VI] SPHENOPSIDA 77 



type of leaf which we associate more particularly with Spheno- 

 phyllum is a much later develojjment, unknown before the Upper 

 Carboniferous. It appears to have arisen by a broadening of the 

 narrow segments of the primitive type on exactly the same lines 

 as we shall see were so frequently followed among Pteropsida. 



We know very little of the fructifications of the earlier members 

 of this race, but those oi Pseudobornia (Fig. 26, 2, p. 54) appear to 

 have possessed sporophylls essentially of a divided leafy type. 

 At any rate in younger types, such as the Lower Carboniferous 

 Cheirostrobus and the Upper Carboniferous Sphenophyllum 

 Daivsoni, the sporophylls are divided, some lobes being fertile, 

 others sterile. In the typical Sphenophyllaceous cone, one 

 segment is sterile and protective and two segments (the sporangio- 

 phores) are fertile. The sporangia, as in all those groups, are 

 simply metamorphoses of parts of a fertile leaf-branch, just as 

 they are in living ferns of to-day. Sometimes the whole segment 

 is thus metamorphosed (when the sporangia are sessile), some- 

 times onty a part, while one portion remains as a sporangiophore. 



As regards the stem, neither in the Sphenophyllales nor the 

 Equisetales is there any trace of a pericaulome origin. The 

 leaves being typically small in both these phyla, the stem appears 

 to have acquired sufficient inherent mechanical strength without 

 any such adaptation. 



In the Equisetales, a group, which on the present evidence 

 appears to be a little later in time than the Sphenophyllales, and 

 unknown before the Lower Carboniferous, the leaves were 

 primitively thalloid and forked. Those of Archaeocalamites come 

 ■\'ery near those of Sphenophyllum in this respect. The prevailing 

 tendency in this phylum since then has been the reduction of 

 such compound structures to a single segment (i.e. the leaves of 

 Calamites), and at later periods to the almost complete union of 

 these reduced leaf bases (e.g. Equisetites and Equisetum). 



As regards the cones the tendency has been to sterilize certain 

 leaf branches (the so-called bracts) alternating with fertile leaf 

 branches. In Archaeocalamites these structures are absent, but 

 in later cones, such as Calamites, they are always present and 

 remain simple, undivided structures with the solitary exception 

 of Cingularia. 



