Strap-Ends (2001 guide)

Please note that this guide has not been fundamentally changed from the original print version of the Finds Recording Guide (Geake 2001), written when the database contained just 8,800 non-numismatic records.


Any fitting which is designed to finish a flat strip of leather or textile, however narrow, should be termed ‘strap end’. By implication, not all strap-ends were dress accessories, almost the majority no doubt were. A fitting designed to finish a square- or round-section leather thong or textile cord should be termed ‘lace tag’.

PAS object type to be used


PAS object classification to be used

The period-specific sections can be consulted for terms to use in the Classification field.  In summary, the terms for Roman strap-ends are: ‘tubular’, and ‘nail cleaner’, as appropriate; for early-medieval strap-ends use the Thomas classification (see below); for medieval strap-ends use ‘composite’, ‘crescentic’ or ‘hollow plate’, as appropriate.

Terms to use in the description

The two ends of a strap-end are the ‘attachment end’ or open end (not split end) and the ‘terminal’ or closed end. The terms open end and closed end are perhaps more useful when there is a decorative terminal beyond; the terms attachment end and terminal are more useful when the strap-end is simply a flat piece of metal.


Strap-ends date from the Roman period onwards, but are rare in the early early-medieval period and seem to have largely fallen out of use early in the post-medieval period.

Roman strap-ends

There are a number of objects in the literature of Roman military equipment which are interpreted as pendants which hung from the ends of legionaries’ apron straps. They are a long pointed oval shape, with oval or square perforations at the top, and are vaguely reminiscent of drop handles from 18th- to 20th-century furniture. Strictly speaking they are strap-ends, and should be recorded as such.

Strap-ends do not seem to have been used in any quantity, however, before the introduction of the so-called ‘soldiers and settlers type’ material in the late 4th century. There is as yet no sensible name for this group of material – calling it after a 1960s article (Hawkes and Dunning 1961) is not good enough! Three types of strap-end have been distinguished, the lancet-shaped, amphora-shaped, and tubular.

The first two are very similar indeed and are only really distinguished by one having openwork decoration and the other not. There is an article covering ‘amphora-shaped’ strap-ends (Simpson 1976). (The article also covers buckles, but these appear to be of a type not found in Britain.) Some have two hinge loops, and can therefore be mistaken for buckle plates; others have two rivet holes. They all have two perforations which look like the handles of an amphora. Strap-ends without these perforations are called lancet-shaped, and illustrations can be found in Hawkes and Dunning (1961, 64 – their type VA). I feel that the distinction between the two types is not particularly helpful for our purposes, and it is not worth putting anything in the Classification field; the date range will flag them up sufficiently: c. 350-450 AD.

The third type of late Roman strap-end, the tubular type, consists of a hollow cylinder with transverse groove decoration which is either cast in one piece with, or soldered onto, a rectangular plate, and runs the width of the belt. Examples can be found in Hawkes and Dunning (1961, 4, 67). Put ‘tubular’ in the Classification field.

Occasionally one of the amphora-shaped or lancet-shaped strap-ends has a forked end, very like a nail-cleaner. These can have a split attachment end and a single rivet, or be hinged through a rectangular perforation. Put ‘nail cleaner’ in the Classification field. Of course, if the attachment end breaks off, you won’t know that it ever was a strap-end, and it will go down as NAIL CLEANER!

Early-medieval strap-ends

Gabor Thomas’s typology should be used to refer to the different types – put ‘Thomas Class A, Type 1’ etc in the Classification field. It is published as two FRG Datasheets (Thomas 2003; 2004) and supersedes his PhD thesis (Thomas 2000); the thesis can still be usefully consulted for illustrations.  Class A is the classic ‘long’ 9th-century strap-end with animal-head terminal, often in the Trewhiddle style; Class B is the narrow type with transverse line decoration and fairly thick cross-section which can date from the 8th to the 11th century; Class C is flimsy and very narrow; Class D is lozengiform and very rare; Class E is the classic 10th/11th century type which is usually fairly broad, with relief decoration, often in the Winchester style, and parallel sides, Class F is a rare double-sided Irish type; Class G is moulded in the largely 11th-century Urnes style.

Medieval strap-ends

There is much variation in medieval strap-ends but, as with buckles, only a few types merit an entry in the Classification field.  A good range can be found illustrated by Egan and Pritchard (2002) figs 83-99 and Egan (2007) pls 23, 24.

The composite strap-end, like the composite buckle, is a 14th-century type. The term ‘composite’ should be restricted (for strap-ends and buckles) to those made up of sheets sandwiched together with a shorter or forked spacer plate. They usually have some kind of decoratively pointed or elaborately cut-out terminal. Put ‘composite’ in the Classification field.

At the end of the 14th century a different set of buckle and strap-end becomes popular and remains so for the whole of the 15th century. They are cast in one piece, with hollow attachment ends and usually incurved sides beyond this. They are a good example of a strap-end where the ‘closed’ end is in no sense a terminal, having a large openwork plate beyond. These are sometimes called ‘lyre-shaped’, but as such analogous terms should be avoided as far as possible it is better to put ‘hollow plate’ in the Classification field.

Crescentic strap-ends are another 15th-century form; Fingerlin (1971, figs 312-317) dates them to the first half of the century. The terminal is usually sub-triangular, and decorated with openwork and low relief to resemble a bush or tree. Engraved lines around the crescent-shaped part represent the roots of the tree. They are made up of two pieces, the cast front plate and a sub-circular sheet underplate held on by rivets and solder. Put ‘crescentic’ in the Classification field for this type only; if other crescent-shaped strap-ends turn up, do use the word in the description, but don’t put it in the Classification field.


Strap-ends: Roman amphora type (left, WILT-B0F593); early-medieval Thomas Class E, Type 2 (centre, NMS-199370); Medieval composite example (right, SUR-5CE4F3). Copyright: Salisbury and South Wiltshire Museum; Norfolk County Council; Surrey County Council; CC-BY-SA licence)
Strap-ends: Roman amphora type (left, WILT-B0F593); early-medieval Thomas Class E, Type 2 (centre, NMS-199370); Medieval composite example (right, SUR-5CE4F3). Copyright: Salisbury and South Wiltshire Museum; Norfolk County Council; Surrey County Council; CC-BY-SA licence)

Search for all examples of strap-ends

Key references

Hawkes and Dunning 1961

Thomas 2003

Egan and Pritchard 2002