DEVONIAN CHIM^ROIDS. 



135 



and upon a closer examination of these remains, I was more than ever convinced 

 that they could not be associated with a Chimaeroid. In the first place, in well- 

 preserved specimens the striae are sometimes continued longitudinally above the 

 "rib sockets," showing, in other words, that they were absolutely unlike vertebral 

 centra. (Cf. fig. 114, /.) Furthermore, and this is, I believe, most convincing, 

 several of the fossils showed a delicate flaring out at one end, like the mouth of a 

 trumpet, which at once suggested the lip of a molluscan shell ; a character in any 

 event distinctly non-vertebrate, not to say un-Chimseroid. I am also permitted to 

 state that it was the view of Professor Cope, to whom my specimens were shown, 

 that the "columns" could have nothing to do with vertebrates. It is probable, on 

 the other hand, that they represent fragments of the shells of mollusks, possibly 

 Cephalopods.* 



Figs. 113 and 114. "Vertebral columns" of "Silurian Chimaeroid," Dictyorhabdus 



priscus Walcott. The first figure after Walcott. 

 t, oblique laminar in the structure of the fossil, suggesting lines of growth. 



Fig. 115. Dental plates of Men- 

 aspis armata = Chalcodus (permi- 

 anus). Kupferschiefer. After spec- 

 imen in Berlin Museum. 



DEVONIAN CHIJVL-EROIDS. 



Chimaeroid remains, or, more accurately, what are generally accepted as such, 

 are widely distributed throughout the middle and especially the upper Devonian 

 rocks of northern Europe and North America. These are referred to the family 

 Ptyctodontidae, Unfortunately for accurate diagnosis the fossils are fragmentary 

 and the best results which can be obtained from them are briefly these: That in the 

 three genera all at present known Ptyctodus, Rhynchodus, and Palaeomylus, 

 dental plates were present which resemble closely those of Chimaeroids. On the 

 other hand, these plates were only four in number and their tritoral characters are 

 puzzling. Within the substance of the plate appear not a few tubercular tritors, 

 but a general series of tritoral points, sometimes arranged in lamellae, which in 

 turn may form a series of flat or curving surfaces tritoral in function. The tritoral 

 points are most conspicuous in Ptyctodus (fig. 116), where they form lamellae. In 

 Palaeomylus (fig. 117) they spread out diffusely, and in Rhynchodus (fig. 118) are 

 drawn together close to the rim of the plates, forming thus an extended sectorial 



*The cephalopod nature of Dictyorhabdus was early commented upon by Hyatt, a reference which I had over- 

 looked and for which I am recently indebted to my friend, Dr. C. R. Eastman. 



