Longstreet Longstreet on Pickett's Charge: "thirty thousand men was the minimum of force necessary for the work"

It seems like the only way in which the effect would be different would be if the Confederate one involved the flame bypassing the fuze channel entirely while the Union one involved the flame breaking through into the whole of the fuze channel (and thus automatically making it be the shortest possible fuzing).

Either way the shock is the same, and my belief is that it's because the fuze was not designed for 10-12 pounder guns. This would explain why the problems weren't found at some point pre-war, because pre-war the US didn't really have any 10-12 pounder guns (or very few, at least).
 
It seems like the only way in which the effect would be different would be if the Confederate one involved the flame bypassing the fuze channel entirely while the Union one involved the flame breaking through into the whole of the fuze channel (and thus automatically making it be the shortest possible fuzing).

Either way the shock is the same, and my belief is that it's because the fuze was not designed for 10-12 pounder guns. This would explain why the problems weren't found at some point pre-war, because pre-war the US didn't really have any 10-12 pounder guns (or very few, at least).
It's hard to say. While the Union fuzes had issues, it does not appear to have risen to the level of the Confederate. The tests by the Army leading to adoption of the Bormann c. 1854 were actually conducted using 24 lb. and 12 lb. (iron) guns. Brig. Gen. Henry Abbot, who commanded during the Siege of Petersburg, reported that the Bormanns used there had a 77% success rate. I haven't seen a breakdown by types/calibers, although he did segregate mortars (which were c. 87%-90%). One issue with the Federal Bormann was a design issue - it had a limited burn rate of 5 seconds in 1/4 second intervals - this limited it to c. 1300 yards or so from a 12 lb. Napoleon.
 
Hmm, segregating out the mortars presumably means that the other ones had a lower reliability rate (well, it states it outright, that's how statistics works). Do we know what the Confederate fuze failure rate was?
Based on the mentioned Petersburg statistics then I'd say that if the Confederate fuze was in the 70%-75% bracket then it'd be statistically indistinguishable from the Union one.
 
Hmm, segregating out the mortars presumably means that the other ones had a lower reliability rate (well, it states it outright, that's how statistics works). Do we know what the Confederate fuze failure rate was?
Based on the mentioned Petersburg statistics then I'd say that if the Confederate fuze was in the 70%-75% bracket then it'd be statistically indistinguishable from the Union one.
I'd have to check Abbot's work to see if the 77% was all-inclusive or was just "non-mortars". Obviously, it also depends how many firings of each were involved. I've never seen any calculation on the Confederate side. Given the lack of really great options and the emphatic level of criticism, I'd guess it was materially worse than the 70% range in order to flat out ditch it at some point - but I don't know.
 
I'd have to check Abbot's work to see if the 77% was all-inclusive or was just "non-mortars". Obviously, it also depends how many firings of each were involved. I've never seen any calculation on the Confederate side. Given the lack of really great options and the emphatic level of criticism, I'd guess it was materially worse than the 70% range in order to flat out ditch it at some point - but I don't know.
I don't think you have to have it be worse than 70% to outright ditch - 30% failures is after all already pretty bad. It reminds me of the Armstrong gun vs Parrott gun, because the Armstrong gun was ditched for unreliability and yet it was more reliable in service than the Parrott gun of similar size (which was kept).
 
I don't think you have to have it be worse than 70% to outright ditch - 30% failures is after all already pretty bad. It reminds me of the Armstrong gun vs Parrott gun, because the Armstrong gun was ditched for unreliability and yet it was more reliable in service than the Parrott gun of similar size (which was kept).
It's all relative. On the Union side, at least, the Bormann was considered better than paper or wood fuzes. Abbot found that the Dyer and Parrott paper time fuzes had success rates of 71%-75%. And the more variable manufacturing issues for paper time fuzes led the Government to reduce the authorized manufacturers. Again without available numbers, the manufacturing issues in the CSA were likely worse. As already noted, it appears that they largely scrapped the Bormanns for paper in 1863 with resulting variations in manufacture. We're both speculating as to the % but Alexander's tone suggests that if it were better than 2/3, he would not have had a problem with it. Again. for the mid-19th century that's probably within the tolerance of "acceptable", given all the other factors.
 
