312 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1918. 



branches perform most of the photosynthesis. They are all sporan- 

 giophoric and strobiloid, although there is a considerable range of 

 variation in morphological and histological details. Some are homos- 

 porous and others heterosporous. 



The phylum is unique in that it attained its maximum development 

 in the Paleozoic and has been practically unrepresented since Triassic 

 times except by the single genus Equisetum, which survives to the 

 present with less than two score ubiquitous and rather uniform 

 species. The stock appears to have been of Pteridophytic origin and 

 to have been primitively megaphyllous. Taxonomically it corre- 

 sponds rather closely with the Articulatae of Lignier and the Sphenop- 

 sida of Scott. As known at the present time, it consists of two classes — 

 the Sphenophyllae and the Calamariae, and the latter includes three 

 rather well-defined groups or orders, namely, the Pseudoborniales, the 

 Equisetales, and the Calamariales, the last including two families, 

 the Protocalamariaceae and the Calamariaceae. 



The most primitive of these subordinate groups constitutes the 

 single order Sphenophyllales of the class Sphenophyllae. This was 

 a synthetic alliance of mostly small forms that combined certain fern 

 characters with those of the Calamariae on the one hand, and the 

 Lepidophyta on the other. They are regarded as representing the 

 specialized descendents of a pro-Sphenophylhun stock, which was 

 more truly intermediate between the Arthrophyta and Lepidophyta, 

 and which is supposed to have flourished in pre-Devonian times. The 

 Sphenophyllums ranged from the Devonian to the Permian and were 

 practically cosmopolitan except for their partial extinction in Gond- 

 wana land during and immediately subsequent to the Lower Permian 

 glaciation. They comprise a considerable number of rather uniform 

 species, based for the most part upon the impressions of the slender 

 jointed ribbed stems with nodal whorls of cuneate leaves (hence the 

 generic name), and long familiar to paleobotanists. 



Their habit was much like that of a modern Galium, although 

 the genus Cheirostrobus, based upon structural cone material, sug- 

 gests that they were not invariably small, weak-stemmed, clamber- 

 ing forms. The ribs did not alternate at the nodes, and conse- 

 quently the leaves of successive whorls were superposed and not 

 alternating. The leaves, normally six to a whorl and cuneate, with 

 entire or toothed apical margins, were frequently dichotomously 

 laciniate, and in some forms with numerous narrow leaves in each 

 whorl, these are legitimately regarded as corresponding to the 

 laciniate segments of the digitately leaved species. The stems 

 branched frequently at the' nodes. The fructifications were fairly 

 large cones, superficially resembling those of calamites. 



The stem anatomy, the elucidation of which we owe in the first 

 instance to Renault, was characteristic and sufficiently unique. In 



