336 
Journal of Agricultural Research 
Vol. XXVII, No. 
Table IV .—Relative rapidity of penetration of the ovary of Pima cotton by Pima and 
by Upland pollen tubes , as measured by the degrees of fertilization attained when the 
pistils were excised at successive intervals 
Pollination with— 
Number of 
hours from 
pollination 
to excision 
of the pistils. 
Pima pollen. 
Upland pollen. 
h'lnwprc 
Percentage 
Mean number 
Flowers 
treated. 
Percentage 
Mean number 
JL 1U If Cl O 
treated. 
of bolls 
developed. 
of seeds per 
boll. 
of bolls 
developed. 
of seeds per 
boll. 
8 . 
IOO 
O 
0 
IOO 
O 
O 
IO. 
IOO 
2. ±0. 94 
8- S± 3 - 58 
IOO 
8. ±1.83 
11. 5±i. 42 
12 . 
IOO 
IO. ± 2 . 02 
5 - 3 ± r - 
IOO 
18. ±2. 59 
7. 8±o. 90 
In Pima cotton, under normal conditions of pollination, from 75 to 
97 per cent of the flowers develop bolls (7, Table 34 , p. 55) and the 
bolls contain an average of from 14 to 18.5 seeds (7, Table 30 , p . 31). 
Comparing with these figures the percentages of bolls developed and the 
mean numbers of seeds per boll as given in Table IV, it is evident that 
relatively few pollen tubes had penetrated the ovaries within 12 hours 
after deposition of the pollen. Yet, it should be noted that of the flowers 
excised at this time, one, pollinated with Pima pollen, developed a boll 
containing 18 seeds and one, pollinated with Upland pollen, developed 
a boll containing 20 seeds. 
The data, so far from showing more rapid development of the like 
pollen, seem to indicate that better fertilization was attained by the 
unlike pollen when the time available for penetration of the ovary was 
shortened by excision of the pistil. In fact, however, the differences 
between the pollinations with Pima and with Upland pollen, in per¬ 
centages of bolls and in mean number of seeds, for either the io-hour or 
the 12-hour period, are in all cases less than three times the probable 
error of the difference. 
The results of this experiment, agreeing in the main with those of a 
similar experiment previously reported, make it reasonably certain that, 
when separately applied, like pollen has no advantage over unlike pollen 
in its rate of penetration of the ovary of Pima cotton. It does not fol¬ 
low, however, that there may not be a difference in the rate of penetra¬ 
tion when both pollens are present on the stigmas of the same flower. 
Jones (6) obtained evidence that in maize, when a mixture of two pollens 
was applied to the stigmas, there resulted a greater proportion of cross¬ 
fertilization in the upper than in the lower half of the ear, from which 
he inferred a slower rate of development of the unlike pollen. The indi¬ 
cated differences in rate of growth seem, however, too small to account 
for the degrees of selective fertilization shown in some of Jones’ com¬ 
binations. 
Maize, with its elongated female inflorescence and extraordinarily long 
styles, is far superior to cotton as a subject for the study of differential 
fertilization. In Pima cotton the stigmas and styles average only 30 to 
35 mm. in length and in the Upland varieties used in these experi¬ 
ments they average not more than two-thirds as long as in Pima. The 
ovary of both types is only 5 or 6 mm. long at anthesis. In view 
of these facts it seemed unlikely that evidence of differential fertiliza- 
