by their accumulation become botryoidal, smaller spheres growing out as it were from the larger ones. Detached specimens that have been washed by the sea, exhibit the appearance of a bunch of grapes, but in their native bed the intervals are filled with an ochry clay. Their fracture shows the uniform aspect of siliceous schist, being neither coated nor radiated; and the rock agrees in every respect, except its structure, with the bed of siliceous schist, which is found not far off.
I have only observed one other instance of a rock similar to this, and it occurs also under amygdaloidal trap at Talisker, in Sky.[1] I have reason to suppose it of rare occurrence. In various parts of the rocks which I have now described there are found natural joints, which on being separated exhibit two smooth surfaces, absolutely flat, and appearing as if they had been cut through by a sharp tool. The globular bodies themselves are divided in various parts, just as they happened to interfere with the section. In the greater number of the sections which I examined, no substance was interposed between the touching faces. But in some I found wavellite, a mineral as yet not so common, but that this new habitat will be acceptable to the mineralogist. It is remarkable that the circles of wavellite occupy precisely the surface of each segment of the several spheres, varying in size according to the varying dimensions of these segments. This bed is covered by one of trap, or to use the term in its present great latitude, of greenstone, forming a parallel bed of a more considerable thickness, but with the same inclination as the subjacent strata. This bed of greenstone weathers so as to exhibit distinct globular concretions, although in its entire state the eye can
- ↑ I have since observed a similar modification in the schist of Devonshire in which wavellite is found.
