Abstract
Blakely's manuscript name Eucalyptus "rivalis" was applied to a specimen taken from a site about 30 miles south-west of Perth, W.A., where at least two trees apparently identical with the original specimen are growing at the boundary of areas occupied by E. marginata and E. megacarpa. On the basis of conditions of occurrence, morphology, and a progeny test it is considered that E. "rivalis" is a hybrid between E. marginata and E. megacarpa. There is morphological evidence that E. megacarpa and the related E. preissiana have more characters in common with the Renantherae than with any other section in the genus and that these species were incorrectly placed by Blakely in the Macrantherae, an error resulting from the use of a single character, anther shape, in determining affinity. From evidence of morphology and occurrence it is considered that E. kalganensis and E. chrysantha probably represent hybrids of E. preissiana with the renantherous species E. marginata and E. sepulcralis respectively, providing cases parallel with E. "rivalis".