446 SKETCH OF PALEOBOTANY. 



The LycopodinecB, now represented by the natural order Lycopodiacew, 

 and constituting little more than one-third of 1 per cent of the living 

 vegetation of the globe, embraced in the Carboniferous epoch the lepi- 

 dodendroid group. About four hundred species of these plants have 

 been described from the Subcarboniferous to the Permian, and during 

 their reign they formed nearly one-fourth of the vegetation of the globe. 

 They were the largest forest trees of their time, and sometimes attained 

 a great size, though, of course, nothing approaching the giants of our 

 present forests. This ancient, or archaic, type disappears entirely with 

 the Permian, and never reappears. Its degenerate descendants con- 

 tinue down to the present, chiefly in the form of club mosses, of which 

 considerable variety exists. 



The two remaining groups of cryptogamic plants, the RhizocarpecB 

 and the LigulatcB, possess little paleontological importance, although the 

 number of species, including spore-cases, that have been referred to the 

 former of these orders has now reached seventeen, four of which are 

 Paleozoic (Devonian and Subcarboniferous) and four Mesozoic. These, 

 as well as most of the Miocene species, belong to the genus Salvinia or 

 one nearly allied to it (Protosalvinia Dawson), although one Pilularia 

 lias been found at (Eningen, and a true Marsilia occurs in an undescribed 

 collection now in my hands, made by Captain Bendire in the Miocene 

 of the John Day Eiver region, Oregon, and which I propose to call Mar- 

 silia Bendirei, should there prove to be no inaccuracy in this determi- 

 nation. 



As regards the IdgulatcB, they are still less frequent in the fossil state, 

 and are thus far represented only by the two very dissimilar genera, 

 Selaginella and Isoetes. Unless, as has been affirmed, the former of 

 these genera has its representatives in the Carboniferous, the group is 

 not found lower than the Ceuomanian of Atane, Greenland, where Heer 

 has detected his Selaginella arctica. Mr. Lesquereux has described 

 three species of this genus in the Laramie group, and the same author 

 has found a true Isoetes in our Green Eiver Eocene, at Florissant, Colo- 

 rado. Two more species of Isoetes from the Miocene of Europe exhaust 

 the enumeration, making in all only seven species of Ligulatse. 



We have thus rapidly glanced at the relative development of each of 

 the cryptogamous types of vegetation, and will next consider that of the 

 phanerogamous types. As already shown, the Gymnosperms stand 

 lowest, and have probably, in some still undiscovered way, descended 

 from the Cryptogams. Of these we place the Cycadacese lowest on ac 

 count of their endogenous growth, circinate estivation, and other char- 

 acteristics which seem to ally them to the ferns. Still, as the lines are 

 now drawn by the best authorities, the Cycadacese cannot be traced be- 

 low the Carboniferous, while the archaic progenitors of the Coniferae 

 extend far down into the Silurian. If we refer the Medullosm to the 

 ferns, as Eenault and Grand' Bury would have us do, only three cyca- 

 daceous plants occur in the Carboniferous ; but one of these is a true 



