If a section of the scale be made (fig. 310. a), its under surface will be found to have a conca vity corresponding with the convexity of the up per. If the section has passed through one of the ridges, it is seen that the osseous tissue of the scale is of two kinds ; a superficial homoge neous-looking, dense, comparatively thin layer, and a deep, thick, laminated portion. If traced from the centre of the scale to its an terior circumference the superficial layer loses its continuity, breaking up into conical bodies, which are the sections of the detached calca reous spots mentioned above ; the deep layer thins out, its laminm gradually becoming fewer, and leaving a soft membranous space between their upper surface and the under surface of these spots. In the centre of the scale again, a series of rounded apertures are seen in a tangential section, the sections of canals which radiate through the scale and become more numerous and wider towards its margin. They are connected below with vertical canals passing through the laminated layer, and anteriorly they pass into the wide membranous space above referred to. There is no histological difference of any importance in the structure of these two layers ; each is composed of true bone with radiated cor puscles ; the upper being more dense and ho mogeneous, the lower less dense and lami nated.
If a section be tnade through several of the ridges of the upper surface, it will be seen that they are entirely composed of the hard homo geneous osseous tissue. On their sides, how ever, and in the valleys between them, more or less of soft integument remains, w hose pigment masses give the valleys a dotted ap pearance. On the other hand, a section of one of the detached tubercles shows, except in its consisting of osseous tissue only, that it is identical with a single spine of the Skate (fig. 310. A). It appears to me, therefore, that there can be no doubt that the ganoid, over lapping scale of the sturgeon commences by an isolated placoid spine; that other spines are developed around this, and their bases uniting, constitute a placoid scale, between whose elevations little valleys, bridged over by the soft integument, remain ; that to the base of such a plate as this, continual additions of osseous laminm are made, the radiating Haver sian canals being left between the first laminm and the superficial plate ; and finally that, extending in size, the anterior face of this complex scale becomes over-ridden by the preceding one. Complicated as it may ap pear, it is obvious that all this structure results from the continued endogenous growth and union of the primary ecderonic calcareous deposits, which constitute, as it were, so many centres of ossification for the large scale. The
final structure, however, is (if we leave out of consideration its histological character), to all intents and purposes, that of a cycloid scale ; and its mode of growth is identical with that of the large cycloid scale described by Prof. Williamson.
The increase of the scale is concentric ; addition being made to its posterior, as well as to its anterior edge and surface ; the only difference being, that in the latter case the development of the upper layer is less rapid than that of the lower, while in the former they are coincident; that soft membranous separation therefore, which exists between the two layers anteriorly, is far less developed posteriorly ; and the soft continuation of the scale which is flat anteriorly, is inflected pos teriorly ; the process of addition being other wise the same. Suppose, now, that each detached calcareous centre of ossification as it is added to the posterior margin of the scale, instead of being flattened, were produced into a spine as in the Rays, then it is perfectly clear that instead of a cycloid scale, the result would be a serrated ctenoid scale. And this appears to be exactly what takes place in the scales of the perch, according to Prof. Williamson's description.
From all this, I think, we arrive at Prof. Williamson's conclusion, that fish-scales are essentially tegumentary teeth ; that like the latter organs, they result not from the calcifi cation of the cellular ecderon covering those folds of the integument, upon which they are developed and which correspond with the dental pulp, but by a calcareous deposit taking place beneath this, in what represents a deep layer of the ecderon ; finally that it is, for the present, an open question whether the deep layers of all scales are produced by a con tinuation of this process, or whether in some cases a deep truly enderonic structure may be added to this superficial ecderonic constituent to constitute the perfect scale. A process of the latter kind would, at any rate, find its parallel in the eventual union of the teeth of many fishes with their jaws, and in that of the plates of the chelonia with the vertebral elements.
§ 3. Histology of the tegumentary organs. — Having thus arrived at a general idea of the mode in which the various forms of integumen tary organs are produced from the primary morphological constituents of every integu ment, we have now to consider their minute histological elements and the mode in which these proceed from the indifferent tissue of which all organs are primarily composed.