270 BOTANICAL GAZETTE [October 



hybrids with other species; but the characters are not always as 

 sharp as in the instances described, or the production of seeds is 

 too insufficient for further cultures. Only one case may still be 

 mentioned here. It was a mutant from 0. lata, discovered in 1914, 

 the self-fertihzed seeds of which gave a dimorphic second gen- 

 eration, consisting of 19 plants of the parental type, 47 of the 

 Lamarckiana type, besides 2 mut. oblonga and 2 mut. lata. Almost 

 all of these flowered in 1915. Those of the parental type were 

 strikingly like one another, constituting a wholly new form, with 

 very long, narrow, dark green leaves, the stems low and scarcely 

 branching, the spikes rich with bright flowers like those of Lamarck- 

 iana, and with a good supply of pollen. The fruits, however, were 

 cylindrical and very thin, containing only a few good seeds. The 

 plants excelled in beauty the species and most of its other mutants, 

 but on account of its slight fertility I do not propose to continue 

 the culture. It may be called 0. superflua. 



0. biennis Chicago mut. saligna. — In the second generation of 

 my race of 0. biennis Chicago^" I found in 1913, among 870 normal 

 individuals, two specimens of a weaker, narrow-leaved tj^e, which 

 differed sufficiently from the former mutants of this species, 

 namely, from 0. biennis Chicago mut. salicifolia and mut. sali- 

 castrum," to be considered a new form. One of these new mutants 

 died before flowering, the other }delded, after self-fertilization, 

 a small but sufficient harvest of seeds. One-half of these seeds 

 were sown, but only 17 specimens germinated and grew up into 

 flowering plants. Of these 9 repeated the type of the parent, but 

 8 returned to the size, vigor, and characters of 0. biennis Chicago, 

 the grandparent. Although the numbers are very small, they 

 point to a splitting into equal parts, as in the splitting mutants 

 of 0. Lamarckiana just described. 



The difference was already evident in March, when the seed- 

 lings were only two months old. In June the rosettes were large, 

 but smaller than those of the species, the leaves smooth and narrow. 

 The stems grew up to about one-half the height of their atavistic 

 sisters, and began to flower in September, having a length of 60-120 



'° Gruppenweise Artbildung, pp. 34, 52, etc. 1913. 

 " Gruppenweise Artbildung, p. 304. figs, no, in. 



