Jerusalem Archaeological Park, Davidson Center

The entrance is on Ma'alot ha-Rav Shlomo Goren Street by the Dung Gate, through an underpass on the access road to the Western Wall. Inside, the ground falls away sharply: the Tyropoeon Valley runs under the park, and the level difference across the western wing of the Umayyad palace alone reaches about 15 metres.

Because the ground holds three thousand years of building, the excavators had to choose which century to leave standing, and the park is laid out in three compounds accordingly. The area beside the Western Wall was taken down to the Second Temple period; the ground along the eastern half of the southern wall was left at its Byzantine level, and the western half at the early Muslim level. Knowing which compound you are standing in is the difference between reading the site and looking at rubble.

One structure crosses all three. Suleiman's sixteenth-century city wall was built partly on the outer wall of the Umayyad palace, which is why it stops at the palace corner and turns north to meet the southern wall halfway along rather than at its end. To keep the garden continuous, the excavators cut a hole through it.

Charles Warren was first. Forbidden by the Ottoman authorities to dig on or near the Temple Mount, he worked by shaft and tunnel instead, and between 1867 and 1870 found a large pier standing some 13 metres west of the wall, opposite the stones Robinson had noticed in 1838.

The excavation that made the park ran from 1968 to 1978 under Benjamin Mazar, with Meir Ben-Dov in the field. Israeli archaeologists had decided not to dig on the Temple Mount itself, so they dug all the way around it and published under the title Excavations of the Temple Mount. Mazar died in 1995 before the final report appeared; his granddaughter Eilat Mazar and then a team under his daughter published the rest.

Ronny Reich and Yaacov Billig cleared much of the collapse around Robinson's Arch between 1994 and 1996. The most recent work is the Israel Antiquities Authority's, under Moran Hagbi and Joe Uziel, between March 2016 and August 2017 and again in 2019 and 2020, ahead of the Davidson Center's expansion. Their report makes a point worth carrying round the site: the conservation work done when the park was first opened hampered their reading of the architecture, and no record survives of the remains before it.

Four courses of stone jut from the Western Wall above the park: a row of impost blocks cut into a dentil pattern and three courses of voussoirs, all that is left of the springing of an arch. Edward Robinson identified them in 1838, after Frederick Catherwood had noticed them, and read them, following Josephus, as the remains of a bridge running west across the valley. The name stuck. The reading did not.

Warren's pier and Mazar's excavation together showed what the structure was. South of the pier the diggers found six more vault-supporting piers, stepping down in height over about 35 metres. The arch and those vaults carried a monumental flight of stairs that rose from the street, turned, climbed the descending vaults and crossed back over the street through the arch into a gate of the Royal Stoa above. It was not a road to anywhere. It went up.

The measurements are large by any standard of classical building: a span of about 15 metres, an arch about 15.2 metres wide, the stair carried some 17 metres above the street, and a pier 15.2 by 3.6 metres preserved to about 5 metres. Four small hollows in the pier's base were probably shops opening onto the street. Fragments of the gate at the top give a width of about 5 metres, though whether it was a single, double or triple opening is not known.

The date it stopped working is fixed by an accident. Excavators south of the Temple Mount found a Roman milestone naming Vespasian and Titus, cut from one of the handrails that had stood on top of the arch. Whoever ordered that milestone had a ruined staircase to quarry, which puts the collapse no later than 79 CE.

• Ancient: the four courses of springing stones in the Western Wall, the pier, the six descending vault piers, the street paving and the shop recesses
• Gone: the arch itself, the staircase it carried, and the gate into the Royal Stoa at the top
• Elsewhere: the handrail block, recut as a Roman milestone naming Vespasian and Titus
• Modern: the raised prayer platform beside the arch, completed in August 2013

The street along the base of the wall is the reason to come. It is the Tyropoeon Valley road, bounded by high stone kerbs and about eight metres wide; its full run from the Pool of Siloam to this corner is around 750 metres, of which only a stretch is exposed here. When it was uncovered it lay under stones thrown down from the top of the wall during the Roman siege in 70 CE. Most of that pile was cleared, but a section was deliberately left in place, the paving slabs cracked and dished under the impact.

One of the fallen stones carries writing. Mazar found it in 1968 near the south-western corner: a block of 84 by 31 by 26 centimetres, inscribed in square Hebrew letters, reading to the house of the trumpeting and then breaking off. Josephus describes a priest standing on the Temple's parapet to sound a trumpet at the beginning and end of the Sabbath, and the stone is read as the sign marking that place. What the broken third word said is not known. The original is in the Israel Museum.

