The Peninsular Campaign

McClellan’s plan for the Peninsula was straightforward: to put a superior army as close as possible to Richmond, the Confederate capital, and break his way through its defences by a combination of strategic outmanoeuvring and material outnumbering. He had originally intended to land at Urbanna on the Rappahannock, the next river northward from the York; when Joseph Johnston, his Confederate opponent, fell back, thus aborting the chance to land in his rear, he had fixed upon Old Point Comfort instead. The plan nonetheless remained the same in essentials and retained its virtue of simplicity: strike for the enemy’s capital.

There were to be three impediments to its achievement: a Southern secret weapon, a Confederate diversionary campaign, and McClellan’s own prevaricating and procrastinating character.

The South’s secret weapon was an ironclad ship, formerly the USS Merrimack, rebuilt as the CSS Virginia, whose appearance on 8 March 1862 threw all Northern schemes to attack Richmond by sea into disarray. The Merrimack, as history knows her, was a cut-down frigate fitted with an iron penthouse mounting ten guns. Its construction had consumed most of the output of the Tredegar Ironworks in Richmond, the South’s main, indeed almost only, foundry for several months. She was sailed as soon as she was ready and in her first day of action sank two conventional Union men o’ war and threatened three others with the same fate. That spelled doom also to McClellan’s gathering fleet of transports and the cancellation of the Peninsular campaign before it began. When Merrimack came out of Norfolk—the South’s naval base opposite Fortress Monroe but inside Chesapeake Bay—on the following day, 9 March, she seemed set to finish off the rest of the Federal fleet and then cruise at leisure in those inland waters, devastating any armada that McClellan or Lincoln sent southwards. That might well have been the outcome; but, by perhaps the most remarkable technological coincidence in the history of warfare, the North had the week before sailed an answer to Merrimack from Brooklyn Navy Yard and got her into Hampton Roads on the night of 8–9 March. Monitor was almost as makeshift as the Southern ironclad, an armoured turret on a steam-powered raft, but she matched her in fire power, fought her to a standstill on 9 March, and so secured Federal command of the Chesapeake for the duration of McClellan’s land campaign. It had been touch and go.

Lord Palmerston, when he had seen the first British ironclad, the sleek, sinister Warrior, lying at anchor beside Britain’s bluff old wooden ships of the line soon after her launching in 1860, had called her “a snake among the rabbits.” Merrimack and Monitorwere both too ungainly to be called snakes; the first was never to venture into the open sea, the second foundered soon after her famous first encounter of the ironclad era. While they were giving battle, however, a real master of the snake’s mesmeric technique with rabbits—his motto was “always mystify and mislead the enemy”—was preparing to attack the flank of the Union armies in the Shenandoah Valley.

Stonewall Jackson’s manoeuvres, which generations of American and British staff college students have subsequently studied for strategic inspiration, were to be an essay in the pure use of terrain for which there are few parallels in military history; campaigns that perhaps bear comparison are Wellington’s backing and filling between the Portuguese frontier passes in the years 1809–11 and Rommel’s “now you see me, now you don’t” dodging among his own minefields in the Western Desert in 1941–42. Jackson, godfearing, hypochondriac, professorial, resembled neither, except in dedication to his profession, and he had much less practical experience of war than either. He enjoyed, however, a cardinal advantage in confronting his Northern enemies: he knew the ground on which he was to fight intimately. His years on the faculty of the Virginia Military Institute in Lexington, which is actually inside the Shenandoah Valley, had familiarised him with its southern end; he was further aided, Professor McPherson points out, “by local spies and scouts, who knew every foot of the country”; and “he had spent many hours studying … maps drawn by his brilliant topographical engineer Jedediah Hotchkiss.”

The valley—for military purposes its length may be taken as some 120 miles—is a region of great, still unspoilt natural beauty which runs from the confluence of the Shenandoah with the Potomac at Harpers Ferry, scene of John Brown’s anti-slavery demonstration of 1859, southwestward into the greater Appalachian chain beyond Staunton. It has no natural outlet in that direction and lacked, therefore, any strategic value to the Northern armies. At its Potomac exit, however, it menaced Washington from a flank and with its “gaps” or passes through the Blue Ridge Mountains which form its eastern side, the Rockfish Gap, the Swift Run Gap, and the Manassas Gap, it offered sally ports through which a Confederate force could fall upon a Union army operating in the coastal lowlands between the Blue Ridge and Chesapeake Bay. Moreover, it is not a simple trench; behind the front range of the Blue Ridge but before the main mass of the Appalachians lying to the west run the intermediate Massanutten Mountains, dividing the North Fork of the Shenandoah from the South Fork, and again pierced by their own “gaps” and passes. The valley, in short, is a ready-made Tom Tiddler’s ground, presenting a resourceful general with the opportunity to play hide-and-seek, up one side of the Massanutten, down the other, through the gaps, and back again, for as long as his enemy would tolerate the game.

This is countryside I have known for nearly forty years, from my first visit to the theatrically beautiful University of Virginia at Charlottesville in 1957. It was my earliest encounter with the high civilisation of the South, and the impression it made has never faded. Here, in physical form, were the aristocratic argument for the slave system, the Palladian grace of Thomas Jefferson’s plantation at Monticello, the scholarly seclusion of its great library and that of the university itself, the serpentine walls of the mown university gardens, the wide-rolling, white-fenced pastures of the Albemarle County horse estates. It is a countryside of stifling noontide heat, but also of cool verandas, dark shadows, old shade trees, wealthy ease, discreet service, and the ever-ready offer of the evening mint julep. Close at hand are the magnificent vistas of the Skyline Drive, which follows the crest of the Blue Ridge, while below it rise and fall the green slopes of long-tended grazing farms in the Shenandoah Valley itself. For all its richness, this is still sparsely settled country. Not long ago, half a lifetme after I first came here as an undergraduate, I drove northward along Interstate 81 from Lexington, home town to the serene campus of Washington and Lee College and the spartan stronghold of the Virginia Military Institute—surely even in Stonewall’s time, the cadets must have enjoyed a little more ease of life than they do today?—and confounded myself with the thought that this lovely and tranquil land had in 1862 been the scene of breakneck marching and countermarching by Union and Confederates and of bitter battle at the little towns of Forest Royal, Winchester, Cross Keys, and Port Republic.

Stonewall Jackson’s problem is easy to describe, its solution almost as difficult to relate as it was to realise. With an inferior force—never many more than eighteen thousand—he had to outface two larger Union armies dedicated to doing battle with him, that of General Nathaniel Banks in the valley itself and General John Frémont’s across the mountains in West Virginia, and at the same time mount such a threat to Washington that he would deter President Lincoln, who was temporarily acting as commander-in-chief, from releasing a third army, McDowell’s, to join McClellan in a convergent attack on the Confederate capital of Richmond a hundred miles to his east. All this, moreover, though he had few cavalry and, as always with Confederate forces in the field, lacked transport, food, and every sort of military supply. His solution to his problem was starkly simple: to run his soldiers as harshly as he ran himself, to march them so fast from strategic point to point that they earned the name of “foot cavalry,” to live off the land, to drub the Union forces wherever found but, above all, to so “mystify and mislead” that the enemy always fought divided and never succeeded in uniting against him.

Thus, starting in the far south of the valley on 8 May, he won a small victory over Frémont’s army at the place confusingly called McDowell (today on U.S. 250) on the slopes of the Shenandoah Mountains. Then, dropping back into the valley proper, he force-marched his troops northward via Harrisonburg (today Interstate 81) to New Market, jumped through a gap in the Massanutten Mountains to Luray (on U.S. 340), and advanced to Front Royal. There he won a small victory on 23 May, so alarming the Union troops that Banks retreated first to Winchester (now in a loop of Interstate 81), where he was beaten again, and then all the way to Harpers Ferry.

The appearance of Confederates in fighting mood at a place only sixty miles up the Potomac from Washington alarmed Lincoln in exactly the way that Jackson—and Robert E. Lee, who was acting in effect as Confederate chief of staff—hoped that it would. Lincoln cancelled his plans to send McDowell forward from Fredericksburg to assist McClellan in the Peninsula and ordered him to turn against Stonewall instead. From Fredericksburg to the nearest of the valley “gaps” at Front Royal is only fifty miles, three, perhaps two days’ march. From Front Royal to Frémont’s current position on the other side of the valley was a good deal less than fifty miles. Jackson, at the moment McDowell turned, was forty miles to the north, which meant forty miles the wrong side of safety. A trap was being sprung in his rear, and every map table calculation suggested that he would be caught in it. Map table calculations did not make allowances for Jackson’s powers of manoeuvre when he was in a hurry. Turning like lightning, and using what cavalry he had as an advance guard to hold Frémont at a distance from his planned line of retreat, he passed through Winchester for the second time on 31 May, only a week after he had fought his battle there and only two days after he had been at Harpers Ferry. He had escaped the immediate trap, but Frémont was marching to pursue him down the west side of the Massanutten Mountains, and another Union force, under James Shields, was marching down the east side, so that he could not tarry. He pressed on to Harrisonburg at the foot of the Massanutten Mountains, where he left his subordinate Richard Ewell to fight Frémont at Cross Keys on 8 May. He could at this point have beaten an honourable retreat out of the Shenandoah Valley through the Blue Ridge Mountains via Brown’s Gap and made his way to Richmond to join Joseph Johnston, since he had now fully accomplished his mission of disrupting the Union concentration against the Confederate capital; but the spectacle of a divided Northern force—Frémont west of the Massanutten Mountains at Cross Keys, Shields east of the mountains at Port Republic—was too tempting to ignore. On 9 June, he arrived opposite Shields at Port Republic, defeated him in a five-hour battle, collected Ewell, organised his train of fifteen thousand troops, two thousand prisoners, and heavily loaded wagons of booty into a column of route, marched through Brown’s Gap out of the valley, and there shortly took the rails to Richmond.

