SEPTEMBER 2–9, 1862
IN THE HISTORY BOOKS THE SECOND BATTLE OF BULL RUN USUALLY figures as the culmination of an offensive that carried the seat of war from Richmond to northern Virginia. For Lee himself the successful campaign against Pope was merely a phase in a larger strategic offensive. The last elements of Pope’s rear guard withdrew to the Washington defenses on September 2, and the decision to invade Maryland was made that evening, with the initial movements ordered on September 3. The Army of Northern Virginia made a leisurely march northward to Leesburg, Virginia, where Lee would reorganize his forces for the invasion of Maryland. Thus Lee “opened the chapter on his next campaign less than six hours after closing the old one.”1 He did so while suffering the painful effects of injuries to his arms, suffered when his horse spooked and threw him to the ground. Both arms were in splints, so he could not use the reins to control his horse, and would have to travel in an ambulance for nearly two weeks.
The defeat of Pope’s army left the initiative in his hands, but his options for exploiting the opportunity were limited. He could stand on the defensive, allowing his army to recover its strength and supplies and gathering reinforcements. However, if he did so he would have to pull his army back to the Rappahannock River, because the region just south of the Potomac had been stripped of food and fodder by the two armies and was exposed to forays by the superior Union forces accumulating in the DC area. Such a retreat would sacrifice much of the territory won by his campaign, and diminish the political impact of his victory at home, in the North, and abroad. In a few weeks the Union armies would have recovered their morale, repaired their organization, and replenished their ranks, and would be ready to begin another offensive with superiority in men, guns, and resources. To forestall such an offensive, he had to keep the Federals on the defensive, maintaining the initiative by offensive operations of his own. He lacked the troop strength and heavy artillery to attack Washington directly, and his army was too small to surround and besiege the city. The only alternative was to threaten some vital point beyond Washington and compel the Federal army to come out and defend it. That left him with the choice of marching away from Washington, to drive the Federals out of their posts in the northern Shenandoah Valley, or crossing the Potomac into central or western Maryland.2
The speed with which Lee made such a critical decision suggests that he had anticipated the situation he now faced. On September 2 he marched the main body of his army to Leesburg, in a gap of the Bull Run Mountains some twenty miles northwest of Chantilly. From Leesburg his army could either move against Federal forces in the northern Shenandoah Valley and Harpers Ferry or ford the Potomac well upstream from Washington and invade Maryland. By the evening of September 3 he had decided on the latter operation, and wrote to inform President Davis and briefly explain his reasoning. He began by stating his belief that “The present seems the most propitious time, since the commencement of the war, for the Confederate Army to enter Maryland.” The crucial point was this: “[We] cannot afford to be idle, and though weaker than our opponents in men and military equipments, must endeavor to harass, if we cannot destroy them.” However, he did not ask, nor did Davis offer, explicit approval for his operations; and that pattern would persist throughout the campaign.3
Some historians have seen Lee as having presented Davis with accomplished facts, in order to preempt anticipated objections. Lee certainly wanted to determine the pace and course of operations without presidential interference. When Davis declared his intention to visit the army in the field, Lee discouraged him. However, there is nothing in the written record, or in Davis’s behavior after August 21, to suggest that Davis was reluctant to grant Lee authority to determine the objectives and pace of military operations. Despite Davis’s earlier concern for the defenses of Richmond, when Lee asked reinforcement for his strike against Pope, Davis sent almost double the number asked for, and stripped the Richmond garrison of its best troops in order to do so. His response to the demands of Lee’s plan of invasion would also be proactive rather than reluctant. It makes more sense to see Lee’s swift and autonomous decision making as a reflection of the mutual trust between Davis and Lee and the closely shared understanding of strategic imperatives they had developed during their long and close association while Lee was Davis’s chief military adviser.
The harmony of views between Davis and Lee was consonant with the public mood, as reflected in Southern newspapers. Calls for an end to defensive warfare and for retaliation on the Northern invaders had been accelerating since May. Lee’s success against Pope, and early bulletins from Bragg’s invasion of Tennessee, transformed the call into a demand. The Richmond Whig condemned defensive warfare as a “humbug” and demanded that Lee “concentrate a force sufficient to strike them to the heart, and deliver the whole continent from bondage.” The Charleston Mercury demanded that “Our victorious troops in Virginia, reduced though they be in numbers, must be led promptly into Maryland, before the enemy can rally,” to win a victory and raze Washington to the ground. The Jackson Mississippian mirrored Davis’s own analysis by astutely noting that the upcoming elections in the North offered an opportunity for Southern armies to overthrow Republican rule. Kentucky was ripe for liberation, and “Maryland would rise with a wild shout of exhultation when they hear the clarion notes of restored freedom and see ‘the gathering of the clans’ in their sister states. . . . Let us then . . . conquer a peace before the year ends. Forward, Forward, FORWARD.”4
In his July 18 letter to Forsyth, Davis had projected an invasion of Northern territory as the logical and necessary culmination of a war-winning strategy. Only by carrying the war to the enemy and defeating him on his home ground could the Confederacy break the political will that sustained the Unionist war effort, and do so before the strain of war inflicted critical or even fatal structural damage on Southern society. Invasion might succeed by simply inflicting punitive damage on Northern farms and factories, comparable to that suffered by the invaded South. Yet Davis also saw invasion as the means by which the Border States of Maryland and Kentucky might be detached from the Unionist cause—a shift that would profoundly alter the balance of forces and resources. Nevertheless, before a successful invasion could be mounted, Southern armies would have to inflict crushing defeats on the “detachments” (that is, field armies) that had invaded Southern territory and seized the critical positions of Corinth, Nashville, and the Virginia Peninsula.5
It was with that strategic design in mind that Davis and Lee had approved Bragg and Smith’s proposal to “crush” Buell’s army in Tennessee and then invade Kentucky. Davis’s instructions to Bragg showed that he not only accepted the risk of a general engagement, he positively urged Bragg to make the defeat of Buell the precondition for his invasion. When the withdrawal of McClellan’s troops from the Peninsula made a similar strike possible in Virginia, Davis accepted Lee’s plan to “suppress” Pope’s detachment. It was impossible to know in advance what kind of opportunities might open up as a result of Pope’s defeat, but Davis and Lee had agreed going in that it would be desirable to exploit any victory by continuing operations in northern Virginia. It was necessary to keep Federal forces on the defensive to forestall a new advance on Richmond, and to prevent the sending of reinforcements to the western theater by keeping the enemy occupied in Virginia. Lee’s plan to continue the offensive by an invasion of Maryland was perfectly consistent with these objectives, and with the strategy Davis had already urged on Bragg. Davis trusted Lee’s judgment that invasion was necessary and success feasible, and left to his general the development of an operational plan. He also backed that plan without reservation, forwarding all the reinforcements he could muster.
