[359] Mineralogical Travels, ii. 90.
[360] Western Isles, see the descriptions of Skye, Mull and Rum.
[361] Berger, Trans. Geol Soc. iii. (1816), p. 190; Portlock, Journ. Geol. Soc. Ireland, vol. i. (1834), p. 9.
[362] Zeitsch. Deutsch. Geol. Gesellsch. xxiii. (1871), pp. 54, 77, 84, 88.
[363] Tschermak's Min. und Petrog. Mittheilungen, 1878, p. 412. The chemical composition of this rock and its place among the rhyolites had already been determined by E. T. Hardman from analysis, Journ. Geol. Soc. Ireland, vol. iii. (1871), p. 32.
This interesting series of rocks embraces a greater variety of petrographical characters than any other portion of the British Tertiary volcanic rocks. On the one hand, it presents thoroughly vitreous masses, some of which in their colour, lustre and microscopic structure remind us of recent obsidians. On the other hand, it affords coarsely crystalline compounds, to which no other name than granite can be assigned, and which, did we not know their geological position, might almost be classed with some of the most ancient eruptive rocks. Between these two extremes abundant gradations may be found, including beautiful spherulitic rocks, felsites and rhyolites.
In dealing with such a series of intrusive rocks, we again encounter the difficulty of reaching certainty as to their relative dates of eruption, since in each case all that can usually be affirmed is that the intrusive mass is younger than that into which it is injected. It is quite possible that protrusions of acid rocks occurred at intervals during the accumulation of the basic masses, as may perhaps be inferred from the rhyolite-tuffs and conglomerates of Antrim and from the occurrence of fragments of siliceous lavas in the gravels near the base of the basalt-plateau of Mull, and in the agglomerates of that island as well as of other districts.[364] It is probable, therefore, that at the time when the basalts of the plateaux were emitted, there existed, within reach of volcanic explosions, masses of granophyric, felsitic or rhyolitic rocks, fragments from which were shot up the funnels of discharge. That portions of these rocks were actually intruded into the basalt-sheets before the building up of the plateaux was completed appears to be proved in Antrim. Elsewhere, however, no evidence has yet been obtained of any such intrusion until after the close of the plateau-period. On the contrary, in every case where the relative ages of the rocks can be fixed, the acid are younger than the basic protrusions.
[364] Reference may also again be made to the agglomerates of Strath, Skye, which contain in some parts abundant fragments of acid rocks that closely resemble some of the masses of granophyre which disrupt these agglomerates.
The only known exceptions to this rule are the latest basalt-dykes. Hence, while amid the large and varied series of acid rocks, which no doubt represents a wide interval of time, some may belong to comparatively early epochs in the protracted volcanic period, the actual available evidence places the emission of these rocks, as a whole, towards the end of the volcanic history. This evidence I shall bring, forward in full detail, since it necessitates an abandonment of what has been the general belief in regard to the relative ages of the rocks.