ON CLASSIFICATION. 
15 
Thus it sometimes happens that one single species 
is so far removed from its nearest allies that natural- 
ists have to place it alone in a distinct family from all 
other species. Consequently there seem to be many 
blanks left in the series, the intermediate forms may 
be unknown to us, or else the existing form may be 
the only one that survives out of a large group, the 
remainder of which have become extinct. 
W e have seen that when two species that are really 
distinct are placed before us, and we are able to com- 
pare them critically one with the other, there is 
seldom any difficulty in discriminating between them. 
There is sometimes difficulty in knowing what group 
a fish belongs to ; this is generally in consequence of 
the imperfection of our classification, and becomes 
less in proportion as our genera form natural groups. 
There is, however, often a difficulty with fishes that 
in reality are of the same species, either they may 
have been described by different authors under dis- 
tinct names, and each author may have laid stress on 
different points : in this case by comparing the de- 
scriptions alone, it is impossible to be certain if the 
fishes are the same or different : or else the species 
may be variable in itself, which often happens when 
it has a large geographical range, specimens from one 
locality differing from those of another, or the same 
individual appearing different at different times, 
according to the food it is living on, the kind of water 
it is inhabiting, or the season of the year. 
Herein lies the great value of a large aud carefully 
