HAIRS, FEATHERS, AND SCALES 35 



distinct layers; the upper one which bears the concentric 

 lines, and a lower clear one which not only fills the radiat- 

 ing bands, but underlies the whole of the lined parts. The 

 concentric lines of the upper layer do not now appear to 

 be edges of successive plates, but irregular canals running 

 through the solid substance. This, however, is illusory: 

 for, by delicate focusing, we perceive that each portion 

 marked by these lines is really in a different plane from 

 the others, that the highest is at the centre of radiation of 

 the scale, and that each is successively lower till we reach 

 the margin. But now, if with very sharp scissors we cut 

 one of these scales longitudinally through the centre, and 

 examine the cut edge, we find that each of these lines 

 forms a distinct ridge. On the other hand, the under layer 

 of clear substance is quite smooth, and always a little ex- 

 ceeds the margin of the concentrically lined portion. The 

 clear substance that fills the radiating slits agrees both in 

 texture and level with this lower layer, and is manifestly 

 continuous with it. 



Hence, I think that, in these slit scales, the upper layer 

 is formed, as commonly believed, by successive deposits 

 from beneath; but that, after a few have been deposited, 

 they begin to slit, probably by contraction in becoming 

 solid ; that the lower layer is formed after each upper one 

 is hardened, exceeding its length by a little, and filling up 

 the slit; that this lower layer becomes the upper layer of 

 the next course, slitting, and turning up its terminal edge 

 as it hardens; that then the lower layer is deposited on 

 this, filling up the slit as before; and that this process 

 goes on as long as the fish lives. 



It is curious that, in the scales of the Pike, the por- 

 tions thus separated by slitting, instead of expanding and 



