22 BULLETIN 883, IT. S. DEPAKTMENT OF AGKICULTUEE. 
yielded that year 6 bushels more per acre than the June 1 seeding-. 
Fair results, however, may be expected from early June seeding; 
in fact, June seedings often yield more than those made during the 
last two weeks in May. The yield is largely dependent upon the 
occurrence or absence in August of hot, drying winds, which usually 
have a tendency to cause the premature ripening of flax sown late 
ha May. Evidence of this was seen in 1915, when a hot, dry spell 
was followed by a late, cool fall. The practice of seeding flax after 
Fig. 2. — View showing plats in date-of-seeding experiment with Damont (C. I. No. 3) flax at the Northern 
Great Plains Field Station, Mandan, N. Dak. The plat at the left was sown on May 15 and that at 
the right on June 1. Photographed July 17. 
June 10 to 15 is a questionable one. A marked decrease in yield 
can be expected, risks of damage by early fall frosts will be taken,, 
and flax diseases are known to develop most severely in hot soils. 
FLAX EXPERIMENTS IN NURSERY ROWS. 
SOURCE AND VARIABHJTY OF MATERIAL. 
In the plat experiments only those varieties were grown which had 
previously shown their adaptability to northwestern conditions. 
These, it has been explained, all conform more or less closely to one 
type. All are erect, sparsely branched tall varieties with blue flowers 
and a relatively small brown seed. Importations obtained through 
the Office of Foreign Seed and Plant Introduction during the winter 
of 1913-14 from widely separated countries showed an interesting 
range of types and characteristics. Many, while not adapted to 
northwestern conditions, have valuable characters. If these are 
