TESTS OF BARLEY VARIETIES IIT AMERICA 15 
METHODS OF COMPILATION 
The aim in this bulletin is to present the yields of barley varieties 
grown in field experiments in the United States and Canada. The 
Canadian yields were included because they have as much value for 
many American farmers as those of our own stations. Yields are 
reported of all varieties which are regarded as significant or which can 
be identified, at least as to group. A few varieties have been in- 
cluded without definite information even as to their group characters, 
but only wdien there was some hope of later identification or some 
particular significance to their yield. The number of varietal yields 
discarded because of lack of information has been very small, and such 
varieties usually have been tested but few years with mediocre results. 
In many cases the individual yields of synonymous varieties are 
not reported. Unpurified Manchuria barleys with various station 
names and numbers are averaged under C. I. No. 244 when these 
trace directly to the Agricultural College of Guelph, Ontario. Like- 
wise, those which trace directly to the origmal introductions of the 
Wisconsin Agricultural Experiment Station are averaged under 
C. I. No. 241. Averages have been made of commercial sorts only. 
Whenever a selection of any variety has been made, for example 
Manchuria, this selection has been given a number and retains its 
identity in all computations. 
With the hybrid variety Horsford the averaging was perhaps more 
arbitrary, as the confusion was greater. Horsford (C. I. No. 507) 
here represents the hybrid type obtained by crossing Nepal with 
barleys of the Manchuria group. The resulting varieties are homo- 
zygous for only a few characters at best. While barleys of more 
than one mating may be listed under this number, they are of the 
same general type, and the procedure probably is justified. Not 
all varieties arising from such crosses are included under C. L No. 
507. In some cases the records of ancestry were clear enough to 
determine the history of the origin. Even in these cases the purpose 
of this bulletin would have been served by including these varieties 
also under C. I. No. 507. The greater number of the yields are 
presented without combining varieties or numbers. At most of the 
stations few, if any, averages were made. 
For points where the data were sufficiently extensive, three tabular 
compilations have been made, although only the first and third are 
presented. The first of these shows the yields of the varieties by 
years. The varieties are listed in the chronological order of their 
first appearance in the experiments. This compilation embodies 
statements of the number of years grown; the average yield, which 
means little; and the average yield of each variety in terms of a per- 
centage of the average of all varieties for the years grown. Suppose, 
for example, that Svanhals barley was grown during the years 1909, 
1911, 1912, and 1913. The yields of all varieties for those years 
are then added together and divided by the number of yields. The 
average of the yield of Svanhals for these years is then divided by the 
average of all varieties for those same years; that is, the yield of 
Svanhals is computed as a percentage of the weighted mean yield of 
all varieties. Any basis of computing is open to objection. A com- 
parison of the yearly average yield of Svanhals with the yearly aver- 
age of all varieties gives undue emphasis to unusual yields in years 
