Notes

une 25, 1863, proved to be a fateful day for the Army of Northern Virginia. As Maj. Gen. James Ewell Brown Stuart, commanding the Cavalry Division of the army, sat outside Haymarket, Virginia, and contemplated the long lines of Union infantry heading northwestward, he had a choice to make: a choice that had major ramifications for the fate of the Confederates in Pennsylvania, where Gen. Robert E. Lee's Army of Northern Virginia was heading. With Stuart were the three best cavalry brigades in the Cavalry Division, nearly 5,000 men, and the main force that Lee had come to rely on for the critical screening duties so vital to an army on the march.1 Stuart himself had intended to ride north through Haymarket, around and between the Union columns, in order to meet the van of Lee's army in Pennsylvania. Speed was of the essence, especially if the Union Army of the Potomac was itself in motion northwards, as the Federals marching through Haymarket manifestly were.2 Denied his intended route, Stuart was left with two options: either a further detour to the southeast in hopes of getting around the Union army; or returning to the west to fall in behind the Confederate infantry filing into Maryland at Shepherdstown and Williamsport.3 Either choice would necessitate a delay-the question was which choice would involve less. In the end, Stuart turned the column south and east. That decision took him out of contact with Lee's army for eight critical days.

    Was speed so critical? Were those eight days' absence truly so detrimental to Lee's performance in Pennsylvania, and ultimately so damaging at the battle of Gettysburg? Stuart's defenders are quick to point out that the battle that began on July 1, 1863, was after all an unplanned encounter, one that none of the participants could have predicted on the 24th of June. As one of the more recent works defending Stuart asks, "How can you be late for an accident?"4  In reality, this question confuses cause and effect: Gettysburg happened the way it did because Stuart was absent, not that Stuart was overdue for a predetermined date. The first shots were fired by a group of Confederate infantry who were probing for Union troops-i.e. performing a reconnaissance mission that more correctly belonged to those troopers riding with Stuart. Worse, once joined, the battle was pursued only haltingly from the Confederate side, with Lee urging caution at every quarter. July 1st was in the end a victory for Confederate arms, but essentially an incomplete one. Lee let the major fruits of victory go ungrasped because he simply did not know how much of the Federal army he was facing.

    The reason Lee did not have a working knowledge of the Union army is that Stuart and those three brigades were not there to tell him. Each day that Stuart remained out of contact was another day that Lee's information on the Union movements grew more stale and out of date. Lee was a general who could take the boldest of risks without blinking an eye, as so aptly demonstrated by Chancellorsville just a few weeks before. How uncharacteristic of him then, to be urging his senior commanders not to bring on a general engagement even after the battle commenced at Gettysburg, or to be so reluctant to give approval for A. P. Hill to join in what was clearly a splendid opportunity late that afternoon when two divisions of Lt. Gen. Richard S. Ewell's corps attacked the Union flank, or finally, to refuse to commit any fresh troops to finish the defeat of the Federals even then rallying on Cemetery Hill. Of course, at Chancellorsville Lee had a good idea of Union movements, thanks to the presence of Stuart and his command. At Gettysburg, however, Lee was almost completely uninformed about the any units of the Union army he could not see deployed in front of him on that July afternoon. That uncertainty was the real cost of Stuart's week-long absence. Of course, Lee was prepared for some loss of contact with Stuart. As Maj. Henry B. McClellan, of Stuart's staff points out, " [Lee] was aware that under the most favorable circumstances Stuart must be separated from the army for at least three or four days...... 5  What Lee was not prepared for, however, was an absence of twice that span.
 
 

    "Stuart's Ride" has become one of the pivotal controversies of the Gettysburg Campaign. Detractors almost immediately blamed Stuart for exceeding his orders and costing Lee the battle.6  Almost as rapidly, defenders rallied to the cause to demonstrate that Stuart did not exceed his orders, and in fact had full authority from Lee and Lt. Gen. James Longstreet to undertake the ride. Such seemingly contradictory charges from both camps have muddied the waters considerably, but not irreparably. In fact, to some measure, both camps are right. Stuart certainly had authority to make the ride, but in choosing to do so he failed to fulfill the other aspect of Lee's intentions: to screen Ewell's Second Corps as it progressed through Pennsylvania, and to protect what would be the entire army's right flank once Lee brought the other two corps into the state as well.

