CHAPTER THREE

Local Knowledge: Stonewall Jackson in the Shenandoah Valley

IN THE MEDITERRANEAN in 1798 Nelson had been much mystified and often misled, at least twice to his crucial disadvantage, even though he enjoyed a superiority of force, intimately understood the geography of his theatre and was pursuing an enemy with a severely restricted choice of manoeuvre. All had come right in the end, but his victory would have been even more complete had he made identifiably different decisions on several occasions. The Mediterranean was a closed and circumscribed strategic arena in which, in optimal circumstances, a fleet commander might have achieved total domination.

In 1862, in a sort of mirror image of Nelson’s Mediterranean campaign, though on land rather than at sea, Thomas “Stonewall” Jackson, also operating within a closed strategic arena, the Shenandoah Valley of Virginia, had consistently mystified and misled his enemy—his catchphrase was “always mystify and mislead”—even though usually inferior in force to the Union armies chasing him and despite severe geographical limitations on his room to manoeuvre. In 1798 Nelson had never, until the last stages of his pursuit, had quite enough intelligence. In 1862 Jackson enjoyed ample intelligence and exploited it to win a succession of victories his objective weakness should have denied him. Better information, keener anticipation and cleverer judgement combined to make Jackson’s Valley campaign a model of an active intelligence victory.

The situation of the Southern Confederacy, at the outbreak of the American Civil War, was intrinsically weak. By every material measure—population, industrial capacity, miles of rail track, to mention but a few essential indices—its capacity to wage successful war was greatly exceeded by that of the North. Of the United States’ thirty-two million people, only five million lived in the eleven seceding states (and four million blacks whom, as slaves, the Confederacy would not arm); of the country’s 30,000 miles of rail track, 22,000 ran in the Northern states; the North produced 94 per cent of the country’s manufactured goods and the vast proportion of its raw materials, including iron, steel and coal. The South was a rich region nonetheless, but it was rich in cotton, tobacco, rice and sugar, crops which brought their planters income in overseas sales and the export of which the North could and did interrupt, as soon as secession was declared, by blockading the Confederacy’s coasts.¹

Had the war been a contest between economic systems alone—as Winfield Scott, Union General-in-Chief, hoped to keep it—the South would have quickly collapsed.² The Southern people, however, were resolute in their determination to preserve “States Rights,” the legal issue over which they had declared separation, and soon showed themselves equally resolved to sustain the deprivations that economic isolation brought them. Hardy and frugal in their rural way of life, they rapidly made it clear that they would have to be beaten in battle if they were to be brought to surrender. President Abraham Lincoln was quick to grasp that, quicker than Winfield Scott. The question was, where to fight? The South might be materially weak, but it was strategically and geographically very strong. Protected on two sides by sea and ocean, it was also shut off from the rest of the United States by unsettled semi-desert to the west and by mountains to the north. Its paucity of internal communications, which in any case connected poorly with those of the North, was a positive strategic advantage. Moreover, in the great valley of the Mississippi, it enjoyed the protection of a sort of secondary internal water frontier, denying Northern armies any easy way forward into its heartland. Above all, the South’s enormous size—its eleven states covered an area as large as Europe west of Russia—was in itself a strength. Even if its outer crust could be penetrated by the Union, there still remained the difficulty of covering the vast distances inside the South between the point of entry and an objective of any value. To get from anywhere to anywhere within the Deep South was a problem in peacetime—there were few railroads, appalling or nonexistent roads, while the inland rivers were too short and usually ran the wrong way. In wartime the problem seemed designed to defy the efforts of a general of genius.

The South, on the other hand, should it choose to attack, faced no such problem. From its Virginian frontier with the Union, Washington, the Federal capital, lay only forty miles distant; not much farther lay the great city of Baltimore. Also within striking distance lay smaller but still desirable urban targets and the rich farmlands of Maryland and Pennsylvania. A successful thrust into the North would also bring menace to industrial New Jersey and perhaps even New York. The South’s strength was the widespread dispersion of its centres of population and production; the North’s weakness was the concentration of similar objectives in the Middle Atlantic coastal corridor and that corridor’s vulnerability to a Confederate offensive.

Critically, the South had a way in, the Shenandoah Valley. The dominant geographical feature of Atlantic America is the Appalachian Mountain chain, which runs roughly parallel to the coast, at diminishing distance, from Alabama to Maine. The Appalachians shut off the enormous interior of the continent from the coastal strip for hundreds of miles and had been used by the French, when they ruled Canada and what they called Louisiana, to deny the Ohio country and the Mississippi Valley to the English colonists in Virginia, the Carolinas and Georgia.

The defeat of the French in 1763 had opened up the trans-Appalachian wilderness to the English and thereby set in train the events that led to the division of 1861. Virginians, Carolinians and Georgians, migrating westward, had taken slavery with them into Mississippi and Tennessee. New Yorkers, Pennsylvanians and New Englanders had established the Midwest as non-slave territory. Disputes over the states of the westerly borderlands had generated the constitutional conflict that resulted in the crisis of secession. Slave or free? That was the issue over the new lands opening up to settlement in the old French region of “Louisiana.” When it could not be settled by debate, the Southerners chose separation.

What would then have happened had the North chosen to adopt Winfield Scott’s passive Anaconda Plan, and had the South chosen to sit inside its formidable natural frontiers, challenges easy speculation. There might not have been a civil war at all. Neither eventuality occurred. The South, as James McPherson convincingly argues, was spoiling for a fight.³ The outraged North, outraged both by the challenge to the Constitution and by the South’s defiant defence of the sin of slavery, was adamant for an offensive. Inspired by the cry “On to Richmond,” the Virginian capital of the Confederacy, the North launched the manoeuvre that led to encounter and defeat at the First Battle of Bull Run (Manassas) in July 1861.

In the aftermath, the Northern leadership pondered a better way. In the west, beyond the Appalachians, local generals tried to open up a new front on the river approaches to the Mississippi. Along the coast, Union admirals began to close off the Confederates’ outlets to the wider world. In Washington, however, Lincoln and his government sought a more direct means to strike at their Confederate enemies. The way, as they recognised, was barred by a succession of water obstacles, the short rivers running off the Appalachian chain between the mountains and the Atlantic that furnished one of the South’s best strategic defences. The course of the Rappahannock, the Mattapony, the York, the James, might have been designed by a friend of slavery to frustrate the advance of Northern armies to the seat of rebellion. At twenty-mile intervals or less, river after river, each easily defensible, stood between the Union forces and the enemy capital.

A solution to the infuriating strategic difficulty was proposed in the spring of 1862 by the man who had recently become Abraham Lincoln’s favoured general, George McClellan. Convinced that a repetition of the “On to Richmond” effort by the overland route would stumble again, at one or other of the water obstacles, McClellan persuaded Lincoln to let him put the Army of the Potomac, the Union’s main force, into troopships, sail it down Chesapeake Bay from Washington and land it at the point of the Virginian Peninsula, between the York and James rivers. There he would enjoy the security of a firm base, Fortress Monroe, one of the great stone citadels of the coast-defence programme known as the Third System, and still in Union hands; from it by easy marches Richmond lay only seventy miles distant. McClellan was confident that he could make the amphibious operation work. As a junior officer marked for promotion, he had been sent in 1855 as an observer to the Anglo-French expedition to the Crimea, so had seen an amphibious operation at work with his own eyes, and had also witnessed the military use of the newly invented telegraph.⁴ As a railroad company executive, which he left the army to become in 1856, he had learnt more about the telegraph as a means of control and also about bulk supply over long distances; both telegraphic control and efficient logistics were to be of central importance in the running of the Peninsula Campaign.

