Once we have located what we take to be their line of junction with the parent snow crest, we must allow a further broad margin in selecting a safe line for our steps. I have known a section of cornice thirty feet deep break away along the line of our steps, when we thought we had allowed a good ten-foot margin from its line of junction. The heavier the cornice, the deeper the margin we must leave; not an easy rule to keep, as the snow curve backing a big cornice will be all the more gradual and solid-looking. Leaders are always apt to cut the margin too fine, for the reason that the farther we move away down the back of a cornice, the steeper becomes the slope, and the harder probably will be the snow for step-making. At great heights the same cold winds which blew the cornice may have blown the snow slope to crusted snow or snow-ice; whereas, where the angle eases off right upon the crest and the back curves over into the cornice, the direct exposure to the sun will often produce a band of softer snow, easy to kick steps in. When we are dealing with huge ridges and long distances, it is very trying to have to continue to cut steps at a snail’s pace, on a steep wall, when three feet above there is a line where we could walk. But these are just the situations which differentiate between good and bad mountaineers. Danger is a matter for our individual judgment, and if a guide, however famous, makes steps at a level that seems to our mountain sense to allow too narrow a margin, we need never have the least hesitation in starting to make another line below. The implied criticism is better than remonstrance.

Occasionally when the cornice is not too big and its back is ice-crusted or exposed to a high wind, it is possible to work along underneath it, below the crest of the wave. This must be done with caution; the shock of step-making may easily loosen a section of the overhang on to our heads.

Where a ridge is surmounted by a double cornice, that is, by two cornices curving opposite ways, one on the top of the other, there is simply no safe way of getting past them. Parties who persist in traversing a great ridge when they find it in this condition must be persuaded that their luck will take pity on their understanding.

In approaching the edge of any ridge, if it is not otherwise obvious that no cornice exists, the leader must test for the cornice every few inches with strong, deep thrusts of the axe shaft. The rest of the party, protecting him with the rope, must remain well below on the face until he has demonstrated that no cornice exists, or until he has had time to locate its line and to return to a point well below the junction. Such a crest should never be approached diagonally, lest two men might find themselves on the cornice together, but always at right angles. If an uncorniced ridge gives us reason to suspect, as we follow along its crest, that a cornice is beginning, at least two of the party should descend well down the face (if rock anchorage is not available), and protect the leader while he investigates.

If we wish to cross a ridge or a pass and the far side is overhung by a cornice, the leader must be strongly protected while he flogs away a section of the cornice, back to the sound edge. A large slice in this case should be cleared away, as adjacent parts of the cornice weakened by the fracture may fall upon the party subsequently as they descend. These crossings should never be made where the cornice is really heavy.

We shall never, of course, approach a crest up a wall crowned by a large cornice; which is equivalent to climbing deliberately up under a toppling sérac. A light cornice we may approach, but preferably at a point where it has broken away. If no such gap exists, we may be forced to cut our own skylight through it. This calls for great caution and skill; a small fragment may easily sweep away any man below us on the slope.

Snow in Couloirs.

In couloirs and chimneys snow can remain at a higher angle than on open slopes, and presents therefore some exceptional features. Certain of these are dealt with under other headings. Snow in a rock couloir, late in the day, must always be tested before we descend on to it. This can sometimes be done with less trouble by throwing heavy stones on to the surface. The way the snow spurts or squelches or rives to receive them tells much to the practised eye. If they start a surface avalanche they may either betray ice below, or they may leave bare a tract of hard under-snow down which we can kick good steps. The presence of ice below snow is often evident at the edges where the snow-ice splays out on to the containing walls. The testing process should be repeated at intervals, as we descend, on doubtful snow. A long couloir, of pure snow above, has frequently ice below snow in its lower sections.

In descending couloirs of uncertain snow, we naturally make use of the rock walls, when possible, for belays or holds. There are often good leg stances between the rock and the snow, where the snow veil has shrunk away, with an ice-trimmed border.

If the snow is of good quality, we proceed as we like; but if it is shallow, lying on ice, or of suspicious consistency, we ascend it in as direct a line as possible, or in steep zags, with a disposition to keep near the sides. On really threatening snow we ascend and descend by means of the vertical ladder of apses or pigeon-holes, climbing face inwards, and using both hands and feet. Our tread in this fashion is more feline, and our weight better distributed.