4788 Fishes. 



acute observation, she proves generally the best of practical natu- 

 ralists. That the system of coloration in the Salmonidae cannot 

 always be depended on in characterizing the species may be ad- 

 mitted without under-rating its value. Human observation is not ex- 

 tremely refined ; it is, in fact, extremely imperfect ; besides, in the 

 case of the Salmonidae, the generic colouring is apt to reappear, 

 though imperfectly, even in the adult : hence a fertile source, no 

 doubt, of error. For my own part, I believe the system of coloration 

 of the Salmonidae to obey fixed laws, and to be constant and regular, 

 the class being free from any influence which obviously affects the 

 colour, namely, domesticity. 



Section III. — The Proportions of the Salmonidas, as compared with 

 each other and with the Generic Animal. 



The adult well-formed individual of every species of animal has its 

 due proportions characteristic of its nature; with these its movements 

 or style of motion is connected. By these movements it may be 

 recognized by man and other animals when alive ; when dead or at rest, 

 its proportions indicate its nature. The distinction extends not unfre- 

 quently to sex, in which case the peculiar proportions are said to be 

 sexual. In wild animals these proportions are remarkably constant, 

 suffering indeed but little change or variety, as in the case of the co- 

 loration ; it is amongst those animals whose nature permits of domes- 

 tication that we find colour and proportions to undergo changes, 

 which, though limited, are yet remarkable in extent. 



Nearly all my early observations, made many years ago, w T ere insti- 

 tuted on the adult specimen, or rather (for it is not so easy to deter- 

 mine what is adult) on individuals which had attained a considerable 

 size. Salmon, sea trout and trout, of various species, were carefully 

 measured and compared with each other, and the results compared, 

 the object being to discover the law of subfamily and species. To 

 these I now add the proportions of the generic animal, as compared 

 with the adult, from which I think it will be manifest that the generic 

 animal has proportions peculiar to itself, yet including the specific, 

 that is, it presents a type out of which all the others may easily be 

 constructed by the comparative enlargement of some measurements 

 and the comparative restriction of others. The tables of these mea- 

 surements I give at the conclusion of this section ; the general results 

 may be stated here. 



