ON THE CETONIIDiE OF SOUTH AFRICA. 



7 



which serves in some degree to insulate it from other groupes. But it is evident that we 

 must previously have arrived at the knowledge of the groupe ; and this is effected by a close 

 watching of the variation of affinity, and by considering the groupe to be complete only when 

 the series of natural objects returns into itself. It is more particularly when we have satisfac- 

 torily applied this test of the groupe's completeness, that we ought to look out for its natural 

 character. Still, after all, we may be wrong ; we may sometimes imagine a groupe to return 

 into itself, when it does not ; or we may consider what in reality are two groupes, to form 

 only one. The best way I know of correcting such evils, is to recollect that the natural 

 system ought to express all the relations which exist between the various objects of our 

 study, and that if any of these are left unrepresented in our arrangement, there must be 

 some latent error, in the formation of our groupes. I repeat, that the process is one of 

 tatonnement. The marsupial animals have clearly some kind of relation to Primates, to 

 Insectivorous Feres, to Edentata, and also to Glires. Now, if a naturalist cannot express all 

 these relations, and at the same time all the various relations which exist among the other 

 groupes of Mammalia, he ought to be convinced that his arrangement is wrong. It is of 

 no use to talk about the importance of this organ, or of that other, in the structure of 

 these animals. If even we were right in any such comparative estimate of the importance of 

 organs in general economy, we ought not to forget that the true question under consideration 

 is, the natural arrangement of the animals themselves ; and that this is to be attained only 

 l^y the expression of every affinity, and every analogy that can be detected. The unnatural 

 groupe Testacea, was founded on the important character of the presence of a shell ; and we 

 shall see among the CetoniidoB, that genera were erroneously separated from this family 

 because they wanted its leading characters, which consist in the membranaceous texture of 

 certain parts of the mouth. 



I have before alluded to the fact, that when a series of affinity is followed, it is observed 

 to return into itself, and thus to form a closed groupe. Now, if we start from the principle 

 that when a few species first agree in some particular character, they combine into a series 

 that will return into itself, we shall probably imagine every such series, so forming a circle 

 in practice, to be in theory the first natural assemblage of species. Yet this will be an 

 incorrect mode of viewing the matter ; for owing to the rarity of its species, the first known 

 circular grouping of the species of Cryptodinus, for instance, is into sub-genera ; whereas 

 the first known circular grouping of the species of Cetoniidce is into certain sub-sections. 

 But, if the above hypothesis were true, the sub-genera of one genus would be equivalent 

 in value to only sub-sections of the other. I conceive, therefore, that in order to attain 

 accuracy in the distinction of groupes, we must proceed in a totally different manner to 

 investigate the subject. In the first place, we must commence from some given point in 

 which all naturalists agree. )Such a point undoubtedly is the animal kingdom. Every one 

 acknowledges this groupe. Now I observe the animal kingdom to resolve itself into five 

 natural groupes, forming a circle. I observe again, each of these circular groupes to resolve 

 itself into five others, and so down, until I arrive at species. Now it is clear that I may 

 term the animal kingdom a genus, or I may call the groupe Vertehrata a sub-genus, and 

 Mammalia a kingdom ; the word given as a name to the class of groupe, being adopted 

 for convenience, and being merely conventional. My aim is not to dispute about the 

 propriety of the names given to these various groupes, but merely to express the proper 

 subordination of one groupe to another, even supposing that they had no names. If, therefore, 



