As this ball travelled down the barrel the irregularities in the bore, where it widened and narrowed, enabled the ball to slip to a new orientation where it took a new signature and it slipped again taking on a third set of marks. In this way the ball took on the signature not just of a stave built gun but a unique signature of that individual barrel, determined by the number of staves, the width of each and the degree to which they were bent. Thus there can be no doubt that this ball was fired from a stave-built wrought iron gun, a type which went out of use during the sixteenth century. Indeed, if another well preserved ball with similar firing evidence was retrieved from the battlefield, then it should be possible to say, beyond reasonable doubt, whether or not it had been fired from the same gun. In addition, the presence of extensive melting right across the lower hemisphere of the ball demonstrates that this was a muzzle loading gun, not a breech loader, and that the ball was fired unwadded.

A similar method was also used to construct barrels of about 70 mm and below, but in these the inner tube typically comprised just two parts. These were thinner and much more heavily curved, looking like gutters, which were set face to face with a slight overlap, and held together with the usual hoops and rings. These barrels are therefore typically described as gutter-built. The construction method is most easily understood when viewing a partly corroded gutter-built barrel (Figure 7.39 and 40). Another variation, in the smallest and shortest of guns, sees an inner tube in the form of a single thin sheet wrapped around and overlapped to form the bore.¹⁵²

In all composite barrels the end product is a tube which is open at both ends and so they were often completed as breech loading guns, although some examples were constructed with a permanent breech and were loaded from the muzzle. The chambers for the breech loading guns could also be of composite construction, but this is a method that tends to be used for those of larger calibre, with the majority being forged as a single piece from wrought iron.

Amongst the Bosworth round shot are examples with bands on which there are ridges or scarps running parallel to the axis of firing which divide the band into facets. This seems to be the signature of round shot fired from a gutter-built gun, as can be seen from our experimental firing (Figure 7.42). The experimental ball has a scarp which represents the lap joints between the two gutters, which presumably would be repeated on the opposing side if the band was intact in that area. There are also a series of less distinct ridges where parts of the curve of the bore seem to have been more angular, which has also resulted in a zigzag upper and lower edge to the band. This appears to be the result of inconsistency in the bending of the two gutters, giving the bore a slightly irregular geometry. Identical markings are seen on several of the Bosworth rounds (Figures 7.42 and 7.43). If such patterns can be consistently repeated in further experiments then it may be possible, as with the stave-built barrels, to distinguish not only the type of barrel but also the signature of an individual gun.

The gun used for our experiments was a breech loader and this created other distinctive attributes on the experimentally fired rounds. First, there is no evidence of melting on the lower hemisphere and the lower edge of the band is sharply defined, confirming that hot gasses did not contact the ball. This is a result of the wooden tampion protecting the ball from the effects of the hot gasses. The compression of the lower hemisphere also seems to be a distinctive signature of firing with a wooden tampion, although the degree of compression seen in the experiments is greater than on the Bosworth round shot. This may be due to the use of modern gunpowder, which burns more quickly than powder manufactured to an early recipe. So, for a given degree of energy release the modern powder creates a higher peak pressure compared to early powder, and it is peak pressure which determines the degree of compression.

The Bosworth rounds which show an identical combination of attributes, though with less compression of the lower hemisphere, were presumably all fired from breech loading gutter-built guns. However there is a problem, in that we lack experimental data for a gutter-built muzzle loader. It is possible that it might produce an identical signature if it was fired heavily wadded. What is not in doubt is where a gutter-built muzzle loader has been fired without wadding, as then extensive melting of the lower hemisphere accompanies the faceted band (Figure 7.43).