Kudos... you folks know way more about civil war era artillery than I ever will. Heck, you probably forgot more than I will ever know. Thanks for sharing... interesting stuff.
 
Kudos... you folks know way more about civil war era artillery than I ever will. Heck, you probably forgot more than I will ever know. Thanks for sharing... interesting stuff.
Or, speaking for myself, at least we pretend to .... :bomb: Seriously, there are some folks out there who may as well have invented the stuff. I'm just a beneficiary of what they generously pass on.
 
Hmm, segregating out the mortars presumably means that the other ones had a lower reliability rate (well, it states it outright, that's how statistics works). Do we know what the Confederate fuze failure rate was?
Based on the mentioned Petersburg statistics then I'd say that if the Confederate fuze was in the 70%-75% bracket then it'd be statistically indistinguishable from the Union one.
So I double-checked Abbot's calculations directly (rather than using a summary). It appears that the Bormann 77% number was solely howitzers and Napoleons. The mortar numbers actually appear to be mostly wood time fuzes (with one paper fuze). He had higher % for Parrott (85%) and Schenkl (82%) percussion fuzes; the numbers I put up previously for Parrott and Dyer paper; 55% for Schenkl combination, and 85% for the Sawyer combination fuze. (The combination fuzes varied, but the idea was a plunger setting off the burn but detonating on contact if the burn had not yet done so first). In terms of numbers, his results involved c, 13,000 Parrotts (both), c. 3,700 Schenkls (both), and 1,312 Bormanns.
 
Right, so that means that functionally the Bormann number's 95% confidence interval is between 75% and 79%. That's enough samples to give a fairly tight confidence interval.


...I admit I sort of wonder about the mechanics of how he measured the number of shells that detonated on time, especially for the Parrotts... I'm imagining a polite letter to the Confederates asking how many times a shell detonated effectively.
 
Right, so that means that functionally the Bormann number's 95% confidence interval is between 75% and 79%. That's enough samples to give a fairly tight confidence interval.


...I admit I sort of wonder about the mechanics of how he measured the number of shells that detonated on time, especially for the Parrotts... I'm imagining a polite letter to the Confederates asking how many times a shell detonated effectively.
A legitimate question - perhaps answered to some extent (or further muddled) by his categories - of the 1,312 Bormanns, 559 were "uncertain"; 508 "burned well"; 37 "burned variably"; and 109 "did not burn". I come up with 79% after excluding the "uncertain", but what's a couple of % points.
 
A legitimate question - perhaps answered to some extent (or further muddled) by his categories - of the 1,312 Bormanns, 559 were "uncertain"; 508 "burned well"; 37 "burned variably"; and 109 "did not burn". I come up with 79% after excluding the "uncertain", but what's a couple of % points.
That's actually kind of worrying, as it makes the results much more uncertain. This is because the 559 uncertain ones (almost half the sample) could have almost any distribution, and while it's legitimate to work with the numbers we have we can't overconclude from it.

For example, it sounds from these categories as though there's no specified category for early fuze detonation (unless that's "burned variably") but if the "uncertain" consists disproportionately of the ones where he didn't discern a detonation then that might mean as many as half of them didn't burst at all. Which would be atrocious.


Quite apart from anything else the confidence interval becomes 74% to 81% by ignoring them entirely, but they represent additional uncertainty so the confidence interval should be wider than that.
 
That's actually kind of worrying, as it makes the results much more uncertain. This is because the 559 uncertain ones (almost half the sample) could have almost any distribution, and while it's legitimate to work with the numbers we have we can't overconclude from it.

