filtration of foreign matter in solution; and that they were caused by the strain produced during contraction of the brittle glass is rendered evident by such examples as the following, in which may be traced nearly every gradation between long rectilinear fissures or joints and those forming the typical perlitic structure. Fig. 1 (Pl. XX.) represents the arrangement observed in a section of perlitic pitchstone from near Meissen. Several roughly parallel joints divide the glass into small columns; and these are again traversed by cross joints, which thus form minute rectangular blocks. In the central one (a) a spheroid is formed by four curved lines, which clearly branch off from the lateral joints, and round off the corners of the rectangle. In the upper compartment there are three distinct spheroids; and in the lower one there are two of irregular shape. Fig. 2 is from another part of the same slice. In fig. 3 the formation differs in this respect, that in both columns several spheroids are piled one on the other without any intervening cross joints. In a typical perlite from Cabo de Gata in Granada, there are numerous large spheroids, in which many of smaller dimensions are enclosed (see fig. 4). In nearly all the specimens examined the glass is divided into areas of various sizes and forms by a number of parallel or diverging straight lines (joints), the intervening spaces being frequently crowded with spheroids, many of which are flattened against each other (see fig. 5, Pl. XX.).
An examination of all the facts leads to the conclusion that the perlitic texture is purely a phenomenon of contraction; and I quite agree with Mr. Rutley[1] that the explanation of the spheroidal structure in basalt recently laid before the Society by the Rev. T. G. Bonney[2] is a closely parallel case. There is, however, this difference—that, in the case of basalt, the comparatively tough texture produced by the interlacing of its crystalline constituents would enable it to resist the actual fracture so frequently exhibited by the more brittle perlites.
The perlites and other vitreous rocks usually contain numerous minute microliths, which have been described by Zirkel under the names of belonites and trichites; their mode of occurrence and relation to the superinduced spheroidal (perlitic) structure deserve special attention.
The belonites are minute translucent prisms, either colourless or of pale yellow or greenish shades; they occur in immense numbers, and are frequently crowded together in stream-like bands, with their long axes lying in one general direction. Whenever a stream encounters any small crystal of felspar, quartz, or mica imbedded in the mass, the belonites are invariably diverted from their course and bend round it, their axes lying parallel with, its sides. Their relation to the perlitic spheroids, however, is totally different; for instead of winding round them, they continue their course uninterruptedly through and across them (see fig. 8). It becomes
