222 ON SYSTEMATIC ZOOLOGY. 



entomologists) so confident in their conviction of the truth 

 of circular affinities, and yet so unconsciously regardless 

 of those principles which must establish this theory in the 

 minds of acute reasoners. The proof of a circle of affinity, 

 as laid down hy its discoverer, rests, in the first instance, 

 upon its complete analysis ; and, secondly, in its con- 

 tents intimately and regularly corresponding in analogy 

 with the contents of a neighbouring circle. There may 

 be seven, ten, twenty, or fifty natural orders, for what 

 we know, and they may possibly be circular, and there- 

 fore natural ; but with the above conditions of a circle 

 before us, we must ever withhold our belief in such di- 

 visions, until they rest upon a more solid foundation 

 than arbitrary opinion. Although somewhat backward in 

 viewing zoology as but a branch of physical science, 

 we are happily so far advanced in its philosophy, as to 

 consider facts more weighty than assertions, and cautious 

 induction more valuable than hypothesis. If, then, the 

 number seven is to be substituted for that of five, let it 

 be made out analytically and analogically in any two 

 groups out of the many which have been assumed as 

 " natural," and we will venture to predict that the 

 learned author of the Horce Entomologies would be one 

 of the first who would proclaim the truth of the demon- 

 stration. We offer these observations generally, and as 

 equally applicable to any determinate number which 

 may be thought the true one of nature. 



(272.) It has been said, in reference to the quinary 

 theory, that in most cases the number of divisions in a 

 natural group is five, but that in many instances there 

 appear to be as many as seven. Now, this may be very 

 true in one sense, and very erroneous in another. 1. If 

 a circular group is to be divided merely according to the 

 fancy of the divider, or according to those marks or 

 characters which he thinks most important, without re- 

 ference to any other considerations, it is obvious that 

 scarcely two persons will agree in the number they 

 eventually fix upon : one may make three, another five, 

 and another seven. But then comes the first test of accu- 



