156 BOTANICAL GAZETTE [SEPTEMBER 
with the total weight in the control plants irrespective of the 
number of primordial leaves formed by the individual plants of the 
tetracotyledonous race. The total number of leaves per plant, 
however, was determined in these four series.© Thus it is possible 
to give the average weights both per plant and per leaf in the two 
series. The results show that in three of the four cases the green 
weight as given in table I of the approximately four primordial 
leaves of the tetracotyledonous race is lower than that of the two 
_ primordial leaves of the dicotyledonous strain. The percentage 
differences in total weight range from +2.1 to —16.4, with a 
general average of —7.37. When the comparison is made on the 
basis of mean weight per leaf, the primordial leaf of the abnormal 
seedling is found to be on the average 54.22 per cent lighter than 
the leaf of the normal seedling. 
For dry weight, given in table II, all four series show lower 
average weight in the tetracotyledonous strain. The percentage 
differences for dry weight of primordial leaves per plant vary from 
—1.6 to —18.0, with a general average of —10.90. On the basis 
of mean dry weight per leaf, the weight for tetracotyledonous plants 
is found to be from 49.6 to 59.9 per cent lower than that of the 
normal seedling, with a general average percentage difference 
of —55.92. Thus the results for these four samples clearly indicate 
that an abnormal race shows the same relationship to the normal 
parental race as do abnormal individual seedlings to the normal 
seedlings in the same race. 
PLANTS CLASSIFIED WITH RESPECT TO NUMBER OF PRIMORDIAL 
LEAVES.—Upon the completion of this preliminary comparison it 
seemed worth while to analyze the relationships more minutely 
by considering individually the results for seedlings of the tetra- 
cotyledonous race with varying numbers of primordial leaves. 
These results were only attained at the cost of great labor, since 
it was difficult to secure considerable numbers of seedlings of any 
given type simultaneously. It was necessary, therefore, to make 
determinations for abnormal and control plants in small sub- 
samples, and to combine these to form samples of 100 seedlings 
6 The average numbers per plant were as follows in the four samples: 226=4.07, 
227=4.15, 228=4.10, and 229=3.91. 
