# Legio IX (VIIII) Hispana on the Lower Rhine: the evidence and its dating

Working copies of the downloaded PDFs, extracted texts and database JSON (`rh_*`, `ru_*`, `edh_*`, `vanderveen2025.*`, `nar71.pdf`, `kemmers2006.*`, `britannia2025.*`, `chiron.*`, `zpe111_281.pdf`) are not included in this folder.

**Verification key used throughout:** `[V]` = I fetched and read the source itself (PDF/HTML/JSON); `[R]` = relayed at second hand from a source I did read (named); `[T]` = from memory only, not checked online. Page numbers for `[V]` items are the printed page numbers unless marked "pdf p.".

---

## 0. Executive summary

* The whole continental dossier for the Ninth consists of **six items**: (1) three stamped roof-tile fragments `LEG VIIII HISP` from the Hunerberg legionary fortress at Nijmegen (all from one die); (2) one stamped mortarium rim `LG VIIII HIS` from the military tile/pottery works at De Holdeurn (Berg en Dal/Groesbeek), 5 km SE of the fortress; (3) a silvered bronze harness pendant punched `LEG HISP / IX` from the villa site Ewijk-De Grote Aalst, c. 9 km W of Nijmegen; (4) the limestone altar to Apollo from Aachen-Burtscheid set up by L. Latinius Macer, *p.p. leg. VIIII Hisp., praef. castr.*; (5) indirectly, the c. 200 `VEX BRIT` tile stamps of a *vexillatio Britannica* in and around the same fortress; (6) very indirectly, a Holdeurn graffito `[...]COH II[...]NICA[...]`. Nothing else: no tombstone, dedication or diploma of the legion is known from Cologne, Bonn, Xanten, Mainz or anywhere else on the continent for the 2nd century (EDH searches for "Hisp" in Germania inferior/superior, Belgica, Lugdunensis; Graafstal 2025 n. 205 `[V]`). The only other continental texts naming the legion are careers of officers (Mersch/Lux., Tivoli, Baalbek) and a vexillation on Domitian's German expedition of 83 (ILS 1025), listed in §4.
* **None of the Nijmegen objects comes from a closed, independently dated context.** The dating is entirely a matter of historical reasoning around the garrison sequence of the fortress: X Gemina leaves c. 103/104; the fortress is then held by small units; the fortress is abandoned c. 125/130 (van der Veen 2025, coins and samian stamps `[V]`); the XXX Ulpia Victrix reaches Vetera c. 120/122.
* Two dating schools: **(a) c. 121–130** (Bogaers 1965/1967; Bogaers & Haalebos 1975; Sijpesteijn 1996; EDH dates the pendant 121–130), resting on the York gate inscription of 108, the British war of Hadrian's accession and the transfer of VI Victrix to York in 122; **(b) c. 104–120** (Haalebos 2000, followed by Kemmers 2006 "current theory", Willems & van Enckevort 2009, Hodgson 2021, van Enckevort 2021, Graafstal 2025, van der Veen 2025), tying the Ninth's men to the *vexillatio Britannica* that replaced X Gemina, and pointing out that the fortress was almost empty of finds after c. 104 and abandoned by c. 125/130.
* **Whole legion or vexillation?** Bogaers 1965/1967 explicitly concluded that the *whole* legion was briefly at Nijmegen (he thought the Ninth and the *vexillatio Britannica* were unrelated). Already in 1975 Bogaers & Haalebos withdrew this: the stone fortress (c. 16.5 ha) "offered insufficient room even for a full legion". Every Dutch specialist since (Haalebos 2000; Kemmers 2006; Willems & van Enckevort 2009; van Enckevort 2021; van der Veen 2025) and the recent British discussion (Hodgson 2021; Graafstal 2025) speak of a *detachment*, most probably as part of the *vexillatio Britannica*. Only Strobel 1984 (the rest of the legion moved to Nijmegen shortly after 120, following a detachment there since 102), E. Birley 1971 (legion to Nijmegen c. 126 after building at Carlisle), Nesselhauf 1967 (legion in Lower Germany 108–before 119) and, per Wikipedia, Campbell 2018 ("the Ninth Legion as a whole ... between 121 and 130") argue for the full legion.
* **No base for a full legion existed on the Lower Rhine after c. 104/122.** The Lower German army was reduced from four to two legions (Bonn: I Minervia; Vetera II/Xanten: VI Victrix, then XXX Ulpia Victrix from c. 120/122); Neuss ceased to be a legionary base c. 103; Nijmegen's stone fortress was too small for a legion and held only small units after 104 (van Enckevort 2021, 197: "it must have been too big to accommodate the smaller units after AD 103–104").

---

## 1. The Nijmegen (Noviomagus, Hunerberg) evidence

### 1.1 The tile stamps `LEG VIIII HISP`

**Number and die.** Three fragments of *tegulae*, all from inside the legionary fortress (none from the canabae or the Holdeurn works), all of **one and the same die**:

* Bogaers & Haalebos, Numaga 22 (1975) 16 afb. 11 `[V]`: "De twee enige fragmenten van dakpanstempels van de Legio IX Hispana die tot nu toe in Nijmegen, op het terrein van de castra, zijn gevonden (1966.1752 en 1959.169); ze zijn van hetzelfde, verder onbekende type: LEG(ionis) VIIII HISP(anae)" ("the only two fragments ... found so far ...; they are of the same, otherwise unknown type"). Drawing by E.J. Ponten, scale 1:2.
* Brunsting & Steures, OMROL 75 (1995) 95 fig. 5, type 167 `[R: van der Veen 2025 n. 518]`; Brunsting & Steures 1997, fig. 6.1 no. 167 `[R: Graafstal 2025 n. 205]` — the catalogue lists the stamps of the Ninth as a single type (no. 167) with three examples.
* Van der Veen 2025, 151 `[V]`: "Three stamps of Legio VIIII Hispana are known from the Hunerberg, all three from the legionary fortress. One of these is illustrated in fig. 6.6,1." Table 6.1 (p. 145): fortress 3 (0.1 %), settlement 0. Van der Veen adds (p. 151): "As no new stamps of Legio VIIII Hispana have come to light" since the 1995 catalogue (his own corpus covers the municipal excavations 1987–2022), i.e. the count has been stable at three since 1995.
* Bogaers 1965, 21 / 1967, 68 `[V]`: the Nijmegen stamp is unusually wide with unusually high relief, letters "fraaier uitgevoerd" than the British VIIII dies; "Der Nijmegener Stempel der legio IX Hispana ist keineswegs der gleiche wie die von Carlisle und Scalesceugh" (1967, 69). So the Nijmegen die is *not* one of the British dies (RIB 2462), although it shares the numeral form VIIII with the Carlisle/Stanwix/Scalesceugh/Leicester stamps, whereas York uses IX (RIB 2462 introduction `[V]`; Graafstal 2025 n. 216 `[V]`).
* The reading: Bogaers 1967, 63 n. 38 `[V]` explains that the last vertical stroke is only partly preserved but certainly belongs to VIIII (note the bar over the numeral); H. Brunsting had first published the 1959 piece in Numaga 7 (1960) 23, 27 Abb. 17 as a stamp of *legio VIII Augusta*. The EDH transcription (HD020257) is `Leg(ionis) VIIII Hisp(anae)`, diplomatic `LEG VIIII HISP` `[V]`.

