As you might guess, when thinking of the Crater at Petersburg, a subject which crosses my mind is the use of artillery in the operation. Not to diminish the other aspects of the battle, but the artillery of the Army of the Potomac played an important role there… and is somewhat overlooked in my opinion. I’m not an expert in the battle. So I would direct you to one of many folks who have written book length treatments of the battle.
My schedule has prevented me from writing up more on Petersburg up to this time. Likely, given the sesquicentennial pace, I’ll have to put that on my “after April 2015” stack. But I did want to mention the artillery’s role and provide a graphic depiction, by way of Brigadier-General Henry Hunt’s map:
The map, and a busy map it is, includes a table breaking down by battery the type and number of guns engaged on July 30, 1864:
For those who are squinting, the roll call is eighteen 4-½-inch rifles, two 20-pdr Parrotts, fifty-two 3-inch rifles (3-inch Ordnance or 10-pdr Parrotts), thirty-eight 12-pdr Napoleons, ten 10-inch mortars, sixteen 8-inch mortars, and twenty-eight Coehorn mortars. Grand total is 164 guns and mortars brought to bear on the Confederate lines in support of the assault.
Some of that number were in the 18th Corps sector and not firing directly in support of the assault. Others were, likewise, firing on the 5th Corps front well to the south of the crater. But all were firing at some time that morning to suppress or pin down the Confederates in conjunction with the assault. For comparison, the “great bombardment” by the Confederates on July 3, 1863 during that “contest” at Gettysburg involved about 140 guns.
Hunt’s map indicates not only the battery positions, but also what the targets were. This adds to the “clutter” on the map. But this is an incredible resource for determining his intent with respect to the fires placed upon the Confederate lines.
The snip above looks at the area of the mine, and just south. Notice there are more dashed blue lines leading to the Confederate redoubt south of the mine than there are the redoubt above the mine. Suppression of the Confederate line was the intent there.
Another Federal position worth noting is that of Company C, 1st Connecticut Heavy Artillery. Battery number 8 on Hunt’s map contained ten 10-inch mortars. Circled here in blue.
Those mortars fired on approximately 1,000 yards of the Confederate front, to the south of the crater (blue shading in the snip above). Recall, these mortars were firing, for at least part of the day, case shot as constructed under Colonel Henry Abbot’s instructions. Battery Number 19, Company B, 1st Connecticut, with six 4-½-inch rifles, located north-east (center-right on the snip above) of the mortars also covered a large section of the Confederate lines.
One problem with these arrangements is that suppressing fire requires a high rate of ammunition expenditure. Suppressing fire cannot be sustained, even by a master artillery chief such as Hunt, for longer than a few hours. At some point, fresh ammunition chests must be rotated in. The assault had to quickly achieve the initial objectives, or lose the suppressing fire support.
Paddy Griffith’s Battle Tactics of the Civil War is nearly three decades old now. Perhaps I’m showing my age by saying the book came out while I was an undergraduate in college. From a personal perspective, it was a landmark book – the first “dive” I made into the nuts and bolts of Civil War tactics coming at a time when my instruction drove me to consider all the footnotes – and chase those footnotes with deliberation. Over the years, my well worn copy of the book has yellowed and aged. Likewise my opinion of the work has aged. I can’t say I accepted all of Griffith’s ideas at the start. Nor can I say I have completely rejected all of it now. But along the way, just as the lines in a pitched battle, the delineation has moved about considerably.
However one point which I can say remained a “salient” throughout, and comes to mind during my recent writings here on the blog, is a passage in which Griffith discusses the ratio of smoothbore Napoleon guns to light rifled guns in the Army of the Potomac. Leading into the paragraph, Griffith noted some disadvantages of rifled guns – shells tended to drill into the ground before exploding, defective shells, small caliber and thus small bursting charge, and, lastly, the limitations of small bore canister rounds. Having established at least a need to retain a mix of smoothbores and rifles, Griffith went on to say:
The limitations of rifled artillery were fully understood at the time, and although the Confederates could never get enough for their needs the Union forces complained of an overabundance of this type of weapon. Early in the war McClellan had decided that the Army of the Potomac needed only one rifle for every two smoothbores, but he never succeeded in bringing the ratio down to less than two rifles to each smoothbore; nor could Hunt force it much lower when he tried to get rid of some of the rifles after the battle of Fredericksburg. At the start of the Wilderness battle smoothbores were still in a minority and it was only when Grant sent home 122 pieces, in order to disencumber his collapsing road network, that Hunt was able to seize a fleeting opportunity.
