Percussion caps and friction primers: Q&A on students in Georgia who found evidence of wild cavalry chase, valiant rearguard stand

Big Buchkead Church, percussion cap (top right) and artillery primer (Camp Lawton project)

Members of the archaeology program at Georgia Southern University have pinpointed where a Ohio cavalry regiment helped successfully hold off charging Confederate horsemen during Sherman’s March to the Sea, the head of the project says.

Students at the university in Statesboro participated in a summer field school from May 15-June 16 at the site in Jenkins County.

Associate professor Ryan McNutt, who heads up the school’s Camp Lawton Archaeological Project, said about 1,000 artifacts were recovered. “I think we’ve made a very good start to confirming the location of portions of the Buckhead Creek battle lines, and this is something that future work will only develop and refine.”

The project for several years has been researching the remains of a nearby Confederate prison camp that was in operation for several weeks in fall 1864.

In 2020, the university was awarded a $116,247 grant from the National Park Service’s American Battlefield Protection Program to document and evaluate the archaeological integrity of two skirmish sites toward the end of Gen. William T. Sherman’s march to Savannah: Buckhead Creek and the subsequent Lawton (Lumpkin’s) Station.

Students tackled the Lawton Station fight first, finding evidence in early 2021 of an engagement.

The program decided to study Buckhead Creek this year, concentrating on property around an historic church caught up in the fighting.

The Battle of Buckhead (or Buck Head) Creek on November 28, 1864, involved cavalry forces under Union Brig. Gen. Judson Kilpatrick and Confederate Maj. Gen. Joseph Wheeler (below, right). It took place across what are now Jenkins and Burke counties.

Kilpatrick (above, left) was in the area to destroy railroad between August and Millen and burn a trestle. Another objective was to release the Camp Lawton prisoners, but Union forces discovered they had been moved to other sites. Federal forces were able to destroy a mile of track.

On the 28^th, Wheeler “almost captured Kilpatrick, and pursued him and his men to Buckhead Creek. As Kilpatrick's main force crossed the creek, one regiment (the 5^th Ohio Cavalry), supported by artillery, fought a rearguard action severely punishing Wheeler and then burned the bridge behind them,” says a National Park Service summary of the fighting. “Wheeler soon crossed and followed, but a Union brigade behind barricades at Reynolds' Plantation halted the Rebels' drive, eventually forcing them to retire.”

The Picket’s questions about Buckhead Creek and McNutt’s written responses below have been edited for brevity and clarity.

Q. Was all of the work at the church? Why there -- any particular historical accounts you wanted to map?

A. All of our work took place around the Big Buckhead (Baptist) Church, on property owned by the Jenkins County Historical Society, for which we had permission to conduct archaeology on. This was for a few reasons. First, primary sources mention the church extensively as a landmark during the conflict over the creek crossing, and the Union withdrawal to Reynolds' Plantation. Secondly, the church’s property encompasses the end point of the causeway from Buckhead Creek Bridge, on line with the historic pilings visible from the modern bridge, and straddles the route of the historic road Gen. Kilpatrick and his 3^rd Cavalry Division retreated along. The (Federal) rear guard artillery action is the bit that likely occurred around the church.

Q. What date was the skirmish or skirmishes your team was studying? Were most of those engaged cavalry?

A. The battle occurred November 28, 1864 -- the action at the church occurred at noon on this date, which we know from a letter from the colonel commanding the 5^th Ohio Volunteer Cavalry who documented the regiment’s involvement in his report. All troops engaged were cavalry, with one light artillery regiment (the 10^th Wisconsin) on the Union side as part of Kilpatrick’s 3^rd Cavalry Division. Wheeler’s Confederate forces are less clear. He seemed to have some artillery attached, and we found some potential evidence of ad hoc Confederate canister shot that seems similar to items recovered from the Battle of Pea Ridge, but this is very tenuous at the moment and needs more lab analysis and documentary research. (Photo of left, Camp Lawton project)

Q. One of your Facebook posts said "pistols were used in the skirmish at Buckhead Creek Church alongside long arms and artillery pieces." Did you find artifacts from all three?

