^{page 198 note 1} The following commentaries will be cited by author's name alone: D., Keyt, ‘The fallacies in Phaedo 102 A–107 B’, Phronesis viii (1963), 167–72Google Scholar. Scarrow, D. S., ‘Phaedo 106 A–106E’, Philos. Rev. lxx (1961), 245–52CrossRefGoogle Scholar. Verdenius, W. J., ‘Notes on Plato's Phaedo’, Mnemosyne, ser. 4, vol. xi (1958), 193–243CrossRefGoogle Scholar. Bluck, R. S., A Translation of Plato's Phaedo, etc., London 1955Google Scholar. Hackforth, R., Plato's Phaedo, etc., Cambridge, 1955Google Scholar. Burnet, J., Plato's Phaedo, etc., Oxford, 1911Google Scholar. Williamson, H., The Phaedo of Plato, etc., LondonGoogle Scholar, 9104. Archer-Hind, R. D., The Phaedo of Plato, etc., 2nd edition, London, 1894Google Scholar. Geddes, W. D., The Phaedo of Plato, etc., 2nd edition, London, 1885Google Scholar. Wagner, W., Plato's Phaedo, etc., Cambridge, 1870Google Scholar. Schmidt, H., Kritischer Commentar zu Plato's Phaedon, 2 Hälfte, Halle, 1850–1852Google Scholar. G. Stallbaum, Platonis opera omnia, etc., Phaedo, editio quarta … curavit Wohlrab, M., Lipsiae, 1864Google Scholar. Bekker, I., Platonis … scripta graece omnia, etc., Phaedo, vol. 5, London, 1826.Google Scholar

^{page 198 note 2} Hackforth, , p. 164.Google Scholar

^{page 198 note 3} Skemp, J. B., The Theory of Motion in Plato's Later Dialogues, p. 8.Google Scholar

^{page 198 note 4} Crombie, I. M., An examination of Plato's doctrines, ii. 164.Google Scholar

^{page 199 note 1} e.g. from 102d6, from of hot or cold, 103e3, from of odd and even, 104c7, from 106d6.

^{page 199 note 2} I use the term ‘particularization’ to express the sensible manifestation of the form of an attribute as well as of the form of a substance. ‘Particular’ used as a noun suggests something substantial and so, as we shall see, in the case of an attribute encourages a false notion of immanent form.

^{page 199 note 3} Burnet's text has been used, except where otherwise stated. Occasional turns of expression have been taken from Hackforth's translation. Once or twice the translation is deliberately literal at the expense of fluency.

^{page 200 note 1} Hackforth, remarks, p. 155, as others have done, that large and small are not qualities but relations. It might be thought therefore that, when Plato introduces fire, which can be only hot, and snow, which can be only cold, he is concerned to draw a distinction between absolute and relative attribution. We shall observe below, p. 228 n. 1, that this difference may in fact have some secondary significance. But the fact that Socrates’ example of accidental attribution has to be expressed in terms of one thing's relation to another is incidental to the main purpose of the argument. The colour of an apple, which is an accidental but not a relative attribute, would have served equally well as a contrast to, and as a preparation for, the essential hotness of fire and the essential aliveness of soul. (It is true of course that relative attribution in this case provides a particularly striking example of accidental attribution. That no doubt is one reason why Plato chooses the example which he does.)Google Scholar

^{page 200 note 2} is repeated at 10362–5, e6, 104a2, a6, b1, b4, b8, d3, and in the case of soul 105d 1, d3, d 10. Variations of ‘never’ occur at 102d6, d8, e6–7, 103b4~–5, c1, 07, d5, d11, 104a3, d10, e1, 105a5, in the case of soul 105d 10–11, and with 106 a g, d 7. These instances include occasions when the opposite form is said to be ‘always’ itself and ‘never’ its opposite. For the point of the analysis is that fire, for example, is like the form of hot in being always hot.