It had been an astonishing month. Jackson’s army had marched some 350 miles between its starting and finishing points, fought five battles, all victories, kept three armies divided, menaced the Federal capital, thrown the enemy’s strategy into disarray, and got off scot-free, causing the commanders of seventy thousand Union soldiers to wonder how a general, whose force had never exceeded eighteen thousand, could so completely have mystified and misled them.

Jackson’s will-o’-the-wisp manoeuvres galled all the more because they cast so unfavourable a light on the pedestrian evolutions of his old fellow cadet of the West Point class of 1846, George B. McClellan, in the Peninsula. More galling still was the fact that McClellan had an opinion of himself, while Jackson appeared to have none. History would make of Stonewall a figure of romance: the nickname itself, won at the first Battle of Bull Run, was part of it, the valley blitzkrieg its substance, his tragic death—shot by one of his own sentries at Chancellorsville after his brilliant flank manoeuvre which decided the battles—would crown it; the poignancy of his dying words, “Let us cross over the river and rest under the shade of the trees,” would live as his epitaph. Their enigmatic beauty supplied Ernest Hemingway with the title for his least-understood novel. Yet, while Hemingway’s central figure was meant to be a romantic hero, Jackson was quite the opposite. Modest and self-effacing, he owes his romantic reputation to history, not to self-seeking.

McClellan was everything that Jackson was not, vain, vainglorious, opinionated, worldly, self-satisfied, ostentatiously busy—but also dilatory and self-doubting. He was a splendid organiser, on the principle of doing everything himself and delegating to nobody, but his gifts were for solving problems presented to him by unsatisfactory subordinates, not by active and contentious enemies. He was a great fault-finder—with his predecessor as general-in-chief, Winfield Scott (“a dotard”—Scott was seventy-six, McClellan thirty-five), with Gideon Welles, Secretary of the Navy (“a garrulous old woman”), with William Seward, Secretary of State (“incompetent little puppy”), with Edward Bates, the Attorney General (“an old fool”), and with President Lincoln (“a well-meaning baboon”). He found no fault with himself. If he were to be compared with other famous American generals, it could be said that he resembled MacArthur in his arrogance and George C. Marshall in his hauteur, but that he lacked, as events would prove, the former’s dynamism and the latter’s strength of character. He was a hollow man who, though events would find him out, refused to confront his own failure and preserved his self-esteem to the very end of a career which encompassed defeat in the field and defeat in electoral contest for the presidency as well.

Grant, a senior at West Point in McClellan’s freshman year, confessed in the aftermath of the war that “[he] is to me one of [its] mysteries.” So he might well have found him. Grant was a tanner’s son who managed to fail at the military academy, fail in the army, fail in business, fail in farming, and succeed only when war offered him the chance to deploy his gritty skills of decision, self-control, and selection of the enemy’s weak spot. McClellan had, until elevated to high command, succeeded at everything. His family was rich, he had graduated second in his cadet class, he had won golden opinions in the Mexican War. He had been sent as an official observer to the Crimea, a plum appointment, but had left the army to become vice-president of the Illinois Central Railroad, of which Lincoln was the attorney, and then president of another. When appointed to the second most senior post in the United States Army in May 1861, he accepted the title of “Young Napoleon” conferred on him by the newspapers as if it were his due. When in November, deferring to congressional pressure, Lincoln appointed him to the post both of general-in-chief, formerly held by Winfield Scott, and of commander of the Army of the Potomac, but with the warning that the dual responsibilities “will entail a vast labour for you,” McClellan complacently riposted, “I can do it all.”

He could certainly organise large-scale military movement. When he eventually persuaded Lincoln that the correct strategy for 1862 was to transfer the Army of the Potomac to the Peninsula, the movement went like clockwork. A fleet of 113 transports of every variety, including transatlantic liners and Hudson River pleasure craft, had been chartered to carry 121,000 men, 14,500 horses, 1,200 wagons, the guns of forty-four artillery batteries, and a vast quantity of stores to Old Point Comfort. The whole force was moved from Annapolis, Washington, and Alexandria in less than three weeks and found camps prepared for it inland of Fortress Monroe near the devastated town of Hampton behind the spit of land on which the great stronghold of Fortress Monroe stood. The value to the Union of the retention of the fort in 1861 now revealed itself. Since the next defensible position was at Yorktown, twenty miles up the river, where General Magruder had in fact planted the Confederate position, the intervening space was a sort of no-man’s-land in which McClellan could deploy his army at leisure before beginning the advance on Richmond. The great natural bastion of Quebec had offered Wolfe no such advantage.

McClellan, indeed, had almost every advantage, most of all a huge preponderance of numbers over the Confederates. He had been less than frank in his assurances to Lincoln that Washington was adequately protected; the President had given him leave to reposition the Army of the Potomac from the outskirts of Washington to the shores of the Chesapeake—potentially at Urbanna on the Rappahannock River, actually to the Peninsula—on the strict understanding that he should “leave said city [Washington] entirely secure.” McClellan had compiled figures to show that 73,500 men would remain. In fact, he had double-counted some men, and included the forces detached to operate around the Shenandoah Valley, so that the true figure was under 30,000. Lincoln’s staff—he was temporarily acting as commander-in-chief, having rightly judged that McClellan could not direct the Peninsular campaign as well, to the latter’s intense annoyance—consequently “agreed in opinion that the capital was not safe.” By then, 28 May, it was too late for second thoughts, because the Peninsular campaign was under way.

Fortunately, McClellan had grossly over-estimated Confederate strength. At the end of 1862 he was insisting that Joseph Johnston had 150,000 men around Manassas, when the true figure was nearer 40,000. It says something about McClellan’s egotism that he nevertheless persisted in arguing for stripping Washington of troops and striking at Richmond via the Chesapeake as the current strategy. Lincoln had seen things the other way: that Johnston might respond to an offensive against Richmond not by falling back to defend it but by mounting his own offensive against the unprotected Washington. Had he had the numbers, Johnston might well have done so. As he actually commanded less than a third of McClellan’s estimate—which had been prepared by the Pinkerton detective agency—it was he who felt overexposed at Manassas, and it was his consequent decision in March to fall back to the more secure line of the Rappahannock and Rapidan rivers, fifty instead of twenty-five miles south of the capital, that objectively robbed McClellan’s Chesapeake venture of risk.

Napoleon had said generals should be lucky. McClellan was extraordinarily lucky right up to the opening of the campaign: lucky that Johnston did not have the strength the intelligence reports alleged, lucky that the Monitor turned up in the nick of time to rescue the fleet off Fortress Monroe from destruction by the Merrimack, lucky that Jackson’s brilliant manoeuvres in the Shenandoah Valley did not panic Lincoln into diverting more of his troops in that direction, lucky even that Johnston withdrew behind the Rappannock when he did, for, though that put the Confederate main army closer to Richmond, it scuppered the Urbanna plan, which, involving as it did the need to cross two, if not three, rivers by a sodden and inadequate road network, would almost certainly have left him floundering.

Then McClellan proceeded to dissipate all the luck he had been given. Instead of proceeding up the Peninsula as soon as his advance guard was ready to march in the first week of April, he began to see difficulties. He had already been warned by the navy that the presence of the Merrimack at the mouth of the James River denied that water route to him as a means of shipping men and guns forward towards Richmond; he himself decided that he could not break into the York River past Gloucester unless he diverted troops to attack the Confederate batteries there, which he was unwilling to do because it broke the principle of keeping his army concentrated. When on 5 April he discovered that General John Magruder was not merely encamped at Yorktown, as he believed, but had thrown a line of defences across the whole Peninsula from the York to the James, he took fright. He began at once to over-estimate the force opposed to him—by 7 April he was informing Washington that he would soon be faced by 100,000 men, when the real number was about 30,000—and to exaggerate the natural strengths of the Confederate position, the badness of the roads leading to it, and the width of the Warwick River behind which some of Magruder’s entrenchments ran. Magruder had only 3,000 men per mile to defend ten miles of front, far too few to oppose a determined assault at a particular point, but exaggerated his numbers by marching units up and down and staging demonstrations by fire. McClellan took the display for reality, declared that he was opposed by a major force, and announced that he must sit down to conduct a siege. By 7 April he had learnt that Lincoln had diverted troops—McDowell’s corps—towards Stonewall Jackson in the Shenandoah Valley; he took that as a reason, while sending forward his own heavy guns along the bad road from Fortress Monroe, to demand both more guns and more troops, pleading that “my force is possibly less than that of the enemy.”

