Dr. R. H. Traquair—Devonian Fishes from. Canada. 149 
extent of the margins of the antero-lateral and anterior ventro-lateral 
plates with which it articulates, it forms no more of an independent 
projection than in the former.’ But as in Phlyctenaspis this spine 
seems divided by a very distinct suture from the interlateral plate, I 
have carefully examined the suite of specimens, both of Coccosteus 
decipiens and C. minor in the Edinburgh Museum, with the result 
that I have convinced myself that the process in question is here 
also a distinct element, to which the name of “lateral spine”? may 
appropriately be given. In both genera it is immovably fixed; in 
one example of Phlyctenaspis it seems indeed anchylosed to the 
interlateral plate. 
Consideration of this element of the cuirass of the Coccosteide 
calls up of course the question of Acanthaspis, Newberry. Mr. 
“Smith Woodward” remarks that the spines of Phlyctenaspis “ are 
indeed precisely similar in external form and appearance to those of 
Acanthaspis; and so far as can be judged from known specimens 
they only differ from the last-mentioned spines, in the circumstance 
that the supporting plate is destitute of the extended oblique pedicle 
observed both in the type specimens from the Corniferous Limestone 
of Ohio, and in the shield assigned to the same genus from Spitz- 
bergen. It thus remains to discover more associated examples of 
the plates and spines from Ohio, to determine whether they actually 
pertain to Ostracoderms, as suspected, or whether they represent - 
part of the armour characteristic of Arthrodira; for the fixed 
spinous appendage is now proved to occur in both of these widely 
-separated groups.” 
It will be remembered that Mr. Woodward, in a previous paper,? 
was inclined to consider Acanthaspis as being ‘‘ most nearly related 
to the Asterolepide, and its spine-like processes to be the homologue 
of the well-known pectoral-paddles of the latter.” But a com- 
parison of Newberry’s figures with the sketch of the ventral cuirass 
of Phlyctenaspis given above will remove all doubt as to the Coccos- 
tean character of the remains from the Corniferous Limestone of 
Ohio; figures 1, 2, 3, and 4 on plate lv. of Prof. Newberry’s work + 
representing obviously the lateral spine in conjunction with the 
anterior ventro-lateral plate of a creature allied to the Canadian 
genus, though the supporting plate is remarkably short from before 
backwards. The same view must, I think, be adopted for Mr. 
Woodward’s Acanthaspis decipiens from Spitzbergen, so that I cannot 
consider the occurrence of a fixed spinous appendage to be as yet 
proved in any genus specially allied to the Asterolepide. Whether 
this fixed spine is in any way homologous to the articulated ap- 
pendage of Asterolepide, or to the pectoral limb of vertebrates in 
general, I need not at present discuss. 
Associated with other plates of Phlyctenaspis Acadica, are several 
fragments of dentigerous bones, showing small, smooth-pointed 
conical teeth in one row. 
1 R. H. Traquair (Guon. Mae. Dec. III. Vol. VI. Pl. I. Fig. 3. 
2 Op. cit. p. 481. 
3 “ Devonian Fish-Fauna of Spitzbergen,” Ann. and Mag. Nat. Hist. (6) vol. viii. 
pp. 1-15. 4 Geol. Survey of Ohio, Paleontology, vol. i. 