The street's date is genuinely open, and it is not a footnote: it decides who built the most photographed pavement in Jerusalem. Herod's reconstruction was ordered in 20 or 19 BCE, and the paving was long attributed to him. Excavation beneath it by the arch found three oil lamps of a first-century type and seventeen identifiable coins, several struck by the prefect Valerius Gratus in 17 or 18 CE, which puts this stretch of wall and arch after that year. A separate count of fifteen coins from under the road gave a latest issue minted under Pontius Pilate.

So some scholars now credit the paving to Agrippa II rather than to Herod. Josephus records that under the procurator Albinus, between 62 and 64 CE, the streets were paved with white stone to find work for the men laid off when the Temple was finished. The slabs show almost no wear.

Round the corner, the southern wall of the Temple Mount runs 281 metres, which Josephus gives as one stadion. This was the pilgrims' side. Two gate complexes, known together as the Hulda Gates, opened in it: the Double Gate to the west and the Triple Gate to the east, each leading into a vestibule and then up a vaulted ramp onto the esplanade. Both have been blocked since the Middle Ages, so a visitor sees the outline of an entrance rather than an entrance.

The Double Gate is the one to look at. Its lintel and relieving arch are Herodian; the elaborately carved cornice above is Umayyad, added six centuries later. Only about half its width, some 3.09 metres, shows from outside, and the three openings of the Triple Gate are about 4.03 metres each.

The stairway that climbed to the Double Gate is largely intact, while the flight to the Triple Gate is mostly destroyed. The treads are deliberately irregular, running roughly 12 to 35 inches deep with risers of about 7 to 10 inches, so that nobody can take them at a run. Ritual baths were found at the foot of the stairs, where pilgrims immersed, and the Mishnah's tractate Middot describes the crowd entering by one gate and leaving by another.

The corner did not stay empty. By the end of the fourth century a dense Byzantine neighbourhood stood over the ruins, and parts of its mosaic floors survive; the excavators think it was destroyed in the Persian invasion of 614 and never rebuilt. Nearby are the remains of a two-storey building identified as a synagogue on the strength of two menorahs painted in red on its walls, a third cut in stone, and recesses at every doorway read as mezuzah niches. It served about sixty years before an Umayyad palace was built over it.

That palace is the most substantial standing structure in the park. Mazar uncovered four large administrative buildings from the reign of al-Walid I covering the whole area of the garden; only one, Building II, is presented to visitors. It measures roughly 80 by 80 metres, a ring of rooms around an open courtyard, built in places from stone robbed out of the Temple Mount's own ruins, standing in places three storeys high, and linked at roof level by a bridge to the al-Aqsa Mosque. It was probably brought down by the mid-eighth-century earthquake.

The Davidson Center sits inside it, mostly underground so that it does not break the skyline. It opened in April 2001 and was built without building permits: the Jerusalem municipality comptroller reported in 2005 that the claim the works fell within the definition of archaeological excavations was baseless. In December 2013 a court ordered the East Jerusalem Development Company, which had built and run it, to hand ownership to the Company for the Reconstruction and Development of the Jewish Quarter. An agreement signed that October would have passed operation to the Elad Foundation, which runs the City of David site in Silwan; the state appealed, and in September 2014 the court struck it down, holding that a site of this significance must stay with a public authority acting as trustee for the public.

The museum closed for three years and reopened in March 2023, rebuilt and expanded by the Israel Antiquities Authority with funding from the William Davidson Foundation, and the Jewish Quarter company now sells the tickets. The galleries lead on the Temple and its menorah. The centrepiece is a coin of Antigonus II Mattathias, the last Hasmonean king, which Yuval Baruch of the Israel Antiquities Authority describes as the oldest known artistic depiction of the Temple menorah, made 107 years before the Second Temple was destroyed.

The other long dispute is over prayer. After a 2003 Supreme Court decision the area beside Robinson's Arch was prepared for services not permitted at the main plaza, and a permanent platform of about 4,480 square feet, named Ezrat Yisrael, was completed on 25 August 2013 with access at all hours. Reform and Conservative movements and Women of the Wall have consistently called the arrangement second-class. Archaeologists object from the other side: Baruch called an earlier expansion plan a threat to the only place from which the Western Wall's courses, the Herodian street and the evidence of the destruction can be seen together.