The form combines in an unusual and remarkable degree reptilian and amphibian characteristics. The leg bones, pelvis, and tarsus are all strikingly reptilian, but the phalanges in the arrangement of elements are so typically amphibian that if we had no other means of diagnosis we would incline to locate this Coal Measures species among the Amphibia. The leg ([plate 23, fig. 1]) recalls in its structure that of another Coal Measures species, Eosauravus copei Williston, which is, however, clearly a reptile. While there is a general degree of similarity between the foot structure of Eosauravus copei and Ichthycanthus platypus, yet there are very great differences in the phalangeal formula and the arrangement of the tarsal elements. These differences are clear and indicate a separation of the two species into distinct classes. The phalangeal formula in the Eosauravus, 2-3-4-5-4, is typically reptilian; while in the Ichthycanthus, 2-2-3-3-3, it is amphibian. The tarsus of the Ichthycanthus is amphibian in the presence of an intermedium, but this is very small and the remaining tarsal structures have nothing which might not be found in an early reptile. There may be a single or even two centralia in the reptilian tarsus among the early forms.
The amphibian nature of the species having thus been established, it remains to give a detailed account of its skeletal anatomy, with comparative references to such other ancient forms as are available. Little can be said of the vertebral column, since only the molds of a few vertebræ remain, and these are so obscured by a closely adherent pellicle of carbonaceous material that their form can not be distinctly discerned. They are high, with relatively broad neural spines. There are no ribs preserved. The pelvis is obscured, but it is possible to determine the presence of an elongate ilium and an ischium. The leg of the left side is the best preserved of all the elements, and it is to this that our attention will be confined. The opposite leg is not so complete, yet all the long bones and a part of the tarsus are preserved with sufficient clearness to corroborate the findings of the left side.
The femur, as has been stated, is reptilian in appearance. This is due to the well-rounded articular surfaces, as though the endochondrium were well developed, and to the large development of the greater and lesser trochanters, which are quite prominent, though these are distorted and depressed in fossilization. The bone is stout and well built and its form suggests an active habit of life. The tibia and fibula are separate, and do not otherwise have sufficiently noteworthy characters to call for a special description in this place, except to note an unusual anterior crest on the tibia. To the lower ends of these bones articulate the first row of tarsal elements, the tibiale, intermedium, and fibulare. The tarsus is composed of 9 elements arranged in 3 rows. The proximal row is composed of the tibiale, the intermedium, and the fibulare. On the edge of the tibiale there lies a portion of one of the caudal vertebræ, so that the form of this tarsal element is slightly obscured. The intermedium is a small, rounded element lying between the larger elements. The fibulare is rectangular and projects a considerable distance out from the tibia, but articulates directly with the large lateral distal tarsal. The centrale is triangular in form and is opposed directly by the tibiale and tarsalia 1 to 3. The phalanges are robust in appearance. The entire foot gives one the impression of a very broad structure. The ungual phalanges were apparently bluntly clawed.
[CHAPTER XXIII.]
SUPPOSED MICROSAURIAN SPECIES OF UNCERTAIN RELATIONSHIP.
The following three species are so unusual and so incompletely known that they can not be considered with any of the above families: Brachydectes newberryi Cope, Linton, Ohio; Amblyodon problematicum Dawson, Nova Scotia; Proterpeton gurleyi Moodie, Danville, Illinois.
Genus BRACHYDECTES Cope, 1868.
Cope, Proc. Phila. Acad. Nat. Sci., 1868, p. 214.