VARIABILITY IN LINKAGE OF CHARACTERS OF MAIZE 49 
are also shown in Table 30. The differences are extremely variable, 
ranging from —19 to 11. The mean weighted difference of 0.43 ± 
0.67 per cent is in the direction of more crossing over in the second 
ears, but obviously this is without significance. 
The three progenies examined all showed differences of compara- 
ble magnitude between upper and lower ears in the proportion of 
crossover to noncrossover seeds where the male was the heterozy- 
gous parent. Two of these progenies had the highest proportion of 
crossover seed on the lower ears; the third being the reverse, with 
a higher proportion of crossover in the upper ears. 
Since these differences are found when the male parent is hetero- 
zygous, and apparently only then, there can be no question of a 
different rate of crossing over in the formation of the gametes, and 
the differences observed could arise only as the result of some sort 
of selective action operating on the male gametes after they were 
formed. 
Only two progenies afforded material for testing differences in 
crossing over between upper and lower ears when the female was 
the heterozygous parent. These progenies were consistent in that 
no differences were found. It would seem, therefore, that even 
though environment has an effect on the proportions of gametic 
classes, the time elapsing between the maturation division in the 
upper and lower ears does not provide a period sufficiently long 
to bring about a change in the rate of crossing over. Furthermore, 
it seems clear that no consistent death rate operates among the 
female gametes, the lower ears having no higher rate of crossing- 
over than upper ears. 
RATE OF CROSSING OVER NOT ALIKE IN BASE AND TIP OF THE SAME EAR 
The seeds of an ear of maize develop under closely similar environ- 
mental conditions. Nevertheless, owing to the length of the ear, the 
seeds from the base and tip are subjected during fertilization to 
slightly different conditions. 
Silks from the base of the ear emerge earlier and in consequence 
are older at the time of pollination than the silks from the tip. In 
extreme cases the difference in time of appearance may be as much 
as two or three days. In cases of this sort it is obvious that the pollen 
tubes fertilizing ovules at the tip of the ear have an advantage over 
those fertilizing basal ovules, since they have not only a shorter 
distance to grow but also younger tissue in which to develop. 
In view of these advantages, if the apparent crossing over in the 
male is influenced by differential mortality of the male gametes or a 
differential growth rate of the pollen tube, one might expect a 
difference in crossing over between the base and the tip of the ear. 
If the crossover class were the weaker it might achieve fertilization 
less frequently in ovules at the base than at the tip of the ear. 
From two sister progenies grown at Arlington, Va., in 1923, 79 
well-formed ears were selected. The seeds from the base and from 
the tip of these ears were classified separately. Since the silks from 
the extreme base of the ear often are delayed in emerging, a few 
seeds from the base of the ear were excluded from the comparison. 
With the idea of disclosing a trend from tip to base in the per- 
centage of crossing over, some of the larger ears were divided into 