The arrival of the Army of the Potomac at Fortress Monroe greatly alarmed the Confederate high command. Entrenchments were hastily dug across the nose of the Peninsula, in places following the line of earthworks constructed during the British defence of Yorktown in 1781. General Joseph E. Johnston’s Army of Northern Virginia was withdrawn from the proximity of Washington, and Richmond was put into a state of defence. The Confederates were right to be alarmed, despite the temporary security these measures provided. They were greatly outnumbered on the spot, by 105,000 to 60,000 in the vicinity of Richmond, and potentially by an even larger number. Three other Northern armies hovered nearby, that of Frémont in West Virginia, that of McDowell near Washington and that of Banks in the Shenandoah Valley. If they could be brought into combination with McClellan’s, Joseph E. Johnston’s Army of Northern Virginia would be overwhelmed and the fate of the Confederacy sealed.

There were only two points of light for the Confederacy amid the encircling gloom. The first was McClellan’s capacity for procrastination. Though objectively superior in strength to the enemy, he was constitutionally incapable of accepting the evidence, constantly petitioned Lincoln for more troops and issued frequent warnings of his inability to proceed unless reinforced. Instead of pressing forward, he hung back, professing to see dangers visible only to himself, thus conferring on his enemies opportunities to strengthen their position which they should not have been allowed. He had landed at Fortress Monroe on 22 March 1862. He then spent a month, 4 April to 4 May, besieging the weak Confederate position at Yorktown. Not until 5 May, after the Confederate garrison had withdrawn, did he advance to fight his first proper battle, at Williamsburg, and not until the 25th did he draw near to Richmond, his proper objective. He had taken over eight weeks to cover seventy miles and had inflicted no damage on the enemy at all. Joseph E. Johnston’s army stood intact and remained to be brought to battle.

The other point of light was the existence of a Confederate diversionary force, poised to unsettle both McClellan and Lincoln, though in different ways. McClellan could be thrown into anxiety by any move threatening to deprive him of the reinforcements he craved. Lincoln, more tellingly, was prone to alarm at any prospect of a Confederate advance against Washington. The Confederates around Richmond lacked the capacity to level either threat. General Thomas Jackson, “Stonewall,” far away though his small army lay in the Shenandoah Valley, was equipped, equally by location and capability, to organise both. Any thrust northward he might make would menace Washington, which Lincoln increasingly believed had been selfishly stripped of troops by McClellan to bolster his Peninsula adventure. Such a thrust would simultaneously lessen the likelihood of Lincoln’s agreement to the redeployment of the covering armies of Banks and McDowell from the Shenandoah Valley vicinity towards McClellan at Richmond. Stonewall Jackson, in the spring of 1862, suddenly found himself in a “swing” position, capable of altering the course of the war, if he handled his force correctly, with decisive effect.

The Shenandoah Valley was a strategic corridor, which worked as a critical anomaly in the military geography of the Civil War. The heartland of the Confederacy, as originally constituted, between the sea, the Mississippi and the mountains, was virtually impenetrable. McClellan had cracked the carapace by finding a maritime point of entry on the Virginian Peninsula, but to enlarge the breach he would need to show a determination and single-mindedness his contemporaries of the West Point class of 1846 had good reason to doubt he possessed. Otherwise, as long as the lower course of the Mississippi was held, there was only one other way in: down the Shenandoah. The Valley is the easternmost feature of the central Appalachian chain. Its southern exits lead into the plains of Virginia, the Carolinas and Georgia, its northern exits into Maryland, Pennsylvania and towards the outskirts of Washington. In the circumstances of the Civil War, it could be used offensively or defensively. Theoretically, the North could use it as a way into the Confederate heartland; in practice, the lack of a north–south railroad within the Valley made that manoeuvre logistically too difficult to undertake, though it was one against which the South always had to be on guard. On the other hand, the South could much more easily use the upper mouth of the Valley as a sally port from which to surprise Northern armies near their major cities. During the course of the war, it was the South which better exploited the strategic potentialities of the Valley and never more so than in the spring of 1862.

There is a large and small strategic geography of the Valley. The large is that of a corridor leading either in or out of what between 1861 and 1865 was the Confederacy; the small is that of its internal features which, if correctly understood, can be put to decisive military use. About 120 miles long and 30 wide, from the headwaters of the South River to the confluence of the Shenandoah and Potomac at Harper’s Ferry, and from the crest of the Alleghenies to that of the Blue Ridge Mountains, the Valley is an enclosed environment of what, in 1862, was rich, cleared farming land. Down the centre, however, runs a dividing ridge, the Massanutten Mountain, which itself divides the Shenandoah River into a North and South Fork. Joined near Front Royal, the forks become the Shenandoah proper, to run northward for forty miles to enter the Potomac at Harper’s Ferry. The tail of the South Fork separates at Port Republic into three minor streams, the North, Middle and South rivers.

Many rivers mean many bridges, and there were, in 1862, at least twelve of military significance, providing crossings at the Valley’s principal townships and villages; like others of that period in the American settlement of the interior, most were wooden and easily burnt. Besides bridges, there were other passages of strategic significance, fords here and there, but also gaps in the surrounding mountain chains. Those leading out into West Virginia, dominated by the North, were few and of lesser military importance. Those giving on to the lowlands of Virginia proper were more numerous—there were eleven in all—and of altogether greater significance, since they provided a Confederate army inside the Valley with the means to dodge back and forth. Equally important was the gap through the central Massanutten ridge between New Market and Luray and the east–west links around the Massanutten’s headlands at Front Royal and Port Republic.

The Valley’s internal geography determined its road network. It was better west of the Massanutten, where the Valley Pike—an all-weather macadamised road of impacted gravel—led from Williamsport on the Potomac, via Winchester, Strasburg, New Market and Harrisonburg, to Staunton, between the South and Middle Rivers. East of the Massanutten, an inferior road ran from Front Royal through Luray to Port Republic and eventually to a junction with the Valley Pike at Staunton.⁵

Few, if any, in the Northern armies understood the Valley’s geography. There were two reasons for that. The first was that in peacetime, the Valley’s communications with the outside world had been almost exclusively by river, up and down the Shenandoah and its branches to Harper’s Ferry; so important were the waterways that Valley people described the northward passage to the confluence as going “down,” the southward as “up.” The North therefore knew the Valley only as a river system, and then at its external points of connection. The second was that there were virtually no Valley maps. That was a prevailing condition of warmaking between the Union and the Confederacy. The Federal government had, before 1861, invested considerable sums in mapping the United States’ coasts; one of the branches of the United States Army, the principal instrument of the government’s internal administration, was a Corps of Topographical Engineers. It had also sponsored a major exploration of the west, as a support to its sponsorship of settlement beyond the Mississippi. It had done nothing similar in the old Thirteen Colonies or the eastern states founded since 1782. The result was that the generals of the Civil War embarked on their operations with wholly inadequate cartographic resources.

No accurate military maps existed. [The Union] General Henry W. Halleck was running a campaign in the western theatre in 1862 with maps he got from a book store. With frenetic haste, the general set topographical officers and civilian experts to work, making maps, but the resulting charts were generally incorrect. Benjamin H. Latrobe, the civil engineer, drew a map for a general going into Western Virginia, but the best he could promise was that it would not mislead the expedition. General George B. McClellan had elaborate maps prepared for his Virginian expedition of 1862 and found to his dismay when he arrived on the scene that they were unreliable; “the roads are wrong,” he wailed. Not until 1863 did the [Northern] Army of the Potomac have an accurate map of northern Virginia, its theatre of operations.⁶

