FOSSIL PLANTS. 493 



an internal ovule or seed, covered with three different envelopes like those of 

 Trigonocarpum Nceggerathu, of which a fine specimen from Mazon creek has 

 been figured in this Report. They resemble fruit of palms, and have been re- 

 ferred by authors to species of Cordaites or Nceggerathia. Still less can it be 

 proposed for winged seeds or fruits generally described under the generic name 

 of Cardiocarpus and Rhabdocarpus. These fruits, whose place in the vege- 

 table kingdom is still problematical to palaeontologists, are far better known 

 from American than from European specimens, and their analysis can, there- 

 fore, be pursued with chances of more satisfactory results. One of them is de- 

 scribed and figured in its whole, in the Arks. Geol. Kept., vol. i, p. 311, pi. 4, 

 fig. 4, as Cardiocarpus ingens, Lesqx. Another, still more remarkable, has 

 been published by Dr. Newberry, in the Annals of Science, of Cleveland, May 

 1853, p. 152, N. 2, as Cardiocarpus samarceformis. A third has been obtained 

 in good and numerous specimens from the shale overlaying the coal of Coshoc- 

 ton, Ohio, by Rev. H. Herzer. This peculiar fruit, Ptilocarpus bicornutiis, 

 Lesqx., (1) is composed of a small oval seed, pointed downwards, rounded or ob- 

 tuse at the top, obscurely ribbed in the length, attached to the inside of an 

 oval scale, elongated upwards, diverging at its base into two short horns and 

 overlapping the seeds by its border. The seeds, though generally found con- 

 nected to the winged scales, are easily detached from it, and indeed all the 

 specimens which I have examined, show the seeds already half detached from 

 the top downwards, and to prevent them becoming lost, I had to take them 

 out of the specimens and preserve them separately. This connection of a small 

 oval seed to one side of a winged scale, point out evidently the relation of this 

 fruit and of others related to it, and mark their places as belonging, if not to 

 true Conifers, at least to the Gymnosperm family. However peculiar they may 

 be in their form, though different from seeds of the species of our time, it is 

 scarcely possible to compare them to any other family of vegetables. As the 

 seeds are generally found separate from their scale, a number of fruits of our 

 Coal Measures are probably referable to the new genus, Carpolithes multistri- 

 atus, for example. For some of the numerous specimens from Colchester, 111., 

 bear evident remains of scales overlapping the seeds like a broad rim, more or 

 less lacerated and partially destroyed, especially towards the point. And in 

 the nodules of Mazon creek, where these seeds are preserved in their original 

 form, they appear merely tumid in the middle, as compressed under a scale, 

 and not cylindrical. Rhabdocarpus clavatus is, perhaps, also a seed of the 

 same kind, as are evidently the species of Cardiocarpus published by Dr. New- 

 berry, loc. cit., and many European species like Rhabdocarpus mammilla 'tus, 

 Artis, etc. 



(1) As the name indicates, this new genus Ptilocarpus is established for the special de- 

 scription of winged fruits having an affinity to those of the Conifers. 



