Hoia Baciu Forest: The Twisted Trees Were Made by Hand

Halfway up the trunk of an old hornbeam in Hoia Baciu Forest, the wood swells into a fist.

Above the swelling, five or six stems rise together, each about as thick as an arm, splaying outward before they turn for the light. Below it the trunk is unremarkable. The fist itself is grey and knotted and layered over with healed tissue, and it sits at roughly the height of a person’s shoulder.

The forest is full of trees like this. Hornbeam, lime and maple, mostly, and the older they are the stranger they look, until a mature one and a young one of the same species barely read as relatives. Some of them corkscrew. Some fork at right angles and then fork again. Some are bent almost to the ground and still growing, at an angle no wind on this hill has ever managed.

They stand on a low ridge a few kilometres west of the centre of Cluj-Napoca, across about three hundred hectares of hillside, between the Someșul Mic on one side and the Nadăș on the other. On a Saturday the paths carry joggers, dog walkers, families and teenagers, most of whom have come out from a city whose 2021 census population was 286,598 for an hour of air, and who go home again without incident.

It is also, according to a great deal of the internet, the most haunted forest in the world.

The claims stack up quickly. A shepherd who walked in with two hundred sheep and was never seen again, leaving no bones, and giving the place its name. A five-year-old girl who came back out five years later in the same clothes with no memory of the interval. A clearing at the heart of the wood, close to a perfect circle, where nothing will grow. Rashes, nausea, headaches, sudden fevers, batteries that empty themselves, cameras that record shapes nobody standing there could see. Voices. A woman in white. And on the afternoon of 18 August 1968 a military technician from Cluj named Emil Barnea photographed something bright and silent above the trees, and the pictures went around the world and are still going.

Every item on that list points away from the ground. Upward at the sky, sideways into another dimension, outward towards a magnetic field that has never been located twice in the same place.

The hornbeam is not pointing anywhere. It is standing where it grew, holding the shape it was given, and it has held that shape for longer than anyone now photographing it has been alive. Something put the shape there. The better question in Hoia Baciu Forest may not be what came down into the clearing, but who used to stand on this slope with a blade in their hand, and when exactly they stopped coming.

The Place: What Hoia Baciu Forest Actually Is

The hill sits at 46.77 north, 23.52 east, immediately west of the city. Its southern border runs along a ridge oriented east to west, and stops short of the steep face that drops away to the Someșul Mic. To the north the ground falls more gently towards the Nadăș. On the eastern side the hill has been cut clean through by an artificial valley called Tăietura Turcului, the Turk’s Cut, which now carries a road. Westward the wood runs out onto the north-eastern slope of Dealul Melcilor and meets the Mujai forest beyond it. Along the north-eastern boundary, Valea Lungă has worn its way down through Eocene limestone into a small gorge with lopsided walls, the Cheile Baciului, and upstream of the gorge there is a lake.

Three square kilometres, give or take. About the area of a mid-sized city park, which in practical terms is what it is.

The canopy is hornbeam, lime and maple, with oak on the drier margins. In the Bongar valley along the southern edge there is a stand of downy oak that botanists treat as a relic of steppe woodland and value accordingly. The ground is uneven and heavily tracked. Signposted trails cross it, a mountain bike park occupies part of it, and there are cleared areas used for paintball, airsoft and archery. Against the eastern edge stands the open-air section of the Ethnographic Museum of Transylvania, where reconstructed wooden houses and a wooden church from villages across the region have been reassembled among the trees.

Near the south-western part of the wood the canopy stops. The gap that people come to see is called Poiana Rotundă, the Round Clearing, and its dimensions are reported so inconsistently that the reporting is itself informative. Thirty metres across in one account, two thousand square metres in another, which would put it nearer fifty. One widely shared English source gives it a diameter of two hundred and ninety metres, which is roughly the length of three football pitches and can be dismissed by anyone who has stood in it.

