Search Our Site

Article Index


Tham Huai Klong Ngu - the Snake River System and Swallow Cave, Kanchanaburi Province, Thailand.

The string of rattling, third class coaches winds across the creaking, decrepit looking, wooden trestle bridge, clinging precariously to the limestone cliff, high above the Mae Nam Khwae Noi.  Below, on a great bend of the river, houseboats of bamboo with palm thatch drift lazily down with the brown current.  Dense tropical rain-forest stretches away to the distant, surrealistic shapes of tall, karst towers.  The rhythmic clatter slows and deepens as the train reaches the other branch of the river the Khwae Yai - and passes at walking pace between the forty year old steel girders of the famous bridge.  We are travelling along the Burma-Siam railway, built by POW's and coolies during the Second World War, and are crossing the Bridge on the River Kwai.

Seventy kilometres to the North West the waters of the eastern branch of the Kwai are held back by the Sri Nakharin Dam, creating a sixty kilometre long artificial lake, hemmed in by jungle clad hills.  Along the western branch much of the old railway line has been torn up, and where the old road to Burma used to cross mosquito infested swamplands, Thai Electricity have erected a second dam, flooding another huge area, including the road and the original town of Sangkhlaburi. A brand new road, a masterpiece of engineering built in spite of torrential monsoon downpours, skirts the lake, twisting and climbing through extraordinarily rugged limestone and lush forest, to reach the border with Karenni and Mhong occupied Burma at Phra Chedi Sam Ong - the Three pagodas Pass, two hundred and forty kilometres from the bridge.

In between the two lakes is a sixty kilometre wide plateau, much of it limestone, lying at a height of around six hundred metres, with several karst towers rising to eight hundred metres and more.  Last year a small French expedition mapped several hundred metres of a huge cave which had previously been explored by Germans working at a nearby lead mine. We knew nothing of the Germans, but we did have a map produced by the French, and a few, flimsy details, including a mention of the mine.  The map was simple - it had a major series of gorges, deep dolines and large towers marked on it (one with a name, albeit incorrect) and it showed water, sinks, resurgences and karst windows (this last is where subterranean water can be seen crossing the base of deep shafts).  The map had no other details - no roads, no contours, not even a location.  John Dunkley, the Aussie caver who instigated our visit to Thailand, had sketched in probable road locations onto an old map of the area but the site of the cave was rather vague.  Perhaps someone at the mine could tell us more.

John, Jane and I met up in Kanchanaburi, the town near the River Khwae Bridge.  The tourist office personnel were helpful: half way up the eastern lake, on the western shore, a little national park has been set up to cater for (rich) visitors to some waterfalls.  A public bus goes as far as the Sri Nakharin Dam, and from there we seemed to have a choice: a fledgling tourist business ran a boat up the lake to a hotel for the night, and then across the lake to the national park, for which we could pay a small fortune; or we could hire a pick-up for around eighteen pounds a day and attempt to reach the park, and thence the mine, via a rough, dirt road to the west following the latter route.  We opted to hitch instead.


Traffic was somewhat thin from the dam to the park.  Only four vehicles used that road all day.  Fortunately we got a lift in each one.  When a two seater Willy's jeep came by, with three people in it and the whole thing overflowing with provisions, gallons of diesel and a tractor tyre, we could not expect any more than a friendly wave, but this was Thailand.  Somehow we got the three of us on too, plus our huge packs.

The road was deeply rutted from recent rains - this was the end of the dry season and parts of it were hair-raisingly steep.  Large areas of previously virgin jungle had been recently burned and cleared, and poor farmers from the arid and infertile north-east had moved in and were making a go at some ephemeral agriculture.  The land lasts for two or three years, during which time the nutrients are used up and the soil eroded. The farmers have to move on and the jungle does not return.  It seemed that this road only existed for the farmers.

The national park ranger took good care of us, letting us sleep in the park headquarters, providing us with an excellent, very cheap meal, and lots of information. Unfortunately he knew of no caves. The park's waterfalls descend steeply as a whole series of dramatic, travertine cascades, and we thought it quite likely that the stream emerged from a cave further up the edge of the plateau. However, after a perfunctory recce we contented ourselves with a wander down the well trodden tourist path, and a swim in the deep, blue plunge-pools under the cool, green canopy of the forest. That night we shared the park H.Q. with a million flying ants, beetles, moths, roaches and mosquitoes, and two exceedingly fat, foot-long geckos who were happy with only the largest and tastiest insects.

