Lutz, Triploid Mutants in Oenothera. 395 



deVries' 1895 mutant. Is it not therefore possible that the years 

 of inbreeding which have followed have reduced the stature of 

 this form and have altered other characters to some extent as well? 



Further careful study, however, demonstated conclusively that 

 f>f)89 was not 0. gigas. The nodes were more widely separated, 

 and while the stoutness of the plant was much increased over that 

 of 0. Lamarckiana, it was still less pronounced than that of 0. 

 gigas, and represented an approach towards, rather than a duplica- 

 tion of the gigas condition. 



A third interesting mutant, 4453 (Fig. 5), appeared in a culture 

 grown in 1909, derived from an 0. lata mutant (known to have 

 15 chromosomes), self-pollinated. 



The rosette leaves of this plant combined the lata crinkledness 

 and general contour of leaf with the thickness of 0. (/if/as . They 

 were, however, less finely crinkled than is characteristic of 0. lata. 

 The adult foliage was distinguished in the same manner. It 

 produced 10 basal branches, although all can not be counted in the 

 photograph (Fig. 5), and completed flowering at a height of slightly 

 less than 7 dm. The increased stoutness of the stem, branches 

 and all parts of the plant was very noticeable here as for the 

 21 -chromosome mutant, and mutant 5589. The shape of the bud, 

 however, differed markedly from that of the two other forms. While 

 the buds of 4453 were much larger than those of 0. lata, the 

 first had the irregular shape so characteristic of the first buds of 

 0. lata, though later ones of both forms became more regular and 

 tapering. Although the early buds had practically no red on the 

 sepals, little fleckings appeared on the later ones, and by the close 

 of the season those of the branches which had been flowering 

 longest were covered with a fairly even thin red. This, however, 

 was much less intense than the coloration of the buds of the 21- 

 chromosome mutant. It will be recalled that those of 5589 remained 

 yellow until the close of the season. The absence of red on the 

 sepals is also characteristic of 0. lata. 



In 0. lata the length of the calyx of the first buds is insufficient 

 to accomodate the full growth of the petals and style without much 

 compression. It is this condition that causes the irregular shape 

 of the early buds in this form. When the flower expands, the 

 petals are unfolded in a crumpled condition, and the style is 

 frequently found to be bent into a single or double elbow. The 

 crumpling of the petals in 0. lata is one of the causes of the earlier 

 fading, for, with the rising of the sun, the flowers of this plant are 

 among the first in the garden to droop. 



A condition similar to that just described for 0. lata was found 

 in the flowers of 4453 (which were much larger than those of 

 0. Lamarckiana or 0. lata), though somewhat less pronounced. 



