520 MR F. R. COWPER REED ON 



the type specimen * and other examples from the same horizon and locality, it is 

 abundantly clear that this " keel" is really a narrow longitudinal sharp fold appearing 

 on one side as a ridge or keel and on the other side as a groove. Euedemann (op. cit., 

 pp. 88, 89) likewise has noticed that the apparent carina in Lepidocoleus Jamesi is the 

 reverse side of a groove. In T. Peachi the groove, or rather fold, is on the outer side of 

 the plate, and as it is the substance of the lamellated plate itself which is folded, the 

 lamellae are affected ; there is not the least evidence of the presence of a median rod or 

 anything resembling the main rib of a leaf such as Etheridge and Nicholson indicated 

 in their figures of T. Peachi and T. scotica, and described as a ridge-like keel. In both 

 species the lamellae which compose the plates are of regular and equal size, and the 

 longitudinal fold which has been mistaken for a keel traverses them at right angles and 

 is not strictly in the middle of the plate. Moreover, all of the kite-shaped plates in the 

 type of T. Peachi show a narrower, less deeply impressed fold or groove above the 

 main one, usually about half-way between it and the anterior margin, running similarly 

 to the apex parallel to the edge of the plate. This groove is stronger in some detached 

 plates than in the type. It should also be noticed that the lamellae on the upper 

 (anterior) half of the plates, i.e. above the median fold, do not curve forward so sharply 

 or overlap to such an extent marginally as on the lower (posterior) half, but meet the 

 upper margin nearly at right angles instead of at a very acute angle. This cannot be 

 seen in the specimen figured by Etheridge and Nicholson (op. cit., fig. 8), as it is the 

 inner surface of the series of plates which is exposed to view ; for the plates themselves 

 overlap from behind forwards in such a way that the anterior half of each plate lies 

 above the posterior half of the preceding plate, so that the anterior portions of all the 

 plates are consequently hidden in an inner view of the series. (Plate, fig. 4.) 



T. Peachi is the only species from G-irvan in which the plates have been found in 

 natural serial position, and in the type specimen from Whitehouse Bay (Plate, fig 4) 

 there are five pairs of the kite-shaped plates present, though only those on the right side 

 are well preserved. The longer axes of the plates in this specimen are inclined at about 

 75° to the axial line of the body of the fossil ; in the other figured specimens (op. cit., 

 figs. 9, 10) the divergence of the lateral plates becomes progressively less towards the 

 anterior end of the body, as Etheridge and Nicholson remarked. The apical ends of 

 these lateral plates are free, and in the type the surface of each plate seems not to be 

 quite flat but to have weak broad longitudinal undulations, as well as the two sharp 

 folds. In specimens such as this one, showing the true inner surface of the plates, the 

 lamellae are usually less obvious than on the outer surface. Portions of the thin 

 calcareous shell are still adhering in many cases. 



An important feature, not described or accurately figured by Etheridge and 

 Nicholson, is distinctly traceable in the oft-quoted type-specimen of T. Peachi here 

 figured. This consists in the presence of a double median longitudinal row of strongly 

 overlapping plates between the bases of the lateral kite-shaped plates. The plates of 



* Etheridge and Nicholson, of. cit., pi. xx., fig. 8. 





