628 Validity of the Doctrine of Mutation. 
hereditary properties of the seeds which they produce; 
as, for instance, in Chclidoniuin inajus flore plcno (p. 
336). 
In the majority of cases a more detailed examina- 
tion is urgently needed, before the true nature of even 
the commonest bud-variations can be properly under- 
stood. This is particularly the case in variegated plants, 
amongst which every one is familiar with the phenom- 
enon, on shrubs and trees at any rate ; but even here the 
process has not yet been exhaustively studied. On the 
one hand, some of these cases consist of 'bud-atavism, 
whole branches of a variegated variety reverting to the 
normal type of the species in their color as well as in 
their secondary characters (Fig. 141). 1 On the other 
hand, sometimes halves of leaves become green, or occa- 
sional branches with usually slightly, but sometimes 
finely, variegated leaves arise on green individuals (Figs. 
142 and 143). On a large tree of Morns nigra in our gar- 
den the latter phenomenon occurs almost every year. 
It is not until all these and similar cases have been 
excluded that bud-variations may be regarded as true 
cases of A r egetative mutations. Even then we should 
require the proof that the deviating branches will re- 
produce their type from their seeds, after pure self-fer- 
tilization. In many cases, however, this is not possible 
because the bud-variations in 'question often bear no 
seeds, even when they occur regularly, as in Cephalo- 
ta.vus pcdnncnlata fasti giata (p. 109, Fig. 16) and in 
numerous other conifers, the bud-variations of which 
have been described by BEISSNER; as, for instance, in 
Cryptonieria japomca spiraliter falcata (Fig. 138, p. 619). 
Even from the green Dahlia I was, unfortunately, unable 
1 See above p. in and pp. 272-277. 
