VERIFICATION OF GROUPS. 287 



mary types explained in the preceding pages. If, in 

 short, a zoological group be natural, it will not only 

 bear a comparison with every other in the same class, 

 but will give and receive a flood of light to and from 

 all with which it is compared. 



(348.) It follows, from the preceding remarks, that 

 the verifications of a natural group are three : 1 . The 

 circular series of its contents ; 2. The parallel rela- 

 tions of its parts to other groups ; and, 3. The sym- 

 bolical representation of the primary types of nature. 

 On some of these points we have expatiated * ; but this 

 is the proper place for treating the subject in a more 

 definite, clear, and connected manner. 



(349.) There are no absolute rules, of universal ap- 

 plication, independent of analysis, which can be laid 

 down for the discovery of a zoological circle. We 

 must begin, in fact, by arranging the objects with the 

 nicest attention to their apparent affinities, and then 

 testing the result. If these affinities are real, and the 

 group is natural, there will be an evident tendency to a 

 circle ; and this tendency will be more or less strong, in 

 proportion to the number of objects which enter into the 

 series. When we consider, however, 'that the relations of 

 objects are complicated, and by no means confined to those 

 which precede, or those which follow them, in the series 

 of affinity, it is obvious that false circles may be made ; 

 and that their fallacy can only be discovered by further 

 tests. Before the naturalist proceeds to these, it is 

 absolutely necessary that he endeavours to make out 

 the two immediate circles which pass into that with 

 which he has first begun. If, for instance, he was 

 investigating the genus Picus Sw., as now constituted, 

 after simply tracing the circular affinities of this group, 

 he should proceed to investigate the two others which 

 more immediately join it ; namely, Chrysoptilus Sw. 

 and Melanerpes. Unless this were done, he will have 

 no definite ideas on the probable demarcation of his 

 first circle, at those points where it touches, and passes 



* Preliminary Discourse on Nat. Hist. 



