EXPERIMENTS WITH FLAX ON BREAKING. 13 
partment of the North Dakota Agricultural Experiment Station. 
Frontier (C. I. No. 17) is a variety which that station had distributed 
under the name Russian (N. Dak. No. 155), more recently named 
Frontier. C. I. Nos. 1 and 3 are selections from this variety. C. I. 
Nos. 2, 16, and 18 have originated from locally grown varieties and 
C. I. Nos. 4 and 5 were direct importations by the United States De- 
partment of Agriculture from Russia in 1903 as Seed and Plant Intro- 
duction Nos. 9906 and 9937, respectively. 
Of the three North Dakota resistant varieties, C. I. No. 13 was 
seed obtained direct from Prof. H. L. Bolley and produced on his most 
diseased plat in 1913. C. I. Nos. 8 and 14 were not available from 
this source, and were obtained from farmers whose seed was recom- 
mended by the botany department of the North Dakota Agricultural 
College. 
C. I. No. 15 was a commercial variety, obtained during the winter 
of 1913 from a farmer in Idaho. 
All of the 14 varieties are of the blue-flowered type grown in the 
Northwest. Most of them can be classified as true seed-flax types. 
C. I. Nos. 12 and 13, however, have a more slender stem and com- 
pact panicle, and have previously been referred to as short-fiber 
flaxes. 1 C. I. Nos. 4, 14, and 15 are more or less intermediate be- 
tween the short-fiber and true seed-flax types, in certain seasons 
resembling one and in other seasons the other group. C. I. No. 15 
was a mixture so lacking in uniformity that it was discarded from the 
experiments at the end of the third season. 
AGRONOMIC RESULTS. 
Table IV presents the annual and average yields of the flax varie- 
ties that were grown in plats, in the order of their highest average 
yield, together with the annual and average probable errors. The 
annual probable errors were computed from the formula 
0.8453 
n-s/n — 1 
and the probable error of the average was obtained by the formula 
R=± 3 , 
■> 
when R equals the probable error of the average, and a 1 h, c the prob- 
able errors of the three results. 
While slight variations in yield are noted between varieties from 
year to year, the relative position of the varieties was the same in 
both the 2-year and the 3-year averages. This order classifies them 
automatically, ranging from highest to lowest yielders, from the most 
pronounced seed types through the intermediate varieties to the 
i Clark, C n. Seed-flax production, U. S. Dept. Agr., Farmers' Bui. 785, 20 p., illus. 1917. 
