TRANSACTIONS OF SECTION C. 733 



2. On some remains of Fish from the Upper Silurian Rods of Pennsylvania. 

 By Professor B. W. Claypole, B.A., B.Sc. (Lond.), F.G.S. 



The earliest vertebrate animals yet known from any part of the world are 

 some remains of lish in the Upper Silurian rocks of England. They are for the 

 most part of three types. First, short tin-spines, named by Ag'assiz Onchus 

 tenuistriatus ; second, fragments of shagreen, or the skin of a placoid fish 

 (Thelodus and Sphagodus), belonging probably to the same that carried the spine; 

 and third, ovate, finely striated plates or shields, supposed to be the defensive 

 armour of some fish, unlike any now living-. 



No one has doubted the ichthyic nature of the first and second of these three 

 forms. But as regards the third there has been much controversy. Evidently 

 allied to Cephalaspis, its right to the name of fish has been called in question, and 

 suspicion has been raised in regard to the whole family of the (Jepkalaspids. 



On the whole, however, it seems best to retain them in the class of fishes, and 

 to this conclusion Professor Huxley evidently inclines in the conclusion of his 

 'Essay en the Classification of the Devonian Fish.' One may expect some, or 

 even considerable, divergence of structure from the usual ichthyic types in such 

 early forms. 



These English fossils occur in the lowest beds of the Devonian (Cornwall), and 

 in the highest beds of the Silurian (Shropshire and Hereford). The well-known 

 Upper Ludlow ' bone bed ' has yielded them in considerable quantity, and one 

 specimen is reported by Sir C. Lyell in his 'Elements of Geology'" (18b'o) as 

 ' discovered from the Lower Ludlow, beneath the Ayniestry limestone.' Below 

 this horizon I have never heard of their occurrence. 



The English Ludlow, taken as a whole, has been usually correlated with the 

 Lower Helderberg of North America, and on good grounds, both containing 

 Eurgpterus and Pterggotus. The English Lower Ludlow and the Water-Lime 

 or basal beds of the North American Lower Helderberg are the lowest strata 

 containing these fossils. On both sides of the Atlantic they range from this level 

 upwards into the Devonian. 



The oldest vertebrate fossils yet announced from America are those found 

 in the Corniferous limestone or Lowest Devonian of Ohio. Possibly the beds 

 at Gaspe on the Gulf of St. Lawrence are somewhat lower; as they have yielded 

 Cephalaspis, which is not yet known i'rom Ohio, and Coccosteus, of which Ohio 

 has yielded only a single specimen. No authenticated fish-fossil has yet been 

 announced from the Upper Silurian rocks of America. 



It is true that reports of the discovery of such remains have been published 

 at various times, but investigation has proved them all erroneous. (See 

 ' Palreontology of New York,' vol. ii. pp. 819, 320, pi. Ixxi. ; 'American Journal 

 of Science,' second series, vol. i. p. G2 ; 'Palaeontology of Ohio,' vol. ii. p. 262.) 



During his recent work on the Palaeontology of Perry County, Pennsylvania, 

 the author came upon some fossils which at once suggested relationship to the 

 Ludlow group above described. Among them were a few spines recalling Onchus 

 tenuistriatus, but with some differences. He has named them Onchus Penn- 

 srjlvanicus. With them he discovered abundance of specimens bearing a strong 

 resemblance to Scaphaspis, but larger, and differing in some other respects. These 

 he named Palaaspis (P. elliptica and P. bitruncata). 



Comparing these with Scaphaspis we find them much thinner, not exceeding 

 one-fortieth of an inch in thickness; whereas specimens of Scaphaspis in the 

 author's possession from Cornwall are in some places much thicker. The striation 

 on both is equally fine, but is rather less regular on the American specimens. 

 These also show no trace of the spine in which the shield of Scaphaspis terminates, 

 as shown by Murchison in ' Siluria.' 



No traces of the English fossil shagreen — TheMus and Sphagodus — have been 

 found in the Pennsylvanian beds, though it abounds in the Ludlow rocks. 



The fossils were found in a bed of sandstone about 200 feet below the base of 

 the Water-Lime in Perry county, Pennsylvania ; near the top of the great mass of 

 variegated shale composing the Fifth Group of Rogers in the First Survey of 



