400 BULLETIN No. 127. [August, 



should be greater in the potato than in parts of perennial plants 

 from a single rootstock, from the latter class of which De Vries 

 obtained a great deal of his data, on account of greater diversity 

 of environmental forces. The variation here ought to be analogous 

 to the individual fluctuations of the fungi, or asexual animals which 

 have the power of obtaining food in different places and of being 

 surrounded by diverse conditions. 



A study of the actual amount of difference of fluctuating vari- 

 ability in asexual and sexual reproduction has been made by Pear- 

 son (80), who also makes use of Warren's (102) work on parthen- 

 ogenetic reproduction in Daphnia. Their work makes use of data 

 from both animals and plants which might be criticized as not be- 

 ing strictly comparable, although most great biological laws have 

 thus far seemed to apply to both animals and plants. 



His first proposition is that selecting one parent reduces the 

 variability of the race by only about 5 percent while selecting both 

 parents reduces it about 10 percent, and this is almost the limit of 

 reduction even if the whole back ancestry be selected. The varia- 

 tion then taking place is, of course, from the new type and not from 

 the unselected type. 



This proposition if true for such reproduction* as there is in 

 potatoes, would show the probable amount of reduction of varia- 

 bility which there is in the established variety after it has been 

 selected to type for several years and then placed on the market, 

 leaving out of consideration the lessened variation due to greater 

 age in the variety. 



The next point is that the individualf variability in a fluctuating 

 character after a bi-sexual union is not greatly less than the varia- 

 bility of the race. As an example, is taken the number of stigmatic 

 bands on the capsules of Shirley poppies. The racial variability is 

 1.885 bands, the individual variability based on 300 plants is 

 .8518x1.885, or a reduction of 15 percent. Again, the racial varia- 

 bility of the number of leaflets on the compound leaf of the ash was 

 found by examining two hundred trees to be 1.976; the partial 

 variability is .9181x1.976 or a reduction of only 8 percent.* * 



The last point is made on the variability of mothers and daugh- 

 ters in the purely asexual reproduction of Daphnia. The variability 

 of the mothers for a certain character was 2.221, for their daugh- 



*This fact should be true at least for selectioti from crosses. 



tPearson does not distinguish here as does De Vries between indidividual and partial fluct- 

 uations. 



**The writer does not subscribe to all of Pearson's conclusions on homotyphosis. There 

 are, however, certain fluctuating characters where the individual variation is probably but 

 little less than that of the race. 