What is there is a grassy oval with a distinct edge, low scrub creeping in a few metres from the treeline, and an unpaved track running across it. Grass grows. Nothing else much does. Standing in the middle of it in daylight, the thing that is genuinely odd is how sharply the trees stop, and how flat and open the ground is on a hillside that is neither.

You can walk to it from the city in a morning, and in daylight most people do. The night visits, which are how Hoia Baciu Forest now earns its living, end here.

The History of Hoia Baciu Forest

The oldest thing on this hill is not Hoia Baciu Forest.

Somewhere on its northern side, in the commune of Baciu that gives the wood half its name, lies Gura Baciului, the earliest Neolithic settlement so far identified anywhere in Romania. People were living here by the early sixth millennium BC, during the first farming settlement of Transylvania. Archaeologists associate Gura Baciului with the Early Neolithic Starčevo-Criș sequence, and specialists still describe it as the earliest known Neolithic settlement in Romania. The protected site carries the national monument code CJ-I-s-A-06949. Excavations by Nicolae Vlassa and later Gheorghe Lazarovici and Zoia Maxim uncovered houses or huts, pits, graves, pottery, stone tools, ornaments and evidence of crop and livestock farming. Lazarovici and Maxim published their major site monograph in 1995.

The published site catalogue records substantial damage and loss from erosion and quarrying.

Between the Neolithic and the twentieth century the record thins into the ordinary. The hill was worked. The village of Baciu sits at its foot. What that work looked like is still legible in the trees, though it takes a specific kind of eye to read it, and in February 2021 the ecologist Tibor Hartel, then in the Department of Biology and Ecology at Babeș-Bolyai University, provided one to a local newspaper. He is not a passing commentator on this subject. He is the co-editor of the standard scholarly volume on European wood-pastures, and his research career has been built on Romanian grazed woodland and old trees.

Hornbeam, lime and maple all coppice and pollard readily. Cut a hornbeam off at the base and it will throw up a crown of new stems from the stool. Cut it at shoulder or head height instead, above the reach of grazing animals, and it will do the same thing from a boll that thickens and knots with every return. Both were done here, repeatedly, over generations, and both are visible now in mature trees whose shapes bear no resemblance to young trees of the same species. Hartel’s phrase for them is heritage elements of a silviculture that has already vanished. He also lists the ordinary contributing factors, slope and wind and competition for light and the way water moves from one side of a root system to the opposite side of the crown, and is careful to say that the situation is complicated. But the primary cause is a blade, applied on a schedule, by people whose names nobody wrote down.

Under the communist government the hill became municipal. Cluj residents came out for the May Day and 23 August celebrations, and picnicked here on ordinary weekends.

It was in that period that a man named Alexandru Sift began going into the wood on his own with a camera. Accounts put the start of it in the 1950s or in 1960, and they are not reconcilable. Romanian sources call him a profesor de biologie, which in ordinary Romanian usage means a biology schoolteacher, and English retellings have consistently rendered this as biologist, which is a promotion the record does not clearly support and which matters, because the authority of the founding photographs partly rests on it. He reported being followed by shadows. He photographed what he described as umbrellas, and said that when the film came back it held forms and lights that nobody present had seen. He kept going for the rest of his life and accumulated a large archive.

Then, on 18 August 1968, the forest went international. Emil Barnea, a military technician from Cluj, was spending the afternoon in the clearing with three other people, one of whom, Zamfira Matea, allowed her name to be used afterwards and two of whom did not. He walked off to gather firewood, saw a bright silent object above the treeline, and got two frames of it. Romanian accounts have the photographs reaching Ion Hobana and Florin Gheorghiță, the leading Romanian ufologists of the day, and from there the wider world. They are still the most reproduced images ever taken here, and they were taken in a country where the government regarded belief in the paranormal as, at best, a symptom.

