PHYLOGENY OF THE ANGIOSPERMS 465 



ments as the most primitive angiosperms. The inflorescences of the 

 Amentiferae, with their simple, naked flowers and wind pollination, 

 suggested the cones and pollination of the conifers. The angiosperm 

 stamen and its two pollen sacs were considered the equivalent of the 

 bisporangiate microsporophyll of many conifers. But the flowers and 

 inflorescences of the Amentiferae have been shown to be specialized, 

 rather than primitive. The simplicity of the flowers of the Salicaceae, 

 Betulaceae, Myricaceae, Casuarinaceae, and other families is the sim- 

 plicity of reduction, rather than of primitiveness. Anatomical structure, 

 especially that of the wood, strongly supports the advanced position of 

 these families. 



Gymnosperm-Gnetales-Angiosperm Theory. The Gnetales (sensu lato), 

 commonly called the highest gymnosperms, have been interpreted as 

 transitional from gymnosperms to angiosperms. They combine naked 

 seeds with vessel-bearing wood. Gnetmn has an angiospermlike leaf 

 and a female gametophyte that, in some characters, suggests the 

 gametophyte of angiosperms. The vessels of the Gnetales are, his- 

 tologically, unlike those of angiosperms; their perforations have been 

 derived from round-pitted tracheids, rather than scalariform-pitted 

 tracheids, as have the vessels of the angiosperms. The ovules are un- 

 like in general structure, and the female gametophyte of Gnetum is 

 unique. 



The Anthostrobilus (Bennettitalean) Theory. The anthostrobilus 

 theory is founded on resemblances in the angiosperm flower (magnolian 

 type) to the cone of the bennettitalean cycads ( cycadeoides ) , a Meso- 

 zoic taxon. The anthostrobihis is a hypothetical, conelike reproductive 

 structure that differs from the strobili of lower vascular taxa in its 

 amphisporangiate structure, with megasporophylls above the micro- 

 sporophylls, and in the presence of a distinct perianth. 



The prominence of the Bennettitales in the Cretaceous, at the time 

 when angiosperm fossils are first abundant, seemed to support the 

 theory of a phyletic relationship between the two taxa. In gross struc- 

 ture, the flower and the cycadeoid cone are indeed alike, but sporophyll 

 types are wholly unlike. The microsporophyll of the Bennettitales is a 

 large, pinnately compound organ; that of the angiosperms is basically 

 a simple, laminar organ. The megasporophyll of the Bennettitales is 

 stalklike, with a terminal ovule; diat of tlie angiosperms is laminar, with 

 numerous ovules on the adaxial surface. Separating the megasporophylls 

 of the cycadeoides are many scales, appendages unknown in angio- 

 sperms. The bennettitalean stem has a large pith, thin vascular cylinder, 

 and thick cortex, with strongly girdling leaf ti'aces; the angiosperm 

 stem has a small pith, thick vascular cylinder, and thin cortex, with 

 leaf traces not girdling or lateral ones only slightly girdling. There are 