There is a paragraph with all sorts of exposed lines and open flanks! And those flanks are in the footnotes. Griffith cited two secondary sources for this paragraph – L. Van Loan Naisawald’s Grape and Canister and an unpublished PhD. thesis by Perry Jamieson. While Jamieson’s later works include Attack and Die, Crossing the Deadly Ground, and other works addressing tactics of the era, his thesis is, to my knowledge, still unpublished. So I’ll not address the one page cited from Jamieson’s thesis as part of my rebuttal here.
However, Naisawald’s book should be familiar to any artillery-minded reader. Naisawald’s work came out in 1960. I’ll say it is “dated” and leave the matter there. That, of course, does not allow me to dismiss it as a source. Quite the opposite! The passages cited from Naiswald lead us first to the organization of the Army of the Potomac under McClellan. Naiswald states,
… the short-range, light 12-pounder smoothbore – the Napoleon – was to be the backbone of his artillery; two-thirds of the field batteries were to be equipped with this weapon, and the remaining one-third with rifled cannon – a new innovation in warfare.
So what was Naiswald’s source? He does not offer. I would, however, refer back to Brigadier-General William F. Barry and that initial organization in the summer of 1861. Barry suggested a ratio of guns somewhat dissimilar to that given by Naiswald, and broke distinctly upon the ratio of guns to howitzers, not smoothbore to rifles. Only after the Peninsula Campaign did Barry suggest the howitzers and 6-pdr field guns should be replaced completely by Napoleons.
Naiswald went on to say that Napoleon production lagged early in the war while rifled guns rolled out in large numbers. That, Griffith could have called upon to explain the abundance of rifles. But the overly generalized statement about gun production falls apart when one considers the raw numbers – guns accepted by the ordnance department, specifically looking at Napoleons, Parrotts, and Ordnance rifles:
I probably should give you a fancy, colored chart. But the numbers speak fine enough themselves. Napoleon production peaked twice – late 1861-to mid-1862; then again in the fall and winter of 1863-4. Production of the Napoleons, for the Federals that is, ceased entirely after that. On the other hand, after the initial surge to start the war, rifled gun production remained comparatively steady. Save one quarter with no deliveries, most quarters included delivery of over 60 guns. The totals of all weapons is worth considering here – 2591 guns of these three types, where I would estimate the Federal armies (all armies) only needed 1000 to 1300 field guns of all types. (And that by the way, is one reason we have so many of these guns still around today as memorials!)
Griffith also cites two other passages from Naisawald – one detailing the artillery re-organization made in the Winter Encampment and the other discussing the re-organization made in mid-May. I’ve linked to my blog posts relating to those two specific changes, to keep things brief. Bottom line here, Naisawald never provides an overall count of Napoleons and rifles. Nor does he offer any analysis of the ratio. For good measure, at the time of the reorganization, every corps had an equal number of Napoleons and rifles – 24 of each. The Artillery Reserve included 36 Napoleons and 24 light rifles (and fourteen “heavy” rifles). The rifles predominated in the Horse Artillery where 16 Napoleons paired with 44 rifles. But weight and tactical application was the justification for the disparity in rifles with those horse batteries. In short – the figures do show a slight majority of light rifles, but only due to the horse artillery. There was not a “two rifles to each smoothbore” ratio as Griffith stated.
All of this leads up to those 122 guns “sent home” by Hunt. Did Hunt select rifled guns so as to balance the ratio? I submit that was not the case. Hunt chose to reduce every battery by two guns, which in effect retained the ratio. Aside from that across the board reduction and the 20-pdrs, the batteries that Hunt “sent home” were from the Ninth Corps. If Hunt was seizing any “fleeting opportunity,” it was to send away some batteries he had suspicions about.
There! Thus ends a Paddy Griffith rant. I feel better now.
(Citations from Paddy Griffith, Battle Tactics of the Civil War, New Haven: Yale University Press, 1989, Page 169; L. Van Loan Naisawald, Grape and Canister; The Story of the Field Artillery of the Army of the Potomac, 1861-1865, Oxford University Press, 1960, page 35.)
