58 Wilson. — Observations on Epigcea repens, L. 



ence in the lengths of the styles, and the differences in the 

 lengths of the stamens are looking toward dimorphism. 



It is not known whether the flowers with small stigmas are 

 ever fertile or not. 



In 1891 Halsted examined sixty flowers and came to the 

 following conclusions : ' 



That there is only a tendency towards dimorphism : 



There is no difference between the pollen grains taken 

 from the anthers of long or of short stamens. 



He finds the strong tendency to become unisexual in large 

 part sufficient to account for the differences in length of 

 styles and stigmas. 



During the present year the writer has examined about 

 1,000 plants of Epigcea repens at the time of flowering. Most 

 of them were from western North Carolina. A few were 

 from localities in Canada, Massachusetts, New Jersey and 

 Pennsylvania. 



These flowers readily fall into two groups : 



(1) Those having perfect pistils, with or without rudiment- 

 ary stamens. 



(2) Those having pollen-bearing stamens, with rudimentary 

 pistils. 



Occasionally flowers are found in the first group without a 

 trace of a stamen. Often there are bare vestiges of filaments 

 at the base of the corolla. In rare cases the anthers may be 

 present, but without pollen. 



In the second group the pistil is rudimentary through an 

 undeveloped stigma only. The stigma in the first mentioned 

 form is a five-lobed, star-shaped, terminal body, opening its 

 lobes out nearly at right angles to the style. In the second 

 form the stigma is five-lobed, the lobes being closely appressed 

 into a terminal, functionless, oval enlargement. 



The first form is without perfect pollen-bearing stamens, 

 and is, therefore, pistillate. The second form, having a 

 functionless pistil through an imperfect stigma, but bearing 

 perfect stamens with pollen, is therefore, staminate. The 



1 Bulletin of the Torrey Botanical Club, Vol. xviii., p. 249, 1S91. 



