Tricotyls as Half Races and Intermediate Races. 353 



rangement of the leaves in the exhibition of trimerous 

 whorls on the stems in the later life of the plant. I paid 

 great attention to this point at the beginning of my ex- 

 periments when I visited the great nurseries at Erfurt. 

 Here the ternary individuals of Antirrhinum ma jus in 

 the fields impressed me greatly. They were not con- 

 sidered by the gardeners as worth any attention, but they 

 formed the foundation for my first tricotylous inter- 

 mediate race. 



The difference between tricotylous half and inter- 

 mediate races lies in their percentage composition and 

 not in the visible characters of the individuals. Neither 

 the number nor the cleavage of the cotyledons on a single 

 individual is decisive. As a rule tricotylous specimens 

 of both races tend to produce a richer harvest of the 

 tricotyls than the atavists of the same race; but expe- 

 rience shows that the difference is only a small one ; and, 

 further, that tricotyls, even those of a high productive ca- 

 pacity, are often surpassed in this respect by some of their 

 atavistic brethren. The chief point is, however, that both 

 the half race and the intermediate race are composed of 

 both types of individuals ; in the former the tricotyls are 

 rare, whereas in the latter, under normal circumstances, 

 both forms appear in about equal numbers. Moreover, 

 both races contain all the stages of hemi-tricotyls, and, 

 although these are rarer, of hemi-tetracotyls also. 



It is not possible, therefore, to tell from a single plant 

 to which race it belongs. Only its ancestry can deter- 

 mine this; and if this is unknown, we have to reach the 

 decision by means of subsequent breeding. It is an ex- 

 treme case of the transgressive variability which was 

 discussed in the first volume.^ The forms composing a 



'Vol. I, Part II, §25, pp. 430 ff. 