At the root of the trouble lay the cartographic backwardness of the United States. That might be thought understandable: the United States was an enormous country, still largely unsettled, by no means completely explored and without a central mapping agency; the army had its Corps of Topographical Engineers, the navy a Hydrographic Office and the federal government a Coast Survey, but they were all tiny.⁷ The basis for accurate survey, a comprehensive triangulation of the land mass, was absent. Yet it had been done elsewhere. The British Isles had been triangulated and a comprehensive series of high-quality maps published, at one inch to the mile, by the Ordnance Survey, beginning in 1791; a small undertaking, certainly, but magnificently accomplished. Impressive by any standards was the work of the Survey of India; India, though smaller than the United States, is topographically even more complex, because of the height and extent of the Himalayan chain. Beginning in 1800, the Survey, under the direction of a succession of military engineers, had embarked on a complete triangulation. Triangulation, which supplies measured distances between a series of intervisible points, allowing for the curvature of the earth, provides the grid from which accurate maps can subsequently be drawn. It was largely complete by 1830 but was subsequently extended and corrected, notably under the leadership of Sir George Everest, after whom the world’s highest mountain is named. His team of surveyors and trigonometricians was never more than a few hundred strong but, largely inspired by the challenge of the enterprise itself, they succeeded within seventy years in producing a complete series of accurate maps of a sub-continent equivalent in size to that of the United States west of the Mississippi.⁸

By 1861, no triangulation of the United States had been undertaken. It was a strange blind spot in the American attitude to their magnificent country. George Washington was by training a surveyor; so was Abraham Lincoln. Thomas Jefferson, most intellectual of presidents, had a passionate interest in exploration and sponsored the Lewis and Clark trans-continental expedition to the north-west in 1804. He made it clear, however, that its purpose was to discover “the most direct and practicable water communication across this continent . . . for commerce.” Route-finding, first for commerce, then for settlement, then railroads, defined American official interest in continental geography. In 1836 President Andrew Jackson sent a U.S. Exploring Expedition, under Lieutenant Charles Wilkes, to investigate the territory of the United States, but it was seaborne and largely committed to investigating the coasts. The earliest major exploration of the interior was authorised in furtherance of the Pacific Railroad Act of 1853, “to Ascertain the most Practical and Economical Route for a Railroad from the Mississippi River to the Pacific Ocean,” which designated five possible lines, all to be surveyed by the army’s Corps of Topographical Engineers. The routes were mapped, but no comprehensive and accurate survey of the United States resulted. That lay in the future.⁹

There were, of course, already many local maps of the United States, made necessary by the need to define sub-division of the land for farming under the pressure of immigration and westward expansion. Flatness, so characteristic of the American landscape in the Midwest east of the Mississippi and in the Great Plains beyond, allowed accurate delineation of property boundaries by reference to astronomical observations of latitude early on and longitude, by telegraphic time calculation, by the 1860s. Such mapping, however, was piecemeal. Without comprehensive triangulation, local maps did not accurately connect with one another, nor did they, in the hilly areas of the Appalachian chain and in the coastal regions to the east, usefully depict height, or contour. No wonder that, as late as 1864, Colonel Orlando Poe, General William Sherman’s chief engineer, should complain that the maps of North Carolina he was able to find “vie with each other in inaccuracy.”¹⁰ Traditionally maps had been military secrets, those of one’s own country to be kept from the enemy, those of his to be made with stealth; with reason, for mapmaking was rightly regarded as espionage. Frederick the Great in 1742 established a secret map room (Plankammer) in his palace at Potsdam, which contained maps both of Prussia and of surrounding territories, such as that of Silesia, which he had made before his invasion that caused the Seven Years War.¹¹ The Survey of India ran what was effectively a widespread espionage network in the countries bordering the Indian empire to the north, including Tibet, Nepal, Afghanistan, China and Russian Central Asia, staffed by Indians who were trained to measure distances by counting their steps on strings of prayer beads. Hurree Chunder Mookerjee, one of Kipling’s most delightful creations in the cast of characters in Kim, was such an agent, but he had his models in real life. The most famous, as he became when allowed to emerge from obscurity, was Nain Singh, known as “the Pundit” or sage, who between 1864 and 1875 twice visited Lhasa, then a closed city, covered 1,200 miles of previously unsurveyed country and followed the course of the great Tsangpo River for 600 miles from its source. On retirement from the Survey of India, he was rewarded with a grant of land, the rank of Commander of the Order of the Indian Empire and the gold medal of the Royal Geographical Society, at which he lectured to rapt attention when he visited England.¹² He was perhaps lucky to survive to retiring age. Nepal, when threatened with invasion by the East India Company in 1814, carefully disguised the mouth of the main road leading into the country and the government threatened death for its betrayal.¹³

The South, threatened with invasion in 1861, could not disguise the mouth of its internal roads, since they connected with those of the North. Thereafter, however, their course was often poorly reproduced on such maps as were available to Northern generals, or inaccurately represented, or not marked at all. Local knowledge often counted far more than the plates in a shoddy bookshop atlas. It was much more readily available, inside the South, to Confederate defenders than Union invaders. Without it, confusions accumulated. Even quite good maps could be out of date, while there was no guarantee that the mapmaker’s choice of place-name was that used by locals. “Cold Harbor, Virginia” (the site of one of General Ulysses Grant’s battles in 1864) “was sometimes called Coal Harbor, and there was also a New Cold Harbor and a ‘burned’ Cold Harbor. Burned Coal Harbor was known by the locals as Old Cold Harbor. Many of the roads were known by one of two names: the Market or River Road; the Williamsburg or Seven Mile Road; the Quaker or Willis Church Road. To add to the confusion, there were sometimes other nearby roads with the same or similar names that ran in completely different directions.”¹⁴

Locals knew; invaders did not. That was to confer an almost consistent advantage on the South, which, for most of the war, was campaigning within its own territory and defending it very often with locally raised troops. That was particularly the case in the Shenandoah Valley in 1862. Stonewall Jackson, the commander of the Valley Army, was a Valley man. After retiring from the regular army, he had become a professor at the Virginia Military Institute, the private military academy at Lexington at the Valley’s southern end. Many in the Valley Army were Valley men, particularly those of the Stonewall Brigade, which had won its name at the First Battle of Bull Run, and of the Rockbridge Artillery, largely recruited from students at Washington College, also in Lexington. Perhaps the most important Valley man in the Valley Army, however, was a civilian, Jedediah Hotchkiss. A schoolmaster, he had set up his own school at Staunton in 1847. It flourished and, though a New Yorker, he stayed. He also began to pursue the hobby of mapmaking. General Robert E. Lee employed him as a mapmaker in his campaign in the Alleghenies, west of the Valley, in 1861. In 1862, home after illness, he attached himself to the Valley Army and was introduced to Jackson. The latter was impressed by his local knowledge and on 26 March added him, though he was a civilian and would remain so, to his staff. His first order to Hotchkiss was “I want you to make me a map of the Valley from Harper’s Ferry to Lexington, showing all the points of defense and offense between those points.”¹⁵

Hotchkiss set to work. He was untrained in cartography but methodical. He first surveyed the terrain from horseback, making sketches and notes as he moved around the terrain, then worked his observations up into a finished product. His 1862 map of the Valley still exists.¹⁶ The course of the rivers is shown in pale blue, the road network in red, hills (uncontoured and without spot heights) in black, by hatching. There is no scale though, as the bottom of the sheet has been torn off, it may simply be missing. As a map, it reflects all the defects of those of the Civil War period: the appearance is messy, there is both too much and too little detail, and it has an unfinished, amateur look. Compared to the clear and elegant map of Yorktown, on the Virginia Peninsula, drawn by Thomas Jefferson Cram from an original by one of the French ingénieurs géographes of Rochambeau’s army in 1781, it is a very inferior thing.¹⁷ It is perhaps unsurprising that Jackson disliked the drawing classes at West Point more than any other subject. It seems probable that mapmaking was badly taught at the Academy, and if Federal military mapmaking, which in Europe and particularly Britain set the standard, was defective, it would follow that American mapmaking in general was unsatisfactory.

Still, Hotchkiss provided Jackson with a map based on local knowledge and derived from contemporary observation, and that put the general on a superior footing to his Union opponents. As late as 1864, during Jubal Early’s resumption of the Confederate offensive in the Valley, the Northern general Philip Sheridan was found to be conducting operations against him from an inaccurate civilian map thirty years old. Hotchkiss’ map told Jackson at least plain essentials: where the gaps in the mountains were, distances between inhabited places, compass orientations, crossing points over the waterways, the course of paved routes. It was better than nothing and would serve him well. Positively bad maps of the Valley would lead his Northern enemies into serious error.

