The name comes apart into two Old Norse words: þing, meaning assembly, and vellir, meaning fields. Assembly fields. The same root gives English its “thing,” which once meant a gathering before it came to mean an object, and it survives in place names across the old Norse world, Tynwald on the Isle of Man, Dingwall in Scotland, Tingwall in Shetland. Anglicized, it becomes Thingvellir, which is how most signage and most visitors render it. Both spellings point at the same ground.

Þingvellir National Park lies about 28 miles northeast of Reykjavik, in the southwest of Iceland, spread along the northern shore of Þingvallavatn where roads 36, 361, and 52 meet.

Two things happened on this ground, and they are not related in any way a geologist or a historian would accept. The crust pulled apart here, dropped a block of land between two faults, and left a valley with a wall on either side. Nine thousand years later, a people who had crossed the ocean to escape a king needed somewhere to meet and agree on laws, and they picked the valley. The rock did not intend the parliament. The parliament did not know the rock was still sinking beneath it. But the site works as it does because of what it is: the wall carried a speaker’s voice, the plain held the crowds, the river fed them. And the ground kept dropping the whole time, nearly thirteen feet since the first assembly convened. Both stories have to be told, and neither explains the other. That is what makes the place worth two days.

We woke to rain at the campsite, with the forecast calling for it on and off through the whole day. There was no chore on board that actually needed doing, and there is no such thing as bad weather for people living in a truck, only the wrong clothes for it, so we put on our warmest and driest layers and walked out into the park anyway.

Geology and geography

Four words carry most of the weight in any description of this valley, and they are worth separating before going further.

A fault is a fracture where the two sides have slid past each other. A fissure, more properly a tension fracture, is one where the two sides have pulled apart without sliding. These are different ways for rock to fail, and geologists classify them separately: shearing in one case, opening in the other. The distinction is not a matter of naming convention, it is a matter of mechanics, and it governs how strain is taken up and how fluid moves through the rock.

A rift valley is what forms when the crust is pulled apart rather than pushed together. Instead of buckling upward into mountains, the ground stretches, thins, and breaks along steep faults, and the land between those faults drops.

A graben is that dropped block. The word is German for ditch, and it names the piece of crust that sank between two roughly parallel faults. Þingvellir sits on a graben about 4.7 kilometers (3 miles) wide.

At Þingvellir, though, the two happen in the same place, and the graben is what results. Almannagjá on one border and Hrafnagjá on the other are both normal faults, meaning the ground on the valley side of each has dropped relative to the ground outside it. That is the sliding. But each also gapes: Almannagjá carries about 100 feet of vertical displacement and is locally 210 feet wide. Hrafnagjá runs longer, close to seven miles, with something near 90 feet of displacement, and a shorter parallel fault called Bæjargjá adds roughly another 70 feet on the same side.

The block between them has fallen more than 130 feet over nine thousand years while the two walls pulled 230 feet apart. Shearing along the boundaries, opening across the middle, the graben floor riding down between them. Structural geologists working in Iceland call these massively dilatant faults, precisely because the shearing at depth is accompanied by enormous gaping in the upper few hundred feet of basalt. The reverse also occurs, and it occurs on the very ground we walked: many of the small cracks on the graben floor show no vertical offset at all, and the park’s own research describes them as surface expressions of deeper faults that never reached the top. A fault can look like a fissure. A fissure can be a fault with the offset hidden. Icelandic uses one word for both, gjá, meaning an open or gaping fracture, and Almannagjá, Hrafnagjá, Silfra, and Flosagjá all carry it. That word describes what a person standing on the trail can actually see, and declines to assert what they cannot. On this particular ground, it may be the more honest label.

The subsidence has not stopped. Since the Alþingi was founded in 930, the assembly ground itself has sunk close to thirteen feet, which means every chieftain who ever stood at the Law Rock stood a little higher than we did today.

One caution on the naming, because the park’s own sources conflict. Þingvellir’s website calls the bounding faults Almannagjá and Heiðargjá on one page and Almannagjá and Hrafnagjá on another. We are giving the names as the structural literature gives them, with Almannagjá on the western border of the graben and Hrafnagjá on the eastern, and noting that the park does not fully agree with itself.

