rocks. The pudding-stone of Craig Phadric differs completely from that of Lorn. It contains no fragments of the greenstone amygdaloid, there being no greenstone beds in its vicinity, as there are on the Oban coast. The pebbles which it does contain are of quartz, gneiss, granite, and the other associated rocks. The paste which cements them is of a granular texture, entirely and essentially different from that of the Lorn pudding-stone, and belonging to a very different class of substances. It is agglutinated by adhesion, as the sandstones are, without a common binding paste; and consists of fragments of the same rocks which form the nodules, exhibiting generally a gritty mixture of hornblende, mica, felspar and quartz, with a considerable portion of ferruginous felspar clay. The difference in the vitrification of the wall arising from this cause is obvious, since the scoria of Craig Phadric contain none of that very light and spungy sort capable of floating in water, and which I have shown to arise from the fusion of the calcareous amygdaloid. It differs also in these respects, that it contains no pyritical slate, and that it contains fragments of sandstone. The heat has operated on these stones so as to roast and crack the quartz, granite, gneiss, mica-slate, common slate, and sandstone, producing appearances similar to those of the specimens in Dun Mac Sniochain. The gneiss only which contains much hornblende, and passes into hornblende-slate, is partially fused. The mica-slate, containing also in some instances layers of hornblende, has been split in sunder by the vitrification of these laminæ, and in other cases it is bent and contorted in a very amusing and instructive manner. But the cementation of the wall is produced by the vitrification of the paste which forms the pudding-stone. By this, not only its own pebbles are united, but the neighbouring stones have been entangled in the general mass.
It is plain that no additional argument to support the notion of