22 T. Lyttleton Lyon and James A. Bizzell 



year ending May 31, 1909, the rainfall was 31.53 inches, of which 14.15 

 inches, or 44.9 per cent, percolated thru one of the 6-foot gauges, and 

 13.95 inches, or 44.2 per cent, thru the other. The drainage from the two 

 3-foot gauges amounted to 48.1 and 49.8 per cent of the rainfall, respectively. 



At the Texas Experiment Station, lysimeter experiments have been 

 conducted by Fraps (1914). The lysimeters consist of galvanized iron 

 cans 12 inches in diameter and 24 inches deep, and are buried in the ground. 

 Eight different soils were used. The total quantity of drainage for the 

 years 1911 and 1912, and for eleven months of 1913, is given, also the 

 rainfall for the same period. There is a very considerable range of per- 

 colation, varying in the unfilled and unplanted soil from 10.5 per cent 

 to 41.8 per cent of the rainfall, which for the period mentioned averaged 

 33.45 inches annually. The percolation from the four clay soils was 

 double that from the four sandy and sandy loam soils. 



This is not an exhaustive review of experiments on this subject, but 

 it doubtless summarizes most of the investigations dealing with this 

 phase of lysimeter experimentation. The attempt has been made to 

 obtain, so far as possible, the results from lysimeters having a considerable 

 depth and volume of soil, also of those experiments extending over a 

 considerable period of time. It is questionable whether very shallow 

 lysimeters give reliable results so far as the quantity of percolation is 

 concerned, but the 20-inch gauge at Rothamsted shows almost exactly 

 the same percolation during a period of thirty-five years as does the 

 60-inch gauge. On the other hand, at Cawnpore, India, the 3-foot gauge 

 gave a larger drainage than did the 6-foot gauge, indicating that in a 

 region of great evaporation there is a greater return of moisture to the 

 atmosphere from the deeper soil, due probably to upward capillary move- 

 ment during dry periods. 



Another factor that is presumably of importance is the duration of 

 the experiment, especially when the lysimeters cannot be weighed and 

 consequently when it cannot be known whether the soil-moisture content 

 is the same at the beginning and at the end of the experiment. Under 

 such circumstances it is safer to continue the experiment for a number 

 of years before calculating the percolation. 



In order that the results of the experiments conducted since those 

 reviewed by Wollny may be brought together, they are- presented in table 8 

 arranged in the manner followed by him: 



22 



