814 
Journal of Agricultural Research 
Vol. XXI, No. II 
ANALYSIS OF TABLE III 
It is evident from a casual inspection of Table III that there is a rather 
striking similarity between the two comparison columns, A and B, in the 
several cases. 
Using either A or B as the standard, the deviations added with respect 
to sign give a sum equal to 5.11 culms for the entire number of counting 
periods. Reduced to the mean deyiation for each counting period it is 
less than unity, an inconsequential amount. However, the size of the 
deviations must be considered; so that if they are added irrespective of 
sign and the mean deviation per group determined for each counting 
period, one finds the mean to be 2.92 culms. This mean deviation is 
about 6 per cent of the mean number of culms for each counting period. 
Considered from either of the foregoing standpoints, it appears that the 
agreement between the rate of culm formation of the real brome plants 
and the ideal rate of formation of the conventional plants is reasonably 
close. It is not to be understood that the plants in the field really 
follow the ideal method as outlined. This would be putting the case 
rather too strongly. It is known that the actual and conventional 
plants started with the same number of culms and that culm formation 
proceeds centrifugally in the actual plant as it does by hypothesis in the 
conventional plants. Rinding the results in the two cases so closely 
in accord, one may say without much hesitation that given a certain 
number of groups of plants of Bromus inermis possessed of varying 
initial culmage, one may expect, with normal development, a continual 
increase in number of culms for each group but at the same time with a 
continual increasing diversity in number of culms between the various 
groups according to fixed rates. 
While the correspondence of group A and group B, taking them as a 
whole, can be considered only general in character, it apparently is 
sufficiently close to make it seem unwise with our present knowledge to 
make other attempts to explain the increasing diversity in number of 
culms between the various groups with the advance of age of plants. 
It should be observed that the number of plants in the group having 
four initial culms was somewhat lower than in the other groups. The 
tagging effect exhibited by this group may be attributed perhaps to the 
fact that the plants of the group did not comprise so representative a 
sample as did the plants of the other groups and that more than their 
due proportion tagged behind, thus diminishing the true mean. As a 
matter of fact there were a number of plants which behaved, after the 
first one or two counts, as though belonging genotypically to one of the 
groups having a smaller initial culmage. However, another explana¬ 
tion is possible. It may be that the effect of transplanting was more 
detrimental to the 4-culm plants, which were farther advanced, than to 
the plants not so far developed. Such an effect would be a retarding 
one and likely to be confused with constitutional difference. 
