122 • BOTANICAL GAZETTE [august 



spikes throughout the entire period of bloom. The general ap- 

 pearance of typical spikes from such plants is shown in nos. 55 

 and 56, but the wide range of stamen forms which are present is 

 not clearly shown. The spikes shown in no. 56 also show the 

 excessive growth which stigmas frequently make. It has been 

 suggested that this occurs when pollination and fertilization have 

 not been effected, and that successful pollination inhibits such 

 growth. It is possible, however, that such growth is an indica- 

 tion of loss of femaleness. Studies are now in progress to deter- 

 mine especially the functional potentiality of pistils of plants for 

 which this phenomenon is very general. 



Summary. — The flowers described and illustrated for plants 

 36-47 show that wide variations exist in the development of 

 stamens among various flowers of a plant, or even among stamens 

 of a single flower. The range is in some cases almost identical 

 with the extremes seen for plants as wholes (nos. 1-35). This 

 statement refers to the flowers produced in the lower two-thirds 

 of the spikes. It may be noted that the range is greater for such 

 a plant as no. 46 than for one like 36 or 37. This variation is 

 not identical with the tendency for the last flowers of a spike to 

 be different from earlier flowers. Here there is a marked mixture 

 throughout the spikes. 



NATURAL DISTRIBUTION 



In the fields in and about the New York Botanical Garden 

 P. lanceolata is so abundant that it often dominates the vegeta- 

 tion over a considerable area. Here plants that approach the 

 first form are most numerous; female plants corresponding to 

 the type described as nos. 7 and 9 are abundant; and there are 

 thousands of plants which are in some degree intermediate between 

 these extremes. Many plants with mixed flowers are to be' found. 

 With respect to vegetation characters and to general size and 

 shape of spikes, extremely wide variations are -everywhere in 

 evidence. 



The variations in flower forms noted by Darwin for England, 

 by LuDWiG and Correns for Germany, and by Bartlett in the 

 vicinity of Washington, D.C., indicate that much the same range 



