444 



feet above the permanent water-table. That this small increase in the 

 water of this soil was not due to a small water capacity is indicated by 

 figures in a table showing the water content at two dates when it was 

 very much larger than at any of the times when the other observations 

 were made. 



A second example of slow capillary flow of water iuto comparatively dry field soil la 

 given as follows: On October 28, 1889, samples of soil were taken down to a depth of 

 5 feet from the summit of a rise of ground, and the amount of water content deter- 

 mined. Over this spot was then set a zinc-lined tray, ti by 6 feet, and 8 inches deep, 

 standing on legs 3 feet above the ground. Boards 8 feet long wore then placed, one 

 eud on the tray and the other on the ground, all the way around, and these were 

 covered with corn stalks, to shod water and prevent snow from drifting under the 

 tray. In this condition this section of ground remained uutil April 14, 1890, when 

 the covering was removed and samples of soil again taken, both under the tray and 

 just outside the covered area. [The results are stated in a table which shows the 

 per cent of water in the soil at the different depths.] 



When it is stated that the i)iece of ground under experiment was seeded out to 

 timothy and clover, three important facts are brought out. 



(1) The upper 4.5 feet, shielded from rain, sun, and winds, was drier after the lapse 

 of 168 days, comprising the winter season, than at the time of covering, showing that 

 the rate of cai>illary flow was insufScient to keep pace with the loss of water by evap- 

 oration, and possibly by lateral or downward trauslocatiou. 



(2) Even outside the shelter, with a precipitation of 58.54 pounds to the square 

 foot, the third foot was drier by au average of 0.79 per cent of its dry weight or 

 about 0.7 pounds, while the lower 18 inches showed only the small gain of 1.02 per 

 cent of the dry weight of that soil. 



(3) During dry seasons, like that of 1889, a sandy clay loam, supporting a crop of 

 timothy and clover an^ underlaid by 2 feet of gravelly red clay, followed by quartz 

 sand containing some gravel, may be so dried to a depth of 5 feet that the average 

 water content is only 6.42 per cent of the dry weight. 



Cases of slow downward movement. — On the morning of September 5, after a 

 rain-fall of 1.14 inches, or 5.9 pounds, to the square foot, which occurred during the 

 previous evening and night, it was found that in a corn field the rain had penetrated 

 only 3 to 4 inches and in clover ground not more than 3 inches. Twenty-four hours 

 later the water had reached 4.5 to 5.5 inches on the corn ground aud 4.5 inches on 

 the clover. On the morning of September 7, 62 hours from the beginning of the rain, 

 the water had penetrated only 5.5 to 6 inches in the corn ground and 5 inches on 

 the clover. 



Lateral capillary flow of water in surface soil (pp. 145-147). — The 

 experiments described in this article indicate that in the clay-loam soil 

 nsed the lateral movement of the soil water was slow, and extended 

 only about 3 feet. 



Capillary movement of loater in wet field soils (pp. 147-151). — This is 

 an account of the results of observations made " to determine the rate 

 of change in the water content of field soils following rains when the 

 soils were near their points of saturation. * * * On June 6, 1890, 

 32 hours after a rain of 3.16 inches, or 10.46 pounds, per square foot, 

 which was distributed over 4 days, samples of soil were taken from 

 eight different localities, and the amount of water they contained deter- 

 mined. Samples were again taken from the same localities 72 hours 

 later, aud their water content determined with the following results : 