The Valley campaign of 1862 opened at a moment of strategic equilibrium between Union and Confederacy, after one Union offensive had been checked in the west, but before McClellan’s began on the coast, in the east. During 1861 the Confederacy had lost much territory west of the Appalachian chain, which marked the physical boundary between the two theatres of war. Most of the state of Missouri, largely Southern in sentiment, had been lost in August, despite a technical Confederate victory at Wilson’s Creek. Kentucky, also pro-South, was held for the North by a well-timed advance organised by the junior but aggressive General U. S. Grant. He would be encouraged by his success to embark on an advance into Tennessee, bringing the capture of the strategic river forts of Henry and Donelson, but then leading to the costly pitched battle of Shiloh in April 1862. The Confederates’ western front, consolidated by a new overall commander for the theatre, Albert Sidney Johnston (who died at Shiloh), would be held largely intact for the rest of the year.

In the east, where the Union had begun a campaign, progressively crippling to the South, to secure control of the Confederacy’s coastline, little dry land had changed hands in 1861. Following the Confederate defensive victory at Bull Run, Joseph E. Johnston’s Army of Northern Virginia had remained close to Washington, threatening the Federal capital. Its presence caused constant anxiety to President Lincoln, particularly because its size was consistently exaggerated by his new General-in-Chief, George McClellan. In March 1862 Joseph E. Johnston withdrew it south of the Rappahannock, one of the west–east waterlines that defended Richmond. That move somewhat relieved Lincoln’s concern for the security of his capital; but it objectively complicated McClellan’s plan to take the Confederacy’s by his seaborne invasion, since it put the South’s largest army closer to his ultimate objective.

On a large-scale map—paradoxically, in mapmaking, the larger the scale, the less the detail shown; one mile to one inch, small scale, is much more informative than ten miles to one inch, large-scale, though the latter is the more useful for strategic planning—the situation in March 1862 would have looked thus: Joseph E. Johnston, with 40,000 men in the Army of Northern Virginia, stood on the Rappahannock, forty miles north of Richmond; McClellan, with 155,000 men in the Army of the Potomac, was sailing it down that river to land at Fortress Monroe at the tip of the Virginian Peninsula, sixty miles from Richmond; various Northern detachments, under the command of Nathaniel Banks, amounting to some 20,000, protected Washington. In the Appalachian Mountains to the west, other Union generals deployed detachments of various strength. Implanted in the middle of the theatre, confronting but also threatened by the Union forces in the mountains and around Washington, Stonewall Jackson deployed fewer than 5,000 men to protect Joseph E. Johnston’s flank, to hold the Federals in the mountains at bay and to deter Banks from bringing the Northern defenders of Washington down to assist McClellan in his seaborne advance on Richmond.¹⁸

In unbroken country—the flat, unforested, unwatered terrain of the Great Plains, say—Jackson’s position would have been untenable. He would have been swept up during a few days of fighting in a concentric advance by Banks and the Northerners to the west. Jackson, however, was not in that vulnerable position. He had the mountains and rivers of the Shenandoah Valley on his side and, by employing the accidents of geography, natural and man-made, to his advantage, might overcome the odds confronting him. In the months of March, April, May and June 1862, he defied every probability in the most brilliant exercise in manoeuvre warfare, depending wholly upon superior use of intelligence, in the broadest sense, perhaps ever achieved.

The Valley Army (formally the Army of the Shenandoah Valley District) began its virtuoso campaign of diversion at the head of the Shenandoah Valley, where it had spent a hard winter near Romney, Jackson’s boyhood home. His orders were to avoid pitched battle but to operate in such a manner as to prevent Banks, outside Washington, from reinforcing McClellan as he advanced on Richmond. As events unfolded, he was to fight several pitched battles but nevertheless achieve the spirit of his instructions.

Though tied to Washington, Banks was also under orders to clear the northern end of the Valley and in late February he crossed the Potomac River where it joins the Shenandoah at Harper’s Ferry, then he advanced south. His purpose was to protect the two strategic lines of communication, the Chesapeake and Ohio Canal (connecting the sea to the Ohio River system beyond the Appalachians) and the Baltimore and Ohio Railroad (a principal rail route westward through the Appalachian Mountains), from Confederate interference. Jackson at first proposed attacking his advance guards at Winchester, where a railroad spur terminated at the Valley Turnpike, believing that he could inflict a defeat on the Union forces while they remained dispersed. The plan, however, defied Joseph E. Johnston’s order to decline action; while Johnston was withdrawing his army from Manassas to protect Richmond, he was particularly anxious not to risk a defeat anywhere that would allow Banks to bring his army to reinforce McClellan’s. Jackson’s plan also frightened his subordinates, who were sure they would be beaten. After a heated debate in a council of war, his first, on the evening of 11 March, Jackson gave up the argument. As he rode away into the darkness, he burst out to Dr. McGuire, his chief medical officer, “That is the last council of war I will ever hold.”

He was to be as good as his word; indeed, better. It is a military catchphrase that “Councils of War never fight”—the phrase was to be President Theodore Roosevelt’s but the idea is as old as antiquity—and, after the timidity shown by his brigadiers at Winchester, Jackson withdrew into himself.¹⁹ Famously taciturn even in his cadet days at West Point and much more given to private prayer than conversation, he henceforth kept his thoughts to himself, revealing his intentions only at the last moment and then in peremptory, often cryptic orders. That was not a deliberate security measure, more a reflection of his introverted nature; but it had the highly desirable effect, in what was to be a campaign of repeated surprises, of shrouding the unexpected in silence.

Between 11 and 20 March, the Valley Army retreated southward down the Valley Turnpike, covered by the cavalry force under Turner Ashby. Ashby was a born cavalier, untrained in formal cavalry tactics but a horseman to his fingertips and a dasher and doer. At times during the campaign, his and his troopers’ lack of discipline would infuriate the professional Jackson, but his relentless aggressiveness always restored him to his general’s favour. Meanwhile, as the retreat lengthened, Jackson was pondering his strategy. “Mobility was the essential factor in the Valley Army’s future.”²⁰ The army could manoeuvre successfully in the face of a superior enemy, however, only if it made correct use of the Valley’s geography, forced the enemy to make mistakes and denied Banks the use of essential links in the communication chain. It was a crucial factor in Jackson’s calculations to know that his opponent was not a professional soldier, indeed not a soldier at all; a leading type of the Civil War “political” general, appointed for party reasons, Banks had been a Congressman, Speaker of the House of Representatives and most recently Governor of Massachusetts. Jackson’s calculations essentially turned, nevertheless, on objective, not subjective factors: roads, bridges, rivers, hills. Now that Banks was inside the Valley, he had to keep him there, but without fighting battles he might lose. He also had to keep at a safe distance from the Union forces to the west, in the Allegheny Mountains. Finally, he had to keep open his line of withdrawal eastward towards Richmond, should Joseph E. Johnston send for him to assist in the defence of the city against McClellan’s army in the Peninsula.

His first thought was of bridges: those to be denied to the enemy, those essential to his army’s ability to manoeuvre. There were many in the Valley, most wooden and easily combustible, but some of critical importance. Two were railroad bridges, one over the South River at the southern end of the Valley, which Jackson needed if he were to escape by rail to Richmond, and one at Front Royal on the Manassas Gap Rail Road, a main line in the Northern supply chain. It had already been burnt by Jackson’s headquarters guard, and he had sent the rolling stock beyond it south to prevent Banks from using the wagons in a subsequent advance.

