Temple Mount PDF Print E-mail
Written by Stephen Langfur
Article Index
Temple Mount
First Temple
Second Temple
Red Heifer

Where was the Second Temple?

We cannot be sure whether the square of 500 royal cubits per side belonged to the First Temple, but we may be confident that it belonged at least to the pre-Herodian Second. The reasoning is simple: (1) The Mishna describes a square of 500 cubits on each side, and it can hardly be a coincidence that if we draw a line from the present-day platform's northwest step, which has exceptionally large stones, to the Haram's east wall, the length turns out to be 500 royal cubits. (2) The literary sources indicate nothing, after the time of the First Temple, to which this square could have belonged - except the early Second Temple. Besides, we have read that Pompey, in order to conquer the temple in 63 BC, had to fill a moat on the north side. If this was the ditch detected by Warren (see his map), it was exactly in the right position to defend the 500-cubit platform.

We have also seen that the platform was expanded to the south either by the Seleucids or the Hasmoneans. In the portion south of this expansion, however, we encounter a problem: 

In 2011, archaeologists Eli Shukron and Ronny Reich dug down to bedrock near the southwest corner of the Temple Mount. They discovered various rock-cut installations that had belonged to buildings earlier than Herod's temple. One of them was identified as a ritual bath, in Hebrew a mikveh. The temple builders had filled it, topped it with three large flat stones, and laid the western wall's first course over part of it. The archaeologists of 2011 removed fill from the part of the mikveh that was not beneath this wall's foundation stones. They unearthed three clay oil lamps that were typical of the 1st century AD. They also found coins, of which the latest four had been minted by the Roman prefect Valerius Gratus in 17 - 18 AD. (For an account by the Israel Antiquities Authority, see here.) The excavators concluded that this section of the Western Wall could not have been built under Herod, who died in 4 BC.

This conclusion, which received worldwide publicity, would imply a post-Herodian date not only for the southern end of the western wall, but also for the royal portico that stretched eastward for 920 feet from this southern end. Josephus wrote, "this cloister deserves to be mentioned better than any other under the sun." Its roof would have been about as high as the top of the dome of al-Aqsa today. It had four rows of columns, 162 in all, making three aisles. On either side of the middle aisle, the columns were 27 feet high and so thick that it took three adults to surround them. By comparison, the columns of Italian marble in al-Aqsa, including the capitals, rise 15 feet; the central ceiling, however, is about as high as the original pillars would have been.

In accordance with the coins of Gratus, we would also have to redate the Double and Triple ("Hulda") Gates in the southern wall, as well as the grand staircase. The redating would undermine Josephus' account, for he tells us that Herod celebrated the completion of the temple, including the magnificent royal portico on the south. Finally, one may question whether so huge a structure could have been built during the rule of a Roman procurator. And why, in such a case, would Josephus have called it "royal," given that the last king, Herod, had died more than 20 years earlier? If the portico's funder had been a Jewish ruler, namely, Herod's grandson Agrippa (r. 40 - 43 AD) or his son Agrippa II (who was made responsible for the temple in 48 AD), surely Josephus, born in 36 AD, would have known as much and said so.

Admittedly, we read in John 2:20 that the construction of the temple, which began ca. 20 BC, took 46 years, and Josephus reports that it was completed in 63 AD. This left 18,000 workers idle, he says, and Agrippa II had the city repaved in order to provide jobs. (As often with Josephus, one suspects that a copyist added a zero to the demographic data.) But we have always had these passages, and they never stopped us from attributing the main structures to Herod, including the royal portico. Do the coins of Gratus compel a redating?

These coins, as said, were not found under the wall's foundation stones. In a recent comment, Leen Ritmeyer notes: "Actually the coins prove nothing at all, as the mikveh only project[s] a few centimeters under the wall. The coins came probably from a later repair." For instance, the street beside the wall shows no sign of use, so it was probably one of those repaved by Agrippa II, seven years before the temple's destruction.

Given these doubts about the proposed re-dating, I shall continue to assume that the temple's main structures were Herodian, including the royal portico and the outermost southwest corner. This assumption guides the following text.

