PARTS OF ANIMALS, I. iii. 



differentia wiW belong to ϊλ\ό species. And if so, 

 it is clear that a privative cannot be a valid 

 differentia.'^ 



(d) Now assuming that each species is indivisible : if 

 each differentia also is indivisible, and none is common 

 to more species than one, then the number of differ- 

 entiae \\\\\ be equal to the number of species. (Suppos- 

 ing it Avere possible to have a differentia which though 

 indivisible was common ; clearly, in that case, animals 

 which differed in species ΛνοηΜ be in the same di\'ision 

 in virtue of that common differentia. Therefore, if the 

 differentiae under which the indivisible and ultimate 

 species fall are to be proper and private to each 

 one, it is necessary that no differentia be common ; 

 otherwise, species Λvhich are actually different will 

 come under one and the selfsame differentia.) And 

 we may not place one and the same indivisible 

 species under two or three of the lines of diff"erentia- 

 tion given by the divisions ; nor may we include 

 diff"erent species under one and the same line of 

 differentiation. Yet each species must be placed 

 under the lines of diff'erentiation available. It is 

 evident from this that it is impossible to arrive at the 

 indivisible species either of animals or of any other 

 group by the method of twofold division as these 

 people practise it, for even on their sho\nng the 

 number of ultimate differentiae must of necessity be 

 equal to the total number of indivisible species of 

 animals. Thus, suppose ΛΛ'e have some particular 

 group of creatures whose prime differentiae are 

 " pale " and " not pale " ; by that method these 

 two ΛΛ -ill each give t\vo other differentiae, and so 

 forth, until in the end the indivisible differentiae are 

 reached : these last ones Λνϋΐ be either four in 



85 



