crop harvested in 1949, however, is actual.

Table No. 1 gives the average weight in grams of the sample nuts. The
Duke, (first prize) was the largest nut of all, in 1945, averaging just
over 27 grams; but the Orth, in 1948, averaged almost a gram more. The
Kuhn, which was the smallest of the eight nuts in 1946 and again in
1950, was the largest nut in 1949, and its size in 1949 was exceeded
only four times by any of the other nuts during the contest. The nuts
were large in size during the off year when only a small crop was
produced and they were small when there was a heavy crop.

In table No. 2 the weight in grams of the kernel recovered on first
crack, secured without the aid of nut pick, is recorded. In this
comparison the Duke, because of large size, might be expected to be an
easy winner and it was in 1946 and in 1950; but in 1948, though second
in average weight of nut for that year, it was in fifth place in
recovery of kernel at first cracking.

Table No. 3 records the average weight in grams of the kernels. Here the
Duke, due largely to its size, is a consistent winner in all three years
it produced nuts. However, in 1949, a small crop year for the Kuhn, the
nuts of this variety were large and contained more kernel than the Duke
did in 1948 or in 1950.

The per cent of kernel in the nuts as recorded in table No. 4 is
interesting. The Burson, which was the smallest nut in 1947, had the
highest per cent of kernel and also had the highest total kernel content
of any sample in that year. Evidently the per cent of kernel is higher
in well-filled nuts and this is largely determined by the weather and
available food supply late in the season.

A comparison of the numerical score of the various nuts, figured out
according to the T.V.A. score system, is given in Table No. 5. By this
system, no variety had a consistent high score, but each varied greatly
from year to year.

The nut characters studied so far in charts 1 to 5 inclusive have varied
so much from year to year that any judgment based on these characters
for any one year could not be relied upon.

What characteristic of a black walnut, then, can be used in evaluating
it? In table No. 6 the percentage of the total kernel that is recovered
at first cracking is given. Oliver and Penn show considerable
consistency in that they remain above 91 per cent in all samples, but
look at the Kuhn. It was perfect in 1950 but in 1948 only 65 per cent of
the kernel wa