But one other form needs to be mentioned here as characteristic of the Carboniferous intrusive trap-rocks--that of great amorphous masses which cut through the strata irregularly. They have not the wall-like form of dykes, nor do they conform to the line of bedding of the rocks among which they occur. They are sometimes irregular lumps, lying above or among the strata, and probably connected with some vein or dyke below. In other localities they look like the upper ends of vast pillars which may descend into the very depths of Tartarus, as though a great hole had been blown through the crust of the earth, and a column of melted matter had risen to fill the cavity. Such masses are often called _bosses_, and seem not unfrequently to have been the craters of eruption whence great sheets of lava and showers of ashes were ejected far and wide over the neighbourhood. They serve to connect the intrusive traps, whose age is always more or less uncertain, with the bedded traps properly so called, the geological date of which can usually be sufficiently ascertained. II. The bedded or contemporaneous trap-rocks consist of two well-marked kinds. There are, 1st, the melted rocks, such as greenstones and basalts and 2d, the tufas and volcanic ashes. Those of the first-named class differ in no respect from the traps already noticed, so far as regards mineralogical texture, general structure, and appearance. In hand specimens the intrusive and bedded greenstones and basalts cannot be distinguished, nor even when examined in the field and in masses extending over considerable areas is it always possible to say to which division any particular hill or crag should be assigned. The reason of this resemblance is obvious. Where a trap has either cut through or insinuated itself among rocks of earlier date it is called intrusive, in relation to the rocks so traversed, and of course we cannot be sure to what geological period it should be referred, nor how long an interval may have elapsed between the time when these rocks were forming and the time when the trap was intruded among them. If, however, the igneous rock passed upward through these same strata and then spread out as a flat sheet along the sea-bottom, the part that came to the top would be termed contemporaneous with the deposits going on at the time. Hence it follows that all contemporaneous lava-form trap-rocks are at the same time intrusive as regards the strata passed through in their progress to the surface. If the sheet of melted matter that spread out below the water were in the course of ages worn completely away, along with the strata subsequently piled above it, so as to leave merely a neck or dyke filling up the cavity through which the lava rose, we should pronounce the remnant intrusive, and could form no certain conclusion as to its age or as to whether its site had ever been a crater actively at work in throwing out lava and ashes. The sole difference, therefore, between a