STRATEGIC AIMS OF THE INVASION
In the campaign that followed, Davis would have to play the role of strategic commander without the benefit of Lee’s advice. Close consultation was simply impossible with Lee commanding a field army more than a hundred miles from the capital. Although Lee was certainly aware that his offensive was strategically connected to Bragg’s in Tennessee and Kentucky, only Davis was in a position to see each operation in relation to the other. His orders to Bragg and Lee were nearly simultaneous. He received and approved Lee’s preliminary invasion plan on September 3 and gave his approval to Bragg’s plans on September 4. However, once these operations were launched, Davis could do nothing to control or coordinate the actions of his field commanders. It was more by coincidence than by order that Bragg moved out from Chattanooga on September 5, the same day that Lee crossed into Maryland.6
What Davis could do was formulate the political objectives of the two invasions and develop means for achieving those objectives. The most critical was to attack the political fault lines in the Unionist camp. Davis and Lee believed that the Unionist sentiment of the Border States was weak, and balanced by strong cultural affinities and economic interests with the slave states. The Confederate Congress had made the accession of Maryland in particular a war aim, and Davis had instructed his diplomats in England and France to insist that any mediated peace settlement allow Marylanders to decide by plebiscite whether they still wished to stay in the Union. The appearance on their soil of a victorious Confederate army might produce a political overturn that would carry these states into the Confederacy or, failing that, at least produce an uprising of Southern sympathizers behind Yankee lines.
Lee and Davis also hoped to take advantage of the fact that midterm congressional and gubernatorial elections would be held in two months. A coherent opposition to the Lincoln regime was finally emerging, uniting conservatives of various schools against the increasing “radicalism” of the Republicans. It was perhaps premature to characterize this movement as a peace party, but its leaders certainly favored a settlement that accommodated the slavery interest, and compromises like those that had been offered in 1861 by Senators John Crittenden and Stephen Douglas, which conceded to the South a form of political autonomy within the Union. A successful invasion of Northern soil, demonstrating the Lincoln government’s inability to protect its own territory, let alone conquer the South, coupled with an appeal to conservatives, might produce a powerful revulsion at the polls that would substantially weaken political support for the war effort.
Davis also hoped that a Confederate victory on Northern soil would encourage Britain and France to recognize Confederate independence, and intervene on its behalf with financial and military aid and an offer—backed by force—of mediation. Lee may have shared the hope, but did not believe the Europeans would intervene until the Confederacy had made victory certain using its own unaided power. Davis’s hope was actually closer to fulfillment than he realized: the British government was going to begin actively considering intervention shortly after Lee’s army crossed the Potomac. There had been a brief flurry of interventionist activity in Britain in December 1861, after a Union naval vessel had stopped a British ship on the high seas and seized the Confederate emissaries Mason and Slidell—which ended when Lincoln repudiated the seizure and released the two men. A second wave of interventionist sentiment had risen in Parliament in mid-July, following a false report that the Seven Days Battles had ended with the surrender of McClellan’s army—thwarted when more accurate reports were received. But news of Lee’s triumph at Second Bull Run, following so closely on the heels of McClellan’s defeat, had convinced the British cabinet that the proper circumstances for intervention might be developing.
The politics of intervention were complicated. There was a general conviction among leaders in the British and French governments that the attempt to restore the Union by force was doomed to ultimate failure, and that the Confederacy was already qualified as a nation-state by all the canons of international law and custom. That belief was shared even by those (like Chancellor of the Exchequer William Gladstone and French foreign minister Edouard Thouvenel) who were sympathetic to the Union cause. Both countries were also suffering economic hardship because of the Federal blockade, which cut them off from the supply of American cotton on which their industries depended. The breakup of the Union might benefit the British by weakening a potential maritime Great Power rival; it would certainly aid Napoleon III’s project of turning Mexico into a French client state. Both powers therefore looked forward to the moment when, by diplomatic intervention, they might bring the war to an end and recover their access to Southern ports. However, both powers had to be concerned about the European balance of power, and the threat of war in their own regions. France was heavily engaged in the Italian struggle for national unification, which had already embroiled it in one war with Austria and might well require another. France and Britain were also concerned about the ambitions of Prussia, which under Bismarck was emerging as a power to rival France on the Continent, and was preparing war against Denmark—a move threatening Britain’s access to the Baltic.
These varied considerations put two constraints on those interested in intervention: neither Britain nor France could risk acting alone; and neither could afford a war against the American North, which would draw naval and military forces away from Europe (and in the French case, Mexico). The French would wait on a British commitment to act. The British ministry would watch for clear signs that the North was willing to accept mediation without going to war. It was possible that a series of decisive victories by Confederate armies would make Southern independence a fait accompli, which could be safely recognized. However, a more likely and certain sign would be an electoral victory for anti-administration forces in the November midterm elections. Mason and Slidell made their pitch for early recognition on the ground that that would encourage Northern opposition to the war and produce the electoral victory the British were waiting for. The Europeans preferred that recognition wait on the election. From either perspective, it was clear that the most critical element of Lee’s proposed campaign was its potential effect on the fall elections.7
Lee’s operational plan was shaped as much by political as by purely military considerations, and his ideas on the subject matched Davis’s point for point. He knew that the redemption of Maryland had been a Confederate war aim from the start of the conflict, and that Davis was particularly eager to attempt it. In his September 3 dispatch Lee advised Davis, “If it is ever desired to give material aid to Maryland, and afford her an opportunity to throw off the oppression to which she is now subject, this would seem the most favorable.” To that end, Lee requested (September 4 and 7) that E. Louis Lowe, a former governor of Maryland, accompany his expedition, to advise him on the political questions that might arise, and perhaps to develop contacts with potential allies. In fact, Davis had already decided to send Lowe to the front—another sign of their common understanding, and perhaps prior agreement, on strategic means and ends.8
Lee continued to think through the political potential of his invasion, and one of his first acts on entering Maryland would be to formulate a systematic program of political action to augment the effect of his operations. He would set forth his ideas in a dispatch to Davis sent on September 8 from Frederick, the initial point of concentration for Lee’s invading force. Although the general had never before infringed on President Davis’s area of responsibility, this dispatch was exclusively political, and offered detailed and assertive advice on the political dimension of their strategy. He believed the military advantage he now enjoyed would allow Davis to negotiate from a position of strength a final political settlement of the conflict. “The present position of affairs, in my opinion, places it in the power of the Government of the Confederate States to propose with propriety to that of the United States the recognition of our independence.” Such a proposal, made while Lee’s army was operating on Northern soil, “could in no way be regarded as suing for peace but, being made when it is in our power to inflict injury upon our adversary, would show conclusively to the world that our sole object is the establishment of our independence and the attainment of an honorable peace.” If the Lincoln administration rejected such a proposal, it would prove to the country “that the responsibility of the continuance of the war does not rest upon us, but that the party in power in the United States elect to prosecute it for purposes of their own.” The events of the previous six months proved that the attempt to resist secession had led to devastating losses, “without advancing the objects which our enemies proposed to themselves in the beginning of the contest.” Lee specifically took notice of the upcoming midterm elections. If at that time there was a large Confederate army firmly planted on Union soil, threatening the capital and raiding Pennsylvania, there might be a revulsion of feeling against Lincoln and the war party. In such a context, “The proposal of peace would enable the people of the United States to determine at their coming elections whether they will support those who favor a prolongation of the war, or those who wish to bring it to a termination, which can but be productive of good to both parties without affecting the honor of either.”9
