whose date and provenance are very dubious.
is horizontally directed. Here it refers to the circular expanse of the earth's surface. The area embraced need not of course, be strictly circular; cf. the rect angular walls of Babylon, Herodotus 1. 179: 
Better yet, cf. Critias 118 d.
equals 
In other words, the areas, inhabited or uninhabited, each of them at least potentially an 
, which make up the vast earth, are usually coextensive with the 
cf. 109 b 4–5. In a later passage, however, 112 c 6, d 1, 
seems to refer to a subdivision of our 
The reason for this lack of precision may be explained as follows. When Plato talks about the earth, he has in mind either the vast horizontally oriented expanse with its numerous 
(let us call this: image I), or our particular 
(image II), or a mixture of the two, a horizontal plain sloping downward towards the middle, but divided into zones (image III); for the last, cf. Xenophanes A 41 a Diels = Aëtius 2. 24. 9: 
Also Parmenides A 44 a Diels, espe cially Aet. 3. 11. 4 where we find a similar use of 
It is this mixed image which is responsible for the location of several 
in what appears to be one 
What I have called Plato's lack of precision seems to be connected with a similar multiple reference in earlier scientific writing. In their remarks about 
the pre-Socratics were as likely to have in mind the cartographical arrangement of the 
as the theoretical shape of the body of the earth (Xenophanes B 28 Diels), or again the characteristics of the element 
as contrasted with 
and 
(Empedocles B 37;39).
whatever its significance, has centripetal force, and is thus not suitable to designate linear extension. The confusion has been further compounded by a comparison with Timaeus 40 b-c where Plato does refer to an axis, and Taylor, in his commentary p. 234, note 2, proposes to substitute 
the verb of the Timaeus passage, for Aristotle's 
Thus Aristotle's vagueness has become responsible for a notorious problem of interpretation; cf. 
and 
here do not refer to Tartarus, in spite of Pindar fr. 207 Snell et al., but to the fluid. The words are almost synonymous, denoting here not ‘bottom’ but ‘fixed position’. For 
see 
For 
cf. 
also Solon 1. 20 Diehl.
the underlying conception is image I. ‘The region beyond’ refers comprehensively to other hollows, divided from 
, our own hollow, by what Plato had earlier called the ‘true earth’: 110 a 1, 8 ff. Both the ‘true earth’ and the hollows outside of our own particular hollow are situated 
There are, of course, no rivers of any kind on the true earth; cf. 11 o c 7: 
That this deprives the men who have lived exceptionally well (114b 6-c 2) of a means of transportation does not seem to have bothered Plato.
indicates, with self-conscious emphasis, that our region has earlier been shown to rest at the bottom of a 
cf. 109 c ff. I am inclined to think that 
is an interpolation, though I have retained it in the translation. It dupli cates 
of 112 c 1. If I am right, the plural noun was added to create an antece dent for 
and perhaps also for 
of c 5. 
stands and falls with 
My colleague Prof. John B. McDiarmid suggests that the original read ing of c 2–3 may have been: 
like the one before at 112 c 6, is based on image III; cf. p. 194, note 2.
and its equivalents. Within a 
the ground slopes down toward the middle; this explains the sense of 
are terms referable to image III. 
points to the land on the other side of the Mediterranean which cuts the horizontal land mass into a northern and a southern half, or to the other side of the torrid equatorial zone. Cf. Thucydides I. 136 
for a similar perspective, on a reduced scale. Burnet brackets tlapei; with Robin, I see no reason for the deletion.
in combination with 
112 e 6, seems to point to image I. Oceanus surrounds our 
If we assume that our 
lies towards the nordi of the vast disc-earth (Olympio-dorus in Platonis Phaedonem C III, p. 202, 12 Norvin, seems to preserve an ancient tradition to this effect; cf. also D p. 241, 5 where a misunderstanding of Plato's arrangement causes the commentator some embarrass ment), then Acheron controls an area to the south, from where it flows northward to the centre of the disc where, below the surface, it forms the Acherusian lake.
equidistant from the other two, i.e. according to Olympiodorus, east. The rivers, then, divide the earth of image I into four geographical sectors, or quarters, each controlled by its river, which at one point in the course of its meanderings flows close to the Acherusian lake. The four rivers together form a virtual network of waterways, surface and underground, for the whole earth. The location of the lakes created by Pyriphlegethon and Cocytus indicates that, at least part of the time, the sinners carried along in these rivers find themselves on the surface of the earth. Uninhabitable regions, far away from our 
are thus drawn into Plato's eschatological scheme. (It is to be understood that this arrangement, with the Acherusian lake at the centre of the map, is only one of several likely dispositions.)
is a non-committal term, comprehending movement both above and below the ground. Burnet's note ad loc. and Bluck's translation are needlessly specific. Cf. 
The traditional translation, ‘at a lower point of Tartarus’, cannot be right, in spite of 
‘beyond Delos’; also Aristophanes, Frogs 70.1 suspect that 
is corrupt. Meanwhile I have translated it as if it were a genitive of sphere, i.e. as if = 
Crossref Citations
This article has been cited by the following publications. This list is generated based on data provided by Crossref.
Dicks, D. R.
1966.
Solstices, Equinoxes, and the Presocratics.
The Journal of Hellenic Studies,
Vol. 86,
Issue. ,
p.
26.
Couprie, Dirk L.
2011.
Heaven and Earth in Ancient Greek Cosmology.
Vol. 374,
Issue. ,
p.
201.