134 REPORT OF THE CHEMIST OF 



The action of a tile drain may be continuous or intermittent ; when con- 

 tinuous it is usually the discharge of some spring; when intermittent it 

 is the usual action of a drain in carrying off the excess of rainfall beyond 

 what the soil can hold as capillary water, or else the redundant accumula- 

 tion of soil-water by the indirect action of frost. How long will it take 

 water to percolate through three feet of soil to reach an ordinary tile drain? 

 The time required will depend upon the kind and porosity of the soil, and 

 upon the amount of water already in the soil. If the soil is saturated with 

 water, the addition of water at the top will cause almost immediate flow at 

 the bottom, but none will tiow till the soil contains all the water it can hold 

 by capillary action. 



Not having at hand the desired data in regard to the passage of water 

 through soil, I made some investigations on this subject, and was fortunate 

 to have the means at hand to verify certain points by means of a large bal- 

 ance made for the Experiment Station by Becker Brothers, of New York. 

 By actual trial, I find that when this balance is loaded with 5,000 grammes 

 in each scale-pan, the balance will turn with one-tenth of a milligramme. 

 A balance that will turn with one fifty-millionth of its load is sufficiently 

 delicate for such investigations. 



A number of tin tubes 2 inches in diameter and 20 inches long, part of 

 them closed water-tight at one end, and others closed at one end by perfo- 

 rated tin, were used for the experiment. The tubes were carefully weighed 

 and marked ; then filled with sifted air-dry soil, and again weighed ; then 

 placed in a vertical position and water added till water began to flow from 

 the perforated end of the tube, when the time was noted that was required 

 for percolation, and when the flood ceased the tubes were again weighed. 

 This gave data for the weight of dry soil, the time for percolation of water 

 through 18 inches of dry soil, and the weight of water taken up by dry soil 

 and held by capillary attraction. The tubes with closed ends gave similar 

 data, except that the time of percolation could not be so accurately deter- 

 mined. 



Two kinds of soil were used, a garden soil of sandy loam, and road dust, 

 with considerable clay. The tubes were used in duplicate with both kinds 

 of soil. The weight of the dry garden soil as compared with water, volume 

 for volume, was 148 to 100; that of the road dust was 173 to 100. 



In the perforated tubes with garden soil (air dry), the vertical column of 

 water required to thoroughly wet 36 inches of soil was 12£ inches, and the 

 time required was 20 hours, and the water retained in the capillary condition 

 was 23.47 per cent of the weight of the soil. With the road dust the time 

 required to thoroughly wet the road dust was 24 hours, and the water held 

 by capillary action' was 18.25 per cent of dry soil. It was difficult to 

 determine the exact time when the water reached the bottom of the closed 

 tubes, but as far as could be discovered the time was about two hours longer 

 than in the perforated tube, the air enclosed in the tube escaping only by 

 bubbling up through the water, while in the perforated tube it had free 

 escape at the bottom. That this soil in closed tubes held free water soon 

 after it had all sunk away from the surface was shown by the fact that the 

 percentage of water was increased from 23.47 per cent to 26 per cent. 