Of the road bridges, the headquarters guard had also burnt the one at Front Royal, to impede Banks’ advance down the Luray Valley, east of the Massanutten Mountain into which the North Fork flowed. The three bridges at Luray were essential to Jackson, however, were he to decide to slip across the central mountains through the Massanutten gap, and he also needed to preserve the spans at Port Republic and Conrad’s Store, both crossing the South Fork or its tributary, which carried roads leading through the Blue Ridge gaps and so to Richmond. Finally, there was a wooden bridge at Rude’s Hill, where the Valley Turnpike crossed the North Fork, which was perhaps the most important of all. If destroyed, with Banks to the north and Jackson to the south, its loss would stop a Northern advance dead at that point. Equally, its destruction behind Jackson’s back would terminate his chance of opening a counter-offensive up the Valley west of the Massanutten.

A dispassionate observer, taking his stance in mid-March 1862 at Staunton, Jackson’s main base at the extreme south of the Valley, would have assessed the situation thus: Banks, having failed to follow up Jackson’s retreat from Winchester with energy, was stuck between that place and Strasburg but retained the option of moving down either the North or South Forks; the latter manoeuvre would require bridging at Front Royal but that was within his army’s capability. Jackson, at Mount Jackson on the North Fork, had two choices: he could reverse his retreat and move up the Turnpike to find and fight Banks near Winchester; or he could cross through the Massanutten Gap to enter the Luray Valley and open a new offensive front.

The second choice, however, would take the Valley Army off the macadamised Turnpike onto dirt roads, limit its mobility and expose the main base at Staunton to Federal attack. Jackson therefore decided, even though he thereby kept himself further from contact with Johnston at Richmond and nearer to the remaining Federal forces in the Alleghenies, to retrace his steps and bring Banks to battle at Winchester. Moreover, he was encouraged to reverse his course by Johnston, who, retreating towards the Richmond river lines from Manassas, now expressed the anxiety that Jackson had got too far away from Banks. “Would not your presence with your troops nearer Winchester prevent the enemy from diminishing his force there? . . . I think it important to keep that army in the Valley, and that it should not reinforce McClellan. Do try and prevent it by getting and keeping as near as prudence will permit.”²¹

He had implicitly not encouraged Jackson to seek battle, but Jackson was not prudent when he scented the chance of a successful fight. On receipt of Johnston’s despatch, he immediately turned north again, marched through unseasonal snow on 22 March and, on the 23rd, found contact with Banks’ advance guard at the village of Kernstown, five miles short of Winchester.

Ashby’s cavalry opened the engagement, skirmishing forward during the morning with infantry in support. As the Union troops opposite began to form a line of battle, he fell back, to meet Jackson bringing up the main body. Ashby may have sent word to Jackson that he was opposed by only four regiments; alternatively, the intelligence may have come from local spies. In either case, Jackson was misinformed. The Federals were in much greater number, about 10,000 to Jackson’s 4,000, and with plentiful artillery, which, quickly brought into action from well-chosen positions, began to cause casualties.

Despite his inferiority in strength—and despite the day being a Sunday, on which the pious Jackson always sought to avoid fighting—he decided in the early afternoon to attack. The Northerners were deployed on both sides of the Turnpike but in greater strength to the west, where ridges and hillocks gave commanding views. It was there that Jackson made his effort. To assist him in directing the battle he summoned an officer of the 2nd Virginia, Major Frank Jones, “who knew the countryside: he could look across the Pike and see his front porch.”²² Local knowledge would not on this occasion, however, get Jackson out of a spot. He was about to bite off more than he could chew. Worse, his temperamental taciturnity added to the difficulty of the situation. He issued an unclear command and then lost control of events by leaving his central position to gallop about, trying to restore order. His leading brigade lost direction, came under heavy artillery fire, took cover and then fell back. Jackson brought up guns of his own—of which he had nearly as many as the enemy—and infantry reinforcements, but after a final and bitter exchange of volleys at short range, his men were beaten; many had run out of ammunition. Jackson himself wrote a few days later, “I do not recollect of ever having heard such a roar of musketry”; but the Federal fire was the heavier and at about six o’clock in the evening the Valley Army began to slip away and retreat down the Turnpike.²³

The Battle of Kernstown was a Confederate defeat. Southern losses were 455 killed and wounded, 263 taken prisoner; Union losses were 568 killed and wounded. Proportionately, the Valley Army had come off much the worse. On the other hand, the strategic effect was to its advantage. Even though the enemy had advanced when attacked, they formed only part of Banks’ army; another division had already left to join McClellan at Richmond, and Banks had gone to Washington. McClellan himself ordered Banks, who returned from Washington posthaste on the Kernstown news, “Push Jackson hard and drive him well beyond Strasburg.” He amplified his instructions on 1 April, emphasising that the Kernstown battle had forced a change of plan, requiring Banks to stay in the Valley instead of leaving it and, once the railroad was repaired, to advance to Staunton, Jackson’s main base, at the bottom of the Valley, so as to force “the rebels to concentrate on you and then [you to] return to me.”²⁴

What he did not do was to offer Banks more troops. Lincoln’s anxiety to protect Washington, the pull on his resources exerted by operations in and west of the Alleghenies, all combined to reduce his striking power against Richmond. Given McClellan’s specific orders to Banks to advance down the Valley Turnpike, west of the dividing barrier of the Massanutten Mountain, he thereby spared Jackson the anxiety that he might have to defend the Luray Valley to the east of the Massanutten also. Indeed, once he became aware of the pattern of Northern deployment, Jackson recognised that the opportunity was opened to use the Luray as an avenue for a counteroffensive of his own. He was to take full advantage. Although he was to spend the rest of March and much of April falling back west of the Massanutten, he was already contemplating countermeasures which would take him up the corridor to the east, where he could reopen attacks towards Harper’s Ferry and Manassas—and so heighten Lincoln and McClellan’s anxieties.

Before he would be free to act in that way, however, there was to be much action at the south of the Valley. Jackson, following his retreat from Kernstown, had brought the Valley Army into defensive positions near Mount Jackson, on the North Fork of the Shenandoah, where he reorganised. Banks, following slowly, occupied Woodstock. The actual outpost line between the two armies, from 3 to 17 April, was along a minor stream called Stony Creek. The two sides skirmished across it during two weeks of inactivity, Jackson content to keep Banks in play, Banks hesitating to advance lest Jackson slip through the Massanutten Gap to Luray and strike at his line of communications higher up the Shenandoah. Eventually, however, Banks perceived—with a rare flash of inspiration—that if Jackson could make geography work his way, it could be made to work for him also. He saw that, given the very small distance involved, he might, by a brisk advance down the Turnpike, drive Jackson past New Market, the entrance to the Massanutten Gap, and harry him on south to Harrisonburg or even Staunton. At dawn on 17 April, Union infantry launched a surprise attack, cavalry following. The Confederate defences were driven in, and when Ashby’s troopers tried to stop the Northern advance by burning the bridge at Rude’s Hill, where the North Fork runs in an impassable trench, the Union cavalry were upon them quickly enough to put the blaze out. The Valley Army, outnumbered nearly two to one, had no option but to leg it south as quick as it could go. Two days of forced marching took it out of reach of the pursuit; but, following Kernstown, Jackson knew that he had suffered a local reverse.

Strategically, however, he was still in the ascendant. Joseph E. Johnston, increasingly hard pressed by McClellan near Richmond, had actually sent orders for him to be ready to leave the Valley; his new quarters, in Swift Run Gap, one of the key passes through the Blue Ridge, positioned him to do so. Jackson, however, became increasingly persuaded as April drew out that he could protect Richmond better by staying where he was and using Swift Run as a secure base—the high ground on two sides protected him against surprise attack—from which to strike at Union forces in the vicinity. He calculated that they numbered 160,000 altogether, spread out across eastern, northern and western Virginia, and that most were successfully pinning down their Confederate opponents: McClellan had Joseph E. Johnston fixed at Richmond, McDowell was facing Anderson on the Rappahannock at Fredericksburg, Frémont, in the Alleghenies, menaced the small force of Edward Johnson. Jackson alone had freedom to manoeuvre, for, though the Federals now appeared to dominate the southern Valley, he was confident that he could outwit them in a mobile campaign. The question was whether Banks presented the most profitable target.

