MEMOIRS OF THE NEW YORK BOTANICAL GARDEN. 277 



of the shoot in close organic connection with the inflorescence should 

 be subjected to the action of the rays during some portion of the 

 period. Hypopiiys appears to be indifferent to the action of light 

 and makes both flowers and seeds in darkness. Seed formation is 

 accomplished in darkness if the inflorescence is exposed to illumina- 

 tion during a portion of the earlier stages of development of the 

 flower, and also in the flowers formed earliest in " partial etiolations," 

 but not in later ones. In the last-named instance however, the 

 introduction of pollen taken from illuminated specimens seems to 

 be necessary. The pistils show least atrophy during such partial 

 etiolations, as the stamens are reduced earlier. It would be difficult 

 to maintain that the non-formation of flowers in darkness is due to 

 the lack of some specific formative substance, as suggested by Sachs. 

 The development of the flower and the perfection of the embryo-sac 

 and stamens entail important morphological differentiations, which 

 it has been shown, have failed to occur in the vegetative portion of 

 the plant. It seems much more probable that the highly specialized 

 groups of cells ordinarily differentiated in the development of the 

 pistil, stamens, and the embryo, and in the integration of the seed 

 should fail to carry out their normal procedure from the same causes 

 as those affecting stems or other members of the body. The failure 

 to effect the usual differentiations brings as an enforced consequence 

 the lack of completion of the gametophytes, and of the complicated 

 structural changes following fertilization and terminating in the 

 construction of the complex seeds and fruits. From this point of 

 view, the incomplete pollen cells found in all etiolated plants ordi- 

 narily green is not due to the direct lack of the action of light upon 

 the stamens but to an incomplete state of morphological development 

 which has not brought these bodies up to their functional maturity. 

 It seems justifiable to take the position that flower formation is 

 accomplished only when any given species is exposed to an illumina- 

 tion approximating a specific optimum, and that any deviation from 

 this is liable to inhibit the process entirely, or perhaps alter the 

 method of procedure in such manner that cleistogamic instead of 

 chasmogamic flowers may be produced. The procedure in question 

 would imply but a small departure from the customary activity. If 

 the species which produce subterranean flowers are now taken into 

 consideration it may be seen that a second adaptational type is reached 

 by which flowers are constructed without the aid of direct illumina- 



