6lO EAMES 



has been followed by that of Hydrastis and Glaucidium as other genera out 

 of place. Even twenty-five years ago, a breakup of the Ranunculaceae would 

 have been no more acceptable than that of the seed plants and vascular crypto- 

 gams, but that we accept the change now with little objection is a sign of 

 progress to broader concepts. The Ranales provide another example of the 

 use of characters from several fields ; even the woody families have been shown 

 to be a heterogeneous assemblage of small families which seem to represent 

 at least three independent lines: the Winteraceae, the Magnoliaceae, and four 

 probably related families — the Degeneriaceae, Annonaceae, Eupomatiaceae, 

 and Himantandraceae. 



Fifty years ago, the angiosperms were considered too large and diverse a 

 taxon for sound interpretation. Today, we still have too little information 

 about them; but we recognize even more clearly the vast numbers, the great 

 diversity of structure, and the possible importance of families and genera yet 

 little known, or still unknown. (Note the major changes made in our recog- 

 nition of the primitive members of the angiosperms by the recent critical 

 studies of the woody Ranales.) 



Fifty years ago, the evolutionary line of vascular plants led from the lower 

 gymnosperms through the Gnetales to the angiosperms. Today, we have tested 

 and discarded three possible ancestral lines. The cycadophyte line was long 

 and strongly supported — and still is by some morphologists — but it must be 

 set aside. This line cannot be ancestral to the angiosperms because of major 

 differences not only in body structure — huge pith, thin vascular cyclinder, 

 thick cortex, and even number of encircling leaf traces — but also in structure 

 of xylem, in sporophyll type, and in microsporangium position and structure. 

 The coniferophyte line — never seriously considered in this respect — ^has re- 

 cently been urged as showing connection with the Amentiferae. This con- 

 nection has been considered to be from Juniperus to the Amentiferae because 

 of similarity in cone and flower structure. Here, also, general body type and 

 detailed vascular structure show the impossibility of any such close relation. 

 The similarities are those of simplicity in form of reproductive structure — 

 similarities that in both taxa are the result of reduction. The Gnetales have, 

 until recently, been seen as the obvious step from the simpler gymnosperms to 

 the flowering plants, largely because of the presence of vessels. We now know 

 a great deal more about the vessel and recognize it as a product of high 

 specialization in xylem — a structure developed independently several times 

 even within the angiosperms. The presence of vessels is clearly without value 

 as support for the Gnetalean origin of the angiosperms. 



A new ancestral group for the angiosperms has recently been proposed — 

 the Prephanerogamae. In a contribution from the so-called "New Morphol- 

 ogy," Casuarina is removed from the angiosperms and added to the Gnetales 

 to constitute this new taxon. Superficial resemblance to Ephedra is part of 

 the evidence. But Casuarina is an angiosperm in every detail of structure — 