In the morning the Ranger drove us a short distance north, and thence down to the lake shore. Between the white, sun-bleached stumps of drowned trees and the weedy, gravel slopes of the shore was moored a large, steel ferry-boat.  So this was the route the lead mine trucks used, and from here up to the mine is a fast well graded dirt road.  After another good meal, courtesy of the ferry captain's family, the first truck of the day was brought over from the distant, eastern shore, and we climbed onto the back.  It was already full with equipment and stores for a second mine, plus a couple of dozen laughing and joking locals.  The truck roared away up the stony, dusty track, with us clinging precariously atop the piles of sacks and girders and boxes of provisions, dodging the overhanging branches that tried to pluck us from our perches.

Forty kilometres on we were dropped off at a junction where the truck continued to Kletee Mine.  Our destination, Song Toh Mine, lay just five kilometres away.  We sheltered from a rain shower and watched the massive, isolated karst towers slowly disappear into the murk, then emerge once more, washed and gleaming in the sun. The bigger towers can be a couple of hundred metres high, and quite long, tending to take the form of humped ridges.  The old geological maps suggest that the towers are of Permian or Triassic limestone, while Ordovician carbonates lie beneath, as a plateau.  Although there is a vast difference in the ages of the two rock types, stratigraphically they are the same, and there is no reasonable explanation why the younger rocks should have been formed into towers.  The walls of many of the towers are steep, even overhanging in places, and generally vegetation free, while the summits are a tangled mass of trees, creepers and roots concealing viciously sharp spikes of stone. A few cave entrances are usually visible part way up the towers, but often the longer caves are at the base of the hills, and are thoroughly hidden by the thick undergrowth.  We observed patches of mist, maybe from hidden holes in the forest, and pondered caverns measureless.

The rain died away and, after a short wander beside the dripping forest and among smaller karst towers, and a lift in a pick up, we reached the mine.  What a contrast: only a few minutes down the track was thick green, barely penetrable jungle and wild, jagged castles of stone; here, in the middle of the wilderness, was a town of three thousand people, complete with street lights and suburban type gardens, shop, hospital, offices, all the buildings and paraphernalia required to run the most modern mine on mainland South East Asia, and it is not even on the map!

The Germans who run the mine immediately made us very welcome.  Dr. Gerdt Pedall, the geologist for the company, was particularly interested. His hobby back home is exploring old mines, but there's rather a dearth of them here so, over a number of years, he has investigated many caves.  Several of these are of archaeological value, being sites of ancient human habitation and containing remains of wooden coffins (or, perhaps, water tanks) and potsherds.  However, his greatest caving achievement must surely be the explorations of the Snake River and its associated caves.

Accommodation was provided for us at the mine and, over a superb German supper and Kloster Bier by the litre, it was arranged that we should visit Swallow Cave the following day.  The evening was rounded off with Mae Khong (Thai whisky).


The headwaters of the Snake River (Huai Khlong Ngu) drain in excess of two hundred square kilometres, and the majority of the waters become a single river deeply incised into the older limestones, flowing roughly southwards. Most of the significant ridges and valleys in this region trend just east of south.  Two dolines to the east may also drain into the Snake River, although the likely confluence is not yet known. The main river runs through a deep gorge, walled in by huge cliffs of towering white stone reaching up to the base of a wide, shallow valley.  After several kilometres the canyon stops abruptly as the waters vanish underground, to reappear briefly two and a half kilometres further south at Swallow Cave.

Gerdt had sorted out a guide for us, laid on a four wheel drive vehicle plus driver, and drawn us a remarkably detailed, accurate plan of the entrance region of Swallow Cave and the other caves down river. Initially we drove back to the Song Toh - Kletee junction (where we had sheltered from the rain) and then headed up the Kletee road, still on a good gravel surface which has to suffer the pounding of way-overloaded ore-trucks, each carrying twenty two tons of washed and ground galena.  A kilometre to the north a track led into the forest, eastwards, on sun-hardened red laterite mud.  This gradually deteriorated until we were dodging trees, and bouncing over steepening ground with lumps of limestone protruding wheel-jarringly out of the laterite, a kilometre further on.  From here we would have to walk.

