326 MR. ALFRED HARKER ON THE GABBRO [Aug. 1894, 



only when the concentration at every point is inversely propor- 

 tional to the absolute temperature. Soret has demonstrated ex- 

 perimentally the greater concentration of the salt in the cooler 

 part of the solution. This law, known as ' Soret's principle,' has 

 been applied by Lagorio, Teall, Brogger, Vogt, Iddings, and others, 

 to the case of an igneous rock-magma regarded as a solution. In 

 particular, the relative richness of the marginal parts of an intrusive 

 mass in the more basic minerals has been explained as due to the 

 concentration of the less soluble constituents of the magma, while 

 still fluid, in the cooler region. 



On this point one or two remarks may be made. In the first 

 place, the idea cannot be entertained at all in connexion with 

 Lagorio's theory of a single general solvent. As Backstrbm * has 

 pointed out, differences of temperature could not, on that hypothesis, 

 alter the relative concentration of different dissolved constituents. 

 We are therefore driven to some less precise and more complex 

 view of the nature of the ' solution ' in a rock-magma. Supposing, 

 however, that something analogous to Soret's principle still holds, 

 we may enquire whether this is adequate to explain the degree of 

 concentration actually observed in the case of the Carrock Fell 

 gabbro. The precise law arrived at by van't Hoff, identical with 

 the law for gases, states that, in the condition of equilibrium, the 

 concentration of the dissolved substance varies inversely as the 

 absolute temperature. Now, in our case, the amount of iron ores in 

 the rocks at the margin of the mass is at least twenty-five times 

 the amount in the rocks at the centre ; but it is manifestly 

 impossible that the absolute temperature of the magma at its centre 

 could ever be twenty-five times, or even five times, that at its 

 margin. The explanation is clearly insufficient to account for the 

 facts. 2 



It must be understood that I speak here of differentiation effected 

 in situ in a magma, which may fairly be assumed to have been of 

 uniform composition when intruded into its surroundings. We are 

 not, in this case, concerned with successive differentiations of 

 magmas and partial magmas amidst new surroundings, as conceived 

 by Iddings, or as described by Brogger. On any solution-hypothesis 

 of rock-magmas, Soret's principle does not afford an explanation of 

 the variations observed in the Carrock Fell gabbro. Further, I 

 question how far that principle, which holds good for dilute solutions, 

 can throw light on the physics of a rock-magma near the point of 

 crystallization, which must be compared with a nearly saturated 

 solution. 



The phenomena of differentiation described by Prof. Brogger 

 in the eruptive rocks of the Christiania basin, and especially in 

 those of the Gran district, 3 differ in a fundamental respect from 



1 .Tourn. of Geol. vol. i. (1893) p. 774. The author, however, does not 

 limit bis criticism, as is here done, to this particular view of the solvent medium. 



2 This argument has been advanced by me in a brief note : Geol. Mag. 1893, 

 pp. 546, 547. 



3 Quart. Journ. Geol. Soc. vol. 1. (1894) pp. 15-37. 



