1897.] NATURAL SCIENCES OF PHILADELPHIA. 189" 



average number of flowers on a well-developed cyme may be one 

 hundred and fifty, it is rare to find a dozen fruits mature. Often 

 there are none. I have not been able to satisfy myself that the 

 anthers discharge their pollen on the stigma before the expansion of 

 the corolla, thus insuring self-fertilization beyond all chances, as I 

 have shown to be the case in many other instances. I have not 

 found pollen in an unexpanded flower, nor any anther that was not 

 covered with pollen in an expanded flower. The discharge of pollen 

 and expansion of the petals is probably simultaneous. The stamens 

 are longer than the style, and one may say, almost with certainty, 

 that the flowers receive only their own pollen. Facts might be 

 adduced in support of the proposition that it was an instance of 

 sterility from the lack of pollen from other flowers, but the weight of 

 evidence will, I think, favor the conclusion that its failure is from 

 abortion. In other words, the sudden arrestation of the growth 

 force disperses the energy into other channels, with sterility as the 

 result. 



The winter color of the past season's branches has attracted atten- 

 tion. The rich reddish-brown has given it the popular name of 

 " Red-twigged Dogwood." The manner in which this red tinge is 

 produced is plainly discernible. I have shown in other papers 

 that color in vegetation is mainly an incident in the struggle for 

 life of the various parts of a plant. If we cut off a branch from 

 a tree, it dies without a struggle. If a frost come early in the fall, 

 the leaves blacken and take on no bright tinge. But if we only 

 partly detach a branch from the tree, or if, in the autumn, there is a 

 struggle with physical forces before death finally assumes control, 

 the leaves color. All this I have fully elaborated elsewhere. The 

 same course is evident here. On the young bark, a few white dots 

 appear. These are incipient cork cells. For a while they are in- 

 active. As soon as they begin to develop, the epidermis turns white, 

 and at once a light pink ring encircles the vesicle. When rupture 

 takes place, a pink line extends upward and downward, and from 

 this pink line the reddish tinge becomes gradually diffused. In a 

 short time the whole of the epidermis, both of the young branch and 

 of the petioles, has a reddish-purple tinge. No one who will make 

 a few observations during midsummer will fail to be convinced that 

 the formation of cork cells is a destructive agency, and yet an 

 agency of comparatively slow action ; that color is the result of this 

 protracted struggle for life ; and that the peculiar action of the cork 



