forms but one great deposit, undergo analogous changes when it approximates to this rock. The transition is in fact more easy and the boundary less definable, since many varieties of quartz rock, formed of felspar and quartz in varying proportions with an occasional mixture of mica, only require to assume that crystallized appearance which the vicinity of granite is so apt to produce on the neighbouring rocks, to become indistinguishable from it. The quartz rock in these specimens contains mica, disposed at first in a parallel form, so that it might equally be ranked among the varieties of micaceous schistus or of gneiss. By degrees the mica loses its parallel disposition, and at length the whole assumes the aspect of granite. This transition therefore adds one more to those passages into mica slate, clay slate, and graywacké, which I formerly described as occurring in quartz rock. Loose specimens are found both at Blair and in Ben Gloe containing drusy cavities, of which the siliceous crystals, though more minute than a pin's head, are perfectly defined. These crystals are evidently of posterior formation to the general mass of the rock, and have resulted from the infiltration of a watery solution of silica into previously formed cavities. They prove nothing therefore relating to the chemical nature of this rock, of which, mixed with its mechanical formation, there are every where to be found abundant examples, some of which I have formerly enumerated.
In Glen Tilt I have taken out specimens from the beds exactly similar to those long compressed and smooth cylindrical bodies which are sometimes found in secondary sandstones. This shows another distinct point of agreement between quartz rock and those sandstones. In the same place is found a very interesting variety. It consists of a regular and repeated alternation of quartz with common argillaceous schist; the smoothest variety of clay slate. The quartz at
