A Wilt-Resistant Watermelon for Florida


were eliminated from further trial after the first season. Al-
though resistance has been the principal basis of selection, it
was necessary to pay attention to other desirable characters,
as toughness of rind and flesh quality. The most desirable
types of melons were saved from the strains showing the high-
est percentages of vigorous surviving plants, and occasionally
from conspicuously good plants occurring in strains that were
low in apparent resistance as a whole. The seeds of each
melon were maintained as a unit, and a card record made of
the various characters of the melons from which they were
taken.














 Fig. 6.-View of portion of 1931 experimental field, showing very general
killing by wilt. The whole area was planted, but only a few strains showed
ability to withstand the disease. (Cf. Fig. 1.)

 Since the number of lots of seeds saved each year was too
great to plant all of them in the field the following year, tests
were made in the greenhouse to determine their germination
and comparative resistance in the seedling stage. Those show-
ing up poorly in these tests were discarded. Since only a limited
number could be tested at one time and since it was found
that a single test was often not conclusive, it was not always
possible to test all of the lots of seeds in this manner. Other
eliminations were made on a basis of the melon field trial records
of the previous year. As the work progressed and the total
number of distinct strains decreased, it became increasingly
possible to pay more attention to type, quality, and rind char-
acter of the melons. Through these various means of elim-
inating the poorer strains, the number of original strains main-
tained at the present time has decreased to less than a dozen,