eriment was
not possible. However, the stock was grown in thoroughly mixed soil in a
coldframe and differences in performance of seedlings could hardly be
attributed to soil heterogeneity.

Buds from the five parent trees were placed on the five lots of their
own seedlings in all combinations of budwood and stock. The work was
done during the first week of September when the bark of both budwood
and stock was slipping yet growth had slowed down to some extent. Buds
were placed about two inches below soil level on the one-year-old
seedlings and the soil pulled back to cover the buds. Budding was done
by means of the familiar shield or T-bud method and rubber budding
strips were used as a wrap. Budwood was shipped from Albany, Ga., to
Beltsville, Md., and was damaged somewhat by high temperature in
transit, a factor which may be partially responsible for the overall low
percentage of buds that grew.

In referring to the results presented in table I, it will be noted that
considerable variation occurred in the performance of the five lots of
seedlings as stock, as well as in the take of buds from the five parent
trees. The totals in the last column on the right are all equivalent to
percentage since 100 buds were placed on each lot of seedlings. In like
manner, the totals in the bottom line are all equivalent to percentages
since 100 buds of each parent tree were used.

Seedlings of stock D were decidedly inferior to seedlings of stock C in
take of buds, and both of these lots of seedlings originated from large
nuts. Also, scion e gave a significantly lower take of buds on all lots
of seedlings than scions c or d. The scion e tree produces small nuts
whereas the scion c and d trees produce large nuts. Scions a and b are
intermediate in take of buds, and the source trees both produce small
nuts.

Discussion

At least one significant interpretation may be made from the results of
this experiment, that may partially explain the difficulties encountered
heretofore in propagating chestnuts. Using the take of buds as a
criterion it can be stated that in this experiment the five lots of
seedlings from known parents differed in their performance as stocks.
Moreover, the five parent trees used as a source of budwood differed
among themselves in the capacity of their buds to grow when placed on
comparable lots of stocks. If these results are correctly interpreted it
is clear that both the stock and the scion may influence th