PLANT-BREEDING METHODS 433 
who advocates the organization of a recognized board of review which 
will decide after trial and comparison, whether a submitted sample is 
worthy of standing as a new variety or simply as a strain or stock. 
Myers has conducted extensive strain tests of tomatoes and has 
reached the conclusion that the best way to insure success 1n procuring high 
yielding strains of vegetables is to secure seed a year in advance of the 
time it will be needed and submit it to a preliminary test. The difficul- 
ties encountered in selecting the most profitable strains within a variety 
involve the same sources of experimental error as are met in variety 
testing. The necessity of distinguishing between heritable and non- 
heritable variations calls for individual plant sclection and pedigree 
Fria. 180.—Increase plots of one-thirtieth acre each. Selected strains of wheat from 
the nursery are tested in these plots for 3 years. (After Montgomery.) 
culture methods. Plant-to-row tests and subsequent plot tests of the 
progeny of individual wheat plants are shown in Figs. 178-181. In all 
such work the use of loose leaf record blanks is advantageous. Two 
forms of blanks used in testing pure line selections of oats are shown in 
Figs. 182 and 183. Plant-to-row strain tests are still generally used 
with cotton, corn and other cross-fertilized plants, but Hartley has 
pointed out the importance of reducing experimental error to a minimum 
in testing corn. 
Factors That Affect Experimental Results.—In discussing the 
standardization of field experimental methods Piper prepared the fol- 
lowing list of factors affecting experimental field work with plants, 
advising that they be published in connection with any series of field 
experiments where relative yield is the object sought. 
28 
Digitized by Microsoft® 
