340 PRACTICAL AND SCIENTIFIC ZOOLOGY. 
not enough for demonstration: its circularity must be 
made out, and its contents submitted to the very same 
tests that have been applied to the entire family. If 
the MyvornEertin@, then, form a truly natural, and, there- 
fore, a circular group, its component parts will re- 
present all the divisions of its own family; and we 
shall find the subordinate variations analogically repre- 
senting Merula, Crateropus, Oriolus, and Brachypus. So 
far is this beautiful uniformity of consistent and definite 
variation from being chimerical, that we have, in another 
work *, selected the MyorHrerin# for this especial pur- 
pose, and have demonstrated that they form a natural 
group, capable of the same degree of verification as we 
have been here insisting upon. To that work we must 
refer the reader, who desires to see in what manner 
relations of analogy, in so small a circle, can be made 
out. To pursue the subject farther, on the present 
occasion, will be needless. We have confined our illus- 
trations to ornithology ; but it must be remembered 
that the same laws are applicable to every group in the 
animal kingdom. 
(419.) Natural groups are thus to be detected by 
three different tests: 1. By their simple series of circular 
affinity ; 2. By the theory of analogy ; and, 3. By the 
theory of variation. We draw the first of these proofs 
from affinity ; but the two latter entirely depend upon 
analogy. No group which will not bear these tests can 
be natural ; whereas, if it will stand such an ordeal, it 
has passed all the trials necessary to establish its cor- 
rectness. 
(420.) We trust that the young naturalist will now 
see the truth of the observation long ago made by a 
well-known naturalist, that nothing can be easier than 
to make circles, provided it is not thought necessary to 
prove them: in other words, to give them more value 
than they possess, either from mere assertion, or from 
wearing an appearance, at first sight, of being really 
what they are affirmed. We trust, moreover, that he 
®* Northern Zoology, vol. ii. p. 168. 
