260 MUTATIONS OF OENOTHERA SUAVEOLENS DESF. 



described above, and the typical specimens were hardly to be dis- 

 tinguished from the laeta of the reciprocal cross. I self-fertilized 

 two specimens of the main type and one of the other. 



From the seeds of the typical ones I had 64 and 66 offspring, 

 almost all of which flowered in August 1916. They were uniform 

 and repeated the characters of their parents. There was, however, 

 a large degree of fluctuating variability in the foliage and the petals 

 also showed differences in size, ranging from 2—5 cm. in length. 

 Besides these I found, in the beginning of July, 5 and 3 weak speci- 

 mens of lutescens, giving a percentage of 6 percent, which is much 

 smaller than in the previous generation. 



The seeds of the lutescens of 1915 yielded a uniform progeny of 

 70 specimens, which were intermediate between typical Oe. sua- 

 veolens mut. lutescens and Oe. Lamarckiana, having somewhat nar- 

 rower leaves than the former but large flower buds of a yellowish 

 green. There was much fluctuating variability among them or 

 perhaps a splitting of minor marks, which I have not studied. 



Setting these aside, the main result of our experiment is that 

 the hybrids showed only two types: (1) a typical intermediate one, 

 and (2) that of one of the mutants of Oe. suaveolens. In Oe. Lamar- 

 ckiana x biennis no such mutants appear, and the progeny of the 

 cross is uniform and constant in the next generation. 



Oe. suaveolens x Lamarckiana. — The cross of 1914 yielded 61 

 offspring, which clearly constituted thiee types. Two of them 

 resembled those of the recipiocal cross, and the third was evidently 

 a velutina, having the typical marks of Oe. (biennis x Lamarckiana) 

 velutina combined with those of Oe. suaveolens. There were 58 percent 

 of typical specimens or laeta, 21 percent lutescens and 21 percent 

 velutina, pointing to a proportion of 2 : 1 : 1. I repeated the cross 

 in 1915 and cultivated from it 66 hybrids in 1916. In August I 

 found among them 81 percent laeta, 9 percent lutescens and 10 per- 

 cent velutina. 



In 1915 I self-fertilized three specimens of laeta, one of lutescens 

 and one of velutina, in order to test their constancy. I had 70 plants 

 of each of the four former lots and 32 of the velutina. These cultures 

 proved to be uniform and like their parents in so far as no trace 

 of a repetition of the splitting of the first generation was seen among 

 them. On the other hand the laeta are far from being a wholly uniform 

 group, showing many secondary types, in the characters of size 

 and form of the leaves, hairiness, density of spike, color of the 

 stems, size of the flowers and so on. But these I have not subjected 