For example, it sounds from these categories as though there's no specified category for early fuze detonation (unless that's "burned variably") but if the "uncertain" consists disproportionately of the ones where he didn't discern a detonation then that might mean as many as half of them didn't burst at all. Which would be atrocious.


Quite apart from anything else the confidence interval becomes 74% to 81% by ignoring them entirely, but they represent additional uncertainty so the confidence interval should be wider than that.
Well, I found Confederate stats. I checked my copy of the Alexander article and in a prior statement he says that 4/5 of the shells tested with their Bormanns detonated early. Obviously, that's a 20% success rate - the opposite pole from Abbot's Federal results.

Regarding Abbot's "uncertain", you're now reaching by concluding that while 79% of those that could be determined burned correctly, some materially different and far smaller % of those that could not be determined burned correctly. That's far more in the line of argument than analysis. Nobody's saying that Abbot's sampling is scientifically rigorous but it's what we have, and it correlates to the following: the Confederate Bormanns were so bad that they simply stopped using them and resorted in large part to paper fuzes. The Federals didn't stop using them. Three options: (1) the Federals were stupid; (2) the Federals had no other options; (3) the Federal fuzes were materially better than the Confederate. Gibbon in The Artillerist's Manual is enthusiastic about the Federal Bormann. One other point: when the Bormann was originally tested by the Ordnance Dept in 1854 an unacceptable number detonated early. Refinements were made, the problem diminished significantly, the Mordecai and Ramsey report recommended adoption, and the War Department approved the fuze.

Wiser students of this than I have noted the refined machining capability needed to make these fuzes correctly and the general lack of such capabilities in the South.
 
Regarding Abbot's "uncertain", you're now reaching by concluding that while 79% of those that could be determined burned correctly, some materially different and far smaller % of those that could not be determined burned correctly. That's far more in the line of argument than analysis.
It's more on the line of "while we can compute a confidence interval from the ones which could be determined, that confidence interval must be loosened to account for the fact we do not know what happened for the ones that could not be determined".
There may be some kind of systemic factor at work related to how they cannot be determined (this would be the case if there was response bias - the equivalent effect resulted in the Dewey Defeats Truman headline, because those whose voting intention could be determined in that case where the people who owned phones) but even if we take the default assumption that there is no systemic factor relating to those which could not be determined the fact that they cannot be determined still affects the confidence interval.

That is the nature of statistical analysis.

With that said, of course, the fact that we now have Union and Confederate statistics means that we can be pretty confident that the Union fuzes were better made. Thank you for going to the trouble to find them.
 
Last edited:
It's more on the line of "while we can compute a confidence interval from the ones which could be determined, that confidence interval must be loosened to account for the fact we do not know what happened for the ones that could not be determined".
There may be some kind of systemic factor at work related to how they cannot be determined (this would be the case if there was response bias - the equivalent effect resulted in the Dewey Defeats Truman headline, because those whose voting intention could be determined in that case where the people who owned phones) but even if we take the default assumption that there is no systemic factor relating to those which could not be determined the fact that they cannot be determined still affects the confidence interval.

That is the nature of statistical analysis.

With that said, of course, the fact that we now have Union and Confederate statistics means that we can be pretty confident that the Union fuzes were better made. Thank you for going to the trouble to find them.
No problem. I don't disagree with the "confidence interval" factor standing alone, but that's why I looked at Abbot's "uncertain" category in light of the indisputable fact that the CSA and Federal commands chose radically different approaches to the Bormann - meaning that something almost certainly was different about the two fuzes. Ultimately, the "problem" with the Federal Bormann was not its quality - it was, as noted, the designed 5 second ceiling and resulting range limit which made it less useful for the rifles. There were a couple of attempts to manufacture it with an expanded time limit but by war's end those had not advanced much.
 
