CHAPTER XXIV. 179 



ume of soil can be caused to receive into its interstitial spaces, gener- 

 ally spoken of as its voids. Of the volume that can be introduced 

 into dry soils a portion is susceptible of being removed by drainage, 

 and the remainder, held within the capillaries and as films on the grain 

 surfaces, is removable by processes of evaporation only. 



It requires no effort of thought to comprehend that conductivity 

 is an important factor in storm run-off, and in estimating the probable 

 yield of a ground storage. The literature of the subject is scant and 

 unsatisfactory. A splendid field is open here for systematic experi- 

 ment. Such records of experiments that I have been able to obtain 

 indicate that the tests were made by downward filtration. In my 

 own tests I have found it practically impossible to get uniform results 

 by this method. There is always a conflict between the ingoing water 

 and the outgoing air it has displaced. Very uniform results were 

 obtained by reversing the method and filtering upwards. It is true 

 that this does not exactly meet the conditions of rain soaking down- 

 ward, but it is free from any complications arising from imprisoned 

 air, and furnishes an uncomplicated base from which to start. Then 

 we must remember that the horizontal movements of the underground 

 water are greater than the vertical, and the upward filtration method 

 will give true results for these lateral movements. 



The texture of the soil is the governing factor as to rate of conduc- 

 tivity. It ranges from well rounded pebbles, through the intermediate 

 stages of sand and loam to clay. A study of the mechanics of the 

 granular soils presents some particularly interesting points. Assume 

 the granules to be all true spheres and assorted by sizes. It can read- 

 ily be demonstrated that the voids form the same percentage of the 

 total cubical contents whatever may be the diameter of the spheres, 

 and the area of the passage between the spheres bears precisely the 

 same ratio to the area of the circumscribing cross section irrespec- 

 tive of the diameter of the spheres. The closer a soil approximates 

 to spherical grains of uniform size the greater the capacity for water 

 and the larger the percentage of waterways. This is a property of 

 uniformity of size. By mixing many different sizes together in such 

 proportions that each succeeding smaller size enters the interstices "' 

 the preceding layer a soil may be made impervious to water save by 

 capillary action. Proper mixtures of gravel and sand make good water- 

 tight dams. This feature of mixture will frequently explain the im- 

 perviousness of stream beds in sandy gravel. 



To return to the matter of voids and water passages. Although 



