PARTS OF ANIMALS, I. iii. 



" winged " and " wingless " — and by " wild " and 

 " tame," " is not permissible, for this similarly would 

 appear to divide up species that are the same, since 

 practically all the tame animals are also found as 

 ^vild ones : e.g. Man, the horse, the ox, the dog (in 

 India ^), swine, the goat, the sheep ; and if, in each 

 of these groups, the wild and the tame bear the same 

 name, as they do, there is no division between them, 

 while if each group is specifically a unit, then it 

 follows that " wild " and " tame " cannot make a 

 valid differentiation.'' 



And generally, the same thing inevitably happens 

 whatever one single line of differentiation is taken for 

 the division. The proper course is to endeavour to 

 take the animals according to their groups, fol- 

 lowing the lead of the bulk of mankind, who have 

 marked off the group of Birds and the group of Fishes. 

 Each of these groups is marked offhyjuani/ cUjferentiae , 

 not by means of dichotomy. By dichotomy (a) either 

 these groups cannot be arrived at at all (because the 

 same group falls under several divisions and contrary 

 groups under the same division) or else there will be 

 one differentia only, and this either singly or in 

 combination ^ will constitute the ultimate species.* 

 But (6) if they do not take the differeriiia of the differ- 

 entia, they are forced to follow the example of 

 those people who try to give unity to their prose by a 

 free use of conjunctions : there is as little con- 

 tinuity about their division. Here is an example 

 to show what happens. Suppose they make the 

 division into " wingless " and " winged," and then 

 divide " winged " into " tame " and " wild " or into 



' And this will never completely represent any actual 

 group or species. See below, 644( a 6 ff . 



89' 



