sion advanced on a previous page, that the primary force which causes lava to rise in the conduit of a volcano is not steam pressure.
How molten lava becomes charged with water can only be conjectured. It is well known that many liquids, especially when highly heated and under heavy pressure, will absorb gases. In a similar way we may conceive that liquid or plastic rock, on coming in contact with water, will absorb the steam produced.
When molten lava rising in the conduit of a volcano passes through water-charged rocks and nears the surface, pressure is relieved and the occluded steam escapes. This escape is either quiet or explosive, dependent on the nature of the magma in which the steam is dissolved. If the magma is highly fluid, as in the case of many basic lavas when extruded, the steam escapes quietly; but if the magma is viscous, as is the usual condition of acid lavas when erupted, violent explosions are apt to occur. The quantity of steam absorbed also influences the fusibility of a magma. Apparently the larger the quantity of occluded steam, the more liquid the molten rock becomes. Greater freedom may thus be afforded for the passage of a magma in the upper portion of the conduit through which it rises than obtains at lower levels. Something of the intermittent character of volcanic eruptions may depend on this cause. Probably, also, the quantity of water present in a magma has an influence on the nature of the minerals formed as it cools. For this reason one would expect differences to appear in the mineralogical composition of rocks formed from magma that have cooled near the surface, and those that failed to reach the water-charged portion of the earth's crust.
The intimate connection between subterranean injections and volcanoes leads to the suggestion that the domes above intruded magmas may become fractured and give origin to volcanoes which would be supplied by local reservoirs. Something like "craters of elevation" may be thus formed.
If two or more cisterns of molten rock should be formed in the earth's crust near each other, or at different levels near the same radius of the earth, and fractures formed above them which would admit of the escape of their material to the surface, the striking phenomena of two adjacent volcanoes erupting independently of each other might result. Such an occurrence is rendered more probable by the fact that reservoirs beneath subtuberant mountains are supplied through fissures from below, which might become closed, thus isolating bodies of injected material in the earth's crust. Even if the feeding fissures were not closed, the large cisterns of fused rock to which they lead might discharge some of their material without immediately affecting the plastic central mass of the earth, which in these sugges-