A. We found evidence of .22-caliber rimfires from a Smith & Wesson Model 1, and a lone percussion cap that fits a Colt Army 1860 .44-caliber revolver (or its equivalent). Artillery was identified primarily through fired friction primers, which were all localized in one area of the grid, in a position to deliver enfilading fire on the causeway and bridge across the road. We have some potential evidence of canister tins, and one possible piece of canister shot, but confirming the identification of these is an ongoing process tied to their conservation. Evidence of long arm use so far comes from exclusively percussion caps, in the top hat style, but there are some odds and ends that might be arms-related as well.

Brass rimfire casing for a revolver (Camp Lawton Archaeological Project)

Q. How far below the surface were the artifacts typically found?

A. Fairly deep -- almost all the munitions were recovered between 20 and 25 cms (8-10 inches below surface.

Q. How many in total were recovered? Any personal items or were all arms-related?

A. We recovered probably around 1,000 individual artifacts, and this included some historic glass, lantern parts, ceramics, numerous machine-cut nails, several boot nails and a few potential horse shoe nails, as well as the arms related items.  Many of these domestic items likely relate to the use of the site by the church, which has been continually used a place of worship since before the Revolutionary War and was certainly in existence from 1787. There were no clearly identified personal items from soldiers, but that may be a result of a long-term metal detector activity on the site.

Q. Any related to horse tack?

A. We almost certainly have some iron items that are related to horse tack, but they’re not obviously military in origin.

Reproduction percussion caps have maker's mark less patina (Camp Lawton project)

Q. You found lots of percussion caps. Were those mostly from pistols?

A. In fact, they’re almost all from long arms, and likely Sharps, which were the carbines the 5^th Ohio was issued with -- we found about 23 top hat style percussion caps, in a distinct line with several clusters, that likely indicates a skirmish line (given that the clusters are about 3-4 meters apart across the site, and in an angled line facing the causeway and old bridge site).

Q. The friction primers – Union or Confederate? Their use?

A. These are likely all Union, given their location. Friction primers were a gunpowder-filled copper tube inserted into the touchhole in the rear of the cannon barrel into the gunpowder charge. A roughened wire was fixed into a spur that was filled with antimony sulfide and potassium chlorate, which essentially acted like a matchhead when the wire was pulled through the spur.

A lanyard would have run from the wire to the hands of the gunner and yanking the lanyard pulled the wire out, ignited the friction primer, and fired the gun. The used exploded copper tube was then explosively hurled into the air and to the rear of the gun (depending on the cannon tube’s elevation).

For our purposes, the exact location of these friction primers is quite important, because we have primary accounts describing the presence of artillery at the Battle of Buckhead Creek Church, and general indications of their positioning.

Wartime photo of 5th Ohio Cavalry (Library of Congress)

The 5th Ohio Cavalry Regiment, Company G, had obtained two 12-pound mountain howitzers, and these were positioned to the right and left of the road running from the bridge at Buckhead Creek past the church and were loaded with canister to sweep the causeways leading up to, and away from the bridge. Another possibility is the 10th Wisconsin Light Artillery, who were attached to Kilpatrick’s 3rd Division, and were also involved in the rear-guard action at the creek, and who were also armed with mountain howitzers. Using the location of these friction primers and working out the potential distance these would have been hurled from where the guns were fired, we may be able to actually identify where precisely the Union artillery pieces were placed next to the church and add these to the battle interpretation. Moreover, these small copper pieces of conflict also indicate that there is quite a lot of integrity of the battlefield surviving.

Q. You posted on social media the discovery of buck and ball loads (left). What are those?

A. We found two, possibly three pieces of shot from buck and ball loads. Buck and ball loads were essentially two or more .32-caliber shot loaded into a paper cartridge with a .69-caliber ball. When fired, it had the effect of a shotgun, spreading the lethality of impact across a wider target.

Both of our shot has banding from the barrel of the firearm, and impressions from the .69-caliber ball, which is how we know they’ve been fired, since shot can only pick up those alterations from contact when it goes semi-molten from the powder charge when fired.

Wheeler himself requested buck and ball cartridges to be sent to Millen for his resupply, and when Sherman’s army took Savannah at the end of the campaign, part of the captured Confederate munitions included a total of 11,500 buck and ball cartridges.

Q. What types of equipment were used during the school?

A. We used exclusively metal detecting. There are several potential earthwork features that might be ad hoc Union fortification -- rifle pits, a potential tiny lunette -- that we need to return to and examine with GPR (ground-penetrating radar).