Lincoln judiciously replied with the reminder that he himself had “always insisted, that going down [Chesapeake] Bay in search of a [battlefield], instead of fighting at or near Manassas, was only shifting, and not surmounting, a difficulty—that we would find the same enemy, and the same, or equal entrenchments, at either place.” This gentle reproof stung McClellan not at all. An expert in sieges as he conceived himself to be—he was a siege engineer by training and had been a spectator at the Anglo-French siege of Sebastopol nine years earlier—he settled down to a siege of the old Yorktown position as if he had all the time in the world. Cornwallis had succeeded in defending his, admittedly shorter, lines at Yorktown for only three weeks in 1781. The Confederates were to hold out for a month and then get clean away, the reason for that being a strange inflexibility that settled upon McClellan: having decided upon siege as the means to carry the Yorktown line, it seems as if he thereafter dismissed all others. His programme envisaged the assembly of over a hundred heavy guns including 200-pounder cannon and 13-inch mortars, together capable of delivering seven thousand pounds of shot and shell in a single discharge, and he settled himself to await that moment. When on 16 April his subordinate William F. Smith captured a portion of the enemy lines at Lee’s Mills, McClellan declined to exploit the success.

His mentality, and therefore the likely development of the campaign, had been discussed two days before by his opponents in a remarkable council of war at Richmond, at which President Jefferson Davis and Generals Robert E. Lee, Joseph E. Johnston, and James Longstreet, all of whom had known him in peacetime days, had pondered strategy. The sharpest judgement was made by Longstreet, who predicted that the cautious McClellan would not attack before 1 May. So it was to turn out, almost exactly. McClellan, after a month of demanding more men and guns and meticulously positioning the very many he had, finally set the date for his great assault at 5 May; but two days earlier, when he sent one of his balloons aloft to observe the enemy’s positions, they were found to be empty. During the previous night, Johnston had evacuated his trenches, hitched in his guns, and set off up the Peninsula’s bad roads for Richmond forty miles away.

Moving sometimes as slowly as a mile an hour—but the Confederates were impeded as much—McClellan’s leading division caught up with Johnston’s rearguards where the Peninsula’s two main roads joined just short of Williamsburg. With forethought, Johnston had prepared there a secondary position between Cub Creek and College Creek, tributaries respectively of the York and the James. It held well during a day of fighting on 5 May, in which the Union army suffered twice the Confederate casualties, so well that Johnston was able to slip away, leaving his earthworks empty. This, counting Manassas in March as well as Yorktown, was the third time he had disengaged without detection, and his manoeuvre succeeded as well as it had before. McClellan did not press a pursuit but proceeded with the methodical advancement of his superior force, part by road up the northern bank of the Chickahominy River, which bisects the upper Peninsula before flowing into the James above Williamsburg, part by river up the York, which Johnston’s simultaneous evacuation of Gloucester and Yorktown had now opened. McClellan moved first to West Point, the objective of his aborted Urbanna plan, and then up the Pamunkey River, one of the York’s feeder streams, to White House, almost due east of Richmond and only twenty miles from it; but the advance had taken another fifteen days, during which Johnston had succeeded in further strengthening the earthwork defences around Richmond and finding more troops to man them. By the last week of May his numbers had reached 60,000, to oppose the 105,000 McClellan had concentrated on the far bank of the Chickahominy.

The Confederates had also had another success, which was definitively to close the James River as an avenue of approach to the Confederate capital. The abandonment of Yorktown had meant not only that of Gloucester but also that of Norfolk, the great naval base opposite Fortress Monroe, and it entailed as a result the withdrawal of the Confederate fleet. Merrimack had proved too cranky to make the voyage; the world’s first fighting ironclad had had to be scuttled and burnt off the Elizabeth River, which leads into the wilderness of the Dismal Swamp. The other ships, including an unfinished ironclad, were got up as far as Richmond, and the James behind them was then blocked with scuttled ships and felled trees. The real obstruction, however, was Fort Drewry, an earthwork built on a bluff the previous year, from which eight heavy guns commanded a mile of the James southwards from modern Bensley. The area today is scruffy exurbia, neither quite town nor country, sandwiched between Interstate 95 and the river, and patchily wooded. The fort—known to the Union as Fort Darling—still stands, however, thanks to the National Park Service, and from its parapet the strength of its position is instantly recognisable. Throughout the Civil War it served as the Confederacy’s naval academy, and a miserable time the cadets must have had, crammed into its cramped surroundings; today it seems to be favoured as a weekend trysting place for middle-aged couples, clearly more interested in each other than in the Park Service’s recorded descriptions of the drama that unfolded at its fort on 15 May 1862.

Early that morning five Union warships, including the redoubtable Monitor, got up to within six hundred yards of the bluff, which stands 110 feet above the water, and one, the ironclad Galena, anchored broadside-on downstream to engage the fort with all its guns. It was not to prove a fair fight. In the age of smoothbore cannon, ships had been at a disadvantage in gun duels with land batteries, but the appearance of rifled ordnance had changed that. Deprived, however, of the ability to manoeuvre, as the Union ships were by the narrowness of the river, and dominated by the height of Drewrys Bluff, the flotilla found the odds reversed. Galena was hit forty-four times in the three-hour cannonade and disabled; Monitor, which had accompanied her, could not elevate her guns sufficiently to engage at all; the ironclad Naugacket burst a gun; the unarmoured gunboats dared not come within range. As the Northern squadron dropped back defeated, a Confederate shouted down at Monitor, “Tell the Captain that is not the way to Richmond.”

Richmond lies only five miles north of Drewrys Bluff; a Union breakthrough past the fort up the James would have brought riverborne guns to bear not only on the city’s heart but also on the freight yards and rail junction of five lines, as well as the Tredegar Ironworks, the Confederacy’s chief source of cast and forged metal. It had been a narrow squeak. The reverse for McClellan was decisive, none the less; his way to Richmond now lay from only one direction: across the Chickahominy River, which guards the city’s eastern and northern approaches. He had arrived on the far bank of the river on 20 May and lined his army up on a front between the little town of Mechanicsville and the Chickahominy crossing place of Bottom’s Bridge nine miles to the south. Richmond was only six miles west of his most advanced positions and though protected by earthworks—those at Chickahominy Bluff just off U.S. 360 remain almost in their original state—lay open to a swift attack. Yet again over-caution assailed him. The Confederates, expecting battle in the open field, battle at the gates of their capital which might decide the issue of the Civil War once and for all, were concentrating troops by all means and at utmost speed. McClellan, the siege engineer, had determined another course. The sight of Magruder’s earthworks at Yorktown had decided him that he must conduct a siege there. Johnston’s earthworks on the Chickahominy brought him to the same conclusion. While Johnston gathered reinforcements, and Jackson’s sleight-of-hand in the Shenandoah Valley continued to deprive McClellan of McDowell’s forty thousand men, the Union engineers east of Richmond were turning White House into a great logistic base and rebuilding bridges, some four hundred yards long, to bring the battering train of 200-pounders and 13-inch mortars across country from the York River’s head of navigation. The work would fritter away over a month of the high campaigning season.

In the interval, the Confederates did not remain idle. While Lincoln urged McClellan to action and McClellan renewed demands for McDowell’s corps to be sent to him—though McDowell was, in this last week of May, marching urgently to protect the Potomac crossings against Jackson’s lunge up the Shenandoah Valley—Joseph Johnston was contemplating how he might strike a blow that would set back what he imagined to be McClellan’s timescale. The most advanced of the Union troops could see Richmond’s church spires and hear the church bells chiming the hours; Johnston presumed it could not be many days before he was forced to give battle at a grave disadvantage in numbers. When news—wrong in detail, right in substance—reached him on 30 May that McDowell had turned away to chase Stonewall Jackson, Johnston therefore decided—with a strong prod from Jefferson Davis, fretting in the Richmond White House—to fight a spoiling action. The heavy rains that had dogged all movement for the second month of the Peninsular campaign had swollen the Chickahominy, isolating one of McClellan’s corps on the wrong side of it. It—Keyes’ corps—was to be the target.

Even these short Virginian rivers can turn nasty in spate. I had the luck to see the Rappahannock after a single day of heavy rain in September 1992 and have not forgotten the sight: a mass of brown water, boiling yellow over the boulders that choke its bed at Fredericksburg, was rushing to the Chesapeake, tossing broken trees about and lapping the woodlands that edge its banks. The Chickahominy is a slighter river, but a month of rain had washed away the bridges that crossed it, leaving Keyes’s corps with a real obstacle at its back. The Confederate troops that appeared to oppose it on 31 May, moreover, in a battle the North would call Seven Pines and the South, Fair Oaks, had a clear advantage of numbers—51,000 to 33,000. All was set for a victory. The outcome was lamentably different. Johnston was a good general—he had won the first battle of the war at Bull Run and would be the last Confederate commander to surrender in April 1865—but he failed to keep a grip of events and was let down by subordinates who misunderstood his orders. Perhaps his orders were not clear; he operated the Prussian principle of outlining his intentions and leaving his subordinates to execute them as they saw best. In an army like the Prussian, whose senior officers had trained together at the war academy, the system worked well; but the training of the Confederacy’s officers was uneven, sometimes wholly deficient, and too often negated by individual temperament. James Longstreet, Johnston’s principal subordinate, was notably erratic and temperamental. On 31 May at Seven Pines he misheard or misunderstood his orders, took the wrong road, decided to bridge a swollen stream he might have forded, wasted time so doing, fell into argument with a fellow divisional commander over precedence in using the bridge, and thus succeeded in confronting the Northern troops not with a concentration of force but with a succession of piecemeal attacks. Seven Pines should have been a Confederate victory. It turned into a sort of Union success, its main outcome being a change of command, for in his effort to retrieve the situation by personal intervention, Johnston strayed into the fire zone at Fair Dales, astride the railway line McClellan was using to bring his siege train forward, took a bold stance on a hillock, and was knocked off his horse by shell splinters. He was too badly wounded to continue serving Jefferson Davis as director of operations and was replaced at once by Robert E. Lee.

