596 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



VABIATION AS LIMITED BY THE ASSOCIATION OF 

 CHAEAOTEES. 



By A. WOESLEY. 



I HAVE experimented with several fertile hybrids and analysed the 

 variations which take place in the ensuing generations, and in most of 

 these experiments I found that the hybrids and their self-fertilized 

 progeny are means between their parental extremes, although there 

 are instances in which one parent influences the hybrid and its progeny 

 to an extent exceeding that of the other parent. Now if we take one 

 (suppositious) case and note eight divergent characteristics between the 

 parents {inter se), we shall probably find that the hybrid holds a strictly 

 equipoised position as regards the bulk of these characters, but may 

 incline towards the female in some and towards the male in others. 

 Tk) reach this result it is not enough to examine the first few individuals 

 of the hybrid race that come into flower, but we must wait until the 

 whole of them have flowered and until a critical analysis of them all 

 can be made. 



To be sure of our ground we must also apply this analysis to the 

 whole of those characters in which the parents differ inter se, and not 

 merely to some of them. For if we submit a hybrid to an incomplete 

 analysis we may conclude that the male is dominant, simply because 

 we have taken no count of those other characters in which the hybrid 

 inclines towards the female. 



In the succeeding generations of self- fertilized plants we shall note 

 some variation from the equipoised hybrid type. This variation may be 

 great or small, and follows no known law. For instance a true new 

 species may result from, hybridization, such being the case when the 

 subsequent variations from the hybrid type do not exceed about 3 per 

 cent., and when the preference of individuals for the pollen of other 

 individuals (in bi-sexual flowers) is sufficiently strong to*ensure the 

 breeding out of these variant characters. I have had no personal 

 experience as to whether new species can be artificially produced by 

 hybridization among those bi-sexual plants which are normally self- 

 fertilizing. 



When I have followed these hybrid progeny by critical analysis into 

 the second and subsequent generations I have not been able to satisfy 

 myself that reversion to certain specific characters follows the allega- 

 tions of the Mendelian advocates. In the first place I have never been 

 able to find in hybrids any characters that were absolutely dominant 

 or recessive, but have only discerned a certain relative or partial incHna- 

 tion towards the specific characters. Nor have I as yet found a single 

 instance of absolute reversion to either specific type ; but I have found 