**Find contexts (none closed/dated):**

1. Find no. **1959/169**: Bogaers 1967, 63 `[V]`: "Eine besondere Überraschung bei den Ausgrabungsarbeiten des Jahres 1959 war der Fund eines ziemlich stark verwitterten Bruchstücks einer tegula ... LEG VIIII ---. Dieses Stück stammt aus dem Laufniveau oder der darüber liegenden Zerstörungsschicht, die zur Periode III gehören, und zwar aus der dritten Offizierswohnung nördlich des großen steinernen Torgebäudes an der Ostseite des Lagers" (= Bogaers & Haalebos 1975 n. 47: "uit het woonniveau of de daarop gelegen afbraaklaag van de jongste bouwperiode (5), uit de derde officierswoning ten noorden van de porta principalis dextra"). Period III (1967 numbering) = period 5 (1975 numbering) = the stone fortress, begun "relatief laat in de Flavische tijd ..., na 88" and probably in the first years after 100 (Bogaers & Haalebos 1975, 15–16 `[V]`). Van der Veen 2025, 153 `[V]`: "the first was found in a rubble layer together with many other stamped tiles; these included three of Vexillatio Britannica, but these were far outweighed by stamps of Legio X Gemina."
2. Find no. **1966/1752**: "fragment van een tegula, afkomstig uit de uitbraaksleuf van de westmuur der principia" (Bogaers & Haalebos 1975 n. 47 `[V]`) — a robber trench, i.e. post-abandonment fill; van der Veen 2025, 153: "found in a robber trench associated with the western wall of the principia, together with two stamps of Legio X Gemina."
3. Third fragment (recorded by Brunsting & Steures 1995 from the RMO excavations 1950–1967): "not found in association with other stamps" (van der Veen 2025, 153 `[V]`).

Van der Veen's verdict (p. 153): "The find contexts of the stamps provide no additional information in the matter whether Legio VIIII Hispana was present as part of Vexillatio Britannica or on its own." His distribution analysis (p. 154, 177): one of the three lies in the NE industrial quarter (with 38 VEX BRIT stamps and only one of XXX), one in the principia area.

**Database entries:** EDH **HD014357** is the Aachen altar (below); the tile is EDH **HD020257** `[V]` (findspot "{Hunnerberg}, Lager, Porta principalis dextra, nördlich, 3. Offizierswohnung", year of find 1959, "Ziegel mit Stempelabdruck", dated 71–130, lit. AE 1977, 541; Bogaers & Haalebos, Studien zu den Militärgrenzen Roms II (Xanten 1977) 107–109 Abb. 6; Trismegistos 209140). So the tile's *Année épigraphique* number is **AE 1977, 541**. Not in CIL XIII (found 1959). EDCS: **EDCS-09301104** `[V]` (record verified via the JSON back end: place 420248 Nijmegen; text `Leg(ionis) VIIII Hisp(anae)`; lit. AE 1977, 541; no date assigned).

### 1.2 The Holdeurn mortarium stamp `LG VIIII HIS`

* Bogaers 1965, 17 afb. 6 / 1967, 63–64 Bild 6, Taf. 5,4 `[V]`: rim sherd of a red coarse-ware mortarium, found **May 1938** in the first excavations of J.H. Holwerda and W.C. Braat at Groesbeek-De Holdeurn (the military tilery/pottery of the Nijmegen garrison), Rijksmuseum G.M. Kam (now Museum Het Valkhof), unpublished until Bogaers; stamp complete, impressed *parallel* to the rim: `LG VIIII HIS = l(e)g(ionis) (nonae) His(panae)`. Bogaers argued from fabric (Holdeurn ware, "baksteenachtig") and stamping habit that it was made at the Holdeurn, not imported from Britain (1965, 22–23). Mrs K. Hartley thought the rim form late 1st/early 2nd century but unparalleled in Britain (1965, 23). Van der Veen 2025, 151 fig. 6.6,2 `[V]` reads "LG VIIII HISP". NAR 71 (2021) table p. 81 lists it under Berg en Dal-Holdeurn: "9th Legion LG.VIIII.HIS 104–120 mortar stamp" `[V]`.
* This is the second of the "two" 1967 objects (Eck 1972, 460: "eines Legionsziegels und einer Reibschale") `[V]`. No AE/EDH number found (EDH search returned only the tile and the pendant for Nijmegen).

### 1.3 The Ewijk harness pendant `LEG HISP / IX`

* **P.J. Sijpesteijn, "Die legio nona Hispana in Nimwegen", ZPE 111 (1996) 281–282** `[V, PDF]`: bronze pendant 5.6 × 5.3 cm with hook (1.7 cm), silvered, once attached to a *phalera*; two-line punched inscription `LEG HISP / IX` (letter heights 0.7 / 0.5 cm), "Leg(io) Hisp(ana) IX". Shown to him by a private finder, who "im E-wijk (d. h. im E-Viertel) in Nimwegen gefunden hat" — Sijpesteijn evidently misunderstood the village name **Ewijk** (municipality Beuningen, c. 9 km W of Nijmegen) as a "quarter E" of Nijmegen; the finder withheld details; Sijpesteijn persuaded him to donate it to the Rijksmuseum G.M. Kam. He follows Bogaers' 121–130 dating and remarks that the new piece uses IX while the three older objects use VIIII: "Es sieht so aus, als ob der Besitzer dieses Anhängsels das Stück aus England mitgebracht und in Nimwegen verloren hat."
* AE 1996, 1107; also AE 2009, 925–928 (from Willems & van Enckevort 2009, 152 fig. 54); EDH **HD050314** `[V]` (findspot given as Nijmegen following Sijpesteijn; "Anhänger, der ursprünglich zu einer Phalera gehörte. Schreibtechnik: gepunzt"; bronze/silver; 5 × 5 cm; dated **121–130**; Trismegistos 209684); EDCS **EDCS-03000651** `[V]` (record verified: place 420248 Nijmegen; text `Leg(io) Hisp(ana) / IX`; lit. ZPE 111, 281; Willems & van Enckevort 2009 [Ulpia Noviomagus] p. 152; AE 1996, 1107; AE 2000, +1011 [presumably the AE notice of Haalebos 2000 on the Ewijk findspot — content not seen]; dated 121–130; photo AE_1996_01107.jpg).
* The true findspot is the **Roman villa at Ewijk-De Grote Aalst**: van der Veen 2025, 151 `[V]` ("a horse gear pendant from the Roman villa at Ewijk-De Grote Aalst with punctim inscription LEG HISP IX (fig. 6.6,3). The latter can be dated before c. AD 120 based on its trefoil shape and construction with separately attached, open loop. Haalebos suggests that the horse pendant belonged to a veteran that settled there", citing Haalebos 2000f, 472–473 and 488 Abb. 11); van Enckevort 2021 (NAR 71) fig. 6.11 `[V]`: "Pendant from cavalry equipment from Ewijk with punched inscription reading LEG(io) IX HISP(ana). Width 5,9 cm"; NAR 71 table p. 78: "Ewijk, 9th Legion Hispana, LEGVIIII HISP [sic], 104–120, punched inscription pendant horsegear". Livius.org caption: "Bronze object mentioning LEG HISP IX, from Ewijk" `[V]`; Museum Het Valkhof, Nijmegen.
* Dating basis: typology (trefoil pendant with separately attached open loop = before c. 120, van der Veen) plus the general horizon; EDH's 121–130 simply follows Sijpesteijn/Bogaers.

