Aug. 20, 1917 



Relation of Soil Water Movement 



395 



the moistest (IV), then in III, nearly saturated, although least in the 

 oven-dried material. It is probable that in III the initial moisture con- 

 tent approached but did not equal the hygroscopic coefficient, moisture 

 from the saturated atmosphere being absorbed very slowly unless the 

 exposed layer of soil is very thin; further, the ratio of hygroscopic 

 moisture to water content is much lower than is to be expected from the 

 coefficient (i, p. 353-356). 



Table II. — Rise of water {in centimeters) in two soils in different states of moistness 

 showing influence of the moistness upon the rate of rise. Data of Wollny 



Time. 



Experiment I (loam powder). 



I 



(dried 



at 

 ioo°C.). 



II 



(air- 

 dry). 



Ill 



(approxi- 

 mately 

 at the 

 hygro- 

 scopic 

 coeffi- 

 cient). 



ly 



(mixed 

 with 

 water) 



V 



(mixed 

 with 

 water). 



Experiment II (calcareous sand, 

 pulverized). 



I 



(dried 



at 

 100° C). 



II 



(air- 

 dry). 



Ill 



(approxi- 

 mately 

 at the 

 hygro- 

 scopic 

 coeffi- 

 cient). 



IV 

 (mixed 

 with 

 water). 



Initial moisture content, 

 per cent 



3-83 



7.96 



8.53 



Hours. 



2K. 

 sH. 



24.. 



48.. 

 72.. 

 96.. 

 120. 



6.0 



8-3 



6.8 

 9-7 

 12. 9 



6.9 

 9.8 



12.8 



14.0 

 17.0 



IS- 6 

 19.4 

 23.8 



10.4 

 23-9 

 30.8 

 39-8 

 52-0 

 60.6 



15-6 

 .S6.s 

 51-4 

 60. 7 

 69. 2 

 76. 2 



IS- 5 

 36.7 

 SI- 6 

 60.9 

 69.6 

 76.7 



25.1 



S2-0 



66.5 



76. s 



83-4 

 90.7 



28.0 



54-7 

 68.5 

 77-3 

 S4-S 

 91. 6 



II. 6 

 14.7 



II. 4 

 14- S 

 17.8 

 20.5 



13-3 

 16.3 

 19-5 

 22.0 



13-4 

 16. S 

 19.8 

 22.3 



30-9 

 40. 7 

 47-4 

 56.0 



41.4 



S2. O 



58.0 



6s- 4 



40. 6 

 48. 6 

 52-9 

 sS.o 



41.8 

 Si.o 



Krakow (10, p. 210) determined the influence of the initial moisture 

 content upon the rate of rise, using a sand in four different states of 

 moistness — viz, with 0.21, 0.51, 1.18, and 2.39 per cent of water. The 

 first was the air-dry sand. From our studies (i, p. 353-356) it would 

 appear that we might estimate the hygroscopic coefficient of this as 

 about 0.5 or 0.6, similar to the dune sand O described in the experimental 

 part of this paper. Krakow, continuing his observations for six days, 

 found that after the fourth minute the rate of rise decreased with the 

 moistness of the soil, the heights at the end of the sixth day being, 

 respectively, 67.1, 42.4, 41. i, and 38.1 cm. The deviation of Krakow's 

 findings from Wollny's conclusions Ramann (15, p. 334) attributes to 

 increased friction due to inclosed bubbles of air. 



Briggs and Lapham (5, p. 26), experimenting with a sandy soil whose 

 hygroscopic coefficient computed (6, p. 69) from the mechanical analysis 

 would be 2.81, concluded that when this was very moist the capillary rise 

 was over four times as great as when the soil was dry. From experiments 

 made by Mr. J. B. Stewart with three sands whose computed hydro- 

 scopic coefficients (6, p. 69) would be 0.7, 0.9, and 1.5, the same authors 

 concluded that — 



no constant ratio exists between the capillary limits of soils in a dry and in a moist 

 condition (5, p. 27). 



