My copy of Firearms: A Global History to 1700 by Kenneth Chase has just arrived. Cracking it open near the middle, I landed on the section concerning Guns and Bows. Thankfully, Chase doesn’t repeat the line, common in history books written for a casual audience, that firearms were somehow less effective than bows until the American Civil War.
Chase does, however, compare bows and firearms in terms of reliability, rate of shot, and accuracy.
Chase claims that firearms misfired up to 50% of the time, citing Tallett, War and society in early-modern Europe, p. 23. This assertion is extremely dubious. Nobody would use a weapon that was so unreliable.
Chase then makes a brief comparison of accuracy. Citing page 138 of Saracen Archery, Chase says “The Turkish mamluks were also expected to hit a 38-inch circle at a range of 75 yards with five out of five arrows, and a good archer might hit a target the size of a man on horseback one out of four times at a range of 280 yards.”
By contrast, Chase says that the smoothbore musket “had no chance of hitting a man-sized target at more than 80 yards.”
Really? No chance? Zero percent? Murphey has no problem hitting a smaller than man-sized target dead center at 80 yards:
In a test at the Provincial Armory in Graz, Austria, the muskets were tested at a range of 100 meters (109 yards) against a target 167cm x 30cm, or about the height and width of a short man with no arms. The pistols were tested at 30 meters. The muskets had significantly more than “no chance” of hitting at 100m.
I suspect that period firearms might have even performed better than the results of the Graz test indicate. Three of the muzzleloaders tested were rifled: G 284, RG 272, and STG 1288. Only STG 1288 displayed better accuracy than the smoothbores- but STG 1288 was also the only muzzleloader loaded with a bullet larger than the bore. The report does not mention what sort of wadding or patch was used, if any. Lack of a patch would explain why the rifles provided with sub-caliber bullets didn’t perform better than the smoothbores, as the bullet could not engage the rifling. If the test was really performed without wadding or patch, that is an unfortunate oversight. Also consider, that at the time the Graz tests were performed, the weapons are hundreds of years old. While when the weapons were actually in use on the battlefield, they would have been relatively new. The test writeup mentions that two pieces were rejected because of weakness in the metal. If two pieces had become too dangerous to shoot in the years since they were built, then time, moisture and the notorious corrosivity of black powder residue could have also worn the pieces enough to negatively affect accuracy.
This post deals only with theoretical accuracy against man-sized targets. Tests against targets the size of close-order units, and accuracy under real combat conditions, are another matter.
Bow Vs. Musket 
You seem to react defensively to any sources that criticise early firearms. You are correct in saying the Graz Armoury tests don’t reflect battle accuracy, since it is nothing so accurate as that.
The Graz test involves a gun fixed in place, possessing perfect and unflinching aim, unlike the poor musketeers who only got 2 to 6 rounds of practice ammunition (30, if they were British). Against a block of infantry at ~75m, the early 19th century Prussian study found their men, wielding rifles, could only hit it half the time… 3/4s of the time with the latest rifles. I’ve heard accuracy during the Napoleonic Wars was something like 5 to 7%, but I never saw the numbers crunched.
So, contemporary data indeed shows that musketeers were less accurate than the mamluks, which makes sense when you consider how little training a musketeer gets, compared to learning to shoot from childhood. This isn’t the case for proficient, well-practiced shooters, such as modern practitioners, the highly professional harquebusiers of the 17th century, or the Japanese veterans of the Sengoku Jidai. Those examples can certainly hit men at 100yds, or perhaps even 300. But expecting a gun to aim for you is ridiculous, so of course low quality troops can barely hit the ocean while standing on the bottom.
As for the age of the guns… people still use centuries old antiques to shoot today. They don’t seem any less accurate. So long as they’ve been adequately preserved (as museum pieces generally are), they should shoot straight just fine.
LikeLike
Well, if I was defensive in the post, I wrote it over five years ago. I remember being annoyed with Chase because he hyperbolically exaggerated the faults of early firearms. Back then I was often frustrated because people seemed to derive their understanding of historical weapons from movies, video games and poorly-sourced pop history books, so hyperbole was not helpful. But today people who are interested in history have easier access to primary sources I think the situation is much better.
One of the Graz guns was apparently so inaccurate that they stopped testing it. That makes me wonder if it was decayed. Some people shoot antique guns but only fools shoot antique guns that are in bad condition.
LikeLike
Getting frustrated is fair enough. And I realize you wrote this 5 years ago, but I didn’t notice any newer post on the topic, so I have to bring up these points in response to this one for the sake of future readers.
Nothing’s to say that musket wasn’t that inaccurate to start with (you get duds even today). There’s really no way to be certain, which is why I brought up contemporary data about accuracy. If these guns were in horrible condition, then it’s ironic that they did better than contemporary musketeers.
LikeLike
Something to add. In a contemporary French test with a musket fixed in place, mentioned by Napoleonic Weapons and Warfare: Napoleonic Infantry, they found a musket fixed in place was hitting a 3×1.75m target at 75m 60% of the time. So even with perfect aim, contemporary load or guns were so bad that they could barely hit a block of ~4-8 men (depending on their ranks and spread).
So the museum tests do not sound out of character.
LikeLike