* ‘in the past’, and before the Cimbric invasions 113–101 BC, which they withstood [4].

† ‘Gallia Belgica’ official at latest 13, if not 27 BC.

* See [14]. The Carnutes’ location, Seine/Loire centred on Chartres, is fully known.

† See [16]. Trebonius’s must have been the legion watching their capital at Amiens (Samarobriva), and guarding Caesar himself with headquarters and stores, till he had to march with it against the NE revolt, so replaced it by the legion under Crassus, brought from the Bellovaci 25 miles farther S. He omits Trebonius’s name because he led this march himself [17].

* Or else ‘thought it superfluous’—for which, like the great Nipperdey before him (1847), he could guess no reason [20].

† His ‘advent’ to give battle with ‘legions’ (BG, v, 48, 10), reinforced now by Fabius’s legion (47, 3), makes these sound numerous; but see Holmes [16] on 46: he had started with Trebonius’s legion only. Cf. Rambaud [17], 182–6; ‘speed’, 203, 251–4; ‘advent’, 254–6: favourite emotive themes.

‡ See [27]. Richest in lands, second only to Bello- vaci in fighting strength, 12 fortresses; location, Marne-Oise-Aisne (round Soissons) stretched nearer to Seine by client tribes, Silvanectes (Senlis) NW of Paris, Meldi (Meaux) E of it on lower Marne [28].

§ Farthest out but expected to be safest [30].

|| Along with the Armorican tribes of the NW, BG, 11, 34 [31]. They soon put Roscius in danger, retreating only at news that Caesar had crushed the NE revolt (BG, V, 53, 6-7). Caesar makes revolts disjointed when in fact concerted [32].

* The final words [29] say all ‘were contained within 100 miles’; meaning what? fig. 1 shows the smallest ring they could have formed. Old dispute on the meaning, Holmes [7], 371–3, today needs Ram-baud [17], 179–82 on Caesar’s figures generally, and 182–6 on his slanting facts by blurring them. Here surely, faking 120–150 miles of diameter as ‘100’, he makes ‘contained’ obscure intentionally, to blur his ring into seeming tighter than it was. But no crude lie: his purpose (though he hides its failing) was just what he conveys, and the ring (though loose) was real.

† This old presumption of identity [35], though perhaps not false [‡ p. 10], has thus at least been over-simplified. We need a Celticist’s fresh appraisal of the seeming name-chain, Cassivellaunus-Catuvel- launi-Catalauni; and fresh thinking to suit it.

* See Wheeler and Richardson [35] 8–12; named from the Camp du Canada overlooking Fecamp (Seine-Maritime), where they excavated 1939: 62–75.

† For Armorica ([35], 1–4, 23 ff.; 38 ff.), and in general (159–216) by Mrs Cotton; her classic type as at Avaricum (Bourges: 199–201) seems now to start more plainly pre-Caesar at the Engehalbinsel (Bern: 214–15), Manching (Bavaria: 213–14), and at Châteaumeillant (Cher): † p. 13.

‡ Wheeler and Richardson in Gaul, among the non-’Fécamp’ forts in ‘Belgium’ (ten) and elsewhere in Belgica (list [35], 122), found few with such ditch-form guessable from profile, as at Catenoy (Oise, their no. 82: Bellovaci); but could note (12–14) its neighbour Mont-César nr. Beauvais (no. 81), Vermand of the Viromandui (Aisne, no. 91) and Étrun of the Atrebates (Pas-de-Calais, no. 93), all capitals perhaps since pre-’Fécamp’ times, as similar in scale to ‘Vieux-Laon’ among the Remi (Aisne, no. 89: 2 phases, blunt and sharp V ditch)—outclassed only by the huge ‘Vieux-Reims’ (no. 88)—and among the Catalauni to the Camp d’Attila or ‘Vieux- Châlons’ (Marne), whose vast V ditch they compared with Catuvellaunian Wheathampstead [35], [44].

§ With Camulodunum’s adjacent Pitchbury, cf. meanwhile Cholesbury in Bucks. [45].

* The E coinage belongs to the period of these because lasting through it, and was probably the gold seen in use here by Caesar himself ([47], 20); thus current about his time, so giving room for refugees in its story, it forms a bridge between what followed and what preceded him ([46], 114 paras. 59, 57), and Allen ([47], table I) takes it through to 40 bc from c, 60 or perhaps just earlier. This is fitted by his seeing its thrust into Sussex, and that inland through Surrey, as dislodging westward the prototypes of Dorset silver, which itself started pre-56, so the thrusts have been made by c. 60: [46], 114 para. 6o, 105–7 paras. 32–3, 35–6, (with 118–20, Armorican), 111–13 paras. 49–52, all using the critical Le Câtillon hoard, and giving the case here summarized. E starting from Nervii: 114 para. 55. D (see overleaf) from near Calais coast (111 para. 50) would mean Morini; from modern Belgium (110 n. 25), Menapii; for their boundaries see De Laet [48].

* Allen ([46], 126, fig. 28) gives sketches hatching each group’s principal area(s); of course to be checked from [47], whence my stretching C’s area out (fig. 3c).

† ‘Within’ because the coins were in Britain from before 100, no matter when e.g. Late La Tène brooches appeared in North Gaul after it. Dr Hodson on these [55], citing coins’ distributions but not dates, might even so have mentioned Moberg [56], besides the still shyer Hachmann [57]; Hachmann’s start for Nauheim fibulas (from Werner) is not early enough to fit new Engehalbinsel evidence [58]; cf. Mrs Cotton, [35], 214.

* Piggott [64] includes British map of these 'Belgic' La Tène II brooches; their wrapped or clasping foot is that of Hodson's Münsingen IIa form in Switzerland ([55], 131–8, figs. 5–6) starting by c. 200, though without the moulding on its foot below the clasp (cf. [75], figs. 30–3, exx. from Aisne). His IIb form, with this moulding moved or repeated to emphasize the clasp itself, as a collar, starts appearing c. 100; most such brooches here are in SW (Piggott [64], 72), and begin no doubt pre–50; still more should the wrapped-foot 'Belgic' ones begin pre-50 in SE—however little sooner than La Tène III forms.

† Beside Deal Waterworks on Mill Hill, Walmer [65], whence [66]–[67]. Fox ([59], 111–12, 114, no. 60) observed the native ornament on those from Weston (Bath: 36–7), but not the Deal spoons’ whirligig designs, which to me ‘are just those of the frieze of the Aylesford bucket’ [68]; this, buried in Belgic Kent after c. 50 [75], is surely earlier itself, and in ornament Continental [69].

‡ The ‘South-Eastern B’, ‘Wealden’ and other pottery discussed by Ward Perkins [71], needs fresh appraisal, starting from Frere’s ([42], 90–2) assessment, where for Margate decoration see [72].