Lee, today the romantic hero of the Civil War to Northerners and Southerners alike, still had his reputation to make. Indeed, in the summer of 1862 his name did not stand high in the South. His campaign in West Virginia in 1861 had failed so badly that that homeland of crotchety backwoodsmen and squirrel hunters, defiant fighters for liberty in the War of the Revolution but proud of the poverty that kept them from owning slaves, proclaimed reverse secession and went over to the Union to found a new state. Richmond newspapers had then called him “Granny Lee” and “Evacuating Lee”: now McClellan, hearing news of his appointment, characterised him as “cautious and weak. Under grave responsibility … likely to be timid and irresolute in action.” A finger should not be pointed in retrospect; but during the next month—the third month since the Peninsular campaign had been opened—it was to McClellan that both North and South looked for action, and of action he gave no sign at all. Everything was against him. The roads were sodden; the bridges were down; the railroads needed rebuilding; the battering train of siege guns was not in place; the Seven Pines battle had disorganised his army; McDowell was still not with him (though one of his three divisions arrived during June); above all, he was dangerously outnumbered. On 25 June he telegraphed Washington to say that the “rebel force” stood at “200,000 …. I shall have to contend against vastly superior odds … If [I am] destroyed by overwhelming numbers … the responsibility cannot be thrown on my shoulders.”

In fact, though Lee had, by massing every man in the northeastern theatre, including at last Stonewall Jackson’s valley army, succeeded in raising his numbers near Richmond to 90,000, McClellan still outnumbered him with 105,000; moreover, by cautiously pushing forward to the headwaters of the Chickahominy River, the Northern army was no longer divided by the watercourse as it had been before Seven Pines, and occupied an excellent position from which to attack. More Northern soldiers than ever were able to see Richmond’s spires and hear their bells, only a thin earthwork and a screen of troops away; but still McClellan refused to move. In the end it was the “timid and irresolute” Lee who took the decision for him. Informed by galloping J. E. B. Stuart, one of the last of the Western world’s cavalry swashbucklers, that the Union right flank at Mechanicsville was “in the air”—that is, neither rested on a defensible position nor was supported by an adjacent force—he decided to open a wide turning movement, around and then behind the Union army, on 26 June.

Ironically, he was to be forestalled, by what McClellan was now describing as one of a series of “partial attacks” designed to advance his artillery into a commanding position. On 25 June, McClellan sent forward troops from his right wing to skirmish across the old battlefield of Seven Pines and seize a piece of woodland known as Oak Grove, from which he planned to advance his heavy guns to high ground at Old Tavern and there open the bombardment of Richmond’s defences. Because Lee was preparing major operations which were intended to take place the following day on exactly the same sector, McClellan’s short-range advance bumped into part of the Confederate main deployment, and the encounter flared into a small battle that lasted the whole of the day. At the end, McClellan declared himself satisfied with the result, which brought him the ground he wanted, though at the cost of six hundred casualties. What he did not know was that this little engagement, now known as the Battle of Oak Grove, was to be the first in a week of continuous fighting, collectively to be called the Seven Days Battles, 25 June–1 July, which would culminate in the collapse of all his plans and the ignominious retreat of the Army of the Potomac to its original theatre of campaign in northern Virginia.

It is easily possible today to drive the whole circuit of the Seven Days Battles between breakfast and dinner. I did so in the late summer of 1992, though I had allotted much more time to what I planned as a preliminary reconnaissance. My exploration was a salutary reminder that there is no substitute, in the writing of military history, for going to see for oneself. Thirty years before, I had taught this campaign to my class of cadets at Sandhurst, and the cartographic image of the events of the last week of June 1862 remained impressed on my mind: 25 June, Oak Grove; 26 June, Mechanicsville; 27 June, Gaines’s Mill; 28–29 June, the Union retreat to the James River and the actions at Peach Orchard, Savage Station, and White Oak Swamp; 30 June, Glendale and Frayser’s Farm; 1 July, Malvern Hill. These are battlefields that looked widely separated on the maps I was using, the wonderful maps prepared by Ed Krasnoborski for that magnificent publication The West Point Atlas of American Wars. Years later I met Mr. Krasnoborski in his drawing office at West Point. I had specially asked to do so, for I rightly esteemed him to be a prince among mapmakers. I had imagined that he would occupy extensive, well-lit premises and command a large staff. I found him alone in a small windowless room, perched on a high stool at a tilted board, quite happy to be disturbed at his work but faintly bemused by my enthusiasm for what he did. Every word of my thanks was justified; there is still nothing to compare with The West Point Atlas, which remains the Gray’s Anatomy of Western military history.

Yet, for all the clarity of the Krasnoborski maps, I had not properly grasped from them the special relationship between one place and another; despite the scale of miles at the bottom, they made everywhere look further apart than it actually was. This optical illusion does not affect military historians only; it is so well known among soldiers that they commonly ask each other, without embarrassment, at the map table, “Have you seen the ground yourself? What does it look like: How far [they usually mean in time] is X from Y?” The mistake I had made was the last: because McClellan’s battles with Lee around Richmond had lasted seven days, I had imagined that their circuit would take almost as long. Setting out in my rented car from Chickahominy Bluff for Beaver Dam Creek, one of the most fiercely contested sections of front in the battle for Mechanicsville on 26 June, I had expected an hour of driving. I found it just round the corner. Gaines’s Mill, fought the following day, was three miles down the road; the course of the retreat to the James, which lasted two days and involved four actions, runs along only eight miles of modern roadway; Frayser’s Farm and Malvern Hill lie on the five miles of road—Virginia State Route 156—south of White Oak Swamp. Stops to inspect the sites and to read the series of splendid historical markers—grey-painted cast-iron columns put up by the Virginia State government in the 1930s—protract the tour. My reconnaissance—I was to drive over the route several times again later—lasted, however, a good deal less than a day. It was a vivid lesson in the difference between travelling time and campaigning time, and in the human inclination to take from a map what one expects to see, not what it is designed to tell you.

McClellan, in the immediate aftermath of his small success at Oak Grove, fell into different but familiar misapprehensions. “I shall have to contend against vastly superior odds,” he telegraphed to Washington that night. What had alarmed him were reports of troop movements into Richmond, which he took to be the Confederate army in Mississippi marching up to join Lee. It was a phantom army on a phantom journey. The only troops arriving were Jackson’s, from the Shenandoah Valley, but they were coming from the west, not the south, and by fits and starts. Jackson’s muddled intervention in the Seven Days Battles temporarily undid much of the reputation he had won along the Shenandoah. At least McClellan had grasped that he was to be attacked; instead of contemplating a counterstroke, however, he issued orders to prepare for a retreat to the James—what possessed him to think of such a move, when his main base, constructed with great labour and containing mountains of stores, lay on the York River defies explanation—and warned his corps commanders to prepare to defend their entrenchments. “I wish to fight behind the lines if attacked in force.” Here was timidity become almost paranoiac. A general in command of a superior force in an attacking position could think of nothing better to do with his troops than dispose them in entrenchments—which ought not to have been dug in the first place—astride the headwaters of the Chickahominy and await a weaker opponent’s offensive.

It was late coming and should not have discountenanced a man of resolute will. Lee had fixed on Mechanicsville, five miles north of Richmond, as his centre of effort, intending there to turn McClellan’s extreme right flank, which Jackson was approaching from a northerly direction. Mechanicsville then was not a town at all but a collection of buildings at a crossroads; a contemporary photograph shows some pleasant clapboard houses on a dirt road running through unfenced and worn meadow. McClellan had posted few troops there, and, as 26 June drew into afternoon, none of them expected to be attacked that day. When at four o’clock they were suddenly charged by Confederate infantry, they took fright, legged it to the rear, and did not stop until they found shelter in the nearest friendly entrenchments, which defended the eastern bank of a little tributary of the Chickahominy, Beaver Dam Creek. There, in what remained of the day, one of the nastier small battles of the Civil War was to be fought out until the smooth waters of the creek ran red.

Beaver Dam Creek is, I think, the most sinister little battlefield I have ever visited. Just to the north, at the top of a Park Service track which drops down to a concreted parking place, State Route 156 crosses the creek on a high modern bridge; it is just too far away for the sounds of the trucks using it to reach the visitor’s ear. What I heard was the croaking of swamp birds and what I saw were willows, goldenrod in flower, poplars, and a scattering of those dead, grey, branchless trees standing up from the water in the swamp bottom so distinctive of American wetlands. The setting is intensely green and lush, the vegetation so dense that the battlefield, except in a slot to the north where the truck route shows through the trees, is entirely shut off from the outside world. The creek itself, though quite fast-flowing, is almost hidden under sedge and cresses. It may be about ten feet wide; the whole battlefield, from the bank down which the Confederates attacked to the higher bank on which the Union troops, a division of Pennsylvanians, awaited them behind timber stockades, is perhaps 150 yards across. It was a warm, deserted, oddly beautiful bowl of stillness when I saw it on a fine September noon; on the afternoon of 26 June 1862 it must have been a place of sudden horror.

