Mackintosh — Railiuay Geology in Devon. 399 



action. But from a very particular examination made during stormy 

 weather, I think they ought to be divided into two kinds — pluvial 

 and marine. The first would appear to be formed as follows : — A 

 slight hollow on the upper surface of a rock collects a little rain 

 water which, agitated by wind, and aided by the fall of fresh drops, 

 detaches a little fine detritus. A continuance of the process deepens 

 the hollow until it reaches the lower limit of the agitated water. 

 The small wind- waves enlarge the cavity laterally, forming a minia- 

 ture line of cliffs, which (except by mere accident) are destitute of 

 any curvilinear regularity. In this way shallow flat-bottomed, rough, 

 and steep-sided rain-pools are formed, but they are distinct (except- 

 ing when one kind of hollow has been superimposed on another) 

 from basins with smooth, regular, and perfectly curvilinear sides. 

 The latter occur chiefly on the inclined surfaces or sloping sides of 

 rocks, and are often open on one side, so that they can contain Kttle 

 or no water. These basins are fac-similes of hollows now ground 

 out by waves wielding stones or sand on sea-coasts, or on the 

 projecting rocky islets of archipelagos. The tendency of rain-water 

 is evidently to roughen their bottoms and dilapidate their brims ; 

 in other words, to ruin instead of to form them. On the Dartmoor 

 granitic area these basins have often, in spite of rain, retained a 

 smoothness equal to that of many stone basins chiselled out for 

 domestic use in the neighbourhood. If the marine theory can be 

 disproved, their artificial origin is the only explanation left, for they 

 are generally situated on heights remote from channels of rivers. 



General remarks on Tors. — On any intelligible theory of the ele- 

 vation of the so-called igneous rocks of Dartmoor, the present cannot 

 be the original outline which their surface presented. The hills and 

 ridges are at least partly the effect of the denudation of the inter- 

 vening valleys, and the tors are evidently the remnants of a former 

 lateral extension of the jointed masses of which they are composed. 

 Many of the tors exhibit clear indications of their having once been 

 connected with other tors, and all of them may be regarded as the 

 unscathed monuments of a stupendous denudation by which the 

 neighbouring parts were removed. The denudation may have com- 

 menced in Tertiary if not in earlier times. The rounded forms of the 

 hills, and many of the rocks (see "Boss and Tail"), may have been 

 enhanced, if not caused, by icebergs. But the present shape of most 

 of the tors of Dartmoor can be best explained by the action of 

 waves and currents. They generally occupy the summits of emi- 

 nences where the sea may not only have surrounded them, but 

 washed over them and through them, and where it may be said to 

 have had them at its mercy. They occupy what may be called the 

 waterless centres of watersheds. Rain has not space to acquire an 

 abrading power (except in the case of rain-pools) within their 

 circumference. "What falls innocuously runs off. The grass or 

 heath comes up to the base of many tors, particularly on one side ; 

 and as no gradual shading oif into parts now undergoing decompo- 

 sition is generally apparent, there must have been a sudden cessation 

 of the "sweeping denudation;" it must have finished its work 



