PRINCIPLES OF CLASSIFICATIOX. 3 



fins, as consisting of bony spines or soft-jointed branched rays, is a difference 

 sufficient to constitute an ordinal group. The obliteration or persistence of the 

 pneumatic duct to the air-bladder is an important character in classification. 

 The persistence of a cartilag-inous condition of the skeleton, with bony plates on 

 the body, similarly characterises the Sturgeons. And the lower condition in 

 the Lampreys, in which the vertebral column remains cartilaginous, and without 

 sub-divisions into distinct vertebral parts, may be taken to represent another 

 grade of structure. But these Oi'ders also vary in minor details of structure, 

 so as to present a multitude of types, which are known as genera. And 

 since every fish must be described in its genus and species, it becomes 

 necessary to dwell upon the nature of the characters on which classification 

 is made. For there are rarely broad lines of demarcation in Nature ; and if 

 such exist, we may reasonably believe that intermediate types have formerly 

 lived upon the earth, and become fossilised during its g-eological history. 

 Hence, whenever an assemblage of species can be united by common cha- 

 racters, and similarly separated from other fishes, they constitute a genus, 

 which may comprise one species or a hundred. If the characters which dis- 

 tinguish genera ai'e compared together in the following descriptions, they 

 will be seen to be of a varying nature. Thus the Perch genus has about 

 thirteen or fourteen spines in the first dorsal fin, and has the head free from 

 scales on the upper part. These characters separate it from some South 

 American Perches, which form the genus Percichthys, in which the spines 

 in the first dorsal fin are reduced to nine or ten, and the head is covered 

 with scales. When we turn to the Marine Perches, which form the genus 

 Labrax, no difference in the number of spines is found, but the tongue, 

 which was smooth in Perca and Percichthys, has become toothed, like the 

 palatine bones. 



The species of a genus are distinguished by characters of another kind. 

 It is often difficult to say where a variety ends and a species begins. There 

 are local races of many fishes which, under the changed conditions of physical 

 geography which from time to time affect the distribution of life on the earth, 

 have become isolated from the rest of the race, so as to live on table-lands or 

 low plains, in cold mountain lakes or in shallow swamps, in sluggish waters 

 or rapid torrents, and thus, differently circumstanced, have developed into 

 varieties distinguished by size, form, colour, and certain internal and external 

 differences in the organs and proportions of the body. A genus like Salmo 

 or Coregonus shows a remarkable vitality under varied conditions in the 

 evolution of specific characters ; but these variations are, for the most part, 

 so small, and so slightly differentiate from each other the forms they cha- 

 racterise, that it is open to us either to recognise that there is no clear 



