As one sits at home, one thinks of the scheme of mountain valleys too simply. One thinks of the stream as coming down through a ravine with its head waters appearing below a definite saddle or notch in the watershed. This stream, let us say, is flowing north. One sees on the further side another stream rising just on the other side of the notch and flowing on through a simple valley, going to the south. The crossing of the port between these valleys seems to depend upon no more than physical endurance and fine weather. One goes up one stream to the saddle, crosses the saddle, follows the other stream downhill, and so makes one’s passage from France to Spain.
There are many passes of this simplicity, but there are many more that, both between the lateral valleys and over the main range, present the danger of which I speak, and which consists in a complexity at a summit such that it is difficult in the extreme to know—even when one is certain one has gone up the right part of the hither slope—what one should do on the thither.
This danger of “getting into the wrong valley” cannot be seized without illustration, and in the following rough sketches I give examples of this.
In the first example is a bit of country such as one very often gets in these mountains with summits round about the 2600 metre line and the last valleys under the ports somewhat above the 2000. I have marked with hatching the contours below 2200 and in black the summits above 2600. The main watershed I have indicated by a dotted line.
When one is crossing a port of this type one sees before one from the summit a confused and gentle slope leading apparently to one obvious valley on the far side like the obvious valley out of which one has just ascended. It seems indifferent whether one should come down on to this by M or by N, to the left or to the right, yet the two valley floors to which each leads are quite separate and may lead one round to different river basins. How deceptive such a place is, the rough sketch appended may help the reader to grasp. It shows the kind of thing one sees from the summit of such a pass and how indifferent the choice appears between the ways by which one may descend.
This type of confusion exists sometimes in a still more dangerous form, as in the contour lines of sketch on next page.
A man arrived at the port P climbing up from the valley Q, which is deep and well defined, sees before him another valley R exactly in line with the last, also deep and also well defined. On either side of him, as he gets to the saddle, run high ridges perpendicular to the line of the two valleys. It seems common sense to take the watershed as running along these ridges and across the port, and if Q is the French valley, R will be the Spanish one. As a matter of fact the watershed may not run in this simple way at all, but (as indicated upon the sketch map) take a sharp turn to the right. R may be a French valley after all, and the proper way down into Spain may be over the gradual grassy slopes indicated by the arrow line. A man standing just at the port, and having a rocky ridge A and the rocky ridge B to his left and right, sees before him the obvious trench of the valley R and takes for granted that it is the Spanish valley, whereas his true way is across the vague grassy land towards S, and the watershed which he thinks runs from B on to A really turns round from B and runs on to the distant mountains before him.
