[Footnote 165: See Mr. J. J. H. Teall, _Quart. Journ. Geol. Soc._ xl. (1884), p. 214.] In general, each dyke is composed of one kind of rock, and retains its chemical and mineralogical characters with singular persistence. The difference of texture between the fine-grained chilled margin, with its occasional glassy coating, and the more coarsely crystalline centre is due to cooling and crystalline segregation in what was no doubt originally one tolerably uniform molten mass. The glassy central bands, too, though they indicate a rupture of the dyke up the middle, may at the same time quite conceivably be, as I have said, extrusions from a lower portion of the dyke before the final solidification of the whole. The ramifying veins of finer grain that now and then traverse one of the large dykes are likewise explicable as parts of a stage towards entire consolidation. All these vitreous portions, whether still remaining as glass or having undergone devitrification, are more acid than the surrounding crystalline parts of the rock. They represent the siliceous "mother-liquor," so to speak, which was left after the separation from it of the crystallized minerals, and which, perhaps, entangled here and there in vesicles of the slowly cooling and consolidating rock, was ready to be forced up into cracks of the overlying mass during any renewal of terrestrial disturbance. But examples occur where a dyke, instead of consisting of one rock, is made up of two or more bands of rock which, even if they resemble each other closely, can be shown to be the results of separate eruptions. These, which are obviously not exceptions to the general rule of the homogeneity of dykes, I will consider in the next Chapter. Among the petrographical varieties observable in the field is the occasional envelopment of portions of the surrounding rocks in the body of a dyke. Angular fragments torn off from the fissure-walls have been carried upwards in the ascending lava, and now appear more or less metamorphosed, the amount of alteration seeming to depend chiefly upon the susceptibility of the enclosed rock to change from the effects of heat. Cases of such entanglement, however, are of less common occurrence than those already referred to, where pieces of some deep-seated rock, such as the gabbros of Skye, have been carried up in the ascending magma. Occasionally, where the enclosed fragments are oblong, they are arranged with their longer axes parallel to the walls of the dyke, showing flow-structure on a large scale. Mr. Clough has found some dykes near Dunoon which enclose fragments of schist nearly three feet in length.