    Any starting point of an objective analysis of this situation, then, can only be the orders themselves. On the 22nd of June, Lee sent the following missive to Stuart:

Millwood, June 22, 1863- 7 p. m.

Maj. Gen. J. E. B. STUART,

Commanding Cavalry:

GENERAL: General Lee has inclosed to me this letter for you, to be forwarded to you, provided you can be spared from my front, and provided I think that you can move across the Potomac without disclosing our plans. He speaks of your leaving, via Hopewell Gap, and passing by the rear of the enemy. If you can get through by that route, I think that you will be less likely to indicate what our plans are than if you should cross by passing to our rear. I forward the letter of instructions with these suggestions.

Please advise me of the condition of affairs before you leave, and order General [Wade] Hampton-whom I suppose you will leave here in command-to report to me at Millwood, either by letter or in person, as may be most agreeable to him.

Most respectfully,

Finally, the last pertinent written communication we have on the subject is Lee's order of the 23rd-.

    All three orders, then, contain essentially the same information. Lee's instructions stressed the need to join Ewell and screen his (and what would become the army's) right flank, and do so quickly. The two surviving orders are clear on this, and it can be inferred that the last letter McClellan recalled said the same, since a radical departure from previous instructions would have been noted by the major had it been specified, nor would there have been any need to discuss the likely location of Ewell's right flank had the mission changed. Finally, the screening mission is mentioned clearly in both sets of orders, without any conditional language. This sentiment is especially strong in the second order, where Lee states "In either case, after crossing the river, you must move on and feel the right of Ewell's troops, collecting information, provisions, &c. [emphasis added]" There isn't much uncertainty there about Stuart's objective. Instead, the conditional aspects crop up only in discussing routes, where Lee clearly gave Stuart full freedom and discretion in moving north. This latitude has been consistently misconstrued as ambiguity or confusion, but in fact represents only realism: Nineteenth Century armies lacked adequate communications sufficient to specify rigid planning or micro-management from army headquarters, especially when dealing with events likely to occur more than a day or two in the future. Events changed too rapidly to allow modifications to those plans as they unfolded, and this was never more true than in cavalry operations. Lee was simply granting Stuart latitude to accomplish his mission, and it was Stuart's responsibility to ensure that the objective-joining Ewell and screening the right flank-was fulfilled. Ironically, in modern military parlance, Lee's orders would be termed "mission" orders, a concept very much endorsed by the Maneuver Warfare enthusiasts who where behind many of the 1980's U.S. Army reforms, and in that sense, are not unclear at all. 13

    Nor was Stuart's plan without merit.14  In fact, it offered a number of advantages. First, Stuart's command would not have to share road space with the Confederate infantry, and hence avoid delay. Second, both Lee and Longstreet agreed that it would be more confusing to Union commanders, and less revealing of Confederate intentions, if Stuart were in the Federal rear rather than simply following Longstreet's and Hill's infantry corps into Maryland.15  Third, Stuart's men could both forage in virgin territory and have the simultaneous opportunity to disrupt Union communications and supplies. Lastly, such a significant Rebel cavalry presence in the rear of the Army of the Potomac might well divert much larger Union mounted forces to cope with Stuart's threat, weakening the Union's own reconnaissance screen facing the Army of Northern Virginia. All of these potentially rewarding results were secondary, however, to the principal mission: a timely arrival in Pennsylvania to meet Ewell and screen what would become the entire Army of Northern Virginia's exposed right flank. This is where Lee granted Stuart the greatest latitude, in allowing the cavalry general to reach his own conclusions as to the best route and what might constitute excessive delay. Nothing speaks more clearly of the absolute trust Lee placed in the young Jeb Stuart.