Sift died in 1993. His photographic archive disappeared within days. What survives of it survives because his friend Adrian Pătruț, a professor of chemistry at Babeș-Bolyai, published the remnants two years later in a book cataloguing the forest’s phenomena, and then spent decades continuing the work with magnetometers and radiation counters and soil samples. Pătruț became the authority the television crews came to see, and he gave them a formulation that has aged better than most: we do not know what these phenomena are, but we have enough data to say what they are not.

Pătruț’s academic work is not peripheral to the story. He has published internationally on accelerator mass spectrometry radiocarbon dating of very old trees, including African baobabs and Romanian specimens. That makes the contrast between his work on the age of trees and his long investigation of the forest’s aerial and magnetic phenomena especially striking.

Listen to The Dark Atlas on Spotify. The Hoia Baciu Forest companion episode will appear in the feed when released.

Destination Truth came to Hoia Baciu Forest, and Ghost Adventures, and after them the tour operators. The other thing that came was the chainsaws. In 2020 the Cluj athlete Vlad Pop twice filmed men felling trees in the wood and said he was threatened while documenting it. Local reporting treated the incidents as part of a wider problem of illegal cutting and dumping around the forest.

The felling is the small version of the problem. In 2021 local reporting connected a newly gravelled road through the wood with a residential project of more than seven hundred apartments at the forest edge. A petition gathered roughly ten thousand signatures, former Cluj MP Adrian Dohotaru asked the prefect to suspend the permit, and several hundred people joined a protest against the development. The developer’s representatives said the project had received the required approvals. Academics and campaigners also pressed for wider protection of the forest-pasture landscape. The road was later closed to through traffic, but the underlying argument over access, building and protection did not end with the barrier.

The Legend

Romanians did not originally tell this as a ghost story, and the shape of the older version matters.

What people in Baciu, Florești and Cluj said about the hill, before any of it reached print, was that it was a cursed wood and that you did not go in. Not because a specific dead person walked there, but because of what the place did to whoever entered it. The formulation that recurs is that the forest finds your worst fear and puts it in front of you. That is a property of the ground rather than an occupant of it, and it belongs to a Romanian folk category that does not map onto the English word ghost. Where a named revenant does appear in the older tellings it is a strigoi, which is a specific thing with its own rules, a dead person who does not stay dead and returns to the living they knew, not a free-floating apparition available to a tourist with a camera.

The shepherd came out of that older layer, or is assumed to have. The story is repeated in essentially identical wording by the city tourism board, by every listicle, and by the guides. The detail that no remains were ever found, neither the man’s nor the animals’, is always included, and functions as the proof.

The girl belongs to a later stratum and travels almost exclusively in English. No police report has ever been produced, no newspaper account, no name, no date, no family. Even the most enthusiastic retellings disagree on whether she came back aged five or aged ten, which is the tell of a story that has never had a source to be checked against.

The physical complaints are the most consistently reported material in the whole record and the most consistently vague: nausea, headache, anxiety, a rash, a feeling of being watched, a sense that more time has passed than the watch says. Devices behave oddly, batteries drain, recordings pick up frequencies that were not there. Photographs come back with faces in them.

The clearing absorbed all of it. Barnea’s object was above Poiana Rotundă, so the clearing became a landing site, and the absence of trees became scorched ground, and the scorched ground became a portal. This is where the tours end, at night, in the dark, with torches off. The internationally marketed names arrived to match: the Bermuda Triangle of Transylvania, Romania’s Area 51, the most haunted forest in the world.

Note what the entire structure has in common. Every element of it explains Hoia Baciu Forest by reference to something that arrived from outside it and then left.

What Can Be Explained at Hoia Baciu Forest?

Take the claims in the order the tour takes them.

The trees are the easiest, because a named academic at the city’s own university has already answered the question and his answer has simply not travelled. Hartel’s account, given in the History above, is not a hypothesis waiting on funding. It is standard woodland ecology applied to three species that respond to cutting in exactly this way, and the same shapes can be found in worked woods across Transylvania, across Britain, across most of temperate Europe. The reason they read as anomalous in Hoia Baciu Forest is that the people who look at them arrived from countries where that kind of woodmanship ended within living memory, or a generation before it, and where the surviving pollards are labelled and railed off in heritage parks. The shape is not rare. The literacy is.

