248 clouds of steam, showers of fragments, and a noise utterly indescribable. Or if, on the other hand, the current should encounter a ridge or hill across its path, it will accumulate in front of it until it either finds egress round the side or actually overrides and entombs the obstacle. The hardened crust or shell within which the still fluid lava moves serves to keep the mass from spreading. We often find, however, that the lava has subsided here and there inside its crust, and has left curious cavernous spaces and tunnels. Into these, when the whole is cold, we may creep, and may find them sometimes festooned and hung with stalactites of lava. As a rule a lava-stream shows three component layers. At its bottom lies a rough, slaggy mass, produced by the rapid cooling of the lava, and the breaking up and continued onward motion of the scoriform layer. The central and main portion of the stream consists of solid lava, often, however, with a more or less carious and vesicular texture. The upper part, as we have seen, is a mass of rough broken- up sla.gs, scoriaa, or clinkers. The proportions borne by these respective layers to each other vary continually. Some of the more fluid ropy lavas of Vesuvius have an inconstant and thin slaggy crust; others may be said to consist of little else than scorize from top to bottom. These diver- .gences in texture seem to depend largely upon the amount of interstitial steam imprisoned within the lava, and the conditions under which it can effect its escape. Throughout the whole mass, but more especially along its upper surface, the steam under its diminished pressure expands, and push- ing the molten rock aside, segregates into small bubbles or irregular cavities. Hence, when the lava solidifies, these steam-holes are seen to be sometimes so abundant that a detached portion of the rock containing them will float in water. They are often elongated in the direction of the motion of the lava-stream. But, besides producing a general vesicular texture in the upper parts of the lava-stream, the aqueous vapour gives rise to much more striking features on the surface of the lava. If the outburst takes place from an orifice or fissure on the exterior of the volcanic cone, so vast an amount of steam will rush out there, with such boiling and explosion of the lava, that a cone of bombs, and slags, and irregular lumps of lava, will probably form round the spot—in fact a miniature or parasitic volcano, which will remain as a marked cone on its parent mountain long after the er11ption which gave it birth has ceased. Moreover, even after such abundant discharge of steam, the lava—stream continues to exhale it, as it were, from every pore. Here and there on the surface of the moving mass a fissure opens, and a column of roaring hissing vapours rushes out from it, accompanied as before by an abundant discharge of lava- fragments, or even by the rise and outflow of the lava from beneath. Some lava-streams are thus dotted over with small cones a few feet or yards in height. Besides the steam which, in condensing, makes its presence so con- spicuous, many other vapours entangled in the pores of the lava escape from its fissures. The points at which vapours are copiously disengaged are termed fmnarole. Among the exhalations, chlorides may be mentioned as particularly prominent ; chloride of sodium frequently shows itself, not only in fissures, but even over the cooled crust of the lava, in small crystals, in tufts, or as a granular and even glassy incrustation. Chloride of iron is deposited as a yellow coating at the fumarole, where also bright emerald green films and scales of chloride of copper may be more rarely observed. Many chemical changes take place in the escape of the vapours through the lava. Thus specular-iron, probably the result of the mutual decomposition of steam and iron chloride, forms abundant scales, plates, and small crystals in the fumarole and vesicles of the lava. Sal- ammoniac also appears in large quantity on many lavas, not GEOLOGY [11I. DYNAMICAL. merely in the fissures, but also on the upper surface of the current. This salt is not directly a volcanic product, but results from some decomposition, probably from that of the aqueous vapour, whereby a combination is formed with atmospheric nitrogen. The hardened crust of a lava-stream is a bad conductor of heat. Consequently, when the surface of the mass has become cool enough to be walked upon, the red hot mass may be observed through the rents to lie only a few inches below. Many years therefore may elapse before the temperature of the whole mass has fallen to that of the surrounding soil. Eleven months after an eruption of Etna, Spallanzani could see that the lava was still red hot at the bottom of the fissures, and a stick thrust into one of them instantly took fire. The Vesuvian lava of 1785 was found by Breislak seven years afterwards to be still hot and steaming internally, though lichens had already taken root 011 its surface. The ropy lava erupted by Vesuvius in 1858 was observed in 1870 to be still so hot, even near its termination, that steam issued abundantly from its rents, many of which were too hot to allow the hand to be held in them. Hoffmann records that the lava which flowed from Etna in 1787 was still steaming in 1830. But still more remarkable is the case of J orullo, in Mexico, which sent out lava in 1759. Twenty-one years later a cigar could still be lighted at its fissures; after 44 years it was still visibly steaming; and even in 1846, that is, after 87 years of cooling, two vapour columns were still rising from it.‘ This extremely slow rate of cooling has justly been re- garded as a point of high geological significance in regard to the secular cooling and probable internal temperature of our globe. Some geologists have argued indeed that, if so comparatively small a portion of molten matter as a lava stream can maintain a high temperature under a thin, cold crust for so many years, we may, from analogy, feel little hesitation in_believing that the enormously vaster mass of the globe may, beneath its relatively thin crust, still continue in a molten condition within. More legitimate deductions, however, might be drawn, if we knew more accurately and precisely in each case the rate of loss of heat, and how it varies in different lava-streams. Sir William Thomson, for instance, has suggested that, by measuring the temperature of intrusive masses of igneous rock in coal-workings and elsewhere, and comparing it with that of other non—volcanic rocks in the same regions, we might obtain data for calcu- lating the time which has elapsed since these igneous sheets were erupted. In its descent a stream of lava may reach a water-course, and, by throwing itself as a great embankment across the stream, may pond back the water and form a lake. Such is the origin of the picturesque Lake Aidat in Auvergne. Or the molten current may usurp the channel of the stream, and completely bury the whole valley, as has happened again and again among the vast lava-fields of Iceland. No change in physiography is so rapid and so permanent as this. The channel which has required, doubtless, many thousands of years for the water laboriously to excavate, is scaled up in a few hours under 100 feet or more of stone, and a still longer interval may elapse before this newer pile is similarly eroded. By suddenly overflowing a brook or pool of water, molten lava sometimes has its outer crust shattered to fragments by a sharp explosion of the generated steam, while the fluid mass within rushes out on all sides. Numerous instances have occurred where the lavas of Etna and Vesuvius have pro- truded into the sea. Thus a current from the latter moun- tain entered the Mediterranean at Torre del Greco in 1794, and pushed its way for 360 feet outwards, with a breadth of
1 E. Sehleiden, quoted by Naumann, Ge0l., i. p. 160.