418 MESSRS. HANCOCK AND ATTHEY 



that we can have little doubt they belong to them.* Among 

 these fragments are four or five which exhibit the greater por- 

 tion of the contour of the scale, and one which has it almost 

 entire. These are all parallelogramic in form, are thin and deli- 

 cate, and apparently represent three species, though the distin- 

 guishing characters are slight. 



The first (PI. XVI., fig. 1), the largest and most perfect speci- 

 men, measures two inches and a half long, and upwards of two 

 inches wide. The sides are parallel ; the anterior extremity (a) 

 is a little arched outwards, and the posterior or exposed extre- 

 mity (b) is rounded ; the angles are rounded off; the central 

 area (<?), under an ordinary hand-lens, appears quite smooth, and 

 is bordered by a rather narrow margin (c) having several con- 

 centric undulations or lines of growth, and marked with minute 

 radiating striae ; no growth-lines are visible within the marginal 

 border. On examination with the inch object-glass, the central 

 area is found to be finely reticulated with slightly elevated bony 

 fibres, the meshes being sunk, so that the surface is minutely 

 and regularly punctate. This is undoubtedly the underside of 

 the scale ; the upper surface is revealed on fragments, and, at a 

 rupture (d) near the centre of the rounded exposed extremity, is 

 minutely granular. Of course, in the latter case, it is only the 

 cast of the upper surface that is seen ; and at this point it is 

 evident that the granules are enlarged and become arranged so 

 as to form imperfect and very irregular vermicular grooves. 



* As this paper was passing through the press, we obtained complete proof of the truth of 

 this opinion in a fine specimen of the greater portion of a cranium and part of the trunk of 

 a large Ctenodus with the opercular plates attached : a considerable number of the ribs are 

 exhibited in connexion with the head, disposed in natural order ; and numerous neurapo- 

 physes and apparently interneural spines are scattered along the dorsal ridge. Everywhere 

 mixed up with this interesting specimen these peculiar scales are found, much broken, in- 

 deed, but occupying both sides of the body portion of the fish, in such a manner as to leave 

 no doubt on the subject. The scales are very similar to those described in the text, differ- 

 ing only specifically, the margin being wider; the smooth central area has the same peculiar 

 minute surface-structure, and the upper surface is minutely granulated in the same manner. 

 Moreover this specimen shows the hatchet-shaped bones, or clavicles, described by us on a 

 former occasion, in connexion with the cranium, almost in their natural positions ; so that 

 here we have not only proof respecting these scales, but the true nature of the hatchet- 

 shaped bones is also established. 