That same day, Lee would issue a proclamation, calling on Marylanders to rise and cast off the tyranny that had been imposed on them by Yankee force. His language was politically astute, striking a nice balance between the conservative adherence to states’-rights’ legalism and the essentially revolutionary call for an overthrow of the existing state government. He listed their grievances for them—their capital occupied “by armed strangers,” newspapers suppressed, citizens subjected to arbitrary arrest and imprisonment. “Our army has come among you, and is prepared to assist you with the power of its arms in regaining the rights of which you have been despoiled.” However, if Marylanders declined this offer, the Confederate army would respect their exercise of states’ rights: “while the Southern people will rejoice to welcome you to your natural position among them, they will only welcome you when you come of your own free will.” While Lee did not believe that a “general rising” in the state was likely, he did believe his proclamation might rouse the Southern sympathies of individual Marylanders, and persuade them to either join the Confederate army on its march or supply its desperate wants of food and clothing. Such a movement would be a serious blow to the material strength of the Yankees and would weaken their moral hold on Maryland and other Border States.10
Once again Lee’s proposals anticipated the thoughts and actions of President Davis. On September 12, Lee would receive a dispatch from Davis indicating his intent to issue a similar proclamation; their messages had crossed in the mail, a not uncommon occurrence. Lee would apologize for seeming to preempt his president’s actions, but the embarrassment was slight. Davis’s own proclamation indicated that he had been thinking along the same lines, but it was broader in scope than Lee’s—more like a political manifesto—and was addressed not only to Lee for use in Maryland but to Bragg and Smith for use in Kentucky.
Davis began by explaining and justifying the Confederacy’s reasons for invading the North. The act was neither malicious nor vengeful but a regrettable necessity forced upon a reluctant Confederacy by Federal aggression. He reminded Northerners that the Confederacy fought only in self-defense, not for territorial gain; and that his government was willing to guarantee by treaty the free navigation of the Mississippi. These were issues of particular concern to the states of the Old Northwest, and by including these assurances Davis was suggesting the terms of a potential peace settlement. Davis then appealed to Northern citizens in terms that echoed Lee’s September 8 dispatch. He hoped they would prevail on their government to respond to the peace overtures Davis’s government had already made.
However, instead of Lee’s suggested offer of new peace negotiations, Davis issued a challenge with revolutionary implications. The laws of war entitled Southern armies to respond in kind to the depredations committed by the North’s invading forces, and to punish “those who persist in their refusal to make peace.” This was, in effect, the mirror image of Pope’s decrees, which had also threatened punishment for Southerners who refused to swear allegiance to the Union. To escape Confederate retribution, the Northern people would have to convince their government to end the war, presumably by using the ordinary civil means of petition and election. However, should such efforts fail, Davis invited the people of individual states to “secure immunity” by making “a separate treaty of peace which this Government will ever be ready to conclude on the most just and liberal basis.” This last appeal was implicitly a call for revolution in the North. The treaty-making power belongs exclusively to the Federal government. By inviting individual states to sign treaties of peace, Davis was proposing an assertion of state sovereignty not far removed from actual secession; and he threatened those states that refused to rebel with a ravaging akin to that which the South had been suffering.11
This was Davis’s finest moment as a strategist. He had heeded the suggestions for offensive action suggested by Bragg and Lee, and to the extent possible coordinated their operations. With his generals he had developed what Lincoln had been trying in vain to achieve: simultaneous offensives by the three main field armies spanning the breadth of the Confederacy. Moreover, he had linked these invasions to a coherent political program, which attacked the structures of state power that constituted the Union. The result was a development of Confederate strategy that mirrored the evolution of Union strategy, toward intensified and more punitive military operations and a more radical assault on the enemy’s social and political order. Driven by the demands of a war that put at risk the survival of their nations, both presidents were doing things that would make reconciliation more difficult and make the Civil War a “total” war, a “remorseless revolutionary conflict.”
ACROSS THE ATLANTIC, political developments also appeared to favor the Confederacy, though Davis had no way of knowing this. Because of the blockade it could take months for communications to pass between Davis and his European emissaries, and even without the blockade there was a lag of nearly two weeks between events in America and the receipt of the news in London. England got the news of Pope’s defeat and McClellan’s reappointment on September 10, five days after Lee’s army had crossed the Potomac into Maryland. The British cabinet had been deeply impressed by Lee’s victories, especially Second Bull Run, and were aware that the Confederates intended to press their advantage. Anticipating further Confederate successes and a corresponding loss of Northern morale, Foreign Minister Lord John Russell and Prime Minister Lord Palmerston believed it was time to begin preparing for the inevitable moment of intervention. Russell opened the conversation on September 11, and Palmerston stated the policy options openly in a letter to Russell on September 14. If the Confederates were able to follow Second Bull Run with a victory on Northern soil it might be time for England and France to “address the contending parties and recommend an arrangement upon the basis of separation.”12
LEE’S OPERATIONAL PLAN
The overall purpose of Lee’s plan was, as he told Davis, to “harass, if not destroy” the Federal field army defending Washington. Obviously, the destruction of that army was the most desirable outcome, but Lee was not overinvested in the idea of a Napoleonic battle of annihilation. The plan he developed was flexible and offered several different ways of achieving the desired political objectives.
Lee’s offensive options were limited by the disparity of strength between the armies. His Army of Northern Virginia was outnumbered and outgunned by the Federal forces in Washington, which may have numbered 120,000 all told. However, the only fresh veteran units available to the Federals were the II and VI Corps from McClellan’s army, the XII Corps of Pope’s command, and two divisions of Burnside’s IX Corps that had missed Second Bull Run—about 55,000 troops. They could be augmented by elements of Pope’s army, and by the newly recruited regiments that had arrived in the capital, but these units were of doubtful value. The Federal corps that had fought under Pope had suffered debilitating losses of men and matériel, and Lee believed they were seriously “weakened and demoralized.” The newly recruited regiments were untrained and likely to prove liabilities on the battlefield. Lee believed he could immediately field an army of between 65,000 and 72,000 men, quite large enough to deal with a Federal army of these ill-sorted elements. His own army would have a distinct advantage in both morale, as a result of its recent victories, and organization, since Lee’s reformed command structure had been tested and perfected in the last campaign. The events of the past month showed that a well-led, veteran, and high-morale force could defeat a larger Federal army deficient in those categories.13
While a direct assault on Washington was beyond Lee’s capabilities, Lee could threaten so many vital and politically sensitive areas by moving into Maryland that Federal commanders would be compelled to leave their fortifications and move against him before their forces were fully reorganized.
