1887.] NATURAL SCIENCES OF PHILADELPHIA. 327 



After twenty four hours, a portion of the clear stigma is seen above 

 the mass of pollen, showing that the stigma expanded a little after 

 the growth of the style had been completed. 



I think it is conceded that plants have not a time for the opening 

 of flowers from which they never depart, and that exceptions have 

 been noted in even very regular habits. But it may be stated in a 

 general way that the flowers of this Cephalanthus open rapidly, soon 

 after dark, and never during the day time. 



The odor of the flowers is singularly grateful. They are visited 

 by large numbers of insects for their nectar day and night. No 

 pollen gatherers seem to work on them. There is indeed no chance 

 to collect it from the stigma, no platform affords a facility for stand- 

 ing during collection. The pollen remains on the stigma till the 

 wdiole dries away. It is one of the most complete adaptations for 

 self-fertilization known to me outside of the Cleistogene class. 



Interested in noting how this absolute in-and-in breeding affected 

 productiveness, I subsequently found the flowers remarkably fertile. 

 Numerous seeds were in every head examined. Carefully dissecting 

 one, I found it had 279 flowers, of these 225 perfected seeds, and 

 only 54 failed. 



In describing Cephalanthus occidentalis in Flora Cestrica, Dar- 

 lington describes the flowers as "five lobed ;" most other botanists 

 "foui'-parted." My flowers were mostly four, but many five-parted. 



The glands between the lobes of the corolla appear to have been 

 unnoticed. That the receptacle is "filiferous" has long been observed 

 (Loureir's Flora of Cochin China, 1788), and Gray (Synoptical 

 Flora of North America 1884) notes that these "setiform bractlets" 

 between the flowers are glandular-capitate. Donn (Dichlamydeous 

 Plants, 1834) notes "in the American species glands none in calyx 

 or corolla." They are so easily noted in these specimens before me, 

 that it is singular they should have been over-looked. There 

 are four of these "setiforme bractlets" at Ihe base of each floret, and 

 they are slightly squamiform at the base. They are nearly alternate 

 with the lobes of the calyx, starting from a little to one side of the 

 sharp angle that terminates at the base of the division of the calyx. 

 The setiform bractlet is just the length of the tube of the calyx, and 

 if adherent to the calyx, the gland would be just where it is sit- 

 uated in the lobes of the corolla. No one can fail to see the gland 

 on the corolla is the analogue of that on the setiform bractlet. 

 We have no hesitation in concluding that there is primarily another 



