350 LINDSTROM, GENETICAL RESEARCH WITH MAIZE 
is almost completely cross-pollinated. Experimental evidence (KIESSEL- 
BACH 1922) indicates that there is probably less than one per cent self- 
pollination. For this reason the species is certain to be in a very hete- 
rozygous condition, even in the best commercial varieties. 
Inbreeding of commercial varieties of maize is characterized by an 
astounding amount of defective types in the offspring. It is a safe esti- 
mate to say that at least 40 per cent of selfpollinated ears of commer- 
cial varieties of good agronomic standing will show defective types of 
plants in the progeny. These include numerous sorts of chlorophyll ab- 
normalities, defective or arbortive kernels, dwarfs and other morpho- 
logical defects of seedling, stalk, leaves, ear, or tassel. One cannot ap- 
preciate the vast array of poor types that arise in this way unless he 
has tested it experimentally. 
Practically all of these defective types are recessive in inheritance to 
the normal or standard maize plant. The reason for this is obvious. Na- 
ture of course may produce both dominant and recessive deviations 
from the type. Dominant defects will naturally perish in the struggle 
for existence. Recessive defects on the other hand may be carried on 
for generations by virtue of the heterozygous condition of maize oc- 
casioned by its method of cross-pollination. In this manner the maize 
species gathers and holds manyof the recessive defects that have arisen 
during its evolutionary history. Natural selection against the defective 
types themselves is of no avail in destroying them completely. Their 
4 + 
Fr 
persistence in the species is thus theoretically proven, and a practical - 
test of inbreeding will verify this fact very clearly. 
It is for this reason that the improvement of maize by breeding has 
been found to be so extremely difficult. Ear-to-row breeding is not suc- 
cessful; close selection to type is dangerous; all because of the perni- 
cious effect of inbreeding which is characteristic in reducing the yield 
and vigor of maize. The abnormal or defective types can easily be eli- 
minated by continued self-pollination, but the remaining stock is ne- 
vertheless surprisingly poor in the great majority of cases. 
It may be that by isolating and removing the defective types ae in- 
breeding, we take from the remaining stock, because of genetic linkages 
of good and bad facors, some of the favorable factors for yield. At any 
rate something happens to reduce the yield and vigor. The only hope 
remains in crossing the purified, inbred strains with the hope of resto- 
ring most of the favorable factors for growth by such outcrossing. At 




