122 H. S. JEVONS, H. I. JENSEN AND C. A. SUSSMILCH. 



Probably the most important method of differentiation 

 during consolidation is fractional crystallisation, as it has 

 been denoted by Becker. The crystals of salts crystallising 

 from an ordinary aqueous solution are always deposited 

 upon the cool surface unless either crystallisation is rapid, 

 when they form also in all parts of the liquid and sink, or 

 the solution is viscous. By analogy we can quite under- 

 stand that a fairly fluid magma would deposit the first 

 products of crystallisation on the walls of the chamber it 

 occupied. Of course, the supersaturation of the neigh- 

 bouring magma as regards the constituents separating is 

 exhausted by any accretion of the crystal; 1 and to restore 

 this supersaturation by diffusion alone would require an 

 impossible length of time, as shown by Becker. Hence 

 differential crystallisation on the cooling surface can only 

 take place when mechanical agitation constantly renews 

 the magma in contact with it; and this can happen in two 

 ways, either by convection currents due to the cooling of 

 the magma, or by slow earth movements. Convection 

 currents must exist when the magma is fairly fluid, and the 

 only other condition is that the magma is not cooled rapidly. 

 Probable examples of this kind of differentiation are the 

 Oarrock Pell gabbro described by Harker, 2 (case of Becker's 

 laccolite), the essexite mass of Brandberget with its 

 pyroxenite border, 3 and perhaps the laccolites (?) of Yogo 

 Peak and Bearpaw Peak in Montana. 4 



Oases in which fractional crystallisation results from 

 motion of the magma caused by earth movements are 

 chiefly dykes, sills, flat laccolites, and pipes. The folding 

 of strata is usually a very slow process, and in intrusions 



1 Chemical Crystallography by A. Fock, (trns. Pope) p. 50, (Clarendon 

 Press, 1895). 



2 Q.J.G.S., Vol. 50, (1894) pp. 311 -335. 



3 Brogger, Q.J.G.S., l, (1894), p. 31. 



4 Weed and Pirrson, Am. Jour. Sci., Vol. l, (1895) p. 467 and Vol. li, 

 (1896) p. 351, respectively. 



