With a view to ascertaining the oscillations in length of a glacier, the traveller or surveyor should, if possible, mark on the ground the position of the end of the glacier at the time of his visit, so that the next visitor will be able to measure the movement that has taken place since his predecessor’s visit. Leaving out of question elaborate trigonometrical methods, such, for instance, as have been carried out on the Rhone Glacier in Switzerland, the following means may be adopted for recording the position of the glacier front with considerable accuracy. Paint some signs on large boulders, not too far from the end of the glacier, and measure their distance from it by a tape (Richter’s system), or build a low wall of stones of a few yards in length, and, say 15 to 20 inches in height, some distance from the ice-end, and measure this distance (Gosset’s system). It is to be recommended that the stones of these walls should also be painted. In either case the date and distance should be painted on the stones. If the traveller himself returns after some interval—even after only two or three weeks—he will be able to judge of the movement of the glacier, and he will have laid down a basis for further observations by future travellers. A plane-table sketch on a fairly large scale with contours or form lines of the ice front and its immediate vicinity is of the greatest value, as it records not only the position but also the form of the glacier. The sketch should, of course, show the position of the painted boulders or other fixed points near the ice. To ascertain the recent retreat of a glacier, measure the distance from the ice front to the most advanced terminal moraine, where vegetation first shows itself. The bare ground recently left by glaciers is easily recognisable. The diminution of volume is best measured by ascertaining the height of bare soil left on the sides of the lateral moraines in the portion of the glacier within the zone of vegetation. All photographic representations of the glacier end, and of the ground which has been freed from the glacier ice, are of great value. Those will be of most service that show the position of the glacier-snout with relation to some conspicuous rock or other feature in the local scenery. Each photograph should be dated, and the bearings and distance of the camera with reference to any such feature accurately noted. Neighbouring glaciers often furnish very different results as regards advance or retreat, owing to the fact that steep glaciers anticipate in their oscillations those the beds of which are less inclined. One of the results most to be desired is an exact knowledge of the dates: I. Of the maximum extension of the ice. II. Of the commencement of retreat. III. Of the minimum. IV. Of the commencement of fresh increase. and even information giving approximate dates—say, to within a decade, may prove to be of value.