14 BULLETIN 1175, U. S. DEPARTMENT OF AGRICULTURE. 
EXPERIMENTAL METHODS. 
The main objects of the grain-sorghum experiments herein reported 
were to determine the relative value of different varieties, the best j 
time to sow the crop, and the best stand or spacing between rows and I 
between plants for grain production. These experiments were 
conducted in plats under conditions of good farm practice as to 
preparation of soil and cultivation. 
SIZE AND ARRANGEMENT OF PLATS. 
The land used in these experiments is divided into series, or blocks, 
which are 8 rods wide and extend the entire length of the field from j 
north to south. The series are separated by roads 20 feet wide. \ 
The rows extend across the series from east to west and are spaced j 
44 inches apart. Thus, each row occupies an area 132 feet long and I 
44 inches wide, or one-ninetieth of an acre. There are no alleys 
between plats, so there is no border effect except at the ends of the 
rows. The plats vary from 6 to 12 rows, depending on the nature 
of the experiment and the land available. The row, one-ninetieth 
of an acre, is considered the unit. In sowing the plats, the rows 
are extended several feet into the road at each end. When the 
plants have attained a height of 2 to 3 feet the ends of the rows are 
trimmed to the proper limits. All the rows are harvested ordinarily, 
but when adjacent plats differ materially in habit of growth or 
maturity the border rows of each are discarded. 
CROP ROTATION. 
It has been impracticable to follow a definite crop rotation in 
these experiments because of lack of suitable land. The practice 
has been to use the land for experimental plats for two years, fol- 
lowed by one year of uniform cropping. The object of the uniform 
field crop every third year is to leave the plat land in as uniform a 
condition as possible. * Until 1919 the uniform crop was cowpeas, 
but this crop was found to leave the soil so loose that it blew very 
readily, and during the last two years the crop has been milo. 
METHOD OF SEEDING. 
A 2-row corn drill fitted with sorghum plates was used for sowing 
the plats in these experiments. Seeding has always been done at 
a heavy rate to insure a stand even under unfavorable conditions. 
When the plants were from 4 to 10 inches high the plats were thinned 
by hand to the stands desired in the different experiments. 
METHODS OF OBTAINING DATA. 
The plant and stalk spaces and the occurrence of suckers and 
heads were obtained by actual counts of the plants, stalks, and heads 
in all the rows of each plat for which these data are given. The extent 
of suckering is indicated by the number of stalks per plant. The 
number of suckers per plant is obtained by dividing the number of 
stalks by the number of plants and subtracting 1 from this quotient. 
The percentage of headed stalks is the number of heads divided by 
the total number of stalks per unit. The percentage of erect heads 
in the milos is determined by dividing the number of erect heads by 
