CONSTANCE: SYSTEMATICS OF THE ANGIOSPERMS 429 



lales, and to Rlioedales and perhaps Salicales; (3) cushion-nectaries, confined to 

 Mai vales; (4) toral disk nectaries, leading from perigynous Resales to the epigy- 

 nous Compositae by one sequence and to Myrtales by another; and (5) staminal 

 nectaries, in higher Ranales. Several authors (Wertli, 1941; Jaeger, 1950) have 

 emphasized the diversity of Ranalian nectaries as confirming the central position 

 of this order phylogenetically, and thus as affording a connecting link between 

 dicots and monocots (Porsch, 1913). 



It is generally assumed that, although there may be a well-defined trend from 

 foliar to disk or toral nectaries in dicotyledons, nectaries may also have been 

 contrived from any suitable material at hand — vestigial perianth (Fisher), sta- 

 minodia (Daumann, 1931a, 1931b; Dawson; Moore, 1936a; Mattfeld) or abortive 

 carpels (Woodson and Moore). The doctrine of vascular conservatism has fre- 

 quently been called into play to decipher their derivation (Kausik, Kasapligil). 

 Fahn (1952) has recently undertaken a morphological-topographical classifica- 

 tion, and recognizes the Torus, Perigonal, Stamen, Ovarial, and Stylar types. 

 Although Brown has pointed to the lack of nectaries in Magnoliales (Ranales), 

 Calestani and others have emphasized the secretion of nectar by stigma and 

 style in this group. Both Norris and Stoudt employed nectarine characters in 

 tracing affinity among the constituent families of Rhoedales. Benson (1940) has 

 stressed the different kinds of nectary scales in Raminculus, and has used them 

 taxonomically. Werth (1923) suggested a degradation series from entomophilous 

 to anemophilous fl-owers, in which nectaries shift from a foliar to an axial posi- 

 tion, followed by complete loss of nectaries with the attainment of unisexuality 

 and wind-pollination. A survey of extrafloral nectaries was undertaken by Zim- 

 mermann (1932) but he concluded that these structures were determined physio- 

 logically and ecologically, and had little or no phyletic significance. 



Embryology 



Comparative or systematic embryology, comprising particularly the features 

 of the ovule and the female and male gametophytes, has received serious atten- 

 tion only during the last half -century, according to Maheshwari (1945), greatly 

 stimulated by the work of Schnarf (1931, 1933, 1936, 1937b). Both these au- 

 thors have expressed the belief that such internal characters should be more 

 conservative than others more exposed to external influences, and hence should 

 be of significance in systematic and phylogenetic endeavors. 



A. OvuU: The nature and homologies of the angiosperm ovule have been 

 fully as vexed a question as has the morphological value of tlie carpel, and the 

 widely varying interpretations of the two structures are closely interrelated. 

 Those who have read into the angiospermous gynoecium axial, soral, cupular, 

 or telomic origins, or "involutions," usually with the construction of ingenuous 

 typological seriations, do not seem to have advanced greatly our understanding 

 of ovule structure (cf. de Haan, 1920; Neumayer, Wettstein; Appl, 1939; Un- 

 ruh; Hagerup, 1942; Chadefaud, 1946; Ozenda, 1946; Bertrand, 1947a; Martens, 

 1947, 1951; Fagerlind, 1948; Just, 1948; Emberger, 1949, 1950, 1951; Mangenot; 

 Walton, 1952). Although a direct relationship between the constitution of the 

 ovules of angiosperms and that of the ovule of any particular group of gymno- 

 sperms is not wholly clear, there seems to be rather general agreement that at 



