50 BULLETIN 1155, U. S. DEPARTMENT OF AGRICULTURE. 
the depth of water produces variable results. On high ground that 
can not be continuously submerged there is usually a great deal of 
grass. On low land submerged too deeply the stand of rice is often 
poor, but the grass is usually killed. 
Early, deep, continuous submergence of rice controls to a marked 
extent the early red barnyard grass, which is the most abundant 
variety in California, and some of the other forms, but the late white 
forms do not seem to be controlled by it. These late types are 
usually present on submerged fields without regard to the manner 
in which the water is handled. The grass locally known as white 
water grass survives under such treatment and appears to increase 
from year to year. The levees are always covered with a rank 
growth of water grass on foul fields. 
Notwithstanding these difficulties, deep continuous submergence 
is quite effective in the control of water grasses and enables the 
growing of profitable crops on land which is too foul for rice produc- 
tion under the usual irrigation methods. On clean land it is probably 
best to follow the old system of irrigation, as unfavorable tempera- 
tures and deep water may cause poor stands. On the other hand, con- 
tinuous submergence of clean land may prevent it from becoming foul. 
During the period of high prices for rough rice it was possible and 
profitable in many cases to produce rice on foul land by continuous 
submergence, but in normal times at normal prices this method may 
or may not give profitable results. This matter is now under inves- 
tigation. It is a rather uncertain method of production, because 
(1) it is impossible to maintain a uniform depth of water on each 
check on account of variations in the land surface, (2) water can not 
always be obtained from the canal when wanted and needed, (3) if 
submerged too early poor germination and stands are often ob- 
tained, and (4) if sown and submerged too late in the spring poor 
stands are often obtained. Deep continuous submergence is most 
effective in the control of early barnyard grass, but late-maturing 
water-grass types are more resistant to control by deep submergence. 
EXPERIMENTS IN WATER-GRASS CONTROL. 
In 1921, experiments to control water grass by irrigation were 
conducted at the Biggs Rice Field Station. The seed-bed prepara- 
tion, method of seeding, and method of irrigation on the different 
plats were as follows: 
On plat 1 rice was sown broadcast at the rate of 150 pounds per 
acre on spring-plowed rice stubble on which a reasonably <*ood seed 
bed had been prepared by disking and dragging. The land was 
harrowed after sowing. Plat 2 was given the same seed-bed prep- 
aration as plat 1, but the rice was drilled at the rate of 150 pounds 
Eer acre. Plat 3 was prepared and seeded like plat 1, but was not 
arrowed after sowing. Plats 1, 2, and 3 were sown May 9 and 
irrigated and drained twice before holding the water continuously. 
The rice and barnyard grass had emerged May 30, and at the third 
irrigation the plats were submerged about 6 inches deep. Thereafter 
the water was held 6 inches deep continuously until the rice was 
ready to drain for harvest. 
On plat 4 rice was sown broadcast at the rate of 150 pounds per 
acre on May 16 on spring-plowed rice stubble with a rough seed bed, 
then submerged 2 inches deep on May 21. The water was held at 