What then do we know about the location of Herod's temple? There is no doubt that the great surrounding walls belonged it, for the following reasons. Apart from Josephus' description, there is a Hebrew inscription carved into a stone from the uppermost southwest corner (i.e., found lying directly on the pavement, hence among the first stones to have been battered down). This reads, "For the place of the trumpeting." It corresponds to a passage in Josephus, who wrote of a place at the top of the temple "opposite the lower city," where "one of the priests stood of course, and gave a signal beforehand, with a trumpet at the beginning of every seventh day, in the evening twilight, as also at the evening when that day was finished, as giving notice to the people when they were to leave off work, and when they were to go to work again."

hebron-temenos2.jpgThe walls compose an irregular quadrilateral of 920 feet on the south, 1590 on the west, 1033 on the north, and 1509 on the east. They enclose 36 acres (144,000 square meters or 20 soccer fields by the FIFA standard). In its time this was the largest temenos ever built. It took up a fifth of the city before 66 AD. Judging from the stones that were batterred down (found on the ancient street beneath the western retaining wall), we may conclude that the design was similar to that of the much smaller temenos in Hebron, which is also attributed to Herod (although Josephus does not mention it).

Within the great walls, then, where would the temple have been? We may assume that the builders of the first Second Temple, who started work in 520 BC, would have known the location of the First and would have set their version in the same area. As we have seen, this would have been in the broad level area near the present Dome of the Rock. In the time of the First Temple, however, the king's palace was also here, and it, not the temple, was the dominant building. Since the lack of a king made a palace irrelevant in 520 BC, the temple had this area to itself, and so the builders probably placed it more centrally.

We are aided, on this question, by an old photograph and a map. (For what follows, see Jacobson, to whom I am indebted for much of this description. -- SL) Charles Wilson made a map of the Haram in 1864, and on it we can see a flight of 4 steps that has since disappeared. There is also an old photograph showing these steps. 

To judge from the photograph and Wilson's map, the steps were probably too high for human comfort. There are examples in classical Greek temples of such high steps (perhaps for the god); the pattern is typically interrupted by flights of shallower steps, friendly to mortals.

A broad Herodian staircase has been unearthed before the Double Gate (the westernmost of the two Hulda Gates) in the southern retaining wall. Its steps are parallel to the four on Wilson's plan. They are, however, mortal-friendly.


Near the four high steps on Wilson's map, there must also have been a staircase of mortal-friendly steps. Josephus wrote there were 14 steps from the lower platform to a terrace or rampart (hel ). (The Mishnah said 12.) These would have been the friendlier steps.

The terrace was 15 feet deep, says Josephus (and the Mishnah too). Then, he writes, came a high protective wall with gates in it. One ascended to these gates by another 5 steps. That would bring us to the level of the Court of the Priests, but still not to the height of the upper platform as it is today. Another 12 steps led up from the Court of the Priests to the floor of the sanctuary, which (figuring half a cubit per step, as the Mishnah has it) would have been slightly higher than the present platform.

If we look at Wilson's map of the Haram (below), we see that the Dome of the Rock appears off center. If it is standing where the Herodian temple was, the latter must have seemed central enough for Josephus to remark that it was "in the middle" (Jewish War V 188-212 [5.4]). Supposing therefore that the southern edge of the upper platform was south of where it is today, namely, at the top of the staircase that included the four steps on Wilson's plan, we would have indeed a more symmetrical impression, in keeping with Josephus.


We can see the same centrality on a more recent photo, if we place a line where the four steps used to be (No. 10 in the picture below).

There is another interesting feature. Of the great walls, the eastern and the western are not parallel. This is because the major part of the eastern wall goes back (probably) to the time of the First Temple (and certainly to the pre-Herodian Second). Herod could not build a new wall further to the east, because the Kidron Valley was too steep and deep (serving as a natural glacis). He therefore adopted the existing eastern wall, which makes an angle of 92 degrees to his new southern wall (rather than 90 degrees as in all the corners of the Hebron temenos). But why didn't he place the southern wall differently in order to get a perfect right angle? The answer has to do with his plan for the city as a whole: the streets appear to have been in a grid pattern, i.e., either parallel or perpendicular to one another. The temple mount's southern and western walls were built in accordance with the grid, but the eastern wall, as said, could not be changed.

