16 BULLETIN 1255, U. S. DEPARTMENT OF AGRICULTURE. 



differences. When the numbers of items in the series to be com- 

 pared were about the same, use was made of ''Student's'' method (7), 

 casting out the items most favorable, to the desired result to make 

 the items even in number. None of these methods, however, were 

 satisfactory in most cases. The method of Dr. George F. McEwen, 

 of the Scripps Institution for Biological Research, at La Jolla, Calif., 

 therefore, was used (?) and Doctor McE wen's advice was followed in 

 the treatment of all comparisons. Even when all the methods 

 showed significant differences, these differences were not considered 

 strongly positive unless a majority of the group samples (Tables 3 

 and 4 x i showed the difference. No case was considered strongly 

 positive when any one group of the three or four showed a reversal of 

 the tendency. 



EUREKA VARIETY. 



The samples of the Eureka lemons show few marked differences. 

 The greatest difference is in the acidity of the fruit of the Eureka 

 and Shade Tree strains (Tables 1 and 3). All the monthly samples 

 of the Eureka strain trees had a higher acidity than any of the Shade 

 Tree strain, excepting tree 34-74-13. Of the five samples obtained 

 ?rom this tree, four were unusually high in acidity as compared with 

 other samples from trees of the same strain. 



The next greatest difference is that in acidity between the Eureka 

 strain and the Dense Unproductive strain (Table 3). As no monthly 

 samples of the latter were obtained, the comparison is based on the 

 samples obtained in 1920-21. The difference in acidity between 

 samples of these strains taken in June 1920 and June 1921 is highly 

 significant. Although the February and November samples do not 

 maintain the differences, they do not show a reversal. The best 

 that can be said of this difference is that it is possible. 



At first glance a difference may seem to exist in the specific 

 gravity of the Eureka and Shade Tree strains, but further study 

 shows that this is only barely possible. The results from the monthly 

 samples are not convincing, for while the difference is slightly in 

 favor of a higher specific gravity for the Shade Tree fruit, it is not 

 sufficiently great to be conclusive. The June 1920 and June 1921 

 samples show a difference in favor of the Shade Tree strain. The 

 February 1921 group shows no differences and the November group 

 shows a decided reversal. 



No consistent differences in specific gravity of the fruit, per- 

 centage of rind, or acidity of the juice between the strains of the 

 Eureka variety are shown. Some differences may exist in the con- 

 stituents determined on the monthly samples, but these data were 

 not derived from a sufficient number of trees to make conclusions 

 from them possible. 



LISBON VARIETY. 



The greatest difference shown by the Lisbon variety is that of 

 acidity of juice between the Lisbon and Open strains. As no monthly 

 samples of the Lisbon strain were collected, conclusions are based 

 on the later samples (Table 4). In each of the four groups the 

 chances are very high that the difference occurring is significant. 

 In November 1920 the lowest acidity of the Lisbon strain fruit is 

 higher than the highest of the Open strain. In February the lowest 