Great clumps of bamboo, up to twenty five metres high, towered overhead, and huge, multi-rooted trees swept up to support a vast sunshade of dappled green foliage. Strange flutings of birds, seldom seen, echoed through the forest, and a large squirrel raced nimbly away across the topmost branches.  A small, grey viper wriggled hastily out of our path, while a silent moth, the colour of dead bamboo leaves, simply disguised itself as another piece of forest litter.  Occasionally we glimpsed tall, but narrow karst towers through gaps in the greenery, and to either side of us the ground dropped away to tree-filled dolines, each inviting a more thorough investigation.

After little more than half an hour's hot walking the valley of the Snake River appeared ahead and below us.  We could view across miles of tree tops a wide hollow with no sign of the canyon' or river at the bottom.  The descent was steeper than it appeared, starting with a clamber among rocks, and followed by a laterite slope, still slick from yesterday's rain.  Evidently elephants come this way as we found their tracks, even on the steepest slopes.

Suddenly we dropped into a deep amphitheatre, carved from the rock and linking to the gorge.  The pungent odour of guano filled the air, and hundreds of swifts could be seen circling and swooping, far above the canyon walls, which gleamed white in the sunlight.  Passing through a short cave along one wall of the gorge we emerged onto a wide, flat, sandy ledge on the threshold of a vast portal.  The river, knee-deep and twenty metres wide, filled the floor. From the ledges on either side the cave walls rose straight up for sixty to eighty metres to support a level roof bedraggled with massive stalactites. Above this there appeared to be very little solid rock between the cave and the jungle.

A group of Thais were camped at the cave to collect guano by the sackful, and then drag it, laboriously, up the hill to the road head, an hour’s tough walk away.  At the moment they were relaxing, fishing by that age-old method - a net across the river and an ounce or two of bang upstream.  They gazed at us, unspeaking, as we donned our 'caving gear'.  The only equipment that John and the guide had was CEAG acid cells supplied by the mine. Jane and I caved in T-shirts, shorts and lightweight walking boots.  We had helmets, stinkies (carbide gobblers are not for lightweight trips), Petzl zooms and little Tekna-lites.  Additionally I carried a polythene wine bag (empty, sad to say).  Blown up and stuffed up my T-shirt this served as excellent flotation for me, a natural sinker.

Initially we all tried to stay dry.  After all, I was wearing a rucksack containing an expensive camera.  Even wading across the underground rivers in this region is not straightforward: masses of organic debris accumulate on all the rock surfaces, and underwater this becomes jelly-like and incredibly slippery.  After a couple of crossings and a very slimy traverse above deep water, we had reached the end of the twilight zone, about three hundred metres in.  I dumped my sack among some stal and hoped that the Thai guano collectors were either honest or afraid of the dark.  The next section was most easily passed by swimming, crossing to a long bank of stalagmited rock and big gours.  John found an awkward but dry route along the opposite wall and our guide, determined not to get wet above his waist, followed.  The next section was definitely for swimmers only, and Jane checked it out as far as dry land.  John and the guide decided that they had seen enough, so we two continued alone.

The passage remained wide and high, the roof often being beyond the range of our lights.  After frequent immersions the draught began to chill us, and we were glad to find long, gravel banks where we could put on a bit of speed and get warm again.  Sometimes the river ran deep and swift in a confined channel and we had some awkward climbs to negotiate in order to avoid the waters and their dangerous currents. Occasionally it was deep water over the whole width of the passage and we were forced to swim.  In one place the river did its best to sweep me into a sump beneath an enormous fallen boulder because I had been foolish enough to attempt a crossing in the wrong place.  With awe and muttered expletives we noted the flood debris - whole tree trunks and huge branches throughout, jammed into crevices up to ten metres above our heads - and the wet season was just beginning.

The wild life in the cave was particularly abundant.  We were constantly pestered by millions of small, white flies which were attracted to the light of our carbide flames, and died there like a steady waterfall in front of our noses.  So dense were they that it was difficult to see through the cloud, and we would have been better off at times with hand held torches.  The screaming swifts at the entrance were replaced by various species of bats further in.  On the gravel banks we came across wetas, crickets and long, brown millipedes, and among the rocks lurked centipedes and scorpions.  The centipedes were about ten centimetres long, with yellow and brown striped bodies and long, spidery legs. Thankfully they scuttled into hiding as soon as our lights disturbed them.  Not so the scorpions, who sat tight, usually right on a crucial handhold.  Pale white fish swam in the pools and crayfish stalked along the bottom pretending to be stones. It is highly likely that some of these creatures will be new to science - of three fish collected in the north, one was the first found in Thailand, and one was a totally new species.  A biological collection from caves in this area is bound to be worthwhile.


