PALEOBOTANY BERRY. 301 



evolutionary schemes, are almost entirely unknown as fossils and 

 such as are known are either inconclusive in character or confined to 

 relatively recent periods. These plants were never of large size, 

 nor have they ever become truly fitted for terrestrial existence. 

 Their absence in the geological record can not be attributed to 

 their perishable nature since much more delicate objects have been 

 abundantly preserved at various geological horizons. While they 

 have been recorded from the Paleozoic, especially by the early 

 students, these records usually resolve themselves into fragments 

 of Lepidophyte, Arthrophyte, or Coniferophyte foliage. None are 

 conclusively known earlier than the Mesozoic, and the liverworts 

 antedate the true mosses in the record, fruiting and therefore con- 

 clusive material of the latter group being unknown in rocks older 

 than the Tertiary. 



PHYLUM PTERIDOPHYTA. 



The term Pteridophyta as here understood is restricted to the 

 fern phylum — the Filicales of the older students — the so-called fern 

 allies (club mosses and horsetails) being grouped in the phyla Arth- 

 rophyta and Lepidophyta. The Pteridophyta are a very ancient 

 stock, always megaphyllous and phyllosiphonic. The fructifications 

 are borne upon but slightly modified foliage leaves and were never 

 strobiloid. The known forms are prevailingly homosporous, al- 

 though heterospory must have been evolved early in their history 

 since the Pteridosperms are clearly derived from the ancient, at 

 least pre-Devonian, fern stock. Among existing ferns the Hydrop- 

 terales or so-called water ferns, of somewhat questionable relation- 

 ships, are also heterosporous. 



Living ferns are usually segregated into two major groups — the 

 Eusporangiate and Leptosporangiate, according as the sporangium 

 develops from a group or from only a single cell. While this dis- 

 tinction is not without exceptions it is one not observable in fos- 

 sils, where the mature features must be relied upon. On the whole 

 the Eusporangiate sporangia are large, attached by a broad base, 

 with a wall more than one cell thick, and without a definite annulus, 

 though some of the cells may be thickened; moreover in the Marat - 

 tiales, one of the important orders from the paleobotanical view- 

 point, the sporangia are united to form synangia, which by some 

 students are thought to represent a modification of the sporangio- 

 phore so prominent in the Arthrophyta. and vestigial in the Lepi- 

 dophyta. In the present treatment the Pteridophyta are segregated 

 into the four classes, Coenopteridae, Eusporangiatae, Leptosporan- 

 giatae, and Hydropteridae, which will be considered separately. 



The Coenopteridae or Primofilices comprise a group of Paleozoic 

 ferns known almost entirely from the anatomy of stems and petioles, 



