286 Non-Isolablc Races 



luarchiana. In the normal families of my cultures some 

 specimens that happened to be variegated had, after arti- 

 ficial self-fertilization, given rise to no more than two 

 per cent of variegated offspring and in the next genera- 

 tion the same proportion was repeated, the conditions be- 

 ino- the same. From these I selected in 1898 the four 

 finest young plants, planted them out a meter a])art, and 

 thereby obtained strong, richly-branched individuals, of 

 which some were slightly, and others strongly, variegated. 

 On all of them the flow^ers from which T intended to sa\-c 

 seed were artificially and purely fertilized with their own 

 pollen. On each of the four plants I first fertilized 

 flowers on the pure green and afterwards some on the 

 variegated lateral branches. Amongst 675 seedlings of 

 the former and 1300 of the latter group the seeds pro- 

 duced the following percentages of yellow and variegated 

 seedlings : 



PERCENTAGE IN VARIEGATED SEEDLINGS 

 PLANTS 



GREEN BRANCHES VARIEGATED BRANCHES 



No. 1 0-0 % 1 % 



No. 2 — 3 



No. 3 0-0 4-12-18 



No. 4 0-0 6-9-45-100 



Eacli num1)er refers to a separate l^ranch The six 



greens gave rise, as we see, to green seedlings only, but 



the varies^ated ones to a laro;er or smaller number of 



seedlings with this character. The figures 1, 3, 4, and 



45 in the last column relate to the slightly variegated 



branches ; the rest to those with this character more 



strongly developed. The latter therefore gave a higher 



percentage of variegated offspring. The variegated seed- 



linijs had either vellow or flecked cotvledons, or ^reen 



cotyledons and flecked leaves, and of these three groups 



there were 68% of the first, 12% of the second and 20% 



