by Bruce Catton
A hint of the way things were going to work came on March 23, in the battle of Kernstown.
Ever since the unhappy canal boat expedition McClellan had kept an army corps in the lower Shenandoah Valley, and the commander of this corps, General Nathaniel P. Banks, had been having a pleasantly uneventful war. He had 25,000 men and the Confederates in his front numbered hardly 4500; decisive odds, surely, except that the Confederates were led by Stonewall Jackson, about whose singular capacities neither General Banks nor the rest of the world knew as much just then as they would know a little later. Jackson’s little force had been pushed out of Winchester and had gone, apparently, far to the south, and could be nothing more than a minor nuisance; so when McClellan began his move to Fort Monroe and needed a garrison for the area around Manassas he naturally thought about General Banks. Banks was ordered to leave a division at Winchester and prepare to bring everybody else east of the Blue Ridge, and he promptly obeyed. At Winchester, with its principal advanced line at Kernstown, a few miles south, he posted the division of Brigadier General James Shields; 11,000 men, approximately, whose chief function was to keep the lower valley clear of Rebels so that the rebuilt line of the Baltimore & Ohio Railroad could operate without interruption.
Shields came from County Tyrone by way of Illinois and the west coast; a lean, combative man who had the odd idea that Stonewall Jackson was afraid of him, and who had once challenged Abraham Lincoln to fight a duel, but who otherwise was competent and well-balanced.4 He had led troops with distinction in the Mexican War, and when Banks went east of the mountains Shields stayed on the alert and kept his troops alert also, maintaining as good a lookout as any general could considering the fact that at this time, in Virginia, the Federals were using some of the world’s worst cavalry against some of the world’s best.
Jackson had heard about Banks’s withdrawal, and he seems to have believed that the man was leaving fewer troops around Winchester than was actually the case. Joseph E. Johnston had just warned him to keep close to the Yankees—the sort of order no one ever needed to give Jackson twice—and anyway Jackson considered Winchester his own private bailiwick and wanted to drive the invaders out for personal reasons. On the morning of March 23 he sent Johnston a characteristic message: “With the blessing of an ever-kind Providence I hope to be in the vicinity of Winchester this evening.”5 Then he made his word good by moving in to make a savage attack on the Federal lines at Kernstown.
Probably the least important thing about the battle of Kernstown is that Jackson tried to do the impossible, and failed. Shields had more than twice Jackson’s numbers, and these Federals were good soldiers—Westerners, mostly, plus a few Pennsylvanians, with some regiments which would eventually be listed with the best combat units in the Union Army. Jackson’s line was halted and at last it had to give ground, and by dark its dour commander, furious over the reverse, was leading it up the valley in full retreat. He had lost some 700 men, and Shields (who himself was wounded) had had smaller losses and was entitled to claim a victory.6 For the rest of the war, Shields’s men bragged that they were the only ones who had ever beaten Stonewall Jackson.
But the victory meant nothing at all, whereas the mere fact that the battle had been fought meant a great deal.
The lower Shenandoah Valley was an extremely sensitive area. When a hostile army touched it the Federal government would react vigorously, almost automatically. Kernstown revealed this fact, and both Jackson and Robert E. Lee made note of it for use later on. What followed Kernstown was most instructive to both of these soldiers.
It seemed clear to the Federals that Jackson was much stronger than anyone had supposed—otherwise he would hardly have dared to attack Shields—and there was a hasty reshuffling of troops. Banks was sent back to Winchester, horse, foot, and guns, to drive Jackson away and keep him away. It seemed necessary also to reinforce Pathfinder Frémont, who was just assuming command in West Virginia and who must protect the western segment of the Baltimore & Ohio. (If the Rebels were strong enough to attack Shields they no doubt contemplated aggression a little farther west, as well.) So the division of Brigadier General Louis Blenker was detached from McClellan and sent to Frémont, with orders to tarry a while in the lower valley until Banks had finally disposed of Jackson.