^{page 201 note 1} This way of thinking is clearly represented in Shorey, , see p. 209Google Scholar n. I below. Mills, , Phronesis ii (1957), 139–40Google Scholar, distinguished ‘the opposite in us’ from both forms and ‘sensible participants’ or ‘sensibles’, without making it clear whether he means sensible qualities or sensible substances: as will be seen, the distinction is crucial. Hackforth's, constant references to ‘immanent form’, cf. pp. 143–4Google Scholar, not infrequently cause difficulty: some instances are discussed at the end of this section, see also pp. 212 n. 2 and 217 n. 2. Others who have tried to deal with immanent form in the Phaedo are: Tarrant, D., The Hippias Major (Cambridge, 1928), pp. Ivi–lviiGoogle Scholar; Demos, R., ‘Note on Plato's Theory of Ideas’, Philosophy and Phenomenological Research viii (1947–1948), 456–60Google Scholar; Bluck, , pp. 17–18Google Scholar, and less confidently Phronesis ii (1957). 123Google Scholar n. 1 cf. C.Q N.S. vi (1956), 33–34Google Scholar; Verdenius, , pp. 232–3Google Scholar; Keyt, Scarrow, p. 169 n. 1.Google Scholar

Mills, , op. cit., p. 143Google Scholar, seeks to contrast with the present passage of the Phaedo two passages in the Parmenides, 128e6–129a6 and 130 b, where he finds only forms and ‘sensible participants’. In fact Plato's distinction in the Parmenides between the form of likeness and ‘the likeness which we have’ is precisely the same as the distinction in the Phaedo between the form of large and ‘the large in us’.

^{page 201 note 2} Specific substances clearly described as forms are the shuttle in the Cratylus 389, table and bed in the ‘appendix’ to the Republic 595–8Google Scholar, and, reasonably clearly, fire in the Timaeus 51b–52a, on which cf. Cornford, Plato's Cosmology, pp. 188–91Google Scholar. speaks, Vlastos of the ‘marginal status’ of forms designated by the common noun, Philos. Rev. lxv (1956), 93–94Google Scholar. writes, Ross, Plato's Theory of Ideas, p. 24: ‘Ideas of substances … are not mentioned in the Phaedo, and are nowhere prominent except in the Timaeus, though they were involved in the theory, since it was the theory that there is an Idea answering to every common noun (Rep. 596a6)’.Google Scholar

^{page 202 note 1} We observe below, p. 213, that Plato probably thinks of simply as number. If meant three as an attribute of numbered objects, then three oxen would be like large Simmias. Three oxen would contain as an attribute the particularization of the form of three. Oxen themselves, if they were to be included in the theory, would be the particularization of another form. Thus there would still be no ‘immanent form’ that was other than the form of three and other than the particular three that characterized oxen.

^{page 203 note 1} Later in the sentence we are told of something other than the opposite receiving 10365, where fers to Hackforth acknowledges his difficulty in explaining this phrase if as well as means ‘immanent form’. The phrase is in fact a simple one. Something like snow has the particularized characteristic, of an opposite form,

^{page 203 note 2} ‘Withdraw’ is used to cover, in the present passage, and and later in the argument 106a4, and or 106a10 and e7.

^{page 204 note 1} This is the interpretation of earlier writers and of Burnet, notes on 102d9 and ea; Taylor, A. E., Plato, the Man and his Work³, pp. 204–6Google Scholar (criticized by Ross, , P. T.I., pp. 30–31)Google Scholar; Cornford, , Plato and Parmenides, p. 79Google Scholar (cf. Plato's Cosmology, p. 184)Google Scholar; and Bluck, , pp. 17–18Google Scholar, 24, 118, 191 n. 2, and, in passing, C.Q. N.s. vi (1956), 35. A cause of confusion has perhaps been the mistaken application to the metaphor of Aristotle's criticism of the Phaedo at De gen. et corr., 335b 14–15.Google Scholar

^{page 204 note 2} The result is particularly clear in two remarks by Taylor and by Bluck. writes, Taylor, op. cit., p. 205Google Scholar: ‘When “cold” at tempts to “occupy” fire, or “heat” to “occupy” snow, an essential character of the thing must either “withdraw” or be “annihilated”, and in either case (my italics) the thing, the fire or the snow, is no longer the thing it was’. writes, Bluck, p. 118: ‘If a man who is bad in a certain respect is to become good in the same respect, the “bad in him” must first depart—it must either “flee and give way” or “perish’. Taylor and Bluck have reduced Plato's two events in effect to a single event.Google Scholar

^{page 204 note 3} Apart from the phrase in the present passage, 102d9, Plato consistently writes each time he presents the formula: 102d9–e2, 103a1, d8, d10–11, 104b 10–c1.

^{page 204 note 4} It has seemed slightly simpler in explaining this third image to speak of physical change, Simmias or Socrates becoming actually larger or smaller. This is in effect how the thought recurs at Theaetetus, 155b 6–c1. But very possibly Plato is thinking simply of comparison. Simmias can cease to be compared with Socrates—rather than Simmias' actually growing smaller or Socrates' actually growing larger—and in that case the comparative ‘largeness’ in Simmias ‘perishes’.