To provoke a premature Federal advance, Lee would first seize Frederick, Maryland, only fifty miles west of Washington. His reasons were political as well as military. The town was centrally located in the mainland portion of the state, convenient to Baltimore and its environs, where pro-Confederate sympathies were strongest. It was therefore the best position from which to instigate a popular uprising in the state, to attract recruits, and to obtain supplies. Because of its proximity to Washington, its occupation by the Army of Northern Virginia constituted a damaging demonstration of the Lincoln administration’s weakness. It was also a deliberate provocation to the Union military, a challenge to battle on their doorstep that might goad them into advancing before their army had recovered from its defeats. If the Federals accepted, Lee felt confident of winning a general engagement. If the North declined his gambit, Lee would shift his forces away to the west, behind the high, extended ridges that ran north from the Potomac into Pennsylvania. The passes through these elongated ridges were narrow and easily defended against a Federal advance. These ridges were also, in effect, northern extensions of those that bounded the Shenandoah Valley on the south side of the Potomac. A Confederate army positioned behind them would enjoy direct access to the Valley’s agricultural resources, and the Valley itself provided a mountain-sheltered route for military supplies carried north from Richmond via the Virginia Central Railroad and the Valley’s excellent road system.
Before he could use the Valley as a supply line, Lee had to drive away the large Federal garrison that held the town of Harpers Ferry at the northern outlet of the Valley. This outpost, and the associated positions at Winchester and Martinsburg, were sustained by lines of communication that ran eastward through the Maryland panhandle and southern Pennsylvania. It was an impediment that should be relatively easy to remove. When the Confederate army interposed between Harpers Ferry and Washington, the outpost would be isolated, and its garrison would have no choice but to withdraw or be destroyed. With Harpers Ferry in hand, and a protected supply line running through the Valley, Lee’s army would be able to maintain itself in western Maryland for an extended period of time. Basing itself in Hagerstown, it could harass the Federals by raids into Pennsylvania, compelling Lincoln to defend his capital and its hinterland instead of marshaling a new offensive against Richmond. If the Federal army moved to attack him west of the mountains, it would be acting at a greater distance from its own base in Washington, and in its attempts to strike across the mountain barrier would expose itself to defeat by the kind of rapid maneuvers at which the Army of Northern Virginia excelled. Under these circumstances Lee would have a good chance of inflicting a decisive defeat or even destroying the opposing army. Otherwise Lee could use Hagerstown as the base for a deeper invasion of the North, following the route he would later take in the Gettysburg campaign: north to Harrisburg to cut the Pennsylvania Railroad, the North’s main west-to-east railroad line, then south to threaten Baltimore and Washington from the rear, maneuvering so as to confuse and divide the pursuing Federals and defeat them in detail.14
Fortunately for Lee, while decisive battle was desirable, it was not essential to a successful campaign. By simply prolonging his presence in Maryland, he would dramatically demonstrate the Confederacy’s ability to stymie Federal efforts at invasion and subjugation, and do so at a moment when such a demonstration could have powerful and perhaps lasting political effects.
WHILE HIS ARMY TOOK its brief rest at Leesburg, Lee assessed its strengths and weaknesses and reorganized its elements for the campaign he had in mind. The commands of Jackson and Longstreet were roughly equivalent to army corps. This clarified the chain of command and, over time, would enable the component divisions to perfect their teamwork. Just after Second Bull Run, Lee received three new divisions, commanded by Generals D. H. Hill, Lafayette McLaws, and Richard Anderson, which had been sent from Richmond by President Davis in response to an earlier request. Lee chose not to assign these divisions to Jackson or Longstreet for the moment but to treat them as independent commands reporting to his own headquarters. He could use them as a general reserve or assign one or more of them to either Longstreet or Jackson as circumstances might require. Lee also ordered the movements of Stuart’s cavalry division.15
Although he was heavily outnumbered by Federal forces in the theater, Lee considered his army strong enough in numbers and in morale for the campaign he was planning. “The only two subjects that give me any uneasiness are my supplies of ammunition and subsistence.” The army lacked “much of the material of war, is feeble in transportation, the animals being much reduced, and the men poorly provided with clothes, and in thousands of instances, are destitute of shoes.” However, he had reason to think that the army could get supplies of flour, meat, and forage from the farms of western Maryland. Since he expected the Federals shortly to abandon the Shenandoah, he asked Davis to send ammunition, especially for the artillery, via Winchester. He also expressed concern about the army’s lack of discipline, especially the tendency of troops to straggle or otherwise absent themselves from their commands, and their tendency to pillage the country they passed through. In response, he instituted a beefed-up provost guard system to deal with stragglers, but its efforts would prove inadequate.16
Every unit was to reduce its transportation needs to the bare minimum, taking only enough livestock and wagons to haul basic subsistence and ammunition. All horses and mules freed by this process were to be turned over to the army quartermaster for redistribution. The artillery presented special problems. A powerful artillery was essential to success on the battlefield, and on paper the Army of Northern Virginia possessed an ample supply: some three hundred guns in seventy-four batteries. However, the quality of these units was extraordinarily uneven. The Confederacy had not been able to develop a large-scale manufacturing capacity, so its artillery was a mishmash of modern guns and antiques, warehouse stock that had moldered in Southern arsenals, captured Federal guns, and a few British pieces run through the blockade. Lee subjected his artillery to triage, cannibalizing equipment and redistributing the sound draft animals to bring his best batteries closer to the mark, at the cost of leaving other batteries behind. Even though this part of the plan was not fully carried out, in part because these volunteer units were political as well as military entities, and state pride forbade their surrendering equipment and the right to fight to units from other states, the net effect was to increase the efficiency and mobility of his artillery.17