Yesterday’s post about the employment of Pratt’s 4.5-inch Rifles at Kelly’s Ford ran a little long. And there’s one more interesting angle to look at, given the records from the engagement. The participants provided a remarkably well detailed list of ammunition expended in the engagement. While not a major battle, the action at Kelly’s Ford on November 7, 1863 is a good case for reviewing what ammunition the battery commanders and section commanders selected for the tactical requirements.
Captain George Randolph provided a list of ammunition expenditures, by type, in his report of the action. Let us start with the 10th Massachusetts:
Schenkl case-shot, 3-inch – 300
Hotchkiss percussion shell, 3-inch – 40
Hotchkiss fuse shell , 3-inch – 50
Schenkl percussion shell, 3-inch – 10
The total given by Randolph – 400 rounds – does not match that reported by Captain J. Henry Sleeper – 459 rounds. So either eight ammunition chests, with fifty rounds each, of 3-inch projectiles were used up. Or a little over nine were used. Of course the expenditure does not indicate any canister or bolts were fired. So there were more than just eight or nine chests opened up.
And you are also thinking about Brigadier-General Henry Hunt’s concerns about mixing different rifled projectile types. Sleeper had his Schenkl and Hotchkiss shells all mixed in. Regardless, it was the case-shot Sleeper and his gun-chiefs selected most for their targets. In his report, Sleeper mentions firing on a brick mill building where Confederate sharpshooters were posted. He later replied to Confederate artillery attempting to drive his battery off. After the Confederate artillery retired, Sleeper fired on Confederate infantry that attempted to reform on the hills beyond Kelly’s Ford. Based on the wording of his report, and that of Randolph’s, the battery engaged those infantry targets for the longest period of the engagement. So it is logical to presume that is when most of the Schenkl case-shot were fired -targeting infantry inside a wood line.
As mentioned yesterday, Captain Franklin Pratt’s Battery M, 1st Connecticut Artillery fired but 15 Schenkl shells with percussion fuses. These were fired at brick buildings where the Confederate infantry sheltered and later on the Confederate battery.
Moving next to Lieutenant John Bucklyn’s Battery E, 1st Rhode Island artillery, Randolph indicated they fired a total of 181 shots from their 12-pdr Napoleons:
Solid shot, light 12-pdr – 80
Spherical case, light 12-pdr – 72
Shell, light 12-pdr – 24
Canister, light 12-pdr – 5
Bucklyn’s guns went into battery about 300 yards from the ford itself. Their first targets were the skirmishers on the distant bank. When Captain John Massie’s Confederate guns opened upon Slepper’s battery, Bucklyn turned his Napoleons on that target. Likely most, if not all, of the solid shot fired were expended at those targets.
Later, when supporting the Federal infantry crossing at the ford, Bucklyn fired a few rounds of canister. Again, let me pick at how, and how few of, the canister were used. Five rounds fired to cover the advance of the infantry. Bucklyn’s guns fired those so close that he later lamented the death of one of the friendly infantry, “but they were so nearly between me and the enemy, the accident could not have been avoided.” Or what we’d call today “Danger Close.” Keep in mind the maximum effective range of the canister rounds was between 300 and 400 yards. If Hunt’s earlier complaints were valid, then the canister was designed with engagements at that range in mind. So let’s dispense with the notion canister was only a defensive projectile. At Kelly’s Ford those canister rounds were useful in the offensive because of their “reach.” But of course, with the crossing effected so quickly (as compared to say a crossing at the same point on March 17, 1863), only five canister were needed.
One other note about Bucklyn’s expenditure. In his report he complained, “I found my fuses very unreliable; some shell did not burst at all, while others burst soon after leaving the gun. I could place no dependence on them.” Those 12-pdr shells used Boremann fuses. Randolph seemed perplexed by this issue, “for I have seldom known them to fail.”
Finally, and this is a bonus round, Captain Frederick Edgell’s 1st New Hampshire Battery fired sixty rounds during a separate action on November 8:
Schenkl case-shot, 3-inch – 20
Schenkl percussion shell, 3-inch – 10
Hotchkiss time fuse shell, 3-inch – 30
about a mile north of Brandy Station, a section of Edgell’s guns deployed and opened fire on a Confederate battery at the range of 2,000 yards. After a few rounds, the Confederate battery fell back. Edgell then moved up to the “left of and near Brandy Station.” There at a range of 1,800 yards, Edgell’s 3-inch rifles traded shots with two 20-pdr Parrotts and two smaller rifles. Edgell reported expending 56 rounds, while Randolph recorded an even 60. The preference, Edgell’s 3-inch rifles firing in counter-battery mode, was shell, with some case-shot mixed in for good measure.
