Sax and Gowen : Productive and Unproductive Apple Trees 299 
or even into the same tree. In general 
the scions from the productive trees 
were no more productive, in fact they 
were somewhat less productive, than 
scions from the unproductive trees. 
The difference in favor of the "un- 
productive" scions is probably not 
significant, but at least there is no 
indication that scions from productive 
trees are superior to scions from un- 
productive trees of a clonal variety. 
In other words the differences in 
performance of trees of a clonal variety 
of apples, in the above case at least, are 
apparently not inherent, but are the 
result of environmental factors. 
The above experiments do not indi- 
cate that unproductive types of apple 
trees are due to inherent differences. 
Let us now examine the data on bud 
variations involving productivity in 
the citrus fruits. 
Shamel and his colleagues (f-i) have 
published four bulletins giving records 
and performance of 151 Washington 
Navel orange trees, 105 Valencia 
orange trees, 117 Eureka lemon trees 
and 121 Lisbon lemon trees. In all 
cases trees of a clonal variety may 
vary greatly in performance. Certain 
tree types or "strains" are often asso- 
ciated with high or low productivity. 
An analysis of Shamel's data shows 
considerable grouping of high and low 
yielding trees, indicating that soil is 
probably an important factor in the 
differences in performance of the vari- 
ous strains. Unproductive trees have 
been top worked with buds from 
productive trees and although some 
of these trees have been top worked 
over eight years no records of their 
performance have been presented. The 
only evidence presented to support the 
occurrence of inherent differences involv- 
ing productivity in clonal varieties of 
citrus trees is the fact that among trees 
of the same variety, age, and under 
similar conditions some are low yielders 
and others are high yielders. 
There is no critical evidence in any 
clonal variety of fruit trees that bud 
variation is the cause of differences in 
productivity. Clonal varieties of ap- 
ples are just as variable as clonal 
varieties of citrus trees, but independ- 
ent workers using different varieties 
have all found no significant differ- 
ences in the progeny of buds from pro- 
ductive and unproductive trees. That 
the results in the citrus, when finally 
presented, may be found to agree with 
those in the apple is indicated by the 
latest recommendations of Shamel. 
(U.S.D.A. Bull. 813, p. 84). 
Until it is shown that buds from 
productive and unproductive trees 
of a clonal variety result in productive 
and unproductive trees respectively, 
the conclusion is not justified that 
differences in productivity of trees 
within a clonal variety are inherent. 
If such differences are not inherent, 
unproductive trees, such as our "Type 
3," cannot be made productive by top- 
working with buds or scions from high 
yielding trees. 
CONCLUSIONS 
A study of 881 Ben Davis trees 
shows that productive and unproduc- 
tive trees are closely associated with 
definite types or habits of growth. 
The productive, or type 1, trees are 
large, open and spreading, with short 
laterals and many spurs. The unpro- 
ductive, or type 3, trees are small 
and upright with slender branches and 
few spurs. Between these two ex- 
treme types are a number of interme- 
diate types. 
Differences in productivity of trees 
of a clonal variety may be attributed 
to soil, root stocks, bud variation, or 
to unknown factors. In the Ben 
Davis orchard, soil was found to be an 
important factor in causing differences 
in type and yield of trees. The influ- 
ence of favorable or unfavorable root 
stocks is also an important factor 
in tree growth and productivity. There 
is no critical evidence that differences 
in productivity within a clonal variety 
are due to bud variation. There is 
then no justification for top working 
unproductive types of trees with buds 
or scions from productive types. In 
the selection of root stocks it is advis- 
