450 A CENTURY OF PROGRESS IN THE NATURAL SCIENCES 



tubes, depending upon the inclination of end walls and the nature of distribu- 

 tion of sieve plates; and (3) that the vascular bundles can be arranged in an 

 ascending typological series depending upon the features of the xylem and 

 phloem of which they are composed. Unlike the situation in dicots, in monocots 

 vessels are believed to have originated in the late metaxylem of roots, and to 

 have spread thence to the aerial portions of the plant. In general, these indicator 

 series suggest that monocotyledonous plants which possess the largest amount of 

 secondary tissue (produced by a thickening cambial ring) generally have rela- 

 tively primitive sieve tubes and vessels in their roots. It is noted that typically 

 bulbose or cormose plants usually possess vessels in their roots only. Highly 

 specialized monocots, on the other hand, are those which have vessels with 

 porous perforations and sieve tubes with transverse end walls distributed 

 throughout the plant body. Among the groups which appear to be highly spe- 

 cialized because of the nature and distribution of their xylem and phloem ele- 

 ments are Gramineae, Cyperaceae, J uncus, Cordyline, palms, Pandanaceae, Ty- 

 phaceae, Dioscoreaceae, and most Alismatales. Bailey has gone further than 

 Cheadle in interpreting the classificatory significance of these studies (1953, p. 7) . 



We now know that there has been an independent evolution of vessels in dicotyledons 

 and monocotyledons, and if the angiosperms are monophyletic, that the two great groups 

 of plants must have diverged before the acquisition of such structures in either group. 

 . . . Obviously the structurally more primitive types of monocotyledons cannot be derived 

 from such highly modified and specialized plants as the herbaceous Ranunculaceae or 

 Piperaceae. 



Bailey's and, I judge, Cheadle's concept of a fundamentally primitive monocoty- 

 ledon would perhaps not be too strongly at variance with Lindinger's postula- 

 tion of Dracaena as an TJrtypus. 



Although warned by both Arber and Bailey (1953) that any attempt to 

 bridge the gap between monocotyledons and dicotyledons must be regarded as 

 purely speculative at this time, it may be worthwhile if only to prevent the Ra- 

 nunculaceae- Alismataceae or Nymphaeaceae-Alismataceae hypotheses from hard- 

 ening into textbook dogma. It would seem that the similarities between dicots 

 and monocots are too profound and too numerous to be dismissed as mere analo- 

 gies, parallelisms, and convergences. The Ranales are unique among dicots in 

 possessing members with the monocolpate type of pollen which characterizes 

 most monocots and gymnosperms — the true character and significance of that in 

 Alismatales appears to be debatable (Wodehouse, 1936b; Erdtman, 1953). The 

 Ranales are unique among angiosperms in possessing primitively vesselless sec- 

 ondary xylem. The perianth-bearing flowers of some Ranales, with leaflike 

 sporophylls of both sexes, may be regarded as "standard" also for most mono- 

 cots — even Cyperaceae (Blaser, 1941), Gramineae; Aponogetonaceae, Potamo- 

 getonaceae, Najadaceae (Markgraf); Araceae (Arber), and Palmaceae (Bosch); 

 whatever the true nature of the monocot perianth may be, all the same possi- 

 bilities of derivation have been suggested also for the perianth of Ranales. Fries 

 (1911) called attention to the occurrence in Annonaceae of an arrangement of 

 prophylls which is characteristic of monocotyledons. The many structural paral- 

 lels between herbaceous Ranales and Alismatales have already been mentioned, 

 but it must be remembered that these herbaceous groups are highly specialized 

 in vegetative if not in floral characters. 



