96 LUTHER BURBANK 



If the three or four different iqualities, 

 improvement of which is desired, are not com- 

 bined to best advantage in any single individual, 

 then it is sometimes necessary to select an indi- 

 vidual for each quality, and to carry forward 

 three or four lines of experiment at the same 

 time. 



It will be recalled that in developing a special 

 variety of small sweet canning pea, with the 

 qualities of uniform ripening, small seed, and of 

 seeds of uniform number and equal size in the 

 pod, I was enabled to find these qualities exhib- 

 ited in such combination that the experiment 

 went forward rapidly, so that in the course of 

 six generations I had developed precisely the 

 variety of pea desired. 



But it will also be recalled that half a dozen 

 other lines of experiment were carried forward 

 at the same time, using the same group of peas, 

 that led finally to the production of as many 

 quite different varieties, characterized by large 

 size of seed, by lozenge-shaped seeds, and the 

 like. And these secondary experiments were 

 carried out without in any way interfering with 

 the primary one. It was merely that, in search- 

 ing among the different vines, I could not fail 

 to notice individual plants with interesting char- 

 acteristics, and nothing more was required than 



