FEAGMENTATION — BARRELL 299 



mix the originally separated intermediate and basic shells. The 

 ultrabasic rocks, such as the metallic meteorites, have not been 

 found to contain the radioactive elements, and therefore deeper 

 shells of such composition, if they exist, have not been considered 

 as entering notably into the mechanism of igneous activity. It is 

 possible that the dense, rigid, and strong centrosphere consists of 

 such ultrabasic materials. 



Fourth. The rising magmas break through the strong outer crust 

 only occasionally and in certain areas, notably along the zones of 

 mountain building. F. D. Adams has shown that the great pressures 

 existing within the crust increase the strength of rocks many fold, 

 and this strength is not decreased by temperatures up to 550° C. 

 The fact remains, however, that magmas do rise and the crust does 

 yield in order to permit mountain making and maintain isostasy. 

 This would appear to necessitate the hypothesis that the temperature 

 is high enough to reverse the effects of pressure upon strength and 

 that rock flowage takes place by recrystallization at temperatures 

 approaching those of fusion. At considerable depths these tempera- 

 tures of fusion are far above the limit of 550° used in the experi- 

 ments by Adams. The zone of greatest strength would tend to hold 

 the magmas down until they accumulate in sufficient volume to dis- 

 solve, drill, and break their way through this resistant zone. Above 

 this they would tend to spread out in lopoliths and batholiths, and a 

 portion would rise higher, forming, as volcanoes, safety valves for 

 the internal forces. These relations are all expressed in Figure 1. 

 The localization of fragmentation to certain regions may conse- 

 quently be due, in large part, to the general resistance of the 

 lithosphere, only rarely broken through in great volume. The basic 

 magmas, because of their lower temperature of fusion, lower content 

 of gases, and higher density, would have smaller power of penetra- 

 tion than more acidic magmas. 



Fifth. Downsinking in obedience to the demands of isostasy is 

 facilitated by the general heating of the whole crust below the zone 

 of lopolithic intrusions. The heat of these intrusions must escape 

 upward. Accordingly the rocks solidifj^ and the outer crust becomes 

 stronger; but below, the heat is held in, the upper limit of the 

 asthenosphere necessarily has risen because of this higher tempera- 

 ture, and the crust as a whole has grown weaker to resist the vertical 

 loads that are imposed by tlie new differences in density. 



To sum up the argument, some periods, such as the Keweenawan, 

 and certain regions, sucli as that of Lake Superior, are seen to be 

 characterized by especially voluminous extrusions and intrusions of 

 basic magmas. If the earth is layered in density, such heavy magmaa 

 would presumably originate deeper than those which are more sili- 

 ceous. The siliceous magmas also give off basic fractions by differ- 



