Preseli bluestones – is Milford Haven the wrong place?

by Edward Pegler on 1 November, 2010

Why should the Preseli bluestones have gone via Milford Haven to Stonehenge when they might have taken a less perilous route to the east? And does this have anything to do with an end ‘Neolithic’ Irish copper trade?

map of Pembrokeshire, with the major rivers, standing stones and rocky cliffs indicated

map of Pembrokeshire, with the major rivers, standing stones and rocky cliffs indicated

Newport, on the north coast of Pembrokeshire in Wales, is a small and charming little town, blighted only by having the main road from Fishguard to Aberystwyth going through it’s narrow main street. At Newport the Nevern (or Nyfer) River reaches the Irish Sea in a series of twists. It is a small and unexciting river, just over ten miles long, although it was a route of pilgrimage in medieval times.

Newport itself was established by the Anglo-Normans. However, it is probable that it sits at the site of a much older ‘port’ now lost in the sands. For along the River Nevern are a series of Neolithic ‘Portal stones’ (portal nothing to do with port). The newest of these to be ‘found’ is the Trefael stone, a megalith once part of a more complex structure and engraved with cup and ring marks.

Crossing the Pembrokeshire headland

From Newport the river heads south. After a few miles it becomes unnavigable, even to small boats, as it is followed down toward the eastern end of the Preseli Mountains. To continue the journey south it’s necessary to leave the Nyfer, cross the watershed and link up with the River Taf  heading south.

The Taf River, although also small, is probably large enough to have allowed a small boat to travel down most of its course, past the odd standing stone and burial chamber. It eventually comes out on the south coast of Pembrokeshire at Laugharne.

Notably, the watershed between these two rivers, including Carn Menyn (and Carn Goedog), is also the location from which most of the famed Bluestones of Stonehenge are thought to have come. I think that this connection may be significant. (note – Bevins et al. 2011 report that at least one stone comes from lower ground to the north, but this is still part of the watershed – 2/3/11)

Bluestone transport

Traditionally, those archaeologists who believe in the great ‘Bluestone Transportation Event’ have, led by Richard Atkinson, opted for the River Cleddau, leading down to Milford Haven, as the route for the stones’ transportation. There’s a lot to be said for this route. Relatively gentle slopes down to the river and an abundance of standing stones at the river’s headwaters.

However, once at the river, it is not really any better than the Taf for transporting the stones. And on arrival at the coast, any boatman is faced with the difficult business of getting around some challenging and very exposed, southwest facing rocky headlands. To me this seems like a big disadvantage.

(For either the Taf or Cleddau route, a further 5 km of portage through modern Swansea or across the Gower would save the difficult journey around Worm’s Head too.)


It has long been assumed by most archaeologist that the Bluestones were transported by people from Pembrokeshire to Stonehenge. This may or many not be true. However, the reasons why anybody should have bothered have not been really discussed, other than suggesting a vague sacredness of the Preseli area. I have no problem with the bluestones being sacred. It’s just that there are some perfectly attractive, and potentially just as sacred, stones in the Mendips or Quantock Hills, say, which are much handier for Stonehenge.

But if the bluestones lay on an existing route to and from somewhere important at the time then their significance can be better appreciated. What if, as in a previous post, traders coming from Ireland with copper to trade avoided the difficult rocky coast of the Pembrokeshire headland by taking a route across the headland from north to south. A route via the Nyfer and the Taf, and therefore passing the Bluestone site of Carn Menyn, seems like a good choice for this. It is interesting to note the concentration of Neolithic and Bronze age standing stones at the northern end of this route.

There is also a second possible route across the headland, further to the east, starting along the Teifi River then crossing a broad section of land to join the Tywi in the south. Again between these rivers there is a concentration of standing stones, although this route does not pass the Preseli Hills.

Perhaps somewhere in this data lies part of the reason why anybody would have bothered to collect the bluestones and move them all the way to Stonehenge. After all they were going that way anyway.

Well, it’s only a thought.

