144 The Sugar-Beet in America 
of water or to the time of application. In the weekly 
irrigation tests, the beets that were not irrigated had the 
largest number of forked roots, whereas in the plats that 
had water applied at different periods, the plat receiving 
water only at the first stage had the least number of forked 
roots. The greatest number was on plats irrigated early 
and late. The differences, therefore, are not consistent and 
the idea that any method of irrigation greatly increases 
the tendency toward forkedness seems unwarranted. 
An examination of Figs. 20 and 21 for the effect of 
treatment on the height of tops reveals a rather close 
relationship between this and the yield of tops, but not 
the yield of beets, which has already been discussed in 
connection with Figs. 14 and 15. 
DRAINAGE 
Reasons for drainage. (Fig. 22.) 
Many millions of acres of land in the United States 
contain so much water that crops cannot be raised suc- 
cessfully. Part of this land is in permanent swamps; 
some of it is dry during a part of the year, being water- 
logged only at certain seasons. Much land, having a dry 
surface appearance, contains ground-water so near that 
roots cannot penetrate to any great depth. The chief 
difficulty in the way of successful agriculture in all such 
places is the surplus of water. The only way to make 
this land suitable for crops is to drain it. 
In most arid regions, much of the land contains a high 
percentage of soluble salts. These often accumulate near 
the surface in such large quantities that the growth of 
