'iO AMERICAN HOME GARDEN. 



c;ultivatoi's, gives importance to the production of improved va- 

 rieties. 



This is most commonly effected by careful and long-contin- 

 ued SELECTION, to which high culture should be superadded, 

 in which we continually choose the most perfect or eiu'liest 

 plant, or fruit, or pod from which to obtain our seed. 



Upon this latter principle rests the old familiar rule of tak- 

 ing for seed the cucumber or melon growing nearest to the 

 root, etc. This rule, however, is seldom rigidly adhered to, and, 

 if it were, would natm'ally tend to produce an earlier but small- 

 er-fruited variety than the original. 



Perhaps the most promising course for improvement is to 

 choose the second, and generally finer fruit for seed ; or, if the 

 olyect be simply to avoid depreciating the variety, the wdrole 

 crop, being left ungathered from the first, wall yield satisfac- 

 tory seed. Unless, indeed, it should happen that from pecul- 

 iar circumstances the plant makes a very extended or a sec- 

 ond growth, in which case the earlier product alone should ])e 

 permitted to seed. 



j^ew and improved varieties arc also sometimes obtained by 

 careful and intelligent INTEEJIIXTURE, in which we aim to 

 combine the desirable qualities of Ijoth the old varieties in the 

 new one we expect as the product. This valualile result is 

 also sometimes effected accidentally. Lr such intermixture the 

 general nde is that the product will have the form and ap- 

 pearance of the fertilizer, with the character or peculiarities 

 of the fruit-bearing plant. To illustrate this : A^ery early peas 

 are generally small. Suppose we desire to produce a variety 

 in which the seed should be larger, but the crop not materially 

 later. Then, on the general rule given, we may fertilize the 

 cede nulli with the Spanish dwarf, and expect to accomplish 

 our purpose ; but if we fertilize the latter with the former we 

 ought not to expect success, though it is not inconceivable that 

 we might succeed, from the accidental concurrence of certain 

 occult causes or comljinations connected with the previous proc- 

 esses through which these varieties may have passed iu ar- 

 riving at their present state. 



Intermixture is effected only between kinds that blossom at 



