MUTATIONS OF OENOTHERA SUAVEOLENS DESF. 



247 



own characters. The leaves were narrow, e.g., 1.5 cm. broad by a 

 length of 8 cm.; the fruits were smaller than in the species, reaching 

 only half their length (20 cm. against 40, by a width of 4 mm.). 

 Foliage of a dark green. Stature low, the same as in Oe. mut. jacu- 

 latrix. 



(2) Oe. suaveolens mut. lata 



Two mutants lata (fig. 1) have arisen in 1915 in the third gem- 

 ration of my pure race, from the seeds of two of the four self-fer- 

 tilized specimens of the previous year. The cultures embraced 150 

 seedlings each, giving a percentage of almost 1 percent lata (0.7 

 percent). The two mutants were exactly the same in outward ap- 

 pearance and in hereditary behavior. One of them was weak and 

 did not yield any fertile pollen; the other was very vigorous and richly 

 branched, without pollen during the first two weeks of its flowering 

 but producing afterwards a sufficient supply for artificial self- 

 fertilization. The female flowers were fecundated with the pollen 

 of one of the typical suaveolens individuals of the race and thus I 

 got three sets of seed. From these I had the second generation 

 in 1916; it consisted partly of lata, repeating the characters of the 

 parent, partly of pure suaveolens and partly of mutants. Among 

 these, the lutescens were frequent, the jaculatrix rare and no others 

 were observed. I counted the cultures in the beginning of May 

 when the differences were sharp. They amounted to 54 individuals 

 from the self-fertilized seed and 69 and 70 from the two crossed 

 lots. The percentage composition of the three sets was as follows: 



All the aberrant individuals and some of the type of the species 

 were planted out and controlled until the period of flowering was 

 almost over. 



It is interesting to compare these figures with those given in my 

 book on the Mutation Theory for the analogous splitting of Oe. 

 Lamarckiana mut. lata. Here the mean percentage of lata is 22 