    There is one other aspect of Lee's intentions that is important to consider. Lee's orders to Stuart assigned him to Ewell's flank, and at the time they were issued, Ewell's and the rest of the Army's right flank were not the same thing. Ewell was in Pennsylvania, and, on June 26, started columns towards Carlisle, Harrisburg and York, while the other two Confederate infantry corps were either at Chambersburg or closing rapidly having just crossed the Potomac into Maryland. Lee's sending of Stuart to Ewell, however, reveals two thin-s about the army commander's intentions and concerns. first, that Lee felt he had a comfortable jump on the Army of the Potomac for what would be the next few days, since he was unworried about the passes through South Mountain in Maryland that had been of such concern the year before during the Antietam Campaign. The timing inherent here is interesting too. Stuart left the army on June 25th. Lee personally crossed into Pennsylvania on the 27th, and did not have the remainder of the army fully concentrated around Chambersburg until the 28th or 29th: a four-day window of transit time, and providing us an idea of the time Lee probably expected Stuart to be out of contact, and coinciding neatly with McClellan's previous estimate of how long Stuart was likely to be gone, cited above.

    Secondly, once in Pennsylvania, the Army of Northern Virginia would be close enough to Ewell's advance corps that Ewell's right flank would essentially be the Army's right flank as well, so that when Stuart was in place, he would again be interposed between the Army of the Potomac and Lee's own main body of troops. A glance at a map brings this into focus: if the Confederate Army is located on an arc drawn between Chambersburg, Carlisle, and York, then the critical ground to cover would be Fairfield, Gettysburg, and Hanover, which would place Stuart's men directly in the path of the Federals as they came north. Once completed, the separate movements of Ewell, Stuart, and Lee's main body would be reunited in southern Pennsylvania, again acting in concert and within mutually supporting distance. If properly executed, this was a stratagem both deft and daring.
 
 

     Stuart spent the day of June 24th assembling his three brigades at Salem, and in sending detailed instructions to Brig. Gens. William E. "Grumble" Jones and Beverly H. Robertson, outlining the tasks of those two commands in his absence.16  Since these two units became important to the controversy after the fact, we shall discuss their mission and Stuart's instructions to them in more detail later. Suffice to say, however, that these two commands, 2,700 sabres strong, were left behind to screen the Army's "right and rear" during Stuart's intended three-to-four day absence.17   With these details dealt with, and with the three brigades comprising the raiding force assembled, the column departed Salem some time after midnight of the 24th of June on the first of many night marches over the ensuing weeks. 18
    Certainly Lee expected to lose contact with Stuart for several days. Even with the Army of the Potomac completely inert, Stuart could not hope to cover the intended distance in less than three to four days, all the while deep in the enemy's rear and without any means of regular contact with Lee. As Nesbitt rightly points out:

    Union interference was found right from the start. At Thoroughfare Gap, the column found Federal troops on the march, using roads Stuart had intended to use himself, and the Confederates detoured five miles to the south via Glasscock Gap.20  Worse yet, upon reaching Haymarket, on the east side of Thoroughfare Gap, Stuart met what turned out to be the Union Second Corps, marching north. After a brief skirmish with the Union infantry, Stuart turned his column back to Buckland, where he halted for the day and sent reconnaissance patrols to determine an enemy-free route.21  This delay proved deadly to Stuart's schedule. With Buckland less than fifteen miles distant from Salem, Stuart had covered a mere fraction of his intended distance in his first full day of movement.

    Stuart later wrote in his report of the campaign that he had originally intended to stay west of Centreville, but with Maj. Gen. Winfield S. Hancock's Union troops there in force, he was compelled to detour south and east, crossing the Occoquon Creek at Wolf Run Shoals, and then to ride via Fairfax towards the fords of the Potomac.22  Not only did this add an extra twenty miles to the route, but it necessitated another halt for grazing and rest. The evening of the 26th found Stuart's main force still at Wolf Run Shoals, farther than ever from the Potomac.23  From the standpoint of timing, if the first day's delay was bad, this was a disaster. At the Occoquon, Stuart was still a full day's march from the Potomac, and already 36 hours behind where he would have been if he had been able to simply move directly to the Potomac from Haymarket, as originally intended. Worse, all of the 27th would be consumed in just getting to and crossing the Potomac into Maryland. The first Confederate troopers would not even set foot on Maryland soil until 3:00 a.m. on the 28th, where, after an arduous night crossing, another halt was needed .21 Stuart, going by his intended route, should have been in Maryland by the evening of the 25th, after a march of about 40 miles, the distance from Salem to Seneca Ford. Instead he had marched more than 60 miles as well as lost most of a day reconnoitering outside of Gainesville and Haymarket, which meant that he was two full days behind schedule before he even left Virginia.

