. 
110 FAMILY CHARACTERISTICS. 
mon parlance, the children or the descendants 
of known parents; they understand by this name 
natural groups including different kinds of ani- 
mals, having no genetic relations so far as we 
know, but agreeing with one another closely 
enough to leave the impression of a more or 
less remote common parentage. The difficulty 
here consists in determining the natural limits 
of such groups, and in tracing the characteristic 
features by which they may be defined; for in- 
dividual investigators differ greatly as to the de- 
gree of resemblance existing between the mem- 
bers of many Families, and there is no kind of 
group which presents greater diversity of circum- 
scription in the classifications of animals pro- 
posed by different naturalists than these so-called 
Families. 
It should be remembered, however, that, unless 
a sound criterion be applied to the limitation of 
Families, they, like all other groups introduced 
into zodlogical systems, must forever remain arbi- 
trary divisions, as they have been hitherto. A 
retrospective glance at the progress of our sci- 
ence during the past century, in this connection, _ 
may perhaps help us to solve the difficulty. 
Linneus, in his System of Nature, does not ad- 
mit Families; he has only four kinds of groups, 
— Classes, Orders, Genera, and Species. It was, 
as I have stated in a previous chapter, among 
