Aug. i8, 1933 
Inheritance of Dwarfing in Maize 
315 
The fit is rather better, and it seems safe to conclude that the genes for 
dwarf stature and perfect flowered ears are located in the same chromo¬ 
some with approximately 1 per cent of crossing over. 
The failure to find a crossover class involving the short stature of 
dwarf suggests that such combinations fail to survive, though such an 
hypothesis seems wholly without a logical basis, since all crossovers in 
this group would be in the direction of normal plants and, therefore, 
theoretically have a higher survival value than the noncrossover, unless 
the assumption is made that some or one of the characters associated 
with dwarf stature is closely linked with dominant factors for growth, and 
hence the crossovers with stature would have still a lower survival value 
than the dwarf complex. In some aspects this hypothesis seems worthy 
of consideration. 
Lethal or semilethal characters that recur constantly must be due 
either to frequent mutation or, as seems more probable, must be asso¬ 
ciated with some hereditary element that raises the survival value of the 
heterozygous stock. While mutations are known to occur in certain 
stocks with frequency, if such mutations have a negative survival value 
only, it is difficult to see how they and their mutating stocks persist, 
since their lethal nature insures their rapid elimination. That stocks of 
maize heterozygous for lethal characters are common is well known. 
Perhaps the most widely recognized of these is the albino, or white 
seedling, which, of course, never produces seed. There are others not so 
well understood but equally lethal, while the number of semilethals 
seems legion. Perhaps the most common of these, from the standpoint 
of repeated occurrence, is the andromonoecious dwarf. If, as has been 
suggested, maize possesses dominant factors favorable for growth, the 
problem of the recurrence of lethal factors is simplified by predicating 
that the deleterious variations which reappear frequently are those closely 
linked with one or more of these factors for growth, though the mutations 
need not have arisen in the chromosomes with the favorable growth 
factors since occasional crossovers would permit their survival. On 
this hypothesis it seems clear that lethal or semilethal variations must 
be those closely linked with the factors favorable for growth and, therefore, 
though destined for death in the homozygous condition, have a survival 
value as heterozygotes higher than that of normal plants. If this be 
true, then in a stock heterozygous for lethal or semilethal factors, the 
homozygous normal plants should be inferior in reproductive value to the 
heterozygous ones. 
Breeders long have recognized that certain variations have a sturdy, 
vigorous appearance which belies their inherent defects. Combinations 
of such variations usually make very favorable first generation hybrids, 
and it should be possible to reconstruct a high yielding strain by com¬ 
bining the deleterious recessive variations. If this be true generally, the 
practice of inbreeding and discarding those progenies which show dele¬ 
terious Mendelian variations is resulting in the elimination from the 
stock of the most desirable hereditary elements. 
The survival value of dwarf plants under field conditions is so low that 
a higher death rate for the crossover classes in this group could pass 
unnoticed, and of course there is no possibility of recognizing them in 
the seedling stage. 
