266 University of California Puhlications in Botany [Vol. 5 



increase in successful treatments on the plants which were allowed 

 to "go off their fullest bloom" before starting experiments on their 

 flowers. In the total series of 800 experiments buds as short as 

 11 mm. and as long as 48 mm., were used for castration and mutila- 

 tion. The number of buds under 35 mm. in length which gave 

 fruits and seeds following treatment is almost identical with the 

 number of buds over 35 mm. The size of the bud to be treated has, 

 then, little significance for the production of fruits and seed following 

 castration or mutilation. Similarly, four- as compared with five-parted 

 flowers are not significant for the production of parthenogenetic 

 or phenospermic seed, since the various castration and mutilation 

 experiments described in table 1 involve approximately equal numbers 

 of the two types of flowers. However, seed containing endosperm 

 and embryos, or endosperm alone, seems to have resulted, in practi- 

 cally all cases, from the more normally formed, flve-parted flowers. 

 Similarly the use of a single bud on an inflorescence for treatment 

 seems, as was perhaps to have been expected, to have been more 

 efficient for the production of normal seed than the castration or 

 mutilation of more than one bud on an inflorescence. Of the total 

 800 operations less than one third involved the pinching off of the 

 stigma as well as emasculation. Similarly only one tenth of the total 

 number involved the pinching off' of the stigma only. Thus it is 

 significant that of the nine cases in which normally formed seeds 

 were produced, three should represent treatments in which the stigma 

 was removed. In general for the production of parthenogenetic 

 and phenospermic seed, simple castration seems most effective. It 

 seems probable that certain of the stimulating and irritating agents 

 used by Hartley and Wellington would be more effective for par- 

 thenogenesis in "Nic. tahaccum Cuba" than in the varieties which 

 they employed. No experiments of this sort were attempted. 



I have assumed throughout that parthenogenetic seed was actually 

 produced in a few cases following the castration or mutilation of 

 flowers borne on the variety of N. Tahacum which Mrs. Thomas used 

 in her much more generally successful experiments concerning par- 

 thenogenesis in Nicotiana. If there were not a very remarkable 

 tendency in this iV. Taliacum variety to mature normal fruits and 

 phenospermic seed following castrations and mutilations of flowers, 

 or, on the other hand, if anything approximating full capsules of 

 normal seed had resulted in a number of cases from such treatment 

 of flowers, I should feel entirely willing to assign the production of 