Equally, in the previous image the point may be not so much that Simmias and Socrates stay the same size, though that will in fact have to be so, but that despite the approach in some way of smallness Simmias continues to be compared to Socrates. Possibly the approach of smallness means that Simmias is compared to Phaedo. Despite his now being seen to be smaller than Phaedo, Simmias is still larger than Socrates, because the largeness in him has ‘withdrawn’.

^{page 205 note 1} 102d7–e3, 103d5–8, d 10–12, 104b 8–C3, c7–9, e7–105a5, a6–b3, 105d 10 ff., 106 a 5–6, b3–4, At 106 a 3–6 Plato writes that if were imperishable, The point of this is not that snow would ‘stay behind’ if it were not imperishable. For snow is never hot. It can never ‘stay behind’ at the approach of the opposite. The exclusion of the alternative, depends on the preceding argument, where the metaphor is first introduced, not, as does on the conditional clause.

^{page 205 note 2} This follows Heindorf ap. Bekker.

^{page 205 note 3} This is the interpretation of Stallbaum, followed by Schmidt, , ii, pp. 43–46 and Williamson.Google Scholar

^{page 205 note 4} It follows that it would be clearer to reverse Burnet's punctuation: to print a colon instead of a full stop at 102e3, and a full stop instead of a colon at e 6.

^{page 206 note 1} Even were we to adopt Heindorf's and Hackforth's interpretation then would still be the large in us. For the combination of would show that what was compared to was not the small itself, but the small in us.

^{page 206 note 2} alone, without is applied to opposite forms at 104 b 7–8 and 105a2. Elsewhere the fact that one opposite form will not accept its opposite is expressed in terms of 102d6, 103c 1, or ’always having the right to its own name’, 103e3–4, e6–7, or simply in terms of and 103b4–5, C7–8.

^{page 206 note 3} Whether at this moment, 104c8–9, we supply and so think of the form of three, or whether we think of particular three, does not affect the interpretation of the metaphor. We may either look back to 104 b 8–C4, where Plato is speaking of things that perish, and so must be thinking of particular three; or we may look back to C5 and forward to d 1 ff., where, so we shall argue, pp. 216–21, Plato introduces the form of three. The form of three could properly be said not to ‘stay behind’ at the approach of an opposite form. On the other hand the form of three could not have applied to it the rest of the metaphor. For the form of three, being a form, could no more ‘withdraw or perish’ than could the opposite forms.

^{page 207 note 1} See Hackforth, , p. 165 n. 1, and Scarrow.Google Scholar

^{page 207 note 2} For the meaning of see p. 203. and may be used in the lines immediately before this of the particularization of form, see pp. 202–3.

^{page 207 note 3} Jackson ap. Archer-Hind conjectures

^{page 207 note 4} is used in later Greek in a different sense, see L.S.J., s.v. Plato happens not to mention ‘not-odd’.

^{page 208 note 1} p. 148 n. 3, cf. pp. 155–6 and 164–5. The idea is derived from Cornford and probably from Ross, as cited on p. 205 n. 1 above.

^{page 209 note 1} Shorey allows that in general in Plato's world ‘there are really only two things’, form and particular, but writes of the present passage that it ‘seems to yield three things: the idea, per se, the idea in the particular, and the particular as affected by the idea’, The Unity of Plato's Thought (Chicago, repr. 1960), p. 41Google Scholar n. 284. This obscures the point that the initial idea, largeness, is not the idea precisely of what Shorey calls the particular, namely Simmias. For there are in effect two possible particularizations: the particular attribute, relative largeness or smallness, and the Simmias or Socrates. There fore we have here in effect nothing more than the normal twofold division of form and particularization: largenesss and the largeness in Simmias. Shorey reveals the source of his error when he goes on to speak of ‘the duplication of the idea’ in this passage as ‘a device employed here only … for the purpose of the argument’. For ‘the duplication of the idea’, i.e. the opposite itself as the opposite in us and the opposite in nature, is not a duplication in some strange sense of the form, but the normal division between form and particularization of form Cf. Keyt, , p. 169 n. I.Google Scholar

^{page 210 note 1} The scope of the law of opposites was defined at 70d7–e6 to include all opposites and Strictly perhaps soul does not offend this law. For under the new definition of death soul has no nor is it itself an opposite. But in so far as soul characterized by life is an we might expect it to fall within the scope of the law. (Similarly, fire itself is not an opposite; but in so far as it is characterized by hot it is, we may suppose, an )

^{page 211 note 1} 103 d6. The first example said that an opposite character, ‘the large in us’, could never become its opposite, but must withdraw or perish. Now we have a substance, snow, which in virtue of an opposite characteristic cannot admit an opposite, but must withdraw or perish. Hackforth, , p. 149Google Scholar n. 3, therefore describes as a ‘formal inaccuracy’. But this is too strict an interpretation of Plato has immediately before this deliberately introduced a new element into the argument ( 102e 10): namely, substances such as fire and snow in addition to the particularizations of opposite forms.

