PROGRESS IN PLANT MORPHOLOGY 613 



tion must have preceded this world-wide Cretaceous distribution of basic 

 families. 



Recently I have been assessing morphological variation broadly through- 

 out angiosperms in respect to as many characters as possible, and the variety, 

 extent, and degree of specialization have compelled me to consider great age 

 as the only answer; I see no evidence of "explosive" evolution. 



In the light of such great diversity of form among the more primitive 

 families (compare the woody Ranales, the Ranunculaceae, the Helobiales, 

 the Liliales), must we consider a possible polyphyletic origin for the angio- 

 sperms? We hear this suggestion more and more frequently. I have lived 

 through the period when any suggestion of polyphyletic origin for the angio- 

 sperms was met by the statement that it is inconceivable that so remarkable 

 a structure as the 8-nucleate gametophyte could have arisen more than once. 

 Yet, today, we accept not only embryo sacs of greatly different structure, 

 but those of such extreme morphological origin as derivation from one, two, 

 or four spores. 



Fifty years ago, we could hardly have accepted the breakup of the vascular 

 cryptogams into unrelated taxa, that is, polyphyletic origin. Even recently, 

 we at first considered the breakup of the Ranales a doubtful step, but now 

 we see even the woody Ranales as an unnatural group. We have insufficient 

 evidence yet to come to any well-substantiated decision about a multiple 

 origin for the angiosperms, but there seem to be at least three apparently 

 unrelated primitive lines in the angiosperms as we know them today: the 

 woody Ranales, the primitive Liliales, and the Helobiales. These may, of 

 course, represent extreme diversification in a very old monophyletic line. 



What can we say today of the origin of the angiosperms? Darwin's 

 "abominable mystery" is still unsolved, but we are making progress by the 

 process of elimination. Discarding both the coniferophyte and the cycadophyte 

 lines, there remain among known seed-plant lines only the ancient pteri- 

 dosperms — a great and diverse group almost unknown fifty years ago. This 

 we are rapidly learning to know. The angiosperms resemble the pteridosperms 

 in many fern-like characters of both body type and anatomy. We know little 

 yet about the Permian and early Mesozoic members of this great taxon. 

 Perhaps these forms will shed more light on the origin of the angiosperms. 



Walking with Professor A. C. Seward in his garden during the International 

 Congress at Cambridge in 1930, 1 asked him what he thought of the possibility 

 of finding fossil evidence of the origin of the angiosperms. He replied, "How 

 shall we recognize them when we find them? Perhaps we already know them." 

 This seems prophetic today when we realize that few, if any, of the characters 

 long used to distinguish the angiosperms can be looked for in their ancestors. 

 We have progressed a long way in our study of the morphology of the angio- 

 sperms when we acknowledge this — a long way from the narrow view of 

 single-character classification, of simplicity as always indicative of primitive- 