### 1.4 The `VEX BRIT` stamps and the vexillatio Britannica

* Numbers: "Ruim honderd dakpanstempels, d.i. meer dan 5 % van het totale aantal uit de Nijmeegse legerplaats" (Bogaers & Haalebos 1975, 16 `[V]`; Bogaers 1965, 17: "ruim honderd ... iets meer dan 5 %"). Van der Veen 2025, table 6.1 `[V]`: **161** in the fortress (4.2 %) and **35** in the canabae (3.2 %) = 196; the second most common unit after X Gemina; type 182 of Brunsting & Steures 1995; only three VEX BRIT stamps at the Holdeurn works itself (van der Veen 2025, 172). Two VEX BRIT tegulae lined a cremation burial on the Kops Plateau datable only "after c. AD 80" (van der Veen 2025, 135). Corpus numbers: the pre-1933 finds are **CIL XIII 12553, 1a–c and 2** = EDCS-69700443 (Nijmegen, 12553,1a), EDCS-69700201 (Holdeurn, 1b), EDCS-69700225 (Hunerberg, 1c), EDCS-69700444 (Nijmegen, 12553,2) `[V]`; note that EDCS expands the stamp as `Vex(illarii) Brit(tonum)` — the older alternative reading (a vexillation *of Britons*), not the now standard *vexillatio Britannica*; a further example from Berg en Dal is Schmidts 2018 cat. 8-1, 4g1 = EDCS-80000127 `[V]`, and one from Ewijk EDCS-80000135 (below).
* Composition: Bogaers 1965/1967 argued from CIL III 4466 (Carnuntum: eques of *ala (I Pannoniorum) Tampiana vexillationis Britannicae*, sent to the Danube c. 85/86) and CIL XIII 8805 (Hemmen: *dec. alae Vocontiorum exercitus Britannici*) that the Nijmegen vexillatio consisted of **British auxiliary cavalry**, not legionaries, and that it went back to Britain before 17 July 122 (diploma CIL XVI 69 listing the ala Tampiana in Britain) (1967, 66–68; 74–75 `[V]`). Strobel 1984, 99–101 `[R: van der Veen 2025, 134]` instead reconstructed a Trajanic vexillatio of c. 3,000 legionaries drawn from II Augusta, VIIII Hispana and XX Valeria Victrix plus auxiliaries, sent in 102. Haalebos 2000, 472 `[R: Graafstal 2025 nn. 208–210; van der Veen 2025 n. 578]`: the vexillatio, "of which Legio VIIII Hispana formed a part", came c. 103/104 and was replaced c. 120 by a detachment of XXX. Graafstal 2025, 47 `[V]`: "The strong likelihood is that a detachment of IX Hispana, the British legion logistically closest to the Rhine delta, was part, and in charge, of that vexillatio."
* Date of the stamps: Brunsting & Steures 1995, 104–105 and 1997, 323 put VEX BRIT in the **AD 90s, contemporary with X Gemina** (same die-cutter style of the X; association of 104 X Gemina and 34 VEX BRIT stamps around horreum II) `[R: van der Veen 2025, 134–135; Graafstal 2025 n. 209]`; rejected by Haalebos 2000, 476 (Graafstal: "rightful criticism") and by van der Veen 2025, 135 (rough dies show no individuality; later repairs; the second "reliable" context rests on "a debatable exegesis of handwritten notes"). Van der Veen's own chronological table (p. 295 `[V]`): "AD 104–122 Vexillatio Britannica (incl. Legio VIIII Hispana?) maintenance to industrial area, barracks, horreum II and campus/forum".
* Holdeurn graffito `[...] COH II [...] / [...] NICA [...]`, "completed very tentatively as [...mil(itis)] coh(ortis) II [...ex vex(illatione) Britan]nica: Haalebos 2000f, 475 and 489, Abb. 13" (van der Veen 2025 n. 282 `[V]`); NAR 71 table p. 81: "vexillatio of Second Cohort of the Britons COH II BRITAN 104–120 graffito". Too fragmentary to build on.
* A `VEX BRIT` stamped tile is also recorded from **Ewijk** itself — i.e. from the villa area that produced the IX Hispana pendant: **EDCS-80000135** `[V]` (place 801687 Ewijk; text `Vex(illatio) Brit(annica)`; lit. T. Schmidts, *Gestempelte Militärziegel außerhalb der Truppenstandorte* (Wiesbaden 2018) cat. 8-1, no. 7). Military tile at a villa 9 km west of the fortress fits the pattern of army building material reaching the countryside; it does not date the pendant.

### 1.5 The Hunerberg fortress after X Gemina: what was there, and until when

**Size.** Bogaers & Haalebos 1975, 12 `[V]`: the timber fortress of period 3 was c. 29 ha (room for a legion plus auxilia); the period-4 fortress "ruim 15 ha"; the stone fortress of period 5 "ruim 16,5 ha" (c. 40 m longer and 5 m wider than period 4); "In het nieuwe kamp ... was geen plaats voor auxilia; zelfs voor een voltallig legioen bood het niet voldoende ruimte. Een 'normale' castra legionis had een oppervlakte van minstens 18–20 ha." From this they concluded (p. 17) that the Ninth, having occupied the period-5 fortress, "wegens de te geringe grootte van de castra niet de volle sterkte van een legioen kan hebben gehad" — the first explicit statement that only part of the legion was at Nijmegen.

**Departure of X Gemina.** c. 103/104 to Aquincum (Bogaers & Haalebos 1975, 3; Kemmers 2006, 238: "between 71 and 103/104"; van der Veen 2025, 69: exact date unknown, 101/102–106 possible, "c. AD 104 will be retained"; Strobel 1984: 102). Brohl-valley altars of 101–103 by Lower German detachments and CIL XIII 7718 (leg. X G.P.F., 96–104) show building activity at the very end of the Tenth's stay (Bogaers & Haalebos 1975, 16).

