524 MR F. R. COWPER REED ON 



Dow Hill there are present the two narrow submarginal folds, one on each side of the 

 median one close to the edges (Plate, fig. 7), as in the complete individual of T. Peachi 

 from the Starfish bed. 



The slender attenuated apex of the kite-shaped plates (Plate, fig. 12) and the broader, 

 more rapidly tapering form of the plates, is a more or less marked distinction from 

 T. Peachi ; the curvature of the plates depends on what part of the series they come from, 

 and cannot be considered a constant or reliable specific difference, but the lamellae are more 

 numerous and closer together than in T. Peachi and meet the anterior margin at nearly 

 a right angle, while they curve forwards sharply towards the apex on the posterior 

 margin of the plate (Plate, figs. 9, 11, 12). 



The isolated triangular plates which may be referred to the median series 

 (e.g. Etheridge and Nicholson, op. cit., pi. xiv., fig. 26) are much like those of 

 T. Peachi, but usually have a more distinct and broader rounded longitudinal groove 

 on each side of the rounded median fold or undulation (Plate, fig. 8). 



The plate with the rounded apical end figured by Etheridge and Nicholson from 

 Balclatchie (op. cit., pi. xiv., fig. 27) is also represented amongst Mrs Gray's specimens 

 from Dow Hill (Plate, fig. 10), and as the pointed kite-shaped ones have been shown 

 to belong to the anterior end of the body in T. Peachi, it is probable that this rounded 

 plate belongs to the posterior end. Tt must be regarded as one of the kite-shaped, 

 lateral series on account of its median sharp fold, though its apex has been modified. 

 As Etheridge and Nicholson pointed out (p. 215), such a plate with an obtuse apex 

 was called by Barrande (op. cit., p. 567) the " Valve fenestree " ( = " cancellated plate "), 

 but they did not remark that this plate from Balclatchie differs from all the other plates 

 of T. scotica by being bilaterally symmetrical. The median fold in it is truly placed in 

 the middle ; it ends abruptly in the subcircular depression round which the lamellae 

 curve. The lamellae are parallel to each other and at right angles to the median fold, 

 and curve forward very slightly (but to an equal extent on each side near the margin), 

 while at the front end they sweep round in a circle and unite around the depression at 

 the end. of the median fold. Until a complete individual is found the true relation of 

 this plate to the others cannot be determined. The same may be said of the strongly 

 convex small triangular plate, such as Etheridge and Nicholson figured on pi. xiv., fig. 25. 

 This plate (Plate, fig. 13) has a thick calcareous shell preserved, and the whole surface is 

 strongly bent down on each side of the median line ; it is asymmetrically triangular, 

 and if flattened would be referred to the left row of the double median series. It is 

 difficult to realise its relation to the rest of the plates on the body, as they all seem to 

 be more or less flattened. In shape and convexity it much resembles the plates of the 

 type of the genus, T. Wrigldii, from the Wenlock. As the similarly shaped plates of 

 Turrilcpas from the Orthis argentea zone of Haverfordwest and from the Dufton 

 Shales have an intermediate convexity, it seems likely that the much flattened condition 

 of the plates is largely due to the thinness of the shell or to crushing, or to both 

 Cdmbitied. It should be mentioned also that in this convex specimen from Balclatchie 