And the north wall of the temple mount? Here we lack the courses of stone that enable identification with Herod. The north wall might have stretched along the line of the present-day Haram wall, or it might have kept to the grid pattern, intersecting the east wall at a point north of where the Haram wall does. At this point, Charles Warren noted, the Herodian courses in the eastern wall come to an end! I am tempted, therefore, to follow Jacobson in moving the north wall a bit. On the other hand, the change would leave the southeast corner of a great block of bedrock sticking awkwardly into the temple area. This is the block on which the Antonia Fortress had been built by Herod, probably a decade or so before he started rebuilding the temple. Such an eyesore seems un-Herodian. In short, when Herod cut the block on which he built the Antonia, he did so on a line which, when extended, met the old east wall at a perfect right angle. 

There were, we see, two conflicting units for determining the lines of Herod's temple platform: (1) the old east wall along the Kidron, and (2) the grid pattern of the city as a whole. Since the ancients did not do flyovers, and because of the enormous size, the irregularity of the temple platform probably went unnoticed, at least by hoi polloi.

Given, then, that the outermost east and west walls are not parallel, we cannot construct a north-south line midway between them, as we would like to do in an attempt to find the north-south axis on which the temple was built. On the other hand, following Jacobson, we can take the midpoint between the two Hulda Gates (the Double Gate and the Triple Gate), and then we can construct a line parallel to the west wall. This line crosses through the present-day Dome of the Chain (see No. 12 in the photo above, as well as the enlargeable map above.) Furthermore, if we revise the line of the mount's north wall in accordance with Jacobson, so that it is parallel to the south wall, we can then construct an east-west line midway between them. It too cuts through the Dome of the Chain! On this basis, Jacobson suggests that something important from the time of the temple must have stood at this point. It could not have been the Holy of Holies, because that would have left insufficient space on the east for the Court of Women (where both men and women were permitted, in contrast with the Court of Israel on a higher level to the west, where only men were permitted). But the main sacrificial altar would have fit well here. This would imply, however, that the memory of the altar's position was somehow preserved from the destruction of the temple (70 AD) until the Muslims built the Dome of the Chain about 600 years later. It seems unlikely that this memory would have survived the Roman and Christian centuries when Jews were not allowed into the city (except, for a part of that time, for one day each year) and when the former temple mount served as farmland and as a source of stones.

The following seems a likelier tale. We have evidence that the Muslims built a mosque on the south end of the platform, the end closest to Mecca, around 640 AD, soon after conquering the city. The midpoint in the south wall was lined up exactly between the Hulda Gates, which were then open and active (as we infer from the early Muslim lintel over the Double Gate). At this midpoint was a mihrab, called Omar's mihrab (still visible, and see the map at the bottom of this page). It was probably the line from this mihrab, parallel to the western wall, that formed one of the two axes determining the site of the Dome of the Chain. Slightly west of the latter, in 688 - 691 AD, the Dome of the Rock was built; it determined a new north-south axis, on which the Al Aqsa Mosque was centered in 711.

So what do we know about the location of Herod's temple? It was within the four great walls, and we can be confident that it was on the upper platform, where the underlying bedrock is level for a goodly stretch. If a Jewish pilgrim entered through the Double Gate, she had a chance to immerse herself in a mikveh shown on the map below. (There were two mikvehs on either side, no doubt to separate the sexes.) She then ascended a staircase of normal steps, which was surrounded on both sides by higher, monumental steps like those in the old photo. (Perhaps there was such an arrangement of steps on all four sides of the upper platform.) She now found herself either in the high, broad, level area or in a slightly lower one, built over the gentle slope to the east. We cannot pinpoint Herod's temple more closely. If it was built directly over the series of cisterns that appear on an east-west axis in the map below, then it was north of today's two Domes. Note too that these cisterns were dug into the broadest east-west stretch of level bedrock on the uppermost part of the hill.