June 9, 1923 
Rogue Types in Garden Peas 
835 
may be partially explained by assuming a differential mortality rate 
among the F2 segregates as explained in the paragraph dealing with 
the relation of the Aa and Cc factors in the F2 generation of Gradus 
rogue X Mummy. 
The occurrence of the early flowering segregates offers a partial explana¬ 
tion of the deviation in the number of broad-stipuled plants observed 
from that theoretically expected, as there is in all likelihood a correlation 
in Gradus and Mummy between the shape of the stipules and time of 
maturing. As has been sho\vn, the Graduslike character of the broad- 
stipuled F2 plants was inherited through the rogue parent, which also 
carried the factors for early flowering. The results with Gradus X 
Mummy show that linkage exists in Gradus between the factors y for 
stipule shape and a for flower color. Hoshino (7) has shown the factor a 
to be linked with a factor for early flowering. Necessarily linkage exists 
in some degree between the factor y for stipule shape and the factor for 
early flowering. One is, therefore, justified in assuming that many of the 
white, broad-stipuled segregates would be early in maturing and the 
correlation would account for a part of the deviation of the observed 
number of broad F2 types from the number theoretically expected. 
STATISTICAI. STUDY OF THE F2 GENERATION OF THE CROSSES GRADUS ROGUE AND 
MUMMY 
In 1920 were grown 51 Fj families of the cross Gradus rogue with 
Mummy. The average ratio of stipule width to length in the mature Fj 
plants was calculated in the same manner as for the generation. The 
F2 segregates were also classified according to flower color but not de¬ 
scribed as broads, intermediates, or rogues. 
RELATION OF COLOR FACTORS A AND a TO STIPULATE SHAPE IN TWO ATYPICAL FAMILIES 
With the exception of the plant already mentioned,® all of the Fj 
hybrids were rogues and appeared similar in regard to stipule shape. 
Although somatically alike the F2 frequency distributions of the progeny 
of two apparently normal Fj rogues ai*e different from those of the 
remaining 49 F2 families in both their range of variation and mean 
stipule ratio (Table XV). The means of the two families are 1.999 ± 
0.9106 and 1.905dbo.0126 and for the two together the mean is 1.998^ 
0.0082 as compared with 2.333^0.0050 for the other 49 Fj cultures 
(Table XVI). 
Because of their similarity both in frequency distribution and in the 
value of the mean stipule ratios, the two anomalous cultures have been 
lumped together for a statistical study of the color segregates. The 
means of the three color claSvSes are: For the whites, 1.899^0.0159; for 
the pinks, 2.046±0.0167; for the purples, 2.038 ±0.0095; colored 
plants as a whole, 2.027±0.0085. difference in means of the AA 
and aa as compared with aa plants is 6.7 times the probable error. The 
frequency distribution of the a plants resembles more that of Gradus 
than it does that of Mummy or Gradus rogue. The frequency distri¬ 
bution of the A population resembles that of Mummy rather than that 
of either Gradus or Gradus rogue. The difference in the mean stipule 
ratios of the two color groups, as well as the difference in their ranges 
^ The two exceptional Fa families were cultures numbered 0.1059 and o. 1093 (Table XV), the progeny 
of Fi plants 9.722-1 and 9.729-5, respectively (Table XII). 