The clearing is harder, and worth being honest about. Soil sampling has been carried out here repeatedly since the 1970s, and the reported result is always the same: nothing found that accounts for the absence of trees. That negative finding is what the paranormal reading is built on, and it should be treated carefully for two separate reasons. The first is that the analyses themselves are difficult to locate in any published, checkable form, so what circulates is a summary of a result rather than a result. The second is that the claim being explained is inflated to begin with. Grass grows in Poiana Rotundă. Scrub encroaches from the edges. A track runs across it, which means vehicles, which means compaction. Sharp-edged permanent clearings in worked hill woodland are among the least surprising features in European landscape history, because that is what a threshing floor, a stock pen, a charcoal hearth or a sheepfold leaves behind: ground that was kept open deliberately, for long enough and hard enough that the wood has not reclaimed it. Nobody has excavated it. The absence of an explanation here is not the absence of a candidate. It is the absence of a trench.

The physical symptoms have a name in the literature and a large body of work behind them, most of it built on other sites: expectation effects. Tell a group of people that a place induces nausea, anxiety and a sense of being watched, walk them into it in darkness, and a reliable proportion will experience nausea, anxiety and a sense of being watched. This is measurable, replicable and not a slur on anyone. It also explains the reporting asymmetry that nobody in the paranormal literature addresses. This wood is in daily use by runners, cyclists, families and dog walkers, and they do not report anything, because they were not told to expect anything. The symptoms cluster almost perfectly around visitors who paid to have them.

Electronics are the weakest item on the list. Cold, damp, old batteries and a hillside a few kilometres from a transmitter-heavy city will do everything reported.

Barnea’s photographs deserve better than either the credulous or the dismissive treatment they usually get. Two frames, an object with no reference point for scale or distance, taken in 1968 on film, in a state where the photographer had every incentive not to draw attention to himself. They were examined by the leading Romanian ufologists of the period and pronounced authentic, which meant, then as now, that nobody had demonstrated a fake. That is not the same as evidence of a craft, and the widely repeated claim that the regime punished Barnea for the pictures has no verifiable source behind it. What can be said is that the man appears to have gained nothing at all from the most famous photograph in Romanian ufology.

The girl fails at the first check. There is no record, and there was never a source.

The shepherd fails at a more interesting point. Baci is the ordinary Romanian word for the head shepherd of a fold, and Baciu is the name of the commune at the foot of the hill. A wood named for the shepherd’s village, or for the shepherding on it, would be called almost exactly what this wood is called. The etymology is genuinely uncertain and honest Romanian sources say so. But a story that explains a place name is the commonest kind of folk story there is, and the direction of travel usually runs from the name to the story rather than the other way round.

Which leaves the small honest door. Sift photographed something, for four decades, seriously enough to keep doing it alone with no audience and no profit, and the archive is gone. Pătruț recorded local magnetic fluctuations with instruments and reported them accurately, and a real reading with no established cause is still a real reading. Neither man was a fraud. What we have is a case whose strongest material was destroyed within a week of its collector’s death, which means the thing sceptics most want to examine is the thing that no longer exists.

My Thoughts

I keep coming back to Adrian Pătruț’s other career.

The baobab work is not a hobby running alongside the Hoia Baciu Forest research. It is the same man, the same faculty, the same city, and it is a method for putting a question to a tree and getting a dated answer back. By any reasonable measure he is one of the people on this planet best equipped to walk up to a trunk and establish what happened to it and when.

And for fifty years, in a forest twenty minutes from his university, he has been pointing instruments at the air.

