1 I am deeply indebted to my teachers Professor M. Heidegger of Freiburg, from whom I obtained an insight into philosophical problems in general and from whom I learned to understand the ontological character of Greek philosophy, and Professor P. Von der Muhll of Basel, who taught me the method of dealing with the Presocratics and for whose seminar the outline of this paper was originally written. I wish also to thank Mr. W. K. C. Guthrie for kindly assisting me to publish it in English.

^{page 47 note 2} Diels, , Vorsokratiker 68B 6–11Google Scholar.

^{page 47 note 3} Untersuchungen zu Cicero's philosophischen Schriften I 113.

^{page 47 note 4} Another representative of the same opinion, Brieger (‘Die angebliche Leugnung der Sinneswahrheit bei Demokrit’, Hermes 37, 1902, p. 56 sqq.), is hardly worth mentioning, as his discussion of the problem is not of a high standard, His arguments against Natorp are unsound.

^{page 47 note 5} Farschungen zur Geschidite des Erhenntnis-problems im Altertum, 1884, p. 164 sqq., and the supplementary article in Archiv fiir Geschichte der Philosophic I, 1888, p. 348 sqqGoogle Scholar.

^{page 47 note 6} Forschmtgen, p. 171: ‘Das Wahre liegt keineswegs in den Erscheinungen der Sinne, sondern in Begriffen des Verstandes’.

^{page 47 note 7} ‘Sie verliessen nicht den rationalen Boden der Begriindung gegenständlicher Erkenntnis, welchen die Eleaten zuerst gelegt hatten’.

^{page 47 note 8} P. 177: ‘in welcher Art von Erkenntnis’.

^{page 48 note 1} Die Philosophic der Griechen, 6th ed. I, pp. 1132–34.

^{page 48 note 2} Langerbeck, Δόξις ⋯πιρυσμίη, ‘Studien zu Demokrits Ethik und Erkenntnislehre,’ Neue philologischc Untersuchungen, 1935, p. 90. Dr. Langerbeck rejects the account, taking up the argument that Aristotle is arguing purely logically and thus coming to a wrong conclusion. So he, too, believes that it has no historical value. Nevertheless he thinks that the concept of ⋯λήθεια requires a new interpretation, because to him the context in Met. F 5 seems, even within itself, to lack consistency. That Langerbeck misunderstood the train of thought in this whole passage, and how it should be interpreted, has been sufficiently shown by Professor Kapp. Compare the following footnote.

^{page 48 note 3} Cf. Kapp's, review of Langerbeck's book, Gnomon, 1936Google Scholar.

^{page 48 note 4} We have to understand the ⋯τεῇ δ⋯ ᾰτομα καί κενόν in B9 as the object of the ϒνώμη ϒνησίη in B II.

^{page 48 note 5} B II: κα⋯ σκοτίης μ⋯ν τάδε σύμπαντα, ǒψις, ⋯κοή, ⋯δμή, γε⋯σις, ψα⋯ϭις.

^{page 49 note 1} It is true that in Met. Г5 Aristotle says in a sentence immediately preceding ours, δι⋯ Δημόκαιτός γέ ϕηϭιν ňτοι οὐδ⋯ν εΙναι ⋯ληθ⋯ς ἣ ἣ γ'ᾰδηλον. But this, as we infer from the context, should be expounded as follows. In view of the fact that in the domain of sense-perception one truth will stand against its opposite, and that there is no criterion by which we can give preference to either, there is no other solution than to assume that either nothing is true or truth is not obvious to us, but hidden in the depth. — There can be no doubt that only the second alternative represents Democritus' conviction. Aristotle did not mean to indicate in this sentence that Demoritus was a sceptic, but repeated one of the Democritean arguments against truth to be found in the αίσθητ⋯ (in a narrow sense). In this matter, Democritus was in opposition to Protagoras, for whom all that appears was true. So far all is in agreement with the Sextus fragments. But in the following sentence, the one with which we have to deal, Aristotle conversely places Democritus side by side with Protagoras and with those who deny the ὠρισμένον τι (1008a 34), i.e. those who maintain őτι οὐκ ⋯νάγκη ἢ ϕάναι ἢ ⋯ποΦάναι (1008a 3) and τ⋯ αὐτ⋯ εἶναί τε κα⋯ μ⋯ εῖναι (1009a 12). So he counts him among those who, ‘on the whole’ (őλως). saw truth in the αίσθητά. But here αίσθητά is meant in a new and wider sense, which must be interpreted in our text. The mere fact that Democritus is now placed in the ranks of his opponents indicates that Aristotle turns to a more fundamental view. The conjecture of Kranz, who substitutes őμως for ǒλως (in A 112, D.V. 5th ed., 1935, adopted by Capelle in his German translation of the Presocratics), errs by overlooking the fact that the train of thought here reaches far backwards.We must retain the ὃλως.