The point of his remark is precisely to show that the earlier law applies to these substances.

^{page 211 note 2} 103e 2. This qualification may apply in two ways. 1. Only some opposite forms have things essentially characterized by them: for there is nothing that in Plato's sense is only large and not small. 2. Only some things characterized by opposite forms are characterized essentially by them: as well as fire which is only hot there is water which can be at once both hot and cold. The latter qualification is essential to the understanding of Socrates' question at 105b8–c2, see pp. 224–5.

^{page 212 note 1} thought of simply as number, i.e. three thought of as particular three, but not as associated with a group of sensible objects, should lead, we might think, to a theory of intermediates. We do perhaps see here an indication of how Plato's mind worked to produce such a theory. But in this argument cannot as yet properly be intermediate three, for at 104c 1–3 and by implication at 106 a 1 (cf. p. 208) Plato speaks of the possibility of perishing, whereas according to Aristotle, Met. 987^b16, the intermediates are everlasting. On the question of intermediates in the Phaedo see Ross, , P. T.I., p. 25Google Scholar, which is not entirely consistent with p. 60. For more recent discussion references are given by Verdenius, p. 210. To these add Bluck, Phronesis iv (1959), 7Google Scholar n. 2, and Rist, , Phronesis ix (1964), 27ff.Google Scholar

The alternative notion, of three as the attribute of a group of numbered objects, is one that cannot in principle be easily defended. The idea of number as the property of an agglomeration of things on a level with colour or shape is criticized by Frege, , Die Grundlagen der Arithmetik, §§ 21–25Google Scholar. But Frege's highly speculative considerations need not have occurred to Plato at this point. Plato could have thought of the three in three oxen as simplyanalogous to the large in large oxen. In fact, to provide an example of essential predication, Plato has chosen to concentrate on three simply as number.

^{page 212 note 2} 103e3, discussed above, p. 203. If this phrase meant ‘immanent form’ as Hackforth supposes, there would be no point in the difference between

^{page 213 note 1} Essentially the same qualification is contained in 102e4, 102e8, 103d7, d 12, 104c3.

^{page 213 note 2} Fr. 123m Wehrli. Strata's criticism is repeated with approval by Hackforth, , p. 163Google Scholar, and by some earlier critics. Essentially the same criticism is echoed by Keyt, , p. 169, when he complains that laughter, as well as soul, ‘always comes to whatever it occupies bearing life’. For Keyt's point is in effect that laughter would come to a man, as hotness to fire, ‘so long as he is’.Google Scholar

^{page 216 note 1} The view that before this there has been a form of fire and a form of snow is considered under (iii) below, pp. 219 ff.

^{page 217 note 1} Burnet, note ad loc, denies that Plato's language here describes forms. There will be the distinction between form and particular whichever translation we adopt, cf. pp. 216–17 above.

^{page 217 note 2} Hackforth's interpretations of this passage, if carried to their conclusion, lead to hopeless confusion. We have observed above, p. 203, that Hackforth designates as ‘immanent form’. We shall see in a moment that Hackforth also equates and in this passage as ‘immanent form’. Since the sequence of Plato's thought and language demands that is the same as ‘whatever the form of three possesses’, it follows on Hackforth's interpretation that the same thing, immanent form, will be at the same time object and subject of the verb

^{page 217 note 3} We have already commented, p. 208, on the implication of the ‘unreal’ form of condition at this point.

^{page 218 note 1} It follows that in the preliminary statement of the numerical example, 104b 2–3, is later expressed as in line with 104e8–10.

