10 BULLETIN 1092, U. S. DEPARTMENT OF AGRICULTURE. 
of 90 per cent. It is assumed in all cases that the check row re- 
sponds to the soil conditions in the same way as the other selec- 
tions. If the soil is poor the growth of both will be poor, and if 
the soil is rich the growth of both will be good. If these two 
strains A and B, grown in different parts of the field, were to 
be compared directly as to the actual straw weights, the poorer one 
might, because of better soil conditions, be apparently the better 
straw yielder. The use of the check prevents such a mistake, as 
the check rows measure the soil conditions and when percentages 
are calculated give the true relation between these selections, which 
is as 75 to 90 on straw weight. 
The percentages assigned to the different selections can be re- 
duced to straw weights as follows: Since Selection A yields 75 per 
cent as much as the check row alongside of it, we assume that it 
would yield 75 per cent as much as the average of all the check rows 
sown throughout the field. If the average yield in straw per acre 
of all the check rows is 2 tons, then, because Selection A is three- 
fourths as good as the check nearest to it, it is assumed that it would 
yield three-fourths as much if sown in all places that the checks 
were sown, or u| tons per acre. This gives what may be termed the 
corrected straw yield, as by means of the ciiecks correction for soil 
variations has been made. In the same manner, corrected seed 
weight, corrected stem length, and corrected values for any num- 
ber of other characters may be found. 
USING THE SCORE CARD FOR THE ELIMINATION OF THE POORER 
SELECTIONS. 
Since all the characters are not of equal importance, a score card 
has been devised to record the proper values for them, so that the 
sum total of the good and the bad points of each selection may be 
expressed in one figure. The following characters have been con- 
sidered as sufficiently important to be used in the score card: 
Weight of thrashed straw, weight of seed, length of stem to the 
first branch, resistance to lodging, resistance to disease, and strength 
of fiber. When available the weight of fiber will take the place of 
the weight of the thrashed straw. 
Resistance to lodging has been given a value of 5 per cent, be- 
cause weather that would cause lodging might be expected about 
1 year out of 10, and since the flax that is resistant to lodging 
could only be expected to stand up about half of these times, it 
would have an advantage of about 5 per cent over the other strains. 
Resistance to disease has been given a value of 4 per cent, because 
loss from diseases has been a less serious factor with fiber flax than 
loss due to lodging. 