^{page 50 note 1} =D.V. Leukippus A 7.

^{page 50 note 2} De gen. et corr. A 8, 325a 13=D.V. Parmenides A 25.

^{page 50 note 3} D.V. 68B 9.

^{page 50 note 4} We should interpret in the light of this the fragment of Galen about Democritus (68B 125), that dialogue between senses and intellect, which are, as it were, engaged in a combat as tcwhich of them shall lead. The final meaning, I think, is not contained in the statement that the in tellect's apparent victory is virtually its defeat. The metaphor of the combat is rather meant to show that each is capable of fulfilling its specific task.

^{page 51 note 1} Cf. Theophrastus' full account in de Sens., D.V. 68A 135.

^{page 51 note 2} D.V. 68A 59: οί δ⋯ περ⋯ τ⋯ν Пλάτωνα κα⋯ Δ ημ⋯κριτον μ⋯να τ⋯ νοητ⋯ ὐπενόησαν ⋯ληθ⋯ Ĩναιε.

^{page 51 note 3} There remains a certain amount of obscurity as to how he believed it to take place. The fragments tell nothing about the way in which the true ŏντα are grasped, except that they give the name of Υνώμη Υνησίη. To this we have added as a complement the term λόγοι, found in De gen. et corr. This expression makes it some how more conceivable by what means human understanding reaches the atoms and the void. It does so by a way of thinking which starts from the appearances and refers back, i.e. a discursive thinking. (Sextus says, in accordance with this, in B II, δι⋯ τς⋯ διανοίας.) The term λόγος or λόγοι is derived from the Eleatics. It is probable that this use here is due not merely to Aristotle's interpretation but to Democritus himself. It may be assumed that the Atomists actually developed their philosophy partly in agreement with, partly in opposition to, that of the Eleatics. (As Natorp observed, the philosopher to whom they principally referred is Melissos.) If, then, we are right in assuming λόγοι to be a Democritean term, the fragments would become clearer. In B II the context was presumably this, that the genuine cognizing begins to work where sense-perception is no longer able to proceed towards the more subtle. Now, if we may imagine that that which then begins to act is the λόγος, we can better understand the possibility of this ever more minute analysis of existing things, culminating in the atoms. The λόγοι are an indirect grasping and mediate thus the knowledge of true being. It seems at first puzzling that Democritus, although he knows of a γνώμη γνησίη, emphasizes so strongly that truth is deeply hidden from men. Now I think if we take into account the λόγοι, this proposition of truth being hidden possibly means that man lacks the faculty of seeing the ⋯τεῇǒ ντα by immediate grasping. In any case, we have to keep in mind that the λόγοι are an indirect grasping. We shall see that this is important for the understanding of Aristotle's view of Democritus.

^{page 51 note 4} = D.V. Leuk. A 9.

^{page 52 note 1} Precisely speaking, from the infinity of appearances they inferred an infinity in the number and variety of the shapes of atoms.

^{page 52 note 2} D.V. A 36= De part. anim. A 1,642a 24 and Met. M 4, 1078b 19.

^{pafe 52 note 3} Here we use στϭ in the narrower sense in which Aristotle uses it. He sometimes employs the term in a wider and more formal sense, in which στϭιϰncludes even the ϰεῖον. Therefore even for Aristotle it would be justifiable to call the atoms ⋯ρϰαί, as did Democritus. But this name is used, the essential philosophical difference remains hidden.

^{page 53 note 1} By ‘all that exists’ or ‘existing things’ I try to render rτ⋯ ὑπάρϰοντα. Cf. ⋯νυπάρϰοντος in the above definition of στοιϰεῖον.

^{page 53 note 2} I refer to Met. 6 io. Cp. the following footnote.

^{page 53 note 3} Professor Ross says in his notes on this chapter (Aristotle, , Metaphysics II, p. 275/76)Google Scholar: ‘He (i.e. Aristotle) says in this chapter clearly enough that there can be no falsity with regard to them (i.e. to the “simple”), but he does not say as clearly as he might that there can be no truth either. That which could not possibly be false cannot without tautology, and therefore absurdity, be said to be true… But instead of saying this he says that truth in anothtr than the ordinary sense is possible with regard to incomposites. The fault, however, is only in the expression; the distinction is probably clear enough in his mind’. But the chapter which, indeed, already appeared difficult to the ancients has a definite meaning, as I have been convinced by Professor Heidegger's interpretation, given in his lectures in 1923–26, but not published. According to it Aristotle, in 6 10, means to bring to light the peculiar truth of the ‘simple’. This truth, which excludes any possibility of falsity beside it, constitutes the original and the pre-eminently ‘true’, in the Greek sense of this word, meaning what is not-hidden, or lifted from its hiddenness (⋯–ληθής). Moreover, this primarily ‘true’ is, as such, even the pre-eminent being (τ⋯ κυριώτατα ŏν, 1051b I). Hence we realize that this chapter rightly stands at the end of Book θ. I canno t here expound Heidegger's detailed interpretation at full length, and should only like to add that, according to this interpretation, one feels inclined to eliminate, in 1051b 1, the words ⋯ ψε⋯δος. They are likely to have been inserted by a copyist, in formal accordance with passages which deal with the secondary truth, i.e. truth (and falsity) in propositions. Consequently we should probably read in 1051b 23/24, ⋯λλ Ӗστι τ⋯ μ⋯ν ⋯ληθ⋯ς θιγεῖν καΙ Φάναι ⋯ληθ⋯ς. Christ already bracketed, in his edition of the Metaphysics, ἥ ψε⋯δος, τ⋯ μέν. But, contrary to him, I hold that the ⋯ληθές following ϕάναν should be kept. I owe these textual suggestions, too, to Heidegger's lectures.