**End of the canabae and fortress — the numismatic/ceramic evidence.**
* Kemmers 2006 (diss. Nijmegen 2005; Studien zu Fundmünzen der Antike 21) `[V, PDF]`, p. 106: "Current theory has it that in the period 104–120 (parts of) the Vexillatio Britannica, among which a detachment of the Ninth Legion Hispana, was located in Nijmegen. After their departure detachments from the Thirtieth Legion Ulpia Victrix, stationed at Xanten, took their place. Around the middle of the second century the site seems to have been abandoned altogether, though a few finds indicate some kind of activity in the (middle of) the third century." Her own results for the **canabae** (p. 109): under 1 % of the stamped samian datable after 105; Central/East Gaulish samian only marginal; Trajanic coins after 103 far fewer than before; Hadrianic and Antonine coins few (n = 17) and heavily worn, "explained by a halt of the arrival of fresh coinage on the site around the middle of the first decennium of the second century, the Hadrianic and later coins having circulated elsewhere previously and then brought to the site somewhere in the Antonine period by occasional inhabitants, perhaps detachments of neighbouring legions"; conclusion: "inhabited from 70 onwards until approximately 105, with some marginal traces of habitation after that date"; p. 238: "the inhabitants of the canabae legionis left together with, or shortly after, the Tenth Legion."
* Van der Veen 2025 (Radboud diss., open access) `[V, PDF]`, ch. 4 (coins and samian stamps from fortress *and* canabae): "There are many differing opinions on the end date of the fortress, ranging from AD 103/4 to 175" (p. 112); the Nijmegen samian curve "rises again in AD 125–130 ... leading to a peak around AD 150", which reflects the town of Ulpia Noviomagus, not the Hunerberg: "This strongly suggests that the fortress was abandoned not much later than AD 125–130" (pp. 104, 112); "Structural occupation of the legionary fortress seems to have ended around AD 125/130" (p. 303); the settlement possibly a few years later; no continuity to the late-3rd/4th-century activity. Tile stamps agree (p. 153): 227 post-104 stamps in the fortress, "the vast majority ... dated before c. AD 130 (Vexillatio Britannica, Legio VIIII Hispana and Legio XXX Ulpia Victrix), while only six are dated to the late 2nd and 3rd centuries"; coins AD 150–250 almost absent (n. 586 lists them individually: one sestertius of Marcus 163/164 and one denarius of Elagabalus from the fortress). His summary table (p. 295): "after c. AD 120 Legio XXX Ulpia Victrix: maintenance to principia and courtyard building; c. AD 125/130 abandonment fortress and settlement, with exception of campus/forum and amphitheatre".
* Legio XXX at Nijmegen: 60–64 stamps, nearly all one type `LEGXXX.VV` (Brunsting & Steures 1995 read the single type as one consignment for small repairs c. 121 and denied a detachment); a votive altar of a *frumentarius leg. XXX U(lpiae) [V(ictricis)]* found 2000 east of the principia (Haalebos 2000b, 50; 2000e, 25–26; 2000f, 474, 488 Abb. 12); Ernst 2016 dates the `Ulpia Victrix` stamp type "shortly after the arrival of the legion in Lower Germany c. AD 120". Van der Veen 2025, 138: "it seems reasonable to assume that a vexillatio of Legio XXX was present at Nijmegen for a short period sometime after AD 120."
* Older views of the end: Bogaers 1965, 14 / 1967, 62: terminus ante c. 175 (no *Exercitus Germaniae inferioris* stamps in the castra; Holdeurn kilns restarted c. 175) — already qualified in his 1967 postscript when an EX GER INF stamp turned up in the castra (van der Veen 2025, 70). Brunsting & Steures 1995, 107–109: last repairs in or shortly after 121; fortress demolished shortly after 125. Haalebos 2000f, 465, 477 and Willems & van Enckevort 2009, 57: castra in use at least until Hadrian, possibly into the second half of the 2nd century (on a handful of Antonine coins — van der Veen n. 297: a coin of Marcus 163/164, one of Verus, one of Julia Domna, c. 37 mid/late-2nd-century sherds). Van Enckevort 2021 (NAR 71, 197–198 `[V]`): "After its [X Gemina's] departure, new units were stationed in the castra, probably until well into the third quarter of the second century AD. Tile stamps of the Vexillatio Britannica and a unit of the Legio VIIII Hispana show that the new residents came over from Britain, to replace the original garrison ... Around AD 120, a detachment of the Legio XXX Ulpia Victrix was transferred to the castra, and stayed until the third quarter of the second century AD ... Although the layout of the castra remained unchanged, it must have been too big to accommodate the smaller units after AD 103–104 ... Three stamps of the Exercitus Germaniae inferioris indicate that the regular Roman army returned to the Hunerberg c. AD 160–170. The abandonment of the castra shortly after might have been connected with the pandemic outbreak of smallpox (Antonine Plague) in AD 166, ... the raids of the Chauci ... AD 172–174, the revolt of Maternus in AD 185–186, or ... 193–197." Van der Veen (2025, 156) reads the same six late stamps as "at best ... a small military force briefly stationed in the fortress at that time".
* No dendrochronological dates bear on the end of the fortress (none is cited by van der Veen, Kemmers or van Enckevort); the end date rests on coins, samian stamps and tile stamps.

**Answer to "was any base available for a whole legion on the Lower Rhine after c. 122?"** — No. Neuss ceased to be a legionary base when VI Victrix moved to Vetera II c. 103 (NAR 71 table p. 83: "Neuss (AD 70–103?), Vetera II (AD 103–121)" `[V]`); Vetera II housed VI Victrix and then XXX; Bonn housed I Minervia; Nijmegen (16.5 ha) could not hold a full legion and its finds after 104 are those of small units (Bogaers & Haalebos 1975, 12, 17; van Enckevort 2021, 197; van der Veen 2025, 134: "since only very few finds from the Hunerberg can be dated after AD 104, it is unlikely that such a large force [Strobel's 3,000] was stationed on the Hunerberg after that date"; p. 153: "it is unlikely that Legio X was replaced at Nijmegen by anything approaching the size of a full legion"). Graafstal 2025 n. 213: "The evacuation of Nijmegen by the Tenth was part of a stepwise degrading of Lower German garrison from four to two legions."

