18 
BULLETIN 1004, U. S. DEPAKTMENT OE AGEICULTURE. 
One fact evident at Amarillo is that the yield is correlated very 
little with the harvest date in the different years. This is probably 
because drought injury has been evident in all years. What little 
correlation exists in these curves is obscured by the fact that the 
date of the final sampling does not always coincide with that of 
harvest. For example, the wheat on plat A in 1913 was harvested 
July 1, but the first soil samples after harvest were taken on July 22. 
In general, the soil-moisture results at Amarillo are subject to 
greater experimental error than those for any other station. In sev- 
eral years the samples were not taken frequently enough to show ac- 
curately the water use for the different stages in the growth of the 
wheat crop. The curves are presented because spring wheat grows 
better at this station than at any of the other stations so far south. 
The high rate of water use in good years when water is available 
/s 
/s 
/S 
Fig. 7. — Diagi-am showing the use of water on wheat plat C or D at Amarillo, Tex., for 
the years from 1911 to 1916, inclusive. 
and the evident drought injury in all years are the points most evi- 
dent in the study of the results for the station. 
The results of any one year or any one station are not conclusive 
in themselves, but those of all the years studied clearly point to cer- 
tain conclusions in respect to water use. 
Four types or general classes of seasons in respect to the use of 
water by the crop may be distinguished : 
Seasons in which there is always sufficient water to meet the needs of the 
crop. This type of season is comparatively rare. 
Seasons when the supply of water in the soil and from rains is great enough 
to caiTy a crop almost to maturity, but in which a forced ripening of the 
wheat takes place because of water depletion just before harvest. This type 
of season is common. 
Seasons when drought conditions begin to affect the crop at or near heading 
time. In this case a forced ripening takes place, but there is in addition a wilt- 
ing or firing of the grain for some time before hanest. This is a type of season 
that occurs frequently on the Great Plains, 
Seasons in which the crop at no time has sufficient moisture for optimum 
growth, but in which by reason of its meager straw growth it may be enabled 