What tipped the decision eventually was growing evidence that Frémont was emerging from the Alleghenies to strike at Edward Johnson’s small and isolated force near Staunton, Jackson’s main base, crammed with war supplies and with produce from the Valley farms. To go to Johnson’s aid would require a march of fifty miles along bad roads and across the front of Banks’ army, still stationed near Harrisonburg, on the Valley Turnpike, after its advance from victory at Kernstown the previous month. The risk was sustainable, however, for Banks lay behind the North River, the bridges over which had been burnt on Jackson’s orders to cover his retreat to Swift Run Gap. Hotchkiss was therefore sent to locate Edward Johnson’s exact position and to reconnoitre a route towards him. On 30 April the Valley Army set out.

It would have reassured Jackson had he known that Banks believed the Valley Army was already leaving Swift Run Gap to go to Richmond. His mind, however, was set on his course, so much so that when torrential rain—“great sluices of water running along the road for hundreds of yards”—blocked the route Hotchkiss had chosen, Jackson turned his column about, marched it back into Brown’s Gap, gave his men a night’s rest and then started them west again along a more southerly route. It had the advantage of running parallel to the Virginian Central Railroad, onto which Jackson loaded his sick and stores. Piecemeal by rail and road the Valley Army concentrated at Staunton on 6 May, left the next day to join forces with Edward Johnson, who was marching to meet it, and then pressed westward towards a tiny place called McDowell (also, confusingly, the name of the Union general commanding on the Rappahannock north of Richmond).

During the afternoon of 8 May skirmishers from the two sides found each other and a battle began to develop. Jackson had reconnoitred the heavily broken ground and formed a plan to fall unawares on the Northern force, a detachment of Frémont’s army commanded by General R. H. Milroy. Milroy, however, had got wind of his approach and, though outnumbered, moved to the attack. In the confused fighting that followed, his men inflicted the heavier toll of casualties. Jackson reported to Lee “God blessed our arms with victory,” and in the sense that Milroy broke off the action, and retreated, the Confederates were the winners.²⁵ It was a costly victory, nonetheless, and Jackson later reproached himself for bad management of the battle. It was the last mistake he would make in the Valley campaign.

Its pace was about to quicken. Lee, in Richmond, was increasingly concerned to keep the Union forces surrounding the Southern capital separated; so was Joseph E. Johnston, and both counted on Jackson to operate in a way that would pin Banks west of the Blue Ridge and keep Frémont in the Alleghenies. After the battle at McDowell, therefore, Jackson decided that he must pursue Milroy, meanwhile taking steps to impede Frémont’s ability to manoeuvre. He sent Hotchkiss, with a scratch force of cavalry, to block the routes from the Alleghenies into the southern Shenandoah, while himself following up Milroy’s retreat. By 12 May he had got as far as the small town of Franklin, deep in the mountains, but had not caught up. He decided accordingly to break off the pursuit and return to the Valley. His purpose as before was to keep Banks from leaving, but he also intended to rejoin his subordinate, Ewell, and combine forces so as to confront the enemy in superior strength.

The Valley Army was now adapting to the extraordinary exertions Jackson expected of it. On 8 May, the day of the Battle of McDowell, the Stonewall Brigade had marched, from breaking camp to contact with the enemy, and then from leaving the battlefield to regaining camp, thirty-five miles. Such marches would, in the month that was to follow, become normal practice. Despite dreadful roads, shortage of food and deficient footwear—marching barefoot, often for dozens of miles, became a common experience—the Valley Army would rise to the challenge. Though Jackson concealed his intentions from even his closest subordinates, the Army came to understand during the month of May 1862 that his strategy was to mystify and mislead the enemy by achieving speeds over distance quite outside the capacities of normal infantry. They came to call themselves “Jackson’s foot cavalry” and, on many days, justified the title by marching for as long as horsemen could ride.

On 17 May, after a hard trek out of the Alleghenies, Jackson’s men re-entered the Valley near Harrisonburg, west of the Massanutten. Banks had been there the previous month, his army facing southwards along the North River, but had since departed to Strasburg at the northern end of the Valley, in preparation to move to Fredericksburg. He had already sent ahead Shields’ division. It remained, as before, Jackson’s duty to hold him where he was. In his favour was a shift in the balance of forces; the departure of Shields had left Banks with only 12,000 men; Jackson now had, either directly under command or readily to hand, about 16,000 if the division of Ewell, in the Luray Valley, was included. Also in his favour was the deteriorating quality of Northern intelligence—Banks was unsure of the Valley Army’s dispositions, and his information would get worse. By 21 May he was placing Jackson eight miles west of Harrisonburg, Ewell in the Swift Run Gap, forty miles apart, with the gap widening. In fact, by then, Jackson had transferred to the Luray Valley, via the Massanutten Gap, Ewell had joined him and the combined army was pressing northwards against a weak detachment of Union troops at Front Royal, guarding the Manassas Gap railroad bridges east of Strasburg.

The realignment had not been achieved without difficulty, even creative disobedience. Mid-May was an awful time for the Confederacy. During March and April, defeat had followed defeat all around its frontiers, in the far west, on the Atlantic coast. By early May the defensive line across the Peninsula had been abandoned, the Battle of Williamsburg outside Richmond had been lost and McClellan was laying siege to the defences of the city itself. Between 15 and 18 May, Robert E. Lee and Joseph E. Johnston, both a hundred miles away from the Valley in Richmond, and in touch at a delay of only two to three days, despite having the telegraph and a relay of fast despatch riders at their disposal, had sent a variety of conflicting orders, the impact of which was, nevertheless, to separate Ewell from Jackson and send him to watch McDowell at Fredericksburg. Neither Jackson nor Ewell wished to conform, since to do so would be to rob the Valley Army of its temporarily decisive superiority over Banks, without any guarantee that success could be won elsewhere by the separation. Covertly, they agreed to play on the ambiguity of the orders they were receiving and to use the delay in their transmission to stay together and march on Banks.

Jackson moved on 19 May. His bridge-burning at Harrisonburg, which had protected his sortie into the Alleghenies, now ought to have blocked his own recrossing of the North River into the Shenandoah Valley proper but Hotchkiss, effectively operating as his intelligence officer, discovered a number of large wagons that, positioned to straddle a ford, allowed passage even though the river was in flood. By 20 May, Jackson had reached New Market at the western end of the Massanutten Gap, by the 21st he had passed through the mountain to join Ewell at Luray and by the 23rd his vanguards were on the outskirts of Front Royal. By a forced march of seventy miles in three days, he had arrived in Banks’ rear and was ready to strike a decisive blow.

He then had a stroke of pure luck, though brought by the circumstance of fighting in friendly territory. Advancing to contact, but unaware of the strength of the Union defence at the Front Royal bridges, one of his officers was met by a breathless girl, Belle Boyd, a pretty eighteen-year-old who had just walked through the enemy camp, charmed an officer and discovered that only one Northern regiment was present. “Tell him [Stonewall],” she urged, “to charge right down and he will get them all.”²⁶ In the confused fighting that followed, most of the Northern infantry got away but the Confederate cavalry saved the bridges, which Jackson needed for the next stage of the operation against Banks at Strasburg, and destroyed the telegraph lines which would have warned Banks of the defeat.

On the evening of 23 May, Jackson pondered the situation that his rather ragged victory at Front Royal had won. He correctly concluded that Banks would feel exposed to a further Confederate attack in his position at Strasburg, where his numbers could be calculated to have fallen to about 10,000. He might fall back on Frémont, in the Alleghenies, but that was unlikely, since one of his duties was undoubtedly to protect Washington, which lay in the opposite direction. He might, improbably, go over to the offensive and attempt to recapture Front Royal, perhaps calculating that Jackson would set out northwards towards Harper’s Ferry, assuming that the Northerners were beating a retreat in that direction also; or he might simply do the obvious thing and retreat anyhow.

