New Guinea Impatiens cultivars, I. hawkeri Bull., are susceptible to hot, windy conditions throughout much of the Midwest and Western U.S. Certain Indonesian Impatiens from Java (I. platypetala Lindl.) and Celebes (I. aurantiaca Teysm.) are much more heat-tolerant. Interspecific hybrids involving Java and Celebes Impatiens with the New Guinea species have been produced, but lack of fertility has been a persistent problem, unless amphidiploids are produced. Because selection is difficult in amphidiploid populations, other methods of recovering fertility have been investigated. Some interspecific hybrid fertility has been obtained from crossing Impatiens `Tangeglow' with a Java × New Guinea hybrid. Evidence will be presented on the role of the Celebes genome in female fertility, and the role of unreduced pollen in the Java × New Guinea hybrid. Approaches to understanding and overcoming sterility in Impatiens interspecific hybrids will be discussed.

The effects of chromosome number on female fertility of interspecific hybrids among Impatiens from Java (I. platypetala Lindl.), New Guinea (I. hawkeri Bull.), and Celebes (I. aurantiaca Teysm.) and the role of unreduced gametes in determining chromosome number and fertility were investigated. Two 24-chromosome interspecific hybrids were used as parents, 'Tangeglow' and 1 7851-1'. The F 1 offspring from this cross had chromosome numbers ranging from 28 to 36. Because 'Tangeglow' is seed-propagated and has normal meiosis, '7851-1' must have produced some male gametes that were incompletely unreduced and other gametes that were completely unreduced. The female fertility of interspecific hybrids with 36 chromosomes and 29 chromosomes was compared. The hybrids with 36 chromosomes exhibited greater fertility than the hybrids with 29 chromosomes. These results indicate that completely unreduced gametes provide for greater female fertility in the hybrid progeny because of the pre sence of complete genomes of Java and New Guinea Impatiens.

Fruit yield in cucumber is limited by the nature of its fruiting habit. Manipulation of plant architecture is one way of increasing yield potential in cucumber. We are incorporating a multiple lateral (ML), sequential fruiting habit into a determinate (de) background. Misclassification of mature plants occurs because the de character is difficult to identify in a ML/de background. It would be valuable to develop an inexpensive laboratory test which would allow accurate classification of ML/de plant types. To develop such a test, we have begun the use of a new DNA marker system, random amplified polymorphic DNA (RAPD). We performed a series of experiments designed to test the reproducibility and predictability of RAPDs. Data indicated that RAPD patterns were not significantly affected by tissue age, infection with powdery mildew or bacterial wilt, or presence of fruit. F1 RAPD polymorphisms segregated in predicted 3:1 ratios. A survey of 6 ML lines and 1 de line with 400 primers identified 4 polymorphisms shared by all ML lines and diagnostic for the de line. No association was found between any of these markers and the de character.