Dr Traquair on the Genus Gyracanthus [Agassiz). 95 



lefts," amply justifies the opinion of Messrs Kirkby, Atthey, 

 and Hancock that they were pectoral or at least paired appen- 

 dages. Which are the right and which the left spines, it is, 

 however, at present not very easy to determine. Accepting 

 the sulcated aspect as posterior, it would be necessary to 

 ascertain whether the flat or the gibbous side was superior in 

 order to indicate to which side of the fish it belonged. 



Now, turning to the spines labelled " dorsal " in the same 

 collection, we find that they are smaller in size, varying in 

 length from 4 J to 10^ inches, and almost all lying laterally 

 compressed on pieces of shale. In this way the lateral cur- 

 vature is obscured, though in one, also marked "dorsal," which 

 happens to be only obliquely placed on its matrix, this 

 curvature is quite obvious. Furthermore, all of them show 

 in other respects the same want of lateral symmetry which 

 I have just described in those acknowledged to be pectoral, 

 namely, the possession of a flat and of an inflated and grooved 

 side ; in fact they are rendered still more asymmetrical than 

 the large truncated spines by the much greater prominence 

 and sharpness of the posterior marginal keel, which we have 

 seen is morphologically a lateral structure in the general plan 

 of the spines. This keel is also furnished with a row of small 

 closely-set recurved denticles. The gyrating ridges become 

 very oblique towards the point, and tend to become plain or 

 only distantly nodulose, except perhaps on the front of the 

 spine. On the flat side a space bare of ridges runs down from 

 the point along the posterior margin for about IJ inch, and 

 an analogous appearance is also observable on the grooved 

 side. The groove itself is smooth and marked with delicate 

 longitudinal striae ; and, as Messrs Hancock and Atthey have 

 already noted, the point is much compressed laterally.^ 



If we next compare the proximal or basal end of one of the 

 largest of these supposed dorsal spines with the distal extre- 



^ These young spines of G. tuleraidatus bear an extreme resemblance to the 

 figure of G. denticulatus (Davis), in Ann. & Mag. Nat. Hist. (5), vi,, 1880, p. 

 373, "being similar in shape, in the characters of the gyrating ridges, and the 

 denticulation of the posterior margin, while the same bare space runs down 

 for a little distance from the point. Mr Davis, however, states that his spine 

 has tv)o rows of denticles posteriorly. 



