FOREST TYPES IX CENTRAL ROCKY MOUNTAINS. 123 



to withdraw the desired sample. When the soil was very dry the 

 obtaining of samples at a depth of 2 feet was practically impossible, 

 and samples taken at a depth of 18 inches were substituted. 



Even a depth of 2 feet does not seem adequate when one has 

 noted the deep penetration of occasional roots in this soil. There- 

 fore, when in 1914 this study was extended to a greater number of 

 stations both near Fremont and in other localities, a plan was adopted 

 which had been in use at Wagon Wheel Gap since 1911, made neces- 

 sary there by large rock fragments all through the soil. This was 

 the plan of " soil wells. " 



The preparation of a ''soil well" consists in making an excavation 

 from 18 to 24 inches in diameter and Sh to 4 feet deep, the space 

 being then filled with the same material as has been taken out, 

 with the exception of the rocks or coarse gravel which interfere with 

 soil sampling. As material must be borrowed from some other 

 place to make up for the coarse material sifted out, the resultant 

 soil in the "well" is, of course, very different from that which pre- 

 viously filled the space. 



The following objections may legitimately be raised against soil 

 sampling under these conditions: 



(1) The "well" soil, weight for weight or volume for volume, will 

 possess a higher water-holding capacity than the native soil, on 

 account of the absence of the rock fragments, which are of relatively 

 little importance in water storage. 



(2) The well soil will at the outset be of the same composition from 

 surface to bottom, whereas the native soil may vary greatly with 

 change in depth. 



(3) The well soil, not being protected and held by humus accumu- 

 lations at the surface, will tend, through leaching and transport of 

 fine material, to become coarser at the surface and of finer texture 

 at greater depths; hence the water-holding properties will then vary 

 with depth and in a direction opposite to the variation of the native 

 soil. 



(4) Roots are removed in excavating the well, and constant 

 sampling within such a small radius tends to cut off new roots which 

 may penetrate this special soil zone. 



Whatever the objections to the method, it is the only method so 



far found practicable for repeated sampling of rocky, mountain soils. 



Considerable encouragement may be taken from the assumption, well 



founded on experimental fact, that when two soils, both moderately 



moist, are placed in contact, there is a strong tendency for water to 



move from one to the other imtil the capillary forces are in equilibrium. 



If, then, it may be assumed that the moisture equivalents of the 



native soil and well soil (ME and ME') represent points at which the 



two would be in capillary equilibrium, then the probable moisture 



of the native soil (M) mav be computed at any given time, from that 



If i ME 

 determined for the well soil (M'), by the formula — Jf=- — j^pj 



But also, if it is desired to relate the current moisture of the soil 

 to the moisture equivalent, on the assumption that this ratio ex- 

 presses in some degree the availabilitv of the water, then it is just 



M' M 

 as satisfactory to use ^v™ as jrpf because it has been assumed above 



that the two are equal. 



