8 4 



MY GARDEN. 



sprouted, to place the young plants close to the glass, that they may 

 have the full effect of light. Every seedling is an individual, having 

 certain characteristics, deviating in some degree, and within certain 

 limits, from a fixity of type. By selecting seedlings having particular 

 characters, and again by continually selecting from their progeny, the 

 gardener obtains those deviations from the original type which are 

 known by the name of florists' flowers, or gardeners' fruits. 



A question has arisen whether selection alone is sufficient to pro- 

 cure new plants, or whether it is desirable to cross the seed of one 

 plant with the pollen of others. Mr. Rivers tries the latter plan 

 to improve our peaches and pears, and he crosses a plant having one 

 good property with the pollen of another having other desirable 

 qualities. On the other hand, I am assured by raisers of florists' 

 flowers that it is generally better to rely upon sowing simply the seed 

 of good sorts and upon selecting from their produce. In the present 

 state of our knowledge we had better try both methods, when w r e 

 desire to raise new varieties. Many of the best varieties of fruits 

 are certainly natural sports. 



Sometimes selected plants of high quality are again propagated 

 by seed, such as our choice varieties of peas and beans. More often, 

 however, we have to multiply the original plant : to such an extent 



Fro. 



. Sucker. 



FIG. 104. Layer. 



can this be done, that the delicious Jargonelle pear is supposed 

 to have been raised by the Romans, and handed down to us by a 

 continuous propagation. 



Various methods are practised at my garden for the multiplication 



