REVIEWS 397 



we do not know where the increasing pressure overtakes the effect of 

 increasing temperature and superheated gases, and therefore we may 

 expect some lavas to be very liquid, and to permit considerable molec- 

 ular flow, while in others the viscosity might be so great as to prevent 

 all such diffusion. Therefore is the assumption that the viscosity of 

 unerupted lavas is the same as that of an erupted and cooling sheet 

 warranted, and is there not a possibility of both far greater and far 

 less viscosity than that assumed by Dr. Becker ? 



We have, moreover, positive evidence that lavas may be very fluid if 

 the pegmatitic structure in the rocks of Christiania described by 

 Brogger as igneous be so, as he affirms for his region at least. Also 

 the intrusion of erupted matter for long distances along planes of fis- 

 sibility, in the Archean rocks of the Lake Superior region, demands 

 more or less fluidity. If the alternate layers of banded rhyolites have 

 a different amount of absorbed water vapor, as Professor Iddings sug- 

 gests, this phenomenon might better be considered, as the result of 

 impregnation of the magma by vapor shortly before its eruption. 

 Moreover the fact that in the banded rhyolites we do not have dif- 

 fusion between the layers is no proof that under other conditions we 

 should not have such diffusion. Some magmas have been erupted 

 under such conditions that they are perfectly homogeneous, while 

 others show marked differentiation. 



We reiterate that Dr. Becker is correct in concluding that differen- 

 tiation into miscible liquids could play only a subordinate part in 

 rock segregation. His attack on the theory of differentiation into 

 immiscible liquids seems open to question; for the separation must 

 instantly begin throughout the portion that has reached the critical 

 temperature. This separation will start at many different centers, as 

 in crystallization, and since molecular flow is rapid for short distances, 

 and very slow for longer distances, this operation would be rapid. 

 When these immiscible liquids are once formed, flow, stirring, gravity, 

 etc., will help aggregate like to like. This segregation may be still 

 farther assisted by the temperature sequence of crystallization of 

 double salts. 



His point, that lavas are too viscous to permit such separation and 

 aggregation, has been already answered. His third point, that we 

 cannot have wide enough range of temperature for this mode of 

 separation, following from "the law of fusion," may have force in 

 some instances, but where we have considerable mass of magma there 



