THE PRIMARY DIVISIONS OF GROUPS. oa7 
latter is so much more diversified in its contents (for 
reasons hereafter to be stated) than the other two, that 
many naturalists reckon five greups in all ;_ the number 
Jive being made out by dividing the aberrant group 
into three, instead of considering it as only one. We 
have seen, however, that the first test of a natural 
group is its circular chain of affinities. If, therefore, 
the three divisions of Mr. MacLeay’s aberrant group can 
be shown to form a circle of their own, independent of 
the other two, then we must reckon them as one only, 
thus making the primary divisions of every circle three. 
= We, consequently, have, in 
2 prs 1 every natural group, three 
Subtypical Typical primary circles, one of which 
(the aberrant) is divided into 
_/ three secondary circles. A 
good idea of this disposition 
may be gained by the an- 
nexed diagram. If, on the 
other hand, we adopt Mr. 
MacLeay’s theory, that every 
eroup is first divided into five circles (the three aberrant 
not being united into one), then we may express them 
in this manner : — 
Z- 
3 Aberrant 
(279.) Let us illustrate this first division of a/natural 
group by an instance drawn from the animal kingdouw. 
Q 2 
