246 HEREDITY AND EVOLUTION IN PLANTS 



the Cycadean series (Cycadales, Hemicycadales (Bennet- 

 titales) , Cycadofilicales) which have pithy steins; but some 

 of the Cordaitean forms also have pithy stems and com- 

 paratively large leaves. Here again, as so often, an 

 attempt at a formal classification necessitates drawing 

 an apparently sharp line where in fact one does not exist. 

 As Professor Jeffrey' has said, the term Archigymno- 

 spermse is one of convenience, and like most scientific 

 terms falls short of covering the situation. On the basis 

 of certain criteria {e.g., the structure of the wood), the 

 Ginkgoales appear to be intermediate between the Coni- 

 f erales and the Cordaitales. In fact, as Jeffrey^ has pointed 

 out, the "living fossil," Ginkgo, may be regarded as a 

 connecting link or transitional form between the Archi- 

 gymnospermae and the Metagymnospermae. 



The relationship of Isoetes is one of the most difi&cult 

 to determine among all the vascular cryptogams. Argu- 

 ments for and against interpreting it as derived by re- 

 duction from the Lepidodendron group are given by Lady 

 Isabel Browne.' The secondary growth in thickness of 

 its stem (in such a dwarfed form) must be regarded as a 

 character of long standing, not recently acquired; plants 

 in both groups have mucilage cavities. Isoetes resembles 

 some of the Lepidodendrales {e.g., the so-called Stigmaria*) 

 in the dichotomous branching of its roots. Other facts 

 of structure {e.g., the occurrence of the sporangia on the 

 upper side of the leaves) have also been interpreted 



'Jeffrey, E. C Science, N. S. 47: 316. 1918. 



2 Jeffrey, E. C. The anatomy of woody ptanls, p. 315. Chicago, igiy. 



'Browne, Lady Isabel. The phylogeny and inter-relationships of the 

 Pteridophyta. New Phy tola gist 7: 93, 103, 150, i8r, 230. 1908. 



* The fossil remains to which the generic name Stigtnaria was assigned 

 have long been known to be the root-system of Sigillaria. 



