When comparison is made to the data from the Mary Rose the difficulties become apparent (Figure 7.6), for the smaller calibre peaks on the Towton, Kings’ and Bosworth graphs correspond to some degree to the Mary Rose graph. It may be that the smaller mid-sixteenth century military handguns had a similar bore to later sporting guns. In contrast, the resampled Bosworth graph shows no correlation to the Mary Rose data or to the background noise on the other graphs. If this is not the result of random factors, due to the small size of the sample, then it could be the faint signature of military action in 1485 or 1644.

The limited studies so far undertaken of surviving early handguns suggest that the hand-cannon had a bore extending from 18 mm up to 35 mm.⁶² When windage of just under 2 mm is taken into account, to allow the ball to be introduced into the barrel, this gives a calibre of 16–33 mm (Figure 7.8). As discussed below, the upper part of this range merges with guns of intermediate types which were fired from a rest, a wall or a wagon. At the very upper end it also overlaps with true artillery. At the lower end it meets the documented bores of the later sixteenth century arquebus.

What is required is a far more extensive study of surviving early handguns, focusing particularly on those not intended to be used on a rest. The need for this is emphasised by the exploded gun barrel preserved in Grandson Castle, Switzerland, believed to have been recovered from the 1476 Grandson battlefield, that has a bore of just 17 mm (Figure 7.7). Allowing for a windage similar to that seen in later centuries, at just under 2 mm, this gun is likely to have fired a ball of c. 15 mm. This is almost identical to the calibre recorded for the arquebus more than a century later, in England in 1589, which was 20 bore and thus equates to a ball of c. 15.6 mm.⁶³ While background noise tends not to peak in this area, unfortunately it sits squarely within the calibre of the seventeenth-century carbine, a cavalry weapon lying half way between pistol and musket.

Comparison of the calibre graph with that for the major Civil War battlefield at Edgehill confirms that no infantry action took place at Bosworth in 1644, for musket calibre balls are almost completely absent. With a Civil War cavalry engagement both pistol and carbine are likely to have been in use and, as we have seen, this is where a peak appears around 15.5 mm in the calibre graph for the battlefield core at Bosworth (Figure 7.5). It could be argued that the bullets are not from the 1644 cavalry skirmish because one apparent indicator of seventeenth-century cavalry action is missing – the slug. This is an elongated bullet of varying type which normally has a maximum diameter which indicates use in guns of pistol or carbine calibre. But slugs represent only a small percentage of any battlefield assemblage and, given that the Bosworth sample is so small, it would not be surprising for none to have been recovered.⁶⁴ The small number of carbine and pistol calibre bullets present in the assemblage is not a problem, for cavalry action appears generally to have left a very low density of bullets, as can be seen from the survey on the cavalry wings at Edgehill. This would be particularly true if, as the record of the Bosworth skirmish claims, the royalist cavalry turned to flight on the first parliamentarian charge.⁶⁵