20 Rhodora [January 



Paxton/ the polygamy was observed by Maximowicz ; the differences 

 in this species are much more pronounced, because in the staminate 

 flower the pistil is reduced to a small subglobose body, otherwise 

 -they are of the same character, as the accompanying illustrations 

 (fig. 3 and 4) show. Also the anthers in the pistillate flowers seem 

 to remain closed as far as can be judged from herbarium specimens. 

 In the literature of systematic botany I failed to find any reference 

 to the polygamy of Chionanthiis 7'irginica \ in the generic descriptions 

 which include C. retiisa the genus is characterized as having perfect 

 or polygamous flowers, while C. virginica is always described with 

 perfect flowers. The polygamy, however, has been observed before 

 and the first notice of it I found in the Horticulturist of 1857 

 ( 13 : 266), where Th. Meehan in an article on trees and shrubs with 

 ornamental fruits makes the following remarks about the Fringe-tree : 

 ■" Many trees do not bear and others imperfectly .... for though it is 

 classed .... with the perfect flowering plants, it is in reality polyga- 

 mous, as much so as the Ash." A similar statement in an unsigned 

 note, probably also by Th. Meehan, appeared in the Gardeners' 

 Monthly of 1885, (27 : 228) . Two years later Meehan '^ gives a short 

 account of his -observations on the polygamy in Chionanthus 7>if- 

 gin/ca, accompanied by two figures showing the different styles, and 

 he remarks that Gray notes in "the later edition of his Manual" 

 that Chionanthus is occasionally polygamous. I could, however, find 

 no allusion to it in Gray's Manual nor in any of the more recent 

 American floras, and it seemed to me therefore not useless to draw 

 again attention to the fact that the flowers in Chionanthus virginica 

 are not monoclinous, but are. what probably would be the best term 

 for it, andro-dioecious, though they could be called perhaps as well 

 imperfectly dioecious. These terms will apply to the whole genus, 

 for there is no real difference between the two species in this respect, 

 only the Asiatic species represents a more advanced state of dioecism. 

 Arnold Arboretum. 



'Brit. Fl. Gard,3:85, f. 273(1^53). 



C. chiiifiisis, Maximowicz, Bull. Ac. Sci. St. Petersb. 20: 430 ; Mel. Biol. 3 : 393 



(1874). 



== Proceed. Acad. Nat. Sci. Philad. 1887 : 280 (1888). 



