8 REPOKT ON THE KLAMATH MARSH EXPERIMENT. 



Early in the season one of the three parallel open ditches was 

 blocked up at either end and kept filled with irrigation water, 

 while the adjacent ditch 200 feet away was kept empty. This con- 

 dition was maintained with but few interruptions throughout the 

 season from May to October, with the object of leaching out the 

 excess salts in the land between these two ditches. 



Water samples were collected each week from the open wells in 

 this tract, and the salt content was determined with the electrolytic 

 bridge, check samples being collected from time to time for chemical 

 analysis. 



At the end of the season samples of water collected at depths of 

 2, 4, and 6 feet at various points in the subirrigated tract showed 

 conclusively that there had been no material leaching of the salt 

 except in the upper foot. 



In the latter part of the season an effort was made to leach out the 

 salt by protracted surface flooding of the land from which the furrow 

 slice had been removed. Borders were thrown around two checks, 

 each containing one-eighth of an acre, and each of these two checks 

 was bordered on two sides by an open ditch nearly 5 feet deep. One 

 of these checks was kept continuously flooded for two weeks and then 

 allowed to dry out for two weeks, while the other was flooded. This 

 flooding process was repeated on each check so that each was flooded 

 twice for two weeks at a time with an intervening period of two 

 weeks without flooding. Samples of the subsurface water were 

 secured at depths of 2, 4, and 6 feet in each of the checks after each 

 period of flooding and, although an aggregate of 5 feet in depth of 

 water was applied to each of these checks during two weeks of flood- 

 ing, it was found that no appreciable lessening of the salt content had 

 occurred at the depth of 2 feet or below. All of the water applied 

 to these checks that had not evaporated had apparently seeped out 

 laterally through the more porous surface layer. 



In view of the very light texture of the tule mat and underlying 

 material, none of which weighs over 40 pounds per cubic foot when 

 dry, it seems difficult to understand the cause of its relatively high 

 impenetrability. When saturated this material holds a large pro- 

 portion of water, it being not uncommon to find samples containing 

 5 or G parts of water to 1 part of dry material. In other words, 

 when approaching the saturation point this soil contains 500 or (JOO 

 per cent of water. Chemical analyses of this subsurface water and 

 of the tule mat itself show the presence of large quantities of soluble 

 salts, chiefly carbonates of sodium, calcium, and magnesium; in other 

 words, lime and black alkali. The tule mat when moist gives a pro- 

 nounced alkaline reaction to litmus, and the subsurface water on 

 boiling gives an alkaline reaction to phenolphthalein. In this re- 



[Cir. 86] 



