^{page 25 note 1} (126). Professor Fitton-Brown points out to me that artistic selectiveness (which I find in the Phaedo's account) may have determined Aristophanes' choice of symptoms also. The whole section plays on comic oppositions, as Dionysus criticizes various routes to Hades: hanging is stiflingly hot ( 122), it chokes the throat, hemlock is much too cold (, 125), it freezes the legs. The choking in the throat that hemlock also induces is omitted, to sharpen the contrast with hanging.

^{page 25 note 2} 186–94. Nicander: Poems and Poetical Fragments, ed. with tr. and notes by Gow and Scholfield (Cambridge, 1953).

^{page 25 note 3} Dwight, E., Toxicology (Philadelphia, 1904), p. 259.Google ScholarKobert, R., Lehrbuch der Intoxication (Stuttgart, 1906), ii, 1079–80.Google ScholarWitthaus, R., Manual of Toxicology (New York, 1911), pp. 924–5.Google ScholarMcNally, W., Toxicology (Chicago, 1937), p. 517.Google Scholarvon Oettingen, W. F., Poisoning (New York, 1952), p. 317.Google Scholar

^{page 25 note 4} e.g. McNally: ‘Poisonous doses produce the general symptoms of weakness, languor and drowsiness, but not actual sleep. Movements are weak, slow and unsteady, and the gait is staggering. Generally there are nausea and vomiting, with profuse salivation … There are tremors and febrillary contractions of the muscles, with occasional convulsions.’

^{page 26 note 1} This disparity is noted by John Burnet, Plato's Phaedo, App. 1, pp. 149–50. He finds it puzzling; and, since he presupposes that the account is historically accurate, he wonders if the drug was indeed hemlock. He is reassured on this point by ‘an eminent pharmacologist, my colleague Professor C. R. Marshall’, who affirms: ‘… Personally, I am decidedly of opinion that his death was due to conium (i.e. hemlock-juice). It is difficult to be absolutely positive on this point, as conium is somewhat peculiar in its action, and the symptoms produced vary with the dose and probably with the individual.’ This vague reply satisfied Burnet but leaves the puzzle where it was. Poisonous doses of hemlock, to judge from the toxicologics, regularly produce a number of symptoms: some of these Plato omits. That is the problem. To say ‘the symptoms … vary … probably with the individual’ (my italics) is not enough.

^{page 27 note 1} 66 b ff., 80 d ff., 114e.

^{page 27 note 2} e.g. Witthaus.

^{page 27 note 3} Cf. Claus, D., Psyche: A Study in the Language of the Self before Plato (diss. Yale, 1969).Google Scholar

^{page 27 note 4} e.g. Philol. 13, Diog. Apoll. 4, cf. Arist. de An. 405^b11.

^{page 27 note 5} Phd. 79 c, Tht. 184 c ff

^{page 27 note 6} Contrast the rapidity of the exit of the Homeric psyche: (II. 16. 856, 22. 362), (Od. II. 222).

^{page 27 note 7} 118 a 3–4. Cf. 515 d 3–4.

^{page 27 note 8} 115 d. Cf. 69 b–c, 84 a–b, 106 e–107 a.

^{page 28 note 1} The usual interpretation of the remark, 118 a 8–9.

^{page 28 note 2} 67 c 6–d 4.

^{page 28 note 3} Compare Sophocles' calm pleasure at having escaped, through old age, the of lust, Rep. 329 c 4.

^{page 28 note 4} It also reinforces the argument. This measured and ordered picture encourages us to see the psyche as an entity separable from the body, not just the ‘harmony’ of the physical parts while they are all working together (85 e ff.), but the inner centre of sensation and cognition visibly separating itself from its shell.

^{page 28 note 5} This article was read as a ‘communication’ at the Classical Association A.G.M. in Bristol, April 1972, and is summarized in the Proceedings, 1972. I am grateful to Professor E. A. Havelock for suggesting, in conversation, the basic idea of this article.