240 STUDY. OF NATURAL HISTORY. 
the group before him is distinguished from all others ; 
and if he can do this effectually, it matters not by 
what means the object is accomplished. He is not, 
however, to expect that he can so far isolate a 
natural group, as that there shall be no exceptions 
to the characters he assigns it; or that each of the 
individuals composing it shall possess those charac- 
ters in the same degree. This would imply not only 
the existence of absolute divisions in nature,— 
which all experience is opposed to, —but would be 
directly at variance with what has been just said on 
the chain of continuity or gradation; for where 
there is gradation in structure, there must be gra- 
dation in character. It will be sufficient, if the 
greater portion accord with his definition, and if the 
others present a gradual diminution of the same set 
of characters, fading and blending into others be- 
longing to adjoining groups. In searching, there- 
fore, for such characters, we must take into the 
account every circumstance that is known regarding 
the economy and the structure of the objects them- 
selves; and from all these make a selection of such 
as are most constant, universal, and obvious. It 
will almost always be found, that a peculiarity of 
internal organisation is accompanied by a corre- 
sponding difference in external structure, and that 
both these are adapted for that particular mode of 
life which the afiimal pursues. As there is a con- 
stant harmony between the conformation of an 
animal and its peculiar economy, we should study 
the former with a constant reference to the latter, 
but yet draw our characters from the jirst rather 
than from the second ; because form can be always 
