452 PART IV. CLASSIFICATION*. 



acese ; the gyna?ceum, in some Alismacese and Butomacese. In 

 some cases, one series becomes polycyclic at the expense of 

 another : thus in the acyclic flowers of Clematis, Anemone, and 

 Caltha, the petals are replaced by stamens so that the number of 

 turns of the spiral ^ = whorls) in the androecium is increased whilst 

 the corolla disappears. The " doubling" of flowers is commonly 

 due to the polycyclic development of the corolla, the additional 

 whorls being either new formations, or the result of the more or 

 less complete replacement of the sporophylls by petals. . 



An important case is that to be found in several Dicotyledonous 

 orders ^Ericaceae, Crassulaceae, Saxifragaceae, some Caryophyllaceae, 

 Onagraceae Fig. 270, Geraniacese/ Oxalidaceae, Rutaceae Fig. 

 268) where the flower is diplostemonous, and the androecium is 

 apparently dicyclic : but the flower is not simply diplostemonous 

 v as in the Monocotyledons^, because the whorls do not alternate 

 regularly ; the stamens of the apparently 

 outer whorl are directly antipetalous, conse- 



quently the stamens of the inner whorl are 

 antisepalous, and the carpels (in eucyclic 

 flowers^ are antipetalous. Such flowers are 

 said to be ot>diplostemonous. 



Oligotaxy, or a decrease in the typical 

 number of whorls in a flower, is frequently 

 FIG. 268. Diagram of due to suppression. For instance, owing to 

 the suppression of one whorl of stamens in 

 some Monocotyledons, either the outer (some 

 Hfemadoraceae, also Cypripedium), or the inner (Iridaceae, most 

 Orchidacea?\ the androecium is monocyclic. In some cases a whole 

 series is suppressed : for instance the corolla may be absent 

 (e.g. Glaux, among the Primulacese ; Alchemilla, Sanguisorba, 

 among the Rosaceae : some Caryophyllaceae, such as Sagina 

 apetala, Scleranthus, etc.) : or the androscium or gyna?ceum 

 (diclinous or dioecious flowers, such as those of Sedum Rhodiola, 

 Rhamnus cathartica, Hydrocharidaceae, ray-florets of Compositse, 

 etc.) : or the whole perianth (Frajcinus excelsior}. 



Although it is true that both oligotaxy and oligomery are 

 frequently due to sujjpression, in the one case of one or more 

 whorls, in the other of one or more members of a whorl, it must 

 not be assumed that this is the only possible explanation. On 

 the contrary, it is very probable that the simple structure of 

 the flower in some plants ( c.g. Urticales and Amentales among 



