84 BOTANICAL GAZETTE [august 



justification of this in the intimate and constant association of one 

 species (C. novaculatus) with a species of '' Lepidodeitdron''' with 

 small square or rhombic, closely packed bolsters, which we have 

 suspected of being a part of the same plant. At a horizon two- 

 thirds of a mile east of Westernport, Maryland (190 ft. below the 

 horizon of the Davis seam), C. novaculatus is replaced by C. rohustus, 

 a very closely related species, and it may be significant in this 

 connection that the small-bolstered species of Lepidodendron just 

 mentioned is here replaced by a very similar one. 



The only groups in the phylum Lepidophyta as it stands today 

 are the Lycopods and the Psilotales. The latter constitute a small 

 recent group of disputed position, believed by Bower (6). Thomas 

 (32), and Scott (28) to be related more closely to the Sphenophyl- 

 lales than to the Lycopods. They are characterized by lepido- 

 phytic leaf habit and furcate sporophylls supporting bilocular or 

 trilocular synangia upon the adaxial surface. Concerning the 

 former Bower has said: "There is perhaps no character which 

 marks off the plants of lycopodinous affinity from others so clearly 

 as the constancy of the solitary sporangium .... it stamps this 

 series of Pteridophytes as peculiar from all others"; to which may 

 be added a statement from Scott (27): " The- Lycopods constitute 

 a wonderfully homogeneous group so neatly rounded off as to give 

 httle hold for any hypothetical hnk with other classes of plants." 



The number and attachment of the sporangia and the position 

 and the nature of the sporangiophore, when present, are matters of 

 prime importance in the present classification of the vascular crypto- 

 gams, and at once the impracticabihty of placing Cantheliophorus 

 with its conspicuous bisporangic sporangiophore among the Lycopods 

 becomes evident. The objections to placing it among the Psilotales 

 are almost equally insuperable, and there remains no alternative 

 but to establish for its reception a new order, Cantheliophorales 

 of the phylum 'Lepidophyta. 



Phyletic relations 



The question whether this group is either more or less primitive 

 than the others of the phylum now arises for consideration. There 

 has been a growing disposition in late years to regard the simple 



