ricotylous Races Do \ot Arise by Selection. 407 
by peculiarly favorable circumstances which were not 
repeated in the following year. 
The repetition of the experiment in 1899 gave prac- 
tically the same result. 
All in all, in the ten generations of my experiment, 
there occurred neither a sudden nor a gradual transition 
to an intermediate race. 
And lastly let me summarize the whole culture in a 
simple table which gives only the mean and the highest 
value for each generation as well as the value for the 
seed-parent selected for the continuation of the race. 
VALUES FOUND IN THE FOLLOWING 
GENERA- YEAR OF gpRING 
TION FLOWERING 
MEAN MAXIMUM SEED-PARENT 
I. 1889 4.5 
II. 1890 4.5 
III. 1891 13 
IV. 1892 3.5 11.8 6.3 
V. 1893 4.7 7.4 7.4 
VI. 1894 1.8 3.6 3.6 
VII. 1895 2.5 5.5 5.5 
VIII. (450 Ex.) 1897 4 25 16-25 
IX. (400 Ex.) 1898 2 11 
The whole series, with the exception of the maximum 
figures for 1897 rather indicates a fluctuating around a 
constant mean value than a steady progress under the 
influence of selection. 
Scrophularia nodosa. Of this species a series of 
forms occurs. Their characters have been given in the 
Prodromus of DE CANDOLLE. The leaves are broadly 
cordate or only rounded at the base, with little teeth of 
almost uniform size, or very coarsely and irregularly 
indented, and the fruits are large or small. The form 
with cordate, uniformly toothed leaves is very common 
in this neighborhood, and, so far as I know, the only 