Meanwhile McClellan was trying to get the Army of the Potomac down to Fort Monroe. At the end of March his move was well under way. In the four army corps commanded by General Keyes, Heintzelman, Sumner, and McDowell—those commanders whom he had not chosen but who had been thrust upon him by the President—he had 126,000 men, present for duty. Keyes and Heintzelman were already on the peninsula, Sumner’s corps was in transit, and McDowell’s was ready to leave. Now this disturbance at Kernstown knocked McClellan’s arrangements slightly out of line and led to a major miscalculation.
President Lincoln had consented to the peninsula move with grave misgivings, and had laid down two firm conditions to govern it. First, there must be a strong force around Manassas to keep the Confederates out of northern Virginia; second, there must be, in Washington and its circling camps, a garrison powerful enough to make the city secure against any sudden Confederate thrust. The arrangements by which McClellan made provision for these two requirements while he also moved his army men to the lower Chesapeake were necessarily intricate, and Kernstown deranged them; revising them, McClellan invited trouble and speedily got it.
Departing for Fort Monroe, McClellan sent Secretary Stanton a tabulation of his strength, pointing out that he was leaving Banks with 35,000 men to hold northern Virginia, and that General Wadsworth, commanding in Washington, had 20,000 more. The total thus remaining to meet President Lincoln’s stipulation was, accordingly, 55,000 men, which struck McClellan as ample and was only 10,000 short of the number which the council of corps commanders had fixed as proper to defend the capital.7 Accordingly, on April 1, McClellan himself left for Fort Monroe, and Stanton took the figures to the White House to show them to the President.
When the President and the Secretary of War examined the figures they felt that there were serious holes in McClellan’s arithmetic. The 55,000 who were to hold northern Virginia had originally been figured as being in addition to General Banks’s corps. Now it developed that this corps made up the larger part of the whole, and Banks had nearly all of his men over in the Shenandoah, moving down west of the Massanutten Mountain far from the vital area east of the Blue Ridge. To be sure, McClellan had told Banks to bring everybody but Shields back to the vicinity of Manassas “the very moment the thorough defeat of Jackson will permit it,” but that moment had not arrived; as far as Mr. Lincoln could see, the region which he had ordered held in force was hardly being held at all. To make things worse, the 20,000 left with Wadsworth included some levies that were already ticketed to go elsewhere and in any case were made up largely of untrained troops, imperfectly equipped. Joe Johnston was known to have his army somewhere below the Rappahannock, and it struck Mr. Lincoln that what had been done would present to Johnston “a great temptation … to turn back from the Rappahannock and sack Washington.” He had accepted the plan McClellan gave him, but it seemed to him now that “that arrangement was broken up and nothing was substituted for it.”8
It seemed, in short, that McClellan had simply disobeyed clear orders, and the President acted promptly. On April 3 he ordered Stanton to have either McDowell’s or Sumner’s corps held in front of Washington, to operate in the Manassas area. Since Sumner’s corps was already on the move, Stanton chose McDowell, and McClellan was notified that this corps-33,510 men present for duty, by the latest returns—was detached from his command and would not join him on the peninsula. McDowell henceforth would get his orders not from McClellan but from the War Department.9 As far as McClellan was concerned, McDowell was an independent operator, and the Army of the Potomac had just lost a quarter of its strength.
… which might not have been so serious if anyone at army headquarters had been able to say how strong
the army really was, either before or after this loss. But it is just here that one begins to encounter that fantastic uncertainty about numbers which was to hang over the Army of the Potomac like a fog too heavy for the winds to lift. McClellan’s headquarters was handicapped by a singular inability to determine the size either of this army or of the army which it was about to fight: a shortcoming which made victory impossible and which bewildered no one as long or as profoundly as it bewildered the commanding general himself—with whom, indeed, much of it originated.
