62 PAUL J. ANDERSON 



and it is quite likely that some of the blossoms were already 

 fertilized when the dust was applied. It is not certain but that 

 in the latter case the fruit would set normally. 



It should be mentioned in this connection that for some time 

 after the falling of the petals, no difference could be observed be- 

 tween fruits on treated and untreated branches. Both kinds 

 grew at the same rate. Often it was ten days to two weeks after 

 blooming before any difference became noticeable. Then the 

 unfertilized fruit stopped growing and soon dropped. 



Nature of the mjunj. Having demonstrated that dust did pre- 

 vent setting of fruit, the next question to arise was: what is the 

 injurious constituent in the dust and how does it act on the flower? 

 A "cement dust solution" was made by shaking up a quantity 

 of dust in distilled water and then letting it set for several hours 

 in order to allow any soluble parts to go into solution. This 

 was then filtered and the filtrate applied to the flowers as a spray. 

 It prevented setting of fruit almost as effectually as the cement 

 dust itself. As indicated by the analyses there is only one thing 

 in the dust that goes into solution to any appreciable extent and 

 that is lime. It was reasonable then to infer that the latter was 

 the injurious part of the dust. A solution of pure lime was made 

 and applied in the same way. The results were the same as 

 those secured by spraying the flowers with cement dust solution. 



In another series of experiments, the blossoms of some large 

 lillies in the greenhouse were used. Cement dust was applied 

 to the stigmas only and then they were pollinated by hand. 

 Checks were pollinated but not dusted. One hundred per cent 

 of the checks set but none of those flowers which were dusted 

 produced seed. From these experiments it was concluded that 

 the lime of the dust affected the stigma or the stigmatic secre- 

 tions in such way as to interfere with fertilization and thus pre- 

 vent setting of the fruit. 



It is a well known fact that the stigmatic secretions of most 

 plants are acid in character. One may easily demonstrate this 

 point by moistening a piece of blue litmus paper and merely 

 touching it with the stigma of a cherry blossom. A pink spot 

 will appear on the paper at every point touched. If a minute 



