A Hill of Mothers

What stood on the Esquiline before the church, and what the record can and cannot establish about why the hill stayed sacred.

1. The sequence

Four of those layers are excavated, four are inscribed, and the rest are literary. The distinction runs through the whole memo and is stated at each point rather than in a caveat at the end.

2. The ground

The Esquiline is the largest of the seven hills and carries three spurs: the Cispius to the north, the Oppius to the south, the Fagutal to the west. The Fagutal’s extent is uncertain and remains disputed.

The name may derive from ex colere, to live outside — the Esquilini as those beyond the settled core, against the inquilini within it. An alternative derivation is from aesculus, the Italian oak.

The hill was a burial field. Trench graves and cremations are attested from the early Republic, and the ground served the poor of the city until Augustus covered it and Maecenas laid gardens over the necropolis.

The Cispius lies between the Vicus Patricius toward the Viminal and the Clivus Suburanus toward the Subura. The Servian Wall and the Porta Esquilina closed its eastern end.

3. The king on the hill

Livy and Dionysius both record that Servius Tullius brought the Esquiline inside the city and moved his residence to it. Livy gives the motive: to increase the hill’s respectability.

The tradition adds that he would not permit patricians to live on the height and assigned them the valley below, which is the reported origin of the name Vicus Patricius. The king took the high ground and put the aristocracy at the foot of it.

By the same tradition Servius decreed that a coin be deposited at Juno Lucina’s shrine for every birth, and the child’s name entered in the record of all births in the city. The birth register and the temple were the same institution.

The reign ends on the same ground. Livy has Tullia driving up the Clivus Urbius toward the Esquiline, meeting her father’s body in the road, and driving over it; the street was afterwards called the Vicus Sceleratus, and Livy notes that the place still carries the memory. The hill keeps events in its street names as well as in its strata.

4. The groves

The oldest recoverable layer is woodland. Three named precincts belong to this spur.

Lucus Iunonis Lucinae, a grove of lotus trees. Varro dates the cult to Titus Tatius, the Sabine king, which places it before the city rather than within it.

Lucus Mefitis, grove and temple of an Italic goddess of vapours and the underworld, toward the Vicus Patricius. Festus derives the name of the Cispius from this zone.

Lucus Poetelius, named only in Varro and the Argei list. Some topographers place it on the height now occupied by the basilica. That is a conjecture and an old one.

None of the three has been excavated. Their positions are literary throughout, and the memo carries them as such.

5. The trees

Pliny lists both in his account of the oldest trees in Rome. The lotus in Lucina’s precinct he gives as about five hundred years old at the time of writing, which places its germination around 430 BC — before the temple.

His argument for its antiquity is etymological rather than observed: the tree must be older than the shrine, he says, because the goddess takes her name from the grove. That is reasoning from a name, and the memo records it as an argument, not a measurement.

The second tree is the lotos capillata, older still and of uncertain age even to Pliny, named because the Vestals’ shorn hair was brought to it.

Where the Capillata stood is not established in the sources consulted. The Vestals lived in the Atrium Vestae in the Forum, and the hair offering is not securely located on this hill. The resemblance between hair given to one tree and hair loosened at the other is a resonance, not a link.

Neither death is recorded. The trees are last attested around AD 70, and nothing states when they went, or whether the houses already pressing the grove in Varro’s day took them. By contrast the Ficus Ruminalis at the Lupercal is recorded as standing in Livy’s day and reduced to a vestige by Ovid’s. Rome noted the death of one sacred tree and not of these.

6. The temple

On 1 March 375 BC matrons dedicated an aedes inside the grove. Lightning struck it in 190 BC, damaging gable and doors. In 41 BC the quaestor Quintus Pedius built or rebuilt a wall around grove and shrine. Inscriptions show the cult still working under the Empire.

No temple walls have been found. The working location is the southern slope of the Cispius, west of Santa Prassede and north of the Clivus Suburanus. A minority view places the grove under the basilica itself.

The epithet Lucina was read two ways in antiquity: from lux, the light into which the child is brought, or from lucus, the grove. Both readings were current and the sources do not settle between them.

7. The festivals

The Matronalia joins two goddesses on purpose. Juno Lucina of the Esquiline is not Juno of the Capitoline; the first governs birth, the second marriage. Romans married in the second half of June, so a first-time bride would be near term by the calends of March — which was also the old New Year and the birthday of Mars.