From the expenditure figures for these four batteries in two engagements, consider these preferences:
3-inch rifle firing on troops in the woods – case shot.
3-inch rifle firing counter-battery – shell
12-pdr Napoleon firing counter-battery – solid shot, though the preference cannot be stated for a fact.
12-pdr Napoleon firing in direct support of infantry advance – canister, within range limitations.
4.5-inch Rifle – shell at anything.
There’s a lot more I could suggest or speculate towards. But what I see with the artillery employment and ammunition expenditures is a lesson in how Civil War era armies effectively employed artillery in the offensive. The guns firing over the Rappahannock on November 7, 1863 (and those later firing around Brandy Station on November 8) succeeded in pushing the opposing forces back and then kept them back. That accomplished, the infantry was able to conduct their most important mission on the battlefield – occupy terrain.
(Sources, OR, Series I, Volume 29, Part I, Serial 49, pages 566-574.)
Mentioned earlier, this piece on the Five Forks battlefield is interesting for several reasons.
Before the normal “walk around,” a bit about A.B. Reading & Brother. Vicksburg, Mississippi plantation and businessman Abram Breech Reading operated a foundry and machine works near the river along with his brother C.A. Reading. As one might expect, the firm turned out products supporting steamboats and light industry. Shortly after the start of the war, the firm turned to military products. But later that year, the firm leased out much of its equipment to the Atlanta Arsenal and ceased cannon production themselves. Yet, between December 1861 and May 1862, receipts credit A.B. Reading & Brother with delivery of 45 cannons. All were bronze field pieces – 6-pdr guns, 12-pdr howitzers, and 3-inch rifles. It is the 3-inch rifles this post will focus upon.
Reading delivered at least fourteen 3-inch rifles. But there were some variations in the production lots. Compare the values provided on a receipt issued in January 1862 –
… with another in June 1862 –
Hard to read? Here’s the summary:
December 14, 1861 – one 6-pdr weighing 844 pounds.
December 31, 1861 – one 6-pdr weighing 844 pounds.
December 31, 1861 – one “6-pdr rifle” weighing 957 pounds.
January 6, 1862 – three 6pdrs averaging 844 pounds.
January 6, 1862 – three “6-pdr rifles” weighing 956, 659, and 955.
March 25, 1862 – three 6-pdrs averaging 808 pounds.
March 25, 1862 – three 3-inch rifles averaging 875 pounds.
April 12, 1862 – three 6-pdrs averaging 809 pounds.
Noting these variations, the writers of Field Artillery Weapons of the Civil War speculated there were at least two different casting patterns in use. The early batches of 6-pdrs is about forty pounds lighter than a US regulation Model 1841 6-pdr, but within tolerances. One surviving Reading 6-pdr is a trophy at West Point and conforms generally to the Model 1841 pattern. The heavier weight listed for the “6-pdr rifle” are within the range expected for a 6-pdr bored out as a 3-inch rifle. The weights are just twenty pounds or so heavier than that recorded for early Tredegar Bronze 3-inch rifles using the Model 1841 envelope. Given those weight figures, Reading likely used the Model 1841 casting pattern for both 6-pdrs and 3-inch rifles.
But for the later batch, those weights are much lower than expected for Model 1841 or derivatives. And that might easily be explained by a reduction in length, either to simplify the casting or a reduction in precious bronze. The Model 1841 guns were 65.6 inches long, while that piece pictured above at Five Forks is only 63.5 inches overall (61 inches without the knob). The shorter length and reduced profile accounts for some of the 75 pound difference, if not all, in both the 6-pdr smoothbore and 3-inch rifles from the later batches. Enough circumstantial evidence to argue Reading used two different casting patterns.
All four surviving Reading 3-inch rifles match the later pattern. All but one have an erratic set of stampings on the right trunnion.
