1897.] NATURAL SCIENCES OF PHILADELPHIA. 177 



brought into contact with the disrupted pollen sacs, and receive an 

 additional supply of fertilizing material, as if nature was taking a 

 double care in this instance that the flower should be self-fertilized. 

 The corolla at this stage seems firmly attached to the receptacle, but 

 very soon afterward it falls at the slightest touch, indicating that 

 fertilization has been perfectly accomplished. An examination of 

 the flowers at this stage will also show that the seeds are wholly 

 mature, and we have to conclude from this examination alone that 

 the fertilization was accomplished in the unopened flower. 



CI«ISTOGAMY IN UMBELLIFERJE. 



So far as I know no record has been made of cleistogamy in the 

 Urnbelliferse. In 1893 a plant of Cryptotcenia Canadensis in my 

 garden indicated cleistogamy, but as the flowering period had ad- 

 vanced considerably, further observation was left for the present 

 season. 



Cleistogamy, as usually understood, perfects seeds when the gen- 

 erative organs are enclosed in the calyx only, no attempt at form- 

 ing a corolla being made. In this strict sense the Cryptotcenia 

 would not be cleistogamous. But the plant has two classes of flow- 

 ers : one in which the stamens and corolla are highly developed, 

 with the gynoecium abortive, the other with a highly developed 

 gynoecium, stamens with comparatively short filaments, but with 

 polliniferous anthers all enveloped in a corolla extremely fugacious, 

 beneath which fertilization is accomplished before opening. The 

 fine anthers are pressed tightly against the stigma, accomplishing 

 fertilization and inducing carpellary growth before the corolla has 

 reached perfection. When this period has been reached it falls at 

 once, carrying along with it the stamens which have already per- 

 formed their functions. Though differing somewhat from the usual 

 character of cleistogamy, the action is cleistogamic surely. 



The corolla in the male flowers is more enduring than in the 

 female flowers ; indeed, as a general rule, it ia only the male flowers 

 that we notice in examining the plant. It is extremely rare to find 

 a female flower with an expanded corolla, as it falls very soon after 

 opening, and it was this seeming absence of corollas in the first in- 

 stance, yet with an abundance of fruit-bearing pedicels, that led to 

 the suspicion of cleistogamy. 



The numerous male flowers with even more perfect stamens that 

 are more abundantly pollen-bearing than those in the seed-bearing 



