BANANA CULTURE IX HAWAII 7 
Like a number of other fruits of commercial importance, the 
seedless banana is distinctly parthenocarpic and does not require the 
stimulus of pollination to insure the setting of fruit (#, p. 293). 
However. 15 of the :i7 species of the subgenus Eumusa produce fruit 
containing seeds [11. p. 263). A number of well-established vari- 
eties of seedless bananas, including Gros Michel (Bluefields) and 
Apple, have been successfully pollinated by hand and caused to 
produce seed (1L preface, p. 8). The Ice Cream variety of Hawaii 
when growing near abaca plants is occasionally naturally pollinated 
and produces seed. W. J. MacNeil, Oahu College, Honolulu, reports 
that in 1909 he succeeded in crossing flowers of the Ice Cream vari- 
ety with pollen from seed-producing banana, presumably abaca. 
More than one bunch produced seeds, some fruits having 1 to 15. 
Ten of the seeds, collected at random when dry. weighed 0.6 gram, 
and were 0.8 by 0.6 by 0.4 to 0.6 by 0.5 by 0.1 centimeter in dimen- 
sions. Of 50 seeds planted on two different occasions none germi- 
nated. 
In nearly all commercial varieties the pistillate flowers with im- 
perfect stamens are open and their stigmas past the receptive stage 
for pollen before either the neutral or the st animate flowers are 
released by the opening of the bracts (pi. 3, A). This fact indicates 
the need of cross pollination to produce seed. Insects, which are 
attracted to the flowers by the nectar, might possibly be the occasion 
of cross pollination, but since seeds are rarely produced, the pollen 
of most of the commercial varieties either does not develop, or it 
is infertile. 
It is believed that the original wild banana bore perfect flowers; 
that is, having both stamens and pistils, as is still represented in some 
of the local varieties, and that pollination of flowers of the same 
cluster was aided by external agencies, such as the wind and insects. 
In isolated specimens of the local Borabora (pi. 3, B), Maia Oa. and 
abaca varieties, it was observed that the flowers becoming seed- 
bearing fruits were pollinated before the staminate flowers, located 
farther along the rachis, were released from under the bracts. It is 
also probable that there once were species, if not varieties, whose 
floral parts were modified, requiring cross pollination to fertilize the 
ovules and produce seeds capable of germination, just as is now the 
ca-e with most of the Hawaiian commercial varieties, which usually 
are parthenocarpic. Examination of a number of these varieties 
failed to show the presence of pollen on their flowers. Several 
varieties, such as the Ice Cream and Chamaluco, had some pollen, 
the surface of whose grains was found to be collapsed, however, sug- 
gesting infertility. S. C. Harland. of Trinidad, British West Indies, 
reports crossing successfully the Gros Michel, resulting in the produc- 
tion of seeds which germinated, giving in one instance a hybrid of 
highly desirable qualities. Some varieties are known to be bud 
sports; for example, the Green Red banana of Panama, which O. A. 
Reinking. pathologist of the United Fruit Co.. states is a bud sport 
of the well-known Red variety, i- also a bud -port of some yellow 
variety. A bunch of the fruit of the Green Red banana was found to 
contain 150 seeds scattered throughout the linger-, only about half a 
dozen of which were perfect. Doctor Reinking call- attention to 
