993 



1879.] -J- 1 '-* [Lesquereux. 



pedicel, five millimeters broad, distinctly joining the rounded subtruncate 

 apex of one of the bolsters. The dividing of the shale in exposing and 

 detaching the upper side of the fruit distinctly shows its mode of attach- 

 ment by the continuity of the pedicel to the top of one of the bolsters. 



The Plate I represents : Fig. 1, the fossilized part of the branch trans- 

 formed into shining cannel coal. It is flattened to about two millimeters 

 in thickness, the bolsters and their elongated base are in relievo. In the 

 counterpart, fig. 2, the bolsters, concave and marked especially by the pro- 

 jections of the borders, are a little enlarged by compression. The part e is 

 broken, as is also the base of the fruit in d. But the counterpart, fig. 2, 

 has the impression of the whole fruit with its outside envelope, and the 

 base is seen fixed upon the top of one of the bolsters, or rather of its im- 

 pression in e. 



This diagnosis and the figure of this branch of Cordaites, explain the 

 position of the fruits as placed in spiral around a narrow branch, in a kind 

 of long strobile. The close position of the bolster, serving as a support, 

 and narrower than the fruits, at least when in maturity, could not allow 

 them to remain upon the stem even until entirety ripe. They were forced 

 out and soon falling off. They do not appear to have been axillary, for 

 the bolsters do not bear any scar indicating the position of a leaf of 

 Cordaites under the points of attachment of the fruits. The flowers were 

 monoecious, even perhaps dioecious. A branch of the same species, C. cos- 

 tatus, figured PI. LVIII, fig. 1, lb, of the U. S. flora, bears racemes of 

 male flowers whose pedicel is too slender for the axis of a strobile. The 

 fertile flowers were in separate racemes, either on different trees or pend- 

 ing from another part of the same tree. 



The discussion on the systematic relation of the Cordaites finds its place 

 in the U. S. Coal flora. I will merely remark now that the position of 

 these fruits in a kind of agglomeration in loose strobile, is comparable to 

 the fruitifications of Cycadese, rather than to those of Conifers. 



The Gingko, it is true, has nutlets resembling the. fossil fruits by their 

 shape. They are, however, much smaller, supported by a basilar cupuli- 

 form disk, with longer axillary fasciculate pedicels. Separate fruits of Cy- 

 cadese have the size and form of the Cordaicarpes. Compare among many 

 others, Cycadosperum sulcatum, in Heer, fl. Helv., IV, PI. LVI, fig. 18-20 

 of the Jurassic. 



I do not suppose, however, that the Cordaites are positively referable to 

 Cycadese. They had compound characters, which have been later sepa- 

 rated and diversely distributed in other vegetable tribes. They constitute 

 therefore a peculiar separate group with analogies of various kinds. We 

 find the same in the ferns of the coal, of which the multiplied references 

 made to genera of our time, have proved a series of failures especially 

 considering the fructifications. It is the same also with the Lycopodiacese 

 as represented by Lepidophlois ; with the Sigillarise, the Calamodendrse, 

 Asterophyllites, Sphenophyllum, Annularis and others. 



Columbus, 22d March, 1879. 



