sures in different directions, but most commonly tending to the perpendicular and horizontal. By those it is divided into masses of a cubical and prismatic shape. Of the exceptions to this rule there is one among many other instances, in Shaugh rick near Plymouth. If we examine a rock of this kind near the surface of the soil, we shall find that the fissure is a mere mathematical plane, separating the two parts, and that the angles are sharp and perfect. If we turn our attention to granites which from their greater elevation above the present soil appear to have been longer exposed to air and weather, we shall find, as the first step to change, a gentle rounding of the angles, such as is exhibited in the drawing last cited, the Vixen Tor. By degrees the surfaces which were in contact become separated to a certain distance, which goes on to augment indefinitely. As the wearing continues to proceed more rapidly near the parts which are most external, and therefore most exposed, the masses which were originally prismatic acquire an irregular curvilinear boundary, and the stone assumes an appearance resembling the pieces which constitute the Cheese-wring. If the centre of gravity of the mass chances to be high and far removed from the perpendicular of its fulcrum, the stone falls from its elevation, and becomes constantly rounder by the continuance of decomposition, till it assumes one of the various spheroidal figures which the granite bowlders so often exhibit. A different disposition of that centre will cause it to preserve its position for a greater length of time, or in favourable circumstances may produce a logging stone.
It is not necessary to call in the aid of long continued friction or distant transportation to account for the rounded form of these granite bowlders. The changes which they undergo in their places of rest, by their more rapid disintegration at the angles than at the sides, are sufficient to prove that this spheroidal shape may be pro-