Improving Plants by Selection or Breeding. 



549 



logical adaptations, are not hereditan', and are lost as soon as the plant 

 is transferred again to its normal environment. If, for instance, we 

 desire to produce a bush cow-pea and the selection is carried on in the 

 south with a viny variety, we should search among the plants for the 

 individual which approached most nearly to the bush type, and it is 

 probable that this plant would be just as likely to transmit this char- 

 acter to its progeny as a similar bushy type selected under northern 

 conditions. As ■ a matter of fact it may be that we could recognize this 

 tendency much more clearly in a southern location, where the plants 

 normally produce vines, than in a northern location. 



Another kind of variation about which we as yet 

 know very little, is the so-called bud variations, 

 sports, or bud mutations. Chr}'Santhemum and 

 rose growers know that it is not a very uncommon 



a b 



Fig. 4. — Red Cedar; a, columnar form; b, spreading form. 



thing for a plant to produce a branch which will be entirely different 

 from the remaining portion of the plant. They seem in a large meas- 

 ure to be comparable to mutations except that they originate in a bud 

 change instead of a change occurring in the sexual reproduction. It is 

 probable that they will ultimately be found to be due to similar causes, 

 being produced in the same way. 



Principles of Selection 



The keynote of improvement by selection is the choice of the very 

 best individuals. The discovery of the best individual in any crop 

 under consideration, requires the growing of a large number of individuals 

 under as uniform conditions as possible, so that the experimenter may 

 have opportunity to examine and select the best. Two methods of 



