INHERITANCE THROUGH VEGETATIVE PROPAGATION. 3 



productiveness and shape of fruit are inherited from tree to tree, or 

 transmitted 'by strains of trees to their progeny, perhaps specific 

 gravity, percentage of rind, and acidity of juice are also transmitted 

 properties. 



Elements of composition are more readily influenced by environ- 

 ment than are many physical characteristics of fruit, so that it is 

 difficult .to prove that composition is inheritable. Decided variation 

 in the composition of fruit from a single tree occurs, two fruits from 

 the same spur often differing markedly in composition. The position 

 of the fruit on the tree, the quantity of water used in irrigation, 

 fertilization, cultivation, and the stage of maturity of the fruit when 

 gathered, all influence its composition. By careful methods of sam- 

 pling, and the selection of healthy trees which have received identical 

 treatment, the influence of these factors maybe reduced to a minimum. 



EXPERIMENTAL PROCEDURE. 



SAMPLING. 



The original plan was to select a few trees of each of the well- 

 defined strains isolated by Shamel and his coworkers (o, 6) , together 

 with some single trees of sporting strains, and to ascertain by the data 

 obtained from monthly sampling and analyses the differences which 

 might be expected in the progeny. This plan, however, had to be 

 modified, for it soon became apparent that trees of nonproductive 

 strains would not mature enough fruit at certain seasons of the year 

 to permit satisfactory sampling. Sampling was continued in all 

 cases, however. Although some of the data obtained may not be 

 strictly comparable where strains of trees are being considered, they 

 may be useful in studying inheritance in individual trees. 



Study of the data derived from analyses of samples over an ex- 

 tended period showed that not enough trees had been selected to 

 make certain that the errors due to variation between trees of the 

 same strain had been obviated. As it was impossible to increase 

 the number of trees and maintain the analyses on the original scale, 

 the number of determinations on each sample was reduced, samples 

 were taken less frequently, and as many trees as possible were in- 

 cluded. 



The monthly samples consisted of from 18 to 24 fruits. These 

 were as nearly as possible representative of the commercial fruit upon 

 the tree and were selected for size without regard to color. They 

 were packed in cartons and mailed to the laboratory at Los Angeles, 

 where analysis was begun, usually within 48 hours after picking. 

 When any delay was necessary, the fruit was left in the cartons, 

 which were placed in storage at from 35° to 45° F. 



In the second period of the work, samples from several trees of 

 each strain, consisting usually of 25 or more fruits, were taken. These 

 samples were subdivided into lots of 5 to 7 fruits each, and each lot 

 was analyzed separately. This method afforded a good opportunity 

 to observe the variability of fruit from individual trees. 



SIGNIFICANCE OF DETERMINATIONS MADE. 



Several points were involved in selecting the elements of com- 

 position to be determined. The feasibility of making the determina- 

 tion on a large number of samples was considered, as well as whether 



