248 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



fruit-growing districts, or in each county, to find out which varieties 

 are self-sterile and which are more or less self-fertile. With regard 

 to self -fertility, roughly speaking, if one blossom out of six were to set, 

 that would give sufficient crop — i.e., 16 per cent. We want to know 

 approximately the relative self- fertility, as if only one blossom in 1,000 

 set, it is for practical purposes self-sterile. 



Then we want information as to what varieties are good pollinizers 

 for our best pears, apples, plums, and cherries. 



It would be of interest to note how long individual flowers of the 

 different kinds of fruit remain open, and in self -fertile varieties to ascer- 

 tian whether the result is identical, whether the pollen it is fertilized 

 with, is of the same flower, the same tree, or a distant tree of the same 

 variety. 



Experiments in the pollination of fruits have the advantage of being 

 inexpensive in material (muslin or paper bags, brush or forceps, methy- 

 lated spirit, labels, notebook), and could be carried out by botanists, 

 fruit-growers, or gardeners. 



With our present knowledge it seems advisable in planting to alter- 

 nate the variety in each one or two rows, choosing varieties that 

 blossom at about the same period. 



In gardens do not plant single trees of a kind, all trees of one 

 variety together, but rather intermix the varieties. 



Keep hive hees, especially in suburban gardens and on large fruit- 

 farms, and in districts where large areas of the same kinds of fruit are 

 grown. 



