168 rROCEEDIXGS OF THE ACADEMY OF [1892. 



fruit, on plants which seem to have only male flowers. Supposing 

 the English holly to liave the same characters as the American, 

 the female flower, solitary on the pedicel, will easily be distinguished 

 from the bi- or tri-florous staminate ones. 



Twenty Ilex opaca trees on my grounds were carefully examined on 

 May 30th. Eight of these are purely pistillate plants. Thegynsecium 

 is large and perfectly developed in every respect, and although there 

 were apparently four stamens, they were membranous and function- 

 less. The pistillate flowers were easily recognized by being solitary on 

 stout pedicels. There w^ere but four or five flowers on each branch, 

 and one might almost pass a tree without knowing it was in bloom, 

 unless the flowers were sought for. The male flowers on the other 

 hand were bi- or tri-florous, and often two common peduncles 

 arose from the same axis. The stamens were large and the anthers 

 abundantly polliniferous, the ground beneath the trees being 

 thickly strewn with the fallen blossoms. The gymecium remains at 

 the base of the flower in a wholly undeveloped condition. By these 

 characters one can tell at once, without any critical examination, the 

 fertile from the infertile tree. It is not improbable that there are 

 some trees that may produce male and female flowers on the same 

 tree — may be monfficious, — but these twenty trees, thoroughly 

 dioecious, would indicate this to be its prevailing characteristic. 



A slight jarring of a branch indicates that the female tree may 

 have their flowers pollinated by the agency of the wind. Honey 

 bees were, however, busily collecting nectar indifterently from the 

 flowers of both sexes, and nuiy aid in pollination. 



On the Stamens of Ranunculus abortivus. 



Of all plants we should hardly expect to find definite stamens in 

 RanuncAdu^ ; but in R. abortivns I find them uniformly in three 

 series of five each, 15 in all. The first five mature contempora- 

 neously with the opening of the flower, and the large full anthers of 

 this series set, as they should be, alternately with the petals, contrast 

 so greatly with the undeveloped ones, that our first impression might 

 be that we were examining a five-stamened flower. In Rannnculus 

 hulbosus, blooming among these plants, no such striking difference 

 could be noted. If other species have this peculiarity it might be 

 useful as a sectional character. 



My object in examining the flowers closely was to note their habit 

 in relation to pollination. As every flower, and we might say every 



