CEREAL EXPERIMENTS IN MARYLAND AND VIRGINIA. 7 
is available. If a variety fails to give evidence of superior merit or 
adaptability after repeated trials in nursery rows, it is usually dis- 
carded, although sometimes it is retained for some character that may 
be of value for breeding purposes. 
Many hundreds of varieties and strains from various sources have 
been tested in the nursery. These have included importations from 
many foreign countries, new and untried sorts, and numerous local 
varieties. After several trials, in nursery rows most of them have 
failed to warrant more extensive tests. Only a small proportion has 
been grown in the plats. 
The plat tests at College Park have included 107 varieties and 
strains of winter wheat, 13 of winter oats, 12 of winter barley, and 
several each of winter spelt and emmer. At Arlington Farm they 
have included 43 varieties and strains of winter wheat, 19 of winter 
oats, 56 of winter barley, 12 of winter rye, and 6 of winter spelt and 
emmer. A rate-of-seeding test with wheat has also been conducted 
at Arlington. With some of these cereals the varietal trials have not 
been of sufficient duration to be conclusive. With others some very 
reliable results have been obtained. In general, those varieties that 
were discarded, more especially those that were grown only two or 
three years, were mixtures of some of the varieties already in the 
tests or were decidedly inferior in some character or characters. 
During the period of crop growth, rather complete notes are taken 
on the different varieties. The date of seeding, rate of seeding, date 
of emergence, and stand usually constitute the data recorded in the 
autumn. In the spring the winter survival (estimated) , dates of head- 
ing, ripening, and harvesting, the percentage of smut and rust infec- 
tion and of lodging, and the average height of plants are recorded. 
At harvest time samples of the heads and of the whole plants are taken 
and stored for reference and study. 
A small separator, especially designed for the purpose, is used in 
thrashing. The total weight of the crop is obtained just before 
thrashing, and the weight of the thrashed grain immediately there- 
after. After thrashing, the grain is fanned and the weight per bushel 
determined. The yield per acre in bushels is figured at the standard 
weight per bushel for each crop. 
CONDITIONS OF THE PLAT EXPERIMENTS. 
DIMENSIONS. 
At College Park the varietal tests have been conducted almost 
exclusively on twentieth-acre plats (10i feet by 208 feet). They lie 
in series running north and south, with 18-inch alleys between the 
plats in the series. The roadways dividing the series (or fields, as 
they are called at the Maryland station) are 16^ feet in width. 
