466 MORPHOLOGY OF THE ANGIOSPEKMS 



important differences also in wood structure. The wood rays of the 

 Bennettitales are simple, consisting of one kind of cell only, and lack- 

 ing marginal cells; the wood rays of the angiosperms are made up of 

 two major types of cells, and marginal cells are prominent. The 

 bennettitalean cone and the angiosperm flower seem to represent paral- 

 lel development in unrelated lines. 



The Caytonialian Theory. The Caytoniales, a small group of Mesozoic 

 fossil gymnosperms, known chiefly by their fruiting structures, have 

 been considered possible ancestors of the angiosperms, because their 

 seeds are enclosed in a carpellike sheath. They are described and their 

 morphology is discussed in Chap. 6. Possessing closed carpels, they can 

 hardly be considered ancestral to the angiosperms, in which the carpels 

 are still incompletely closed in many genera. 



The Pteridosperms. The seed ferns are frequently suggested as possible 

 ancestors to the angiosperms. But, in addition to ovule differences be- 

 tween the two groups discussed in Chap. 7, wood structure does not 

 uphold this seed-fern origin of the angiosperms. Scalariform-pitted 

 tracheids in secondary xylem, prominent in the wood of primitive 

 angiosperms, are absent in the pteridosperms. 



The Prephanerogamae and Chlamydospermae. The terms Prephan- 

 erogamae and Chlamydospermae have both been applied to seed plants 

 that are intermediate in some characters between gymnosperms and 

 angiosperms. Authors differ in the use of the two terms and of their 

 coverage — the Gnetales alone; the Gnetales and Ginkgo; and Casuarina 

 and Sarcopus, sometimes called the "Protangiosperms." The various 

 bases for classification are so confused and artificial that the terms 

 should be discarded. 



Angiosperms, Polyj^hyletic or Monophyletic. Regardless of the ances- 

 tral stock from which the angiosperms arose, the question is frequently 

 raised — are they monophyletic or polyphyletic? In the primitive orders 

 of dicotyledons and monocotyledons, there are groups of families or 

 single families that seem to have no close relationship to other families. 

 In the Ranales, there are about ten of these lines (Fig. 147); in the 

 monocotyledons, there seem to be three or four. This great diversity in 

 the primitive families suggests a polyphyletic origin, a long specializa- 

 tion of a primitive stock, or origin en masse from a diverse ancestral 

 stock. But absence of close relationship among these primitive families 

 need not indicate independence of origin; differences may be the result 

 of modification over long periods. 



Support can perhaps be found for the hypothesis of a polyphyletic or 

 en masse origin of the angiosperms ( 1 ) in the variety of perianth in 

 the more primitive families — a perianth of one, two, or several bud 

 scales, with or without petaloid organs; (2) in the evidence that the 