On April 7, three days after he had begun to move up the peninsula, McClellan was unable to come within 17,000 of stating the number of men he actually had with him. He told Brigadier General John E. Wool, the white-haired veteran of the War of 1812 who commanded at Fort Monroe, that he had just 68,000 men present for duty; on the same day he sent to President Lincoln a telegram stating that “my entire force for duty only amounts to about 85,000.”10 To be sure, this need not have mattered much, because the Army of the Potomac just now had a prodigious advantage over its opponent. When McClellan’s divisions marched up from Fort Monroe there were fewer than 15,000 Confederates on the scene (fewer enemies, altogether, than the margin of McClellan’s doubt about the size of his own army) and more than a week would pass before that number could be increased to any great extent. Yet this was of little help, because army headquarters always credited Confederate commanders with having from two to four times as many men as was actually the case. In addition to fighting the tough Confederates who were physically on the scene with loaded weapons in their hands, the Army of the Potomac had also to contend with scores and scores of thousands of enemies who never existed. As a result, it tended to move very slowly.
The business began with the Army intelligence section—Secret Service, as it was known then. McClellan had confided his Secret Service to a man who was carried on the books as “E. J. Allen,” but who in real life was Allan Pinkerton, the first and most famous of America’s great private detectives.
Pinkerton was a diligent detective and a first-rate organizer, and he set up a network of spies, messengers, and observers all over Virginia, submitting to McClellan periodic reports which were highly convincing because they contained such a wealth of detail. Running through each report that went to McClellan would be revealing thumbnail comments: “coffee getting scarce … plenty of lead … salt scarce … a good supply of tents and camp equipment except camp kettles … plenty of wagons and teams, generally impressed … on an average arriving at Richmond 3 companies daily.” Here would be a note on the number of shooting galleries in and around Richmond and on the extent to which Confederate soldiers used them; there, a remark that Southern regiments averaged from 700 to 800 men in strength; next, a statement that almost all fortifications were built with slave labor. From each area came just the sort of detail that would show that methodical observers had been taking notes on the spot.
With these comments, of course, came the estimates of troop numbers. These were based on the reports from the Pinkerton operatives, on the examination of Confederate prisoners and deserters, and on things said by contrabands, the figures finally presented being drawn up by Pinkerton himself. Pinkerton carefully worked out percentages so that the totals assembled for any unit could be reduced to a proper present-for-duty level, following his percentages so faithfully that now and then this or that Confederate general would be credited with having “6,346⅔” men in his division. When he drew up his grand totals Pinkerton arbitrarily raised the numbers slightly, on a system previously discussed with McClellan, to make certain that there was a margin for error to cover new arrivals or units that somehow had been missed.
He gets credit, nowadays, for having been worse than he was, and some of his estimates indeed were grotesquely unreal—such as an autumn report that Beauregard’s command contained 100,000 men, and that the Confederates had thirty-three regiments on the peninsula. But at times his carefully calculated totals were fairly close to the mark. On November 15, 1861, for instance, Pinkerton worked it out that there were in Virginia, from Norfolk all the way to the western mountains, approximately 117,100 armed Confederates. The official returns for December 31 (six weeks later, in other words) show an “aggregate present” for all Confederate forces in the state of 118,306. On the face of it, Pinkerton late in 1861 was keeping very fair track of Confederate numbers in Virginia.11
Yet the result was disaster. The “aggregate present” figure, which came so close to the Pinkerton total, actually had very little relation to the number “present for duty”: the number, that is, that would be of use in battle. It was always substantially higher, including sick men, men under arrest or on noncombat details, men from disorganized units awaiting reassignment, men without weapons—all the multitude of military extras who had to be fed, paid and reported on but whom no foe would ever have to face. Even when he came closest to accuracy, Pinkerton made a paper army look real. In addition, his reports got worse instead of better as time went on, and his estimates of the numbers McClellan would have to face finally lost all touch with reality. In the end Pinkerton was persuaded that the Confederacy had between 100,000 and 120,000 soldiers on the peninsula, and that their available forces around Richmond came to more than 180,000.12
These wild guesses would have done less harm, however, if there had not been at army headquarters (where such matters can be crosschecked) a will to believe them. This will McClellan had and never lost. Long after the war, when the truth about Confederate Army strengths in Virginia was clear to everyone, he clung to the belief that he had been beset everywhere by superior numbers: a belief which had no base in fact or in logic but which, if held hard enough, might perhaps justify the paralyzing indecision which governed the direction of the Army of the Potomac.
