STUDS' UF IGNEOUS KOCKS. 279 



With such an understanding of the causes of heterogeneity in 

 lock solutions the great variability in the composition of igneous rocks 

 as shown by chemical analyses, and by a quantitative study of their 

 mineral composition, appears as the natural, as well as the logical 

 result of their mode of formation. 



Mineralogical and constitutional facies of igneous rocks are 

 readily comprehended; and the absence of fixed types of magmas, or 

 of frequently recurring bodies of igneous rocks with definite or invari- 

 able composition becomes ■' natural " and is the thing to be expected. 

 Variations in texture Avithin one rock mass, and among rock bodies 

 having various ] nodes of occurrence are readily understood as the 

 results of variabihty in the conditions attending volcanic eruption. 



As to the possible character of volcanic eruption, some conception 

 of it may be derived from a consideration of the probable condition of 

 highly-heated rock material under great pressure deep beneath the 

 surface of the eai'th, as well as its probable experience in moving 

 upward and out upon the earth's surface. 



The high temperatures of volcanic lavas when they reach the 

 atmosphere ; the fact that they were losing heat continually from the 

 time of their first movement upward, the evidence that they were 

 completely liquid at some stage in their eruption ; together with the 

 obsei-ved gradient of increase of temperatui^e downward from the 

 surface of the earth, all combine to show that rock magmas come 

 from some region where the temperature is considerably above the 

 melting point of igneous rocks. The behaviour of the earth as a rigid 

 globe, and the known effect of pressure in countei'acting that of heat, 

 together with its estimated high gradient of increase downward within 

 the earth, force the conclusion tlfat at sufficient depth magma, though 

 hot enough to be liquid, behaves as a solid. Such conditions of heat 

 and pressure can not vaiy abruptly from place to place, but must be 

 nearly the same for large volumes of material ; and differences of 

 tempei-ature and pressure must olitain very gradually ; chiefly in 

 vertical directions. Magma in such a position must be m a virtually 

 static condition until it experiences change of pressure or stress. 

 Whatever its composition it must remain unclianged. 



A change of stress may come about by movement in the over- 

 lying portion of the earth. Orogenic movement, readjustment of the 

 upper rigid rock mass, from whatever cause, when profound, must 

 affect the sti-esses in still deeper parts. The known crustal movements 

 behave as bendings of the upper rock mass, which in places at the 

 eai-th's surface appear to result in tensile stresses; in places, in com- 

 pressional stresses. Beneath each of these the effective stresses must 

 be of the opposite kind ; under the tensile, compressive stresses ; and 

 andei- the upper compressive ones, tensile stresses. Tensile stresses 

 should occur at some distance below ocean beds, and more especially 

 along the borders of oceans and continents. Compressive stresses 

 siiould occui- in general beneath continental masses. 



Tensile stress, as at the bottom of a synclinal arcli, operating in 

 a rigid mass must commimicate downward as far as the mass behaves 

 rigidly. Where the hot mass is potentially fluid — that is, is kept solid 

 by pressure^ — change of stress must be followed bj^ change of position 

 of the mass. A tendency to pull apart, or sti'etch in the potentially 

 fluid mass must be followed by a yielding of the mass. At a point 



