The Missing Screen

During the final days of June 1863, two large armies moved north through Maryland and Pennsylvania without a clear view of one another. Robert E. Lee had crossed the Potomac to relieve Virginia and subsist in the North. His columns now extended from Chambersburg toward Carlisle and York. The Army of the Potomac followed east of the mountains, changing commanders in the middle of the pursuit. Neither army had selected Gettysburg as a battlefield. It was one road junction in a campaign governed by mountain gaps, crowded turnpikes, and incomplete reports.

That setting made cavalry important before it ever charged on a battlefield. An army needed cavalry to find enemy roads and conceal its own movements. It also needed it to guard routes and trains, carry reports, and keep separated columns in contact. The approach to Gettysburg was therefore a contest over information: what was known, how recently, and whether commanders received it in time. For Lee, that problem became urgent at Chambersburg on the evening of 28 June.

Late on 28 June 1863, Robert E. Lee received the report that changed the Gettysburg campaign. James Harrison, a civilian scout employed by James Longstreet, had learned that the Army of the Potomac had crossed the Potomac River and was moving north from Frederick, Maryland. Lee’s three corps were spread from Chambersburg to Carlisle and York. His own headquarters had received no useful report from J. E. B. Stuart for several days. Harrison’s warning compelled Lee to abandon any immediate movement against Harrisburg and draw the Army of Northern Virginia together east of South Mountain.

Stuart’s absence did not leave Lee without sources. Reports had already indicated a Federal movement toward the Potomac. Scouts, civilians, prisoners, and infantry patrols supplied further fragments. Harrison gave those fragments decisive operational force by confirming that the Federal army had crossed and was advancing. What Lee lacked was a mounted screen that could follow the leading columns and report their movement day by day. It also had to guard his army’s routes and maintain contact among its separated commands.

The difference between a report and a screen is the subject of this essay. A report fixed the enemy at one point in time. A working cavalry system could identify his columns, test new information, and give warning when a probe became a battle. Both armies still worked through rumor, fatigue, and an unfamiliar road network. Buford’s division did not eliminate those frictions, but it made better use of the information it gathered. Lee’s cavalry, dispersed by assignment and weakened by failures of communication, did not give him the same continuity during the crucial change from raid to concentration.

What cavalry was for

Civil War cavalry’s distinctive value lay in mobility. It scouted, screened, delayed, pursued, and raided. It also carried messages, guarded trains, and protected an army’s flanks and rear.^[1] These jobs shaped how an army approached battle as much as cavalry combat did.

Reconnaissance sought positive information about an enemy force: its position, composition, direction, and purpose. A screen denied such information to the enemy. The two functions could overlap, but they were not identical. Stuart’s cavalry screen in the Loudoun Valley illustrates the distinction. Pleasonton forced it back at Upperville on 21 June, yet did not break through the Blue Ridge gaps to locate Lee’s infantry. Stuart had lost ground without losing the larger screening mission.^[2]

Observation had to be followed by transmission. Couriers, guides, staff officers, and liaison parties connected a patrol to the headquarters that could act on its report. They also kept friendly columns aware of one another. Cavalry served a second practical purpose on the roads: watching approaches, protecting trains, and detaching hostile raiders from them. When no mounted force performed this work, the infantry had to be held back for it. A covering force added another use for mobility. It could fight from successive positions, give warning, and trade ground for time. Buford would use his division in exactly that way. Raiding could disrupt an enemy’s communications, but it often competed with these more immediate tasks when conducted far from the main army.^[3]

How Lee’s cavalry system came apart

Lee entered Pennsylvania with seven cavalry brigades available in the broadest sense, but they were not concentrated under Stuart for a single mission. Stuart selected the experienced brigades of Wade Hampton, Fitzhugh Lee, and John Chambliss for the movement around the Army of the Potomac. Beverly Robertson and William “Grumble” Jones remained east of the Shenandoah Valley to watch the Blue Ridge gaps and the Federal rear. Albert Jenkins accompanied Ewell’s advance into Pennsylvania. John Imboden operated well to the west, protecting approaches and gathering supplies. This was not the disappearance of Confederate cavalry. Some of its strength was assigned to screening and some to raiding. Other elements had reconnaissance or security duties, but no mechanism was strong enough to reunite them when the campaign changed.^[4]

