512 ON THE DISTRIBUTION OF MUTANT CHARACTERS AMONG 



described originally under this name (de Vries 1918, p. 13) and Oe. 

 distorts and Oe. diluta have been derived from it (p. 485). 



Full descriptions of the stature and characters of the new types 

 have been given in "Flora" and the "Zeitschrift fur Botanik". 

 {Opera VII, p. 292 and 451). 



From this review we conclude that there are three large groups 

 of mutants and probably four small ones. Among the first, one has 

 only mutants with the normal number of chromosomes, and their 

 characters are mostly of a recessive nature. It is probable that this 

 group includes also the factors which determine the difference 

 between the laeta and velutina gametes. We will call it the "central" 

 group. This gives for the other divisions the name of "lateral" ones. 

 Two of them are large, consisting of several externally similar or 

 genetically related races. The four remaining groups are very small, 

 each including as yet only one or two types. 



In his classical researches on Drosophila Morgan has compared 

 the groups of the mutants of this fly with the relative size of the 

 chromosomes and concludes that the larger rods carry the factors 

 of the large associations, whereas the smaller groups are located in 

 the smaller chromosomes. Now if we extend this principle to our 

 case, we are led to expect three large chromosomes and four small 

 ones in the haploid nuclei, or three large and four small pairs after 

 copulation. 



Unfortunately, the size of the chromosomes in Oe. Lamarckiana 

 and its derivatives is so small, and the irregularities in their reduction 

 phenomena are so prominent, that until now no direct answer can 

 be given to our question. 



But one of the Californian species, which are most intimately 

 connected with Oe. Lamarckiana, gives better results. Ralph E. 

 Cleland (1922) has described the reduction divisions in the pollen 

 mother cells of Oe. franciscana Bartlett. In mid-diakinesis the chro- 

 mosomes are arranged in rings, each ring consisting of one or more 

 pairs, as shown on Cleland's plate XXV 1 1,, figures 25—29. Differences 

 in size are easily seen and most clearly in figure 28. Three pairs are 

 large, but four are evidently smaller. The three first ones are on 

 one side and the small ones on the other, constituting two sharp 

 groups. In clearness and distinctness these preparations far surpass 

 those obtained by previous authors from other species and from 

 mutant races. 



Therefore, it seems permissible to apply the results of Cleland 

 provisionally to our question. If we do so, our three large groups 



