18 THE BANANA 



1000 years, numerous varieties could be continuously 

 raised, and a selection made of such as seemed promising. 

 Numerous experiments were carried out at Hope 

 Gardens, Jamaica, with this end in view. Cross -pollina- 

 tion is easy enough to carry out, and the experiments 

 were made under varying conditions in soil, in water 

 supply, &c. No success was obtained in a great number 

 of experiments by using the pollen of the Jamaican 

 banana, and at last the pollen of the red banana (var. 

 rubra), dusted on the stigmas of the ordinary cultivated 

 banana, led to the production of seed. Unfortunately the 

 hurricane of 1903 levelled the banana plantation before 

 the seeds were ripe, and the experiments were for the time 

 abandoned. It was thus definitely proved at Hope 

 Gardens that the female flowers which produce the com- 

 mercial banana have not become sterile after ages of 

 vegetative reproduction of the plant, but are capable of 

 producing seeds. 



If experiments are made again, it is suggested that 

 pollen be used not only of the cultivated varieties, but 

 also of distinct species with edible fruit. 



As far as could be ascertained, the pollen of the common 

 Jamaican banana was quite inert on female flowers of that 

 variety. But it was not proved that this was due to 

 sterility in the pollen itself. Experiments might be made 

 to determine whether this pollen is fertile by using it to 

 pollinate the female flowers of some of the seed-bearing 

 species. It is well known that pollen is inert in many 

 species on flowers of the same plant, although fertile on 

 flowers of another plant of the same species. This may be 

 the case with species of Musa. Now, as seedless cultivated 

 bananas are propagated by off -shoots (suckers) they are 

 practically all derived from one individual, and many 

 varieties are only sports from the same individual stock ; 

 if the pollen is inert on the same individual, it would 

 naturally be also inert on all plants derived, however 

 remotely, from that individual, even although they may 

 have sported. 