(P.S. by chance “pres” means copper in Welsh. There is no significant copper deposit in Pembrokeshire and this put me in mind of an exciting survival of some ancient name in Preseli. I suspect, however, that the word “pres”‘s etymology can be traced back to the English word “brass”, and so is much younger than Stonehenge. Oh well.)


Atkinson, R.J.C. 1956 Stonehenge, Penguin, pp256.

New References

Ixer, R.A. & Bevins, R.E. 2009 Missing Stonehenge circle did not come from Preselis (with retraction), British Archaeology 109.

Ixer, R.A. 2007 Waiting by the river: Stonehenge and the Severn Estuary. Abstract. Stone artefacts as material and sybolic markers in cultural landscapes. An international perspective.Implement Petrology Group Meeting. York. (not read but mentioned in comments below)

Bevins, R.E., Pearce, N.J.G. & Ixer, R.A. 2010 Stonehenge rhyolitic bluestone sources and the application of zircon chemistry as a new tool for provenancing rhyolitic lithics, Journal of Archaeological Science 38, 605-622. (only read abstract)

Stonehenge rocks definitely came from Wales, but how?

{ 16 comments… read them below or add one }

Richard Armstrong March 5, 2022 at 7:02 pm

Hi Ned, It’s nearly 12 years since you wrote these posts, so I very much doubt you will be checking the blog for replies! Nonetheless, I stumbled on these blog posts of yours on ‘tinternet – and I think they are fascinating. So thought I’d drop you a line.

I’m a total armchair amateur, but to amuse myself during lockdown, I started trying to trace ancient travel routes to/from Hillforts near my house in Devon. Firstly using OS maps and a bit of logic, then on foot. Hillfort Highways, as my blog would be called if I ever wrote it up. Now I’m muckin around doing the same for Stonehenge and the Bluestones. I might take a week off work this year, and try and drive+cycle+walk the route I think is the most logical. Something to do!

My “school” of thought on this whole megalithic transport/travel is basically summed up by: boats and ridgeways.
Coastal hopping (I’m a small boat coastal sailor, so its just logical to see the sea as the main highway for that time);
and then ridgeways to avoid dense, impractical, arduous, wet+boggy, up+down travel when on land.

(I deviate from you slightly on the use of inland river valleys. My experience from the hill forts in Devon is – ridgeways, always the high roads. The river valleys can be impenetrable – but you do have a good point about using them to go floating down!)

Your suggestions of (1) cutting the corner off St David’s, and (2) cutting the corner off Lands End – totally opened my eyes – jaw-drop a-ha moment. Of course! And when I started to look at the map, OMG! the evidence is there (well, it always is when you’re seeking to prove yourself right ha ha).

I’ll have to write it up all proper etc, but for now, check out this route – best on on Bing Maps (it has OS as a layer) and Strava for the elevation (I sound like a geek writing that, but it keeps me amused of an evening!)…

1. Waun Mawn to Haverfordwest (highest navigable point of the estuary, and closest southern port to quarry). From Rosebush (on the crest of the hill just above the Waun Maun circle) there is a lovely ridgeway route all the way down to the estuary at a gentle decline, apart from the very first 1.7km up to Rosebush on the crest of the hill above the circle – and that itself is a manageable and fairly gentle up-slope.

NB (Mike P-P’s claim that “because Waun Mawn is on the north-facing side of the Preselis, it rules out the sea-route” is slightly odd, if you see just how close to the top of the Rosebush ridgeline the MW circle is. Pop up to the top with yer stones, and then its a simple, 21.5km gentle gradient path all the way down to the port at HFW). Like you say, I think that would have been a VERY well-trodden path (see below on connection to Ireland/Wexford).

2. At Haverfordwest, jump on a boat / raft and use the outgoing tide to paddle through the sheltered estuary to Pembroke (the most SW point of the estuary and up a river creek)… when you get to the end of the Pembroke creek on the high tide, can you believe what you find…???….the start of a …

3. Ridgeway! I couldn’t believe my eyes. An almost dead-flat ancient highway, (with one short 50m elevation at the start over a 1km distance so not too bad) but otherwise flat, with a gentle down slope at the end to the sandy beach. It goes from Pembroke directly to an amazing natural harbour at nr Tenby/Caldey Island (which was probably still attached to the mainland in those days, creating a haven harbour of beautiful proportions)… thereby cutting the corner, avoiding the whole of the dangerous SW St David’s coastline just as you suggested.

