24 o VULCANISM 



the melting point of either. Something of the same sort appears to take place 

 when rock becomes liquid. The distinction between such solutions and molten 

 rock is not very sharp, but it is essential to know that the order in which the min 

 erals crystallize from lavas is not dependent on their melting temperatures. It 

 appears rather to depend on the order in which the solution becomes saturated 

 with the constituents of each of the several minerals. For example, quartz, which 

 has a very high melting-point, may crystallize out from a lava much later than 

 minerals which have lower melting temperatures. The solutions are exceedingly 

 complex, and include a wide range of chemical substances. Chief among them 

 are silicates of aluminum, potassium, sodium, calcium, magnesium, and iron 

 (Chapter X) , with minor ingredients of nearly all kn , wn substances. Gases as well 

 as rock materials enter into the composition of the igneous rock. When lava is 

 cooled suddenly, the result is glass, every part of which has essentially the same 

 composition that the liquid had, but even in this case some of the gases of the 

 lava escape. If the cooling is slower, the various substances in the mixture crys- 

 tallize out into minerals in the order in which they severally reach saturation. 

 This involves the principle that solubility is dependent on temperature, and that 

 as the temperature sinks the degree of solubility declines, and the saturation-point 

 for some constituents of the solution is reached earlier than that for others. With 

 sufficiently slow cooling, all the material passes into the solid state by the crys- 

 tallizing of the several minerals in succession. This does not mean that two or 

 more minerals may not be forming at the same time, but it means that some 

 minerals may be crystallized out while the surrounding material is still fluid. In 

 most igneous rocks, nearly perfect crystals of certain minerals are common, while 

 other minerals, crystallizing later, adapt themselves to the space left between older 

 crystals. This conception is supported by the fact that some lavas, while still in the 

 fluid condition, contain well-formed crystals, very much as water in certain 

 conditions may be filled with crystals of ice. 



Temperature of lava. Accurate determinations of the temperatures of liq- 

 uid lavas have not been made; but it is clear from the white heat of some lavas 

 that their temperatures are appreciably above the melting-point. This is 

 also a necessary inference from the length of time lavas remain fluid, in spite of its 

 contact with cooler rock, through its miles of ascent. From various facts it is 

 probably safe to assume that the original temperatures of lavas as they rise to 

 the surface are in some cases considerably above 2,000 Fahr. (1,093 C.). Even 

 such a temperature must be somewhat below the original temperature of the lava, 

 because some heat must be lost in rising, both by contact with the cooler rocks 

 through which it rises, and by the expansion of tha gases within them. 



Depth of source. Attempts have been made to determine the depth from 

 which lavas rise, by calculations based on the earthquake tremors accompanying 

 eruptions; but such calculations really tell very little concerning the true point 

 of origin of the lava. At most they probably tell merely where the ascending 

 lavas begin to rupture the rock through which they pass, and rupture may not be 

 possible below the zone of f racture,which is probably not more than eleven miles deep. 1 

 In the zone of flowage below, where the pressure is too great to permit fracture, the 

 lava not improbably makes its way by some boring or fluxing process, which 

 might not be capable of giving rise to seismic tremors. The tremors perhaps com- 



1 Adams: Jour. Geol., Vol. xx (1912), pp. 97-118. 