Lee’s orders of 22 and 23 June reflected the tension. If Federal movements permitted, Stuart could take three brigades across the Potomac east of the mountains, damage the enemy, and collect supplies. But he was also to place himself on Ewell’s right and communicate with him. He was to guard that flank and keep Ewell informed. After crossing the river, Lee wrote, Stuart “must move on and feel the right of Ewell’s troops.” The two brigades left behind were to watch the enemy and guard the passes. Once the Federals departed, they were to retire west of the mountains and close on the rear of Lee’s army.^[5] The instructions granted Stuart discretion over his route and authorized activity against the Federal rear. They also made renewed contact with Ewell, the operational end of the movement.

Stuart’s route did not unfold as expected. Near Haymarket on 25 June, Winfield Hancock’s Federal Second Corps blocked the direct road. Stuart could have abandoned the movement around the Federal army and returned toward the Shenandoah route, but doing so promised a long countermarch and no quick reunion. He persisted east and south, then crossed the Potomac at Rowser’s Ford. Enemy movements and difficult roads imposed real constraints. His horses needed forage, and he knew too little of Federal dispositions. Yet Stuart also accepted delays that served the raiding element of his mission more than the requirement to find Ewell. Near Rockville on 28 June, his men captured a large Federal wagon train. Keeping roughly 125 wagons gave the expedition material value but reduced the column’s speed and tactical freedom. Prisoners required processing. Wrecked wagons had to be burned, and scattered regiments had to reassemble.^[6]

The captured train did not by itself account for every lost hour. The most important obstacle was the Army of the Potomac and its cavalry. At Westminster on 29 June, a small Federal detachment resisted Stuart’s advance. On 30 June, Judson Kilpatrick’s division reached Hanover first, fought Stuart, and forced him to protect the train and detour east before resuming his northward movement. These encounters show why cavalry operations cannot be judged as though one side moved across an empty map. Federal horsemen were screening Meade’s army and contesting Stuart’s routes. The enemy had a vote in when and where Stuart could rejoin Lee.^[7]

Still, the movement’s deepest failure was not that Stuart saw nothing. At Fairfax on 27 June, he learned that the main Federal army had moved toward Leesburg. He sent the intelligence to Lee by courier and sent a duplicate to the Confederate War Department in Richmond. Neither report reached Lee. The messenger failed, while Richmond filed the copy without confirming that army headquarters had received it. The episode isolates the difference between reconnaissance and intelligence. Stuart had collected the essential fact and tried to transmit it, but the Confederate command system did not deliver it to the commander who needed it.^[8]

Meanwhile, Robertson and Jones did not perform the mission left to them. Federal cavalry withdrew from the Loudoun front on 26 June to cover the Army of the Potomac’s crossing at Edwards Ferry. Under Stuart’s instructions, that movement should have triggered Robertson’s withdrawal toward Lee’s rear, with continued pickets and reports. Instead, the two brigades remained near Ashby’s and Snicker’s Gaps until Lee ordered them north on 29 June. There is little evidence of energetic patrolling toward Edwards Ferry or the South Mountain gaps. Jones had a well-earned reputation as an outpost officer; Robertson did not. But seniority placed Robertson in command, and neither Stuart’s detailed instructions nor Lee’s authority converted the force into an active reconnaissance screen.^[9]

Responsibility was therefore distributed across the command. Stuart chose the route, the three brigades, and the captured wagons. Robertson failed to move and report as circumstances required. The Richmond War Department failed to forward critical information. Lee had approved an arrangement in which his best cavalry commander and three best brigades might be incommunicado for days, yet he did not ensure that the remaining brigades covered the obvious observation points. The system had multiple components; by 28 June, none was supplying Lee with a continuous picture of the enemy.

