10 



PEOOEEDINQS OP THE ACADEMY OF 



[1886. 



as the speaker knew, no botanist regards the stamen as an axial 

 development, 3'^et there are occasional phenomena that seem to be 

 inexplicable on any other hypothesis. We have to admit that a 

 flower is not merely composed of modified leaves, but is a modified 

 branch ; the branch, arrested in its development, produces sepals 

 and petals in the order and in the place where leaves might have 

 been. Occasionally, however, the usual order of phyllotaxis seems 

 broken. Stamens will spring from the base of petals, and oppo- 

 site, where we looked for them to alternate ; and then for the 

 sake of consistency with the phyllal hypothesis, we have to 

 assume that one theoretic whorl has proved abortive, or that there 

 has been a multiplication of whorls, the superimposed one being 

 the extra. This introduction of an extra series, immediately over 

 the lower, not provided for in the original phyllotaxy, has, I 

 think, never been seen in the normal condition of the branch, 

 and it is difficult to conceive how this could occur under the 

 arrestation of axial growth that transforms the branch into a 

 flower. On the other hand, if we take the petal to be the analogue 

 of the leaf on the elongated branch, there seems no reason why 

 there should not be, theoretically, an axial bud to the petal ; and, 

 should this bud develop, it would be the superimposed stamen. 

 Branching and articulated stamens are frequent in those families 

 that have these organs spring, as it were, from an axial bud at 

 the base of the petal, as in a diminutive or suppressed secondary 

 branch we might expect them to do. 



The flowers of Mahernia verticillata Cav., a well-known Byttneri- 

 aceous plant from the Cape of Good Hope, common in cultivation, 

 which he exhibited this evening, seem to indicate that its super- 

 imposed stamens are really arrested branches. The genus is 

 separated from Hermannia chiefly by a cup-shaped gland at the 

 middle of the stamen (see Fig. 2). A comparison with the 



axial development of the inflores- 

 cence shows the stamen to be formed 

 on precisely the same plan as the bi- 

 flowered peduncle (Fig. 1). Really 

 the flowers are axillary a single 

 flower being produced from the axil 

 of each leaf. What appears to be 

 the " two-flowered jjeduncle " of 

 authors is simply a diminutive 

 branchlet. After forming one node 

 the longitudinal development be- 

 comes nearly arrested, and we have 

 only a shortly-pedicillate and slowly 

 developing flower, representing the 

 shorter of the two in the cut. The 

 bud in the axil of the bracteolate 

 leaflet, however, makes another and stronger attempt at growth, 

 and pushes up and over the one which terminates the first growth. 



2. 



Mahernia verticillata Cavanilles. 

 1. Two-flowered branchlet. 

 2. Stamen, magnified. 



