124 BOTANICAL GAZETTE [august 



from sterility of the type classed as impotent (Stout 25), which 

 results very frequently from hybridization. In sterility of hybrids 

 there is poor development of both sets of sex organs; stamens 

 and pistils are both affected very uniformly, and the tendency 

 is to give complete sterility. In intersexuality loss of sex develop- 

 ment for one sex is not necessarily associated with similar loss in 

 the expression of the other sex. In fact, the opposite condition is 

 the normal one for such cases. 



In P. lanceolata fhe so-called ''first form" is very high in its 

 grade of maleness, and it is in these plants apparently that seed 

 production is noticeably low. As already stated, such plants may 

 fail to set any seed. They have maleness well developed, but func- 

 tional femaleness may be lost, although pistils are present. Like- 

 wise in the most marked cases of loss of maleness the degree of 

 femaleness may be high, as is seen in the plants classed as females. 

 Darwin (7) reports that females in certain gynodioecious species 

 {Thymus serpyllum, T. vulgaris, and Satureia hortensis) are much 

 more productive of seed than the hermaphrodites, and that thus 

 the species produces more seed than if all were hermaphrodites, 

 a condition to which he attaches evolutionary significance in the 

 formation and separation of the two sex forms. Correns (5), 

 however, reports that the hermaphrodites of S. hortensis are more 

 productive of fruit and seeds than the females. 



If it is found that in P. lanceolata femaleness also varies in 

 the degree of its expression, it is quite probable that increased 

 maleness is correlated in the individual with decreased female- 

 ness. Still it is also possible that the variations are such that 

 both decreased maleness and femaleness may be present in the 

 same individuals, that individuals may be intermediate for both, 

 and that both maleness and femaleness may be well developed, 

 giving full hermaphrodites. All these conditions, it appears, are 

 represented in the groups of intersexes studied by Goldschmidt, 

 Banta, and by Davey and Gibson. Such facts go far toward 

 establishing the fundamental similarity between sex characters and 

 every other class of structures as functional hereditary characters. 



It is the tendency to a differential loss of one sex that distin- 

 guishes intersexuality from sterility (impotence) resulting from 



