It was generally necessary to remain for at least a week on the mountain summits which formed the main triangulation stations, for only on about one day in seven was the air clear enough for sighting the beacons on the longer lines. Occasionally the entire landscape was blotted out for ten or more consecutive days by clouds surrounding the summit, while at other times it was possible to see only for a limited distance round the station owing to haze. Such times were made use of to map all visible detail within a moderate range (say within a radius of twelve kilometres) round the station. In this work the first stage was to find a small base, one end of which was the station itself. Usually a minor peak of the same range, 500 to 1,000 metres away from the station, was fairly easily accessible, and was chosen for the other end of the base. The six-inch theodolite being at the main station, the five-inch tacheometer was set up at the auxiliary station, and all noteworthy hill tops, as well as a few points along each main line of drainage, were triangulated off this small base. The length of the base was found by including one or two main triangulation points in the round of angles. These minor triangles were conveniently reduced by the slide rule, and the points plotted at once on the plane-table by means of the alidade and the calculated distances. The base being short, it was necessary to observe to fine marks; cracks in the rocks, and the droppings of birds on the peaks, and the centres of selected tree trunks in the wadis, were usually chosen. The levels of these minor points were determined by vertical angulation in the ordinary way. Usually about thirty points were thus fixed round each high station. Once a number of points were fixed in the wadis, the levels of these gave the slope, and the difference of height between any other parts of the wadis and the station could be estimated to within a few metres by means of the knowledge thus obtained. A sketch being now made of the wadis, which appeared spread out almost like a map below the station, a hundred or more points along them were selected, and their depths below the station being very approximately known from the wadi slope, their distances were found by observing depression angles to them and reducing the vertical triangle by means of the slide rule on the spot. In all, therefore, measurements were usually made of the distances of from 100 to 200 conspicuous points in the area round the station, and when these were plotted with the alidade on the plane-table sheet it was not difficult to sketch in all the detail with considerable accuracy. Usually it was not possible to see all round the mountain from the station itself, so that subsidiary plane-table stations near the main one were necessary. In other cases more than one small base was measured in order to get good angles to various points by the minor triangulation.