[Footnote 38: That "the attribute is in its subject," is explained by Aristotle only by saying That it is _in_ its subject, not as a part in the whole, yet as that which cannot exist apart from its subject (Categor. 1, a. 30-3, a. 30). Compare Hobbes, Comput. or Logic. iii. 3, viii. 3. Respecting the number of different modes [Greek: tou= e)/n tini ei)=nai], see Aristot. Physic. iii. p. 210, a. 18 seq., with the Scholia, p. 373 Brandis, and p. 446, 10 Brand. The commentators made out, variously, nine, eleven, sixteen distinct [Greek: tro/pous tou= e)/n tini ei)=nai]. In the language of Aristotle, _genus_, _species_, [Greek: ei)=dos], and even _differentia_ are not [Greek: e)n u(pokeime/nô|], but are predicated [Greek: kath' u(pokeime/nou] (see Cat. p. 3, a. 20). The _proprium_ and _accidens_ alone are [Greek: e)n u(pokeime/nô|]. Here is a difference between his language and that of Plato, according to whom [Greek: to\ ei)=dos] is [Greek: e)n e(ka/stô| tô=n pollô=n] (Parmenid. 131 A). But we remark in that same dialogue, that when Parmenides questions Sokrates whether he recognises [Greek: ei)/dê au)ta\ kath' au)ta/] he first asks whether Sokrates admits [Greek: dikai/ou ti ei)=dos au)to\ kath' au(to/, kai\ kalou=, kai\ a)gathou=, kai\ pa/ntôn tô=n toiou/tôn]. Sokrates answers without hesitation, _Yes_. Then Parmenides proceeds to ask, Do you recognise an [Greek: ei)=dos] of man, separate and apart from all of us individual men?--or an [Greek: ei)=dos] of fire, water, and such like? Here Sokrates hesitates: he will neither admit nor deny it (130 D). The first list, which Sokrates at once accepts, is of what Aristotle would call _accidents_: the second, which Sokrates doubts about, is of what Aristotle would call _second substances_. We thus see that the conception of a self-existent [Greek: ei)=dos] realised itself most easily and distinctly to the mind of Plato in the case of _accidents_. He would, therefore, naturally conceive [Greek: ta\ ei)/dê] as being [Greek: e)n u(pokeime/nô|], agreeing substantially, though not in terms, with Aristotle. It is in the case of accidents or attributes that abstract names are most usually invented; and it is the abstract name, or the neuter adjective used as its equivalent, which suggests the belief in an [Greek: ei)=dos].] [Footnote 39: Aristotel. Metaphys. [Greek: D]. 1015-1016, I. 1052, a. 29 seq. [Greek: ta\ me\n dê\ ou(/tôs e(\n ê)\ suneche\s ê)\ o(/lon; ta\ de\ ô(=n a)\n o( lo/gos ei(=s ê)=|; toiau=ta de\ ô(=n ê( no/êsis mi/a], &c.