28 June: warning without definition

Harrison’s arrival was strategically and operationally decisive. He reported that the Army of the Potomac had crossed the Potomac and was advancing from the vicinity of Frederick and Middletown. Lee’s army was extended over a wide area: Ewell’s divisions threatened Harrisburg and occupied York, while Hill and Longstreet lay nearer Chambersburg. A Federal advance through the South Mountain gaps could threaten the dispersed Confederate columns and their communications through the Cumberland Valley. Lee responded by ending the outward movement and ordering concentration.^[10]

Yet Harrison did not place the Federal Corps precisely. He could not provide the kind of recurring reports that showed the head of each column, distinguished cavalry from infantry, or calculated when those columns could reach Gettysburg, Cashtown, or the mountain passes. Nor could he screen Lee’s concentration from Federal observation. His report answered the largest question—whether the Union army had followed Lee north—but left the operational questions unresolved.

Lee had received indications that a Federal crossing at Edwards Ferry was imminent. His campaign design also assumed that the Army of the Potomac would follow him. Yet those indications did not establish when the crossing had occurred, how much of the army had crossed, or which route it would take. Harrison supplied the confirmation that required immediate action.^[11]

Lee’s later statement that Stuart’s silence led him to infer that the Federal army had not left Virginia exposed a faulty assumption. Because Lee had authorized Stuart to pass around the Federal rear, he should have expected a period in which that column could not communicate. Silence did not prove Federal inactivity; it more plausibly indicated that Stuart was beyond the enemy’s lines or that his couriers had failed. The inference also placed excessive weight on Stuart personally when Robertson and Jones still had the mission and opportunity to observe the Federal withdrawal. Lee was not blind because one man was absent. He was poorly informed because no redundant source was following the enemy closely enough to replace him.

By the end of 28 June, Lee knew that the Army of the Potomac had crossed north of Washington and was advancing generally toward South Mountain. He could order his scattered forces to concentrate and expect the Federal army to use several northbound roads. He could not reliably identify the leading Federal corps or measure their distance from the road network in south-central Pennsylvania. Nor did he know which gaps and roads Meade intended to use. He did not know where Stuart’s main body was, and he had no mounted screen between his own concentration and the Federal advance. This was enough knowledge to make a sound, broad decision, but it did not give him control over the encounter that was approaching.

29 June: concentration under uncertainty

Lee’s concentration orders were more flexible than the standard story of an accidental march suggests. Hill moved from Fayetteville toward Cashtown, with Heth reaching the eastern side of South Mountain on 29 June and Pender following on the 30th. Longstreet was to follow Hill. Ewell received an initial recall from the Carlisle area, then a revised direction toward Heidlersburg. From there, he could move south to Gettysburg or west to Cashtown while remaining east of South Mountain. The letter preserved in The Wartime Papers of R. E. Lee told Ewell that Hooker had crossed and that the head of his column had been reported at Middletown; it then offered Gettysburg and Cashtown as alternative destinations.^[12]

The dating of that letter is confused in the headquarters record—it is entered as 28 June although the strongest chronological evidence places the revised order on the morning of the 29th—but its operational logic is clear. Lee used parallel routes, avoided sending all of Ewell’s corps back through Chambersburg, and kept two divisions positioned to respond on either side of Gettysburg. Rodes and Early could approach the town from the north while Hill approached from the west. This was not a detailed battle plan. It was a prudent branching concentration designed for an uncertain enemy location.^[13]

The same uncertainty imposed costs elsewhere. On the night of 30 June, only four of Lee’s nine infantry divisions were east of South Mountain: Heth and Pender near Cashtown, Rodes and Early near Heidlersburg. Anderson, Johnson, most of Longstreet’s corps, and Pickett remained west of the ridge or on the roads through it. Artillery and long wagon trains crowded the Chambersburg pike and Cashtown Pass. If battle began suddenly, the road network would determine how fast the army could mass. Lee had no cavalry report that gave him a detailed timetable for the approaching contact.^[14]

