SFERMATOPHYTA 411 



the Cordaitales^ — is the older. Except for the presence of seeds, however, 

 the Cycadofilicales are so fern-like as to leave little doubt that they have 

 been derived from some ancient fern stock. The Cordaitales, less fern- 

 like, may have branched off from the Cycadofilicales early in the Paleo- 

 zoic, although it seems more likely that both groups have had a common 

 origin. The Cj^cadofilicales very probably gave rise to two divergent 

 lines of descent, one represented by the Bennettitales of the Mesozoic, a 

 specialized offshoot that became extinct, the other leading to the Cyca- 

 dales, a group that still survives. 



There is strong evidence that the Cordaitales were ancestral to both the 

 Ginkgoales and the Coniferales, orders that reached a climax in the Meso- 

 zoic. Except for Ginkgo hiloha, the Ginkgoales are extinct, while the 

 Coniferales are still so abundant as to constitute the largest order of living 

 gymnosperms. The Gnetales are a group of obscure origin. They may 

 represent a specialized offshoot from the Coniferales, but any relationship 

 to the angiosperms is very doubtful. 



The oldest undoubted angiosperms appear in the deposits of the Lower 

 Cretaceous. Their characters are so distinct that they must have orig- 

 inated at a much earlier time, but it is not known when or from what 

 older group they arose. There is a possibility that the angiosperms orig- 

 inated independently from the pteridophytes, but this is remote in view 

 of the many common features existing between gymnosperms and angio- 

 sperms. There is no convincing evidence, however, indicating from what 

 group of gymnosperms the angiosperms may have sprung. 



One theory holds that the angiosperms have been derived from the 

 Gnetales. Their compound strobili resemble the inflorescences of certain 

 angiosperms with monosporangiate (imperfect) flowers and a simple 

 undifferentiated perianth. The presence of vessels in the secondary wood 

 is a character shared by both groups. If the resemblances between the 

 Gnetales and angiosperms are a result of parallel evolution, there can be 

 no direct relationship between them, although they may have come from 

 a common ancestry. 



Another theory claims the derivation of the angiosperms from the 

 Bennettitales. Their strobilus, which is bisporangiate, somewhat resem- 

 bles the flower of a magnolia, but the stamens, and particularly the car- 

 pels, are very different in the two groups. On the whole, the cycadeoid 

 strobilus is so specialized that it is very unlikely that an angiosperm 

 flower could have evolved from it. It seems more reasonable to suppose 

 that the resemblance between them is a result of parallel development and 

 does not denote any direct relationship. The fossil record has not pro- 

 duced transitional forms connecting the gymnosperms and angiosperms. 

 In their absence, speculation concerning the ancestry of the angiosperms 

 seems futile. 



