May 8. 1916 OccuYYence of Sterile Spikelets in Wheat 243 



10.4 per cent for the two varieties; phosphoric acid and nitrogen com- 

 bined gave 10.3 per cent of sterile spikelets, while nitrogen and potash 

 gave 5.4 per cent. Since phosphoric acid gave the highest percentage 

 of sterile spikelets when used alone, it, would seem that this element of 

 plant food is largely responsible for the sterile spikelets, as in every 

 combination in which it is used the number of sterile spikelets is greater 

 than where nitrogen and potash are used singly or in combination. The 

 untreated plot gave 6 per cent of sterile spikelets, the lowest for the 

 series except where nitrogen and potash were used in combination, 

 which gave 5.5 per cent. The complete fertilizer gave an average of 

 7.6 per cent of sterile spikelets. From these data it would seem that 

 there is a tendency for phosphoric acid to produce a larger percentage 

 of sterile spikelets than either potash or nitrogen. However, the fairly 

 high percentage of sterile spikelets in the case of the wheat treated 

 with phosphoric acid did not afFect the yield per plant or spike. Under 

 this treatment the yield and quality of the grain surpassed that under 

 either nitrogen or potash. 



CORRELATIONS 



In order to determine what relation might exist between the total 

 number of spikelets per spike and the number of sterile spikelets, the 

 readings constituting the averages for the 25 spikes of each variety were 

 arranged in correlation tables. The beardless varieties form one table 

 and the bearded the other. Thus, the readings were the average of 

 each variety and the array or distribution in the table was made up of 

 varieties. The data were secured from the plants in hills. Since the 

 number of spikelets per spike in a large measure determines the length 

 of spike, the relation found will be closely associated with the length of 

 the spike. In Table V, which includes the beardless varieties, the coef- 

 ficient of correlation between the number of sterile spikelets and the 

 total number of spikelets is 0.543 ±0.054. The bearded varieties show a 

 correlation which is expressed as r = 0.598 ±0.041. It appears that the 

 number of sterile spikelets per variety bears a direct positive correla- 

 tion to the total number of spikelets or the length of head. The varie- 

 ties with the shorter spikes have decidedly fewer sterile spikelets. The 

 relation betvv'een the number of spikelets and the length of spike may 

 not be close, inasmuch as there may be more or less range among varie- 

 ties as to the condensation or closeness of the spikelets on the spike. 

 However, the long spikes are made up of a relatively larger number of 

 spikelets than the short ones, and the actual percentage of sterile spike- 

 lets may be smaller in the long spikes, as will be pointed out later. 