### 1.6 The argued date ranges and their bases — scholar by scholar

| Author | Date for the Ninth at Nijmegen | Whole legion or part | Basis |
|---|---|---|---|
| Bogaers 1965 (Numaga 12, 10–37) = 1967 (BJ Beih. 19, 54–76) `[V]` | "um 121" for a "verhältnismäßig kurze Zeit"; end of legion "not before c. 130" (1965, 25); possibly lost in the Jewish war 132–135 | **whole legion**: "diese Legion kurze Zeit in ihrer Gesamtheit im Lager von Nijmegen gelegen hat" (1967, 73; Dutch 1965, 24 "in zijn geheel"); the Ninth and the VEX BRIT auxiliaries unrelated | York inscription 107/108 (legion still in Britain; not moved under Trajan); Hadrianic British war ended by 119 (coins); Platorius Nepos legate of Lower Germany c. 121, Britain 122; VI Victrix from Vetera II to York 121/122; VEX BRIT back in Britain before 17 July 122 (CIL XVI 69); CIL X 1769 (Aelius Asclepiades, Cilician, miles leg. IX) implies later service in the East; Holdeurn kilns restart c. 175 = final evacuation |
| Nesselhauf & von Petrikovits 1967 (BJ 167, 268–279) `[T/R]` | altar (and legion in Lower Germany) **108–before 119** "aufgrund der Truppenstationierung" | (whole legion implied) | relayed by de.wikipedia "Porcetum" citing the article (which I could not open: Heidelberg journals server blocked); Eck 1972 n. 8 and E. Birley 1971, 75 f. discuss Nesselhauf's reconstruction |
| E. Birley 1971 (in Butler ed., Soldier and Civilian in Roman Yorkshire, 71–80) `[R: van der Veen 2025, 152; Graafstal 2025, 27]` | legion to Carlisle in early 122 to build a fortress, then to Nijmegen **c. 126**, later East (Judaea or Elegeia 161) | whole legion | VI Victrix's arrival at York 122; VIIII stamps at Carlisle |
| Eck 1972 (Chiron 2, 459–462) `[V]` | "wohl im zweiten Jahrzehnt des zweiten Jahrhunderts" the unit belonged to the garrison of Germania inferior | not discussed (speaks of "diese Einheit") | Bogaers' tile + mortarium, confirmed by the Aachen altar; the cursus of Q. Numisius Iunior (cos. Feb. 161; tribune of IX not before c. 140/141) shows the legion existed under Antoninus Pius |
| Bogaers & Haalebos 1975 (Numaga 22, 1–20) `[V]` | VEX BRIT c. 104–121, then "vermoedelijk van ongeveer 121 tot 130, de Legio IX Hispana", then XXX to c. 175 | **not full strength** (fortress too small) | stamp counts (two IX, >100 VEX BRIT) and size of stone fortress |
| Strobel 1984, 99–101, 444, 453 `[R: van der Veen 2025]` | detachment of IX in the VEX BRIT from **102**; the rest of the legion to Nijmegen **shortly after 120** to replace it | detachment, then whole | Dacian-war troop movements |
| Keppie 1989/2000 `[R: en.wikipedia]` | tiles "cannot be securely dated", suggests **c. 105** during a temporary absence of the legion from Britain | (temporary absence) | — |
| Sijpesteijn 1996 `[V]` | follows Bogaers: 121–130, "als sie von Britannia nach Nimwegen versetzt wurde ... als sie nach Osten geschickt wurde" | whole legion | Bogaers |
| Haalebos 2000, "Römische Truppen in Nijmegen" (Le Bohec & Wolff 2000, 465–489) `[R: Graafstal 2025 nn. 17, 205–210; van der Veen 2025 nn. 570, 578; NOT accessed]` | **c. 103/104–120**: detachment of the Ninth as part of the vexillatio Britannica, replaced c. 120 by a detachment of XXX (p. 472); Macer perhaps *praefectus castrorum* designate (p. 472); pendant = a veteran settled at Ewijk (pp. 472–473, 488 Abb. 11); VEX BRIT not Domitianic (p. 476, contra Brunsting & Steures); castra possibly into the 2nd half of the 2nd c. (p. 477) | **detachment** | departure of X Gemina; stamp associations; historical plausibility |
| A.R. Birley, Roman Government of Britain (2005) 228–229 `[R: de.wikipedia; en.wikipedia via Graafstal n. 206, 217]` | vexillation at Nijmegen "um 105, oder zwischen 113 und 117"; triggered by X Gemina's departure | vexillation | — |
| Kemmers 2006, 106 `[V]` | "current theory": 104–120 (parts of) the vexillatio Britannica incl. a detachment of the Ninth | detachment | coins/samian: canabae largely abandoned by 105 |
| Willems & van Enckevort 2009 (JRA Suppl. 73) 56–57, 152 `[R: en.wikipedia; van der Veen 2025 n. 296; Raepsaet's review in AC 80, 2011 `[V]`]` | after 104 a vexillation from Britain "consisting of parts of VIIII Hispana and other units"; replaced by a detachment of XXX "not long after AD 120"; castra at least until Hadrian | detachment ("vexillations détachées de la IXe Hispana et de la XXXe Ulpia Victrix") | — |
| Campbell 2018, The Fate of the Ninth, ch. 11 `[R: en.wikipedia only]` | senior officers such as Macer "convinced several scholars that the Ninth Legion as a whole was based there between 121 and 130" | whole legion (as summarised) | Aachen altar |
| Hodgson 2021 (Britannia 52, 97–118) `[V, abstract only]` | "no more than a detachment of the legion went abroad, and that only for the period c. 105 to c. 120; ... no prosopographical or other evidence which proves that the legion existed after the early 120s" | **detachment** | Nijmegen archaeology; prosopography |
| van Enckevort 2021 (NAR 71, 196–198) `[V]` | new residents "came over from Britain" after 103/104; XXX from c. 120; castra too big for the small units; legion table (p. 83): "Nijmegen, Legio IX Hispana, AD 104/106–AD 130, temporarily at Nijmegen after transfer from Britannia" (the table's "AD 130" and "Later to Danube" are loose) | small units / "a unit of the Legio VIIII Hispana" | stamps; size |
| Graafstal 2025 (Britannia 56, 25–55) `[V]` | detachment with the VEX BRIT from c. 104; "It cannot be excluded that a vexillatio Britannica ... was still based at Nijmegen by the death of Trajan"; any remaining men repatriated by c. 119/120 for the Wall project | detachment ("the archaeology of the legionary fortress does not support the presence of a complete legion after c. 104", p. 47) | Haalebos 2000; van Enckevort 2021; York rebuilt in stone 107/8 (RIB 665) shows York remained the home base |
| van der Veen 2025 `[V]` | with the VEX BRIT 104–122(?); fortress abandoned c. 125/130 | detachment; "a full legion" excluded by the finds | coins, samian stamps, tile-stamp counts and contexts |
| Russell 2011 (BBC History Magazine) `[R: en.wikipedia]` | the Nijmegen tiles "all seem to date to the AD 80s" | — | contradicted by the contexts: the tiles lie in/over the stone fortress built after 88 and probably after 100, and the Ninth had no known presence in Lower Germany in the 80s (only vexillarii on the Chattan campaign, ILS 1025) |

---

## 2. The Aachen (Aquae Granni / Burtscheid) altar of L. Latinius Macer

**Text** (EDH HD014357 `[V]`; EDCS-10700516 `[V]`; identical in both):

```
L(ucius) Latinius L(uci) f(ilius) Publilia / Macer Ver(ona) p(rimus) p(ilus) leg(ionis) VIIII Hisp(anae) /
praef(ectus) castr(orum) pro se et suis / Apollini / v(otum) s(olvit) // l(ibens) m(erito)
```
Diplomatic (EDH): `L LATINIVS L F PVBLILIA / MACER VER PP LEG VIIII HISP / PRAEF CASTR PRO SE ET SVIS / APOLLINI / V S / / L M`. EDH expands `p(rimi)p(ilus)`, EDCS `p(rimus) p(ilus)`; `Ver(ona)` = origo (tribus Publilia/Poblilia is Verona's tribe).

**Publication numbers:** **AE 1968, 323** (from P.-M. Duval, REA 70, 1968, 430 no. 43, 14, reporting H. Nesselhauf & H. von Petrikovits, "Ein Weihaltar für Apollo aus Aachen-Burtscheid", **Bonner Jahrbücher 167 (1967) 268–279** — the volume/year is confirmed by Eck 1972 n. 8 `[V]`; de.wikipedia's "BJ 179, 1979" is an error); U. Schillinger-Häfele, BRGK 58 (1977) 528–529 **no. 144** (EDCS photo files "TR_Schillinger_00144_1-4"); AE 2005, 1061 (EDH) / AE 2005, +1060 (EDCS), from C. Keller, Archäologische Forschungen in Aachen. Katalog der Fundstellen in der Innenstadt und in Burtscheid (Rhein. Ausgr. 55, Mainz 2004) 80, no. 648/019, Abb. 52. **Not in CIL XIII** (found 1957). EDH **HD014357** (Trismegistos 208961); EDCS **10700516** (the EDCS text search for "Latinius Macer" returns exactly this id; record verified: place 420002 Aachen/Aquae Granni; date 101–200). Bogaers 1967, 76 "Nachtrag" `[V]` records that von Petrikovits announced the find at the 1964 Limes Congress discussion ("vor einigen Jahren in Aachen-Burtscheid eine Reliefdarstellung des Apollo (höchstwahrscheinlich des Apollo Grannus) ... von einem zum 'praef(ectus) castr(orum)' beförderten 'p(rimus) p(ilus) leg(ionis) VIIII Hisp(anae)' geweiht ... Die Inschrift soll demnächst veröffentlicht werden").

**Findspot and object:** EDH: "Burtscheid, {Burtscheider Markt 22}", found **1957**, limestone (de.wikipedia "Porcetum" `[V]`: Lorraine limestone, found "bereits 1957 bei Ausschachtungsarbeiten in der Rehaklinik Schwertbad", five-fold profiled base, 87.5 × 57 cm; relief of Apollo seated with kithara, bow and quiver on his back), height (87) cm, width 57 cm, depth (43) cm; now **LVR-LandesMuseum Bonn**. Burtscheid = the Roman *Porcetum* spa quarter with thermal springs; foundations of a nymphaeum of Apollo Grannus were found in 1969 at the Burtscheider Markt, the altar being taken as part of that spring sanctuary (de.wikipedia, citing H.G. Horn, Die Römer in NRW (1987) 328 f. and Keller 2004).