The stampings are in two different sizes. The top line, curved with the trunnion edge, reads “A.B.R. and Bro.” Early “cannon hunters” failed to see the period after the “R” and interpreted that as “A.B. Rand Bro.” and thus could not correctly identify the firm. The second line notes the firm’s location in Vicksburg, Mississippi. In a smaller font is the year of manufacture “1862” and below that is the gun’s foundry number – 24.
The left trunnion displays three letters – “COL.” This appears to be a post-delivery stamp and might be post-war. Notice the trunnion face is a bit recessed from the carriage cheeks. And the trunnions require a spacer to fit properly on the carriage. The Reading Rifle’s trunnions are slightly smaller than those of the Quinby & Robinson Rifles of the same caliber.
The breech profile matches that of the Quinby & Robinson 3-inch rifles at Petersburg – well-rounded knob, thick fillet, rounded breech face, and a base ring. Notice the vent is bouched.
The base ring is about 1 1/8 inches wide. The stamping to the right of the ruler is an Army depot tracking number.
The trunnions and rimbases also match that of the Quinby & Robinson 3-inch rifle.
The muzzle is straight, lacking any swell. The front sight post sat directly on top, where a tapped hole is now.
The bore diameter is, as advertized, 3-inches.
But look a little closer at that rifling, particularly at the edges on the muzzle face. Those are “sawtooth” grooves, often used by Confederate cannon makers.
The grooves are left-handed. This particular gun has a bit more bore wear than the Quinby & Robinson gun.
The rifling pattern is the only significant difference between the guns produced by Quinby & Robinson and A.B. Reading & Brother. These are “cousins” in most other respects. As seen with the James series, bronze was not the best metal for rifled field pieces. The bronze rifles compared dis-favorably to 3-inch Ordnance Rifles encountered on the battlefield. No doubt some of the 3-inch rifles were melted down by the Confederates for casting into more useful 12-pdr Napoleons.
One that did survive the war and post-war scrapings is A.B. Reading & Brother’s number 24. Today that rifle sits a long way from its place of origin, guarding Five Forks – an obscure gun guarding a famous crossroads.
In the last cannon post, I closed with mention of similar 3-inch rifle patterns used by both Memphis gunmaker Quinby & Robinson and the Vicksburg firm of A.B. Reading & Brother. You saw this photo in the earlier post:
Compare to this 3-inch rifle on display at Five Forks:
The two rifles share a common external form. Although records are silent, the most likely explanation is Confederate officials distributed documentation for the patterns. But there is also a formal connection between the two companies, with a sub-contracting twist.
Quinby & Robinson and A.B. Reading were among several companies in the Mississippi Valley which answered the Confederate calls for ordnance early in the war. Even without any formal experience in the field, both firms made credible progress. But of course, like many other Confederate gunmakers, the efforts suffered with the lack of raw materials. And like other Western Theater companies, the two firms were endangered by Federal advances.
Quinby & Robinson suffered one additional setback. On September 30, 1861 a fire destroyed much of the Memphis firm’s facilities. The timing of the fire was unfortunate. Prior to the fire, Quinby & Robinson was able to deliver up several cannon a week. After the calamity, the firm had several unfinished castings on hand without the means to complete. So the Memphians turned to other vendors. A receipt from February 20, 1862 indicates Quinby & Robinson turned to at least three different firms. Two of which were in Vicksburg.
Listed on the receipt, A.B. Reading & Brother completed four 12-pdr howitzers and one 6-pdr gun for Quinby & Robinson.
A line below, A.M. Paxton, another Vicksburg firm, received credit for finishing three more 6-pdrs. There was more work for Paxton, but I’ll save that as a tangent to follow on a follow up post. The only 3-inch rifles listed on the receipt were two completed by the Memphis entrepreneur George Washington Grader.
Still, if Quinby & Robinson sent unfinished castings to Vicksburg, is it also plausible for them to send detailed plans and casting models? If so, that might account for the nearly identical guns provided by Quinby & Robinson and A.B. Reading & Brother.
Nearly identical, I say. There are some differences between the rifles from the two companies. To see those differences, one has to look down the bores.
In February 1862, Major William Richardson Hunt approved receipt of over $2500 of ordnance from the Memphis firm of Quinby & Robinson.
The third item listed on the receipt records “1 6 pdr 3 in Rifle Gun” received on February 6 at a cost of $687.43. (Recall the nomenclature used for other Confederate 3-inch rifles incorporated similar references to the base 6-pdr caliber.) The 3-inch rifle was one of only a handful, perhaps only three, produced by Quinby & Robinson before the fall of Memphis that spring. Remarkably two of the guns survive today in Petersburg National Battlefield.
