157. Look through the telescope; you see it crossed by two fibres of the finest spider's thread. In actual work we first direct the telescope across the glacier, until the intersection of the two fibres accurately covers some well-defined point of rock or tree at the other side of the valley. This, our fixed standard, we sketch with its surroundings in a note-book, so as to be able immediately to recognise it on our return to this place. Imagine a straight line drawn from the centre of the telescope to this point, and that this line is permitted to drop straight down upon the glacier, every point of it falling as a stone would fall; along such a line we have now to fix a series of stakes. 158. A trained assistant is already upon the glacier. He erects his staff and stands behind it; the telescope is lowered without swerving to the right or to the left; in mathematical language it remains _in the same vertical plane_. The crossed fibres of the telescope probably strike the ice a little away from the staff of the assistant; by a wave of the arm he moves right or left; he may move too much, so we wave him back again. After a trial or two he knows whether he is near the proper point, and if so makes his motions small. He soon exactly strikes the point covered by the intersection of the fibres. A signal is made which tells him that he is right; he pierces the ice with an auger and drives in a stake. He then goes forward, and in precisely the same manner takes up another point. After one or two stakes have been driven in, the assistant is able to take up the other points very rapidly. Any requisite number of stakes may thus be fixed in a straight line across the glacier. 159. Next morning we measure the motion of all the stakes. The theodolite is mounted in its former position and carefully levelled. The telescope is directed first upon the standard point at the opposite side of the valley, being moved by a tangent screw until the intersection of the spider's threads accurately covers the point. The telescope is then lowered to the first stake, beside which our trained assistant is already standing. He is provided with a staff with feet and inches marked on it. A glance shows us the stake has moved down. By our signals the assistant recovers the point from which we started yesterday, and then determines the distance from this point to the stake. It is, say, 6 inches; through this distance, therefore, the stake has moved. 160. We are careful to note the hour and minute at which each stake is driven in, and the hour and the minute when its distance from its first position is measured; this enables us to calculate the accurate _daily motion_ of the point in question. The distances through which all the other points have moved are determined in precisely the same way.