Q. Regarding open and public days, any particular questions or themes raised by visitors?

A. One of the consistent themes raised is how heavily metal detected the area has been in the past, and how surprised they were that were finding items. And also a deep appreciation for us working in the area, and raising the profile of both the battlefield, and the church. Despite its great historical importance, it’s not listed on the National Register (of Historic Places), and part of the final reporting of this project will be to nominate the church itself to the National Register as historically significant.

Dr. McNutt with visitors to the excavation site (Camp Lawton Archaeological Project)

Q. I imagine is analysis is to come, but do you have any takeaways from what was found/observed during the five weeks? Anything become clearer about the fighting around the church?

A. I think we’ve likely pretty much confirmed that we uncovered evidence of the 5th Ohio Volunteer Cavalry’s actions at Buckhead Creek. They were a regiment who served in the 2nd brigade of General Kilpatrick’s 3rd Division. Originally, the entire 5th Ohio formed the rear guard, dispersing Confederate attempts to cross the bridge over the creek with canister shot from two 12 pounder mountain howitzers from Company G commanded by Capt. John Pummill.

John Pummill may be individual at left with 5th Ohio howitzer (Library of Congress)

At noon on November 28 , 1864, they opened fire at the charging Confederates charging over the bridge and ‘when the smoke of [their] discharge cleared away’ the causeways were swept clean. Co D of the 5th Ohio destroyed the bridge, but Wheeler himself said his command used the pews from Buckhead Creek Church to rebuild it.

The 5th Ohio began a slow withdrawal to join Kilpatrick at Reynolds' (Bellevue) Plantation, leaving the 3rd battalion consisting of companies E, I, H, and K lead by Capt. (Alexander) Rossman to defend the rear at the creek crossing. Finally, only Company K was left, fighting dismounted as skirmishers. By this point in the war, Company K held only 61 men, who tenaciously held off Wheeler’s forces. And this accords well with what we’re uncovering in the archaeological record. Contrary to popular belief, only three regiments in Kilpatrick’s entire 3rd division were armed with Spencer repeating rifles. All the rest possessed either Springfield rifle muskets, or carbines of various effectiveness.

The 5th Ohio, which is our most likely candidate for the troops whose positions we’re investigating, were armed with Joslyn carbines at the start of their service, transitioning to Burnsides and then Sharps. As well as the standard issue Colt Army 1860 .44-caliber revolver, most importantly for our purposes, all of these carbines used top hat style percussion caps to ignite their breech loaded cartridges -- the exact style we’re finding in abundance, spaced at regular intervals facing the still visible causeway, strongly indicating a skirmish line. Potentially, the line of Company K. I think we’ve got really good evidence for the position of at least one of the 5^th Ohio’s artillery pieces, and accompanying skirmish line. (Damaged percussion cap in photo)

Q.  What do you think students most experienced/learned at this field school?

A. I think we’ve made a very good start to confirming the location of portions of the Buckhead Creek battle lines, and this is something that future work will only develop and refine. Most importantly, I think we’ve clearly demonstrated that portions of the creek battle site have really good archaeological integrity, with surviving artifacts, battle lines and detritus from the action, despite the extensive metal detecting activity in the area. This is incredibly important for these aspects of the conflict that spun off from Sherman’s March to the Sea, given that the battle of Waynesboro, and so many other of the smaller skirmish sites have vanished under development and urban expansion. I think this publicity and confirmation of intact historic terrain and battlefield material should help heighten the visibility of the battle, and encourage tourism to Jenkins County to witness a landscape in very good condition for visualizing one of the last major battles in Sherman’s March to the Sea.

Dr. McNutt (back row, left) with members of the summer field school team

Our students learned how important this local history is to the people of Jenkins County, and how it intersects with heritage tourism and real world impact. And of course, they also learned the value of hard work in 111 degree heat, and how much team work and strict scientific approaches are necessary to uncovering the past. Most of our percussion caps were in areas that were untouched because they covered over with extensive modern garbage, and it’s only through dedication and discipline that we uncovered them.

Moreover, I think the battle at Buckhead Creek Church really drove home to the students the concept of the Civil War as the first industrialized war, with a stark contrast between black powder muzzle loaders and buck and ball loads utilized by the Confederates, and breech-loading carbines and lever action repeaters and pistols with cased ammunition being used by the Union.