**Dating arguments:**
* Nesselhauf & von Petrikovits 1967: "die Zeit von 108 bis vor 119" (de.wikipedia `[R]`: "Aufgrund der Truppenstationierung datierten ... den Altar auf die Zeit von 108 bis vor 119") — i.e. they assumed the whole legion was in Lower Germany between the York inscription of 108 and the British troubles/VI Victrix's transfer; E. Birley 1971, 75 f. criticised Nesselhauf's reconstruction (Eck 1972 n. 8). I could not read the BJ article itself (the Heidelberg digital journals server refuses non-browser access); this dating is therefore relayed, not verified.
* EDH/EDCS: 101–200 (generic). No published letterform-based dating found; the databases give none.
* Eck 1972, 460: the altar "weiter bestätigt" the Nijmegen finds — presence of the legion in Germania inferior in the 110s.
* Haalebos 2000, 472 (relayed by Graafstal 2025 n. 205): Macer "perhaps designate" as camp prefect — i.e. a *primus pilus* of the Ninth's detachment who was to take up the *praefectura castrorum*.
* Graafstal 2025, 47 n. 213 `[V]`: the posting of a detachment of the Ninth "had been a temporary arrangement from the outset, perhaps intended to finish a series of building projects that had been started by the Tenth. Hence perhaps the unexpected presence of a praefectus castrorum" — i.e. the camp prefect (the legion's chief engineer/quartermaster) headed the building detachment; Macer's altar is "likely connected" to the Nijmegen presence (n. 205).
* Hodgson 2021: abstract only; his treatment of Macer not verified.

**Serving camp prefect or veteran at the spa?** The stone itself only says that a *primus pilus* who had become *praefectus castrorum* of IX Hispana paid a vow to Apollo "for himself and his family" at the healing springs of Aquae Granni/Porcetum; no "veteranus"/"evocatus"/"missus" formula, and *praefectus castrorum* was normally a post held in a legion by ex-primipili (Wikipedia's aside that the primipilate led "on completion of their single-year term" to the prefecture is a simplification). The mainstream reading (Nesselhauf–Petrikovits; Eck; Haalebos; Graafstal; the EDH commentary chain) is that he was the *serving* camp prefect (or designate) of the unit then in Lower Germany, taking the waters c. 130 km from Nijmegen. The alternative — a retired or seconded officer visiting the spa from Britain (York–Aachen is a long way; Aachen lay off the Britain–Rhine route only marginally) — is not, to my knowledge, argued in print by any of the sources read, but nothing in the text excludes it. Note that the origo Verona and tribe Publilia say nothing about his posting.

---

## 3. Other continental attestations of the legion (none from the Rhine garrison)

* **Hemmen (Betuwe, NL), altar to Dea Vagdavercustis**, CIL XIII 8805 = ILS 2536 = EDH HD082661 `[V]` (found 1842 in the river Linge; dated 51–150 in EDH): `Deae Vagdavercusti Sim[p]licius Super dec(urio) alae Vocontior(um) exerci[t]uus Britannici`. The key text for the *vexillatio Britannica* (Bogaers 1967, 66–68). Does not name the Ninth. EDCS **EDCS-11100895** `[V]` (place 420129 Hemmen; dated 51–150).
* **Carnuntum (Petronell), tombstone of T. Flavius Crescens**, CIL III 4466 = ILS 2515 = EDH **HD073876** `[V]` (EDH date 89–101; Vorbeck, Militärinschriften aus Carnuntum no. 283): `T(itus) Flavius Cre{n}sce(n)s equ(es) al(a)e Tam(pianae) vex(illationis) Brit(annicae) an(norum) XXX st(ipendiorum) XV dom(o) Durocor(toro) Rem(orum) h(ic) s(itus) e(st) Flavius Silvanus dec(urio) al(a)e eiusd(em) h(eres) f(ecit)`. The only inscription naming a *vexillatio Britannica* as such; Bogaers' comparandum for Nijmegen (auxiliary cavalry, not legionaries). Does not name the Ninth.
* **Mersch (Luxembourg), CIL XIII 4030 = AE 1973, 361 = EDH HD011186** `[V]` (Belgica; fragmentary epitaph/cursus of a Treveran notable): `... flam[en] Augu[sti] / flamen Leni M[art]is quinq[uennalis] / praef(ectus) cohort(is) [II His]panor(um) eq[uitatae] / tribunus militu[m le]g(ionis) VIIII Hi[spanae] / praefect(us) e[quitum alae Augustae] / Vocon[tiorum ---`. An equestrian *tres militiae* career including the angusticlave tribunate of the Ninth; undated in EDH (G. Kremer 2007 discusses the monument); 1st–early 2nd century. Attests only that a Treveran served as tribune in the legion (presumably in Britain), not a Rhine posting of the unit.
* **Tivoli, CIL XIV 3612 = ILS 1025 = EDH HD030763** `[V]`: L. Roscius Aelianus Maecius Celer (cos. 100): `trib(uno) mil(itum) leg(ionis) IX Hispan(ae) vexillarior(um) eiusdem in expeditione Germanica donato ab Imp(eratore) Aug(usto) militarib(us) donis corona vallari et murali vexillis ...` — a **vexillation of the Ninth served in Domitian's German (Chattan) war of 83** under its tribune. This, not the Nijmegen tiles, is the only evidence for men of the Ninth on the Rhine in the 80s.
* **Baalbek/Heliopolis, AE 1903, 368 = ILS 9200 = IGLS VI 2796 = EDH HD031653** `[V]`: C. Velius Rufus, *praef(ecto) vexillariorum leg(ionum) VIIII: I adiut(ricis) II adiut(ricis) II Aug(ustae) VIII Aug(ustae) VIIII Hisp(anae) XIIII gem(inae) XX vic(tricis) XXI rapac(is)* — a Domitianic task force of nine legionary vexillations including the Ninth's (context Danube/Germany, 80s).
* Nothing from Cologne, Bonn, Xanten, Neuss, Mainz or Trier names the legion; EDH searches for "Hisp"/"VIIII" in Germania inferior (16 hits), Germania superior (31), Belgica (3) and Lugdunensis (1) yielded only the Aachen, Nijmegen (2) and Mersch texts `[V]`. (The EDH also holds HD020257/HD050314 under Nijmegen; the Holdeurn mortarium has no database record.) The Pompeii/Naples epitaph CIL X 1769 (Aelius Asclepiades, natione Cilix, miles leg. IX), used by Bogaers for an eastern transfer, is Italian and its attribution to IX Hispana is doubted (Jarrett & Mann 1970, 185, per Eck n. 7).
* Britain (York gate RIB 665 of 107/108; Carlisle/Stanwix/Scalesceugh VIIII stamps; Vindolanda) is covered in `britain_inscriptions.md`; note only that RIB 665 and the Scalesceugh dies spell the numeral VIIII like Nijmegen (Graafstal n. 216) and that Bogaers saw the Carlisle/Scalesceugh dies as typologically related to, but not identical with, the Nijmegen die.

