1922] BUCHHOLZ—VASCULAR PLANTS 283 
classed as the ones affected by some form of developmental selection 
are many of the lethal factors that. have been described in some 
plants. It has been shown by Lirtte that some kind of embryonic 
selection is responsible for the non-appearance of homozygous 
yellow mice whose inheritance was studied earlier by CUENoT. 
No attempt will be made to discuss here the lethal factors of 
Drosophila in this connection, but doubtless many things whose 
inheritance would follow the lethal type, could be caused by a kind 
of embryonic selection mechanism. 
The origin of the variations or mutations is another problem. 
Chromosomal phenomena, such as gene mutations, the chromosomal 
mutation of non-disjunction, etc., are probably able to account 
for the actual origin of many variations. Similarly bud mutations 
and other heritable vegetative variations would also be accounted ~ 
for by some type of nuclear or intracellular phenomena. They 
would be played upon by developmental selection even before their 
outward manifestations are recognized. These mutations may also 
be acted upon by natural selection if they possess some very marked 
advantage or disadvantage. It is in this connection that develop- 
mental selection has a very definite réle in the origin and heritability 
of some mutations. Only those variations which are not affected 
by developmental selection could reappear regularly or give 
consistent Mendelian results. The outcome of developmental 
selection may be so decisively against a mutation that it may 
seldom recur. On the other hand, if the mutation is closely linked 
with a factor greatly favored by developmental selection, it may 
reappear as if fixed, even though the pollen produced is heterozygous 
for it. Thus we can understand how even the discovery and 
recognition of the mutations themselves depend upon whether 
they reappear in the next life cycle, and thus pass the censorship 
of the developmental selection machinery.. Developmental selec- 
tion is doubtless responsible for the recognizability of some varia- 
tions as mutations, but we have no evidence that it could be held 
responsible for the chromosomal phenomena themselves. 
Summary 
1. The process of developmental selection is a normal event 
or succession of events in the life cycle of vascular plants, where it 