The unit pride that prevented Lee from completely reorganizing his artillery was otherwise his army’s greatest strength. Although he faulted their lack of discipline in the camp and on the march, Lee rightly regarded his soldiers as superb combat troops. As combat veterans the soldiers knew what to expect on the battlefield and were unlikely to panic or break without good cause. Although conscription had been in effect for four months, these men were still volunteers. Whatever their reasons for enlisting—for adventure, to be part of a great community engagement, to vindicate Southern rights, to defend their homes from invasion—they now identified community, pride, and patriotism with their units. Those who would make the march were also, for the most part, men who had proved their devotion by sticking to the flag through grueling marches and terrible combat. More than 60 percent of the regiments in Lee’s army had already participated in three or more general engagements. More significantly, most of the brigades and divisions were composed of units that had served together through the Richmond and Second Bull Run campaigns. Their officers were familiar with each other and with each other’s troops, an invaluable kind of experience that made for efficient communication and liaison on the battlefield. The corps organizations were of more recent vintage, but most of Longstreet’s and Jackson’s divisions had made at least one campaign together.18
However, the actual troop strength of the Army of Northern Virginia at the start of the Maryland campaign is difficult to assess. Historians have estimated Lee’s initial force at anywhere from 50,000 to 75,500. The latter estimate is based on Confederate records, so it is possible that Lee had a similar estimate in mind. However, Confederate paperwork was notoriously incomplete and inaccurate. Moreover, Lee himself could not have had a clear or accurate knowledge of his army’s numerical strength, since he had not yet instituted the most basic of accounting systems, which requires unit commanders to report their strength up the chain of command. Nonetheless, he did know that he had a huge number of regiments under his command. With the accession of the divisions from Richmond, the Army of Northern Virginia contained 205 regiments—fully a third of the units then under arms in the Confederacy. If Lee supposed these to be at anything near half their authorized strength, he might have estimated his army’s strength as near the 75,500 figure.19
There are also reasons to doubt the high estimate of Lee’s force. The report Lee made after the Battle of Antietam asserted that he had fewer than 40,000 in that engagement, which occurred only two weeks after the start of the campaign. Only about 3,000 men were lost in the battles and skirmishes leading up to Antietam. If Lee’s initial force had been 75,500, then his army would have lost 32,000 men, or 43 percent of its strength, to nonbattle causes (mostly straggling and desertion) in less than two weeks’ time. With such a rate of loss, officers would have seen their commands melt away before their eyes; it suggests a degree of demoralization that no officer observed at the time. Yet most of the reports made after the campaign suggest that the extent of straggling was not generally realized until the battle was about to start, with some officers failing to note its effects till the army was back in Virginia. Lee himself would later remark that straggling had reduced by a third the force he was able to field at Antietam. If that estimate is correct, then his army probably mustered 65,000 to 70,000 at the start of the campaign.20
Whatever the actual numbers, the salient fact is that Lee began his campaign without knowing his army’s exact strength—and most probably with an exaggerated idea of its numbers. As the campaign went forward, it would also become apparent that he had overestimated his soldiers’ physical endurance and failed to properly anticipate the consequences of their lack of food, clothing, and especially shoes.21
Although the troops were not informed of the army’s plan of operations, they were aware that an invasion of enemy territory was in the works. For most of the army’s veterans that news would have been a boost to morale. Some units were thought to have objected on ideological grounds to fighting for any purpose but homeland defense, but this does not appear to have affected any significant number of regiments. However, there was a good deal of reluctance felt by troops who had been marching and fighting almost without letup for the past month. The vast majority of these stayed with their units, but many would fall out by the way because of physical and mental exhaustion.22
THE INVASION BEGINS, SEPTEMBER 4–10, 1862
Lee’s initial moves were provocative. On September 4 General D. H. Hill crossed the Potomac at Point of Rocks, to break the B&O Railroad line, disrupt Federal communications between Washington and Harpers Ferry, and make a reconnaissance in force toward Frederick. If the implied threat of Hill’s movement did not force the abandonment of Harpers Ferry, the concentration of Lee’s whole army north of the Potomac would do the trick.23 On September 5 Jackson’s command crossed the Potomac at White’s Ford, only twenty miles from Washington, while Stuart’s cavalry crossed lower down and aggressively probed Federal positions. On the sixth Jackson occupied Frederick and Longstreet’s Corps forded the Potomac and marched to join him.
The current was swift where Jackson crossed, and in one Georgia Regiment “to keep from washing down four men would get in line up and down the river and hold on to each other for support.” This “kind of dammed up the water” behind their line, and guaranteed them all a good soaking. They savored the cooling effect on what turned into a hot, dry, dusty daylong march.24
Despite eighty-degree heat, the march up to Frederick was otherwise a pleasant one, and the troops were in high spirits, anticipating better and more plentiful food supplies to be gathered from Maryland’s rich farms and happy to be hitting the Yankees where it hurt. Frederick itself was a prosperous and pretty market town of redbrick and white clapboard houses set among trees and broad fields.
By holding Frederick, Lee put his army at the center of the region’s road network and cut the B&O Railroad at Monocacy Junction south of town. From Frederick he could then threaten movements against the northwestern defenses of Washington, raid into Pennsylvania, or bluff an advance on Baltimore, whose pro-secession population might be ripe for an uprising. With the belligerent and impulsive Pope in command, Lee could hope that the Federals would move to counter his offensive before their battered army was fully prepared, allowing Lee to fight them on his own chosen ground. If that failed he would also have an open route for retreat across the mountains to Hagerstown, where he hoped to establish a long-term presence.
The prosperous civilians of Frederick thought the Rebel army were “the dirtiest men I ever saw . . . a most ragged, lean, and hungry set of wolves.” One observer compared them to a barbarian horde, a resemblance reinforced by the presence of a large number of Negroes, serving as teamsters or ambulance aides or officers’ servants, dressed in rags but often armed with “rifles, muskets, sabers, bowie-knives or dirks.” The vast majority of soldiers were clothed in homespun garments, made of various combinations of cotton, linen, and/or wool, haphazardly dyed in shades of gray, or the varied browns and ochers produced by the use of butternuts, a variety of white walnut, in the dye-stuff. Some wore the natty-looking French kepi, a leather-billed cap that could be given a rakish tilt; but most wore “slouch” hats, broad-brimmed and high-crowned, stained by sweat and battered by weather. Those who had had the luck to raid the Yankee depots at Manassas had shoes or boots to wear. A great many started out barefoot, and many more would march their footgear to pieces over the next week. Their equipment was patchwork as well. Wagon trains and artillery batteries were pieced out with vehicles and guns captured from the enemy, many still bearing the “U.S.” marking.
