112 CARNEGIE INSTITUTION OF WASHINGTON. 



viduals of an F3 family in which a 3 : 1 ratio occurs, only two kinds 

 of F4 families arise, namely, with triangular capsules only, or with 

 the two kinds of capsules in the ratio 3:1. 



(e) Extracted dominants in an F3 family displaying a ratio 3 : 1 

 are not the same genotypically as the original type used in the crosses, 

 though indistinguishable from it by simple inspection. Crosses be- 

 tween these extracted dominants and heegeri have yielded only F2 

 families with a ratio 3:1, instead of 15:1, as found in the original 

 F2 families. 



A new and unequivocal case of duplication of determiners is thus 

 added to the very few which have been found in other plants, as 

 the results here summarized are supported by about 140 pedigreed 

 families and can find no other logical interpretation. There is some 

 indication that duplicate determiners exist for one of the rosette 

 characters also in Bursa as well as for the distribution of sterile 

 flowers on the racemes, but these cases require further investigation. 



Hybrids in Oenothera, hy G. H. Shull. 



Dr. Shull has continued his hybridization of Oenothera lamarckiana 

 and the several biotypes of Oe. cruciata, and the progeny are truly 

 remarkable in their complexity. 



The first generation hybrids between lamarckiana and the biotypes 

 of cruciata, and between the several possible pairs of the biotypes of 

 cruciata which were reported last year as consisting of polymorphic 

 progenies and strikingly unlike reciprocals, have been repeated, 

 together with parallel series in which another strain of lamarckiana 

 and a ruhrinervis were used respectively in the place of the lamarck- 

 iana involved in the original series. The 22 types found last year 

 have all been confirmed in this year's cultures and several new types 

 have been added, the corresponding forms from the two new series 

 being indistinguishable from those originally secured. The second 

 generation has been grown extensively and more than half of the 

 possible combinations of Fi types with the parents and related 

 hybrids have been studied, and additional crosses have been made 

 to fill partially the gaps still remaining. In the crosses between the 

 biotypes of Oe. cruciata the results are relatively simple, the most 

 striking features of the situation being the facts : (a) that more than 

 100 F2 combinations have added only one new type in addition to 

 the occasional aberrant individuals which one is apt to meet in pure- 

 bred as well as hybrid stocks; (b) that in most cases the results in 

 the Fj are what would be expected if it be assumed that the hybrids 

 produce only the eggs of their mothers and only the sperms of their 

 fathers. There are several rather striking exceptions to the last 

 proposition, but the exceptions occur with a degree of consistency 

 that promises to make possible some rather simple interpretation. 