From there, you could hop from beach to beach, using rafts and the tides (cutting across the Gower to miss out Worms Head – there is a very short, low lying route across top (N) of the Gower to Swansea (use the topographic map online to see it clearly) that suggests a fairly obvious solution) – beach+cove+inlet hopping all the way to Barry, where the natural stepping-stone islands of Flat Holm and Steep Holm take you over the estuary to Brean…Another (IMHO) major neolithic port/harbour sticking out where the high land of the Mendips meets the sea.

I won’t recount the route from Brean to Stonehenge here, but its essentially initially flat along beside the Mendips, a short ramp up at Burrington Gorge onto a flat ridgeway path that goes all the way (just to the north of) Shepton Mallet, then cuts across the valley to Warminster and up the slope at the back of Warminster to reach the ‘megalithic motorways’ that run along the chalk escarpments… before turning south along one of the N/S ridges to Larkhill (above Durrington and The Henge itself)…

On the West Wales / Ireland connection – I have a theory (of course!)…The high-point right beside Waun Mawn is a hill called Foel Eyry (check it out on YouTube for the 360-degree views – amazing!). From where you can see Ireland. And in particular, the Wexford area – Wexford is the closest sea crossing from Ireland to Fishguard. And in particular, there is only one hill in the Wexford area – Forth Mountain. From my desktop calculations on the map, Forth Mountain is (theoretically) visible from Waun Mawn. It is perfectly positioned to see Forth Mntn. What if… the two communities were linked by regular crossings and trade. What if … they were linked by family too? Why not commemorate your ancestors who died in Wales, but perhaps originated from Ireland in previous generations – by placing a stone circle at Waun Mawn from where you can actually see their ‘homeland’ at Forth Mountain? It’s a long way away (125km) – but fire beacons would be visible? I need to check that bit….

Anyway, thanks for the inspiration! Hope you get to read this one day…!! Cheers, Richard A


Edward Pegler March 6, 2022 at 5:55 pm

Dear Richard

Thank you for the excellently long comment. I’ll need to take it all in, and will get back to you when I have, but I take your point about ridgeways.

best wishes



Richard Armstrong October 10, 2022 at 10:22 pm

Hi Ned, just a quick update – I have finally had some time (Covid and other issues got in the way) to get out and explore my random theories by foot, bike and van. Have been trying to trace the trade route as you describe from West Wales / Ireland to Salisbury (Old Sarum). Between the (real) landscape and the (OS) map, I’m even more convinced about the ridgeways. So far, have tracked a surprisingly flat and high route from Brean Down to Salisbury, with one (convincing-to-me) theory about boats being navigable all the way to nearly Wells – Bleadney Port. This results in a land journey of just 42miles to Old Sarum, with Durrington just 6miles on from there. Total climbing on this route is about 450m, with about 80% of the journey downhill.

Even if I’m wildly wrong, or someone has already thought all this before, it is quite thrilling to discover places like Mere Down which leads onto the Great Ridge to Salisbury; or White Sheet Hill – standing like the pillars of Hercules alongside Long Knoll for anyone cresting the ridge-line at Gare Hill to see – like enormous signposts.