The Pennsylvanians, 9,500 strong, were on the east bank, hidden on rising ground among trees, through which gaps had been cut to open fields of fire for their artillery; on the west bank, where the visitor stands today, more trees had been felled to form an entanglement. As so often in North American warfare, the abundance of timber was to influence directly the outcome of the fighting. A Confederate artilleryman judged the Union position to be “absolutely impregnable to a front attack.” The Southern infantry, North Carolinans and Georgians, attacked none the less. They were repulsed and attacked again—and again. A Georgian carrying his regiment’s colour had to wrap it round the staff to get through the entanglement; next day he found ten bullets in the fabric and another in the wood. “We were lavish of blood in those days,” a Confederate general later wrote of a battle which was their first for many of his soldiers, “and it was thought to be a great thing to charge a battery of artillery or an earthwork lined with infantry.” There were six Union batteries of thirty-two guns opposite the Confederates at Beaver Dam Creek. By the end of the afternoon’s fighting, which Lee directly commanded himself from the western bank, 1,484 Confederates had been killed or wounded. The 1st North Carolina lost 142, including its colonel, lieutenant colonel, major, and six captains; the 44th Georgia lost 335 men, 65 per cent of its strength. It had been a terrible day; “nothing could be heard in the black darkness of that night,” a survivor recalled, “save the ghastly moans of the wounded and the dying.” It had also been Robert E. Lee’s first battle as a commanding general and it had profited the Confederacy nothing at all.

A robust enemy would have profited from this local success to press home the advantage, or at least to reconsider plans for retreat. McClellan did no such thing. Though Jackson had signally failed to intervene at Mechanicsville, to the north of which he hovered about in uncharaceristic inactivity, while the grim battle of Beaver Dam Creek was fought to a close, his presence confirmed McClellan in all his fears of being outmanoeuvred and outnumbered. On the night of 26–27 June he ordered Fitz-John Porter, commanding the corps which had held the Beaver Dam Creek position, to fall back four miles to another line on high ground, behind Powhite Creek and Boatswain’s Swamp, south of a tiny place called Gaines’s Mill. The position was strong, but by retreating further down the Chickahominy, where the river widened, while leaving his army astride it, he risked physically dividing his force. The retreat also surrendered a long stretch of the railroad—the Richmond and York River—by which he had planned to bring his siege artillery forward from the landing place at White House on the York. These circumstances, avoidable though they were, were the justification for the decision, already half-formed in his mind, which he now definitively took to retreat to the James River. He called it “shifting his base of operations”; but, as he had an excellent base on the York and none on the other river, he was unarguably abandoning the offensive and embracing withdrawal.

Given that McClellan was going, he did so with efficiency. Indeed one of the foreign observers attached to his army, the Comte de Paris, noted that he showed “a firmness of decision” in issuing orders for retirement that he had never shown in the advance. Retreat in the face of the enemy was a technical problem; echoing the classic masters, McClellan now called it “one of the most difficult undertakings in war.” Since, however, it could be done by the rule book—the rules laid down that a general should fight a firm delaying action while clearing the routes to his rear, evacuating the army’s train of transport, and then falling back behind strong rearguards—and adherence to rules calmed, we may surmise, the anxieties to which McClellan’s mind was prey whenever he had to grapple with the uncertainties of pressing forwards into ground controlled by the enemy, he solved the problem very efficiently. His engineers had built and rebuilt four bridges over the Chickahominy: the Grapevine Bridge and, named for their constructors, Alexander’s, Woodbury’s, and Duane’s. The position McClellan had ordered to be held between Gaines’s Mill and the Chickahominy covered all four. It rested on a plateau between Gaines’s Mill and the Chickahominy, some forty feet high, two miles wide, and a mile deep, and much of it heavily wooded.

There is an extraordinary calm about this piece of the Peninsula battleground today. Perhaps the woodland is denser now than it was then—nineteenth-century photographs generally show, outside the swampland, a more open Peninsula—and it certainly presses hard against the escarpment on which Porter’s troops took their stand to repel the Confederates on 27 June 1862. At the Watts House, one of the few scattered dwellings which occupied the tabletop of the plateau above Boatswain’s Swamp, the forest so closely surrounds the farm and its mown enclosure that nothing of the outside world intrudes at all, no sound or sight of the twentieth century. Unlike Beaver Dam Creek, where a miasma of past blood-letting hangs in the hot air of the little amphitheatre, the Watts House clearing is a clean, cool, and tranquil place, with no hint of old violence about it. Yet the Battle of Gaines’s Mill, part of which was fought on the lip of the ground that falls away here to the stream below, was a far longer and bloodier action than that of 26 June. In the opening stages, the two sides were more or less equally matched: Fitz-John Porter’s corps, separated from the rest of McClellan’s army by the Chickahominy, faced part of Lee’s army, the rest of which was also south of the river. Porter, however, had a marked superiority in artillery, twenty batteries to nine, and over a hundred guns to less than fifty; on the other hand, while McClellan was leaving Porter to fend for himself, Lee was counting on Jackson to reinforce his line and so achieve a preponderance of numbers.

The oddity of the situation was that Lee did not grasp the moral superiority he enjoyed. He had no inkling of McClellan’s collapse of will, imagined that his opponent was still fighting to protect his line of communications with the great Union base at White House on the York, and had not detected from any source of intelligence that McClellan was intent on retreating to the James River and so effectively on giving up his offensive altogether. Had he done so, he would have pressed events. As it was, he failed to get the battle going at full swing until noon, failed to energise Jackson, who was still behaving as if combat-fatigued by the Shenandoah Valley campaign, and so failed to bring all available force against the plateau on which the Watts House stands until late afternoon. Porter and his men showed a determination in defence which McClellan had abandoned altogether. While they rode out the storm of repeated Confederate attacks across Boatswain’s Swamp and up the wooded slopes beyond it, McClellan was commanding the battle from south of the Chickahominy by telegraph in a bizarre anticipation of the generalship of the First World War. An aide who came forward on horseback found Porter watching the stream of wounded making their way back from the firing line, while Confederate shells dropped near him. “We’re holding them but it’s getting hotter and hotter” was all he said. When the aide returned to McClellan’s headquarters, he found only the telegraph office still operating, the tents packed, horses saddled, wagons loaded, and McClellan sitting silent by himself on a tree-stump, awaiting the outcome.

Porter very nearly won the Battle of Gaines’s Mill. His 35,000 men, who included several units of the tiny pre-war regular army, fired volley after volley, salvo after salvo, throughout the long hot afternoon, inflicting terrible casualties on each Confederate regiment that approached their earthworks. The 1st South Carolina Rifles lost half its strength killed and wounded, 309 men. The 20th North Carolina lost 272, the 4th Texas 253. The colour of the 7th North Carolina changed hands five times, as those carrying it were shot down in turn; thirty-two bullet holes were later found in the fabric. When bullet hit flesh, but particularly bone, the effect was terrible. Both sides were using some variant of the Enfield rifle, which threw a solid cone of lead, half an inch in diameter and weighing a full ounce, to a range of four or five hundred yards. The small-calibre bullet of the high-velocity rifles of the First World War would travel farther but usually made a neat puncture wound. The ball of the old smoothbore musket of King George’s War would kill in close-range combat but otherwise, if it hit, would often track along a bone without breaking it or lodge in muscle; some of the Southern troops at Gaines’s Mill were still carrying smoothbores. The Enfield bullet, by contrast, being both large and fast-moving, could do catastrophic damage, destroying large volumes of tissue or reducing bone to shattered fragments. To suffer wounding in a battle like Gaines’s Mill, therefore, was not a passport off the battlefield, but might mean disablement for life or, indeed, death in the immediate aftermath.

Until mid-afternoon on 27 June, the weight of casualties fell on the Confederates. Lee had been finding the troops, however, to thicken his front, while McClellan sent almost none, though he had many to spare. At one stage he promised “the whole army” but released only two brigades, one-tenth of his available force; nor did he go to Fitz-John Porter’s battle line himself. Lee, by contrast, and Stonewall, who had at last arrived, put themselves squarely behind the centre of the firing line as the light began to fade, urging their subordinates to carry the Union position at whatever cost. Some 35,000 men were now ranged against each other on either side, but in the Union line the heat and losses of the day were beginning to tell. So, too, was the burden of defence. The Confederates, buoyed up by the spirit of the offensive, gradually began to overcome. It would never be agreed which of the Confederate formations eventually achieved the break. When it came, in the centre of the line, the collapse was so rapid and widespread that credit for victory was difficult to apportion. An index of the scale of the Northern defeat was the number of guns lost, twenty-two pieces or nearly a quarter of the artillery, a clear indication that the retreating Union infantry passed through their own battery positions too rapidly for the gunners to hook in the teams and get the guns away. Only the descent of darkness, in which the attacking Confederates held fire for fear of hitting each other, saved Porter’s corps from complete rout.