One is on display near the visitor center.
The other is located at Colquitt’s Salient opposite Fort Steadman.
At first glance the gun presents a clean appearance, with minimal moldings confined to the base ring and knob. The cylindrical rimbases attach directly to the gradually tapering barrel. Small numbers on top of the breech (#33 on the piece in front of the visitor center and #34 on the gun in the field) should correspond to a foundry numbers. The stamps on the right trunnion indicate the guns are indeed from Quinby & Robinson.
The year stamped on the left trunnion of each piece, 1862, puts the guns are in the range corresponding to the receipt shown above.
The thickness of metal at the muzzle suggests the original casting pattern was intended for a larger caliber weapon.
The bore features twelve left-handed twist lands and grooves. Remarkably, neither gun exhibits significant wear of the rifling.
The bore measures out at the prescribed 3-inches.
The breech profile incorporated a base ring, rounded breech face, and a rounded knob with rather thick fillet connecting to the breech.
The gun sight mounts are no longer attached. But the fittings indicate the use of a standard hausse seat in the rear and a spike front sight above the muzzle.
Of the pair, #33 definitely has more “character.”
The divot under the lower left of the barrel looks like a battle scar. But it could also be the result of mishandling. But it sounds so much more exciting to say some Yankee solid shot ricocheted off the barrel in the heat of some artillery duel. The damage deformed the interior of the gun and actually warped the bore.
Needless to say, #33 won’t be firing any more rounds.
Up until the recent refurbishment of the Petersburg artillery display, #33 sat on the rails between a James Type 2 14-pdr rifle and a Wiard 2.6-inch rifle, allowing for convenient comparison.
The Confederate rifle measures 61 inches long, compared to 74 inches for the James rifle and 52.5 inches for the Wiard.
The external appearance of these two Quinby & Robinson rifles, even if breaking with established patterns, is not unique. Another pair of 3-inch rifles at Petersburg, produced by A.B. Reading and Brother, from Vicksburg, Mississippi. I will examine them next.
Last fall I profiled a 3-inch iron rifle, located today at the Arlington County veterans memorial in Clarendon, Virginia, made by the Noble Brothers & Company of Rome, Georgia.
The profile of the Clarendon 3-inch rifle resembles contemporary Tredegar guns of the same class and caliber. No doubt among the earliest actions of the Confederate Ordnance Department was to forward the designs to J.R. Anderson and the Noble brothers. The Clarendon rifle dates to November 1861. Noble brothers turned to bronze for both 6-pdr smoothbore and 3-inch rifled calibers, in addition to a fair number of 12-pdr howitzers.
While the howitzers remained, generally speaking, faithful to the established patterns, the field guns and rifles exhibited several deviations. One of which was a flattened knob resembling an empty spool. The knob is unique to Noble guns. Although unmarked, a 3-inch iron rifle at Manassas National Battlefield Park has just such a knob.
The 3-inch rifle is noticeably shorter than other weapons of its class, with the elevating screw resting at the very edge of the breech. This gun also lacks the muzzle swell seen on the Clarendon Noble gun.
This particular gun is among a set representing Confederate massed batteries which fired upon the Federal attack at the Deep Cut on August 30, 1862. However, until a few years ago the gun sat near the Stuart’s Hill / Park Headquarters at an overlook of the southwestern portion of the park.
As you can see from the wide view, the muzzle face is very rough. Almost appears to be cut or trimmed after manufacture.
But we can rule such “surgery” out, as the gun is balanced on the trunnions – despite being much shorter than other weapons of its class.
A look down the bore shows six groove rifling, although a little corroded… and now used as a dwelling for a feathered friend of the battlefield.
The gun also bears a “scar.”
The slash of that scar appears to be from improper handling as opposed to battle damage. There are significant lateral scrapes to go with the scar, also indicative of some bad handling.
Aside from the trophy number plaque, the upper breech shows a “ghost” of a sight bracket.
Lacking any markings to read, the best one can do is associate the weapon based on external features. The little field rifle at Manassas has the look of the Noble brothers’ work. But pending the cleanup of layers of paint and decades of oxidation, attribution is tentative.