It keeps me interested at least!
Next trip I’m hopefully heading to West Wales to check out the ridgeway to Tenby.
One day, I’ll write it all down as a blog – working title “Megalithic Motorways”…

As for ‘why’ they moved the bluestones… have another idea about that, building on your point “they were going that way anyway”… One for another time.
Kind regards


Dr Mark Hylton March 1, 2011 at 2:09 pm

Hi Ned,

Funnily enough I’ve just come across the Bevins et al. article which reminded me of your post on this side. I’ve emailed you a copy of the full article, but for the purposes of this discussion they comment that:

“The correlation between the Stonehenge/Cursus and the Pont
Season samples verified by this study has important archaeological
implications, in particular around the debate as to how the bluestones
were transported to Stonehenge. What is important is that
Pont Saeson lies in an incised valley location, on the lower ground
to the north of the Mynydd Preseli. Previously, the most convincing
source locations for Stonehenge bluestones (the spotted dolerites)
were located on the highest ground in the Mynydd Preseli, and all
attention to transport mechanism had focussed on a southerly
transport direction, either by human or glacial means. The Pont
Saeson source could still be consistent with a southerly transport
direction by ice, but would open a debate about a northerly
transport route to the sea by humans”

Unfortunately while this research provides some more data on ‘direction’ it doesn’t really help with the big ‘why’ question!



Edward Pegler March 1, 2011 at 3:20 pm

Dear Mark

You know I start to feel like I haven’t got a clue, really. People could have taken a northern route but I wouldn’t fancy being in the boat that came round the Pembrokeshire headland. As for a land route it’s not impossible. After all, the first evidence of carts is from the fourth millennium BC so they probably had them in Britain. The stones aren’t that big. The only problem is whether they had anything that passed for a long distance road. Probably not.

Then of course it could be ice that transported them, at least onto land just east of the Severn vale. That would help. But I’d need to see more evidence of glaciation there (relics of erratics in the Buildings and the streams). Maybe it’s there.

If it’s not ice then it comes back to the archaeologists favourite word, ‘ritual’, because there’s no other reason I can think of why they would be at Stonehenge. Of course that ritual could be in different forms, say the yearly tribute of one stone a year from the home of the gods/fairies, or a mad king/priest’s stealing of stones to spite the gods/king of wales, etc etc etc.

I have a feeling that this debate will never end. Great, isn’t it.

love Ned


GCU In two minds February 19, 2011 at 3:56 pm

Milford Haven has been dead in the water for almost a decade now.
Read Waiting by the River Stonehenge and the Severn estuary. R.Ixer
It was a red herring due to misreading of the original sources. Totally old hat.


Edward Pegler February 19, 2011 at 10:36 pm

Dear Rob

Is there any chance that you could explain your comment a little more fully?

Best wishes


Edward Pegler February 20, 2011 at 7:33 pm

As an addition to this mini (and now, sadly, one sided) debate, see the articles written in “New References”, written by Richard Bevins, Nick Pearce and Rob Ixer (who wrote the above comment). As far as I can tell (and I await correction), these can be summarised as:

Some of the rhyolites (a kind of acid igneous rock formed in lava flows) found in the area around Stonehenge (including the Altar Stone) come from other areas of South Wales. (I admit that I knew of this evidence already). Contrary to initial speculation, there are no known sources of rhyolite from areas north of this, but this could, of course, change.

Loose boulders found in the Severn Sea/Estuary and Milford Haven are unlikely to be the result of movement by people and are probably erratics, placed there during the ?Anglian or Devensian glaciations. This seems entirely reasonable.

My personal opinion, at this stage at least, is that the peculiar concentration in Stonehenge of large boulders from the Preseli area makes a glacial origin for these boulders unlikely, although, of course, not impossible. See “bluestones and bell curves” for further discussion of this point.



Ed Watson November 15, 2010 at 5:16 pm

I do apologise, there was no offence meant but I clearly seem to have touched upon a nerve. I was merely trying to be suggestive of further possibilities, but i think it’s probably best I keep my thoughts to myself in future.
Great blog, keep up the excellent work.
Best wishes,


Edward Pegler November 16, 2010 at 4:24 pm

Dear Ed

I am so sorry if you thought I was offended. No, not at all. I was genuinely interested. Unfortunately, I’ve never mastered emoticons but probably should. If you want to post your idea here I’d love that.

