448 A CENTURY OF PROGRESS IN THE NATURAL SCIENCES 



Lotsy, Campbell, and Lam, while a possibly multiple origin has been advocated 

 by Suesseng'uth, Engier, Ponzo, Calestani, and Gaussen. The seeming necessity 

 for more than a single origin is owing, as in dicotyledons, principally to re- 

 luctance to accept derivation of apetalous and petalous, anemophilous and ento- 

 mophilous, woody and herbaceous plants from each other. 



The selection of an ancestral group largely determines, of course, which rep- 

 resentatives of living monocotyledons are to be regarded as retaining the maxi- 

 mum number of primitive characters. These authors seeking a single source 

 for the monocots have turned overwhelmingly to Ranales — living or extinct. 

 Such features as plasticity and indefiniteness in number of floral parts, the regu- 

 lar or occasional occurrence of acyclic or hemicyclic arrangement, trimery, and 

 a differentiated perianth, numerous stamens, successive pollen division and other 

 embryological similarities, an apocarpous gynoecium, parietal placentation, the 

 frequent lack of vascular cambium, and the preference for an hygrophytic way 

 of life have all been given importance. Thus the supposedly closest ties have 

 usually been found between Alismatales (Heliobieae, Butomales) and either Ra- 

 nunculaceae (Salisbury, Wettstein, Wodehouse; Werth, 1941; Metcalfe) or Nym- 

 phacaceae and Ceratophyllaceae (Nitzsche, Worseck, Parkin, Ankermann, Troll, 

 Eber, Andreanszky, Puri). Turning the tables, Schellenberg thought dicots 

 might have been derived from monocots via Alismatales-Ranales. Worsdell sug- 

 gested that both Ranunculaceae and Nymphaeaceae might have descended from 

 monocotyledons; Earle (1938) found reasons for considering Nymphaeaceae, 

 and both he and Mez and Ziegenspeck, Ceratophyllaceae as much monocots as 

 dicots. Several authors have, however, disagreed with the postulation of affinity 

 between Ranunculaceae and Alismataceae, contending that they are not so close 

 as has been generally assumed (Hallier; Meyer, 1932; Troll, 1932; von Tuzson) 

 and that neither family is by any means primitive (Corner). Cuenod (1932) 

 believed monocotyledons to be derived monophyletically from dicotyledons, but 

 was undecided whether the junction should be with Ranales or Caryoph^dlales. 

 A possible connection between Ranales, in the vicinity of Berberidaceae, and Lili- 

 ales (Liliiflorae) was postulated by Hallier. Sargant (1905), also, thought Lili- 

 ales basic in the monocots, and Markgraf and Granick proposed Alismatales as 

 a connecting link. Nicotra (1909-1910) regarded Alismatales as related with 

 Nymphaeaceae but at the same time deemed Cyclanthaceae as being nearer the 

 base of the monocotyledons. 



Reluctant to derive woody plants from herbaceous forebears, Lindinger 

 saw Dracaena as the JJrtypus of monocots, a view apparently shared by Boureau. 

 A similarity between "pachycaul" monocots and cycadophytes was mentioned by 

 Corner, and a possible line of ascent from bisexual palms to Liliales was sug- 

 gested on the basis of pollen by Wodehouse and on evidence from floral nectaries 

 by Brown. Advocates of diphyletic, triphyletic, or polyphyletic derivation of 

 the monocotyledons have tended to reconcile some of these differences, as well 

 as to submit additional possibilities. Hill saw a derivation of aroids (Spadici- 

 florae) from Piperaceae, and of Liliaceae from Ranunculaceae, and Lotsy main- 

 tained a similar diphylesis. Descent of palms from Bennettitalean ancestors, 

 and of Alismatales from Nymphaeceae, was offered by Calestani. Schaffner 

 thought that Yucca and Dracaena are primitive in habit, Alismataceae and 

 Palmaceae in floral organization. That the monocots had three independent 



