A METHOD OF DETERMINING THE PULL OF SOILS 13 



be cited. Tests of sand, and of* adobe, separate and mixed, 

 were made. When mixed, equal parts of each were taken. In 

 all cases the largest soil particles, or, more exactly, intrusions, 

 like small pebbles, were removed by passing the soils through a 

 coarse sieve. After this each kind of soil tested was separated 

 into two parts by sifting through sieves of two sizes, namely, 

 20-mesh and 40-mesh (that is, 20 and 40 meshes to the inch). 

 Repeated tests were made in every instance.- For the mixed 

 soil, the highest point reached by the mercury was 22.5 mm. 

 The 20-mesh sand registered 6 mm., and the 40-mesh sand, 9.5 

 mm. Unsifted adobe, '20-mesh, stood at 18 mm., and 40-mesh, 

 25 mm. The capillary pull of the sifted adobe soil is not easily 

 determined, owing, probably, to its peculiar physical nature. 

 An interesting and characteristic quality of the tests with sand, 

 especially the coarse sand, is that the mercury rises very quickly 

 at first, but for a brief time, after which it maintains an even 

 height for a long period. For example, the mercury in a test 

 with 40-mesh sand remained at 9.5 mm. for over three hours. 



Other results might be given as well, but those presented are 

 sufficient to show the behavior of the soils mentioned as regards 

 the quality in question. It is recognized that the method is 

 not without theoretical, and, possibly, practical objections. Of 

 these, possibly the most serious is the inevitable difference be- 

 tween soils, even of the same kind, when treated in such tests, 

 as regards their compactness. To this it can be suggested that 

 settling under water, and parallel treatment otherwise so far as 

 possible, makes the objection possibly less serious for this method 

 than for other methods. The objection holds, in other words, 

 quite as well for the direct observation of the capillary rise of 

 water in soils. Errors from such a source, in the manner of de- 

 termining the capillary pull in the way here described, can 

 be minimized also by making the tests several times, and under 

 parallel conditions of temperature, which is not feasible by the 

 direct method referred to. Finally, the manometer method 

 commends itself as a convenient and striking means of demon- 

 strating an important physical characteristic of soils, whether, 

 which is unimportant, it is an entirely new method or not, and 

 independent of the probability that it is theoretically not ideal. 



