An interesting case of surface (but one which is rarely met with in this country) common in dry countries where the rare rainfall is sudden and intense, and where temporary water courses carve out the friable soil, is the inconvenience due to what are called in some parts of the East “nullahs”--that is, the dry beds of such water courses or the sudden depressions made by what were formerly water courses now dried up through a change of climate. The banks of these are often so steep and their depth so considerable that the making of a plain, straight trajectory across such a country would, even under modern conditions, not be worth the labour expended. It would mean continual bridging, or continual embankment. One of the effects of this type of surface is the inordinate winding of all the roads, and even, alternatively, the absence of roads perpendicular to the fall of the land, and the establishment of communications along the line of fall rather than across it. One can see this very conspicuously in Morocco, where there are whole districts, a couple of days’ march across, the trails of which are determined by this accident. A special example of the same kind of thing is to be found in any hill range where a number of narrow spurs project towards the plain. The Road hardly ever runs parallel to the range across these spurs. It nearly always runs down the valleys or along the plain at their foot, and that although there be, as there usually are, in each valley centres of population which need to be linked up with the neighbouring parallel valleys. ii GRADIENTS. The obstacle of gradient the “minimum of vertical effort” is the most evident of all the factors which modify the trajectory of a road; yet it is, upon the whole, the most complex. To determine the minimum of effort you have to find a formula consisting of many factors, some of which I have already enumerated in the opening words of this essay. In the first place, you have to consider the _average_ nature of the travel to be served. The Road used by men on foot without burdens, by men on foot with burdens, by pack animals, by wheeled vehicles, etc., must conform itself, on the whole, to the _least_ gradient useful to those who travel by it, but that “on the whole” least gradient is a factor by no means easy to determine. It depends not only upon the nature of the instruments of travel, but upon habit, upon vigour, and to some extent upon surface. It depends also on the proportionate use of the Road. You cannot sacrifice ninety-nine travellers to the special weakness of one.