592 



VBRTEBEATED ANIMALS. 



beautiful St. Andrew's cross ; and if a plate of selenite be placed 

 behind it, and the analyzing prism be made to revolve, a re- 

 markable play of colors is presented. 



409. In studying the structure of the more highly-developed 

 scales, we may take as an illustration that of the Carp ; in which 

 two very distinct layers can be made out by a vertical section, 

 with a third but incomplete layer interposed between them. 

 The outer layer is composed of several 

 concentric laminae of a structureless trans- 

 parent substance, like that of cartilage ; 

 the outermost of these laminae is the small- 

 est, and the size of the plates increases 

 progressively from without inwards, so 



Fio. 307. 



Fig. 306. 



Sf>^ i 



Poruon of Skill of Sole, viewed as an opaque object. 



■ill ''JA^ 



Scale of SoU, viewed as a 

 IranspareiU olijecl. 



that their mai'gins appear on the surface as a series of concentric 

 lines; and their surfaces are throAvn into ridges and furrows, 

 which commonly have a radiating direction. The inner layer is 

 composed of numerous laminae of a fibrous structure, the fibres 

 of each lamina being inclined at various angles to those of the 

 lamina above and below it. Between these two layers is inter- 

 posed a stratum of calcareous concretions, resembling those of 

 the scale of the Eel ; these are sometimes globular or spheroidal, 

 but more commonly "lenticular," that is, having the form of a 

 double convex lens. The scales which resemble those of the 

 Carp in having a form more or less circular, and in being desti- 

 tute of comb-like prolongations, are called cycloid ; and such 

 are the characters of those of the salmon, herring, roach, &c. 

 The structure of the " ctenoid" scales (Fig. 307), which we find 

 in the sole, perch, pike, &c., does not differ essentially from that 

 of the " cycloid," save as to the projection of the comb-like teeth 

 from the posterior margin; and it does not appear that the 

 strongly-marked division which Prof. Agassiz has attempted to 

 establish between the "cycloid" and the "ctenoid" orders of 

 fishes, on the basis of this difference, is in harmony with their 

 general organization. Scales of either kind may become con- 

 solidated to a considerable extent, by the calcification of their 

 soft substance ; but still they never present any approach to the 