Route security further diluted combat strength. With Imboden still west of Chambersburg and Robertson absent, Lee retained Pickett’s division at Chambersburg and Law’s infantry brigade near New Guilford to protect the rear and watch approaches from Maryland. The instructions Lee sent Imboden on 1 July show what mounted security was expected to do. Imboden was to relieve Pickett, guard the roads, control stragglers, and hold key crossroads. He was also to picket the approaches to the army and protect the reserve trains, “upon the safety of which the operations of this army depend.”^[15] These were not ornamental duties. Until cavalry assumed them, infantry that might have advanced toward the battlefield remained tied to the communications zone.

Imboden could not perform that relief immediately. On 30 June, his command was still around McConnellsburg, far west of the concentration, and it did not reach Gettysburg until 3 July. The problem was not simply that Lee lacked scouts in front of Hill. Mounted troops assigned to security were also too distant to release the infantry already performing their work. The army’s cavalry shortage therefore, affected both reconnaissance and the rate at which combat divisions could be made available.^[15-1]

Ewell’s own cavalry illustrates the difference between possession and employment. Jenkins had roughly 1,600 troopers with the corps, but much of the brigade lagged on the return from the Harrisburg area and contributed little to the march toward Heidlersburg. A smaller battalion under Elijah White, attached to Early, did more useful work. White sent parties toward Gettysburg and the northern approaches. They watched roads and detected signs of Federal movement. Yet Buford’s growing screen prevented those patrols from obtaining a precise view. Lee’s army had mounted men in the theater. It lacked a unified forward reconnaissance effort capable of overcoming the Federal cavalry and reporting directly to army headquarters.

30 June: two intelligence systems meet

By 30 June, the campaign had narrowed from a strategic pursuit to a contest for information around a road junction. Confederate scouts established that the Army of the Potomac was somewhere in northern Maryland and not making a major move into the Cumberland Valley. That negative information reduced the immediate threat to Lee’s rear. The more urgent report came from Henry Heth’s division. James Pettigrew took most of his brigade toward Gettysburg, encountered Federal horsemen near the town, and withdrew rather than force a fight. His reconnaissance by accident proved that a sizable enemy force was present, but not what it was. The Federals might be militia, a cavalry screen, or the advance of the Army of the Potomac. Confederate observers had a contact without an identification.^[16]

That limitation reflected method. Sending an infantry brigade down a road could answer whether the road was physically blocked. It was a poor instrument for wide reconnaissance. Infantry could not rapidly patrol parallel routes or turn a position. Nor could it pursue withdrawing pickets while remaining in contact with the rear. Once Pettigrew encountered Buford, the Confederate probe stopped at the screen. No mounted patrol worked around it to discover the infantry columns approaching from Emmitsburg. No chain of reports updated Heth, Hill, and Lee through the afternoon and night. The observation remained static: there were Federals at Gettysburg.

Buford’s division performed the same task at a different level. Moving north from Maryland on the morning of 30 June, Buford sent patrols along the approaches from the west and north. At Fairfield he found Confederate infantry but obeyed orders not to become heavily engaged. A northern patrol captured a man from Rodes’s division. Another party heard of the cavalry fighting around Hanover. Buford met John Reynolds in person at Emmitsburg and reported that the enemy was strong behind Cashtown. One of Reynolds’s staff officers remained with the cavalry column to preserve liaison. Before Buford reached Gettysburg, the information was already moving toward the infantry commander whose First Corps could support him.^[17]

At about 11 a.m. Buford entered Gettysburg as Pettigrew’s men withdrew. He continued the reconnaissance. Patrols went out again, civilians were questioned, and a pass signed by Lee that morning at Chambersburg was examined. His reports mixed firm observations with rumor. One placed Anderson’s division on the wrong route, and accounts differed over Lee’s location. Buford’s advantage was not perfect knowledge; it was the continued collection and transmission of qualified reports.