Yet this indecision was more than the simple result of a belief that the enemy was the stronger. It preceded that belief, displaying itself in a baffling lack of capacity to drive a chosen plan through to its conclusion, and it became visible before the spring campaign was a week old.
On March 19, McClellan had sent Secretary Stanton an outline of his strategic design. He would go up the peninsula to make an advanced base at West Point, where the Pamunkey and Mattapony Rivers meet to form the York, fifty miles northwest of Fort Monroe. Somewhere between West Point and Richmond, he said, the Confederates would concentrate to fight the great, decisive battle, and so it was all-important for him to reach West Point as quickly as possible. About one third of the way up the river was the historic town of Yorktown, where Cornwallis had come to grief, and Yorktown was powerfully fortified by the Rebels. It could be taken by siege, of course, but that would involve a delay of weeks, and there were no weeks to spare; no days to spare, even. At all costs, Yorktown must be taken at once. This, said McClellan, could be done by a joint Army-Navy attack. The Navy “should at once concentrate upon the York River all their available and most powerful batteries.” If it did, Yorktown should fall in a few hours; if not, the business might take weeks. Since speed was essential, McClellan insisted, full naval co-operation was “an absolute necessity”; Yorktown was the key to the entire campaign.13
What McClellan was talking about, of course, was an operation after the Fort Henry model—a pulverizing naval bombardment, with the Army coming in when the dust settled to mop up pockets of resistance and take full possession. The plan was definite enough and it made perfectly good sense—and nothing was ever done to put it into effect.
McClellan appears to have assumed that the President or the Secretary of War or somebody would tell the Navy that its warships were supposed to reduce the fortifications at Yorktown, but nothing of the sort happened. Assistant Secretary Fox insisted afterward that the Navy had never been asked to bombard Yorktown, and he added that the Confederate works there were so strong, and were situated on such commanding ground, that they could not have been reduced by naval gunfire anyway. Flag Officer Louis Goldsborough, commanding at Hampton Roads, understood that Washi
ngton simply wanted Monitor to keep Virginia away from the transports and the disembarkation area around Fort Monroe. Goldsborough talked to McClellan the morning the general reached Fort Monroe, and McClellan said nothing to him about any bombardment. Instead, he asked the Navy to help in the reduction of a Confederate fort at Gloucester, which lay on the north side of the York just opposite Yorktown.
Between Yorktown and Gloucester the York is only 1000 yards wide, and with forts in both places the Confederates had the mouth of the river firmly closed. McClellan told Goldsborough that he wanted to land troops on the banks of the Severn River, a few miles north of Gloucester. Gloucester then could be stormed from the rear, and if that was done nobody would need to bombard anything. If Gloucester fell, Yorktown might fall also; or, at the very least, warships would be able to enter the York and harass the flank of the Confederates on the peninsula. At any rate, Goldsborough assigned seven gunboats to help with the Severn River operation and believed this was all the Army wanted. Apparently he was right; on April 3 McClellan notified Stanton that he had talked to the flag officer and was confident that the Navy would crush Virginia if the ironclad came out, adding that he hoped to advance the next day and that “my only trouble is the scarcity of wagons.”
The Federal advance began, as anticipated, on the morning of April 4, and it was keyed neither to a bombardment of Yorktown nor to the capture of Gloucester but to the belief that the troops could simply force their way past Yorktown and isolate it by getting into its rear. McClellan sent Heintzelman and the III Corps straight up to Yorktown to pin the Confederate garrison in the fortifications there, while Keyes and the IV Corps swung to the left and headed for a place known as Halfway House, four and one half miles beyond Yorktown on the road to Williamsburg, at the narrowest part of the whole peninsula. Once Keyes reached Halfway House, Yorktown would be cut off and must fall. As his army reached the Confederate outposts McClellan sent word to Stanton: “I expect to fight tomorrow, as I shall endeavor to cut the communication between Yorktown and Richmond.”14
