548 Heritable Spiral Torsions. 
A B %A %B 
Twisted main stems . . 5 2 20 10 
Ternary main stems 1 1 4 5 
Decussate mam stems 19 17 76 85 
Total 25 20 
A and B are two groups which had been produced 
from seeds separately harvested from two seed-parents 
of 1892. 
The extent of the culture is, as we see, too small to 
allow of an exact determination of percentage values. 
Seed was saved in the autumn of 1894 from the four 
best plants which had been isolated from the rest before 
they flowered. 
The sixth generation (1895-1896) again produced 
a better result, viz., 42% twisted main stems, a result 
due in part at least to the increase in the distance between 
the plants in the culture, which was in other respects the 
same as in the previous year. The seed of 1894 was 
sown in the middle of March in pans in the greenhouse. 
The seedlings were transplanted to pots in April and into 
the bed in May Altogether there were only 33 plants 
on 4 square meters, that is about 8 per square meter. 
In October I found 14 rosettes with a spiral arrangement 
of their inner leaves, whilst seven were ternary and 12 
decussate, that is 42% spiral, 21</c ternary, and 36% 
decussate. This result was confirmed in May and the 
decussate and ternary individuals were thrown awav. The 
* m 
six individuals with the finest torsions were selected as 
seed-parents and isolated before flowering. 
The seventh generation (1897-1898) was grown in 
practically the same way. Seeds of 1896 were sown in 
the greenhouse in May 1897, were transferred to pots 
later, and to the beds at the beginning of July at the rate 