At 10:30 p.m., Buford could tell Reynolds that Hill’s corps was massed behind Cashtown and that Confederate pickets stood within four miles on the Chambersburg road. He had evidence that Pender had arrived and that Rodes was coming from the Carlisle direction. Other reports, including the supposed approach of a force from York, remained uncertain. Buford also reported his own condition: his men and horses were exhausted, and he lacked forage and rations. Reynolds and Pleasonton received parallel summaries.^[18] The system had gaps and errors, but it produced successive reports that identified what was known and what remained doubtful.

The wider Federal cavalry deployment was no seamless machine. Pleasonton’s divisions were spread across a broad arc, from Buford near Gettysburg to Kilpatrick near Hanover and Gregg farther south and east. Buford could not watch every Confederate road, and his division’s lack of forage reduced its endurance. Even so, the division nearest the likely point of contact had a working connection to Reynolds. That local relationship, rather than a perfect corps-wide intelligence system, gave the Federal advance its advantage on 30 June.^[17-1]

Buford then used terrain to gain time. Gettysburg itself was not the best defensive position; the commanding ground lay south of town on Cemetery Hill and Cemetery Ridge. The ridges west of town could, however, delay an advance from Cashtown. During the night Buford’s outposts covered the western and northern approaches. This screen did not seal off the town. White’s patrols detected Federal activity, and Pettigrew saw Buford’s cavalry. It did keep Confederate scouts from defining the infantry support close behind Buford. It also gave the Federal commander warning when Heth advanced the next morning.^[17-2]

Lee’s knowledge that evening was much thinner. He knew the Union army was east of South Mountain and that some Federal force occupied Gettysburg. He could reasonably expect a major engagement somewhere along the Chambersburg–Gettysburg axis. He did not know with confidence that Buford’s division was the screen of Reynolds’s wing, that the First Corps was within supporting distance, or how quickly additional Federal corps could converge. Nor did he have a cavalry force on Hill’s front to answer those questions before infantry advanced. He possessed strategic orientation, but the tactical space immediately ahead belonged to Buford.

1 July: cavalry turns knowledge into time

Shortly after dawn on 1 July, Heth sent two brigades east from Cashtown. The advance is sometimes explained as a search for shoes, but the more important fact is that Heth and Hill knew Federal troops had been seen near Gettysburg and still pushed a substantial reconnaissance down the pike. Lee had instructed his corps commanders to avoid a general engagement until the army concentrated. That order did not forbid reconnaissance or every skirmish, but it required subordinates to recognize when a probe was meeting something larger. Without cavalry ahead, infantrymen had to discover the answer by marching into Buford’s outposts.^[19]

The first shots drove in Federal videttes west of town. Buford’s troopers fell back through their warning line and formed mainly dismounted on McPherson Ridge. Their carbines gave them useful firepower, but the popular image of repeating rifles mowing down Confederates is wrong; most of the division carried single-shot breech-loading Sharps carbines. Their decisive advantages were preparation, dispersion, and the ability to trade ground without losing cohesion. Small groups had already watched the approaches. Brigade and regimental reserves knew where to form. Horses were held to the rear while skirmish lines used fences, woods, and ridges. The division could resist without pretending to be infantry strong enough to win the field alone.^[20]

Buford’s stand was a covering action, not an independent attempt to hold Gettysburg. His division numbered roughly 2,700 men and faced the leading brigades of a Confederate infantry division. Its task was to slow that advance and keep its commander informed until Reynolds arrived. The roughly two hours won west of town were therefore valuable because they brought Federal infantry onto the field before Heth could seize the approaches without opposition. Once the First Corps deployed, Buford’s cavalry continued to serve on the flanks rather than claiming the ground by itself.^[20-1]

As Heth’s brigades pressed forward, Buford’s communications continued to work. At 10:10 a.m. he reported that A. P. Hill’s force was advancing and driving his pickets rapidly. A force from Heidlersburg was also pressing his northern posts. Reynolds’s leading division was within three miles. The message was not perfect—Ewell’s infantry had not yet arrived on the northern field—but it accurately conveyed the essentials. The western contact was a major advance, and the northern flank was threatened. Federal infantry support was close.^[21] Cavalry reconnaissance had become battle management.