Eventually Jackson decided, correctly as it turned out, that Banks would go back towards Harper’s Ferry. He therefore ordered his army to follow the presumed line of Banks’ retreat, up the Valley Turnpike towards Winchester, by a converging route along the less good road leading through Cedarville and Ninevah. The distance each had to cover was about twenty miles, but while Banks was encumbered by a large wagon train, crammed with stores the Confederates coveted, Jackson was able to cover the countryside with a cloud of reconnoitring cavalry. On the morning of 24 May, the Confederate cavalry found Banks’ wagons, almost unprotected by fighting troops, jammed nose to tail in the Valley Turnpike at a point where it ran between stone walls. The Southerners brought guns up to fire into the mass, and Ashby’s cavalry charged in. Though the Federals set fire to as many of their wagons as they could, the Confederates captured a rich prize. Meanwhile, Jackson pressed the pursuit. On the evening of the same day, his army was arrayed outside Winchester, tired and footsore but prepared to give battle.

The appearance of the Valley Army outside Winchester, only twenty-five miles from Harper’s Ferry, only seventy from Washington, caused acute alarm in the Federal capital. It seemed to threaten a direct attack at worst, at least the need to dilute the campaign against Richmond. Lincoln, like Jackson, was studying a map—a less good map than Hotchkiss’—of the theatre.²⁷ Between four and five on the afternoon of 24 May, he ordered Frémont to abandon his plan to move west out of the Alleghenies against the rail centre of Knoxville, in Tennessee, and to march east to the relief of Banks. He also ordered McDowell, who was preparing to join McClellan in the Peninsula, to send half his army to the assistance of Banks as well. “At that moment, 5 p.m., May 24, the Valley Army won its Valley campaign.”²⁸

Jackson still, however, had much fighting to do, both up and down the Valley. On the morning of 25 May, in thick early mist, his advanced guard found Banks’ men positioned outside Winchester on hills that protected the town from the south. Jackson’s local intelligence had for once failed him. He thought the Union forces were behind, not ahead, of him, and he was expecting to cut them off from Harper’s Ferry. In the confusion that followed the initial encounter, brought on not by Jackson’s own brigades but by those of Ewell, marching to meet them, the Union troops at first inflicted heavy losses. Their batteries were well positioned on high ground. As the Confederate concentration grew, however, the Northerners found themselves outflanked to both left and right, their batteries brought under direct rifle fire and their infantry forced to fall back. Soon Banks’ men were in full retreat. They tried to make a stand in the streets of Winchester itself, but the townspeople, producing hidden weapons and shouting information to the advancing Southerners—many in the 5th Virginia came from the town in any case—undermined their resistance. By noon Banks’ army was streaming up the Valley Turnpike towards Harper’s Ferry with Jackson’s infantry—his “foot cavalry”—hot on their heels.

Had Jackson had his full force of horsemen under his hand at that moment, the destruction of his enemy might have been complete. Ashby, his cavalry leader, was elsewhere at the critical moment, on some cavalier venture of his own, the besetting fault of Southern riders. Banks, as a result, got clean away, managing to keep just ahead of Jackson’s vanguards until he reached Harper’s Ferry, where he crossed the Potomac on the night of the 25th, leaving the Valley in Confederate hands.

For how long? Lincoln, acutely alert to the dangers of the changed situation, and accurately reading it, was determined to prevent Jackson from disrupting the Union convergence on Richmond. He accepted General McDowell’s analysis: “Jackson will paralyse a large force with a very small one.” By correct disposition of his own forces, however, he hoped to crack the paralysis and re-establish the dominance that the North’s superiority in numbers ought to confer. Jackson’s advance to Harper’s Ferry appeared, on the map, to represent a threat to Washington. It could also be seen as an entrapment in a potential envelopment, and from three sides: by Frémont, advancing out of the Alleghenies to the west, by McDowell from the east and by Banks, if he recrossed the Potomac, from the north. The president sent the necessary orders to McDowell and to Banks on 29 May. To McDowell he wrote, “General Frémont’s force should, and probably will, be at or near Strasburg [on the upper North Fork of the Shenandoah] by 12 noon to-morrow. Try to have your force or the advance of it at Front Royal [on the two forks] as soon.” Lincoln, in short, was arranging a pincers behind Jackson’s back, which would cut him off from the Valley, and from Johnston’s army at Richmond, and expose him to defeat in isolation.

Jackson was not to be caught. Acutely sensitive to danger in any case, he was alerted to its correct reality by a succession of reports—that of a loyal Southerner, who had ridden from the Blue Ridge Mountains with word of a move by Frémont, then the transcript of the interrogation of a Northern prisoner who said Shields was marching on Front Royal, finally news of actual contact with Federals near Front Royal. By noon on 30 May, Jackson could no longer ignore the signs that his advanced position just short of Harper’s Ferry was overexposed and that prudence required a retreat into the Valley proper.

What followed might have been a rout. The Valley Army, rich with plunder, was encumbered with hundreds of wagons, some its own, some civilian, some captured from the enemy. They occupied eight miles of road. Military caution dictated that they should be abandoned, so that Jackson’s men could disengage as quickly as possible. Their commander was set on keeping his plunder, however, and counted on his soldiers’ ability to outmarch their pursuers to avoid entrapment. They also retained the capacity to deploy rapidly into battle formation off the line of march. On 1 June, as Frémont staged a thrust to cut the road, Jackson reversed the march of one of his brigades to drive the Union sally back. Some of his troops had marched as much as thirty-five miles in sixteen hours, snatching sleep on wet ground in wet blankets at intervals, but, with skilfully organised bursts of artillery, they succeeded in holding the enemy at bay. Jackson was in frequent conclave with Hotchkiss, who was reconnoitring energetically and measuring off relative distances on the map. He calculated, and so persuaded his commander, that the Valley Army could by quick marching just keep out of danger. On the afternoon of 1 June the Army was beyond Strasburg and still heading south, leaving the vanguards of the armies of Frémont and Banks closing hands on empty space.

During the next two weeks, Jackson would escape from a real or suspected trap several more times. As he headed south from Strasburg, just out of the enemy’s reach, his acute sense of danger alerted him to the makings of another. With Frémont, as he believed, hard on his heels and Shields advancing down the westward side of the Massanutten Mountain, he foresaw the two encircling him lower down. That was to overestimate Shields’ rate of advance; but, with the barrier of the Massanutten between them, his anxiety was understandable. His solution was to hurry a cavalry force ahead, with orders to pass through the Luray Gap and burn the surviving bridges across the Shenandoah at Luray, thus blocking Shields’ way southward.

Jackson’s own way south, towards New Market, was impeded by the constant harrying of Union cavalry and by appalling weather, which turned the surface of even the macadamised Valley Turnpike to glue. Men linked arms to keep their footing in the great press of traffic, swollen by the convoy of wagons which Jackson refused to abandon and by the complement of Union prisoners who, sensing how close their own side followed, dragged their feet and had to be bullied onward. The bridge at Rude’s Hill, which Ashby had failed to destroy in April, was burnt in the face of the enemy on 3 June.²⁹ The Valley Army was now running out of room to manoeuvre. Robert E. Lee, who had succeeded the wounded Joseph E. Johnston in command around Richmond, actually contemplated stripping his forces of troops to strengthen Jackson, with a view to his leading an invasion of the Northern states of Maryland and Pennsylvania, as would happen a year later. In the circumstances of 1862, however, such a démarche was impossible. Jackson, like it or not, was still bound to the retreat. His problem was to find the means to continue drawing Frémont and Shields after him, without becoming entangled in a costly battle, and then to disengage on favourable terms.