My basic reason for setting up this blog is that I had ideas that none of my friends were interested in, so I’d get them off my chest by putting them on a blog. I find it really frustrating that there don’t seem to be that many interesting and thoughtful blogs on prehistory or ancient history out there (with notable exceptions such as Geoff Carter’s blog, Mathilda’s Anthropology, Diakenes). I just like the idea of ideas being out there with the possibility that someone might discuss them. Please please don’t keep your thoughts to yourself. That’s the last thing I’d ever want.

best regards



Ed Watson November 14, 2010 at 10:06 am

Ned, not sure if you’ve covered this in a previous posting as you have covered much ground on ancient trade routes, but I’m inclined to think the Romans targeted Anglesey because it was an important point on the gold route that passed through the island on its way from the Wicklow Hills in Ireland to the east of Britain and then across the North Sea to Europe.
Any thoughts on this?


Edward Pegler November 14, 2010 at 10:30 pm

Dear Ed

Interesting idea. Have you thought of posting it? No, I’ve not thought about your idea before and I’m probably out of my depth. Which way do you think trade went after reaching Anglesey from Ireland? Also I seem to remember there’s also a source of gold in the mountains of Mourne near Newry, used in the Bronze age. Is that relevant?

Would you say that the Romans targeted Wales/Britain in general to control the metals trade or perhaps did it for more traditional explanations like grain supplies, controlling troublesome influences on Gaul or pure egotistical Empire stretching?

Also, what and how extensive were the other sources of gold in the empire and is there any chemical analysis that suggests the use of Irish gold in the empire or northern Europe at the time?

Sorry for all these questions. I’m just curious. To be frank with no written sources for the period that I’ve covered I have much less evidence at my disposal than you do. If you’re not sure about the first century AD then I probably have no justification at all for any of my posts.

I look forward to hearing from you if you’d like to reply.

best regards



Edward Watson November 12, 2010 at 3:12 pm

I can’t agree with the comments on your previous post (Stonehenge, Avebury, Ross Island and the perils of the Cornish coast) with regard to Parys Mountain on Anglesey being one of the largest copper mines in Europe; indications are that mining activity during the Bronze Age barely scratched the surface, the scars we see today are evidence mainly from mining activity in the last couple of centuries, and has no bearing whatsoever on the Island being a centre of Druid power. For this we only have the words of one historian, Tacitus; significantly, there is no mention of Druids in any of the classical accounts covering the earlier British campaigns in Britain until Paulinus’ attack on Môn in 60AD. The myth of the British Druid, and their association with megalithic temples, is largely the creation of the 17th – 18th centuries, led by one William Stukeley.

Whereas the copper mine just along the North Wales coast from the Menai Strait at the Great Orme was probably the largest of its kind at that date. With tunnelling to a depth of 70 metres (230 feet) below ground with some five kilometres (3 miles) of passageways and galleries; similar in concept to later day deep shaft coal mines. It is thought that the combined mines in Ireland produced some 370 tons in total of finished copper, estimated from the cubic area of extracted ore against end product. By comparison the single source of the Great Orme copper mine is estimated to have produced between 175 to 238 tons.

Movement of such vast quantities of raw material must have opened up these trade routes around the coast and through navigable rivers as you suggest, evidenced by the Iron Age logboat found near Shardlow, Derbyshire on the course of the River Trent, in addition to the number of Bronze Age plank boats found around the coast of the British Isles, such as at Dover and Hull , the latter thought to be the oldest of the kind found in western Europe.

Interesting that during the Neolithic axe trade, material removed from source sites as ‘rough-outs’ was then moved to Causewayed Encosures, (as opposed to megalithic henge monuments), for final working. Oddly, most of these finished highly polished stone axes, which must have consumed many hours in labour to manufacture, were far too brittle to be used in a practical sense and must have had, (dare I say), a ritual significance.

With all this movement of materials on such a large scale, I find it incredible that some people remain of the opinion that moving four ton Bluestones, from Preseli to Wessex, whether via Milford Haven or down the Taf, would have been beyond the capabilities of Neolithic man.


