6 BULLETIN 697, U. S. DEPARTMENT OF AGRICULTURE. 
senior writer in 1909 revealed the fact that there existed in the grove 
a striking diversity in the characteristics of the trees. Out of a total 
of about 500 trees, 123, or nearly 25 per cent, produced fruits having 
from 30 to 90 seeds in each fruit. The trees bearing these very seedy 
fruits were found to have a characteristic drooping habit of growth, 
from which it soon became possible to identify them wherever they 
occurred in the grove. The fruit from this grove had been guaranteed 
by the owner to be commercially seedless, and the occurrence of the 
seedy fruits in the crops, when noticed by the buyers, led to a con- 
siderable depreciation in their value on the markets. Upon the dis- 
covery that the seedy fruits were borne only by trees of a certain 
character, these fruits were eliminated, except those occurring as 
individual fruit and limb variations, by top-working the trees which 
produced them, using for this purpose buds selected from trees of the 
true Marsh strain, which is commercially seedless. One of the trees 
of the Marsh strain in this grove is shown in Plate II and one of the 
trees of the Rough Seedy strain in Plate III. 
An interesting characteristic of the production from individual trees 
was revealed by the tree-census studies in the Marsh grove first studied. 
Some of the trees bore very heavy crops, while others, very similar in 
general appearance to the productive trees, bore only a few fruits. 
Further observation of this condition during the following years re- 
vealed the fact that some of the trees possessed the habit of bearing 
heavy crops one season and light ones the next. The trees having 
this habit have been called alternate bearers, from the fact that they 
usually bear full crops only every other year. 
Another strain of trees was found in this grove bearing fruits having 
a modified pyriform, or bell-like, shape in contrast with the flattened, 
or oblate, shape of typical fruits of the Marsh strain. These trees 
have been called the Bell strain on account of the shape of their fruits. 
This shape, from the market standpoint, is not so desirable as the 
flattened one of the Marsh strain, and the fruits are likely to be later 
in ripening than those of other strains. 
The fruits borne by the trees of the Marsh strain have a compara- 
tively thin rind of very smooth texture and waxy-white appearance. 
Some of the trees in the performance-record plat were found to produce 
fruits having very thick rinds with a rough or coarse texture and yel- 
lowish color. The fruits of this Rough strain, as a rule, have a shape 
more nearly round or globular than those of the Marsh strain, while 
the flesh, juice, and other characteristics are somewhat similar to 
those of that strain. 
Individual-tree census observations made in other groves located 
in various citrus districts of southern California have brought out 
clearly the fact that the trees in these orchards vary in about the same 
manner as those in the grove described above. More than 25 per cent 