    The march proceeded little better once into Maryland. While the main body of the Union army had passed, time was still lost in encountering the infamous wagon train outside of Rockville, which resulted in the capture of 125 wagons mostly loaded with fodder and forage.25  These wagons became the focus of most of the criticism of Stuart after the war, but in reality they were likely a less critical delay than simply getting out of Virginia had been. The morning of the 28th found Stuart's vanguard at Cooksville, where they halted and damaged the Baltimore & Ohio Railroad .26  It was here, during the 28th and 29th, that Stuart devoted the most effort to fulfilling Lee's secondary objectives. Numerous prisoners were paroled, mostly from the captured wagons. Stuart's men interrupted traffic on the railroad for most of a day, and managed to cut many telegraph lines as
well .27

    Still, time was slipping away.  June 29th saw Stuart close to the Pennsylvania border, fighting his way through Westminster after yet another skirmish with Union forces. The column halted for the night just north of Westminster at Union Mills for further rest and to pasture the horses .28  June 28th and 29th each saw solid, successful marches of 20-25 miles, but now a different problem had arisen. The last word   ' Stuart had heard of Ewell's intended whereabouts was Lee's message of late on the 23rd, in a dispatch now missing but recalled by Major McClellan as described above. This message outlined Ewell's probable location as York, Pennsylvania, but that information was now six days old. Subsequent events had rendered this information obsolete by the 30th, the day Stuart at last crossed into Pennsylvania.29  Of course, no one should have been surprised at this: both armies were moving, and six days is a long time in a fluid campaign situation. Now, however, Stuart was too far to the east.

    June 30th found the Confederate cavalry at Hanover, clashing with a strong Union cavalry force in and about the town, and the ensuing action used up most of the day.30   Stuart was left with no choice but to sidestep eastwards, and more importantly, he had heard nothing useful in Hanover to update Ewell's whereabouts. That night, Stuart began a hazardous night march to Dover, where he hoped at last to find solid information leading to Ewell or Early.31  The march to Dover proved extremely difficult, since neither the cavalry nor the draft animals had been grazed, rested, or watered; and McClellan's description of it paints a vivid picture of troops on the frayed edge of exhaustion.32   Here, at least, the captured wagons proved more trouble then they might have been worth, and should have been destroyed. Unfortunately, while Stuart's men eluded the Union troops at Hanover, their arrival in Dover brought word only that Early's division of Ewell's corps had been called to rendezvous with Ewell west of South Mountain.33 Stuart decided to go to Carlisle, hopeful of finding Ewell, and with him supplies and a chance to rest.34

    Stuart's command was essentially groping blindly for the rest of the army at this point. From Dover, Stuart sent two couriers-Maj. A. R. Venable of his own staff, and Capt. Henry Lee of Fitz. Lee's to Shippensburg and Gettysburg respectively, in hopes of locating Lee's men. Both couriers were instructed to bring back word to Stuart at Carlisle, where the main body was headed after a short pause for rest.35   Carlisle, however, proved no better a destination. Late afternoon on July 1st found the Rebel cavalry on the outskirts of town, only to find Ewell gone and the town occupied by Union troops, most likely militia.36   With his men exhausted by the almost continuous movement since the 30th, however, Stuart wanted the town for the food and rest it represented, and so ordered Brig. Gen. Fitzhugh Lee to capture the place. Fitz. Lee's demand that the town's surrender was curtly refused by the Union commander, Maj. Gen. William F. "Baldy" Smith. Lee then proceeded to shell the town with artillery.37   About 1:00 a.m. on July 2nd, that effort was suspended, however, because Stuart had at long last received definite word from Gen. Robert E. Lee. Major Venable had returned, carrying sealed orders from Lee describing the July 1st encounter at Gettysburg, and ordering Stuart to join him there via the towns of Hunterstown and Heidlersburg.38   By 2:00 a.m., having directed his brigades to head south, Stuart himself set out to ride directly to meet Lee at Gettysburg. He arrived that afternoon, the 2nd of July.39   Contact between Lee and his cavalry was at last restored.

