[Footnote 159: _Trans. Roy. Soc. Edin._ vol. xxii. (1861), p. 650.] 1. The prevalent tendency of the dykes to take a north-westerly course. There are exceptions to this normal trend, especially where the dykes are small and locally numerous; but it remains singularly characteristic over the whole region. 2. The increasing abundance of the dykes as they are traced to the west coast and the line of the great Tertiary volcanic plateaux of Antrim and the Inner Hebrides. 3. The rectilinear direction so characteristic of them and so different from the tortuous course of local groups of dykes. The exceptions to this normal feature are as a rule confined to the same localities where departures from the prevalent westerly trend occur. [Illustration: Fig. 233.--Dyke on the south-east coast of the Island of Mull.] 4. The great breadth of the larger dykes of the system and their persistence for long distances. This is one of their most remarkable and distinctive characters. 5. The posteriority of the dykes to the rest of the geological structure of the regions which they traverse. They are not only younger than the other rocks, but younger than nearly all the folds and faults by which the rocks are affected. 6. The manner in which they cut the Jurassic, Cretaceous and older Tertiary rocks in the districts through which they run. At the south-eastern end of the region they rise through the Lias and Oolite formations, in the west they intersect the Chalk and also the Tertiary volcanic plateaux together with their later eruptive bosses. 7. Their petrographical characters, among which perhaps the most distinctive is the frequent appearance of the original glass of the plagioclase-pyroxene-magnetite (olivine) rock, of which they mostly consist. This glass, or its more or less completely devitrified representative, often still recognizable with the microscope among the individualized microlites and crystals throughout the body of a dyke, is also not infrequent as a black vitreous varnish-like coating on the outer walls, and occasionally appears in strings and veins even in the centre. It is the assemblage of dykes presenting these features which I propose to describe. Obviously, the age of each particular dyke can only be fixed relatively for itself. But when this remarkable community of characters is considered, and when the post-Mesozoic age of at least a very large number of the dykes can be demonstrated, the inference is reasonable that one great system of dykes was extravasated during a time of marked volcanic disturbance, which could not have been earlier than the beginning of the Tertiary period. And this inference may be maintained even when we frankly admit that every dyke within the region is by no means claimed as belonging to the Tertiary series. [Illustration: Fig. 234.--Fissure left by the weathering out of a dyke.]