Harris — Polygamy and Certain Floral Abnormalities. 189 



As to the relative number of perfect and sterile flowers 

 occurring at one time, the following list shows the condition 

 of flowers selected at random on the grounds of the Missouri 

 Botanical Garden and on vacant lots within a radius of two 

 or three miles of the Garden : — 



July 20-August5, 1902, 863 185 



This gives 82.2 per cent, of perfect flowers and 17.6 per 

 cent, of those with undeveloped pistils. At Lawrence, 

 Kansas, in July, up to the 7th, of 205 flowers examined, 166 

 were perfect and 39 sterile. In 182 flowers examined at 

 Thayer, Kansas, August 5, 177 were perfect and 5 sterile. 

 While few tabulated countings were made, practically the 

 same conditions were found in material examined during: the 

 summer of 1903. 



The lower flowers of a cluster are almost always perfect, 

 while those near the end are much more likely to be simply 

 staminate. The reduction in the pistil may occur in any or 

 aU of the flowers of the inflorescence. To gain some idea of 

 the relative position of the sterile flowers, 100 clusters were 

 selected as nearly at random as possible, the only limitation, 

 and that a rather large one, being the necessity of selecting 

 such material as had lost none of its individual flowers and 

 was yet of such maturity that the sexual condition of the 

 terminal buds might be determined with certainty. The 

 flowers were numbered from the base to the tip of the main 

 axis, which produced, in the material examined, from two to 

 fifteen flowers, with an average of eight. The inflorescence 