Had Jackson’s Shenandoah Valley troops intervened in mid-afternoon, as Lee intended and expected, all would certainly have been up. Porter’s troops would have been pushed against the Chickahominy and been slaughtered or forced to surrender. As it was, they pressed over the bridges in the dark and got back to the main position, where the sixty thousand men McClellan had not committed to their assistance waited about their campfires. In the interval before the pursuing Confederates cut the wire, he got off to Washington one last self-pitying telegram. “I have lost this battle because my force was too small. I again repeat that I am not responsible for this.… The Government has not sustained this army.… If I save this army now, I tell you [the Secretary of War] plainly that I owe no thanks to you or any other person in Washington. You have done your best to sacrifice this army.”

A staff officer in the Washington telegraph office sensibly censored the last two insubordinate sentences, leaving Lincoln, still unapprised of the full extent of McClellan’s moral collapse, to assure him of his continuing support. McClellan meanwhile was busying himself at the engineering work which he did best, covering the crossings of the Chickahominy with battery positions, having trenches dug, and directing the bridging of White Oak Swamp, the tributary of the Chickahominy which stood between him and the landings on the James River to which he planned to make his final retreat. Bridging was an operation at which the Army of the Potomac was now skilled. It had bridged to bring the railroad forward from White House and bridged the Chickahominy. At high speed its bridging columns now constructed two crossings, one in the remarkable time of two hours. Across them throughout 28 June the Union army pulled back. It remained a formidable fighting force. Though it had lost 894 killed, 3,114 wounded, and 2,829 captured, Lee’s losses were higher. He had given no prisoners, but his dead totalled 1,483 and his wounded 6,402, from a smaller army. He continued to misinterpret McClellan’s intentions, not crediting that he could be in full retreat and believing that he might either recross the Chickahominy lower down and resume the offensive against Richmond or stand where he was and fight to regain the rail link to White House. It was only when a cavalry patrol brought news that Union troops had destroyed the railroad bridges in their faces and set White House ablaze that he grasped the truth and organised his army to press the Northerners in their retreat.

A series of small battles fought on 29–30 June on the roads south to the James River were the result. They form a confused episode, which the modern traveller follows with difficulty, for the armies were deployed to fight to the left and right of the roads, which therefore usually cross the lines at right angles. In some sectors the countryside is open, particularly at Savage Station on the old Richmond and York railroad, where the ground is bumpy, abandoned-looking heath rather like some nasty patch of the Western Front in France. In others the woodland is dense—is this second growth, invading old, worked-out farmland, as one suspects?—and along the watercourses the vegetation is very thick and lush. Yet the whole area is also strangely uninhabited, without town, village, or even farmstead to give a traveller bearings. Richmond is only ten miles away to the northwest, but this landscape seems to have no social or economic connection with it at all. In Europe the surroundings of a major city would be dotted with small centres of habitation, a pub or bistro or Wirtschaft every mile or so to offer hospitality. Henrico County is eerily deserted, only the odd gas station–cum–country store at a crossroads giving any sign of human settlement at all. The traveller here can easily go hungry and thirsty, as I did, and all the more readily understand why the American tourist in the European countryside is so charmed by the wayside inn, the husband-and-wife estaminet, the beer garden in a forest clearing. America might be like that, had time allowed; Europe’s pattern of settlement reflects centuries of slow, short-range movement by horse or wagon or on foot, which required frequent stops for refreshment, sustenance, fodder, or stabling. In much of America, by contrast, the railroad often antedated the trunk road, and then the appearance of the motor car robbed the neighbourhood halting place of point. Those that existed in places of old settlement like the Peninsula must have disappeared over the last century, leaving nothing to replace them but those charmless caravanserais of Wendy’s and Burger King restaurants, filling stations, and efficiency-unit motels into which the weary vacationer falls by default at the end of a day’s driving. It is one of the greatest oddities of life in the richest country in the world that its hinterland is so bereft of civilisation’s small comforts.

I did not pick up clear traces of the course of the Seven Days Battles again until I passed Glendale, now little more than a country store on State Route 156, and approached Malvern Hill, the site of the last battle of the Seven Days. On the way I had found Grapevine Bridge over the Chickahominy, almost destroyed in the May flood, rebuilt by McClellan’s engineers, then destroyed by his engineers in the retreat from Gaines’s Mill, then rebuilt by Stonewall Jackson’s engineers, today a graceless county engineer’s concrete span on Route 156. I had also found Savage Station, which McClellan briefly made his headquarters on 28 June, and White Oak Swamp, where the Union fought its main delaying action on 29 June; a wet wilderness in 1862, its banks have been drained today and it is little more than a sedate stream south of a modern railroad track. On the spot where Confederate and Union soldiers were busy killing each other 130 years earlier, I found an auction sale in full swing at a decrepit farmhouse, with neighbours bidding enthusiastically for tattered hymnbooks, Marshall Field catalogues, and long-superseded items of hand-agriculture equipment; an enterprising confectioner was selling ice-cream out of a small truck to a crowd seated on folding chairs on what might once have been a lawn, rapidly going back, as everything does in America, to secondary forest.

These were brutal and bitter battles none the less. During the rearguard actions of 29 June at Savage Station and White Oak Swamp the Union army lost nearly two thousand men and the pursuing Confederates nearly a thousand. The next day, 30 June, there was a pitched battle at Glendale and Frayser’s Farm, fought by McClellan to cover his taking of position on the feature he had correctly identified as the spot where he could deploy his superior numbers to fight an action that would halt his pursuers and allow him to get the army to safety down the James River. That spur was Malvern Hill, the highest elevation on the Peninsula, which covers the roads leading to the landing places on the river behind the ancient Shirley, Berkeley, and Westover plantations. The army of the North would make its escape through a spot called Harrison’s Landing opposite Bermuda Hundred, where another Union attempt on Richmond mounted from Fortress Monroe would fail in 1864. During 30 June McClellan at first retreated to the river bank, where he lost touch with his own telegraphic link to the army five miles away and then, without appointing a subordinate to deputise for him in command, boarded the Galena, one of the gunboats which had fought the artillery duel with Fort Drewry on 15 May. As it cruised about the river firing on Confederate troops ashore, McClellan was in a strange state of elation; an observer who appreciated how complete his demoralisation was noted that, underneath, he was “a man … cut down.… He was unable to do anything.”

His army had much to do. Some of it was still leaving the battlefield of White Oak Swamp, getting the long train of wagons on to the Quaker Road (today State Route 156) that leads to the James, and felling trees to impede the Confederate retreat; that American plethora of timber was once again playing its part in an operation of war. The rest of the army was assuming positions in which to defend the retreat to Malvern Hill, both against Jackson’s troops pursuing it on the direct route south down Quaker Road and to form a flank against the rest of Lee’s men, who were angling in from the left. The centre of the position was at Glendale, the crossroads where I stopped to refuel but failed to find lunch on Saturday, 26 September 1992, but the focus of the fighting was to lie at a nearby homestead, Frayser’s Farm. This is high, dry, rather sandy countryside, where tall-standing corn intermingles with forest, fine but empty country now and not much changed, it appears from contemporary descriptions, since 1862. The Confederates certainly complained of the difficulty of spotting the enemy among the trees—two days earlier the 1st U.S. Sharpshooters had skirmished “Indian-style” in the woodlands along the Chickahominy—and now Union riflemen and artillery batteries took a heavy toll as their opponents pressed forward in an effort to cut their retreat to Harrison’s Landing.

The results of the fighting can be found in the Glendale National Cemetery, in between Glendale itself and Malvern Hill, from which the bodies of those killed the next day have been brought also. They lie under little white rounded markers, inscribed with the barest of information—name, regiment, date of death. The starkness, matter-of-factness, of the cemetery took me aback. Glendale was the first American military cemetery I had visited, and I expected something more like those built and maintained by the Commonwealth War Graves Commission in France, in Flanders, and, indeed, in almost every country in the world. The British—and the soldiers of the Commonwealth and Empire who died with them in the wars of the twentieth century—are scattered wide. I have found War Graves Commission cemeteries wherever I have travelled, as tourist, correspondent, or historian, from Lebanon, where I was covering the civil war in 1983, to South Africa, where I went to investigate the Namibian war in 1984, to Pakistan, which I visited during the Afghanistan War in 1985; in Germany, in Israel, in Turkey, where the dead of Gallipoli lie within sight of Homeric Troy, in Australia, in Canada, and in the United States itself.

Commonwealth war cemeteries signal themselves from a distance. First, there are the indications of a gardened landscape—a mown verge to the roadside, a stretch of park wall in stone. Then, above the wall, there is a glimpse of flowering shrubs and of garden sculpture, a memorial stone, a cross, a classical gateway. Inside, the ideal English garden unfolds, roses, rosemary, bay, myrtle, and shaven turf, across which stretch the rows of white Portland headstones, carved with a cross, the regimental badge and title, the name, the age, and the date of death of the fallen soldier. At the bottom of the stone is space for a family inscription, two lines that read all the more poignantly for the effort made, often by simple people, to cram heartbreak into no more than a dozen words. The British war cemeteries move the visitor readily to tears, as they were intended to do. Edwin Lutyens, Britain’s greatest modern architect, designed the monuments, Rudyard Kipling wrote the inscriptions that appear everywhere, from the Arctic to the South Pacific: “Their Name Liveth for Evermore”; “A Soldier of the Great War Known unto God.” Death and beauty intermingle in a cunningly contrived Arcadia, eloquent of the ease with which the British fall into romantic communion with the ideas of self-sacrifice and love of country.