The infantry would get a few days’ rest at Frederick and have an opportunity to gather supplies while it waited to see how the Federals would respond. The Twenty-second Georgia rejoiced in the opportunity to wash their clothes in the Monocacy River, “to drown some of the lice of which we had plenty. We had not washed our clothes for about a month, and the bugs were getting unbearable.” The problem was not exclusive to the Georgians. A local doctor noted “the penetrating ammoniacal smell,” a compound of piss and sweat both horse and human, which pervaded the air wherever the Rebels camped, and announced their coming if the wind was behind them. “A dirtier, filthier, more unsavory set of human beings never strolled through a town—marching it could not be called.” However, the doctor was mistaken in thinking their informal marching style a sign of indiscipline. As veterans they had learned to prefer the loose and natural stride to the measured pacing of the drill ground; and their discipline was generally good, as reflected in their obedience to the orders that forbade looting or seizing goods without at least offering payment—in Confederate currency. The troops understood that the purpose of their invasion was to win the hearts and minds of Marylanders for the Confederacy. Even the Frederick doctor who sneered at their rags admitted, “They all believe in themselves as well as in their generals, and are terribly in earnest.”25
To keep the enemy confused as to his purposes and movements, Lee ordered Stuart to scout and skirmish actively in two directions: toward Washington, to keep close tabs on Federal advances; and toward Baltimore, to threaten that city and compel the Yankees to divide their attention and their forces. Lee himself needed the rest to recuperate from the injury to his hands and arms. He used the interval (September 7–8) to complete his reports to Davis and issue his proclamation inviting Marylanders to rise in revolt.26
It soon became apparent that their hopes for a revolutionary uprising in Maryland were misplaced. Confederate sympathizers freely visited the camp, many of them relatives of Marylanders already serving in the Army of Northern Virginia. However, these were mainly from Baltimore and the Eastern Shore, where Southern sympathies were strongest. Unionist sentiment became stronger as one moved west from Baltimore—in Frederick, forty miles from Baltimore, the majority were either pro-Union or indifferent to the Southern cause. There would be no substantial accession of recruits from the state. Barely two hundred enlisted while the army was in Frederick.
The expectation of enjoying ample supplies of foodstuffs and clothing was also disappointed. There was not sufficient livestock to supply either draft animals or meat in the quantities required by the army. The grain harvest was not far advanced, and the millers demanded payment in U.S. currency for turning what grain there was into flour. Instead of bread or hardtack, the troops ate cooked grain and uncooked or even unripe corn. The result was that they began to suffer from diet-related diarrhea, in addition to the bowel problems endemic to all Civil War armies from drinking polluted water and living and eating in unsanitary conditions. For troops already worn down by hard marching and harder fighting, this was physically disabling. Even at this early and easy stage of the campaign in Maryland, Lee was newly concerned about further straggling, though Lee attributed it to “cowardice” rather than debility.27
On September 9 other misjudgments became apparent. Contrary to both expectations and all military logic, the Federal garrisons at Harpers Ferry and Martinsburg had not been withdrawn. As long as Federals held Harpers Ferry, the Valley supply line was closed. Lee had originally counted on that line for the resupply of ammunition and military equipment he would need once fighting began. Now he realized he would also have to draw on the Valley for foodstuff and livestock. So rather than linger in his exposed position at Frederick, Lee decided to fall back on his alternative plan to retreat west of the Catoctin Mountains and establish a base of operations from which he could “harass, if not destroy” the Federal army.28
The new plan made it imperative to open the Valley supply route by driving out or capturing the large Federal garrisons that held Harpers Ferry and Martinsburg. Their continued presence there was a serious tactical problem and something of an astonishment. When Lee’s army crossed the Potomac it had cut the line of supply linking these garrisons to Washington, and if Lee sent a force of any size into western Maryland they could be completely isolated and compelled to flee or surrender. By all the canons of military science the troops there should already have been withdrawn.
In fact, the canons of military science had nothing to do with the case. The garrisons were immobilized by the confusion or incompetence of their commanders, by the tangled webs of bureaucracy, and by the continuing fuss over lines of authority between Halleck and McClellan.
The Harpers Ferry garrison was not attached to the field army but to the Military Department of Maryland and Southern Pennsylvania, a purely administrative entity whose commander was General John Wool, a superannuated veteran of the War of 1812. Wool was out of touch with the garrison, his office in Baltimore was more than a hundred miles away, and he wanted to pass responsibility for the post to General McClellan. McClellan, too, wanted the garrisons placed under his command, partly for the usual reason—he was “outnumbered” and needed every available man—but also because in its present position it was uselessly imperiled. Halleck refused his request for no discernible reason and threw the responsibility for ordering an evacuation back on Wool, relying on his nonexistent “experience and local knowledge.” Wool was unwilling to take that responsibility and passed the decision down to the garrison commander, Colonel D. S. Miles—an incompetent who had been sent to rusticate at Harpers Ferry after a disgraceful performance at First Bull Run. While Miles would not take responsibility for abandoning the post without positive orders, he also failed to take the necessary measures for making the place defensible. It would take nearly a week for McClellan to pry command of the garrison away from Halleck—from the sixth to the eleventh of September—and by then it was too late for Miles to do anything but hunker down.29
The Martinsburg garrison, commanded by Brigadier General Julius White, was the larger of the two, but it seems to have been an administrative orphan. His command was far out on a limb, holding an indefensible position in the West Virginia foothills, ostensibly protecting the B&O Railroad line west of Harpers Ferry. While Halleck, Wool, and McClellan fussed over Harpers Ferry, White was left without instructions, and he, too, chose to stay where he was.
This tangled web of blunder and misfeasance would doom the garrison to hapless defeat. But it also threatened to derange Lee’s whole plan of campaign, and confronted him with an extremely complex tactical problem. In trying to solve it he would expose his army to destruction.
LEE DIVIDES HIS ARMY, SEPTEMBER 10–13, 1862
The combined garrisons of Martinsburg and Harpers Ferry mustered nearly thirteen thousand Union soldiers. The force sent against them would have to be at least twice as strong. Stonewall Jackson was the obvious choice to lead the expedition. Jackson excelled in independent operations and had the most thorough knowledge of the Harpers Ferry terrain. Lee met with him on the evening of September 9 to work out the details.30
The operation was designed to surround and capture the Federal force in Harpers Ferry, not simply to drive it away. The peculiar character of the terrain would complicate the maneuvers required to assault it. Harpers Ferry sat at the northern entry to the Shenandoah Valley, among the high, wooded ridges of western Virginia. The town was sited on a triangular peninsula formed by the confluence of the Shenandoah and Potomac rivers, with the Shenandoah Valley opening out to the west. A ridge named Bolivar Heights ran across the base of the peninsula, and a trench line there defended the town from troops attacking out of the Valley. The town was overlooked by two other ridge lines. Southward on the other side of the Shenandoah River was Loudon Heights and northeastward, on the other side of the Potomac, was a high bluff called Maryland Heights. If an enemy were to capture these heights, their artillery would dominate the entire peninsula and the garrison would have to surrender. Colonel Miles posted a brigade and artillery on Maryland Heights, but Loudon Heights was weakly held.
