394 REVIEWS 



However this law only holds for dilute solutions, and he attempts to 

 show that magmas are such dilute solutions. 



Applying the foregoing to the rate of diffusion under the assump- 

 tion that the solvent fluid is kept at constant composition and the 

 dissolved substance is diffusing through it, he calculates the speed of 

 the diffusion of various salts in water. He finds that common salt 

 will diffuse fast enough to semi-saturate water at a distance of i"™ in 

 one day and loo meters in 270,000 years. From certain observations 

 he infers that lavas are at least fifty times as viscous as water, and there- 

 fore diffusion in such a substance would be fifty times as slow as that 

 in water. Taking copper sulphate as the salt whose speed of diffusion 

 should be multipled by fifty to represent the rate of diffusion in an 

 average magma, he finds that in a million years there would take place 

 a diffusion sufficient to impregnate the magma for forty-nine meters, 

 and semi-saturate it to a distance of twelve meters. Finally he dis- 

 misses the hypothesis of the diffusion of miscible liquids because (i) 

 diffusion is too slow a process ; (2) a higher temperature is postulated 

 at the top than at the bottom, and (3) because more or less convec- 

 tion is unavoidable, and he proceeds to consider diffusion of immisci- 

 ble liquids. 



He gives some examples of liquids which, above certain tempera- 

 tures peculiar to each, are perfectly miscible, and below these definite 

 temperatures separate into two immiscible liquids. This separation 

 is accompanied by a sudden contraction of volume, and therefore 

 all such separation is aided by increase of pressure. His points 

 against such separation appear to be as follows: (i) The temperature 

 of separation is first reached along the walls containing the magma, 

 and therefore the separating component will condense on the contain- 

 ing walls much as frost and dew on good conductors of heat. Such a 

 separation, however, involves molecular flow from the interior of the 

 fluid to the walls, and therefore no large portion of even a moderately 

 viscous magma could thus separate. (2) If the fluid does not con- 

 dense on the sides, it must aggregate as a fog in the center of the 

 fluid. This fog cannot condense farther, as a fog does in the atmos- 

 phere, on account of the great viscosity that he assumes the lava to 

 have. (3) Magmas are not heated much above the melting point, and 

 therefore the range in their temperature is not great enough to present 

 much likelihood of their crossing the critical temperature of separa- 

 tion. This follows from the law of fusion that if a melted magma 



