OE THE AMEEICAX TTEKASPIDIAN, PAL^ASPIS. 549 



Museum.' This palaeontologist has there figured Pteraspis^ armour- 

 clad both above and below. 



It is not yet possible to refer all the diiFerent species of Scajphaspis 

 directly to the several species of Pteraspis and Cyathaspis to which 

 they must severally belong, and probably the line of argument 

 indicated by Lankester will be the only available one until some 

 geologist is fortunate enough to discover the various species of 

 Scaphasj)is in juxtaposition with their corresponding species of 

 Pteraspis and of Cyathaspis. 



YI. The two Shields of Pal^aspis. 



Eepeated and careful study of my specimens, both new and old, 

 since 1883, has from time to time led to new observations regarding 

 them, but it was not until I had seen the paper by Yon Alth, quoted 

 above, that I examined them in the light of this new theory of 

 their structure. Considering that almost all my fossils are preserved 

 in a sandstone-matrix, I was not sanguine of finding any confirma- 

 tion of his view among them. The total result of the studies that 

 I have made since the publication of the original description it is 

 the purpose of the present paper to set forth. 



Among the mass of material in my possession I have discovered 

 two specimens in which the same relative position of the shields 

 may be observed as that which was noted by Kunth and Yon Alth. 

 This had not altogether escaped my notice when several years ago I 

 first extracted the specimens from their matrix. But the fact, to 

 which Lankester, with wise caution, alludes in his tract on Holaspis, 

 that the shields of these fishes are frequently found packed 

 together as so many shells, had so often come to my notice in 

 Pennsylvania that I naturally attributed this fact also to the same 

 cause. In this I was somewhat hasty, as it now appears, inasmuch 

 as I did not lay due stress on the circumstance that in these two 

 cases the relative position of the shields was contrary to that of the 

 great majority. Though nothing could be more natural than that a 

 number of saucer-shaped shells should pack themselves together 

 just as a pile of plates is packed, yet it was far less likely that any 

 two of these should come to rest with the concave sides facing each 

 other and with great exactness. Moreover, at that time I had not 

 had an opportunity of reading Kunth's paper, which, in spite of his 

 erroneous interpretation, might have raised a suspicion of the 

 true significance of these two specimens. But, considering the 

 sandstone-matrix and the scattered position of the great majority 

 of the specimens, were it not for the confidence afi'orded by the two 

 already-mentioned fossils, I should even now hesitate to put 

 forward my own as evidence. However otherwise doubtful, the 

 occurrence of the shields of Pcdceaspis in the same relative position 

 serves now only to bring the American species into line with their 

 co-ordinates in Europe, and so to lend confirmation and to gain 

 probability at the same time. I feel therefore that no reasonable 



