HOW FISH ARE CLOTHED. 29 



plates we should find if it was taken from a 

 roach that it partook somewhat of the shape of 

 a human nail. Furthermore, we should probably 

 have noticed before removing it that its anterior 

 end was thrust deep down into a sort of pocket 

 in the skin, whilst its posterior, or hinder end, 

 was free, and could be easily raised by any 

 pointed or blade-shaped instrument. Such an 

 arrangement of horny plates or scales may be 

 taken as typical of the majority of living fishes. 

 Any variation of this arrangement may be re- 

 garded, roughly speaking, either in the direction 

 of further specialisation or of degeneration. 



Instances of degeneration are numerous. Pro- 

 bably we should be correct in regarding the first 

 indication of degeneration to be the isolation of 

 the scales. In such cases the scales, instead of 

 overlapping, remain perfectly distinct from one 

 another. The African lung-fish (Protopterus), and 

 certain wrasses, are instances of this kind. In 

 some carp known as " mirror-carp " this isola- 

 tion of the scales is very marked. Those along the 

 sides of the body have assumed relatively enor- 

 mous proportions; those along the top of the 

 back are smaller, but all are widely separated 

 from their neighbours. In many eels, as we 

 have just remarked, the scales have become so 

 reduced in size that they must be sought for 

 with a microscope, and then are found to be 

 deeply embedded in the skin. 



Of instances of specialisation we have a great 

 variety. Thus, to take a few of the most 

 striking. One of the cat-fishes of tropical South 

 America, known as Callichthys, has the scales 



