GABRIELLE L. C. HOWARD. 41 



number of flowers on two individuals in nearly all the fifty-one 

 types of Indian tobaccos, and also in some Fj hybrids, were 

 castrated. About fifty to one hundred flowers were prepared 

 on each plant under everj'^ possible condition. In some the 

 anthers were removed, in others, both anthers and stigma. 

 The plants used included types which self -fertilize with great ease 

 and those which will set hardl}^ any seed unless selfed, as it was 

 thought that the latter would be the most likely to produce 

 parthenogenetic seed. Plants were chosen at all periods of their 

 growth — when in full seed formation, when full of capsules and 

 going off their bloom, and when very nearly over. In most cases 

 the plants were heavily pruned, all capsules, flowers and buds 

 other than the castrated ones being removed (such heavy 

 pruning ordinarily induces rapid seed formation), others were 

 lightly pruned. The same methods were adopted in 1911, but 

 here the number of kinds employed was smaller, onh^ those used 

 as parents in the hybridization experiments were tested again, 

 namely. Types 9, 51, 16, 35, 23 and 38. The castrated flowers 

 were enclosed in parchment bags and these were taken off 

 at frequent intervals in order that any newly-formed buds 

 might be removed. In the earlier experiments the bags were 

 not applied after the corollas had withered, but in the later 

 experiments bags with perforations were placed over some of the 

 branches. A great difference was found between the capsules 

 formed from the castrated flowers and those formed by ordinarj^ 

 pollination. In the latter case the capsule swells quickl}^ and 

 remains firmly attached to the plant. No difficultj^ is 

 experienced in removing or replacing bags, and the peduncle 

 would have to be broken before the capsule could be removed. 

 This is always the case, whether the flower be self- or cross- 

 pollinated. The capsules of the castrated flowers, on the other 

 hand, although they also became swollen at first and simulated 

 the fertilized ones, were very easily detached from the plant. It 

 was exceedingly difficult to remove the bags, which finally had to 

 be cut away carefully. The capsules thus exposed to the air were 