Reynolds reached Gettysburg while Buford was still holding west of town. Because the two commanders had met the day before and remained in communication, Reynolds did not arrive at an unknown crisis. He understood the direction of the Confederate advance, the ground Buford intended to hold, and the urgency of bringing the First Corps forward. The cavalry’s resistance bought the hours necessary for the leading Federal infantry to deploy on and north of the Chambersburg road. Reynolds’s death soon after arrival disrupted Union command, but not the flow of troops toward the field. Oliver Otis Howard’s Eleventh Corps followed, and Meade’s headquarters increasingly understood that the developing action required reinforcement.

The military effect is sometimes described too simply as “Buford chose the battlefield.” No cavalry commander could compel two armies to fight where he wished. Buford did something more precise. He recognized the operational value of the road junction and the defensive value of the heights south of it. He established contact with the enemy far enough west to protect those heights from sudden seizure, and he maintained contact with Reynolds long enough for infantry to convert a cavalry covering action into a corps battle. His division shaped the timing and initial geometry of the engagement. It denied Heth an unobserved approach and denied Lee the chance to learn what lay near Gettysburg without paying for the information in combat.

On the Confederate side, the battle expanded faster than Lee’s information could mature. Heth’s first attack met Federal cavalry, then First Corps infantry. Pender’s division entered after Heth. Reports and gunfire drew Lee and Hill toward the field. From the north, Ewell’s carefully chosen routes proved their value: when he learned that Hill was moving on Gettysburg, Rodes and Early could turn toward the sound without a major countermarch. Their arrival helped break the Federal First and Eleventh Corps and drive them through town. The Confederate concentration plan was not a failure. Within the forces available east of South Mountain, it produced converging attacks and a clear tactical success.

But the concentration was incomplete. Road congestion and the single main passage through South Mountain delayed the divisions west of the ridge. Anderson arrived too late to affect the decisive phases of the first day; Longstreet’s divisions were farther back; Pickett remained behind until Imboden could assume security duties. A cavalry screen could not have widened Cashtown Pass, but earlier and more precise warning might have changed marching priorities, moved security tasks to mounted troops sooner, or allowed Lee to decide whether to reinforce or avoid the Gettysburg contact before the battle drew in two corps. The absence of warning compressed choices that already depended on the road network.

Nor did victory on the afternoon of 1 July erase the information problem. Confederate forces had driven the Federals from the ridges north and west of town, but they had not destroyed them. A mounted screen close to army headquarters might have reconnoitered the Federal withdrawal more rapidly and watched the roads south and east. It could have secured exposed flanks and helped determine how much of Meade’s army was arriving behind Cemetery Hill. Jenkins’s troopers reached the area with Ewell but were tired and unevenly handled; Stuart’s three veteran brigades were still absent. Pursuit and post-battle reconnaissance—two standard cavalry tasks—were therefore weaker at the moment Lee had to decide what the first day’s success meant.

Stuart, meanwhile, was searching for the army through a landscape whose information was already stale. After Hanover, he moved toward Dover, learned that Early had gone west, and sent a staff officer to find him. Stuart then continued toward Carlisle, expecting to find supplies and Confederate troops, only to discover Union militia in possession. The detour brought no compensating operational result. On the night of 1 July, contact was finally restored when Stuart learned that Lee was fighting at Gettysburg. His exhausted column did not join the army until the afternoon of 2 July. The most experienced part of Lee’s cavalry had spent the opening battle collecting prisoners, wagons, and damage in the Federal rear, while the task of locating and screening the Army of the Potomac remained undone near the army it was supposed to serve.^[22]

What Stuart’s absence did—and did not—mean

The fairest assessment begins by rejecting two extremes. The first is that Stuart’s absence left Lee completely blind. Harrison identified the Federal advance, while staff scouts and civilians supplied other reports. Pettigrew found Buford, and White’s cavalry detected Federal activity. Lee’s own map and experience showed the likely roads. By 30 June he understood that battle was near. The second extreme is that Stuart’s absence was inconsequential because Robertson, Jones, Jenkins, and Imboden still existed. Mounted strength counted only when placed where it could perform the necessary mission. Those brigades were delayed, misallocated, poorly coordinated, or committed to other tasks. Lee had cavalrymen but no effective army-level cavalry screen on his eastern front.

