PRB-0ARB0NIFBR0U8 PLANTS. 



48 



The flattened stems of Arthroatigma gracile often show distinct indica- 

 tions of a slender central axis, probably of scalariform vessels, though 

 the structure is obscure. 



. 



Genus Cyclostigma — Haughton. 



66. Cyclostigma densifolium, S. N.,— (PI. VIII., Figs. 92 to %.)— 

 M.D., Gasp^. 



Stems slender, covered with circular or transversely lengthened leaf-scars, spirally 

 arranged, and hearing short curved leaves with broad bases. 



The specimens on which this species is founded are fragments of stens 

 an inch or less in diameter, found in beds associated with the small bed of 

 ^coal, near Tar Point, Gasp6 Bay. At first I was in doubt whether to 

 regard these stems as belonging to Lepidodendron or Stigmaria, but the 

 form of the scars, in connection with that of the leaves, places them in the 

 genus Cyclostigma, as somewhat aberrant members tending towards 

 Lepidodendron* 



Genus Cordaites.— Unger. Pychnophyllum. — Brong. 



64. Cordaites Robbii, Dn.— (PI. XIV, Figs. 156 to 162.)— Can. Nat. 

 VI. J. G. S., XVIII, 316, PI. XIV.— M. D., St. John, New 

 Brunswick. M. D., Cazenovia, N. Y. 



" Leaves elongated, lanceolate, sometimes three inches wide and a foot in 

 length. Veins equal and parallel. Base broad, clasping the 

 stem, point acuminate." 



These large striated leaves are so abundant in some of the shales at St. 

 John, that the name " Cordaite shales" has been given to the beds by Mr. 

 Matthew. I have nothing further to add to the full discussion of the affi- 

 nities of this plant in the papers above cited. Like the Cordaites of the 

 Coal-formation it sometimes has on its surface shells of Spirorbis. f (Fig. 

 161.) 



•Since writing the above, I have seen in London, through the kindness of Messrs. 

 Btheridge and Carruthers, specimens from Kiltorcan, collected by Mr. Bailey, and very near 

 to mine from Gasp6. They seem, however, to be branches of the remarkable tree referred 

 to above under Stigmaria. At the same time, they are evidently identical with Haughton's 

 Cycloitigma; which genus, I think, must stand, though founded on imperfect specimens; 

 sines the plants in question are of a very distinct type, not to be included with the Lepidoden- 

 droid or Sigillaroid trees, though they exhibit characters in some respects intermediate. 

 |i t These shells are attached to some of the leaves of Cordaites Robbii in Prof Hartfa 

 eollectiona, and I have noted the same fact as occurring at Gaap^, though the specimens 



f ' .:: - ■■» 



