In the group of strata between the two series of basalts in Antrim, some of the tuffs consist chiefly of rhyolitic detritus, both glassy and lithoid. Where the tuffs become fine-grained and free from imbedded stones, they pass into variously-coloured clays. Among these are the "bauxite" and "lithomarge" of Antrim, probably derived from pale rhyolitic tuffs and conglomerates (p. 204). Associated with these deposits in the same district, is a pisolitic hæmatite, which has been proved to occur over a considerable area on the same horizon. Many of the clays are highly ferruginous. The red streaks that intervene between successive sheets of basalt are of this nature (bole, plinthite, etc.). The source of the iron-oxide is doubtless to be traced to the decomposition of the basic lavas during the volcanic period. (_d_) There occur also grey and black clays and shales, of ordinary sedimentary materials, containing leaves of terrestrial plants (leaf-beds), with occasional wing-cases of beetles, sometimes associated with impure limestones, but more frequently with sandstones and indurated gravels or conglomerates containing pieces of fossil wood. These intercalated bands undoubtedly indicate the action of running water, sometimes even of river-floods, and the accumulation of sediment in hollows of the exposed flows of basalt at intervals during the piling up of the successive lava-sheets that form the plateaux. The alternation of fluviatile gravels with volcanic tuffs, fluviatile conglomerates, and lava-streams, is admirably displayed in the island of Canna, as will be narrated in detail in Chapter xxxviii. The vegetable matter has in some places gathered into lenticular seams of lignite, and even occasionally of black glossy coal. Amber also has been found in the lignite. Where the vegetation has been exposed to the action of intrusive dykes or sheets, it has sometimes passed into the state of graphite. The remarkable terrestrial flora found in the leaf-beds, and in association with the lignites, was first made known by the descriptions of Edward Forbes already referred to, and has subsequently been studied and described by Heer, W. H. Baily, and Mr. Starkie Gardner.[229] It was regarded by Forbes as of Miocene age, and this view has generally been adopted by geologists. Mr. Starkie Gardner, however, contends that it indicates a much wider range of geological time. He believes that a succession of floras may be recognised, the oldest belonging to an early part of the Eocene period. Terrestrial plants, it must be admitted, are not always a reliable test of geological age, and I am not yet satisfied that in this instance they afford evidence of such a chronological sequence as Mr. Gardner claims, though I am convinced that the Tertiary volcanic period was long enough to have allowed of the development of considerable changes in the character of the vegetation.