734 EXPERIMENT STATION RECORD. 



practiced the yield was 1534 lbs. greater iu 1895, the year of scanty 

 rainfall, than in 1S94. On the other hand, the subirrigated soil yielded 

 1,074 lbs. less iu 1895 than iu 1894, a result which is explained by the 

 fact that in this case the surface soil did not obtain sufiicient moistm^e 

 to give the maxiiuuui growth of com. 



The data reported show "that the yield of dry uuitter per acre was 

 more than one fifth, or 23.36 per ceut, larger on ground surface irri- 

 gated than it was ou that subirrigated. This difference of yield, iu the 

 writer's judgment, is due to the fact that a large part of the water 

 introduced into the subirrigated laud percolated below the level of 

 most effective service in the root zoue, while at the same time a large 

 part of the surface soil, where the vital processes are most active, was 

 left too dry." 



It Avas observed that when 15.03 acre-inches of water was pumiced 

 into the tiles only about 4 ft. of the surface soil immediately above the 

 Hues of tile was saturated, while midway between the tiles the soil 

 was wet only to within 9 in. of the ton. It is estimated that at least 

 5 acre-inches more of water would hav^e been necessary to completely 

 saturate the soil of the whole plat. It is evident, therefore, that by 

 applyiug equal amounts of water by the surface and subirrigation 

 methods at stated intervals the amount of moisture in the surface foot 

 of soil will average much larger in the former than in the latter case. 



"It follows from these observations that instead of recjuiring less water for sub- 

 irrigatiou than for surface irrigation, as has been generally assumed, it does reqni e 

 much more when applied at a depth of 18 iu. on lands where the ground water is 6 

 or more feet below the surface irrigated." 



It is possible that a greater economy of water would have l)een 

 realized if the tile had been placed nearer the surface of the soil. 



Irrigating through systems of tile drains. — As in the experiments of 

 the previous year (E. S. E., 8, p. 295), the outlet of the system of tile 

 drains was closed and water pumped into them in order to determine 

 the effect of raising the water level on the yields of clover, barley, and 

 corn. 



On May 24 and 25, 2.84 acre-inches of water was pumped into the 

 system. The irrigation was repeated June 24, 25; July 10, 11, 12, 22, 

 23, 24j and August 8, 9, 10, 21, 22, and 23, aggregating iu all about 

 17.04 acre-inches. Besides this, the clover plat was surface irrigated 

 June 17, July 12, aud August 19. 



Two crops of clover were cut on the irrigated area (3.2 acres), the 

 first yielding at the rate of about 4 tons per acre, the second about 1.8 

 tons. The third crop (on the 3.2 acres) furnished pasturage for 58 

 sheep for 31 days. 



The unirrigated soil yielded a single crop of about 1.5 tons per acre. 

 The increase due to irrigation on the total area was 5.5 tons of hay, 

 which, with the pasturage, is valued at -§39. The cost of production of 

 this increase is estimated at $18, leaving a net i)rolit of $21. 