    In retrospect, those eight days with no contact between Stuart and Lee proved very damaging indeed. Stuart's lost time was not due to any single cause, but the fact remained that he was at least four days longer in transit than anyone had expected before the march commenced. Certainly the most critical delay happened the day he set out. Not being able to move via Gainesville and stay west of Centreville cost Stuart two full days right at the start. Had he chosen to give up the overly ambitious goal of circumnavigating the Army of the Potomac at that point, there might still have been enough time to choose a different route and join the Confederate main body without too costly a delay. Of course, two of Stuart's staunchest supporters, John S. Mosby and Major McClellan, take some pains to explain why Stuart had no recourse but to turn south to Wolf Run Shoals, but their arguments fall short on several grounds. Mosby claimed that Lee's orders allowed Stuart to either cross the Potomac via Shepherdstown, or cross east of the Blue Ridge and behind Hooker's army.40  Furthermore, he claimed the only route to Shepherdstown was roundabout, through rough terrain, and crowded with Confederate infantry.41   Manifestly, all these things were true, but they were also true for the route Stuart adopted. The path via Wolf Run Shoals was certainly not direct; nor was the terrain, with crossings of the Occoquon and the Potomac ahead, an easy jaunt. Finally, there was the matter of infantry potentially blocking this route as well-only it wore blue, not gray, and could be considered a lot more difficult to navigate around than friendly troops.

    Major McClellan's arguments are similar. Focusing specifically on the night of the 25th at Buckland, McClellan states:

    Both men completely ignore what would have been the most critical benefit of Stuart choosing to turn back to the west and seek an alternative to riding around Hooker's Federal army: the chance to remain in much closer contact with Lee's headquarters. Traveling via Shepherdstown would have allowed Stuart to remain in constant contact with Lee, and to adjust his mission as circumstances changed. When Stuart arrived in Hanover he was relying on information that was six days old to find Ewell/Early: had he been at South Mountain on the 28th, as McClellan speculates, his knowledge of the Confederate Second Corps' position would have been no more than a day or two old. At least a day in fruitless searching would have been saved, including the two most difficult marches of the entire ride: those to Dover on the night of the 30th and to Carlisle on the Sit.

    Finally, a direct route from Sharpsburg to Ewell's area of operations around Carlisle and York would take a column directly through the Gettysburg-Hanover area. Had Stuart's troopers been operating in that area in the critical period of June 29th to July 1st, they would have been directly in the path of any Union advance and covering Lee's most exposed point, instead of behind the Army of the Potomac and struggling to catch up. Even incidentally, the information of Union movements they could have provided Lee would have been invaluable.

    The next most critical delay-that time lost in looking for Ewell's men once the Rebel cavalry arrived in Pennsylvania-can also be largely attributed to that initial 48-hour period of the 25th and 26th. As previously eluded to, Stuart's information of Ewell's intended whereabouts was, by June 30th, almost a week old. Not surprisingly, Early's division of Ewell's Second Corps which was in York from June 28th until early on the 30th-had been forced by rapidly evolving- circumstances to move on." Stuart reached Hanover, by 10 a.m. on the 30th of June, but it still took until the afternoon of July 2nd before Wade Hampton, John R. Chambliss, and Fitz. Lee's brigades securely rejoined the main body of Lee's army.48   Even worse, this was the most grueling 48-hour period yet, with little rest and extensive marching. When the column did rejoin the army, it was badly in need of sleep and forage, which can only have negatively effected the efficiency of the mounted arm over the remainder of the battle. In essence, by arriving even a day earlier, Stuart would have linked up with Early and avoided much of the ensuing groping for Lee's army.

    Lt. George W. Beale of the 9th Virginia Cavalry, in a letter home, gives us a look at the mindset of these exhausted troopers at Carlisle:

    Interestingly enough, the wagons captured outside of Rockville proved the least delay. Even McClellan faulted Stuart for not abandoning the wagon train by the 29th .47  Certainly the wagons proved a serious hindrance, but the column still made good time when it needed to, and in one way, the wagons helped increase its speed. Among the provisions captured in the wagons (and at Westminster) was grain for the horses-a significant capture because of the high-energy, quick-feeding nature of grain versus grazing. Capt. William W. Blackford's comments are especially informative here: "Here was a godsend for our poor horses, for every wagon was loaded with oats intended for Meade's army and it did one's heart good to see the way the poor brutes got on the outside of those oats."48   Grain could be fed rapidly, and a pound of grain was worth several times the nutritional value of a pound of fodder gained from grazing."

