112 BOTANICAL GAZETTE [august 



Standard botanical treatments describe this species as having 

 only perfect flowers. For over 50 years, however, sex polymor- 

 phism has been recognized as present in the species. The tendency 

 has been to group the individuals in 3 classes (Ludwig 19), most 

 recently designated by Bartlett (2) as (i) first form hermaph- 

 rodite, (2) second form hermaphrodite, and (3) female. Correns 

 (4), however, groups plants of this species grown from seed col- 

 lected near Leipzig, Germany, into 5 classes, in two of which there 

 was variation in single spikes (a) from hermaphrodite flowers to 

 flowers with imperfectly developed stamens (=^^), and (h) from 

 more or less perfect flowers to flowers only female (±^ and ?). 

 In thus making these classes recognition is given by Correns 

 to variations in sex organs which include various grades of 

 gynomonoecism already observed in this species by Schulz (24). 



The difficulty of making any adequate classification, expressed 

 in some degree by Correns (4) and by Bartlett (3), has been 

 very apparent from the observations which the writer has made. 

 In 191 2 Bartlett very kindly supplied me with plants which he 

 classed under the 3 forms just noted. Seed progenies have been 

 grown and observations made of plants growing wild in the fields 

 in and about the New York Botanical Garden, where P. lanceolata 

 is exceedingly abundant. Study of this material reveals that 

 there is present a wide range of variations in the development of 

 sporophytic sex organs, which in its general aspects is quite iden- 

 tical with the phenomenon of intersexualism especially described 

 by GoLDSCHMiDT, Banta, and by Davey and Gibson. 



description of the three forms 



Flowers typical of the forms most generally recognized may 

 first be described, as illustrating the two extremes and one inter- 

 mediate. ' The flower drawn for a plant was in all cases selected 

 from the middle portion of a spike, and was typical of a large 

 number of flowers in bloom. The flower was placed on a glass 

 slide, a large cover glass was placed over it to bring the various 

 parts into somewhat the same plane, measurement was made of 

 the flower parts under very low magnification by ocular microm- 



