80 BULLETIN 813, TJ. S. DEPARTMENT OF AGRICULTUEE. 
tion of the "undesirable strains has been the natural tendency of the 
bud cutters to secure bud wood from the trees producing the most 
suckers. In other words, they cut their bud wood from the trees 
where they could most easily and quickly secure the largest supplies. 
The method of using fruit-bearing bud wood for the propagation 
of the citrus varieties, as shown in figure 18, has been evolved in the 
course of these investigations and has largely eliminated the danger 
of the unintentional propagation of the vegetative strains of the 
Eureka lemon. The use of fruit-bearing bud wood naturally results 
in bud cutters securing most of their bud wood from the most heavily 
fruited trees, because more fruit wood is available on such trees than 
on trees of the poor fruiting strains. When performance records are 
used as a basis for parent- tree selection and only that bud wood used 
for propagation which bears typical fruits, the danger of propagating 
the undesirable strains is largely eliminated. 
THE ISOLATION OF STRAINS THROUGH BUD SELECTION. 
The orchards of the Eureka lemon variety in California are made up 
of trees of many diverse strains which have originated from bud 
variations. The description and characteristics of some of the 
important strains have been presented herewith. It is obvious from 
these data and observations of the behavior of the trees of the different 
strains that some of them are not worthy of propagation, because of 
low and inferior production. Other strains mature their fruit during 
poor marketing seasons when the prices for lemons are usually low. 
The trees of some of the strains produce fruits low in acidity, of un- 
desirable shapes for packing, of coarse ugly texture, with little or no 
juice, or with some other undesirable characteristics. It has been 
proved that a mixture of strains in orchards is very undesirable and 
is likely to make the maintenance of such orchards unprofitable. 
After determining the relative value of the different strains for 
commercial production in California, the next step in these investi- 
gations was the attempt to isolate each of the important strains 
through bud selection. This study .was made by propagating from 
typical trees of each strain, selecting the bud wood on the basis of 
tree-performance records and intimate tree knowledge. Only fruit- 
bearing bud wood was used for this pupose, with typical fruits of the 
strain attached to each bud stick. 
The first trees grown from these propagations are now 4 years 
old from planting. Individual-tree performance records are being 
secured from some of these progenies. While the performance- 
record data obtained in the course of this work are incomplete as yet, 
there is sufficient evidence in hand to warrant the statement that 
each of the important lemon strains has been isolated through bud 
selection. This means that the strain characteristics in the progenies 
