362 ON THE ORDERS 



The distinction between the Mandibulala and Ilaus' 

 tellata is, I repeat, clearly natural, because it directly af- 

 fects those habits of the animals upon which their exist- 

 ence immediately depends. It may, indeed, be sup- 

 posed at first to be in contradiction to the acute observa- 

 tions of M. Savigny; yet, so far is this from being true, 

 that, in his first Mtmoire, he himself recommends the di- 

 vision of Insects into Uroyeurs and Suceurs, His work 

 in fact affords no more than a happy demonstration of that 

 analogy which neighbouring groups generally, if not al-r 

 ways, bear to each other. 



When in the distribution of the animal kingdom into 

 classes my efforts to detect any marked relation of analogy 

 were foiled, I contented myself with pursuing those con-, 

 siderations of affinity Avhich served to reconduct me to the 

 point from which I had started. I never can persuade 

 myself, however, that this scarcity of analogical relations 

 between the classes of Zoology has resulted from any other 

 cause than my own inability to detect them. To me their 

 rarity argues nothing more than how much remains to be 

 done, before the order of affinities I have pointed out can 

 be considered as perfectly correct. It is impossible to 

 draw any other inference than this, since in every inves- 

 tigation I have made by analysis, and of the accuracy of 

 which therefore I am most sure, relations of analogy have 

 not failed to be conspicuous. Nay more, when these 

 could be detected, they have always supplied the most 

 convenient testimony of the affinities with which they 

 were connected being real, and thus have given me some 

 •reason to suspect that no affinity can be true which is not 

 connected with a relation of analogy. 



Suppose the existence of two parallel series of animals^ 