Between the two boundary faults, the graben floor is cut by a great many smaller fractures, and they are not all the same kind of thing. Silfra is a true fissure, opened in an earthquake in 1789, largely submerged, with no vertical offset. Flosagjá and its branch Nikulásargjá, nicknamed Peningagjá after visitors began throwing coins off a bridge built there in 1907, are also openings. Bæjargjá, by contrast, is a fault in its own right. Others carry names most visitors never hear: Kattargjá, Skötutjörn and Skötugjá, Kolsgjá, Leiragjá, Sleðaásgjá, Vallargjá, and Brennugjá, the last of which comes up again in the history section. Which of these have measurable throw and which are pure extension is not something a walker can determine by looking, and in several cases the published record does not say either.

That brings up the question of how accurate the tourist framing actually is. Almannagjá is commonly described as the edge of the North American plate, with Hrafnagjá marking the Eurasian edge across the valley, and in a loose sense the labels hold. But the real plate boundary through southern Iceland is not one clean line. It splits into two branches: the western volcanic belt, which runs from Hengill north through Þingvellir toward Langjökull, and a second branch further south along the transform zone where Iceland’s largest earthquakes originate. Between them sits a wedge of crust belonging to neither the North American nor the Eurasian plate, which geologists call the Hreppar microplate and which drifts independently of both. Almannagjá is a genuine, active fault at the edge of a real rift. But calling any single fracture “the” plate boundary, as if the whole separation happened along one crack, is a convenience built for visitors rather than a description of the structure. The boundary is the entire rift zone, several miles wide. Everything we walked past is one fracture within that system, not a candidate for the one true line. The spreading rates make the same point: the Mid-Atlantic Ridge as a whole widens by something on the order of an inch a year, but the western branch through Þingvellir carries only a fraction of that.

The water in the fissures comes from Langjökull glacier, about 30 miles north. Meltwater filters through decades of lava rock before surfacing in fissures like Silfra and Peningagjá as spring water clear enough to see straight to the bottom, then drains into Þingvallavatn, Iceland’s largest natural lake, which the valley wraps around to the south.

History

To understand why the Alþingi was founded here in 930, it helps to start with why Iceland needed a national assembly at all. Most of the people who settled Iceland from around 870 onward came from Norway, and a good number were leaving specifically to get away from centralizing kings. Norway at the time was being pulled together under Harald Fairhair, who spent his reign subjugating local farmers and military leaders who had previously answered to no one above them. Snorri Sturluson, writing centuries later, described Harald’s consolidation of power as the reason many chose to leave rather than live under it. So when these families set up new farms in Iceland, they deliberately avoided recreating a monarchy. Instead, each district organized itself around a local chieftain, called a goði, who ran a regional assembly and settled disputes for his own followers, called thingmen.

That system worked at the local level, but it had an obvious gap: nothing bound the different districts together. If a dispute crossed from one chieftain’s followers to another’s, there was no shared court, no shared law, and no higher authority to appeal to. As the population grew, this started to matter more.

A regional assembly at Kjalarnes, set up by the son of Iceland’s first settler, took the first concrete step by sending a man named Úlfljótr to Norway to study its laws. He spent three years there, adapting the Gulating law of western Norway into a framework meant for Iceland’s very different circumstances, no king, no central enforcement, and returned with what became the country’s first shared legal code. Historians do debate how much of that returning code was genuinely Norwegian import versus material worked out locally once he was back, since Iceland’s later law collections do not match Norwegian sources especially closely, but the broad outline, that he went, studied, and came back with a workable framework, is well documented.

While Úlfljótr was away, his foster brother Grímur geitskór spent that same stretch of years traveling Iceland to find a site that could host a truly national assembly, neutral ground that did not belong to any single chieftain’s territory. He had a stroke of luck: the site he eventually chose had recently become public land after its owner was convicted of killing a slave, a crime taken seriously enough at the time to cost him the property. Beyond the legal convenience, the location worked on practical terms too. The cliffs formed a natural amphitheater for addressing a crowd, the Öxará river supplied drinking water and fish, and the surrounding plains had enough grazing for the horses that chieftains and their followers rode in on. Iceland’s own history office also notes a simpler, more logistical reason: Þingvellir was roughly reachable from the populated north, south, and west, and even the longest ride, from the glacier bound east, took no more than seventeen days.

