Variegated Leaves. 277 
way in this case as in that of the striped flowers. The 
buds in the axils of the leaves on the variegated sectors 
usually produce variegated branches, but those of the 
green sectors preen ones. Breeders take this fact into 
o o 
consideration in the choice of buds for use in the multi- 
plication of variegated forms, as we have already seen. 1 
It seems that the bud-variations, that is to say both the 
progressive (producing variegation) and the retrogres- 
sive or atavistic ones, are generally the result of a pre- 
ceding sectorial variation. But in most cases all traces 
of the latter are soon lost. In Quercus pcdunculata I ob- 
served, as I have already mentioned, a variegated twig 
on a green bush. In this case the variegation extended 
on to the bark and the color of the branches of the pre- 
ceding years could thus still be recognized. The main 
branch proved to be unilaterally variegated and the varie- 
gated twig arose from this side. The branches at the 
other side were green (Hilversum 1886). AYhen the 
leaves are arranged in two rows as in Castanea vcsca, 
Ulmns campestris etc. the leaves on one side of the 
branch may be variegated and on the other, green. In 
such cases I found the lateral twigs on the older parts 
on the green side to be entirely green and those on the 
variegated side entirely variegated ; but I might repeat 
here that there is still a tendency in the green branches, 
even if only a slight one, to produce variegated leaves. 
The contrast between green and variegated is therefore 
not nearly so great as would appear at first sight. 
We have no\v to consider the question of the in- 
fluence of external conditions on the degree of variega- 
apical cells would be of great interest, especially in the Conifers and 
vascular Cryptogams (e. ., funiperus, Adianthum, Selag'mella etc.). 
1 See SALTER'S method, Vol. I, Part I, p. 147. 
