108 BULLETIN 697, U. S. DEPARTMENT OF AGRICULTURE. 
Trees of the Bell. Rough, and Unproductive strains are all undesir- 
able from the commercial standpoint, and should be eliminated in 
bearing orchards and avoided for propagation, except perhaps for 
purely experimental purposes. Fortunately, there are but few trees 
of these strains in most of the established California orchards of the 
Marsh variety, so that their elimination is a comparatively simple 
and inexpensive matter. 
It is probable that by means of individual-tree records in bearing 
orchards covering two years most of the trees of inferior strains can 
be located, so that they can be top-worked with select buds of the 
Marsh strain. 
r/vr ft/GffrsT p/towcm TPrrs , /7r <. rw /owrsT ppoduo/mc TPrrs 
prp crAtr or a pa oari> box prp ma ^ /Z£r ^ pm cavr or a PAcxrp box prp rprr 
65.2 J/2 28.6 mSA 
$6 /S.S B 
80 /Z5HB 
64 25.0 mwm 
48 2 s.o mm 
32 28/ mwm 
24 58.4 mwmmmm 
4.ps PACxrr soxrs \ AVf * A p§/rMr£ C * 0P \ ^6 packo? boxes 
146 PAOPrP 30X55 OP | 7ff£ /X05T D5S/PABL5 S/ZJ5S \ 0.68 PACKED B0X5S OP 
5.44 P5P- J C5//T Or r/ff OPCP I SO, 64, AND 48 \ 34. 1 P5P CFNT OP TH5 CPOP 
340.0 PACXrPSOXcS^ 00 " 07 ^™^ \ /S6 6 PACX5P BOXrS 
t [ ACTUAL VALU5 P5P ACPB } t ?A4 /^ 
2 85J.UU [ pfi/VrPrO AT T/ir PACX/NG /1QU55 I * JQ * Jt ° 
F20S27HP 
Fig. 14. — Average number of fruits of the various commercial sizes produced annually during the 6-year 
period, 1910 to 1915, inclusive, by the five highest and the five lowest yielding trees of Marsh grapefruit 
in the investigational performance-record plat B in a grove planted in the fall of 1903. This production 
is expressed as percentages of a packed box calculated from the number of fruits of the different sizes 
contained in a box. The percentage of most desirable sizes is also stated, and the production on the 
acre basis is shown and its value calculated from the actual returns received for the fruit from this grove 
during the 6-year period. 
THE UNINTENTIONAL PROPAGATION OF UNDESIRABLE STRAINS. 
The trees of undesirable strains in established orchards have been 
propagated unintentionally through a lack of careful selection of 
bud wood. Heretofore, all trees of the variety have been con- 
sidered to be equally good for propagation, and little or no selection 
of trees or bud wood has been practiced. 
As an illustration of this condition, the experience of the senior 
writer in 1910, in the first grapefruit orchard selected for individual- 
tree performance-record work, may be cited. In this orchard, as 
soon as the owner found that the seedy fruits were mainly borne by 