The assembly that resulted, the Alþingi, met here every June for about two weeks, for close to nine hundred years running. It was never just a courtroom. Attendees built temporary lodgings called búðir, walled with turf and stone and roofed with woven wool, and roughly fifty of these booth ruins are still visible in the park today. During the two weeks of assembly, Þingvellir functioned as something close to the country’s temporary capital: old law books mention tanners and peat cutters with their own booths, sagas describe beer brewers and food sellers setting up alongside them, and merchants, sword sharpeners, and entertainers all did business with the crowd. Iceland’s national culture, language, and literary tradition are widely credited to have taken shape here as much as in any single farmstead, simply because this was where the whole country gathered and told its stories once a year. That said, one popular claim is worth qualifying.

After independence, a strain of nationalist history writing described Þingvellir as not just an assembly site but Iceland’s greatest marketplace. A systematic study of the written and archaeological record found comparatively little evidence of large scale trade goods at the site, suggesting that while some commerce certainly happened alongside the legal business, the image of a major medieval market here is likely overstated.

Free men could attend the assembly, cases were argued and settled, and the Lawspeaker stood at Lögberg, the Law Rock, reciting the law of the land from memory before any of it was written down. In the year 1000, the assembly voted to adopt Christianity as the law of Iceland, a decision credited with avoiding a civil war between the country’s pagan and Christian factions at a moment when both sides had armed supporters present. Iceland’s independence lasted until 1262, when infighting among chieftain families led the country to accept the Norwegian crown, and legislative power drained gradually toward Norwegian, then Danish rulers, ending in 1662 when the Danish king took absolute authority over Iceland.

Punishments handed down here darkened over that period. Brennugjá, one of the fissures named above, is said in local tradition to be where nine men, not women, were burned in the 1600s, a detail worth noting since it runs against the usual assumption that Iceland’s witch trials mainly targeted women. Drekkingarhylur, which we walked past later in the day, is more firmly documented: it was used from around 1600 into the mid 1700s to drown at least 18 women convicted of crimes such as adultery, incest, or infanticide, while men convicted of comparable offenses were beheaded or hanged elsewhere in the park instead.

The last Alþingi session at Þingvellir was held in 1798, after which the assembly was abolished and later reestablished in Reykjavik in 1845. The site had one more chapter left. On June 17, 1944, the Alþingi convened here a final time to declare the Republic of Iceland, ending Danish rule for good, and the date is still Iceland’s national day.

The visit: Langistígur

We started the morning on the short trail to Langistígur, one of two entry points into Almannagjá. The path is old cobblestone rather than boardwalk, worn from centuries of use as a bridle trail, and it still permits horses alongside walkers. Inside it, the fault reads plainly: the wall on one hand, the dropped block of the graben floor on the other. This is the boundary, and the ground we stood on had fallen about a hundred feet relative to the rim above. The trail carries a grim local nickname, the Walk of Death, since this is where men condemned under Iceland’s harsher codes were beheaded or hanged. A scene from Game of Thrones was filmed in the same corridor.

Öxarárfoss

From there we continued to Öxarárfoss, which drops about 44 feet off the edge of Almannagjá where the Öxará river spills into the gorge. The waterfall exists because the fault exists, water meeting the drop the fault created. The river’s course is not natural to this spot either. According to the Sturlunga Saga it was diverted into the gorge in the assembly’s early days so that people gathered at Þingvellir would have drinking water, which makes the park’s best known waterfall a piece of medieval infrastructure.

Hakið

We went to the main park entrance and up to Hakið, the viewpoint above the valley. This is where the graben becomes visible as a single object rather than a set of features. The platform sits on the Almannagjá rim, and from it the valley floor runs out flat toward Þingvallavatn with the far wall standing across it, about three miles off. Both boundary faults in one frame, the sunken block between them.

Flekaskilin

From Hakið we walked into Flekaskilin, whose name translates directly to “the plate boundary.” The name is the simplification discussed in the geology section, and standing in it is the clearest argument against the name.

