The Nantymwyn Mine is geographically divided into two sites, lying above and below the village of Llandovery, Dyfed. The mine was worked mainly by adits driven into the mountain known as Pen Cerrig Mwyn and, on the face of it, being the largest metal mine in South Wales, should be a mine explorer's paradise. Unfortunately, however, at the end of its last working period, in 1932, all of the entrances were blown in. Spoil was later removed for road building and the surrounding area forested, with a forestry road cutting through the site. This, together with massive falls of ground, means that today only a small part of the underground workings are accessible.

The lower workings were served by the Deep Adit of the mine (known as the Deep Boat Level) and the collapsed entrance can be seen adjacent to a row of cottages known locally as Dray Cottages, an obvious corruption of the word "Dry". About 60 metres above the entrance there is a collapsed shaft but access may also be possible from a shaft by the roadside, opposite a house called Erw Hwch (NGR SN780440). As this shaft has been used for dumping rubbish in the past, I felt disinclined to go down it.

The upper workings are centred around the engine house at Angred Shaft and it used to be possible to drive up there in a car. The gate on the mine road at Nant y Bai, however, is now kept locked so the nearest vehicle access is at the top of Pannau Street, north of the village. Here, a gate gives access to a parking area on the spoil from Level Pannau, the lowest drive on Roderick Lode. This is the final level reached on a descent for which the modern explorer will need about 400ft of rope - more on this later.

Underground Remains

I have been able to discover 4 open adits in this area :-

a) Just in the forest adjacent to Angred Shaft (NGR SN785445). It is about neck deep in water, 20 metres to a collapse, so this can be immediately discounted.

b) Above the top spoil heap above Angred Shaft (NGR SN790445). It is a long crosscut, upper thigh deep in water, leading to an area of workings.

c) The level known to cavers as Level "Cadno" (NGR SN788441). This is just below the forestry road and is easily reached from the "Car Park" by following the path until a gate is reached. Then follow a wire fence on the right and, when you see spoil above, go up in a straight line. It is situated at the end of a cutting above a pile of stones.

A short distance inside, you reach a shaft in the floor which can be crossed on two springy planks. The level continues to a collapse, passing another shaft in the floor and a blocked rise. Back at the shaft with planks, there is a choice of either a straight SRT pitch of about 80ft or a short 6ft pitch down some rotten ladder to a ledge. From the latter, you go round a dog leg incline to another ladder of about 12ft with many missing rungs. Here I have left in situ a hanger for a belay.

This pitch gives access to part of Lewis's Level and a large stope is reached to the north. Looking down, water can be seen about 80-90ft below and Lewis's Level can be seen to continue beyond. The level here must once have been timber floored but this has now all gone down, together with a lot of rock. This dissuaded me from an SRT descent, although there are some belay points left there by others. Heading south leads to the second pitch and, beyond this, a collapse can be seen and there is a length of "vintage" caving ladder with wooden rungs. As there was no proper belay point, I have left two hangers in situ. The pitch itself is about 120ft, possibly free-climbable but with two short vertical bits. Some old ladder here should be avoided.

This gives access to part of Maescarhyg Level and heading south leads to a stope in the floor. Flooded workings can be seen about 20ft below this, the water overflowing into an adjacent shaft which must connect with the level below. Beyond the stope, a loose collapse can be seen with what may be a level. Heading north, you reach a shaft which is almost the full width of the floor. I have built a bridge out of some old ladder uprights from the last pitch, so it can now be passed on the left. Beyond, are a few more yards of level ending in a collapse but with a few minor artefacts. There was a single in situ hanger here, set back along the passage for some reason. I have fitted another at the pitchhead and recommend the use of both.

This pitch is about 30ft to a dog leg, where there was a dangerous mass of rotten ladder, old timber and rock which I have been able to clear. This was done by containing my rope in a tackle bag hanging from my harness, drawing rope out as I abseiled down and clearing the way as I went. The dog leg lead to another 30-40ft of vertical descent with some nearly sound ladder in it.

The next level reached is a short section of Angred Level and heading north leads to a collapse. Heading south leads to the final pitch. Although there was a length of scaffold pole wedged across the level as a belay, I have left two hangers in situ at the pitch head. This pitch is bad enough as it is without "foul rope" condition on the lip of the shaft. Beyond the shaft, the level can be seen to continue and I may return some day and put in a traverse line or build another bridge. As to the pitch itself, be advised that there is a near vertical section of about 40-50ft to a ledge, then about 30ft free hanging. On this section DO NOT look up (your helmet is stronger than your face and eyes).

Pannau Level is the next one reached and, heading south, you reach a small chamber before a collapse - this area is thigh deep in water. To the north is a well decorated passage with, believe it or not, secondary copper formations. This leads to the base of a stope where plankways and platforms can be seen above, possibly climbable with the proper tackle. It is not the same stope as seen from Lewis's Level above.

One I have called for the purposes of this article "Wet Level" (NGR SN789441). The entrance is chin deep in water (be advised that I stand about 5ft 10ins) and at about the same elevation as Level Cadno about 100 metres further to the east. Walk east along the forestry road and look for a flat area below through the trees.

The first 30 metres are very wet, with less than 1ft of air space in places, but after climbing over a fall it is down to about 2ft. The passage continues and eventually ends at a complete collapse with an air pipe sticking out. Before this is reached, however, there is a passage to the left. Unlike the other workings described, these passages from now on are all low and narrow and the one now entered appears to have been driven without the use of explosives.

On the floor, many clog prints can be seen and some are small enough to be those of children. There are a few minor artefacts. The passage eventually reaches a long east-west crosscut, which soon reaches a small narrow stope to the east. This stope is very narrow, little more than 1ft wide, and at one point higher workings can be seen which a good climber could possibly reach. Heading west along the crosscut, a complete collapse is reached after about 100 metres. I started digging this on Boxing Day 1993 and a strong draught can now be felt. In my opinion this point is on Roderick Lode, somewhere inbye of the fall in Level Cadno, it certainly appears so from my survey.

It was interesting to note that this crosscut had been driven from the direction of the fall, ie the west. The water runs west and the shotholes point east. If my theory is correct, to continue the dig straight ahead would eventually strike the solid rock of the hanging wall, so the way on would be either right or left. A possible route for future exploration.

There are two piles of backfill at the easterly end of this crosscut. This would indicate that at some time this passage had fallen out of use, the dumped material probably being from fossicking operations in the late 1920s. It is also probable that my Wet Level is in fact the level marked as "Old Level" on some of the old plans, being a drive on Old Vein.

extracted from "Nantymwyn Lead & Zinc Mine - a Modern Exploration", Roy Fellows, SCMC Journal No.2



Last revised: 21 April 1998