220 STUDY OF NATURAL HISTORY. 
that these are the very groups, above all others, 
wherein the principles of enquiry we are now re- 
commending can be most successfully and most 
easily pursued; and for this purpose, as such, they 
should consequently be selected. When, therefore, 
we can draw any general deductions from the 
contents of several such groups, whether as re- 
gards the mode of variation in their subordinate 
forms, the characters of the forms themselves, or 
their definite number, we may rest assured of 
having committed no great error in their natural 
arrangement ; and may safely assume the inductions 
thus obtained, as instruments to facilitate our further 
progress. Having now stated those primary con- 
siderations which appear necessary to determine, on 
sound principles of inductive science, the two lowest 
groups of nature, namely, sub-genera and genera, 
we may proceed a step further, and enquire into 
higher combinations. 
(155.) The groups next in rank to genera, modern 
naturalists agree in calling swb-families. Thename, 
however, has nothing to do with our present object, 
further than that it is necessary to give some 
designation to groups which are next in rank, 
or in comprehensiveness, to those last discussed. 
The determination, therefore, of a group of this 
sort,—no matter by what name we choose to call 
it, — must be regulated by the law we set out with 
assuming ; that is, by the union of a certain number 
of genera, which, thus combined, produce a circular 
series. Here, again, the question of mumbers 
arises. Now, bearing in mind, that the greater the 
degree of harmony and unity we can produce in our 
