F. R. Cowper Reed — Fossils from Girvan. 293 



perfect individual occurs amongst them. Some of these fossils have 

 a slightly convex or domed surface composed of small plates, which 

 much resemble in appearance those described and figured by Rauff ^ 

 as belonging to the wall . of H. orhis, Eichwald, as preserved in 

 a limestone pebble of Ordovician age from AVartin in Pomerania. The 

 Girvan organism must have been elliptical or subcircular in shape, and 

 the more or less marked convexity of the upper surface, which these 

 ■plates cover, seems to be an original and natural feature and not due 

 to distortion or pressure. The plates themselves on the upper surface 

 are small, numerous, of regular rhomboiclal shape, and of equal size, 

 about seven plates measuring 10 mm. along their greater diameters, 

 and about nine plates measuring also 10 mm. along their shorter 

 diameters. The plates are fitted closely together into a pavement, as 

 usual in the genus ; the boundaries of the individual plates are not 

 very sharply marked, being partly obscured by their ornamentation 

 and perhaps fused to a certain extent along their edges. In one 

 specimen there is a pore at each angle between the plates, as Hinde ^ 

 has shown in his figures of H. occidenfalis. Each plate is gently 

 convex, rising slightly towards the centre, in which is a more or less 

 indistinct small depression, from which radiate to the sides a number 

 of short, broken, irregular, and sinuous ridges and vermiculate grooves, 

 interspersed with or replaced by small closely-set pits and tubercles 

 of the same size and elevation, which by their fusion seem to produce 

 the above-mentioned ridges and grooves (PL XII, Pig. 1). On many of 

 the plates only pits and tubercles are present, and no radial arrangement 

 is noticeable. Prom the examination of both the natural casts and 

 the external impressions of the same specimens which have been 

 carefully collected by Mrs. Gray, and from the regular occurrence of 

 these surface-features, it becomes clear that they are not due to 

 weathering, though Rauff was in some doubt on this point in 

 connection with his specimen of R. orhis. Our Girvan species 

 difPers from the latter by the less distinct character of the radial 

 ornamentation, by the more traversely rhomboidal shape of the plates, 

 and by their reduced overlapping. 



A fragment of the lateral (lower) surface and margin of the same 

 species comes from the same horizon and locality. It shows a few of 

 the outer series of upper plates, apparently similar to those above 

 described but rather larger, situated near the edge of the upper dome, 

 and passing over into the flattened, smooth, and larger subquadrato 

 plates of the lower surface. These plates decidedly overlap, like 

 tiles, from above downwards, the lower overlapping angle being 

 generally rather prominent and thickened. A shallow median 

 depression or pit is generally noticeable on each plate, indicating 

 the point of attachment of the vertical ray or column of the spicule. 

 Ptauff has figured closely similar plates (op. cit., t. i, fig. 10) from the 

 upper surface of R. Neptuni. Occasionally traces of tubercles seem 

 to be visible on the surface of these plates, but this is doubtful. 



1 Eaufi: Abliandl. k. bayer. Akad. Wiss. Math. Phys. Kl., Ed. xvii (1892) 

 p. 688, t. iv, figs. 2-6. 



- Hinde, op. cit., pi. xxxvii, figs. Zd, 2>e. 



