Analysis of Wild American Races of Oenothera (Onagra).

HERE was a time when the evening primrose was the despair of the T geneticist. It still is the despair of the taxonomist, who finds it difficult to classify and delimit species in this genus on the basis of phenotypic criteria alone. This difficulty is now known to be due to the fact that most Onagras are, from the phylogenetic point of view, not single, but dual.entities. They are usually composites of two distinct, separable, and, as a rule, phylogenetically diverse components or genoms known as the RENNER complexes, one of which may largely dominate over the other phenotypically. These complexes are continuing entities of a more or less permanent nature, being passed intact from generation to generation in selfed lines. The peculiar cytogenetic behavior of Onagra, which is responsible for this duality, is now familiar to all geneticists and will not be discussed here. The fact should be emphasized, however, that the genoms which are associated to form a given species are, in most races, very different from each other genetically, and probably of diverse phylogenetic origin. Thus, in tracing the affinities of a given race, one must consider two distinct lines of phylogenetic descent. The relationships of both genoms must be taken into consideration. This is not an easy thing for the taxonomist to do, however, since one genom, by reason of its dominant genes, often effectively masks the other. Furthermore, races composed of two diverse genoms breed as true as do those whose genoms are essentially similar, thanks to their balanced lethals. Hence a taxonomist cannot tell by observation, or even by inbreeding, which races contain genoms of diverse phylogenetic origin and which contain genoms which are closely related. This can only be determined by cytogenetic means. One example will show the necessity of the cytogenetic approach in tracing the phylogenetic position of a race in Onagra. “Iowu I” is a race of the biennis type. It has broad, dark green crinkly leaves, small flowers and other so-called “biennis” characters, and it breeds true to type. So far as one can see phenotypically, it is biennis and nothing else. If one crosses it, as egg parent, to the kookeri of DEVRIES, the F1 hybrids resemble “Iowa I ” very closely. But if one crosses it as pollen parent to hookeri, a very different sort of hybrid results, one which looks in no respect like a biennis, which differs from its reciprocal in almost every external character. Both