42 
BULLETIN 1164, U. S. DEPARTMENT OF AGRICULTURE. 
A very different condition is encountered in the second veneration 
Absolute sterility was shown by 7 per cent of the F 2 pop°ulation, 15 
of the 215 individuals having produced no seed, even from open- 
pollmated flowers. Four of these sterile plants developed bolls 
which, although containing only shriveled ovules, reached a o- 00 d 
size As compared with the averages for fully developed bolfs of 
the Fima parental race, they had a length of 67 to 82 per cent and 
a diameter of 90 to 105 per cent. -The seedless condition was 
common to all of the bolls borne by the 4 plants in queston On 
the remaining 11 sterile plants the bolls did not attain measurable 
size, although the numbers which had apparently set on these 
individuals ranged from 31 to 389 and exceeded 100 in 7 of the 
plants. 30 
In addition to the 15 completely sterile F 2 plants, a very low degree 
of fertility was not infrequent, one of the plants having produced 
onlv 7 bolls and others very small numbers. Shedding of the flower 
buds in various stages of their development and of the young bolls 
was the proximate cause of the low fertility. Deficient fertilization 
resulting in a very small number of seeds in the bolls which matured! 
was also of common occurrence. One obvious cause of deficient 
fertilization in many of the plants was the imperfection of the 
anthers (PI. VIII, fig. 8), especially of the uppermost ones upon 
which seli-iertilization mainly depends. Of 55 F, plants (not includ- 
m §-^ y n the 15 com P letel 7 sterile individuals above mentioned) on 
which flowers were bagged to prevent cross-pollination, 8 plants pro- 
duced no seed from such flowers and 9 plants produced fewer than 
10 seeds each, or fewer than are usually contained in a single boll 
from a bagged flower of either parental variety. It is probable that 
some of the plants of which the anthers were ^apparently well devel- 
oped produced defective pollen, but no investigation of this condition 
was made. 
The seeds produced by many of the F 2 plants were of low viabihtv 
and there was a heavy mortality in the seedling stage amono- the F 3 
hybrid progenies. Unfortunately, adequate data on this point were 
not obtained, but m approximated 25 per cent of the hills which 
were planted with selfed seed of the F 2 parents not a single plant 
developed beyond the seedling stage, although three or four seeds 
had been planted in each hill. 
The 22 F 2 plants of which F 3 progenies were grown had differed 
considerably m fertility, as is indicated by the fact that in the num- 
ber of bolls of all sizes counted on October 15 they ranged from 55 
to 326 and averaged 167. The average number of bolls per plant in 
the I 3 progenies was m every case much smaller than the number 
borne by the corresponding F, parent, the average for approximately 
400 plants m the 22 F 3 progenies having been only 27, with a mini- 
mum of 17 and a maximum of 50 bolls. The verv low fertility of 
the F 3 populations may have been due partly to the fact that these 
progenies were grown on the same land on which the F„ was grown 
the year previous, and the soil may have become partially exhausted. 
I he weather conditions also were less favorable for cotton in 1920 
than in 1919. It is highly improbable, however, that these were the 
sole causes of so great a decline in fertility from the F, to the F 3 . 
;o The highest number of bolls on any Fj plant was 406. 