Other countries do things differently. The central point of the French national war cemeteries is the tricolour, symbol of liberty, equality, fraternity, floating above a field of white crosses, symbol of France, eldest daughter of the Church. The Germans bury their war dead among sombre evergreens in terrible mass graves; at Langemarck in Belgium, site of the slaughter of the German student volunteers of October 1914, 36,000 bodies, the same number as were killed in battle in the seven years of the Vietnam War, lie intermingled in a single plot no bigger than a stage on which Wagner’s Twilight of the Gods might be sung; those who stand at its edge are brought to wonder at what dark forces underlie the German urge for dominance over the Old World. America’s military cemeteries arouse different emotions altogether. The headstones are so small, the landscape which surrounds them so little interrupted by the digging and clearing which preceded their planting: this is a place, the visitor feels, which might go back to nature in a generation, as so much that is human interruption of the vastness of America has already gone back. I have found cemeteries elsewhere in the United States already relapsing into the wilderness: at Durham, Connecticut, where garter snakes have made their habitation among the elaborate, overgrown headstones of soldiers of the Revolution; at Princeton, where a tiny graveyard of Washington’s battle of 1777 is reverting to scrub and briar.

Glendale, I felt as I walked over its rough-cut grass, was fated for the same forgetting. Its headstones commemorated an America which belongs to the past, an America of exclusive Anglo-Saxonism and simple state patriotism. The list of names I noted, though it contains a sprinkling of Irish and German, is preponderantly English of the two centuries of colonial settlement: Wilson, Burton, Atkins, Pottle, Hall, Cole, Lawrence, Wheldon, Long, Angell, Tarbell, Prosser, Denson, Goodrich, Tibbets, Cleeland, Coombs. There are Lawrences and Coombses in my Wiltshire village in England where, the parish record books testify, they have been born, married, and died since the sixteenth century. I feel little doubt that genealogy would establish a connection. Americans and Canadians come to our churchyard each year to peer at gravestones and mark names in their notebooks of family history. What I was doing at Glendale, as I walked among the tiny headstones, was pursuing a similar quest; but while our Coombses and Lawrences spread themselves over no more than a dozen or so square miles in places called Witham Friary, Milton-on-Stour, North Brewham, Marston Bigot, the American Coombses, Lawrences, and fellow settlers from the old country who lie at Glendale had come with their regiments to fight and die for the Union from places as far away as Michigan, Wisconsin, Maine, Utah, and California. The English are a people of fierce local loyalties; hereabouts villages only a mile apart insist on their difference from each other. The English of Michigan and Maine who lie at Glendale had preserved that emotion but invested it in the vast tracts of the continent their settler ancestors had taken as home. There will be no real sadness if Glendale eventually goes back to nature. That would be in the spirit of the English escape from the narrow constriction of their Domesday fields to the wideness of the New World, where space enlarges the individual and his progeny, and time and history are dimensions to be left behind.


Richmond, the capital of the Confederacy, 1861

(photo credit i3.1)


(photo credit i3.2)

Robert E. Lee, after the Confederacy’s defeat, 1865

(photo credit i3.3)

The Battle of Williamsburg

(photo credit i3.4)

“Stonewall” Jackson

(photo credit i3.5)

The Battle of Malvern Hill

(photo credit i3.6)

The Union retreat to Harrison’s Landing

(photo credit i3.7)

McClellan faces Lincoln, after the failure of the Peninsular campaign, October 1862

(photo credit i3.8)

The ruins of Richmond in 1865, with the Capitol in the skyline

(photo credit i3.9)

The Virginia State Capitol today

(photo credit i3.10)


Plains Indian hunting buffalo, by George Catlin

(photo credit i4.1)

Indians catching wild horses, by George Catlin

(photo credit i4.2)

Cheyenne village on the Great Plains, c. 1870

(photo credit i4.3)

Custer in the uniform of his own design

(photo credit i4.4)

Sitting Bull, chief of the Hunkpapa Sioux

(photo credit i4.5)

Custer with his Crow Indian scouts before the Battle of Little Bighorn, 1876

(photo credit i4.6)

Otto Becker’s reconstruction of Custer’s Last Stand; the landscape remains much the same today.

(photo credit i4.7)

How news of the disaster reached the nation: the New York World, 6 July 1876

(photo credit i4.8)

Temporary graves on the Little Bighorn battlefield

(photo credit i4.9)

Glendale, a little clearing in the forest, holds the bodies of 1,192 soldiers, mainly Northerners, who died on 30 June or 1 July 1862. The majority were killed at Glendale and Frayser’s Farm, a day of fighting in which McClellan’s undirected army was at a disadvantage. At Malvern Hill, McClellan had the battlefield he wanted: Andrew Humphreys, his topographical engineer, wrote of “the splendid field of battle on the high plateau where the greater part of the troops, artillery etc. were placed.” It has been called an American Waterloo, and there are indeed similarities: the Union position on the high ground, bisected by a good road, the long fields of fire sweeping down towards the positions from which the enemy had to launch an attack, and the strong flanks. Malvern Hill is, if anything, a better position than Waterloo, where Napoleon had high ground of his own on which to position artillery to oppose Wellington’s, whose flanks were protected by some broken ground on the left and the orchards of Hougoumont on his right. Two streams, running through dense woods, absolutely define the Malvern Hill position, funnelling any enemy attack into a corridor no more than a mile wide. All this is grasped instantly by the visitor’s eye, which sees a landscape that can have changed little in 130 years. “It was as beautiful a country as my eyes ever beheld,” wrote Lieutenant Charles Haydon of the 2nd Michigan. “The cultivated fields, interspersed with belts and clusters of timber … wheat was in the shock, oats were ready for harvest, and corn was waist high.” It was strikingly beautiful when I saw it in the late summer of 1992; then the corn had been cut and the fields burnt golden by three months of sun, but the woods stood tall and silent on each hand, their dark depths only hinting at the stream bottoms they concealed, and the Quaker Road down which McClellan’s men had marched to deploy across the plateau and range their batteries on the crest was as empty as a country lane in the West of England. Not a sound broke the drowsy, noontide air, and not a sign remained to warn that on 1 July 1862 this had been a terrible place, except for the National Park Service’s little marker and the recording machine’s description of how, on the morning following the battle, enough of those wounded “were alive and moving to give the field a singular crawling effect.”

McClellan should have won every fight he fought on the Peninsula. He did win Malvern Hill. He was, moreover, present in person. The morning of the battle, as at Waterloo, and the early afternoon were occupied with a great artillery duel. The Confederates might have given as good as they got, had they brought forward all the guns available, but for no good reason, except poor staff work, they did not. As a result, their two “grand batteries” positioned left and right of the Quaker Road (today at places off State Route 156 just south of Glendale Cemetery) were gradually overwhelmed by the greatly superior concentration of Union artillery on the Malvern Hill plateau. By three-thirty, when McClellan arrived after another morning cruising in the James aboard USS Galena, Lee had accepted that he was not going to win by gunfire and had sent forward his infantry. The point of assault was entrusted to Lewis Armistead, who had come bottom of the West Point class in which McClellan graduated one from the top and who was to die a year later at Gettysburg placing his hand on the muzzle of a Union cannon on Cemetery Ridge. He might have died at Malvern Hill; terrain and tactical situation were not dissimilar. In the event, his brigade was so devastated by Union gunfire that it was driven to ground and dropped early out of the battle. Its advance was succeeded by a succession of others, launched as fast as troops could be put into action. Lee’s problem was not, as it had been earlier in the Seven Days Battles, to find formations. All stood ready to hand, including Stonewall Jackson’s men; the problem was to find space on which to deploy them. In the tunnel of fire the ground and the Union positions contrived, the Confederates were reduced to mounting one narrow frontal assault after another, at fearful cost and without productive effect. Lee, who had placed himself in the centre of his battle line, tried to open the engagement to the flanks, but, even had his men succeeded in finding a way over the two streams that hemmed them in, Turkey and Western Runs, it seems doubtful that the pattern of the fighting could have been altered to productive effect. The battle had acquired a momentum and logic of its own, rather as the dreadful trench-to-trench assaults of the First World War were to do. As darkness fell, a last Confederate charge reached the summit of Malvern Hill, forcing one Union battery to limber up and retire. The gun line itself held firm, however, and under its unrelenting cannonade the Confederate line fell back. When one of his subordinates proposed making a last attempt on the iron front of Union batteries, Stonewall replied, “I guess you had better not try it, sir.” It was the admission of defeat.

Next morning the two sides, separated by the “singular crawling effect” of a no-man’s-land full of wounded survivors, counted the cost. Lee’s army had lost over 1,000 dead, McClellan’s, which had fought largely an artillery battle, about 400. The toll of wounded was very high also, more than 4,000 Confederates, nearly 2,000 Union. Altogether, the Seven Days Battles had caused 36,000 casualties, and the total for the Peninsular campaign as a whole, from the landing at Fortress Monroe to the final evacuation, would reach 56,000. This from a force, it was later calculated, which approached a quarter of a million combined, North and South.