---

## 4. Legionary chronology of the Lower Rhine, c. 100–125 (where the Ninth fits)

| Base | c. 100 | 103/104 | c. 120/122 | Sources |
|---|---|---|---|---|
| **Nijmegen (Hunerberg)** | X Gemina (from 71; PF from 89, PFD 89–96) | X Gemina leaves for Aquincum c. 103/104 (Bogaers & Haalebos 1975, 3; Kemmers 2006, 238; van der Veen 2025, 69; Strobel: 102); replaced by the *vexillatio Britannica* (161 + 35 stamps) incl., most probably, a detachment of IX Hispana (3 stamps; mortarium; Ewijk pendant) | detachment of XXX Ulpia Victrix "shortly after 120" (60–64 stamps; frumentarius altar); fortress abandoned c. 125/130 (van der Veen); small late presence c. 160–170 (EX GER INF stamps; van Enckevort 2021) | `[V]` as cited |
| **Neuss (Novaesium)** | VI Victrix (from 70) | VI Victrix moves to Vetera II c. 103 (NAR 71 table: "Neuss (AD 70–103?), Vetera II (AD 103–121)"); Neuss reduced to an auxiliary fort | — | NAR 71 p. 83 `[V]`; Livius VI Victrix page `[V]`: "At the end of the first century, the sixth legion was transferred to Xanten, where it replaced XXII Primigenia" |
| **Xanten (Vetera II)** | XXII Primigenia until 92/93 (to Mainz), then VI Victrix | VI Victrix | VI Victrix to York with Hadrian and A. Platorius Nepos in **122** (Bogaers 1967, 75: "121/122 oder vielleicht etwas eher"; Livius: "Next year [122], the emperor visited Britain and took VI Victrix with him"); XXX Ulpia Victrix from Brigetio "in AD 118/119 or 122" (Reuter 2012, 11–12, per van der Veen 2025, 136), "seit etwa 120" (Bogaers 1967, 75), "after 122" (Livius), "AD 120 ... from AD 122 permanent at Vetera II" (NAR 71 table) | `[V]` |
| **Bonn (Bonna)** | I Minervia (from 82/83) | I Minervia (away in Dacia 101–106 under Hadrian as legate) | I Minervia (to 401) | NAR 71 table `[V]`; `[T]` for the Dacian absence |
| **Where the Ninth fits** | York (RIB 665: gate of 107/108) | men of the Ninth at Nijmegen with the VEX BRIT, c. 104–c. 120 (school b), or | the whole/part legion at Nijmegen c. 121–130 (school a: Bogaers; Sijpesteijn; Campbell) | see §1.6 |

Graafstal 2025, 47 `[V]` on the systemic point: "The evacuation of Nijmegen by the Tenth was part of a stepwise degrading of Lower German garrison from four to two legions." Livius `[V]` on the "swap": "At the same time, VIIII Hispana came to Germania Inferior; it seems that the two legions traded places" (VI Victrix page) — the traditional 1970s view (Bogaers; E. Birley), now abandoned by the Nijmegen excavators.

---

## 5. Two further questions

**(a) End date of military occupation of the Hunerberg.** Best current estimate: **c. 125/130** for structural occupation of fortress and canabae (van der Veen 2025, ch. 4 and 10: coin and samian-stamp dating curves; 221 of 227 post-104 tile stamps belong to units of 104–c. 130), with the canabae already "largely abandoned by 105" (Kemmers 2006, 109; van der Linden 2011: only 1 % of samian stamps in the western canabae after c. 105) and only "marginal traces" of later habitation; a brief small military presence c. 160–170 (three EX GER INF stamps, one VEX EX GER INF, one LEG I M ANT; two or three Antonine coins) — van Enckevort 2021 still writes "probably until well into the third quarter of the second century", Bogaers' old c. 175 survives only as this residual phase. No dendro-dates exist for the end. Consequently there was no base for a full legion on the Lower Rhine: after 122 the province's two legionary fortresses (Bonn, Vetera II) were occupied by I Minervia and XXX, and Nijmegen's 16.5-ha fortress was a caretaker post (van Enckevort 2021, 197: "Upkeep of the fortress ... may have been the main purpose, so that in case of emergency it could be quickly made operational again").

**(b) Dies and contexts of the three tiles.** One die for all three (Bogaers & Haalebos 1975: "van hetzelfde ... type"; Brunsting & Steures 1995 type 167 with three examples; van der Veen 2025, 151 fig. 6.6,1); the die is not a British one (Bogaers 1967, 69), and the numeral form VIIII matches the north-western British group (Carlisle/Stanwix/Scalesceugh/Leicester) and RIB 665, not York's IX. Contexts: 1959/169 from the occupation level or overlying destruction/rubble layer of the stone fortress (period III/5), third officer's house N of the porta principalis dextra, mixed with X Gemina (majority) and three VEX BRIT stamps; 1966/1752 from the robber trench of the west wall of the principia with two X Gemina stamps; the third without associations. None is stratigraphically dated more closely than "in or above the stone fortress" (built after 88, probably after 100; abandoned c. 125/130). The 104–120 vs 121–130 alternatives are therefore both compatible with the stratigraphy; the choice is made on the garrison history and on the near-absence of finds after c. 104/105.

---

## 6. Bibliography (with URLs)

Primary database records
* EDH HD014357 (Aachen altar): https://edh.ub.uni-heidelberg.de/edh/inschrift/HD014357 — API: https://edh.ub.uni-heidelberg.de/data/api/inschrift/suche?fo_modern=Aachen*&atext1=Hisp `[V]`
* EDH HD020257 (Nijmegen tile, AE 1977, 541): https://edh.ub.uni-heidelberg.de/edh/inschrift/HD020257 `[V]`
* EDH HD050314 (pendant, AE 1996, 1107): https://edh.ub.uni-heidelberg.de/edh/inschrift/HD050314 `[V]`
* EDH HD082661 (Hemmen, CIL XIII 8805): https://edh.ub.uni-heidelberg.de/edh/inschrift/HD082661 `[V]`
* EDH HD011186 (Mersch, CIL XIII 4030): https://edh.ub.uni-heidelberg.de/edh/inschrift/HD011186 `[V]`
* EDH HD030763 (Tivoli, ILS 1025), HD031653 (Baalbek, ILS 9200) `[V]`
* EDCS-10700516 (Aachen altar): https://db.edcs.eu/ (new front end; JSON: https://edcs.hist.uzh.ch/data/monument/107/10700516.json) `[V]`
* EDCS-09301104 (Nijmegen tile, AE 1977, 541): https://edcs.hist.uzh.ch/data/monument/093/09301104.json `[V]`
* EDCS-03000651 (Ewijk pendant, AE 1996, 1107): https://edcs.hist.uzh.ch/data/monument/030/03000651.json `[V]`
* EDCS-80000135 (Ewijk VEX BRIT tile, Schmidts 2018, 8-1, 7): https://edcs.hist.uzh.ch/data/monument/800/80000135.json `[V]`
* EDCS-11100895 (Hemmen altar, CIL XIII 8805): https://edcs.hist.uzh.ch/data/monument/111/11100895.json `[V]`
* EDCS-69700443 / -69700201 / -69700225 / -69700444 (VEX BRIT stamps, CIL XIII 12553, 1a–c, 2) and EDCS-80000127 (Berg en Dal, Schmidts 2018, 8-1, 4g1): https://edcs.hist.uzh.ch/data/monument/697/69700443.json etc. `[V]`
* EDH HD073876 (Carnuntum, CIL III 4466): https://edh.ub.uni-heidelberg.de/edh/inschrift/HD073876 `[V]`
* Trismegistos 208961 / 209140 / 209684 (not reachable from this environment: TLS/503).
* RIB 2462 (tile-stamps of the Ninth in Britain, intro on VIIII/IX dies): https://romaninscriptionsofbritain.org/instrumentum/2462 `[V]`

