Pillars of Hercules (Heracles)
I'm not about to join in with the arguments, it has always seemed obvious to me. For a long time I have identified Atlantis as being in the Aegean Sea. Sea levels rose considerably after the end of the last ice age (10,000 BC), and much of the now submerged Aegean Sea area was previously above sea-level. So what does Plato tells us about location? First off he doesn't say Atlantis is 'beyond the Pillars of Hercules' as is popularly thought, and even if he did it could still apply to pretty much anywhere.
Pillars of Hercules could refer to the pillars of the temples of Hercules, of which there were many but Plato actually referred to the stele of Hercules. A stele can just be a flat slab of rock, the ancient Greeks used them as boundary markers. Part of ancient Tyre (in modern day Lebanon) was the Island of Hercules (now submerged) perhaps the stele was there.
To further confuse things, Plato's pupil Aristotle states they were called the steles of Briareus in earlier times. Briareus was regarded as one of three Hecatoncheires, and was also known as Aegaeon. Briareus is associated with Ogygia where he guarded Cronus. Then there are several Homeric references, including Calypso and Odysseus in Odyssey Book V that gives Malta - or more specifically Gozo - as being Ogygia. On the positive side, both options could easily apply to an Aegean location for Atlantis.
The diagram shows Tyre (Tyrus) with the Phoenician mainland (Lebanon) to the east, and the now submerged Island of Hercules to the south.
There are references to the Atlantic Sea or Ocean, but this could easily be written as Atlantis or Atlas and is most likely the Mediterranean. Furthermore the 'ocean' is said to be surrounded by a vast continent (Europe & Africa). If the steles of Hercules were on the Island of Hercules, the 'straits' would be the channel between Crete and the Aegean. There are mentions of smaller seas which could be the north of the hypothesised Aegean landmass, then there are the Sea of Marmara, the Black Sea, Red Sea, and Persian Gulf, all bodies of water that could be attributed to the text.
If this numeric anomaly was applied to the date of Atlantis and it was 900 years (as opposed to 9000) before Solon's time, it would tie in nicely with the eruption of Santorini. Experts cannot even agree on the dates of the Santorini eruption with opinion divided between 1650 BC and 1450 BC. In my experience volcanoes have a habit of erupting more than once. Didn't anybody consider they might both be correct. There is also archaeological evidence of another possibility. There was an earthquake and flooding (tsunami) on Crete in 1450 BC, perhaps there was no eruption at that time. After the undisputed 1650 BC eruption, the magma chamber may have ebbed back into the mantle leaving Atlantis situated above a potential sinkhole. An earthquake could have triggered the collapse or perhaps it was just a matter of time. Evidence from the region that indicates an even earlier volcanic cataclysm around 2500 BC and a similar collapse could have occurred in nearby Milos, a site I favour as the home of Hades.
Now we have established Atlantis did exist, where it was, and when it was destroyed, we can take a look at the inhabitants.