of perched blocks and roches moutonnées are the dark hollows of Loch Coruisk in Skye, and the more accessible region south of Killarney, along the road from Kenmare to Glengariff. Let us travel thither, and to Switzerland also, to continue our studies when we can. Meanwhile, any simple mountain-hollow will have taught us much in an afternoon.

And now, as we descend in the twilight, the clouds are already gathering on the crests behind us. We come down among the great loose blocks again, critically inquiring which of them have merely fallen from above and which have been stranded by ice-action. Some, like the grey flinty boulders so common on the hills around Llangollen, are obviously strangers, and must have been ice-borne from a considerable distance. When we at last strike the highway and tramp homeward through the dark, we can still hear the far sound of waters falling across the blackness of the rocks. And we know now that work is going on up yonder, and that next morning the sun will send his rays into the hollow, already shining upon new things and spying out the changes of the world.

CHAPTER III

DOWN THE VALLEY

THE course of any river is full of variety, as we follow it down from its source in the hills, through the long windings of the valley, across the open plains, and away to the distant sea. On every side there are things to interest us, and the river itself is the moving spirit of it all. Our great cities would not be where they are were it not for the rivers, which gave water to the early inhabitants before the days of deep wells and costly aqueducts. In many cases the town was established where ships, the light sailingvessels of former days, could come up and land their merchandise direct upon the quays-just as great steamers still do below London Bridge, or at the Broomielaw in Glasgow, or along the Custom House quay in Dublin. In other cases the river divided hostile tribes, and the town grew up at some ford, where it was necessary to plant a fortress and to keep a constant watch. Whether we travel down the Thames, the Loire, the Danube, or the Rhine, we see how the flow of the water and the form of the riverchannel have had their influence upon history. And then, besides all this, long before man came into the country, the river was working out an elaborate history of its own.

Many streams have simple beginnings, just as we saw among the barren highland rocks. The rain falls, and leaps in tiny cataracts down every crack and groove; on some shelf the water is stopped, and gathers as a little lake, or forms with the mosses a soft mountain-bog; from this a true streamlet rises, finding its way down among the grass-slopes and the fallen stones.

In some places, however, this water sinks in and disappears. Rain falls on sandstone or common limestone, where it becomes soaked up, and may sink to considerable depths. Such rocks, through which water can flow, are called permeable rocks; the water finds its way between the

mineral particles, or even, in the case of limestone, dissolves the material and works out little channels for itself. In our summer, the air is so warm and the days are so long that the rain which occasionally falls and sinks in becomes dried out again in the form of invisible vapour, and thus does not help to form an underground supply. But the winter rainfall, on the other hand, produces a quantity of water which flows slowly through the permeable rocks, and this is the supply that is often reached by the wells in old villages and farms. The ground becomes, in fact, waterlogged, because the water cannot sink indefinitely towards the centre of the earth. It is soon stopped by some impermeable layer of rock, such as clay or granite, and continues to gather in the mass above. If the layers of rock have a slope in any particular direction, the water will flow

Sea

B

L

FIG. 3.--SECTION ILLUSTRATING THE ORIGIN OF SPRINGS.
A, inland heights; L, limestone; c, clay. Springs will emerge at B.

down this as long as there is some means of escape at the other end.

Imagine (fig. 3) a great layer of limestone resting on a layer of clay, and both tilted up so that the higher end forms the crest of a range of hills; and let the lower end come down to the sea-side, forming a limestone cliff, with the clay just showing on the shore beneath. Rain will fall heavily upon the heights, and will, in the colder months, accumulate as a great body of water in the limestone. The tilt of the rocks will cause this water to flow towards the sea, and it will produce a number of little streams gushing out over the clay surface on the shore. Here we have a line of springs formed, which will flow steadily so long as a good rainfall—especially a good winter rainfall-is kept up on the higher ground inland.

Springs arise where underground waters find a con

S

venient outlet. When the gathering-ground of the water
is high, we may sink a well upon some much lower
portion of the permeable rock, and thus form an artificial
spring. A "head" of water occurs in the distant mass at a
higher level, which presses the water upwards in the outlet
provided by the well, and forces it to flow out upon the
surface.

Sometimes we may find a permeable rock lying between
two impermeable masses, and the whole three layers may be
bent into a basin-shaped form (fig. 4). Thus on all sides the
permeable rock is exposed at the surface by this upward
bending, and receives rain, which sinks into it. But the
water runs down and accumulates in the centre between the
two impermeable layers, thoroughly waterlogging the mass.

FIG. 4.-SECTION SHOWING PRINCIPLE OF ARTESIAN SPRINGS AND WELLS,
AND ORIGIN OF BOURNES.

B, pervious stratum; A and c, impervious strata; s, sea-level. Bournes
may arise at x and x', and an Artesian well may be made at w.

If we now bore down through the upper impermeable layer,
anywhere near the centre of the basin, the water will rise
freely; and, if the sides of the basin have been bent up high
enough, it will flow out copiously as a fountain on the sur-
face. Wells made under these conditions are called Artesian
wells, from the fact that successful ones were constructed in
the French province of Artois as early as 1126 A.D.

It will be clear that any permeable layer of rock will be.
drier near its upper surface, and that the water stored up in
it will rise to a certain level in the mass, this level being
nearer the surface at some times than at others. Any natural
excavation or hollow which reaches down to the lowest of
these variable levels will be occupied by a permanent spring.
Many other hollows will be full of water in certain seasons
only-usually after a succession of wet winters—and may

S