PARTS OF ANIMALS, I. ii.-iii. 



(a) Some '^ groups will get only one differ ejilia,^ the 

 rest of the terms being superfluous extras,'' as in the 

 example : footed, two-footed, cloven-footed ^ — since 

 this last one is the only independently valid differentia. 

 Otherwise the same thing * must of necessity be 

 repeated many times over. 



(b) Again, it is a mistake to break up a group, as 

 for instance the group Birds, by putting some birds 

 in one division and some in another, as has been done 

 in the divisions made by certain wTiters : in these 

 some birds are put in with the water-creatures, and 

 others in another class. (These tw^o groups, each 

 possessing its own set of characteristics, happen to 

 have regular names — Birds, Fishes — but there are 

 other groups which have not, e.g. the " blooded " 

 and " bloodless " groups : there is no one regular 

 name for either of these.) If, then, it is a mistake 

 to break up any group of kindred creatures, the 

 method of division into two will be pointless, because 

 those who so divide are compelled to separate them 

 and break them up, some of the many-footed animals 

 being among the land-animals and others among the 

 water-animals. 



III. (c) Again, this method of twofold division 

 makes it necessary to introduce privative terms, and 

 those who adopt it actually do this. But a privation, as 



" i.e. all terms except the final one can be dispensed with, 

 because none of them constitutes an independent (/cupta) 

 differentia ; one line of division yields one valid differentia 

 and no more (c/. 644 a 2-10). 



<* Cf. 644 a 5 and Met. 1038 a 32. 



• In this case, " -footed " {cf. Met. 1038 a 19 if.).^^ But 

 Aristotle does not explain how Si'tt-ouv is " superfluous." 



8L 