At the rite women unbound their hair and untied every knot on their person, so that nothing should hinder the opening of the body.

The Argei station on the Cispius is identified not by the temple but by a dwelling: apud aedem Iunonis Lucinae ubi aeditumus habere solet — at the temple of Juno Lucina, where the temple-keeper usually lives. Somebody was resident on that spot.

8. Houses eat the groves

Varro lists the groves on this height while complaining that private greed had narrowed their boundaries. Dense housing is implied by 88 BC, when Sulla’s forces were stoned from the rooftops.

Excavation under and around the church confirms it: Republican and Hadrianic houses in two strata, Augustan-period walls beneath the basilica, a Roman road running diagonally under the later nave, rooms of the first to fourth centuries including a painted agricultural calendar of the late second century, and to the north the retaining walls and the Balnea Naeratii Cerealis, restored in 385.

The hill did not become a sealed precinct. It became a neighbourhood with a working shrine inside it.

9. The consul’s hall

Just off the summit, on the estate of Junius Bassus, consul in 331, a civic basilica was built. Its opus sectile panels survive in museums: Bassus in a chariot, Hylas, hunting groups.

In the fifth century the Gothic notable Valila gave the hall to the Church and it became Sant’Andrea. It was destroyed in 1930 for the Pontifical Oriental Institute.

Six halls of that size would fit inside the nave of Santa Maria Maggiore. The summit had the basilica form before it had a Marian church.

10. The columns

The colonnade is the clearest evidence on the site for assembly rather than construction. It is certain that the columns are spolia and probable that they came from more than one building, because they were not originally a matching set.

The fifth-century builders used them as found. They stayed mismatched for thirteen centuries, until Fuga pared down the thicker shafts and shortened the longer ones to bring them into line, and supplied new bases and capitals. The colonnade a visitor sees is 1750 work on fifth-century assembly of older stone. The original Ionic capitals are not preserved and are known only from historical description.

A claim in circulation holds that the columns came from a temple of Juno on the site. No source supports it. It is an attribution built backwards from the hill’s history, and it is the exact error this memo is written to avoid.

11. The church

Tradition speaks of a Liberian foundation and of snow falling on 5 August. What can be dated is Sixtus III, 432 to 440, immediately after the Council of Ephesus in 431 defined Mary as Theotokos.

He built, or rebuilt at monumental scale, on this summit, and lined the triumphal arch with infancy mosaics: Annunciation, Presentation, and an unusual Adoration in which the Child sits enthroned between two women while the Magi approach like foreign kings.

The programme is Christological and Marian after Ephesus. It does not reference the hill’s earlier cults and was not designed to. The later crib relic in the confessio continues the birth theme in a different register.

No pagan cella sits under the high altar. What sits under the floor is the Roman city: houses, a road, painted rooms, then the fifth-century church.

12. The throne

The arch mosaic is not a Marian scene. At its apex sits an empty throne bearing a jewelled crown and a jewelled cross, with a scroll sealed with seven seals resting on a stool at its foot. Peter and Paul flank it, with the four living creatures of the Apocalypse.

The motif is the hetoimasia, the prepared throne: the seat Christ is to occupy on his return. This is the earliest surviving Christian example, in the earliest major scheme of church decoration to survive. The arch measures roughly nine metres by sixteen.

The dedication of the church is Marian and the argument on the arch is Christological. The Council of Ephesus had not ruled on Mary for her own sake; Theotokos was a claim about the Child, that he was fully divine from conception and his mother therefore God-bearer. The scenes chosen demonstrate that divinity rather than narrate the infancy in order.

The Child is enthroned in the Adoration, not held, and the Magi approach as foreign kings approach a throne. The infant at the foot of the arch and the returning judge at its apex occupy the same seat.

Beneath the throne runs the inscription XYSTVS EPISCOPVS PLEBI DEI — Bishop Xystus, for the people of God. The builder put his own name directly under the seat reserved for Christ.

The arch has moved. It was originally the apsidal arch; in the thirteenth century the apse was set back five metres to make room for a transept, and this became the triumphal arch. The original apse mosaic is lost and what it showed is not recorded.

The arch is treated panel by panel in Reading the Arch, with the restoration record that governs any close reading of it.

Torriti’s replacement of 1295 gives Mary a throne of her own, crowned Queen of Heaven and seated beside Christ. Eight centuries separate the two thrones on the same axis.