I do not say that to mock him. I say it because it is the clearest illustration I can find of what has actually gone wrong at Hoia Baciu Forest, which is not credulity. Credulity would be easy to argue with. What has gone wrong is a question that got asked in the wrong direction and then everybody kept asking it that way, for six decades, with increasing sophistication and better equipment, while the answer stood in the ground at shoulder height with a fist of scar tissue on it.

Because that fist is a date. Every one of those bolls is a record of a decision made by a person on a specific morning, and then made again, and again, on a schedule, for as long as somebody on this hill needed poles and firewood and leaf fodder more than they needed a tall tree. A pollard is a contract between a household and a tree, renewed every eight or twelve or fifteen years, and it ends when one party stops turning up. The trees are still holding up their end. The last cut on any given one of them could be dated to a decade, possibly closer, by exactly the technique the country’s most celebrated tree-dater has spent his life refining, and as far as I can establish nobody has ever done it.

That is what I would want to know. Not whether the field readings fluctuate. When did the shepherds stop coming, and did they stop all at once.

Because underneath the trees is Gura Baciului, and roughly eight thousand years of human use are written into this hillside, and the two things on it that are genuinely irreplaceable are both being subtracted right now while everybody watches the sky. The quarry is eating the settlement. The chainsaws are taking the trees, and the man who filmed them was threatened for it. Neither process needs a single unexplained phenomenon to proceed, and neither one has a night tour. The legend says a shepherd vanished into this forest, and that is exactly the wrong shape for what is happening. The shepherds did not vanish. They are all over the hill, in every knotted crown, and we are cutting them down for firewood because we cannot read what they left.

And then there is the part I cannot get past.

Hartel, asked directly what he thought about the legends, said he would keep them. If the choice were between the stories and construction on the hill, he would take the stories, because they are part of what makes the place worth protecting. He said that in the same season the road dispute was at its loudest, which means the choice he was describing was not a thought experiment. He is right, and it is unbearable. The false account of this wood is currently doing more to defend it than the true one, because the true one is a chapter in landscape history and the false one sells a hundred-dollar night walk that funds a guide who, by his customers’ own accounts, spends part of the evening telling visitors about the illegal logging.

So the ghost is the bodyguard. The story that gets everything wrong about Hoia Baciu Forest is the only thing standing between it and a housing development, and it works precisely because it is wrong, because nobody would fly in from another country to look at a coppice stool.

I do not know what to do with that. I am fairly sure I have just spent four thousand words dismantling the only argument that has ever protected this hillside. If a place can only be saved by being misunderstood, is telling the truth about it a service or a demolition?

Visiting Hoia Baciu Forest

Hoia Baciu Forest sits a few kilometres west of Cluj-Napoca city centre. There is no ticketed forest entrance or central visitor centre. The most straightforward eastern approach is near the Romulus Vuia open-air section of the Ethnographic Museum of Transylvania, which publishes current bus and visitor information. Journey times vary with traffic and the chosen trailhead, so check a current map before setting out. Do not assume that a road into the trees is open to public vehicles: access has changed during local efforts to keep through traffic out of the wood.

Routes and starting points vary, but allow two to three hours for a visit that reaches Poiana Rotundă and returns without rushing. Carry an offline map: the path network branches repeatedly and signage is inconsistent. Trails are unsurfaced, root-crossed and muddy after rain, so wear shoes with proper grip and carry your own water. Facilities are available in the city and at established attractions, not reliably inside the forest.

Daylight is when the trees can actually be seen, which is the argument for going in the morning. Autumn gives the best light and the worst mud. Winter strips the canopy and makes the old pollards legible in a way summer hides completely, and is the season I would choose.

Night visits are the commercial product, and several Cluj operators advertise guided walks with transport and equipment. Prices, durations and inclusions change, so compare the current details rather than relying on an old quoted figure. Going alone after dark means unlit, uneven ground and a branching path network; a reputable local guide is the sensible option.

On conduct, two things. This is a working suburban wood that people use every day, and it is under real pressure from illegal felling and dumping, so take everything out with you and stay on the paths. And the Gura Baciului archaeology on the northern side is a protected site that has already lost ground to the quarry. It is not a place to dig, detect or take anything from.