The practical loss can be stated in five parts. First, reconnaissance lost continuity. Lee received snapshots rather than a moving picture of the Federal advance. Second, counter-reconnaissance weakened. Buford’s patrols collected unit identifications and watched Confederate approaches, while Confederate patrols could not see through his screen. Third, route security consumed infantry and contributed to the separation of Pickett and Law from the main concentration. Fourth, communications between distant Confederate columns became episodic. Stuart could not find Ewell, and Robertson did not close on Lee. Reports also failed between field commands and Richmond. Fifth, the Confederates lacked a strong covering force able to probe, delay, or disengage without committing infantry. Heth’s brigades had to find the enemy by collision.^[23]

Stuart’s presence near Lee would probably have improved several of these functions, but the counterfactual has limits. A cavalry screen could have identified Buford and Reynolds earlier; it could not ensure that Lee chose to avoid Gettysburg. Lee was concentrating on the enemy and expected a battle. Better reports might equally have encouraged him to mass more rapidly for a deliberate attack. Nor could Stuart make Confederate orders unambiguous, eliminate road congestion, or guarantee correct estimates of Federal strength. Cavalry reduced uncertainty; it did not abolish it.

The responsibility question is similarly conditional. Stuart was authorized to ride around the Federal army and was genuinely delayed by enemy movements. He also made choices that prolonged his separation. He persisted after Hancock blocked the intended route, kept the captured train, and continued toward Carlisle after learning that Early had moved west. Lee created the command arrangement and gave mixed incentives to raid and rejoin. He accepted the risk that Stuart would be incommunicado, then failed to employ the remaining cavalry aggressively. Robertson disregarded the operational trigger in his instructions. His inactivity compounded the failure, but it did not create it. Lee had already accepted an arrangement that depended on reports from distant and disconnected elements. Once Robertson remained in the gaps, it had no redundancy. The War Department did not forward Stuart’s 27 June intelligence. Federal cavalry fought to keep him away. No single failure explains the result; together they broke the mounted network on which Lee had become accustomed to rely.^[24]

That network perspective also explains why Buford deserves attention without turning him into a prophet. He did not know Lee’s complete plan or foresee three days of battle. He made errors and repeated rumors. He worried about forage and asked for a route of retreat. His strength was procedural. He patrolled several roads and identified prisoners. He kept higher headquarters informed, maintained liaison with Reynolds, and established depth in his outposts. He also selected ground suitable for a delaying action. When his first estimate changed, his system produced another report. Cavalry succeeded not because its commander guessed once and guessed right, but because it kept learning while events moved.

Lee’s system did the opposite. It produced important facts but did not sustain them. Stuart learned that Hooker had moved toward Leesburg, but Lee never received the message. Harrison found the army, but he could not maintain observation. Pettigrew found troops, but could not identify or turn their screen. Ewell’s small patrols found signs of movement, but their reports could not give army headquarters a full picture. The distinction between “knew” and “did not know” is therefore too blunt. Lee knew enough to concentrate and anticipate battle. He did not know enough, soon enough, to regulate how the battle began.

Conclusion

Gettysburg was not caused by an empty saddle. It developed from two armies concentrating across a constrained road network, the decisions of several commanders, and active enemy resistance. Stuart’s ride was part of that process, not an all-purpose explanation for Confederate defeat. The better question is what knowledge and control an army expected its cavalry to provide.

