1060 
and the other with broader blue flowers. These groups do not, 
however, correspond to the white and blue P-varieties, as is the 
case in the cross between the narrow-petalled white and the common 
blue flax. The white ones are in general broader than the narrow- 
petalled white flax and the blue ones are on the whole narrower 
than the Egyptian. In both transitional breadths occur between the 
two P-varieties. Nevertheless the phenomena do not agree with those 
observed in the cross between the common white and the Egyptian 
flax, for, although transitions do occur, the white and the blue F,- 
individuals are not identical in breadth, as in this last named cross. 
The white ones have no very broad, and the blue ones no very 
narrow petals. In order to determine the extreme limit, which both 
groups can reach, some of the white /,-plants with the broadest 
petals were cultivated further. From the third generation so obtained 
a few of the broadest were again selected for further culture, and 
similarly from the fourth and the fifth generation. Although the 
nutritional conditions were always most favourable, the greatest 
breadth cbserved in the 786 plants from the second to the sixth genera- 
tion was 11.4 mm. This was in the fourth generation. Among the off- 
spring of this plant in the fifth and sixth generation even this 
breadth did not oceur again; the maximum in /, was 10.5 mm. 
Evidently therefore the very broadest white ones barely surpass 
the minimum breadth of the Egyptian flax and do not even reach 
its mean breadth. 
In the same way the narrowest blue ones were cultivated further 
for some years, the conditions being made less favourable. The 
extreme minimum was found to be 5.7 mm. Blue flowers with 
petals of the breadth of the narrow-petalled white flax did not occur 
at all among the 722 plants investigated up to the fifth generation. 
In this crossing then two groups are indeed formed, one with 
narrow white and the other with broader blue flowers. This was 
also found to be the case on further cultivation of the heterozygotic 
F- and F,-plants; the whites were always on the whole narrower 
than the blue ones formed at the same time. 
The way in which the breadth behaves in this crossing, is at 
first sight not in accordance with Merpri’s segregation-law. The two 
forms crossed show a great difference in breadth and among the 
offspring intermediate forms are also found. Yet the second genera- 
tion does not show the ordinary phenomenon of Mendelian segrega- 
tion, as is observed with characters which vary in a fluctuating 
manner, where the limits between the different groups formed in 
fF, are obscured. In such cases the /,-individuals together give a 
