74 Nebraska Agricultural Exp. Station, Research Bui. IS 
of 100 seeds of wheat to one of corn is normal in planting 
under field conditions in this region. Accordingly in com- 
paring corn and wheat in potometers they were planted re- 
spectively at the rates of one plant and 100 plants per pot. 
Under these conditions the respective water requirements 
for grain production of the corn and wheat were 743 ±48 
and 1017 ±60. However, when the corn was grown at the 
rate of six plants per potometer these relative water require- 
ments were 3481±389 and 1017±60. 
Applying the general rule of ''three times the probable 
error," we may be fairly confident from the one comparison 
that Hogue's Yellow Dent corn uses considerably less water 
than Turkey Red winter wheat, and from the other compari- 
son we may be equally confident that corn uses more than 
double the amount of water for grain production than the 
wheat. 
In the first comparison the degree of cropping for this 
quantity of soil corresponded well with normal field conditions 
for each crop. In the second test, however, the corn was 
planted relatively much too thick, and for this reason the 
ratio of grain to vegetative growth was greatly reduced. As 
a result the water requirement for grain production was in- 
creased. 
EFFECT OF CHANGE IN METHODS ON A(iRONOMIC EQUIPMENT 
Replacing the single-row nursery test plat planted in du- 
plicate with five-row test plats replicated 10 times increases 
the land requirement 25 times for such nursery testing. In 
testing hoed crops the substitution of three-row plats, repli- 
cated five times, for single duplicated rows requires 15 rows 
rather than two rows. The replication of small grain field 
plats five times, rather than twice, greatly increases the land 
requirement. 
Fertilizer and tillage experiments which frequently are 
conducted in unduplicated plats should probably be at least 
triplicated. Reduction of error by replication is more effec- 
tive than the use of check plats alone. 
The introduction of check plats every fifth plat in itself 
occupies one-fifth of the land. The more refined methods of 
securing comparable stands of corn upon which to base the 
yields at harvest require much greater labor expenditure than 
formerly. 
