16 BULLETIN" 618, U. S. DEPARTMENT OF AGRICULTURE. 
Fargo, N". Dak. The evaporation, however, is much higher at Ama- 
rillo than at these stations in the northern part of the subhumid 
division of the Great Plains area. 
Results Obtained. 
The results of varietal experiments with durum wheat at 30 
different experiment stations are presented herein. At most of these 
stations the experiments have been conducted by the. Office of 
Cereal Investigations, either cooperatively with the State experi- 
ment station or some other agency or independently. At the remain- 
ing stations the data have been obtained independently by the 
State or Province. The source of the data is stated in connection 
with each station. 
The experimental conditions, such as crop sequence, size of plat, 
width of alleys, number of replications, etc., vary somewhat at 
different stations. The results obtained at one station, therefore, 
are not necessarily comparable with those obtained at any other 
station. In most cases, however, they probably are directly com- 
parable. The results from different varieties at the same station 
were nearly always obtained under similar conditions and may be 
compared directly. Any known exceptions to this fact are stated 
in the text. In all cases the crops are grown with only the natural 
rainfall and under conditions approximating good farm practice 
for the district. 
The value to crops of any given rainfall depends on the time and 
nature of its occurrence, on the kind of soil, and on the quantity 
and nature of the wind, as well as on other factors. Not one factor, 
but the total influence of all factors, must govern the classification 
of land for agricultural purposes. 
The results from the durum varieties under experiment are shown 
for each station. Usually all varieties that have been grown for 
more than one or two years are included. In the tables they are 
separated into the major groups given in the key to groups and 
varieties. 
The yields of the durum wheats are compared with those of the 
standard common wheats at each station. The highest yielding 
variety in each important group of common wheat is chosen for 
this comparison. Occasionally two varieties are given in one group, 
as in the Fife group, where such old standard varieties as Power, 
Glyndon, and Rysting are now being outyielded by Marquis. 
In figures 8, 12, and 13, the vertical column showing the yield of 
durum wheat is based on the yield of only the highest yielding 
variety in the period of years indicated, except when otherwise 
noted. 
