e inferior seedlings were discarded and limited data only are
available. Furthermore, the filbert breeder is interested primarily in
the worthwhile material that may be taken from populations of known
parentage.

Assuming that we have a fairly good nut productiveness is the most
important characteristic in a filbert. If the plant is productive it
must of necessity be reasonably vigorous and hardy. For that reason much
emphasis has been placed on productiveness in the final evaluation of
the selections.

The selections in Table 2 are from the U.S.D.A. Bixby plants which were
the first to fruit at Geneva. Considerable variation in productiveness
is evident in the different populations. Rush x Kentish Cob and Rush x
White Aveline selections were only about half as productive on the
average as Rush x Barcelona, Bollwiller, Red Lambert, and Daviana. Rush
x Italian Red also failed to produce high-yielding selections. In a
later planting in the same orchard, as shown in Table 3, the Rush x
Kentish Cob selections performed no better, the Rush x Red Lambert
selections outyielding them by a substantial margin. The Barcelona x
Italian Red selections were very low yielding.

In orchard 22, as shown in Table 4, where Rush and Barcelona are crossed
with the same varieties, the resulting selections from the Rush crosses
are about one third more productive if mean yields are considered, or
one-half more productive if only highest yielding selections are
considered than with the Barcelona crosses. Cosford has been outstanding
in transmitting productiveness in crosses with Rush, Italian Red, and
Nottingham. Rush x Kentish Cob selections in this orchard as in the
other planting, were only about one half as productive on the average.
In the crosses with Barcelona the combination with Medium Long, Red
Lambert, and Italian Red were considerably more productive than crosses
with Purple Aveline, Halle, Daviana, and Bollwiller.

The Kentish Cob x Cosford cross was less productive than most of the
other combinations made. Kentish Cob definitely appears to transmit
unproductiveness when crossed with Rush, Barcelona, and Cosford.

In orchard 8 as shown in Table 5, the trees soon became very crowded as
the discards were not removed and the yield records were less reliable
than in the other plantings.

Winterkilling of catkins were recorded on the selections for several
years. In early April the percentage of winter-killed catkins was
recorde