Staying in Cluj-Napoca? Compare places to stay in Cluj-Napoca. If you book through this link, Shin Stories may earn a small commission at no extra cost to you.

Before You Go

This is Location 13 of twenty in Season 1 of The Dark Atlas, immediately before Bhangarh Fort. The companion episode will appear in the Spotify player above when released, and it does something the article cannot: it reads the hill as a single continuous document, from the Neolithic pits to the chainsaw. Continue through the archive with the Highgate Vampire, and follow The Dark Atlas on Spotify so the next location arrives without you having to look for it.

Sources

Books and Academic Works

Lazarovici, Gheorghe, and Zoia Maxim. Gura Baciului: Monografie arheologică. Bibliotheca Musei Napocensis XI. Cluj-Napoca: Muzeul Național de Istorie a Transilvaniei, 1995.

Luca, Sabin Adrian, and Cosmin Suciu, eds. The First Neolithic Sites in Central/South-East European Transect, Volume II. The site catalogue documents damage and quarrying loss at Gura Baciului.

Mathieson et al., “Paleogenomic Evidence for Multi-generational Mixing between Neolithic Farmers and Mesolithic Hunter-Gatherers in the Lower Danube Basin”. Includes the excavation history and material record of Gura Baciului.

Pătruț, Adrian. Fenomenele de la Pădurea Hoia-Baciu. Cluj-Napoca, 1995. The surviving Sift photographs and Pătruț’s early measurements.

Patrut, Adrian, Karl F. von Reden, Daniel A. Lowy, et al. “Radiocarbon dating of a very large African baobab.” Tree Physiology 27, no. 11 (2007): 1569–1574.

Patrut, Adrian, László Rakosy, Roxana T. Patrut, et al. “AMS radiocarbon dating of the large pedunculate oak of Mercheașa, Romania.” Studia Universitatis Babeș-Bolyai Chemia 66, no. 3 (2021).

Lange, Houran, Harte and Havens, “Context-induced paranormal experiences”. A controlled study of how information about a supposedly haunted setting can influence reported experiences.

Journalism and Long-form Reporting

Lucuț, Carmen. “Pădurea Hoia Baciu, de la legendă, la adevăr”. Monitorul de Cluj, 4 February 2021. The Tibor Hartel interview on pollarding, coppicing, the legends and illegal felling.

Monitorul de Cluj reporting on the forest-edge development and access road, March 2021, including the 700-apartment proposal and request to suspend its permit.

Monitorul de Cluj, “Drumul prin Pădurea Hoia-Baciu, închis traficului auto”. Reporting on the later closure of the route to through traffic.

Newsweek România. “Misterele pădurii Hoia-Baciu: OZN-uri, apariții bizare și fenomene stranii.” 30 March 2023. A detailed Romanian-language account of the Sift and Barnea material and the hill’s pre-1968 reputation.

Official and Institutional Sources

Repertoriul Arheologic Național and the Lista Monumentelor Istorice: Gura Baciului archaeological site, code CJ-I-s-A-06949.

Ethnographic Museum of Transylvania, Romulus Vuia open-air park. Current visitor hours and public-transport information for the eastern approach.

Cluj Tourism, Baciului Gorge nature reserve. Official description of the protected gorge on the northern edge of Hoia Forest.

Institutul Național de Statistică, 2021 census: Cluj-Napoca, 286,598 inhabitants.

Further Reading

The book links below are affiliate links. As an Amazon Associate Shin Stories earns from qualifying purchases, at no extra cost to you.

Rackham, Oliver. Woodlands. Collins New Naturalist, 2006. The standard work on reading worked woodland and understanding pollards and coppice stools.

Hartel, Tibor, and Tobias Plieninger, eds. European Wood-Pastures in Transition: A Social-Ecological Approach. Earthscan from Routledge, 2014.