of the mass, the rock appears to present the same general characters throughout, specimens from the northern end being quite similar to some of those collected at the opposite extremity. It is for the most part a hard compact rock of dark red or brown colour, and is the compact felspar and hornstone of Murchison, who also refers to one variety as a porphyritic clinkstone.
In order to show clearly the true character and structure of these rocks, it will be necessary, in the first place, to give a short account of their recent analogues; as rocks of this peculiar type are by no means of wide distribution, and some of them have not been previously observed in these islands.
II. Structure of Modern Perlites.
Perlite, spherulitic perlite, perlitic pitchstone, and perlitic obsidian belong, as is well known, to the glassy group of acid rocks, their average amount of silica being at least 70 per cent. In typical unaltered specimens the mass consists of a true glass which has no action on polarized light.
The spheroidal and ellipsoidal balls by which perlite is chiefly characterized have been described by Zirkel[1], Rosenbusch[2], and Lassaulx[3] as consisting of concentric laminæ arranged like the coats of an onion—a comparison which may, I think, possibly lead to erroneous ideas as to their real character; for however close may be the resemblance, as seen in thin sections, there is no real analogy between the structure of these perlitic spheroids and that of a tunicated bulb built up of broad scales which surround each other in a concentric manner.
An examination of typical specimens from the old volcanic districts of Schemnitz in Hungary, Meissen in Saxony, and Cabo de Gata on the south coast of Spain shows that the spheroids have not been formed by the superposition of successive laminæ: they are not concretions in any sense of the term; nor is there any thing about them suggestive of any process of progressive construction. When a rather thick section is examined under a low power of the microscope, it is at once seen that the little spheroidal balls are merely portions of the homogeneous glass which have been partially separated from the general mass by the formation of a number of small curved planes of fracture; these are more or less concentric with each other, but vary greatly in size, and are irregularly disposed in various directions round the centre. As these curved planes lie at various depths in the section, some of them appear with a convex or concave surface, according as the slice happens to cut through the upper or lower half of a spheroid. Such being the general arrangement, the lines seldom form closed curves when seen in thin sections. That these lines and curved planes are really fine cracks is clearly shown by the way in which they are frequently filled more or less completely by the in-
