Oct., 1922] KLAPHAAK AND BARTLETT RESISTANCE TO MILDEW 447 
corn and grasses (Erysiphe graminis DC), and to the hop mildew {Sphaero- 
theca Humiili (DC.) Burr.). He proved the existence of biologic strains 
among the powdery mildews, especially in his work on Erysiphe graminis DC. 
That in many cases resistance to mildew is inheritable is without doubt, 
though little is known of the quality of the resistance or of the genetics of 
the situation. 
No previous work has been reported, as far as we know, on the mildew 
problem as it is presented by the Oenothera cultures. Atkinson (i, 2) 
made observations upon immunity and susceptibility to a downy mildew, 
Peronospora Arthuri Farlow, in connection with his genetical studies of 
Oenothera pycno car pa (susceptible) and Oe. nutans (immune). He published 
his results on the hybrids produced from these crosses, but made only one 
statement in regard to their susceptibility or immunity to this downy 
mildew, namely, that the Fi of the cross Oe. pycnocarpa X Oe. nutans was 
susceptible (i). The results, however, suffice to show that the markedly 
antithetic characters of the Oenotheras, as concerns disease resistance, ex- 
tend to fungi of other groups than the true mildews with which the present 
paper is concerned. Of course, the observations of de Vries (12) upon the 
relative resistance of the mutations of Oe. Lamarckiana to infection by 
Micrococcus are well known, and especially interesting because of the 
mutational origin of disease susceptibility in the case of mut. nanella. 
Material 
At the outset, for the sake of clarity, it will be well to state that general 
observation had indicated that susceptible species when crossed reciprocally 
with immune ones gave only one immune cross. It was not possible to 
get immune hybrids by crossing susceptible parents, and in the case of 
crosses between immune strains, both reciprocals might be immune, or one 
of them immune and the other susceptible. The results could be formulated 
in accord with the hypothesis of heterogametism, already set forth in 
several papers (3, 6). Each species of Oenothera is supposed to produce 
two types of gametes called a and /3 gametes. The a gametes are generally 
female and the ^ gametes generally male, although other conditions occur, 
as will be shown later in the discussion of the phenomenon of metacliny. 
If the immune strains carry a factor I for immunity (i will then represent 
the absence of the factor for immunity, or presence of a factor for sus- 
ceptibility) in only one type of gamete, and if only combinations are 
viable, then it can readily be seen that such a strain will breed true for 
immunity, but will give a susceptible hybrid, one way or the other, when 
reciprocally crossed with a susceptible strain. If I were a dominant factor, 
all the breeding behavior would be clear, providing it were possible for I to 
be an attribute of the a gamete in some strains, and of the gamete in 
others. This hypothesis has been borne out by the results obtained, and 
