath these walls of loose pebbly alluvium, which perhaps rise 100 feet above us, the more we shall believe in the carrying power of the torrent. Its action is so violent that even the hardest rocks are converted into smooth round pebbles by friction one against another; and Professor Bonney has told us that 3000 feet of descent in the upper reaches of an Alpine stream is sufficient to round pebbles of ordinary hard rocks, such as granite. As we noted upon the mountain-slope in our own islands, the smaller materials become more slowly rounded than the coarser, owing to their being buoyed up and kept apart by the flowing water. A stream travelling at four miles an hour at the bottom of its bed can move forward pebbles nine inches in diameter, and two-inch pebbles can be moved by a flow at the bottom of only two miles an hour.

The steep slopes of the stream-beds in the upper region, the region of excavation, greatly assist the rapid denudation of the valley-head. A Bulgarian student, M. Baüff of Shipka,2 has recently given us some interesting results of a year's observation of the Arve, one of the most striking torrents of the Alps. He finds that the solid material brought down by this river in suspension is greatly increased by only a slight increase in the volume of the water, while the matter carried invisibly in solution is much less variable in amount. The level of the river is highest in summer, which is likely to be true of any stream resulting from the

1 "The Rounding of Pebbles by Alpine Rivers," Geol. M 2 "Les Eaux de l'Arve;" Geneva, 1891.

ag., 1888, p. 60.