The time-limits imposed by the κλεψύδρα on speakers in Athenian trials have been much discussed, but a valuable distillation of the ancient evidence and modern interpretations of it has recently been made by P. J. Rhodes, A Commentary on the Aristotelian Athenaion Politeia (1981), pp. 719–28. He prudently states his own conclusions in a cautious manner, but I find them convincing. One khous of water took 3 minutes to run out; this is indicated by the length of time taken by the κλεψύδρα found in the Agora (first published by S. Young, Hesp. 8 [1939], 274–84), which holds 2 khoes and takes 6 minutes, and it is also consistent with the evidence of Aiskhines about the διαμεμετρημένη ⋯μέρα. In a ‘measured-through day’, used only for public cases, the total amount of time allowed for the speeches in a trial was 11 amphoreis (Ais. 2.126), equivalent to 132 khoes, taking 396 minutes; one third of this time was allocated to the prosecution, one third to the defence, and one third to the speeches on the assessment of the penalty (Ais. 3.197). Time taken for other proceedings, including the allocation of jurors to courts, voting, and payment of jurors at the end of the day (this last is not mentioned by Rhodes, but it was surely completed well before dusk, because the jurors had time to go shopping afterwards; cf. Ar. Wasps 303–11, 788–9), was additional.