226 FIRST PRINCIPLES OF NATURAL CLASSIFICATION. 



But when a part of the series is perfect, and the other 

 part presents the idea of a chain where several of the 

 links are wanting, then the group !s called imperfect. Now 

 this imperfection arises from two causes : either these 

 absent links have not yet been discovered, or they have 

 been destroyed in the revolutions which have agitated 

 our globe. This is the first great law of the natural 

 system ; it is that upon which all others repose, and which 

 has been already demonstrated in almost every de- 

 partment of zoology, but more especially in ornitho- 

 logy. If the reader wishes to see this theory made 

 good in the animal world, we must refer him to the 

 Horce Entomologicce and to the Northern Zoology. We 

 may refer him, in the last-mentioned work, to the genus 

 Picus, and to the sub-family Piciance, as examples of 

 perfect groups ; and to the family Picidce (of the same 

 volume) for one that is imperfect. The circle of the 

 animal kingdom (p. 203.) is also a familiar illustration 

 upon a large scale. Commencing with the Polypes, we 

 pass on to the Mollusca; from these we are led to verte- 

 brated animals ; thence to insects and radiated animals ; 

 and, finally, arrive once more among the polypes. Our 

 course has thus been circular ; the two ends of the series 

 meet ; and we have, theoretically, a natural group. 



(277-) II. We now pass to our second proposition ; 

 viz. The primary circular divisions of every such group 

 are three actually, and five apparently. 



(278.) As it is manifest that every group, according to 

 its magnitude, will exhibit more or less variety in its con- 

 tents, the first question which suggests itself is. Are 

 these variations regulated by any definite number ? And 

 is that number so constant, in all such groups as have 

 been properly investigated, as to sanction the belief that 

 it is universal ? The answer is in the affirmative. Every 

 group, whatever may be its rank or value, (that is, its 

 size or its denomination,) contains, according to our 

 theory, three other primary groups, whose affinities are 

 also circular. One of these is called the typical, the other 

 the suh.typical, and the third the aberrant group. This 



