CONSTANCE: SYSTEMATICS OF THE ANGIOSPERMS 425 



tation to the demands of pollinators (Thomas, 1931; Stebbins, 1951), whether 

 insects (Robertson, 1904; Worsdell; Ames, 1937, 1946; Pennell, 1948; Grant, 

 1949, 1950a, 1950b, 1950c, 1952; Li, 1951), birds (Porsch, 1931, 1932, 1933), or 

 even bats (Porsch, 1934-1935, 1937, 1942). Conversely, reduction in all floral 

 structures may arise as a consequence of a change to anemophily (Wirth, 1923; 

 Fagerlind, 1948; Porsch, 1950; and cf. "Amentiferae," below). 



B. Androecium: The general homogeneity of staminal structure is, accord- 

 ing to Parkin, one of the principal arguments for the monophylesis of angio- 

 sperms. This uniformity does not, however, extend to the attempted explana- 

 tions of the origin and morphological homologies of stamens, as might be antici- 

 pated from the conflicting concepts of the flower mentioned above. The stamens 

 have frequently been interpreted as cauline structures, representing: (1) the 

 axes of male flowers (Neumayer, 1924); (2) the condensation products of di- 

 chotomously branched systems bearing terminal fertile "telomes" (Thomas; Wil- 

 son, 1937, 1941, 1942, 1950; Reece, Bertrand, Emberger) ; and (3) the sporo- 

 geneous emergences from a "staminal ring" (Plantefol). Ehrenberg (1945) has 

 suggested that tepals and stamens spring from tangential division of the same 

 primordia. 



The classical view that stamens are "phyllomes," homologous with leaves, has 

 been more generally accepted (Arber and Parkin, Hallier, Bessey, Eames, Troll; 

 Schaeppi, 1939; Gunderson; Baum, 1949c; Parkin, 1951; Ozenda, 1952). Of 

 special significance to this latter view is the work of Bailey and his associates 

 in calling attention to the occurrence of broad, microsporophyll-like stamens in 

 more than a dozen families of the Ranales (Bailey and Smith; Bailey and Nast, 

 1943a, 1945a; Bailey, Nast and Smith, 1943; Bailey, 1949; Bailey and Swamy, 

 1949; Canright, 1952). These stamens are characterized by a three-trace vas- 

 culation, elongate linear sporangia embedded in tissue of the sporophyll, and 

 the lack of any clear division into filament, connective, and anther. Canright 

 regards the stamen of Degeneria as the closest living epitome of the primitive 

 angiosperm stamen, and gives a synopsis of trends of specialization in staminal 

 form occurring within Magnoliaceae. This reviewer finds no difficulty in ac- 

 cepting the last author's conclusion that, 



... the preponderance of evidence seems to support the hypothesis that these broad types 

 of microsporophylls are primitive and, as such, should be considered as relatively unmodi- 

 fied phyllomes. With this concept in mind, the conventional stamen with its narrow 

 filament and protuberant terminal anther should be recognized as an extreme specializa- 

 tion of this primitive type of microsporophyll (Canright, 1952, p. 487). 



Evolutionary trends from spiral or hemicyclic to cyclic arrangement, from 

 two whorls to a single whorl, from numerous to few members per whorl, from 

 free to connate, from hypogynous or perigynous to epipetalous or epigynous, 

 from tetrasporangiate to bisporangiate, and from longitudinally to poricidally 

 dehiscent, are generally accepted. The distinction between the usual centripetal 

 development of stamens and their centrifugal maturation in orders centering 

 around the Parietales is regarded by Corner (1946) as of systematic impor- 

 tance. Effective classificatory use of characters of the androecium has been made 

 in such groups as Ericales (Matthews and Knox, 1926; Matthews and MacLach- 

 lan, 1929; Copeland; Doyel and Goss, 1941; Palser, 1951; Kavaljian, 1952), Mal- 

 vales (Edlin, 1935; C. V. Rao, 1952), and Melastomaceae (Morley). 



