CEREAL EXPERIMENTS AT DICKINSON, N. DAK. 9 
Table IV shows that the highest average mean, maximum, and 
minimum temperatures have been recorded in July, and that in this 
month alone frost has not occurred. This table also shows that the 
average mean temperature for the different seasons varied but little 
from the 7-year average of 57° F. 
Table V gives the dates of the last spring and first fall frosts and 
the number of days in the frost-free period during each year from 
1907 to 1913, inclusive. The latest date on which frost has occurred 
in spring during the 7 years was June 5, in 1910 and 1912; the earliest 
frost in the fall during this period was on August 19, 1907. The 
average frost-free period for the 7 years is 99 days. 
TaBLe V.—Dates of: the last killing frosts in the spring and the first killing 
frosts in the fall at the Dickinson substation for the years 1907 to 1913, in- 
clusive; also the temperatures recorded and the length of the frost-free period 
for each year. 
[Data from the records of the Biophysical Laboratory of the Bureau of Plant Industry. } 
Last frost in spring.| First frost in fall. 
eas Frost- 
Year. | free 
Temper- Temper- | period. 
Dave. ature. Date: ature. 
Scale oh eke Days. 
LG) prec rere ena i Me ahaa Be eas we eA! June 4 30 | Aug. 19 34 76 
TO S'S 5 SiS eG Ses Ae ee ees a a May 21 29 | Aug. 21 27 92 
IO) nes cadets oS 55s Shim tare Ee ee re re ree Re nee May 17 32 | Sept. 14 34 120 
OW eae oe BeOS SOE Sac pe eee tee iets oan a anes June 5 29 | Sept. 8 28 95 
TIL es ies BG SiS ise ae eo haa Ae Sm aS eed May 27 27| Aug. 27 28 92 
NOI ean GaSe aaa ore SaaS a eae See ea June 5 32 | Sept. 21 32 108 
OD. Bas Se wicks Foes ce Peo Seen tea ee a a May 21 32 | Sept. 10 30 112 
INSISTS oo se sao 4 soe ge Or ee ee eee See INGE Vine 2 Sialeee eee SI] Dis 45. |Seece cues 99 
NATURE OF THE WORK. 
In all the varietal tests at Dickinson the primary object has been 
to determine the relative yielding power of the different varieties 
and to discover kinds better adapted than those ordinarily grown. 
Efforts have also been made to improve the yield of the leading 
varieties by selection and to find reasons for the existing variation 
in yields between different groups of varieties of the same cereal. In 
order to make these studies complete, it was necessary to divide the 
work into plat and nursery experiments. The plat experiments were 
designed to conform as nearly as possible to the best farm prac- 
tices in the northern Great Plains region. On these plats the crops 
were studied on an extensive scale. In the nursery tests, which were 
confined to short rows, the crops were studied intensively. The rows 
were uniformly spaced and a weighed quantity of seed was planted 
in each row. This method made possible the testing of a large num- 
ber of varieties and offered a means of making a study of a consider- 
able number of plants of each variety. 
62800°—Bull. 33142 