From 28 June to 1 July, Lee’s information improved in scale but not in resolution. Harrison told him that the Army of the Potomac had crossed and was approaching South Mountain. The report justified a well-conceived concentration through Cashtown and Heidlersburg. Scouts later put Federals near Gettysburg. Lee never received a continuous account of the leading corps, their roads, and their marching times. He also lacked a clear sense of their supporting distances. That missing operational picture left infantry to perform reconnaissance and security, held other infantry back, and allowed the Gettysburg contact to escalate faster than Lee could control it.

This does not lessen Harrison’s contribution. His warning allowed Lee to change the army’s strategic posture in time. It explains why Lee could respond sensibly to the campaign’s broad direction while still learning the immediate tactical situation through infantry collision.

Buford showed what cavalry could do when its functions remained connected. His patrols collected information; his screen denied it; his couriers and liaison linked cavalry to infantry; his videttes gave warning; and his dismounted defense converted warning into time. The effect was not to predetermine the battle’s outcome but to ensure that the Army of the Potomac entered the fight with a clearer understanding of the immediate field than its opponent.

The contrast makes clear what Stuart’s absence costs. Lee was not blind. He was looking through reports that arrived late, stopped at the enemy’s screen, or could not be verified. On the road to Gettysburg, cavalry’s decisive work was less the charge than the maintenance of an army’s awareness. It let a commander trace an enemy’s approach and his own army’s routes. It also kept friendly columns in contact before a movement became a battle.

Those same problems of reconnaissance, screening, road security, and uncertain contact are central to Campaign Gettysburg. Across more than 620 scenarios and variants, players are repeatedly placed in situations where the enemy is not necessarily where historical hindsight says he should be. Forces may approach along different roads, arrive on altered schedules, or concentrate in unexpected places. Under those conditions, cavalry becomes essential not simply as a combat arm, but as the means of finding the enemy, screening friendly movements, securing routes, and providing the warning needed before a march becomes a battle. The game’s large campaign maps therefore recreate something close to the operational problem faced in June 1863: acting on incomplete information while trying to discover what lies beyond the next ridge or road junction.

Bibliography

Coddington, Edwin B. The Gettysburg Campaign: A Study in Command. New York: Touchstone, 1997.

Gottfried, Bradley M. The Maps of the Cavalry at Gettysburg. El Dorado Hills, CA: Savas Beatie, 2020.

Lee, Robert E. The Wartime Papers of R. E. Lee. Edited by Clifford Dowdey and Louis H. Manarin. Boston: Little, Brown, 1961.

Longacre, Edward G. The Cavalry at Gettysburg: A Tactical Study of Mounted Operations During the Civil War’s Pivotal Campaign, 9 June–14 July 1863. Lincoln: University of Nebraska Press, 1986.

Martin, David G. Gettysburg, July 1. Rev. ed. Conshohocken, PA: Combined Books, 1996.

Mosby, John S. Stuart’s Cavalry in the Gettysburg Campaign. New York: Moffat, Yard, 1908.

Starr, Stephen Z. The Union Cavalry in the Civil War. Vol. 1, From Fort Sumter to Gettysburg. Baton Rouge: Louisiana State University Press, 1981.

Thomas, Emory M. Bold Dragoon: The Life of J. E. B. Stuart. Norman: University of Oklahoma Press, 1999.

U.S. War Department. The War of the Rebellion: A Compilation of the Official Records of the Union and Confederate Armies. Series 1, vol. 27, pt. 1. Washington, DC: Government Printing Office, 1889.

Wert, Jeffry D. Cavalryman of the Lost Cause: A Biography of J. E. B. Stuart. New York: Simon & Schuster, 2008.

Wittenberg, Eric J. “The Devil’s to Pay”: John Buford at Gettysburg: A History and Walking Tour. El Dorado Hills, CA: Savas Beatie, 2014.

Wittenberg, Eric J., and J. David Petruzzi. Plenty of Blame to Go Around: Jeb Stuart’s Controversial Ride to Gettysburg. Havertown, PA: Savas Beatie, 2006.

Footnotes