To surround Harpers Ferry and force its surrender, Confederate troops had to put a large infantry force west of the town to attack Bolivar Heights, and at the same time seize both Loudon and Maryland Heights. However, because of the high mountains, deep rivers, and bad roads of the region there was no single route by which a united Confederate army could reach its objectives. Infantry could only attack the town by getting into the Shenandoah Valley west of Bolivar Heights, but to get there from Frederick, Jackson’s troops would have to march past Harpers Ferry on the opposite bank of the Potomac, pass through the little crossroad town of Sharpsburg, cross the Potomac at Shepherdstown Ford five or six miles upstream from Harpers Ferry, then work their way down through the hills. En route Jackson would have to detach a large force under the command of General Lafayette McLaws to capture Maryland Heights, which could only be approached from the north. To complicate matters, Loudon Heights could only be approached from the east, on the Virginia side of the Potomac. It would therefore be necessary for Lee to further divide Jackson’s force, sending J. G. Walker’s infantry division with some cavalry back into Virginia to march on Loudon Heights from that side. Once they separated, Jackson, McLaws, and Walker would be unable to communicate with each other until all three arrived at Harpers Ferry. It would therefore be impossible for Jackson to control and coordinate their movements or compensate for setbacks or delays they might suffer. All three columns were supposed to converge on and take Harpers Ferry by September 12.
While Lee and Jackson were conferring, General Longstreet arrived. He was initially opposed to the idea of dividing the army, preferring the more cautious plan of advancing against Harpers Ferry as a united force. Lee rejected this idea, for reasons that remain conjectural. However, Lee modified his plan in accord with Longstreet’s suggestions. Lee’s original plan called for Walker’s Division and a cavalry brigade, about five thousand troops, to cross the Potomac back into Virginia, then march west to seize Loudon Heights. Jackson was to march west through the gaps of South Mountain with his own three divisions, about twenty thousand men, followed by McLaws’s Division of about seven thousand, which was supposed to peel off and capture Maryland Heights. At Longstreet’s urging Lee strengthened McLaws’s column with the three brigades of R. H. Anderson’s Division, raising the strength of his detachment to seven thousand. This addition raised to about thirty-seven thousand the troops committed to seizing Harpers Ferry—more than half of Lee’s total force of between sixty-five thousand and seventy thousand of all arms. In the meantime Longstreet’s command, reduced to three infantry divisions and screened by Stuart’s cavalry (about twenty-six thousand total), would fall back on Boonsboro, near one of the gaps in the Catoctin Mountains, where it could offer earlier resistance to any Federal advance out of Washington and secure control of Hagerstown. This was consistent with his overriding desire to retain the offensive initiative: even as he moved to secure his supply line he would be establishing the base for his next offensive.31
LEE’S DECISION WAS probably made before he learned from the Baltimore papers of McClellan’s assumption of command, but when it came the news would have reassured him. He believed he could count on McClellan to undertake an extended period of reorganization, to be followed by the most cautious of advances. Meanwhile, with his supply line cleared by the capture of Harpers Ferry, Lee would reunite his army either in Boonsboro or Hagerstown. From that point, his campaign would take on the desired form. Catoctin Mountain and South Mountain would be a defensive bulwark between his army and McClellan’s. He would be in an ideal position to harass the Federals by raiding into Pennsylvania or to strike for a decision by marching in force against Harrisburg and seeking a general engagement.
The maneuver plan was extraordinarily complex and risky. Its success required the coordination of three converging columns, operating on separate lines, unable to communicate with or support each other. The timetable was also aggressive: Lee expected Jackson to have Harpers Ferry in hand by September 12, less than three days after he marched out of Frederick. The plan made no allowance for the appearance of unanticipated difficulties that might slow the rate of march or require a change of route. Its success also depended on the dangerous assumption that enemy responses would conform to Lee’s expectations. Until Harpers Ferry surrendered or was abandoned, the separate elements of Lee’s army would be far apart. Lee had taken similar risks in the campaign against Pope, but then he had played for the chance to destroy the main force arrayed against him. Here the objective was the capture of an outpost, which could have been taken by the safer (albeit slower) means suggested by Longstreet. If the garrisons of Martinsburg and Harpers Ferry chose to fight rather than flee, Lee’s timetable would fall apart. If McClellan was not as slow to advance as Lee expected—if Lincoln’s goading or some unforeseeable turn of events led him to move with unwonted energy—he might catch Lee’s army while it was still divided and defeat it in detail.
The directives that would set the several columns of the army in motion were written out on September 9 in Special Order No. 191, and copies were sent by mounted courier to President Davis and to the commanders of those divisions not yet incorporated with the commands of Jackson and Longstreet. Two copies were directed to General D. H. Hill, one from Lee’s headquarters and another from Jackson’s. Protocol demanded the latter, because Hill had been temporarily attached to Jackson’s command. The copy sent by Jackson never reached Hill. It was lost by the courier, along with three cigars, in a field outside of Frederick. The courier never reported the loss and Hill did not miss it, since he received the original from Lee’s headquarters.32
ALL OF THE CONFEDERATE COLUMNS, with the exception of Walker’s, would move west by the same route, the well-engineered National Road that ran twenty-five miles west from Frederick through Turner’s Gap in South Mountain to Boonsboro. But Jackson’s command moved slowly. His troops had been nicknamed “foot cavalry” for the speed with which they covered great distances, but they were worn out from months of such marches, and the macadam and packed-stone road was agony to march on with the broken shoes or bare feet most were reduced to. Jackson halted at Boonsboro so that his scouts could reconnoiter the route he planned to follow. For all of September 10 and part of the eleventh the rest of the army was queued up along the road behind Jackson or waiting in Frederick until he cleared the way.
Boonsboro was a small town, a way station on the National Road, a few taverns and stores along the highway. Skirmishes with Union patrols revealed that the Sharpsburg–Shepherdstown road was watched, and Jackson received reports indicating that the Martinsburg garrison (about six thousand men) was still holding its ground. Jackson therefore altered his line of march to cross the Potomac upstream at Williamsport, ten miles farther from Harpers Ferry than Shepherdstown, so he could cross unopposed and close off the Martinsburg garrison’s escape route to the west. Instead of marching a bit over twenty miles to reach an abandoned Martinsburg, as Lee had planned, Jackson’s troops would now have to march thirty-five miles, then deploy for action against the garrison. Jackson pressed ahead on the eleventh and by evening had most of his command across the river and within two hours’ march of Martinsburg—with absolutely no chance of reaching Harpers Ferry on the twelfth, let alone capturing the place. Moreover, his men were exhausted, and he had lost heavily in stragglers.
