EEMAEKS UPON FERTILISATION. 



38 



This experiment was repeated several times, insects of 

 all kinds being most carefully excluded from the flowers; 

 and though diflering in some cases in the number of seeds, 

 yet in each the proportions were about the same. Now, 

 although we are not prepared to assert that the like results 

 can be obtained in the case of every species or variety, yet, 

 from other experiments we have made, we are certain that 

 the yield of seed wiU be larger and finer when the flowers 

 thus receive the benefit of a cross with pollen from a separate 

 plant, and where insect agency is deficient, it should always 

 be supphed by artificial means. To avoid self-fertilisation, 

 the flower should have the stamens removed as soon as pos- 

 sible after it is open ; this may be done by cutting them out 

 with a small pair oi scissors. According to some authorities, 

 certain flowers fertilise themselves before expanding, but this 

 is no doubt very exceptional. 



c 3 



