92 EXPERIMENT STATION. [Jan. 



iiR'ters.* Tlic rdiuiiniiiii' Hue C bc-iiig, as aln^idy staled, ulivi- 

 ously a distinct variety, has a vine length much greater than 

 any of the others. 



The two gron})S ahove designated may he spoken of as strains, 

 their component parts being known as lines, each of which is, 

 as before explained, composed of the descendants of a single 

 plant. The means of the several lines vary greatly from year 

 to year, dne to the varying conditions of weather and of the 

 soil of the difierent plots on which the crops were grown. The 

 relations of the mean lengths of the several lines in the same 

 years also vary greatly. Much of this is obviously due to vary- 

 ing soil conditions. They are more consistent with the four- 

 year averages, in 1911, than in the previous years, the only 

 very marked departure being the case of line K, which is much 

 below the average. They are extremely variable in 1910, when, 

 as already stated, the unwise method of applying fertilizer in 

 the row was followed. 



Whether the slight departure of the averages of the different 

 lines of either strain have any significance in inheritance is 

 questionable. Only further testing under more uniform con- 

 ditions would determine this. 



Great diiferences are shown in the mean number of pods per 

 vine. They follow the mean vine lengths only in a general 

 way, and do not show very clearly the segregation into two 

 strains as do the mean vine lengths. This might be expected in 

 consideration of the slight coefficient of heredity of pods per 

 vine already shown. Nevertheless, the average number of ]iods 

 in tlie long-vined strain is about 15 ])er cent, greater than in the 

 other, while the vine length is only about IG per cent, greater. 

 We have here a result of the greater variability of pods per 

 vine over vine length that will be more fully discussed later in 

 this paper. 



' Thi3 explains the division of Excelsior I. in Table VI. 