Edward Pegler November 13, 2010 at 5:20 pm

Dear Ed

I think what you say is excellent. I suspect that the previous comments were simply a misremembering of the Great Orme mine. To my memory the Great Orme copper mine is the only major mine of significance in Wales from that early period, although I believe that there are other, later mines to the south of here. However as the evidence stands it is of later date than all but the last phases of Stonehenge. Unless other evidence comes to light it has to be put to one side for now.

I suspect that these mines are especially late compared to others in Europe as all of these areas (North Wales, Ireland, Cornwall) were heavily glaciated during the last ice ages. Weathered zones of easily extracted native copper or copper oxides would have been scraped into the sea by ice during these times so mining would not have commenced until people could see the worth of difficult to extract sulphide ores. This stands in stark contrast to Iran and Turkey for example where copper must have been collected long before real mining took place.

As for boats, I suppose there’s still a part of me that wants a Neolithic boat to be recovered from the sea bed, perhaps in the Severn or off the south coast of England in a place of net sediment deposition. The technology to transport animals from the continent (even if young) was there by 4,500 BC after all. It would be interesting to see what such a boat or different boats looked like.
As for the axe trade I think you’re spot on. The evidence of Henges such as Thornborough indicates (at least to me) strong river trade links across Britain from the region of the Lake District (a source of axes) across to the Humber estuary even during Henge times. I haven’t given too much thought to the distribution of the earlier causwayed enclosures but I believe their role must have been in some way similar for an earlier generation (see one of my earlier, perhaps more incautious posts on Avebury).

And as for ‘ritual significance’ of polished stone axes you must be right. They seem to have been the faberge eggs of their time. Almost completely useless but great to impress the little people with.

thanks for a very interesting reply, Ned


Ed Watson November 11, 2010 at 10:35 pm

Interesting article Edward. Milford Haven probably is the wrong place for disembarkment of bluestones but I don’t think the Nyfer or Taf, although useful navigational aids across country, are probably not as navigable as far as you surmise.

I can’t see the significance of Preseli to the copper trade; The modern Welsh word for copper is ‘copr’; I can’t find a translation as ‘pres’? According to A Dictionary of British Place-Names by A. D. Mills (2003), “the name Preseli Hills renders Welsh Mynydd Preseli, probably from Welsh prys, ‘wood’.”

We find a village called ‘Prees’, south of Whitchurch in Shropshire, said to be derived from ‘Prys’, (bush/thicket) as in a large portion of the land is heathy as shown in a number of nearby villages which share the name of ‘Prees’: Prees Village, Prees Green, Prees Heath and Prees Wood. All were formerly known as ‘Prys’ = Wood.

Just over a third (36%) of the total amount of lowland heathland in Wales occurs in Pembrokeshire, which indicates the meaning of the name as ‘heath’ would appear to be quite apt.

However, the name Preseli is clearly anglicised, so there must have been an ancient Welsh name, perhaps now lost.


Edward Pegler November 12, 2010 at 10:26 am

Dear Ed

Yes, I understand your reservations. I suspect that all of the rivers in the area are rocky bottomed for a fair bit of their length, which would make them difficult for transport. However, I suspect that they may offer the easiest paths through a once very difficult landscape.

Though admittedly only a Collins Gem dictionary, the translation of the word ‘Pres’ into English from Welsh gives “Brass; Bronze; Copper; Money”. This may well not be the best translation as copper comes in at number three. However, it’s also entirely irrelevant I suspect as the ancient word for copper in the third millennium is probably something unrecognisable and entirely unlike any modern word. So I’m with you all the way about the actual origins of Preseli. What I found in the dictionary just amused me.

As for the significance of Preseli for any copper trade, this is largely hypothetical, but based upon the opening of Ross Island and associated mines in Ireland around 2500BC. If any of that copper was exported to the continent, Pembrokeshire would have been a likely landfall and its difficult headland would possibly cause people to avoid the headland. A similar argument could, I suppose, be applied to the Great Orme copper mines around the turn of the second millennium BC. However, with the rise of Cornish tin mining I suspect that the whole pattern of trade may have changed by then.



Leave a Comment

Previous post:

Next post: