664 PROFESSOR F. O. BOWER ON 



pseudo-leaves, each with several veins, which however do not appear to have them- 

 selves branched. This is seen in Schizoneura* For our comparisons the most 

 important fact is the existence of equal dichotomy of the linear leaves within the 

 Equisetales (Asterocalamites) , though in most of them the leaf is unbranched. 



In this connection must be mentioned two organisms of great antiquity but of 

 rather uncertain affinity. Pseudobornia, which was described by Nathorst from 

 the Upper Devonian of Bear Tsland, had jointed stems and whorled leaves, according 

 in this with the Equisetales and Sphenophyllales. But the leaves are palmately 

 branched with repeated dichotomies, the segments having serrate edges and the 

 whole presenting a fern-like habit, with a more complex leaf-structure than is 

 found in the Equisetales or Sphenophyllales. t Another early type with elaborate 

 leaves borne in whorls on a jointed axis is Cheirostrobus, of which only the fertile 

 cone is known. Each sporophyll is composed of three equal sterile lobes and three 

 sporangiophores. In so compact a branching as it shows it is difficult to be sure 

 of the method of branching. It suggests trichotomy. But there is no doubt that 

 the distal end of each of the sterile lobes shows a dichotomous branching. It thus 

 appears that, though in both cases the knowledge is limited and development 

 unknown, still there is clear evidence of dichotomy in the construction of their 

 foliar organs.J 



Sphenophyllales and PsiLOTACSyE. 



But dichotomy is much more prevalent in the leaves of the Sphenophyllales and 

 their living correlatives the Psilotacese. The latter bear in the vegetative region 

 simple leaves with an unbranched vascular strand ; but their sporophylls show 

 normally a single dichotomy associated with the production of a sporangiophore. 

 In abnormal cases the dichotomy may be repeated. A possible interpretation of 

 these facts would be that the vegetative leaves are simple by reduction, while the 

 forked sporophylls represent the more primitive state. This would bring them 

 into line with the fossil Sphenophylls, in which branching exists both in the 

 vegetative leaves and in the sporophylls, and is carried further than in the 

 Psilotaceae. This is particularly well seen in Sphe?iophyUum majus, where the 

 sporophyll may show two successive, equal dichotomies ; but these are repeated 

 still further in S. cuneifolium, and especially in the relatively large-leaved 

 Trizygia. 



There is a great variety in the state of " webbing " in this family. All grades 

 of it can be illustrated, from narrow linear dichotomising segments with a single 

 vein, as in S. tenerrimum, to webbed segments with a marginal tooth at the distal 

 end of each vein of the dichotomy, as seen in S. cuneifolium. The most advanced 

 state of webbing is seen in those with dichotomous venation, but entire margin, as 



* See Feistmantel's drawing reproduced in Engler and Prantl, i, 4, fig. 345. 

 t Nathorst, K. Svensk Vetmskaps-Akad. Hand., xxxvi, No. 3. 

 I Scott, Studies in Fossil Botany, i, pp. 114-123. 



