290 University of California Publications in Geology [Vol. 13 



Pointing out the laboriousness of computation from this equation 

 as it stands he then undertakes its simplification. He says : 



The abscissa of the probability curve is not the natural independent variable 



v 



for calculating diffusions. Greatly preferable is the quantity — What 



is needed to facilitate computations of diffusion in a neat form, such as will not 

 require diagrams to render the relations clear, is a table in which q, or better 

 2q, is expressed in the terms of v/c. Such a table is given below, but only in 



skeleton for intervals of v/c of .05 Knowing 2q for a stated value of 



v/c the value of x in centimeters follows easily from the equation, x—2q\l let. 

 . . . . The values of 2q are given in part because they are the distances in centi- 

 meters after the lapse of one particular time, that, namely, which is the reciprocal 

 of the diffusivity. Thus in the case of salt with a diffusivity of .00001, after a 

 lapse of 100,000 seconds, or some 28 hours, let = 1 and 2q represents the distance 

 answering to v/c. 



In computing the table referred to, his value for the diffusivity of 

 salt, k = .00001, refers to a concentration of 0.14785 gms. per cubic 

 cm. at a temperature of 7.7°, then for the dilution of v/c = 0.5 per 

 cent he finds that in 100,000 years salt would diffuse a distance of 

 731 feet. 



In rock media water must be considered as dispersed in films 

 through the pore spaces, and a diffusing substance would therefore 

 have to travel a relatively great distance to obtain only slight dilution. 

 The tendency would probably be for it to move more rapidly through 

 space, but this would be to some extent offset by the tortuousness of 

 its path. In the face of these considerations the evaluation of the rate 

 of diffusion can not be greatly assisted by mathematical calculation, 

 but must depend upon experimentation. This, however, is a difficult 

 matter, since the duplication of geological conditions in the laboratory 

 would be very dubious, especially in view of the necessary magnitude 

 of the time factor. It is therefore necessary for the present to depend 

 chiefly upon the results obtained in nature's laboratory, knowing that 

 diffusion must occur wherever solute and solvent are in contact, and 

 bearing in mind the quantitative results obtained by G. F. Becker for 

 diffusion in a homogeneous and concentrated solvent. 



In the dispersion of a solute through the ground water filling the 

 pore spaces of rock, diffusion can not be supposed to operate alone. 

 Osmosis would undoubtedly play an important part in many instances. 

 For example, suppose that a solute has obtained a fair degree of con- 

 centration in a rock chamber, either by diffusion or by convection, 

 and the surrounding rock is saturated with comparatively pure ground 

 water. If the pore spaces of the chamber walls are too small to per- 

 mit the passage of the ions of solute, and yet are large enough to 



