BUD VARIATION IN THE WASHINGTON NAVEL OEANGE. 9 
available for consideration in the location of similar plats of other 
fruits. 
The performance-record plats of the Washington Navel orange 
trees usually have included 100 or more trees. In some cases a 
smaller number, varying from 5 to 50 trees in a plat, has been used. 
All the plats on which detailed performance records have been ob- 
tained are situated in two adjoining orchards, and in their selection 
all known interfering environmental factors in individual-tree com- 
parisons were avoided as far as possible. By the selection of these 
plats in uniform soil areas and in sections with the same elevation, 
like exposure, and comparable drainage conditions, it is believed that 
the results presented are less affected by fluctuations due to variations 
in soil fertility, moisture, or other environmental factors than would 
be the case were the records of all the trees in the orchards used for 
comparison. Records of a total of 743 Washington Navel orange 
trees are now being kept in this work. 
METHODS OF KEEPING PERFORMANCE RECORDS. 
In describing the methods developed in the course of these investi- 
gations for recording the behavior of individual trees in the perform- 
ance-record plats, the methods of procedure are discussed in the 
order in which they are carried out. This arrangement will be of 
interest in planning similar work in other locations or with other 
fruits, in order that the performance-record work may be done in the 
most natural and logical manner. 
TREE NUMBERS. 
The individual trees in the performance-record plats were given 
numbers, so that the data secured from the study of each tree and its 
fruit could readily be kept separate during the entire period of 
observation. In the beginning, the trees were given consecutive 
numbers. These numbers were stamped in metal tags which were 
attached by wires to the limbs of the trees. This method of num- 
bering soon became cumbersome and impracticable. 
As a result of a careful study of this matter, a method of tree num- 
bering was evolved which has proved to be satisfactory and has been 
adopted extensively in individual-tree performance-record and re- 
lated work. By this method each tree is given a number consisting 
of three parts: (1) The number of the block or division of the orchard ; 
(2) the number of the row in the block; and (3) the position of the 
tree in the row, always counting from some fixed point, as, for in- 
stance, the irrigation head. A tree located in block 5, row 6, and 
the seventh tree in the row has the number 5-6-7. Where several 
different orchards are under observation, the tree number in the 
performance-record notes is preceded by the number or n^ame of 