Lee was aware that the physical strength of his men had been seriously worn by months of campaigning, hundreds of miles of hard marching, and heavy combat. Jackson’s command, which had the longest line of march to Harpers Ferry, had already done more forced marches than the rest of the army. Lee therefore directed Jackson to limit his rate of march to three miles an hour and allow frequent rests. But these measures were inadequate. Jackson had to cover a considerable distance in a limited time, and while he may have slowed his usual rate of march he could not come near meeting his schedule without extending the hours of march from predawn to postdusk. His men were desperately hungry, poorly supplied by their commissary, and unable to find any quantity of decent food by foraging. Soldiers with blistered and bleeding feet who could not keep up fell out by the roadside. Most would struggle to rejoin their commands, but many gave up and went back to Virginia, where the sandy roads were gentler and the rations more reliable. Other men fell out because they were racked with diarrhea and dysentery, which left them too weak to catch up. One of Jackson’s brigadiers was bitter about the fact that “[o]ur men march and fight without provisions, living on green corn. . . . Jackson would kill up an army the way he marches and the bad management in the subsistence Dept.” Jackson’s command may have lost nearly a third of its numbers just marching to Harpers Ferry.33
The regiments that fell back to Boonsboro with Lee and Longstreet on September 11 would similarly leak stragglers all the way. Lee’s troop strength—which was, to begin with, less than he supposed—was diminishing every day. This wing consisted of two divisions of Longstreet’s command plus D. H. Hill’s Division, accompanied by the army’s reserve artillery and the large wagon trains that carried its supplies and ammunition. Lee also disposed of the cavalry concentrated under Stuart’s direct command, which lingered in and around Frederick to screen the army’s movements. On September 11 Longstreet’s divisions, with the army’s artillery and supply wagon train, completed their march to Boonsboro, while Hill’s Division lingered as a rear guard in the vicinity of Turner’s Gap, where the National Road from Frederick crossed the high ridge of South Mountain. Shortly after their arrival, Lee received reports that a Federal force of unknown strength was marching down from the north, threatening to occupy Hagerstown some fifteen miles north of Boonsboro. The report was probably an exaggerated account of the militia and home-guard units then forming to resist any incursion into southern Pennsylvania, but Lee was concerned by it.
Hagerstown played a significant role in Lee’s new plans. It was one of the larger towns in the region, well stocked with provisions, clothing, and shoes. It had good road connections in every direction and was also the southern railhead for a line that ran north into Pennsylvania, which made it an ideal base for future offensives into the north. Lee also wanted Hagerstown as a place, well beyond McClellan’s reach, in which to park his army’s large train of supply wagons and its reserve artillery. He therefore sent a reluctant Longstreet to secure Hagerstown, and divided his command yet again, leaving only D. H. Hill’s Division to defend the South Mountain gaps and the army’s headquarters and base of supply at Boonsboro.34
While Jackson was marching to Williamsport and Lee to Boonsboro, the columns led by Walker and McLaws had made little progress. Walker had crossed back into Virginia on the tenth, but his men were so worn-out that he decided to let them rest all the next day before moving on Loudon Heights on September 12.
On that day McLaws’s command, consisting of his own division and R. H. Anderson’s, was preparing to cross South Mountain at Crampton’s Gap, which was twelve miles south of Turner’s Gap. His was the most crucial assignment: to make Harpers Ferry untenable by seizing Maryland Heights, the high bluff that rose at the southern tip of a long ridge called Elk Mountain. But McLaws had no experience in independent command. The terrain before him was difficult and unfamiliar, held by an enemy whose strength and position were unknown, and the army had not provided him with scouts who knew the area or cavalry sufficient to clarify the situation. McLaws had not only to organize an assault on a position of unknown form and strength, in his approach he had to prevent the garrison from breaking out and escaping to the north. He therefore spread his command in a broad line, spanning Pleasant Valley and straddling Elk Mountain as it advanced. The difficult terrain and the need to maintain the cohesion of his line slowed his advance, so that his lead brigade did not locate the Federal force on Maryland Heights until the evening of the twelfth. Since that force appeared to be sizable, and holding a strong position, McLaws could do nothing until he had concentrated most of his command against it.
Meanwhile, unbeknownst to any of the Confederate commanders, on the night of September 11–12, General White had decided to escape Jackson’s impending assault by abandoning Martinsburg and retreating to join Miles’s command at Harpers Ferry, some fifteen miles south. Thus on the twelfth of September there were nearly thirteen thousand Federal troops concentrated in Harpers Ferry, twice as many as McLaws and Walker expected to meet; and they were preparing to fight, not flee. Because the three prongs of the Confederate advance had no way of communicating with each other, none could call on any of the others for support in case of attack. Under an enterprising commander, a force of that size could have held McLaws at bay, crushed Walker’s Division, and made its escape before Jackson could even reach the scene. It was Jackson’s good luck that the Union commanders, White and Miles, were not at all enterprising, and that Miles (who had primary responsibility for the Union defense) was barely competent.
On September 12 the three independent Confederate columns were converging, more or less simultaneously, on their objectives. Walker’s Division seized undefended Loudon Heights on the evening of September 12. McLaws’s advancing troops discovered that Colonel Miles had entrusted the defense of Maryland Heights to an inadequate infantry force under an incompetent commander. McLaws would attack and carry the position the next morning. Meanwhile Jackson’s command was marching over the hills from Martinsburg and would soon begin deploying across the face of Bolivar Heights. The Union garrison was now trapped in the town of Harpers Ferry, and once Walker and McLaws got their guns in position the town would be exposed to devastating fire from high ground on two sides.
However, although the capture of Harpers Ferry was ultimately assured, the operation was already a full day behind schedule and far from complete. It would take another day to bring up the guns and put the town under fire, and there was no telling how long the garrison might hold out before surrendering. Then there would be further delay while the victors dealt with their large haul of POWs, distributed the captured food, uniforms, and ammunition they so badly needed, and recovered from days of hard marching and the strain of combat. Until all that was done, Lee’s army would remain badly split: more than half of it around Harpers Ferry, the rest divided among Turner’s Gap, Boonsboro, and Hagerstown. In these circumstances the army’s safety depended on Stuart’s cavalry screen, and on General McClellan staying true to type, advancing slowly and cautiously, reluctant to press hard.
McClellan’s initial movements justified Lee’s belief that the Union advance was in fact extremely deliberate. The Federal cavalry never seriously tested Stuart’s screen, and McClellan’s advance guard did not retake Frederick until September 13, by which time Jackson was at Harpers Ferry and all of Lee’s infantry was west of Catoctin Mountain.
Then, on September 13, Lee’s campaign encountered bad luck so egregious that it would have nullified even the most perfect tactical plans. A pair of Union infantrymen, preparing to bivouac outside of Frederick, found under a tree an envelope containing three cigars and a piece of paper—which, when opened, proved to be the lost copy of Lee’s General Order 191. The finding of that lost order gave McClellan detailed information on Lee’s plans and exposed the dangerous division of his army. It compromised Lee’s plan of campaign and encouraged the cautious McClellan to dare the offensive that culminated at Antietam.