THE SPERMATOPHYTA : GENERAL INTRODUCTION 649 



they leave the stele. The strands which supply branches do, however, leave 

 gaps in the stem stele, filled with parenchyma. In the Filicales and in all the 

 Spermatophyta the leaf traces themselves make leaf gaps in the stele. The 

 leaves in these groups are relatively large, and a physiological reason for the 

 leaf gaps may lie in the fact that through them there is continuity between 

 the parenchyma of the cortex and of the medulla, thus throwing open the 

 medulla as well as the cortex for the storage of carbohydrates elaborated in the 

 leaves. Jeffrey called the Lycopodialean type microphyllous and clado- 

 siphonic, because only the branch traces cause stelar gaps, while he called the 

 Filicalean type megaphyllous and phyllosiphonic. 



He laid great stress on the systematic importance of the distinction, and 



Fig. 655. — ^Comparison of the shoot architecture in the Lycopsida, Sphenopsida, 

 and Pteropsida. A, Lycopodium type. B, Eqidsetum tA-pe. C, Fern type. {After 

 Jeffrey.) 



he called the chief microphyllous alliance the Lycopsida and the whole 

 megaphvllous alliance the Pteropsida, which includes not only the Ferns 

 but all the Spermatophyta which share the megaphyllous character and which 

 are assumed to be more or less directly of Filicalean descent. The Fquisetales 

 and the microphvllous fossil order Sphenophyllales formed the Sphenopsida 

 (Articulatae) (Fig. 655), while the leafless Psilophytales and Psilotales are 

 combined under the name Psilopsida. Goebel has objected to this view 

 that the diff'erence is merely one of relative leaf size, and he has pointed out 

 that in some of the smallest Ferns the leaves make no stelar gaps. This is 

 not a conclusive argument, for in Azolla on the other hand, which is one of 

 the most reduced Ferns, there are leaf gaps present. It must be accepted 

 that the difference is a ver}' ancient one, that the two types have been separated 



