1862.] DAWSOIT — DEYONIAN PLANTS. 317 



often of a regularly oblong form. They have numerous equal 

 parallel nervures, which were probably fibre -vascular, like those of 

 Ferns, as they present precisely the same appearance as the ner- 

 Yures of the plants of this familj preserved with them, and which, 

 in these beautiful graphitized specimens, are traced in deeper lines 

 of graphite than the film of the same material which represents the 

 intervening parenchyma. In the best-preserved specimens, the 

 leaf is quite smooth ; but in some the space between the nervures 

 rises into little ridges, so as to give a striated appearance. These 

 different aspects, however, often occur on different portions of the 

 same leaf. The present species so closely resembles G. horassi- 

 folia of the Coal-formation that it might readily be mistaken for 

 it ; but it differs somewhat in the form of the leaf, and still more 

 in the venation, the nervures in the present species being perfectly 

 equal*. 



In the paper already referred to, I have stated at length my reasons 

 for preferring, in the case of this plant and C. horassifolia, the generic 

 name Cordaites, to Poacites, Flahellaria, and Noeggerathia, all of which 

 have been appKed to such plants, together with others having no affi- 

 nity to them. To the name PyclinopliyTlum, proposed by Brong- 

 niart, this objection does not apply; but Cordaites, I believe, has 

 priority, and is due to the describer of the typical species. 



I associate the genus Cordaites with Lycopodiaceous plants without 

 hesitation, notwithstanding the peculiar character of its foliage, 

 because Corda has shown that its stem is strictly acrogenous in 

 structure, and of the same type with those of Lomatofloyos and 

 Ulodendron — a fact which excludes it alike from association with 

 Monocotyledonous plants and with Perns. 



It is worthy of notice that, while the leaves of Cordaites, unlike 

 those of Sigillaria and Lejndodendron, were not attached by narrow 

 bases, but clasping, they were still, like those of nearly all other 

 Devonian and Carboniferous plants, deciduous and capable of dis- 

 articulation, as is proved by the immense abundance of fallen leaves, 

 while the stems, probably remaining attached to the soil, are rare. 

 It is further to be observed that these leaves were rigid, and long 

 resisted decomposition ; on which account, no doubt, they formed a 

 favourite base of attachment for the little Sjnrorhes which swarmed 

 both in the Devonian and Carboniferous Periods. At St. John many 

 of these leaves are covered with these little shells. 



The leaves of the present species are very abundant in the shales 

 of the vicinity of St. John, and indeed are eminently character- 

 istic of them ; and on this account I regard the dedication of it to 

 my late lamented friend, Dr. Eobb, as specially appropriate. I have 

 not recognized this plant in the specimens from Gaspe or Perry ; 

 and the only indication of it in the New York collection is a frag- 

 ment of a leaf from the Hamilton group of Cazenovia, New York, 

 not sufficiently perfect to render its identification certain. 



* The nervures in C. borassifolia are alternately thick and thin ; but there is an 

 undescribed species in the Up])er Coal-measures of Nova Scotia which has equal 

 nervures. 



