ON FOUR DOORYARD FLOWERS 



pollen and dust it on the stigmas of plants that 

 have already shed their pollen. To make abso- 

 lutely sure about guarding against the self-fertili- 

 zation of the flower, it would of course be neces- 

 sary to remove the stamens before ripening. 



Some of my experiments in hybridizing have 

 been conducted with the idea of producing fra- 

 grant races of geraniums. The chief difficulty in 

 this work is that most of the fragrant geraniums 

 have been grown for such a length of time from 

 cuttings that they have for the most part lost the 

 power of producing seeds. This makes it obvi- 

 ously difficult to secure seeds from the plants that 

 are precisely the ones it would be desirable to use 

 for the purpose. 



Nevertheless I have produced a number of vari- 

 eties having fragrance of very attractive quality. 



One of these fragrant varieties is developed 

 from a compact growing Australian form which 

 produces an enormous amount of seed. If this 

 form were crossed with the other fragrant varie- 

 ties a valuable type should be produced, as this 

 plant has recently come from the wild and would 

 instill vigor into the specialized and long cultivated 

 plants. 



A line of work that I carried out at one time 

 involved the crossing of the Pelargoniums with 

 variegated leaves with those having ordinary green 



[131] 



