530 Journal of Agricultural Research voi. xviii, No. lo 



anthers dehisced naturally. In the sixth stage the anthers had dehisced 

 the same morning the filaments were extended. The base of the anther 

 still contained considerable pollen, which was freed readily. The seventh 

 stage was represented by anthers which had dehisced 24 hours previously. 

 In this and in the eighth stage, in which the anthers had dehisced 48 hours 

 'previously, only a few still contained pollen. Only the best-appearing 

 of this was used. 



The period in which the pollen was found to be viable was very short. 

 The best pollen was secured from anthers just breaking. Pollen from 

 anthers ready to break gave a slightly lower percentage of successful 

 pollinations. In 191 8 the anthers used at this stage were slightly less 

 mature than those used in 191 7. Stamens with green anthers were used 

 in both 1 91 7 and 191 8, but the two following stages were omitted in 191 8. 

 No seed was secured in either year from the use of immature pollen. 

 These results are contrary to a widespread impression that detached 

 anthers will complete the process of development, ripening, and fertiliza- 

 tion when placed in a flower. By immature is not meant rudimentary 

 pollen but pollen from anthers which, if left undisturbed, would not break 

 for several hours. Pollen frequently can be dusted from such anthers by 

 forcible rupturing of the walls. It will be recalled that this pollen swelled 

 in water but did not burst. No artificial germination of such pollen was 

 obtained. 



The pollen remained highly viable in the field during only a few hours. 

 Pollen taken from the anthers two or three hours before natural dehiscence 

 effected fertilization in about 30 per cent of the attempts. Pollen taken 

 from anthers which were dehiscing resulted in over 60 per cent of success- 

 ful pollinations, while pollen remaining in anthers which had dehisced 

 two or three hours previously effected fertilization in less than 40 per 

 cent of the flowers pollinated. 



In both years some fertilization resulted from pollen which had 

 remained viable 24 hours after dehiscence. In one instance in 191 8, 

 not here reported, seed was secured from pollen of anthers which may 

 have dehisced 48 hours previously. The age of the pollen in those 

 classes after dehiscence is not certain. The stamens were left in the 

 flowers, and the pollen remaining in the base of the anther was used. 

 The age of the flowers was reckoned by means of observations on adjacent 

 flowers and by other observations which have enabled the authors to 

 determine the stage of development to within close limits. 



In 1 91 8 shrunken seed followed the use of pollen taken from anthers 

 24 hours after dehiscence. The numbers secured were not sufficient to 

 indicate that the poor development was due to poor pollen. 



It is the opinion of the authors that the success obtained with over- 

 ripe pollen was due to the especially favorable conditions that prevail in 

 Idaho. It is improbable, to the verge of certainty, that pollen could 

 have remained viable in ruptured anthers 24 hours after dehiscence in a 