The trouble was that the earlier cunning of his bridge-burning was now telling against him. He needed routes to fall back on his main base at Staunton and towards the Virginia Central Railroad, which led from it towards Richmond: one was his point of resupply, but also the spot where he could disembarrass himself of his hundreds of wagons, liberating the army for a counterattack if necessary; the other was his line of escape. A key point was Conrad’s Store, to which a road ran from the Valley Turnpike, and a way through the Swift Run Gap (in the Blue Ridge) to the railroad. The necessary bridge had been burnt, however, and Jackson’s engineers advised him that the Shenandoah, swollen by the exceptional rains, could not be bridged with any safety.

The only way out, therefore, was that followed by the Valley Army the previous month, after its foray into the Alleghenies: a bad track leading to the village of Port Republic, short of Conrad’s Store, from which there was a way via Brown’s Gap to the railroad at Mechum River Station. It was risky. Shields, moving south from Luray, might catch the Valley Army in column of route and, in its exhausted state, defeat it. Hotchkiss was, however, acting energetically as Jackson’s eyes. From a lookout position at the southern tip of Massanutten Mountain he observed Shields encamp his army on the afternoon of 5 June near Conrad’s Store. Jackson, having done the distance a month earlier, reckoned that the Union force could not outpace him to Port Republic, now his touchstone of safety.

Not until 7 June, however, did Jackson bring his headquarters into Port Republic, after two days of desperate fighting which had left several regiments shattered and Ashby dead on the field. Union forces were pressing harder than he had anticipated. Pressure was shortly to bring on a battle that threatened to cut off his line of retreat. The Port Republic position was complex. Compressed between the southern tip of Massanutten Mountain and the Blue Ridge, it was also the junction of several key roads and the site of the confluence of three rivers, the South Fork of the Shenandoah, the North River and its tributary, the South River. A surviving bridge at the top of Main Street crossed at the junction of the North River and South Fork, while the South River was fordable at two points, Upper and Lower Fords. Jackson needed to dominate the whole scene of action in order to outface the enemy—Frémont, advancing from the northwest, Shields, advancing from the northeast—and still to preserve his options of retreat southwestwards.

His advantage in intelligence was played out. He was at close quarters with the enemy, who could read the situation map as well as he could. They dominated two sides of the battlefield, including the northern high ground. Unless he could fight them off, they might envelop him to left and right, cut off his line of escape and achieve the Union victory he had staved off for the last four months.

The exit strategy was to fight, and between 8 and 12 June, the Valley Army fought with terrible ferocity. The action of 8 June, conducted in the streets of Port Republic itself, was brought on by an uncharacteristic failure of attention on Jackson’s part. Tired himself, from days of marching, and with a tired army, he allowed the need for rest to overcome watchfulness. On what was hoped to be a peaceful Sunday morning, Federal cavalry got into Port Republic and surprised the sleepy Confederates. They were chased out with only difficulty and loss. Meanwhile, Frémont’s army, advancing above the South River in strong but unsynchronised support, was defeated at the village of Cross Keys.

Jackson might now have used the time he had won to break contact and retreat in haste to Brown’s Gap, the railroad and Richmond. That was what prudence dictated. Instead he decided to engage the enemy again, in the hope of inflicting a conclusive victory but at the risk of falling into a final trap.

The trap almost closed. The jaws were kept apart only by the harshest of fighting on two fronts, at right angles to each other. On the high ground north of the Shenandoah’s South Fork, Frémont was held at bay, while on the low ground between the river and the Blue Ridge the bulk of Jackson’s Valley Army, recovering from initial disorganisation, eventually formed a strong point and drove Shields back. In several hours of fighting in the early morning of 9 June, the heaviest of the whole Valley campaign, the Confederates eventually drove both Frémont’s and Shields’ men from the field, at a cost of over 800 killed and wounded to each side. A Confederate survivor later recorded, “I have never seen so many dead and wounded in the same limited space.”³⁰

The Valley Army was nevertheless the unquestionable victor of these culminating battles, so much so that Frémont and Shields did not merely leave it in possession of the battlefield, traditionally the mark of defeat accepted; each peremptorily withdrew northwards into the Valley, to positions from which neither could resume an offensive. After their defeat at Port Republic “[they] were terrified of the Valley Army. On June 19, with the rebels seventy miles away, Banks fretted that they were upon him.”³¹ In the face of such moral feebleness, Jackson actually re-entered the Valley, from his positions beyond the South Fork, and rested and refitted while the enemy retreated.

There was even in mid-June a revival of the suggestion that he should march northwards into Pennsylvania, opening an invasion of the North. Lee, now in overall command of the Army of Northern Virginia, organised reinforcements to send him and took no trouble to disguise the disposition from the enemy. It was, in truth, no more than a feint. The Confederacy, at a high crisis of its existence, was not really ready to take war to the enemy. It needed to assure the security of its own capital, not to menace that of its enemy. On 13 June Lee noted on a letter from Jackson, “I think the sooner Jackson can move this way [to Richmond], the better. The first object now is to defeat McClellan. The enemy in the Valley seem at a pause. We may strike them here before they are ready to move up the Valley. They are naturally cautious and we must be secret and quick.”³²

Lee, in short, had decided that he needed Jackson on the Virginia Peninsula, where the bulk of the Union Army was now deployed. On 16 June, therefore, he sent orders for Jackson to bring the Valley Army to the vicinity of the Confederate capital. Jackson departed on 18 June, riding ahead to meet Lee near Richmond on the afternoon of the 23rd. His army, following behind, crossed the Blue Ridge Mountains and arrived at Richmond shortly afterwards, to take part in the Seven Days Battles, which brought McClellan’s Peninsula Campaign to defeat. Richmond was saved and so, for a time, was the Confederacy.

In the second half of 1862 and during 1863 the Confederacy, under Lee’s generalship, went over to the offensive, which culminated in the Union’s flawed victory at Gettysburg. The progressive destruction of the South’s defences followed. What part had Jackson and his intelligence operations played in the postponement of the South’s defeat?

Jackson, first of all, had worked on the uncertainties and anxieties of his opponents, at every level from the field commanders to President Lincoln in Washington. He had threatened Frémont and Shields with defeat in open battle in the theatre of the Valley. He had also menaced Lincoln with the danger of an advance across the Potomac to Washington itself. Secondly, within the Valley theatre, he had confronted each of the Northern armies in turn, drawing them deeper into the Valley at the outset of the campaign, when the need to distract Union strength from the Peninsula was paramount, later risking battle when it could not be avoided but almost always on his own terms. He had tried his own army hard—his “foot cavalry” achieved, as its endurance increased, almost unparalleled feats of marching, at times covering as much as seventy miles in a hundred hours—but those who could stick the pace remained able to fight even at the end. Losses from disease and exhaustion ran as high as 30 per cent but battle casualties were surprisingly low, only about 2,000 in forty days of fighting. When it left the Valley, the army, which had admittedly been reinforced, was actually larger than when it had begun the campaign.

Jackson’s success was due in large measure to his ability—reinforced by his natural taciturnity and secretiveness—to think faster and more clearly than his opponents and to calculate more moves ahead, making good choices, rejecting bad. That ability rested, however, on his possession of superior knowledge of the Valley’s geography and of superior local intelligence, constantly refreshed by the work of a busy intelligence chief, Jedediah Hotchkiss, and a friendly population. The best generals have always valued detailed knowledge of topography, almost above any other sort of intelligence. Jackson was a better general than any of his opponents, and his operations in the Valley, assisted by McClellan’s refusal to profit by any of the advantages the North’s material superiority gave him, assured the successful defence of Richmond and the Union setbacks of 1862–63 which flowed from McClellan’s retreat. The proof of his generalship was demonstrated above all, however, by his exploitation of the secrets of place and passageway in the complexity of the Shenandoah Valley, which he possessed and the enemy did not. He deserved his triumph.