Not really, no, there's very little improvement in weapon employment. Almost the entirety of the greater effectiveness of rifled weapons comes from greater range, and the open fire range of the defenders at Pickett's Charge was in keeping with the range of muskets (which means that they were, in effect, being used as smoothbore muskets). There is no improvement in infantry weapon effectiveness, and indeed something of a reduction because the rifle muskets load slower.

Of course, my understanding is that it's the Confederates who were universally armed with rifle muskets at this point; the Union hadn't quite finished changing their weapons out for 1st class rifle muskets yet.
Thank you for the information. I am not the really informed as to armaments . With the information which is available as to the charge ,one question, was there really any chance of this charge succeeding ? If the artillery bombardment had been effective and if the distance which they had to transverse had not been so long, allowing for Union artillery to have the destructive effect it had ,would it have been effective? Meade had the troops which he could move into that area before Pickett reached the Union line, remember that Meade could observe Lee's movements from Round Top. What if the attacks on the flacks had been successful, would that have had any effect? I have read that the worse part was that the lay of the land had a large part to play, the roll of the ground and the wooded fence which they either climbed over or take down slowed the movements of the troops . IF the Confederate army had marched on after Manassas to Washington ,then this would all be irrelevant, Maybe ,Prehabs. One last inquiry ,who had the better staff officers/division command Lee or Mead? I believe that these men played a large part in the outcome. I take Mead
 
Thank you for the information. I am not the really informed as to armaments . With the information which is available as to the charge ,one question, was there really any chance of this charge succeeding ? If the artillery bombardment had been effective and if the distance which they had to transverse had not been so long, allowing for Union artillery to have the destructive effect it had ,would it have been effective? Meade had the troops which he could move into that area before Pickett reached the Union line, remember that Meade could observe Lee's movements from Round Top. What if the attacks on the flacks had been successful, would that have had any effect? I have read that the worse part was that the lay of the land had a large part to play, the roll of the ground and the wooded fence which they either climbed over or take down slowed the movements of the troops . IF the Confederate army had marched on after Manassas to Washington ,then this would all be irrelevant, Maybe ,Prehabs. One last inquiry ,who had the better staff officers/division command Lee or Mead? I believe that these men played a large part in the outcome. I take Mead
I believe this to be the case:

- The attack had at least some chance of succeeding, including with the distance of the approach march.
- A key factor which caused the failure of the attack was that some of the Confederate units involved got misaligned and had to oblique under fire.
- Meade did not have any remaining reserves. (If you think otherwise, name the brigades - I believe it to be the case that every brigade Meade had was deployed, which means they were not available as reserves. Pulling troops out of the line to then move them in as reserves is much more time consuming than deploying troops that were already in reserve.)
- I understand that Meade was out of comms (his HQ had been hit by "overs") and was unable to influence the battle at that point anyway.
 
Kudos... you folks know way more about civil war era artillery than I ever will. Heck, you probably forgot more than I will ever know. Thanks for sharing... interesting stuff.
I agree. Their knowledge on artillery is very impressive. That area is not my expertise, so my head is spinning... but I have greatly enjoyed reading their posts.
 
Longstreet wrote this some years after the war, although he knew at the time that 12,500 men would most likely not be able to break the Union center. If Lee had sent 30,000 men across that field, would the charge have been successful?
 
Last edited:
Yes, it's the mid-19th century. That doesn't change that, once a shell is fired, whether or not it will overshoot is nothing to do with the fuze and everything to do with the trajectory. The fuze may detonate long (which doesn't matter for a common shell, but results in reduced effectiveness or no effectiveness for a shrapnel shell) but there is no way I am aware of in which the fuze being defective can make the round fly past the target if the shot was a good one without considering the fuze.

If that matches your understanding, then we don't have an argument, but I certainly thought that that's just what I'd been saying all along...
You are correct about the trajectory = distance=sighting the gun for proper trajectory. Has nothing to do with the fuse.
 
Back
Top