R. A. Daly — Mechanics of Igneous Intrusion. 291 



to a magma layer as dense as itself. Yet it appears not to 

 stretch the probabilities too far, to assume that pressure solidity 

 would not characterize, in general, the lower basic part of a 

 magma basin above the level where a normal siliceous sediment 

 or schist block would come to rest. As a rule, that would 

 mean that the magma must become acidified by the assimila- 

 tion of the more siliceous rock. The equilibrium would thus be 

 disturbed and an upward current of the locally modified magma 

 would be set up. Such action would be specially notable in 

 the case of gabbro and peridotite magmas intruded into any 

 staple member of the fundamental crystalline schist complex. 

 Downwardly-directed currents in the magma would also be 

 developed during the sinking of the blocks. 



Such currents would tend toward the rapid formation of 

 horizontal homogeneity at the various levels in the chamber, 

 and of vertical heterogeneity (on a larger scale than the similar 

 heterogeneity expected by the older assimilation theory), 

 governed by the law of increasing density with depth. The 

 more acid layers would concentrate at the top, the more basic 

 remain at the bottom, of the magma basin. The layers would, 

 it is believed, be more or less definitely composed by the laws 

 of differentiation operating by diffusion. It is certain that 

 diffusion is aided by agitation,* and in the density currents 

 and those induced by the falling blocks we have two efficient 

 causes for agitation. Further, Braun and Alexejew have 

 shown that enhanced pressure stimulates the differentiation of 

 a complex magma into distinct fluids. f Increased pressure is, 

 as we have seen, felt in the magma by the conversion of each 

 fallen block into liquid rock. Finally, certain other facts of 

 recent discovery seem to bear out the conclusion that differen- 

 tiation by the gravitative effect actually occurs. Morozewicz 

 cites instances of the process in his fusion experiments and 

 in the study of glass furnaces.;}; Doelter has stated that such 

 results adhere to special cases both in his own experiments 

 and in those of the Russian investigator; yet their significance 

 is still great, since they agree with Gony and Chaperon's 

 theoretically deduced principle of gravitative stratification in 

 saline solutions,! as well with some positive field observations. ' 

 For example, Sir A. Geikie describes the separation of a lower 

 layer of picrite and an overlying layer of olivine basalt in the 

 same lava-flow, and finds it probable that similar differentiation 

 has taken place in basic sills. |, Becker's physical researches^ and 

 Eosenbusch's conclusion from more purely penological consid- 



* Becker, this Journal, iii, 25 (1897). f See Becker, op. cit., p. 32. 



JTscher. Min. u. Petrog. Mitth., xviii, 170, 233 (1898). 

 ?; Ann de chimie et de physique, ser. vi, vol. xii, 1887, p. 384. 

 I Ancient Volcanoes of Great Britain, London. 1897, vol. i, pp. 419 and 

 442, and vol. ii, p. 310. *[ Op. cit., p. 37. 



Am. Jour. Sci. — Fourth Series, Vol. XV, No. 88.— April, 1903. 

 20 