Nijmegen and Aachen literature
* J.E. Bogaers, "Romeins Nijmegen. De bezettingstroepen van de Nijmeegse legioensvesting in de 2de eeuw na Chr.", Numaga 12 (1965) 10–37. PDF: https://repository.ubn.ru.nl/handle/2066/26454 `[V]`
* J.E. Bogaers, "Die Besatzungstruppen des Legionslagers von Nijmegen im 2. Jahrhundert nach Christus", in: Studien zu den Militärgrenzen Roms (6. Limeskongress, Süddeutschland 1964), BJ Beih. 19 (Köln/Graz 1967) 54–76. PDF: https://repository.ubn.ru.nl/handle/2066/26473 `[V]`
* J.E. Bogaers & J.K. Haalebos, "Romeins Nijmegen. De Nijmeegse legioensvestingen vanaf 70 na Chr.", Numaga 22 (1975) 1–20. PDF: https://repository.ubn.ru.nl/bitstream/handle/2066/26331/26331.PDF `[V]`
* J.E. Bogaers & J.K. Haalebos, "Die Nijmegener Legionslager seit 70 nach Christus", in: Studien zu den Militärgrenzen Roms II (10. Limeskongress, Xanten 1974), BJ Beih. 38 (1977) 93–108 (AE 1977, 541 from pp. 107–109 Abb. 6) `[T; cited by EDH and Sijpesteijn]`
* H. Brunsting & D.C. Steures, "De baksteenstempels van Romeins Nijmegen I. Opgravingen castra 1950–1967, opgravingen Kops Plateau c.a. 1986–1994", OMROL 75 (1995) 85–117 `[R]`; iid., "The brick stamps and the occupation history of the legionary fortress at Nijmegen", in: W. Groenman-van Waateringe et al. (eds), Roman Frontier Studies 1995 (Oxbow Monogr. 91, Oxford 1997) 323–329 `[R: Graafstal 2025; van der Veen 2025]`
* W. Eck, "Zum Ende der legio IX Hispana", Chiron 2 (1972) 459–462. PDF: https://publications.dainst.org/journals/chiron/article/download/332/4940 `[V]`
* H. van Enckevort, "The Roman military complex at Nijmegen (19 BC–475 AD)", in: W.A.M. Hessing, W.K. Vos & E.J. van Ginkel (eds), Romans on the Waterfront. Evaluation of archaeological interventions (1997–2020) along the Dutch part of the Lower Rhine and Coastal Limes (NAR 71, Amersfoort 2021) 185–202. PDF: https://cultureelerfgoed.info/download/pdf/NAR071_Limes.pdf `[V]`
* E.P. Graafstal, "What Happened Next? Hadrian's Wall, the expeditio Britannica and the Fate of the Ninth Legion", Britannia 56 (2025) 25–55. PDF: https://www.cambridge.org/core/services/aop-cambridge-core/content/view/2939F9DCC38AC465ED8D2D3B13A3A044/S0068113X2510038Xa.pdf/ `[V]`
* J.K. Haalebos, "Römische Truppen in Nijmegen", in: Y. Le Bohec & C. Wolff (eds), Les légions de Rome sous le Haut-Empire (Lyon 2000) 465–489. Repository record without file: https://repository.ubn.ru.nl/handle/2066/124815 `[NOT accessed; relayed]`
* J.K. Haalebos, "Traian und die Hilfstruppen am Niederrhein. Ein Militärdiplom des Jahres 98 n. Chr. aus Elst", Saalburg-Jahrbuch 50 (2000) 31–72. PDF: https://repository.ubn.ru.nl/handle/2066/27348 `[V; on the Hemmen altar/ala Vocontiorum and the vexillatio Britannica, p. 39–40]`
* J.K. Haalebos & W.J.H. Willems, "Recent research on the limes in the Netherlands", JRA 12 (1999) 247–262 `[NOT accessed: Cambridge paywall; Leiden/willemwillems.com copies unavailable]`
* N. Hodgson, "The End of the Ninth Legion, War in Britain and the Building of Hadrian's Wall", Britannia 52 (2021) 97–118, doi:10.1017/S0068113X21000015 `[V: abstract only]`
* F. Kemmers, Coins for a Legion. An analysis of the coin finds of the Augustan legionary fortress and Flavian canabae legionis at Nijmegen (diss. Nijmegen 2005; Studien zu Fundmünzen der Antike 21, Mainz 2006). PDF: https://repository.ubn.ru.nl/handle/2066/40928 `[V]`
* H. Nesselhauf & H. von Petrikovits, "Ein Weihaltar für Apollo aus Aachen-Burtscheid", Bonner Jahrbücher 167 (1967) 268–279 (Heidelberg digitisation: https://journals.ub.uni-heidelberg.de/index.php/bjb/article/view/73450 — blocked to automated access) `[NOT accessed]`
* U. Schillinger-Häfele, "Vierter Nachtrag zu CIL XIII ...", BRGK 58 (1977) 447–603, no. 144 `[T]`; C. Keller, Archäologische Forschungen in Aachen (Rhein. Ausgr. 55, 2004) 80 no. 648/019 `[R: EDH]`
* P.J. Sijpesteijn, "Die legio nona Hispana in Nimwegen", ZPE 111 (1996) 281–282. PDF: https://www.uni-koeln.de/phil-fak/ifa/zpe/downloads/1996/111pdf/111281.pdf `[V]`
* K. Strobel, Untersuchungen zu den Dakerkriegen Trajans (Antiquitas I 33, Bonn 1984) 99–101, 444, 453 `[R: van der Veen]`
* V. van der Veen, On the legion's doorstep. Studies into the Nijmegen castra and canabae legionis (Radboud Dissertation Series, Nijmegen 2025), doi:10.54195/9789465150116. PDF: https://repository.ubn.ru.nl/handle/2066/319111 `[V]`; id., "Chronology and spatial distribution of terra sigillata potters' stamps and coins within the Nijmegen castra and canabae" (2020/21; https://repository.ubn.ru.nl/handle/2066/319921) `[not opened; incorporated in the 2025 thesis ch. 4]`
* W.J.H. Willems & H. van Enckevort, Vlpia Noviomagus. Roman Nijmegen. The Batavian capital at the imperial frontier (JRA Suppl. 73, Portsmouth RI 2009) 56–57, 152 fig. 54 `[R]`; review G. Raepsaet, AC 80 (2011): https://www.persee.fr/doc/antiq_0770-2817_2011_num_80_1_3805_t16_0662_0000_2 `[V]`
* Wikipedia (en) "Legio IX Hispana" (wikitext fetched; footnotes Willems & Enckevort 2009, 56; Campbell 2018 ch. 11; Hodgson 2021; Keppie; Russell 2011) `[V]`; Wikipedia (de) "Porcetum" (altar description, N&P dating) `[V]`; Livius.org "Legio VIIII Hispana", "Noviomagus", "Legio VI Victrix", "Legio XXX Ulpia Victrix" `[V]`

## 7. Caveats / not achieved
* Haalebos 2000 ("Römische Truppen in Nijmegen") and Nesselhauf & von Petrikovits 1967 could not be opened; their positions are relayed from Graafstal 2025, van der Veen 2025 and de.wikipedia respectively. Hodgson 2021 only via its abstract and Graafstal's summary. Campbell 2018 only via Wikipedia.
* EDCS: the legacy form endpoints now return the new single-page app; the JSON API at edcs.hist.uzh.ch works but is throttled. All five relevant EDCS records were verified from the JSON back end: EDCS-10700516 (Aachen), EDCS-09301104 (Nijmegen tile), EDCS-03000651 (Ewijk pendant, filed under Nijmegen), EDCS-80000135 (Ewijk VEX BRIT tile), EDCS-11100895 (Hemmen). The VEX BRIT stamps are in EDCS as CIL XIII 12553, 1a–c and 2 (EDCS-69700443, -69700201, -69700225, -69700444, expanded there as Vex(illarii) Brit(tonum)) plus Schmidts 2018 entries (EDCS-80000127 Berg en Dal, EDCS-80000135 Ewijk). EDCS has no record for the Holdeurn mortarium (as far as the 755 downloaded records from the "VIIII Hisp", "VEX BRIT" and "Britannic/exercit" searches show).
* Brunsting & Steures 1995/1997 (the tile-stamp catalogue) is not online; the die/type information is taken from van der Veen 2025 and Bogaers & Haalebos 1975.