16 



we are tolerably well acquainted are divisible into seven ; we 

 shall never find the number greater, and when less, we shall 

 invariably perceive that the deficiency exists in groups of 

 which our knowledge is particularly limited, for the perfection of 

 a septenary distribution of any particular group will depend 

 entirely on our acquaintance with that group : thus the groups 

 at present kno'wn by the names Mammalia, Aves, and Insecta, 

 resolve themselves instantly into sevens. No ingenuity can frame 

 eight good groups of either, and no scheme, however plausible, 

 can reduce the number to sixes or fives. An attempt to reduce 

 birds into five groups has been made in this country; I cannot 

 do better than refer the reader to it as a triumphant confirmation 

 of the predominance of the number seven.* The great Linnseus 

 assigned to Mammalia seven orders, to Aves six, and to Insecta 

 seven, in a system which, though capable of improvement in 

 many of the orders, evidently points to the truth, and considering 

 his limited means of reference, compared with what the naturalist 

 now possesses, was a remarkable and magnificent monument of 

 human talent.-j- 



To go back two thousand years before the birth of Linnaeus, 

 may be thought rather an unlikely mode of obtaining proof of the 

 value of a modern theory in natural history; yet at that time we 

 find a system of insects;]; divided so accurately into seven groups, 

 that every attempt to improve it has, as far as regards these great 

 groups, proved an utter fallacy. Now this array of names, 

 Aristotle, Linnaeus, Cuvier and Kirby, thus corroborating Holy 

 Writ, even in direct opposition to our own observations, is en- 

 titled to a good degree of confidence ; but how much more cheer- 

 ftilly is that confidence given when our own unbiassed judgment 

 must thoroughly coincide ! 



Presuming, therefore, that a septenary and circular arrange- 



* By Mr. Vigors. Linnaean Transactions. 



f It will be observed that in the Mollusca, Radiata, and Acrita of Mac- 

 Leay, all attempts to employ a particvilar number in grouping will be found 

 futile, a circumstance obviously attributable to our ignorance ; and the only 

 conclusion to be drawn from it is this : that, as these tribes can never be 

 rendered available tor any numerical distribution, so they can never be fairly 

 and satisfactorily adduced in refutation of such a distribution. 



1 That of Aristotle. 



