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



The fruit in all three groups of the later Bull strain samples shows 

 a higher percentage of rind than the Lisbon fruit. This difference 

 is greatest in the May 1921 group but is also decidedly marked in 

 the January 1922 group. The samples taken in February 1921 

 show the difference but to a smaller degree than the others. Al- 

 together, it seems that a probable difference exists. 



The only difference in specific gravity between the strains of the 

 Lisbon variety occurs between the fruit of the Dense unproductive 

 and the Bull strains. In May 1921 and January 1922 the specific 

 gravity of the lowest Dense unproductive samples was higher than 

 that of the highest Bull samples. In the February samples the 

 difference was not maintained, but the monthly samples show a 

 significant difference in this respect. Therefore a probable difference 

 exists. 



SUMMARY. 



From the data presented it may be concluded that the following 

 probable differences exist : 



In the Lisbon variety. — The Dense Unproductive strain has a 

 higher specific gravity than the Bull strain. The Bull strain has 

 the highest proportion of rind found in any of the strains of this 

 variety. The Lisbon strain is more highly acid than either the Open 

 or the Dense Unproductive strain. The Dense unproductive strain 

 is also more acid than the Open strain. 



In the Eureka variety. — The Eureka strain is more acid than the 

 Shade Tree strain. Other differences may exist, but the data at 

 hand are inconclusive in these cases. 



LITERATURE CITED. 



(1) Chace, E. M., Wilson, C. P., and Church, C. G. The coniDosition of Cali- 



fornia lemons. U. S. Dept. Agr. Bull. 993 (1921), 18 pp. 



(2) McEwen, G. F. Unpublished manuscript. 



(3) Mellor, J. W. Higher mathematics for students of chemistry and physics. 



New ed. Longmans, Green & Co., London, 1919. 



(4) Pearl, Raymond, and Miner, J. R. A table for estimating the probable 



significance of statistical constants. In Maine Agr. Exp. Sta. Bull. 226 

 (1914), pp. 85-88. 



(5) Shamel, A. D., Scott, L. B., Pomeroy, C. S., and Dyer, C. L. Citrus-fruit 



improvement: A studv of bud variation in the Eureka lemon. U. S. 

 Dept. Agr. Bull. 813 (1920), 88 pp. 



{6) . Citrus-fruit improvement: A studv of bud variation in the 



Lisbon lemon. U. S. Dept. Agr. Bull. 815 (1920), 70 pp. 



(7) "Student." The urobable error of a mean. In Biometrika, 6 (1908): 



1-25. 



(8) Wilson, C. P., and Young, C. O. A method for the determination of the 



volatile oil content of citrus fruits. In J. Ind. Eng. Chem. (1917), 9: 

 959-964. 



(9) Wood, T. B. The interpretation of experimental results. In J. Bd. Agr. 



(London) Sup. 7 (1911): 15-37. 



