248 MUTATIONS OF OENOTHERA SUAVEOLENS DESF. 



percent, deduced from above 50 cultures. Thus we see that the 

 proportion is as narrowly the same in both types of lata as could 

 be expected, and this fact constitutes a strong argument for their 

 exact parallelism, and for the assumption that they are due to the 

 same internal change of the hereditary material. Moreover, we may 

 deduce from our table that the pollen of Oe. suaveolens mut. lata, 

 if present, has in this respect the same qualities as that of Oe. suaveo- 

 lens and that the mutation changed only the maternal side. 



There is still another analogy. Oe. Lamarckiana mut. lata is more 

 mutable than Oe. Lamarckiana itself, producing especially a larger 

 number of the pale and weak form Oe. albida (2 percent, Gruppen- 

 weise Artbildung 1913, p. 314). In the same way Oe. suaveolens 

 mut. lata produces far more mut. lutescens than Oe. suaveolens itself 

 usually does (about 1—2 percent). But here the difference is much 

 larger. Moreover the chromosomes have been counted by my assis- 

 tant, Mr. C. van Overeem, in the roots of the young plants of the 

 second generation. The number was fifteen, exactly as in Oe. La- 

 marckiana mut. lata. 



Already in May, about ten weeks after the sowing, when the 

 seedlings are still small but ready for planting out, the distinguishing 

 marks are clear. The leaves of lata are smaller and broader and 

 have a rounded tip and smaller petioles. In June the plants make 

 their stems, which are much lower than those of suaveolens of the 

 same age. The foliage becomes less broad, but the tips of the leaves 

 are always more or less rounded. The margin is usually turned 

 upward. All of my plants flowered as annuals, but this is also the 

 case with the parent species. 



In August and September the lata reached only a height of 60-80 

 cm., whereas the suaveolens reached 1.5 meter and more. They 

 were densely branched and of a dark green. The flowers differed 

 in size on different individuals, the petals measuring ordinarily 

 2.0—2.5 cm. in both directions. The lobes of the stigma were thick and 

 sometimes partly concrescent with the style. In 1916 the largest 

 number of the flowers had some good pollen, but far less than Oe. 

 suaveolens itself. The fruits were small, less than half the size of 

 the species, and hairy, with the same form as those of Oe. Lamarcki- 

 ana mut. lata. 



(3) Oe. suaveolens mut. sulfurea 



The flowers are of a pale yellow, exactly as in Oe. biennis sul- 

 furea. Otherwise all the characters are the same as in the species. 