13. What the sequence supports

The hill held female cult continuously from the archaic period to the fifth century: Juno Lucina and Mefitis on the Cispius, Diana attested on the Vicus Cyprius nearby, Minerva Medica giving her name to a district of it. Three of the four concern birth, bodies, or passage.

It also held the dead. Birth registry and burial field on the same height, outside the old wall, on the highest ground.

What the record supports is that a place used for one purpose remained in use for a related purpose across a change of religion. What it does not support is substitution. There is no text in which anyone decides to put Mary where Juno was, no reused cella, no dedication that names the older cult. The continuity is of ground and of function, and the memo claims nothing beyond it.

14. Sources

Reference codes read source.work.passage and resolve on the Master Source Register, which carries every source used across this series.

1. The triumphal arch — the hetoimasia at the apex, with crown, cross, the sealed scroll, Peter and Paul and the four creatures; the inscription XYSTVS EPISCOPVS PLEBI DEI; the thirteenth-century setting back of the apse by five metres. The reading of the inscription as linking Christ’s throne to the pope’s is an interpretation offered in the literature, not a statement the mosaic makes. The original apse mosaic is lost and its subject is not recorded. Supports §12.
2. Livy, Ab urbe condita 1.44 and 1.48 — Servius bringing the Esquiline into the city and moving his residence; Tullia on the Clivus Urbius and the naming of the Vicus Sceleratus. Supports §3.
3. Dionysius of Halicarnassus, Roman Antiquities 4.13 — the same expansion, independently. This is the second channel for §3 and the reason the residence is carried at a higher grade than the surrounding tradition.
4. Varro, De lingua Latina 5.45–54 — the Argei shrines, including the Cispius station apud aedem Iunonis Lucinae ubi aeditumus habere solet; the groves of the Esquiline; the complaint that private greed narrowed their boundaries; the attribution of the Lucina cult to Titus Tatius. Supports §4 and §7.
5. Platner and Ashby, A Topographical Dictionary of Ancient Rome (1929) — the Argei sequence at twenty-seven shrines, of which eleven of Varro’s twelve described stations can be located with confidence and the rest are conjectural; the working position of the Juno Lucina temple on the upper southern slope of the Cispius. A modern reconstruction, not ancient testimony. Supports §6 and §7.
6. Pliny the Elder, Natural History 16 — the lotus in Lucina’s precinct at about five hundred years old; the lotos capillata of uncertain age and the Vestals’ hair offering; the argument that the tree must predate the temple because the goddess is named from the grove. Supports §5.
7. Ovid, Fasti 3 — the Matronalia, the temple founding, and the unbinding of hair and knots. Supports §6 and §7.
8. Festus — the Cispius named from the Mefitis zone; the Septimontium list in which Fagutal, Oppius and Cispius appear as three of the seven montes. Supports §2, §4 and §7.
9. Lost or absent: the temple of Juno Lucina. No walls have been found and no excavated remains are attributed to it. Its position is established from the Argei sequence and from literary description alone. Everything this memo says about the temple rests on texts about a building nobody has seen.
10. Lost or absent: the original apse mosaic, replaced when the apse was set back in the thirteenth century. Its subject is not recorded and the dedication makes a Virgin and Child likely but unevidenced. Supports §12.
11. Lost or absent: the original Ionic capitals of the basilica colonnade, removed by Fuga and known only from historical description. Supports §10.
12. Destroyed: the Junius Bassus hall, demolished in 1930 for the Pontifical Oriental Institute. Its opus sectile panels are in museums; the building is gone. Supports §9.
13. The basilica fabric — forty-two Ionic columns across twenty-one bays, thirty-eight of Hymettian marble and four of grey granite, certainly spolia and probably from more than one building; the disturbance of the spacing by two campanili, probably in the twelfth century; Fuga’s paring, shortening and recapping for the Jubilee of 1750. Supports §10.
14. Research assistance. The initial topographical survey for this memo was prepared with an AI research assistant and has been checked against the sources named above. Where the survey and the primary source disagreed, the primary source is followed — the Capillata is a lotus, not a nettle tree, and the Argei shrines number twenty-seven, not twelve.
15. Verification register. Carried as reported and not settled: the extent and location of the Fagutal; the site of the lotos capillata; the positions of all three groves; the tradition that Servius excluded the patricians from the hill; the Servian birth-registry decree; and the claim that the basilica columns came from a temple of Juno, which is in circulation and which no source supports; and the subject of the lost apse mosaic. The dates of the trees’ deaths are not recorded anywhere.