WATER LOSSES FROM WET AREAS 43 



surface and the evaporation test was continued. There was immediate 

 response in loss of water from tlie supply tanks, but indications were 

 that the large initial losses from the JMariotte tank were partly absorbed 

 by the dry soil as capillary moisture, and that only a small part was 

 lost by soil evaporation. This adjustment of soil moisture continued 

 for a period of four to six weeks until the capillary demand was satis- 

 fied, after which a small but rather uniform rate of evaporation con- 

 tinued. 



The evaporation test with water tables at a 3-foot level continued 

 for the following 12-month period to October, 1930, at which time a 

 number of changes in depth to water tables in some tanks were made 

 and all soil evaporation tanks were transplanted to salt grass. It was 

 rather hard to get this started and in some cases light surface irriga- 

 tions were applied as the tank surfaces were generally quite dry. 

 Moreover, the time of year was not the best for starting new growth 

 and the grass was slow in developing a root system. Consequently the 

 recorded use of water during the winter or dormant season by those 

 tanks in which grass was newly transplanted was almost entirely 

 caused by evaporation from the soil surface rather than l)y consumptive 

 use. The increase in use of water beginning in jNIarch, foUoAving trans- 

 planting, shows definitely that this was the end of the dormant period 

 for the salt grass in this set of tanks. 



The second set of tanks, Nos. 4, 5, and 6, containing undisturbed 

 soil, were first used for measurement of evaporation from the soil 

 surface with the water table at a depth of 2 feet. As this depth was 

 well within the limit of capillary rise, evaporation began immediately 

 after the soil received water and continued until October, 1930, when 

 the soil evaporation tests were completed. The first two months of 

 record in 1929 showed a high rate of loss from the iMariotte supply 

 tank, which may be accounted for as ad.justment and increase of 

 moisture held in the soil following establishment of a fixed water table. 

 Such losses were observed in every case where the water table was 

 raised to a higher level. 



Immediately following completion of the evaporation test, the tanks 

 in this set were transplanted to salt grass, the water table remaining 

 unchanged. As with the first set of tanks, the grass was slow to start 

 and some early surface irrigation was necessar}^ Consequently, the 

 water used during the following winter when the grass was in the dor- 

 mant stage was almost entirely soil evaporation. Increase in growth 

 began in the following February and these tanks were soon covered 

 with a luxuriant growth of grass which completely shaded their soil 

 surfaces. Kecords were made until April 30, 1932, when the investi- 

 gation was completed. 



The third set, including Tanks Nos. 7, 8, and 9, was first operated 

 with a 2-foot water table, but with salt grass sod instead of bare sur- 

 faces. Tank No. 9 was the only one to have an original crop of grass at 

 the outset. In this tank the grass root system was fully developed 

 and remained undisturbed, exceept where the shell of the tank cut off 

 lateral roots as it was forced into the ground. The other two tanks 

 were bare, and it was necessary to transplant grass and tlevelop root 

 systems before the maximum use of water was attained. The trans- 

 planted grass showed a very heavy and healthy growth of 6 inches or 



