In all these volcanoes we see the tendency towards the eruption of intermediate lavas in the first instance, and of basaltic and acid lavas at a later date. Valuable, however, as are the early generalisations of Scrope, and the more precise law enunciated by Von Richthofen concerning the 'natural order of succession of volcanic products,' we must not forget that there are to be found a considerable number of exceptions to them. There are some volcanic centres from which only one kind of lava has been emitted, and this may be either acid, basic, or intermediate in composition; and on the other hand, there are districts in which various kinds of lava have been ejected from the same vents within a short period of time, in such a way as to defy every attempt to make out anything like a law as to the order of their appearance. Nevertheless the rules which we have indicated appear to hold good in so great a number of cases that they are well worthy of being remembered, and may serve as a basis on which we may reason concerning the nature of the action going on beneath volcanic vents. From the study of the external appearances of volcanic mountains, combined with investigations of those which have been dissected by denudation, we are able to picture to our minds the series of actions by which the great volcanic mountains of the globe have been slowly and gradually built up. In the first instance the eruptions appear to have taken place at several points along a line of fissure, but gradually all of these would become choked up except one which became the centre of habitual eruption. From this opening, ejections, firstly of lavas of intermediate composition, and afterwards of basic materials, would take place, until a volcano of considerable dimensions was built up around it. But at last a point would be reached in the piling up of this cone, when the volcanic forces below would be inadequate to the work of raising the liquid lava through the whole length of the continually upward-growing tube of the volcano. Under these circumstances the expansive force of the imprisoned steam would find it easier to rend asunder the sides of the volcanic cone than to force the liquid material to the summit of the mountain. If these fissures reached the surface explosive action would take place, in consequence of the escape of steam from the glowing mass, and scoria-, tuff-, and lava-cones would be formed above the fissure. In this way, as we have already pointed out, the numerous 'parasitic cones' which usually abound on the flanks of the greater volcanic mountains have been formed. The extrusion of these masses of scoriæ and lava on the flanks of the mountain tends, not only to increase the bulk of the mass, but to strengthen and fortify the sides. For by the powerful expansive force at work below, every weak place in the cone is discovered and a fissure produced there; but by the extrusion of material at this fissure, and