Colchester

Each has an integral bilateral spring and hook issuing from the back of the bow, the latter to hold the external chord.

1 sf. 338 The hook is very short with just a hint of a bend along the top of the bow. The wings and bow are plain, the bow having an octagonal section. Its profile is almost straight, the suggestion of a recurve may be the result of damage. The catch-plate is largely missing, along with the bottom of the bow.

338

A Colchester one-piece type brooch of length 47mm with a spring of 6 coils behind small plain wings. The very slender, plain bow is of oval section and is straight except for a sharp angle a short distance from the head. The bow tapers to a point with a very small catch-plate which extends 9mm up the length of the bar. This brooch appears closely related to the small angular La Tène III type. (See Nos. 33 & 35).

2 sf. 206 The hook is moderately long. Each wing has two wide flutes divided by a groove. The bow is plain, an octagonal section and a well curved profile. The catch-plate has three piercings separated by dog-legs. The bow has a line of rocker-arm on each side next to the catch-plate.

3 sf. 69 Brutally stripped during conservation, none of the original surface is left. The wings are now plain. There is evidence for there having been a wavy ridge down the front of the bow which may have had a hexagonal section.

4 sf. 75 Brutally stripped during conservation, none of the original surface is left. The wings had been fluted and the bow, probably with a hexagonal section, once had a wavy ridge down the front.

51

A Colchester one-piece brooch of length 54mm. The spring coils and catch-plate are largely absent except for part of the first coil and the rather rough remnant of catch-plate which does, however, clearly indicate the point of convergence with the bar, that being 21mm from the tip of the bow. The head of the brooch has short wings and there is no visible evidence of any decoration on the bow. There is a hint of a perforated catch-plate in that three of the indentations in the edge of the catch-plate remnant appear to be rather more regular than the rest. It is by no means certain however, that these are not the result of soil chemistry and natural corrosive processes. Hull's type 90.

6 sf. 175 Badly pitted, only general details can be seen. The wings and bow appear to have been plain, the latter certainly had a hexagonal section and is now bent.

None of these brooches displays any sign of being late in the overall sequence: they are long, all the ornament is hand made, where the catch-plate survives, the piercing is carefully made and where only the stubs survive, there are traces of rectangular holes and not circular ones. As Colchesters like these were not being made at the time of the conquest, in fact, the latest types had probably passed out of manufacture by then. However, many survived in use after the conquest, but these are almost exclusively the late types, short in length, fully cast whose catch-plates have circular holes or none at all. The only true guide to the development of the Colchester available at the moment is the King Harry Lane cemetery (Stead and Rigby 1989)

The dating of the phases into which the cemetery is divided is not entirely secure. While the earliest likely date for the cemetery is said to be 15 B.C. (ibid., 83), the authors preferred to be more conservative in their dating: Phase 1, A.D.1-40; Phase 2, 30-55; Phase 3, 40-60; Phase 4, 60+ (ibid., 84). Over half the burials should therefore be statistically later than the Roman conquest. However the samian report (ibid., 113) contains the comment that it is surprising that there should be only six vessels: three earlier than 25, none dating to 25-50, two dating 45-65 and the last being much later. Looking at the published dating, it is surprising is that there is only one Colchester Derivative (G316,4) and no fully formed Hod Hill, both types well represented in Verulamium scarcely 500 metres away. If the dating is, however, taken back to 15 B.C., and the divisions between the phases adjusted accordingly, most of the basic problem largely disappears. The following ranges are suggested: Phase 1, 15 B.C.-A.D.30; Phase 2, 20-40; Phase 3,35-50/55; Phase 4,45+. Most of the burials now become pre-conquest. The end-date of 50/55 for Phase 3 was arrived at by using general evidence for the dating of Colchesters and is relevant to the present brooches in that they would be subject to the same rules of residuality which governs all material not melted down or otherwise removed from a normal site assemblage. In the case of such a specialised collection like that from King Harry Lane, the writer would be happier with a terminal date for Phase 3 of 40/45. This should mean a further adjustment in the phases before then, but no suggestions are made here. In terms of Brooch 2, which is the best preserved and only placed in second place because it has a longer hook than that on Brooch 1, the occurrence of similar ones in the cemetery is: Phase l, G.202.9-11, G.206.3, G.397.6; Phase 2, G.143.3, G.231.4-3, G.23.2. The criteria used were size, decoration and the presence of relatively elaborate to very elaborate fretting in the catch-plate. Following these, the result is clear, most occur in Phase 1, some in Phase 2, and none later. In terms of the suggested dating, such brooches range up to about 40, in terms of life in use, then it is possible that one might have been seen in as late as 45, but hardly beyond. As far as all the brooches here are concerned, the end of their life in use should not be later than 45/50.