^{page 218 note 2} Verdenius's comparisons are accepted in effect by Rise, , op. cit., pp. 30 and 34–35, who writes mistakenly, p. 30,in the present passage, just as on p. 27 he writes where in the text cited, 74 c, we have the plain We may also question Rist's p. 35. At least in the Phaedo, 105c6, Plato writes simply Google Scholar

The meaning of in the Phaedo has recently been much debated. The latest additions are the articles by Rist and Haynes already cited, where references to earlier contributions may be found. Most likely do simply replace If so, they are probably exceptional, in that by their very meaning it is natural to think, at least casually, of equality and likeness as pairs of equal and like things, even though strictly it is a form that is being spoken of and a form should be a unity. In the Phaedo Plato has perhaps been encouraged to write by the example of the plurals immediately preceding, (Earlier when he speaks of and Plato uses the singular expression 7439–12.) Plato is then perhaps moved to add simply because he feels that the plural expression is an unusual one for a form. Contrast Bluck, , op. cit., p. 117: the addition of ‘can hardly be because Plato wanted to make it plain that he had referred to the Form’.Google Scholar

^{page 219 note 1} We deal with fire not snow because fire is mentioned more often. Whatever is true of fire will fairly obviously be intended to be true of snow.

^{page 220 note 1} 104b 10 and C7, 104d 9–10 and had hitherto been used of one opposite, whether form or particularization, ‘advancing upon’ or attacking another opposite, or something which contains the other opposite. Now, if our interpretation is correct, 105a3–4 is used without this implication of hostility of a form occupying a particular. Similarly, will be used in a moment, we shall argue, of the form of soul coming to particular soul, 105d3. Later, as one word or two, will be used again in the hostile sense, 106b 3, b8, e5. This shift in meaning does not seem too strained.

^{page 220 note 2} It would be difficult to interpret the definition at 104d 1–3 in terms of the distinction between fire and wood, even if, like Burnet, we were to suppose that and need not apply to forms. For we might perhaps argue that sensible fire ‘possesses’ wood and impresses its ‘character’ upon it. But how could snow ‘possess’ some thing and impress its ‘character’ upon it? For fire can turn something into fire: but it is not obvious that snow can turn something into snow. And yet, for the purposes of the argument, whatever is true of fire should presumably be true also of snow.

^{page 221 note 1} 10538. Perhaps weshould prefer the commoner transitional sense, ‘of course'; so Hackforth.

^{page 221 note 2} 105 a 8. Also, i.e. as well as being characterized by the even and excluding the odd. Other possible renderings are ‘although itself’ and ‘actually itself’. Cornford in his article on this passage, C.Q. iii (1909), 189–91, adopts the latter translation at one point; another time he translates ‘itself also’. Verdenius, p. 235, wrongly interprets: as well as ‘other things, such as, for example, five, which as being “half” is the opposite of ‘double”, ‘not of “even”’.Google Scholar

^{page 221 note 3} 105 b 1–2. Perhaps this is too much meaning to give to Verdenius may be right, that is a gloss.

^{page 221 note 4} The opposites are almost certainly double and half, not double and single, as Hackforth, prefers, p. 153Google Scholar n. 1. See the passages quoted by Verdenius, , p. 235. Cf. also Aristotle, Top. 135^b17–26, 147^a20–31, De soph. el. 181^b27–28, Met. 1020^b26. In the present instance ten is double five; five is half ten.Google Scholar

^{page 221 note 5} This interpretation seems to be suggested by Cornford's article on the passage. However, at least at one point (mere is some inconsistency in his exposition) Cornford speaks of ten ‘qua “double”’ as the subject of the sentence, and so appears to follow earlier commentators, e.g. Wyttenbach ap. Bekker, in making out that it is specific numbers like ten and five, not die opposites double and half, which constitute the exception to Plato's earlier statement of his thesis. If ten were the subject of the sentence, at 105 a 8 would have to refer to which with a singular neuter noun, intervening would be very awkward. The phrase ‘qua double’ destroys Plato's point, that numbers like ten and three are precisely to be distinguished from the opposites odd and even or double and half.

^{page 222 note 1} There would in fact be this difference between the two examples. The halves of double numbers, e.g. 5 the half of 10 and 6 the half of 12, are (except on Cornford's thesis) either odd or even, unlike the doubles which can be only even. The series of odd fractions like the series of even fractions excludes wholeness.

^{page 223 note 1} A less literal translation will be: ‘What makes whatever body it comes to be in hot … holness makes whatever it comes to be in hot.'

^{page 223 note 2} 105 b 7, looks back to 100c–e and 101d2, where is used three times. 105 c 2, looks back to 101 C 8.

