A STUDY OF PLATONIC TERMINOLOGY. 483 the study of geometry is described as having an elevating (o/cbv ty-vxfis) and purifying force ; as capable of rekindling that organ of the soul, which, stained and blinded by all sorts of occupations, is more precious and worthy of being cared for and protected than ten thousand eyes, (opyavov n eKKaOaiperai re KCU ava^wrrvpelrai, d-TroXXv/xez/oz/ real <f)ovfievov UTTO TWV a'XXcoz/ eTTirrjSevfjidrwv, tcpeirrov ov Rep., 527.) No less intimately connected with the subject of the mean- ing of general terms is the use which Plato makes of the word 61877 in the phrase : icar eiSrj Biaipelv, and other analog- ous phrases which appear chiefly in the Dialogues which are considered to be the later ones. It is not always easy to decide when, in such phrases and analogous ones, the word ei8r) is to be understood in the usual sense of "class," like 761/09, /ie/>o5, and when, on the other hand, in the sense of character, as it seems, for instance, that it must be understood in the Cratylus (424 G). Such an ambiguity is rendered still more serious by the fact that the preposition Kara may be interpreted at the same time as equivalent to the Latin in (e.g., in the phrase : dividere in paries) or to the Latin secundum, in the sense of " according to". In the Phadrus, the proceeding of one who divides (opifr- nevos) is represented as correlative with, and complementary to, the other which consists in uniting under one and the same concept facts or objects apparently different, and both are described as contributing equally to constitute the true art of discussing (8iaKTiKrj) (Phadr., 265 D). The same distinction is also referred to in the passage in Rep., vi. in which those who are capable of dividing tear' eiSr) are contrasted with those who, being incapable of emancipat- ing themselves from verbal suggestions (Kara TO ovo/xa Smj/coz/re?), are more disposed to litigate than to argue (epiSi ov 8iaKTa) xpcopevoi), Rep., vi., 454 A, B. Similarly, in the Politicus, the method of dividing by species (rear' e'lBrj Siaipeiv) is described as the best calculated to render people capable of inquiry and discussion. Such a method is there (263) described still more precisely by the citation of examples of cases which do not conform to it, among the rest that of the division of men into Greeks and Barbarians, to which Plato objects as being based on igno- rance of the fact that the nations included under the single denomination (fiia /eX^'crei) of Barbarians differ less from one another in language and race than the Greeks from each of
