CONSTANCE: SYSTEMATICS OF THE ANGIOSPERMS 445 



6. LEiTNERiACEAE. Evidence that this monotypic family is a reduced member of either 

 Resales or Geraniales is found by Abbe and Earle (1940) in characters of the inflorescence 

 and flower. 



7. GARRYACEAE. Evidence that Garryaceae are not primitive l)ut highly specialized 

 and reduced has been adduced by Hallock on data from the infloi-escence and flowers. 

 She suggests that these plants are the "highest of the Umbelliflorae" (1930, p. 810), an 

 affinity accepted also by Hjelmqvist and Sporne. 



The lack of any fundamental unity among the groups which have been re- 

 ferred to "the Amentiferae," the occurrence of such presumably advanced char- 

 acters as syncarpy and epigyny, the ample evidence of extensive reduction in 

 inflorescence and flowers, and the absence of derived herbaceous forms, appear 

 to me to destroy completely the notion that this is a primitive group marking the 

 transition between gymnosperms and angiosperms. Allen (1940), Lewis (1942), 

 Whitehouse (1950), and Darlington (1939, 1952) present genetic evidence to 

 show that the derivation of unisexual flowers from bisexual ones is highly prob- 

 able, the converse essentially impossible, an argument which gives added im- 

 portance to the widespread occurrence of abortive male or female structures in 

 the unisexual flowers of ''the Amentiferae." The parallel between the inflores- 

 cence and flowers of the group and the reductions apparent in those of such 

 other angiosperms as Acer (Hall, 1951), Platanus (Boothroyd, 1930), Planta- 

 ginaceae, Ambrosieae, and Cyperaceae (Blaser), provides a strong intimation 

 of the factors promoting structural degeneration. Similarities with living gym- 

 nosperms, including embryological ones (Mahcshwari, Battaglia), are appar- 

 ently analogous rather than homologous and many of them are very superficial. 

 The vessel studies of Thompson (1918, 1923)— although attacked by Bliss (1921) 

 and MacDufKe (1921) — have apparently been confirmed, and lead to Bailey's 

 conclusion (1949, pp. 67-68) : 



Such fundamentally signiflcant anatomical differences form an insuperable barrier to a 

 derivation of the angiosperms from the Coniferales or the Gnetales. Thus, the presence 

 of vessels in both the Gnetales and the angiosperms, which has so frequently been cited 

 as evidence of relationship, actually negates such relationship. There are similarities 

 between the end products of tracheary specializations in Gnetum and certain of the 

 dicotyledons, but they have arisen by entirely different developmental changes. 



The supposed paleobotanical proof of the comparative antiquity of catkin- 

 bearing angiosperms is inconclusive. As Axelrod remarks (1952, p. 29) : 



The fossil record does not demonstrate whether the primitive flower was generally of 

 a magnolian type ... or whether the simple type of the Amentiferae (oak, willow, alder) 

 comes nearer to the proangiosperms. 



In short there is no basis for the supposition that there exists a natural 

 group, "the Amentiferae," and the systems of Engler, Wettstein, and Rendle 

 are unnatural in so far as the basic status accorded such an artificial group 

 is concerned. 



Origin and Relationships of Monocotyledons 



A great deal of phylogenetic discussion has centered about the monocotyle- 

 dons and their position in "the System." Some of the principal points at issue 

 have been the following: 



1. Is there real affinity between monocotyledons and dicotyledons, or do they 