Snorri’s Booth

We stopped at Snorri’s Booth, one of the búðir ruins described earlier, on the graben floor below the fault wall. It carries the name of Snorri goði, a chieftain who appears in several of the sagas, and some sources connect it to his descendant Snorri Sturluson, the lawspeaker who wrote down much of the Norse mythology that survives.

Lögberg

From there we reached Lögberg, the Law Rock. It sits at the base of the Almannagjá wall, which is what made it work: the fault face behind the speaker returned his voice to the assembly on the floor below. The ground here has dropped about thirteen feet since the assembly first met.

Drekkingarhylur

Afterward we stopped at Drekkingarhylur, the drowning pool, where the Öxará pools against the Almannagjá rock wall before continuing on. There is no plaque naming the women who died there, only a small marker. In 2000, during Iceland’s Christian millennial celebrations, a wreath was placed at the pool as a gesture of atonement. By then the rain had picked up considerably, so we turned back to Carioca for lunch.

Öxarárhólmi

After lunch we walked to Öxarárhólmi, the small islet in the river used as a dueling ground in the assembly’s early years, a practice that ended not long after Iceland adopted Christianity in the year 1000.

Peningagjá

From there we continued to Peningagjá, a branch of Flosagjá, out on the graben floor well away from either boundary fault. This is the other kind of fracture: an opening, not a fault, the lava field pulled apart with nothing dropped. Its deepest sections are documented at around 82 feet, clear enough to see coins on the bottom. The custom of throwing money in dates only to 1907, when a bridge was built across it for a visit by the Danish king.

Þingvallakirkja

We passed Þingvallakirkja, built in 1859 on the site of Iceland’s first Christian church from just after the year 1000, when King Olaf of Norway is said to have sent timber and a bell. The current church seats around 35 and holds two altarpieces: an 1834 painting on driftwood by a local farmer named Ófeigur Jónsson, judged too plain by church authorities and replaced by a more polished 1896 painting from a Danish artist, though the original made its way back to Iceland decades later. The tower holds three bells, one cast for the 1944 declaration of the Republic and known as Iceland’s Bell. Behind the church is the national cemetery, which despite the name holds two graves, both Icelandic poets.

Þingvallabær

Next to the church is Þingvallabær, a cluster of white gabled buildings whose symbolic role traces to two national anniversaries rather than one. It was built in 1930 for the millennium of the Alþingi, originally three gables under a turf roof echoing the old farmhouses, and it housed the park warden and the local priest rather than any head of government. The turf proved too heavy for the roofline and was replaced with metal during a renovation in the 1970s, and at that same renovation, timed to the 1974 celebration of the 1100th anniversary of settlement, two more gables were added and the building took on its current role as the Prime Minister’s official summer residence. Its use as a leader’s house is a recent role grafted onto a building whose purpose from the start was to mark the country’s connection to this ground, which is why it sits here rather than somewhere private.

Silfra

We finished the first day at Silfra, the meeting point for the diving and snorkeling tours and the fissure itself, opened in the 1789 earthquake. Like Peningagjá it is an opening rather than a fault, and it holds the same water: meltwater filtered for decades through Langjökull’s lava, leaving almost nothing suspended to scatter light. Visibility is reported up to about 330 feet. We watched a few groups suiting up before heading back to camp. The rain never really let up, but staying dry underneath it turned out to be the easier part of the day.

Hrafnagjá

The next morning we drove across the graben floor to the viewing area at Hrafnagjá, the fault on the far side. About three miles of dropped ground lie between it and Almannagjá, and standing at one wall after the other is the only way to take the structure in whole. The two are a matched pair: normal faults facing each other across the block that fell between them. Hrafnagjá is the longer of the two, close to seven miles, with something near 90 feet of vertical displacement, and a shorter parallel fault called Bæjargjá adds roughly another 70 feet on the same side. From the platform the valley runs back the way we had come, toward the Almannagjá wall we had walked the day before, with the assembly ground somewhere in the middle of it.

We also visited the Stelpuhellir Lava Tube that sits on a remote area of the park