290 Non-Isolable Races. 
tricotylous specimen and had been raised from seed ob- 
tained in exchange; it flowered in 1893 in complete iso- 
lation and produced 6% yellow seedlings amongst its off- 
spring. I planted out about 60 of the green plants which 
grew to healthy individuals in 1899. I left the pollina- 
tion to insects but saved the seeds of each plant separately 
and then counted the number of yellow seedlings for each 
in a lot of 300. 
Y. S. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 22 24 27 30 
Ex. 27 611100011231 12403 1001 
Or 27 63 4 5 6 31001 
'Y. S." signifies the yellow seedlings in each lot of 
seed and 'Ex." the number of plants which exhibited 
this proportion amongst their seedlings. In the lower 
line from 2% onwards these are arranged in groups of 
3 in order to emphasize the general result. The num- 
bers of this last row are plotted in Fig. 55. This curve 
is similar to the curves of monstrosities which have been 
subjected to a selection extending over many years. 1 
Even as these, it consists of a half curve and of a bilateral 
curve. It indicates therefore the selection of a latent 
character which in this particular case started with the 
choice of a tricotylous plant which happened to be a 
particularly suitable one. 2 
The observations and experiments which we have 
described or passed in review show that variegated plants 
constitute a group of forms which in spite of being se- 
lected for years or even for a century have manifested 
no further improvement in the quality and constancy of 
1 See the second section and Sur les courbes galtoniennes des 
monstniositcs. Bull. sc. de la France et de la Belgique, published by 
A. GIARD, Vol. XXVII, April 1896, p. 396. 
2 See Uebcr cine Mcthode, Zivangsdrchun^cn aufznsuchcn, Ber. 
d. d. hot. Ges., Vol. XII, 1894, p. 25. 