^{page 224 note 1} Hackforth's translation follows Archer-Hind. It is curious that in his commentary, p. 161, Hackforth should go out of his way to criticize Plato precisely for making fire the only cause of heat.

^{page 224 note 2} That fever is not the only cause of sick ness happens to be stated explicitly in the second Alcibiades 140–b. Cf. Rep. 610b1–2.

^{page 224 note 3} p. 158 n. 2, quoting Stob, . Eel. i, p. 22Google Scholar. 19 Wachsmuth, ; and p. 161.Google Scholar

^{page 225 note 1} It may be true, however, that soul is the only cause of life, in that other things are alive only through the agency of soul. This at least is the notion presented in Phaedrus 245 c–e. Perhaps, however, in the present argument Plato thinks of animal bodies as alive and of the soul as essentially alive.

^{page 225 note 2} There is, as we have noted before, p. 212, this difference: that a group of things cannot at the same time be odd as well as even, whereas water, for example, can at the same time be hot as well as cold. Some systems of course allowed for the notion of 105 c 2, looks back to 101 C 8.

^{page 225 note 3} Whatever is true of fire will presumably be true also of fever: but it is best to found our analysis on those examples which Plato has described more in detail. Plato has added fever and sickness, as earlier he has added ‘the form of whole’ (if we take the simpler interpretation of that passage), 105b 1–3, simply to remind us that his laws are of wide application.

^{page 226 note 1} Opinions on the possibility of soul as form have been given above, p. 219. It is perhaps not too cavilling to question Hack-forth's distinction between soul as form and soul as substance, pp. 162, 163, and 165. For Plato forms are substances, if we are using that word in the Aristotelian sense, see, e.g., Met. 1080^al, where the forms of the Phaedo are described as

^{page 226 note 2} There would be the same implication with the less likely alternative translation, although the verbal parallelism would be slightly less exact.

^{page 226 note 3} There is the same implication at Rep. 611a4–6.

^{page 226 note 4} This is especially clear in the argument on the similarity of soul to forms, 78b–81a, although the sinful soul can of course be described as entangled with what is and so forth, 81c.

^{page 227 note 1} The argument concludes with 107a1. Individual souls have been Plato's concern throughout the dialogue, as is evident from such phrases as al see 70b2–4, C4–5, 71e2, 72d9–e1, 76 e6.

^{page 228 note 1} It is perhaps not fanciful to suppose that as well as the development of the distinction between form and particularization, and between accidental and essential predication, there has been a progression, whether conscious or unconscious, in Plato's choice of examples of opposites. I. A thing can be both large and small: at the same time larger than one thing and smaller than another, and at different times larger and smaller than the same thing. There is nothing that is in Plato's sense essentially large or essentially small, only large or only small. (A half, for example, is essentially smaller than a whole, but at the same time it is larger than something else, a quarter; it is not only small.) This therefore is the least exclusive of Plato's examples. 2. A thing can be both hot and cold: at the same time hotter than one thing and colder than another, and at different times hotter and colder than the same thing. Now Plato introduces a new feature: things like fire and snow that can be only hot or only cold. Hot and cold therefore can be manifested as more exclusive opposites than large and small. 3. Numbers can be only odd or only even in the same way as fire can be only hot and snow can be only cold. Numbered objects can be odd and even at different times, in the same way as water can be hotter and colder at different times. But there is nothing that can be odd and even at the same time, in the way that water can be at the same time hotter than one thing and colder than another. In this way odd and even are a degree more exclusive than hot and cold. 4. A body can be alive at one time and later dead. Nothing can be alive and dead at the same time. To this extent alive and dead are as exclusive as odd and even. Soul can be only alive, in the way in which three can be only odd and two can be only even. But life in soul excludes, as we shall see, the possibility of change. In this way the opposite form of life exhibits a new degree of exclusiveness. This development of Plato's thought, if such it is, is not unlike the development of thought and expression in the images of the line, the sun, and the cave in the Republic.

^{page 229 note 1} 105e8. The qualification, which Hackforth leaves untranslated, looks forward to and the addition of

^{page 229 note 2} This view follows essentially from Strata's criticism, see p. 213 above. It is adopted by Keyt, and earlier, for example, by Landmann, T., ‘Tendenz und Gedankengang des platonischen Dialogs “Phaedo” Gymnasialprogramm (Königsberg in Pr., 1871), p. 8Google Scholar, and Schneider, G., Die Weltanschauung Platos dargestellt im Anschlusse an den Dialog Phädon (Berlin, 1898), pp. 106–8.Google Scholar