16 
WATER REQUIREMENT OF PLANTS. 
this is not definitely stated. The pots were apparently weighed only 
at the beginning and at the end of the experiment. Although the 
dry matter increased progressively with the increased supply of 
water, the water requirement reached its minimum with the use of 
from 20 to 22 inches of water, increasing with either a greater or a 
smaller supply of water. 
Table VI. — Effect of the amount of water added on the water requirement of oats, at 
Bozeman, Mont., according to Fortier (1903, p. 107). 
Water added. 
Dry 
matter. 
Water re- 
quirement. 
Inches. 
2 
Grams. 
55 
76 
145 
235 
252 
264 
312 
316 
501 
574 
525 
434 
433 
413 
528 
552 
6 
14 
20 
22 
22 
32 
3S 
VON SEELHORST AND BUNGER's EXPERIMENTS. 
Von Seelhorst and Bunger (1907) conducted a fertilizer test with 
high and low soil-moisture content, using open pots containing 12 
kilograms of soil. This experiment was made on spring wheat. 
The results (Table XXXII, p. 44) indicate a higher water require- 
ment in soil of high moisture content. 
OHLMER'S EXPERIMENTS. 
Ohlmer (1908) also employed two different soil-moisture contents 
in connection with a fertilizer experiment with wheat at Gottingen 
(Table XXXIII, p. 45). The comparison was made with 11 different 
combinations of fertilizers, and in every instance the lower soil- 
moisture content gave the lower water requirement. The average 
of all the tests shows that a decrease in the soil-moisture content 
from 70 per cent of the water-holding capacity to 45 per cent decreases 
the mean water requirement by 12 per cent. Open pots were used. 
wimmer's experiments. 
Wimmer (1908) published the results of an extensive series of ex- 
periments to determine the effect of fertilizer on the water requirement 
of carrots and ray-grass. Open pots were used, the method being 
the same as that employed by Wilfarth and Whnmer (1902). 
The results (Table VII) show no significant differences that can be 
attributed to the different soil-moisture contents employed. In sev- 
eral of the series differences are found which might be considered 
significant if they were at all consistent. 
285 
As a whole, the data indi- 
