56 BULLETIN 14£8 ? IT. S. DEPARTMENT OF AGRICULTURE 
the period of reduction divisions have an effect on crossing over. 
The plants considered in the present groupings flowered during a 
]3eriod of a week or more, and if unfavorable climatic conditions 
affected the gametes this effect would be observed on the early 
gametes of one plant and on the late gametes of another. A com- 
bined population would mask such effects, but a consideration of 
the individual differences shows that none is greater than three 
times the probable error, and all reasonably may be ascribed to 
chance. 
The probable errors are so large, however, that no final conclu- 
sions as to the lack of influence of environmental factors can be 
drawn. The evidence merely indicates that factors operating during 
the period of reduction division are not critical with respect to the 
rate of crossing over. 
CORRELATIONS 
If apparent crossing over is influenced by the condition of the 
parent plants, it was thought this might be reflected in a correlation 
between vigor (as measured by the number of seeds produced) and 
the rate of crossing over. Very definite evidence of this was found 
in the data that had accumulated before 1923. Forty reciprocal 
crosses measuring the linkage between C and Wx made prior to 1923 
showed a positive correlation of 0.51 ±0.08 between the percentage 
of crossing over in the male gametes and the number of seeds pro- 
duced on the ear of the double-recessive - plant fertilized by the 
heterozygous pollen. There also was a correlation of practically 
the same magnitude (0.56 ±0.07) between the rate of crossing over 
in the female gametes and the number of seeds on the double-reces- 
sive plant. With the idea of following up the suggested relationship 
indicated in these early experiments, an effort was made to amass 
a larger population derived from a single progeny in one season. 
The results of this effort were successful in that 47 reciprocal 
back-crossed ears were obtained in a single progeny, but the data 
show the relationship to be much less close than in the earlier experi- 
ment. The correlation between the crossing over in the male and 
number of seeds on the recessive plant was 0.24±0.09. The correla- 
tion is a little less than three times the probable error. That this 
correlation, though small, is significant is indicated by the results 
from the entire progeny, including ears not reciprocals. The cross- 
ing over in the male was measured in 112 individuals, and the corre- 
lation between the crossing over and number of seeds produced on 
the double-recessive plant is 0.28 ± 0.06. This is in close agreement 
with the coefficient found in the population restricted to reciprocals, 
but is more than four times the probable error. 
A further examination of the data prior to 1923 indicates that the 
large coefficients were the result of interprogeny correlations. Ten 
progenies were involved, and when the mean crossing over in the 
male is plotted against the mean number of seeds on the recessive 
plants there is a coefficient of 0.78 ±0.15. The number is so small 
and the means are derived from populations that vary so greatly in 
size that no importance should be attached to this coefficient beyond 
an interpretation of the positive correlation found in the combined 
population. 