More than a kilometre into the cave two small inlets emerged from loose boulders at floor level by one wall - perhaps these originate in the large dolines to the north-east.  There was no obvious passage here, and indeed we saw no passages leading off the main one anywhere in the cave.  However, our lights were not all that brilliant, so who knows?

There is little stal throughout the cave.  There is the stal bank near the start, already mentioned, and part way in is a wall of deep, cup shaped gours, dry at present but, no doubt, full in the wet season. Fifteen hundred metres in a tall, lonely stalagmite dominates the passage, and beyond this the passage continues, big as before, with pebble banks beside the meandering river, interspersed with sections of more turbulent, deep water.  After two kilometres there is a karst window, a sort of skylight where the cave has been unroofed for a short distance, and thence it is but a few hundred yards to the inflow entrance.

Having made our way back to the resurgence and joined the others we set off down river to see where it sinks once again.  The waters meandered in deep pools or ran in rapids over a coarse sand and gravel floor, with huge cliffs to one side.  We crossed over, balancing precariously on the trunk of a fallen tree, and soon crossed back where there were shallows.  After a couple of hundred metres there were cliffs on both sides, the bedding clearly showing a gentle dip to the south.  Flood waters had scoured out big, elliptical scallops in the rock.  Swallow Cave was only just out of sight behind us in, the trees, as we rounded a bend to see the next cave in the system.  While we slipped and slithered among the big boulders of the entrance rockfall our guide excelled himself, dancing from rock to rock-and over the stream, first in flip-flops and then in bare feet.  The river flowed fast and deep between wide, sloping shelves and in a vadose canyon, so we stayed well above it on the ledges.  As sunlight appeared through a narrow karst window marking the far end of this cave we reached an impressive array of deep, cup shaped gours, more massive and extensive than those in Swallow Cave, like tiers of gigantic swallow’s nests.  One of these made a superb, pulpit-like stance from which to view the next cave entrance.

In the section down-river from Swallow Cave there are three such fragments of cave, interspersed by canyon, or unroofed cave, and then the river flows underground yet again, now for the fifth time.  It drops down a short waterfall and appears to sump immediately.  This has not yet been explored - these highly flood prone tropical systems are safest explored upstream.  Some two kilometres down-valley it resurges again, but the cave is quite different in nature from those up-river.  The roof is wide and low over deep water, and progress is entirely by swimming against the strong current.  Gerdt explored this solo on a previous occasion, reaching his own, psychological barrier after about four hundred metres.  The cave was observed to continue in the same fashion.

The map shows the river continuing southwards for a few more kilometres, still entrenched within a deep valley or canyon, and then turning abruptly to the east.  It seems to go underground for the last time towards the end of its eastward course.  The map indicates at least one and a half kilometres of cave, while the contours suggest that the resurgence would be close to lake level at the head of a long, deep and narrow inlet of the Sri Nakharin Lake.  None of this has been explored, and the best access to any possible cave here is clearly by boat across the lake, and then up the inlet.  Hopefully the cave entrance will be above the water.

There is to be a combined French-Australian expedition to the Snake River later on this year.  Although there seems to be little potential for long cave in the region, there is certainly more passage to be found, and it is quite likely that this will be via resurgences within the Snake River canyon. The French have surveyed part of Swallow Cave - this needs completing and the other caves need mapping.  There is at least a couple of kilometres of cave to be explored down-river.  There are several dozen other sites already known, unrelated to the Snake, but close by, and some of these are of archaeological significance.  Obviously there is much work to be done.

There can be no doubt at all that the area deserves the status of a national park, and presenting a case for this would not be difficult.  However, the Thai vision of such set-ups is that they are primarily to attract visitors, and must be altered and managed to cater for this.  Here is a unique, wild and dramatic, true karst landscape.  We can only hope that the Thai authorities do not realize its tourist potential before they come to understand the true meaning of conservation.

Graham Wilton-Jones 20 / 6 / 1987 Kuwait.