METHODS OF TESTING DIFFERENT VARIETIES. 



37 



in the United States and that already considerable confusion in re- 

 gard to the names of the varieties had arisen. With the increase in 

 the number of varieties grown the confusion has become much 

 greater as there are now several or many varieties for every seed 

 color. Some endeavor has been made to ascertain the actual identity 

 of varieties which have been described or mentioned in publications, 

 but in many cases it is no longer possible to do this, and in other 

 cases the identification can be only approximate. 



METHODS OF TESTING DIFFERENT VARIETIES. 



The comparative data concerning the numerous varieties here 

 given are based largely on the testing conducted at Arlington Farm 

 since 1903. During the first two years the work was conducted by 

 Mr. C. R. Ball; during 1905 to 1910, inclusive, by the writer, with 

 the assistance of Mr. H. T. Nielsen, in 1905 to 1908, and pf Mr. W. J. 

 Morse in 1909 and 1910. Most of the testing has been in cultivated 

 rows 3 feet apart (except in 1909 and 1910 when they were 3 J 

 feet) , the rows being usually 8 rods long, but in many cases shorter 

 owing to the small supply of seed. As new varieties have been in- 

 troduced each year the period of testing varies from eight years 

 down to one }^ear. In many cases the variety did not mature seed 

 at Arlington, but of most of these enough seed was grown in the 

 greenhouse to plant a second year. The collection of 1909 was 

 almost complete, and except where otherwise indicated the compara- 

 tive descriptions are based mainly on the behavior during that sea- 

 son, which was an exceptionally dry one, so that the plants were 

 rather smaller than usual. The characteristics of the varieties dur- 

 ing the different seasons held remarkably true, so that many of their 

 distinctive peculiarities of habit could easily be recognized. The 

 principal obscuring fluctuation due to better soil or more favorable 

 season was the tendency of the upright-growing varieties to lodge. 

 Apparently, there are no phenomena exhibited in cowpeas that 

 could be called new-place effects unless changes in seed color (as 

 exemplified under No. 16167) are such, which is at least doubtful. In 

 all cases where such seed color forms or varieties were selected out 

 they bred true. No cases of natural hybridization in cowpeas have 

 been observed at Arlington, notwithstanding the fact that numerous 

 varieties have been grown in close, proximity for several years. In a 

 few cases where diversity occurred from the same lot of seed it 

 seems clearly explainable by previous hybridization. 



While this method of testing does not give comparable yields, it 

 does furnish an excellent method for comparing the varieties as to 

 habit, disease resistance, vigor, fruitfulness, life period, etc, and it 

 is believed that the conclusions reached as to the most desirable 



229 



