250 Journal of Agricultural Research voi. x. no. s 



OBSERVATIONS 



After the tanks were installed and filled, water was applied until a 

 stable level was reached according to the schedule in Table XVIII. 

 Tanks were weighed about twice each week, the platform scales weighing to 

 X pound. Observations were kept on the water level, a morning record 

 being made, starting at 8 o'clock, and in the afternoon beginning at 3.30, 

 for the greater part of the season. During particularly dry periods ob- 

 servations were started at 6.30 a. m., 12.30 p. m., and at 6.30 p. m. At 

 the time of each observation of the level, additions of water were made 

 in even pounds to bring it to the fixed elevation. The depth measure- 

 ments were made to % inch with an ordinary rule. Some difficulty was 

 experienced in keeping the tanks having the surface saturated in exactly 

 the right condition. After a couple of weeks' use these were provided 

 with individual supply reservoirs which allowed a small quantity of water 

 to drip into the tank. Very close attention was required. The entire 

 water losses from the tanks were determined by the weighings and 

 from the amount added to keep the level at the desired point. 



DETAILED DATA AND FINAL RESULTS 



The 14 tanks containing Cherry Creek sand were installed with a view 

 to using them as a so-called standard from which to determine the general 

 form of the curve of evaporation loss at different water levels. Five 

 points were possible: saturation, water at 3 inches, at 6 inches, at 12 

 inches, and at 2 feet. It happened that the final figures showed that 

 loX inches had been used instead of 12 inches. With the use of labo- 

 ratory soil, which is not truly the same type of material, six points were 

 found with the equipment, water at 4, 16, 28, 38, 43, and 51 inches. Of 

 the other types used it was thought that, since the form of the evapora- 

 tion curve could be learned from the so-called standards, two depths 

 would present the desired information. Accordingly, for sands i, 11, and 

 12, two depths, 3 and 12 inches, were used. 



The detailed results of all observations have been tabulated. Sand 10 

 is covered in Table XIX. During the period July 31 to October 16 the 

 final figures show the following results: At saturation the evaporation 

 was 77 per cent of that from evaporation tank 2, a water surface tank of 

 the same area with water 2.75 feet deep; 3 inches below the sand surface, 

 69 per cent; 6 inches, 64.5 per cent; io>^ inches, 57. 7 per cent; and at 24 

 inches, 11. 3 per cent. Table XX shows the data from laboratory soil; 

 Table XXI that from sand 12, Rio Grande; sand 11, Arkansas, is tabu- 

 ated as Table XXII; and No. i, Platte, is shown on Table XXIII. 



