390 BOTANICAL GAZETTE [December 



from taxonomy most of it has centered about the latex and the 

 poUinia. Chauveaud is the only recent writer who has examined 

 the embryo sac. The latex will not be considered in this paper, 

 except incidentally ; and pertinent literature will be noted under 

 the different topics. Schacht (lo) states that in A, CormitiWv^ 

 primordia of each set of floral members arise simultaneously and 

 separately, while the sets arise in centripetal succession. Chau- 

 veaud (3) reports the centripetal order in CynMichiirn L. {Vin- 

 cetoxictim Moench). Payer ^ is referred to in this connection, 



but his book was not accessible. 



From the upper surface of the somewhat swollen mass of 

 meristematic tissue destined to bear the flowers of the umbel 

 many hemispherical projections arise, each of which gives rise 

 to a single flower. The outer ones are the first formed. This 

 agrees with the centripetal order of opening of the buds in the 

 umbel. Each flower grows from the axis of a bract which 

 slightly precedes it in origin. One might expect that a careful 

 investigation of an hypogynous, apocarpous, actinomorphic, 

 almost polypetalous flower would reveal its acyclic ancestry, but 

 in no case could any precedence be noted among members of 

 the same set, nor any confluence of primordia. The umbels in 

 A. tuberosa are terminal, as shown by longitudinal sections 

 tlirough the stem tip {jig. 7), and the same is probably true of 

 the other species. 



About the time the carpels appear, the calyx begins a growth 

 of epidermal hairs {Jig. 2), mostly on the dorsal surface. Later, 

 when the sepals are reflexed, hairs are also formed on their ven- 

 tral surfaces, which are then exposed. The sepals are reflexed 

 with the petals at the time of the opening of the bud, but after 

 the latter drop off, the former fold tightly up against the carpels. 

 In this second closing their function is again probably protec- 

 tive. The reflexion of the sepals and petals seems to be due to 

 turgescence of cells on their adaxial sides near the base. It is 

 suggested that perhaps the exposure of the adaxial surface of 

 the basal part of the sepals, due to the dropping off of the 

 corolla, causes loss of turgescence and consequent infolding. 



'Traitd d'organog^nie comparee de la fleur. Paris. 1857. 



