1887.] NATURAL SCIENCES OF PHILADELPHIA. 325 



tioiis at the time of my observations. The bruised petals of Am- 

 phicarpcea, showed that, at some time during the day or night, in- 

 sects were at work ou the flowers. 



The large vigorous blossoms on these plants were of a brilliant 

 purj^le. The plants which I have seen with whitish-yellow flowers, 

 in other localities, grow weaker than these. They also produce 

 legumes from apetalous flowers on the climbing stems, but the local- 

 ity is not near enough for me to watch, to see whether the petali- 

 ferous flowers bear freely or not. 



Fruiting as these flowers do, on the climbing branches, at least, 

 there are two-thirds fall without seeding. This cannot be from im- 

 potency of the pollen, or none would seed. Few leguminose plants, 

 even when exposed to cross-fertilization, perfect more than a small por- 

 tion of their flowers. Failure is from too great a draft on the nutritive 

 ' powers of the plant. So many flowers cannot be properly fed, though 

 properly fertilized by pollen. The remarks of Mr. Darwin on Voande- 

 sia, are no doubt just, changing the word cultivation so as to read "we 

 should remember how often fertility is aftected by circumstances." 



I have already placed on record that the petaliferous flowers of 

 Viola cucullata rarely produce seed under ordinary circumstances, 

 but freely do so when the plant is growing on a dry rockery. 



The points I regard as rendered certain by these observations are : 



1. That the climbing stems, as well as the trailing ones, of Amphi- 

 carpcea bear apetalous flowers freely. 



2. That these flowers produce a third form of legunie. 



3. That the petaliferous flowers, under circumstances favorable to 

 nutrition, bear legumes as freely as leguminous plants generally. 



4. That the petaliferous flowers are adapted solely to self-fertiliza- 

 tion, and I think the probability is developed : 



That the apetalous flowers are often fertilized by pollen from the 

 petaliferous ones ; and, so far the plant is arranged for as much 

 cross-fertilization as other monoecious plants receive. 



A contribution to the life history of Cephalanthus occidentalis. 

 The pistils of Cejohalanthus occidentalis are exserted more than an 

 inch beyond the flowers, and as there are usually from one hundred 

 to three hundred flowers in the button-like head, the mass has a well 

 known trichodic appearance. I had never been able to see these 

 extending themselves. The flowers were always either closed or fully 

 expanded with the pistils to their full length. Satisfied that the 

 opening took place during the night, I cut a few on the evening of 



