24 EXPERIMENT STATION. [Jan. 



indicated should be chosen at the start, thus increasing the 

 chances of securing the very best that the variety afiords. Ex- 

 perience has shown most conclusively that the character of an 

 individual plant is small indication of the value of its progeny, 

 and the selection that proves finally to be the best nuiy very 

 likely come from one of the less promising plants of the origi- 

 nal selection. 



The foregoing has been written as applicable to the garden 

 peas alone, for our own investigations along this line have been 

 confined to that plant. The question now arises whether the 

 same methods may be applied to other plants as well. 



Our common garden peas are almost invariably self-polli- 

 nated, and there is reason to believe that the same general 

 method will succeed equally well wath all self-pollinated plants ; 

 with naturally cross-pollinated plants some modification will 

 be necessary. Among our common garden and field crops the 

 following are believed to be almost always self -pollinated ; 

 peas, beans, tobacco, wheat, oats, barley. Most others are often 

 or generally crossed in the field, and for such the method must 

 be modified to make sure that the pollen for the improved 

 races comes from the best possible ])lant. 



Assuming that naturally cross-fertilizing plants are made up 

 of slightly differing strains, as seems to be the case with peas, 

 a field of plants will be made up largely of individuals that are 

 crosses between the several strains ; consequently the task of 

 separating superior strains becomes a complicated one. Fur- 

 thermore, it is probable that with some, and perhaps many, 

 naturally crossed fertilized plants, crossing between distinct 

 strains or varieties results in the first generation, in greater 

 vigor and more productive plants. Thus is the problem of 

 securing the very best still further complicated. 



Nevertheless, great improvement can be made by following 

 substantially the scheme outlined for self-fertilized plants, but 

 taking into consideration the pollen parent, and pollinating from 

 superior individiuils cither by hand or planting individuals 

 which are desired as parents together in an isolated location. 

 The latter method is often used with corn and other wind- 

 pollinated plants. Tn any event it is always of the utmost 