^{page 54 note 1} Thus, to speak precisely, Book Г is not. intended to treat the principle of contradiction, i.e. the foundation of Logic, as is often believed, It rather deals with the first and essential onto-logical problem.

^{page 54 note 2} Cp. above, p. 49, note 1.

^{page 54 note 3} I find this meaning of Г 5 especially obvious from 1010a 1–3: x03B1;ῖτιν δ⋯ τ⋯ς δόξης τοότοις ⋯τι περ⋯ τ⋯ν ŏνων BC;⋯ν τ⋯ν λήθειαν ⋯ έσκόπουν, τ⋯ δ' όντα ὑπέλαβον 03B5;Ιναι τ⋯ αίσθητ⋯ μϭνον. For the ‘preeminently true’ compare above, p. 53, note 3.

^{page 54 note 4} Cp. below, p. 55, note 3.

^{page 54 note 5} E.g. 85a 1 sq., 88b 35, etc. Cp. especiallythe final chapter of Posterior Analytics (B 19).

^{page 55 note 1} Especially in Z 6 and, besides, in 1142a 25 sqq. and 1143b 1.

^{page 55 note 2} 1051b 23. Cp. above, p. 53, note 3.

^{page 55 note 3} For the object of καϒάϕασις is necessarily characterized by συμxπƛοκ⋯.Hence it works as combination of Sialpens and aivdcais. We know this as the way of performing a ƛ⋯γειν and δι7alpha;7nu;οῖσθα7iota;.In performing these, we denote something as something. In contrast to them the φ⋯σιςis a mere referring to a ‘single. It develops from the simple becoming aware of a simple and is the way in which the νοεῖνis performed. Compare above, p. 54 and p. 53, note 3.

^{Page 55 note 4} Cp. the following footnote.

^{Page 55 note 5} Compare De An. T 3. where we read about the Presocratics: T⋯ φρονεῖν κɑ⋯ τ⋯ αἰσθ⋯νεαθαι ταὐτ⋯ν εἷ7nu;α⋯ (427a 22). Here the terms the two objects of identification are the same as n i Met. Г5. Aristotle uses here φρ⋯νηɑιςand φρονεῖνin a sense different from his own precise use of these words, according to which πρ⋯νησις denotes reflexion in πρ⋯ξις only. Here the word has its pre-Aristotelian sense as a general title for any kind of thinking. Thus it stands for νοεῖν.

Another point in which De An. Г 3 agrees with Met. Г 5 is that here too failure to see the νο⋯ς is connected with seeing truth in appearance. We as a read that those who identify the νο⋯ς (or theφρ⋯ησις) with αἲσθησις are bound to believe π⋯ντα τ⋯ φα7iota;7nu;⋯7mu;ενα 7epsiv;ἷναι ⋯λ;ηθῇ (427b 3).

This chapter τ 3 of De Anima is indispensable for an interpretation of our passage in A 2. InA 2, Aristotle deals with the soul as ⋯ρχ⋯ κιν⋯σεως. He says that implicitly already the ancients saw the soul in this way, e.g. Democritus, though, meaning the soul, he speaks of the νο⋯ς. But he takes νο⋯ς differently from Anaxagoras. For Democritus definitely means ‘νο⋯ς to be simply identical with soul, for he holds that truth lies inappearance’. We could hardly understand thefor connexion of this without the parallel in Г 3 But with it we can assume that Aristotle has in and view the same problem in A 2. He intends tosay here that Democritus, as he saw truth inappearance, did not know the νο⋯ς as a specific means of conceiving being. Where he mentions, the νο⋯ς, this word means the soul.

^{Page 55 note 1} D.V. 68A 113.

^{Page 56 note 1} De An. A 2, 404a 30/31. Compare the beginning of this article.

^{Page 56 note 2} Cf. above, p. 52.