198 



THE FLO WEB. 



completing the sjonmetry of the blossom and the normal alterna- 

 tion of its members. This explanation of the anteposition of a 

 single circle of stamens is the more readily received, because it 

 well accords with the idea here adopted, that the androscium 

 of a typical flower should consist of two circles of stamens. 

 (324.) The only serious objections to this explanation rise out of 

 the difficulty of applying it to analogous anteposition when both 

 circles are present. 



364. For Diplostemony, the condition of two circles of sta- 

 mens, each of the same number as the petals, is also itself very 

 commonly attended by anteposition. In normal or Direct 

 Diplostemony, that which answers to the floral type com- 

 pletely, the antisepalous stamens (324, note) are the outer 

 and the antipetalous the inner series, and the carpels when 

 isomerous alternate with the latter and oppose the sepals ; the 

 alternation of whorls is therefore complete, as in the diagram, 

 Fig. 382. Such stamens, however, may actuall}- occupy a single 



line or coalesce into a 

 tube, without derange- 

 ment of the type. But 

 it as commonly occurs 

 that the antipetalous sta- 

 mens are more or less 

 exterior in insertion, and 

 then the carpels, when 

 isomerous, are alternate 

 with the inner and anti- 

 sepalous stamens, and therefore opposite the petals, as in the 

 diagram, Fig. 383. This arrangement takes the name of 

 Obdiplostemony. In it the normal alternation of successive 

 whorls is interrupted, so as to produce anteposition, 



365. With Obdiplostemony. This condition prevails, more or 

 less evidently, in Ericaceae, Geraniaceae, Zygophyllaceae, Rutaceae, 

 Saxifragaceae, Crassulaceae, Onagraceae, &c. (but in some of these 

 with exceptions of direct diplostemony) ; also, accompanied 

 by a peculiar multiplication of members (380), in Malvaceae, 

 Sterculiaceae, and Tiliaceae. The explanation is difficult. The 

 Irypotheses may be reduced to three, neither of which is quite 

 satisfactory. There is, first, the hypothesis of St. Hilaire, ap- 

 plied to this as to the preceding case (to Rhamnus, Vitis, &c.), 

 that these exterior antipetalous stamens belong to the corolline 

 whorl ; in other words, that the petal and the stamen before it 



. FIG. 382. Diagram of pattern flower with direct diplostemony, 383. Diagram 

 of similar flower with Obdiplostemony. Both from Eichler'a Bluthendiagramme. 



