The ancient site of Stonehenge is shrouded in mystery - we don't know how it was built, or what it was built for.
One of the strangest mysteries is why those giant stones were dragged to an unremarkable Salisbury hillside instead of being erected where they were originally found. Now it seems we have an answer - some of them were already there.
Archaeologist Mike Pitts - one of few researchers to ever perform excavations within Stonehenge - has found evidence that two of the monument's biggest and most important stones, called sarsens, may have been at the site for millions of years.
In the 1970s, it was thought that sarsens - giant boulders of sandstone weighing up to several tons - didn't occur naturally on the Salisbury Plain.
Researchers concluded that the people who constructed Stonehenge must have schlepped all the stones from Marlborough Downs, 32 kilometres (20 miles) away.
Just why they would do so, when they could have far more easily erected the henge at that location instead, was a real head-scratcher.
But now it looks like the Heel Stone and Stone 16 may hold the answer.
These two stones, unlike the rest of the monument, haven't been carved or shaped in any way. When lined up, these are also the stones marking the horizon where the Sun rises on the summer solstice, and sets on the winter solstice.
If the two stones were already sitting there, pointing at the solstice Sun, that could have given the site its significance for the people who lived nearby thousands of years ago.
Pitts hinges his explanation on two holes very close to the stones.
Near to the Heel Stone, he found evidence of a large hole, up to 6 metres (20 feet) in diameter. This was far too large to have been dug as a "socket" to place a stone, and it's always been a bit of a puzzle - but it could have comfortably contained the Heel Stone.
If the stone had been excavated and raised on the site, that would neatly explain this seemingly misplaced hole.
"The assumption used to be that all the sarsens at Stonehenge had come from the Marlborough Downs more than 20 miles away," Pitts told The Times.
"The idea has since been growing that some may be local and the heel stone came out of that big pit. If you are going to move something that large you would dress it before you move it, to get rid of some of the bulk.
"That suggests it has not been moved very far. It makes sense that the heel stone has always been more or less where it is now, half-buried."
There's only one other buried pit of comparable size at Stonehenge, and it's very close to Stone 16.
The pits, like the stones, lay along a solstice axis - and the entire geometry of Stonehenge could have been built around this naturally occurring coincidence, Pitts explained in a Stonehenge special edition of the journal British Archaeology.
More recently, evidence has emerged that sarsens can be found on the plain, although they're scarcer and smaller than those found at Marlborough Downs, with a different shape.
The stone formed millions of years ago, probably during the Tertiary period, and would have been broken and weathered by the permafrost of repeated ice ages.
This supports Pitts's hypothesis, and could finally help lay to bed the mystery of why the location of Stonehenge was chosen.
"Continued radiocarbon dating may reveal further clusters of middle neolithic ritual features," Pitts wrote.
"But for now, the combination of a little henge, large cattle bones … and perhaps the two largest natural sarsens on the plain aligned with the rising midsummer and the setting midwinter Sun, make the site locally unique.
"It all suggests that Stonehenge didn't so much burst into view shortly after 3000 BCE, as grow slowly over a long time before."