Dr Adrian Goldsworthy

But again, it's a throwaway line.

He just mentions this.

So there's a lot...

Sometimes you have to wonder whether he's actually thinking, well, this is how our army would do it today in the second century AD.

Alexander was great, therefore he must have done it that way.

But still, by ancient standards, as long as you're aware of that great distance in time separating when he's writing, just because it's ancient doesn't mean it's actually close to the events.

But it does give you...

For most of Alexander's campaigns, you can get a fairly clear idea of where he is, what he's doing, some sense of why.

And even the battles, you have a clearer sense than you do of, say, Chaeronea or anything like that.

It's somewhere between 30,000 and 50,000.

So 40,000 is probably a good rule of thumb.

Half of it at most is Macedonian.

Then you have the units of the Pike phalanx that are, and this is the core, the Macedonians with the sort of honorary Macedonians, the Thessalian cavalry, who always, the Macedonian cavalry tend to be on the right, the Thessalian cavalry doing the same thing in similar numbers on the left.

You get some other favored contingents, Illyrians, Paeonians, people like this, who've been part of Philip's army for a decade, sometimes two decades.

These contingents where they're very much part of the team.

Everybody knows what they're doing.

But you're also adding in then people who've come from all the Greek allies, this great league that Alexander's formed.

So it's a little bit hard to be precise about the total numbers, particularly as we're not quite sure whether you're including some of the troops that are already in Asia Minor and not there.

And, of course, the tendency in all periods of history, particularly in the ancient world, is your source knows that a unit is supposed to be of a certain size, 200 men, perhaps 300 men, depending on which sort of companion cavalry it is.

and then assumes that they are always at full strength and everybody's present all the time, which, as the campaign goes on, gets less and less likely.