All for nothing, too, from the Union point of view. On the morrow of Malvern Hill, some of McClellan’s subordinates raged against his determination to behave as a defeated general. Phil Kearny, commanding Third Corps, announced that “I, Philip Kearny, an old soldier, enter my solemn protest against this order for retreat.… I say to you all, such an order can only be prompted by cowardice or treason.” Kearny was indeed an old soldier. He had lost an arm in the Mexican War in 1847 and, as a freelance, won the Legion of Honour with Napoleon III’s Imperial Guard fighting in 1859 at Solferino and Magenta. In the Peninsula he had had the eerie experience of picking up an echo of Washington’s victory over Cornwallis when, on reconnaissance beyond Fortress Monroe, he was told by an aged plantation slave of his memory of hearing cannon firing at Yorktown in 1781. What Washington had achieved he believed McClellan could do also. Richmond lay only ten miles from the battlefield; with resolute will—and the soldiers were willing to fight—McClellan might have used the victory of Malvern Hill to march on the Confederate capital, carry the defensive lines around it, and end the war. McClellan would have none of it. Apparently oblivious of what his subordinates felt, he retired aboard the Galena and issued orders for retreat to Harrison’s Landing eight miles away along the river road. He had reconnoitred it as a safe haven in person, while he might have been commanding in battle, and was determined to withdraw the Army of the Potomac into it rather than exploit his success.

He insisted, particularly in his Independence Day address to the army on 4 July, that the move was a change of base, as he had been insisting to Lincoln since the second of the Seven Days Battles, and he made the new base very strong. Into a space only a mile deep and four miles long, flanked by Herring and Kimmage’s creeks, backed by the James River, which Union gunboats controlled, and fronted by a hastily dug line of strong earthworks and battery positions, he crowded all ninety thousand men of the Army of the Potomac, with their stores and guns. Lee, after one look at it, declared it to be impregnable. McClellan’s claim that it represented the culmination of a strategic redeployment cut less ice with his soldiers. “We are at a loss,” wrote Sergeant Edgar Newcomb of the 19th Massachusetts, “to imagine whether this is strategy or defeat.”

It was defeat in every sense, of course, except McClellan’s willingness to admit it. For the next six weeks he protested to Washington that his plans remained entirely offensive, that he only needed more troops to make victory a certainty, and that his new plan was another change of base: to cross the James to the south bank, advance on Petersburg, the rail junction twenty-five miles below Richmond, entrench himself there, await a convergent attack by other Union forces on the Confederate capital, and so complete its capture. Lincoln came to Harrison’s Landing on 8 July to hear McClellan’s assurances that victory was still possible—given 100,000 men—and Henry Halleck, the new general-in-chief, came on 25 July. That hard-headed professional heard McClellan out, demonstrated that if, as he insisted, Lee now had 200,000 men—pure fantasy, as with all McClellan’s intelligence assessments—the Confederates would defeat both the Army of the Potomac and any other marching to co-operate with it, and gave him a stark choice: either resume the advance on Richmond from Harrison’s Landing with such local reinforcements as were available or quit the Peninsula for good. McClellan seemed to accept the ultimatum; but, as soon as Halleck was back in Washington, he telegraphed him with further alarmist estimates of Lee’s strength and demanded yet more reinforcements if he were to move. Halleck had now heard enough. McClellan, he said, “does not understand strategy and should never plan a campaign.” On 30 July he wired the order to begin evacuating from Harrison’s Landing the Army of the Potomac’s sick, of which unhygienic conditions in the crowded encampment had produced thousands; on 3 August he issued the command to abandon the Peninsula altogether and bring the army north again to open a fresh campaign in upper Virginia in combination with the forces McClellan was demanding should be sent to him.

So the great army began its withdrawal down the river road, today State Route 5, across the Chickahominy, past Williamsburg and Jamestown, through the old Yorktown lines it had besieged so laboriously in April, across the ground on which it had concentrated in March, to Fortress Monroe, there to embark for Aquia Creek and Alexandria on the Potomac. It was the completion of the rondo which had begun almost six months earlier. Then the army had taken ship with every hope and much objective indication that Richmond was about to fall into its grasp. By September, it would be Washington that lay under threat, with Lee north of the Potomac, Harpers Ferry in Confederate hands, and a desperate battle about to be fought at Sharpsburg near Antietam Creek; its result forced Lee to withdraw to the Rappahannock, that de facto military frontier between South and North, but the Union effort to secure a bridgehead across it at Fredericksburg in December ended in defeat and repulse; 1862 was not a good year for the Union cause in the eastern theatre of war.

The landscape around Fredericksburg, that charming little town of grey-painted clapboard houses on narrow, tree-lined colonial streets, bears the evidence of the long ensuing struggle along the hilly banks of the Rappahannock to this day. Trenches follow the defensible contours, protected here and there by restored timber chevaux-de-frise, and battery positions can be discerned among the enveloping woodlands. It is only a few minutes’ drive northwestward to the battlefield of Chancellorsville, scene of Lee’s supreme act of generalship and of Stonewall Jackson’s death in April 1863, and entrenchments follow one all along the road. General Hal Nelson led me along the route in 1992, elucidating with brilliant clarity each step of the campaign, from the rapids, which make Fredericksburg the head of navigation on the Rappahannock, to the unbuilt railroad track along which Jackson drove his battle-winning flank march on the day before he died.

Yet the memory of this narrow theatre of war that remains most strongly with me is of the shoreline that binds it so closely together: the low, lush, overhung banks of the Chesapeake at Annapolis, from which part of McClellan’s army sailed for the Peninsula and along which I navigated in 1983 in a naval academy professor’s sloop; the upper reaches of the Potomac under eighteenth-century, brick-built Alexandria, opposite Washington, where the main body embarked; the densely forested York River; the long, sloping sward beneath Shirley and Westover plantations at Harrison’s Landing, where McClellan lingered after the Seven Days while his thousands of men and horses trampled the light, yellow sandy soil to pulp; the dank, gloomy defile of the James River at Drewrys Bluff, high-water mark of his attempt upon the defences of Richmond; and above all, the grass-grown hummocks of the Yorktown redoubts opposite Gloucester, defending nothing now but the delicate beauty of a forgotten colonial seaside place, but in 1781 and in 1862 the focus of giant military enterprise. “These frowning battlements,” General John Magruder had told his army in the days while he was awaiting McClellan’s assault, “are turned in this second war of liberty against the enemies of our country.” Too flowery; but the recurrence of decisive conflict on “these very plains of Yorktown,” as he called them, re-emphasises the dominance of the shape and size of America over the wars that its inhabitants fight within it. The French conquest and loss of Canada had begun and ended on the great headland of Quebec; the American struggle to win their liberty from the British had begun with an assault on Quebec and ended on the headland between the York and the James, where the English had planted their first place of settlement coevally with Champlain’s arrival in the St. Lawrence; the first great Northern assault on the stronghold of Southern power had begun at Yorktown and petered out in a withdrawal past the original English Jamestown colony; but the culminating act of the Civil War in 1864 would open with a return to the battlefields of the Seven Days at White Oak Swamp and Malvern Hill and a transhipment of much of Grant’s army up the James to the vicinity of Harrison’s Landing and an advance to Petersburg to begin that investment which McClellan had claimed was the object of his “change of base” but which he lacked the resolution to initiate.

Two and a half centuries of the European presence in North America had seen much warfare, but the spaces of the continent had been touched by little of it. The sea and its inlets, the great rivers and their tributaries—St. Lawrence, Mississippi, Ohio, Tennessee, Cumberland, Hudson, Mohawk—the Great Lakes by which they are fed or into which they run, appear in hindsight more significant altogether as determinants of events than any of the human players who acted out the drama of campaign in the narrow corridors made available by nature for their efforts. A few Americans whose intelligence attuned them to accept the smallness of man before the vastness of nature in the New World displayed an ability to work with, rather than against, its forces: Samuel Champlain, with his instinctive grasp of the unity of the lakes, the St. Lawrence, and the Hudson, was one; George Washington, with his understanding of how the hinterland could be used as a barrier against the settled littoral, was another; Benedict Arnold, who possessed a fierce determination to exploit difficulties of weather and terrain as a means of surprising his enemies, was a third; perhaps the greatest of all was Ulysses S. Grant, whose ability to carry in his head a mental map of the Civil War’s nodal points made him the master of manoeuvre between them. Most of the rest, however, those who approached warfare in America in terms of European strategy or tactics, owed the defeats they thereby suffered to their failure to comprehend how fundamentally different the New World was from the Old. No one was more guilty of that fault than McClellan. What remained of the warfare to be fought in America would be directed by pupils and subordinates of Grant, whose gifts were for generalship over wide spaces and in the continent’s interior. “I will take no backward step” was his watchword in the harshest and most impenetrable of terrains. In the era of Grant’s presidency which followed the Civil War, America’s step lay forward, across the wide Missouri and into the heartland of European America’s last enemy, the native Americans of the Great Plains. That was where the final battles on American soil were to be fought.

If you find an error or have any questions, please email us at Thank you!