﻿EXISTING AND FOSSIL CYCADS COMPARED. 237 



de feuille a peine modifiee pour l'autre sexe, conduit M. Wieland, en tenant compte 

 des afnnites de leur appareil vegetatif, a les regarder comiue des termes symetrique 

 d'un meme groupe, et il n'hesite pas, conformement a l'opinion que j'ai soutenue 

 plusd'une fois et qui est cgalement celle de M. Scott, a rattacher les Bennettitees aux 

 Cycadinees, pour constituer parmi celles-ci une famille ou un ordre de meme valeur 

 que les Cycadees et les Zamiees de notre flore actuelle." 



Such is the conclusion reached as based on the known facts. If, however, one 

 passes on to purely hypothetical considerations, other possibilities present them- 

 selves. If after all, we are deceived as to the profundity of structural agreement 

 in the two great cycadean lines, and if some such hypothesis as the following could 

 be shown to approach the truth, it would assuredly be quite logical to erect two 

 cycadophyte classes, the Bennettitales and the Cycadales. 



Count Solms has called attention to the similarity between the synangia of 

 Cycadeoidea and Chorionopteris Corda ; and Potonie considers the latter to resemble 

 Onoclca externallv. Now, suppose Chorionopteris and the Cyeadeoidete in turn to 

 have been derived from some very ancient Onocleiform fern stock, and the Cyca- 

 dacese from some other primitive stock by way of Angiopteris-Yike tree ferns. In 

 such a case the two different ancestral pteridophyte lines leading respectively into the 

 Cycadacese and Cycadeoidese would have passed independently through essentially 

 the same evolution of heterospory by way of two different cycadofilicineau lines, each 

 nevertheless presenting a great variety of genera and presumably families. More- 

 over, the generalized forms of these two cycadofilicinean lines, as thus fundamentally 

 and primitively separated, would always present a homoplastic similarity that 

 would perforce be especially striking at the time both assumed the more consolidated 

 type of armored trunk. The paleontologic record thus far known does not, 

 however, disclose even a hint of the existence of any such complexity in the early 

 history of the present fossil and the existing cycads. 



Contrariwise, let us conceive a second hypothetical group made up of Marat- 

 tiaceous ferns exhibiting, let it be said, to avoid prolixity, double the vegetative 

 and reproductive variation exhibited by the existing Marattiacere ; and let it be 

 further conceived that most of the members of this closed group began to assume 

 more compact trunk and foliar structures, and at the same time to develop hetero- 

 spory, thus producing a great paleozoic "quasi-fern" complex. In such a case all 

 the ecological factors known, bearing in mind the generalized tropical conditions 

 of the Paleozoic, would presumably most favor the series advancing along much the 

 same lines. 



Once more, to be specific, let it be conceived that this heterosporous Maratti- 

 aceous complex eventually gave rise to a variously branched cycadofilicinean type 

 of trunk with a gradual elimination of separate prothallial stages and a reproduc- 

 tive advance from a merely heterosporous to a primitive seed and pollen bearing 

 condition ; and let it be conceived, too, that these primitive seed plants exhibited 

 a great variety of carpellary leaves and of stamiuate fronds tending to organize 

 into various bisporaugiate ovulate and staminate strobilar forms presenting various 

 phases of monoecism and dicccism ; and that eventually, from somewhere amongst 

 this complex, was developed, in the later Paleozoic or Permian, the common 



