112 
these calculated crops figured on the basis of the actual prices obtained 
during the investigational period, and considering the variations in 
grades and sizes produced per acre' by the two groups of trees, would 
be $311.32 and $84.17, respectively. 
This correlation of heavy yield and superior commercial grade in 
the production of trees of the Valencia strain is very important from 
the standpoint of the stabilization of the variety through bud selection 
based on performance records. Figures 4, 5, and 6 indicate that the 
propagation from the highest yielding trees means not only heavier 
production, but also improvement in the grade and size of fruit. 
7257V S/f/IW£-sr&?C?0Cte/A/& 77&2F5 7ZW ZOPKE&T 0&XX/C//VG 77QEZ& 
\32.0 2S& SZ3 ■BHHKEnBHflHH 
/oo.o /oo.o 
\6&.3aVGV4M 4&2\ 
24.0&7?tMQ4/?£3$./ 
Pig. 5. — Diagram showing the variations in the commercial sizes and grades of fruit on the 10 heaviest 
producing trees in comparison with the 10 lowest producing ones in the investigational performance- 
record plat of Valencia oranges. The percentages are calculated from the number of fruits in the average 
crop per tree for a 4-year period. Figure 4 shows similar variations based on the weight of tbe crop. 
Figure 7 shows graphically the relative number of variable fruits 
in the total crops of trees of the various strains of the Valencia orange 
occurring in the investigational performance-record plat. The dia- 
gram illustrates very well the comparative freedom from such fruits 
which characterizes the trees of the Valencia strain of this variety 
and the larger proportions of variable fruits which occur in the poorer 
and more undesirable strains. It is interesting to note that on 
the trees of the Barren strain, in addition to the almost entire lack 
of a crop of any sort, 84.3 per cent of what was produced was made up 
of sporting fruits. In counting the variable fruits on a tree only 
those showing some variation from the general type of the fruits 