    McClellan noted that prior to Westminster, the column halted several times to graze, which required both time and space, forcing the column to disperse quite a bit. 50   Without the train, McClellan estimated that the column would have been past Hanover by the 30th, thereby avoiding the encounter with Brig. Gen. H. Judson Kilpatrick's Union cavalry that morning.51   While by this estimate Stuart would have saved a day, certainly some of that time savings would have been negated simply because, without the grain found in Westminster and carried along in those wagons, additional time likely would have been used in more grazing. Furthermore, while the train made forced marches that much more arduous, it did not preclude forced marching altogether, and might well have provided the extra-high energy food the animals needed to even make those forced marches possible and still possess any effectiveness once Stuart did reach Lee's army.

    The 4,500 troopers riding through Maryland with Stuart were not the entire cavalry of the Army of Northern Virginia. Stuart took three brigades with him on his ride. He left four with the army, leaving, as many historians have pointed out, sufficient cavalry for Lee's other needs. More than half of the available mounted men remained behind. However, raw numbers are deceptive, for the commands Stuart left behind were not-with the exception of "Grumble" Jones' brigade-the equals of the brigades he took. Nor were they sufficient in either strength or quality to fulfill both the important tasks set them prior to Stuart's departure and at the same time fill the void left by Stuart's nonappearance in Pennsylvania at the end of June.

    First, let us examine the numbers. On the 30th of June, Stuart's division reported a strength of seven brigades and one artillery battalion, numbering in all, 12,358 officers and men ready for action .52 Stuart took three brigades with him; Chambliss' with 1,328 troopers,  Hampton's command with 1,978 men, and Fitzhugh Lee's brigade of 2,164 sabers, for a total of almost 5,500 men. Since Stuart culled out those who were unfit or too poorly mounted to make the raid, the actual raiding force was closer to 4,500 men.53    Left behind were Albert. G. Jenkins' brigade, 1,274 men strong; Beverly H. Robertson's new command up from North Carolina, 966 men in two regiments; William E. "Grumble" Jones' veterans from the Shenandoah Valley, 1,743 troopers; and John Imboden's mixed command of cavalry, mounted infantry, and partisan Rangers, 2,245 men. 54  In addition, the artillery battalion was divided among the various brigades, with Stuart himself taking one battery and an additional section of the six artillery batteries in Maj. R. F. Beckam's battalion, for a total of six guns. Several of these brigades were not normally members of Stuart's division, notably Jenkins' and Imboden's commands, attached from the Department of Southwest Virginia, and Beverly Robertson's newly recruited brigade brought up from North Carolina.55  Hence, while Stuart likely had a good grasp of the numbers and experience of his regular commands, he did not have such a clear picture of some of these new arrivals. For example, he grossly overestimated the strength of Jenkins' brigade at 3,800 men, more than three times the actual strength .56  This particular oversight was most unfortunate, because Lee certainly knew of that brigade's numerous detachments when it left Southwest Virginia, and by assuming it was much stronger than in reality, Stuart also mistakenly assumed that the brigade had ample strength on hand for any extending scouting duties Ewell might need."

    One of the things to bear in mind when examining the dispositions of the Confederate cavalry during the campaign is the orders we do not have. While a number of specific communications between Stuart and Lee, and Stuart and his subordinates survive, there is a good deal that is only implied. Part of the problem is the informal nature of command in the American Civil War, before the creation of a staff college or TRADOC, with it's creation of manuals and standard operating procedures for all aspects of military operations.58   Cavalry in the midnineteenth century was expected to screen an army's main body, performing the dual role -of discovering the enemy's positions and intentions while simultaneously hiding'one's own. While some manuals on tactics and outpost duties did exist, for the most part, each cavalry commander evolved his own procedures informally, and rarely were these details written into orders the way a modern army would establish tactical doctrine. Certainly this is the case for Stuart's command-we lack formal orders from Lee establishing a full cordon around the Army of Northern Virginia, but we do have strong clues to Lee's intentions, and Stuart's responsibilities. First and foremost, Lee's two surviving orders to Stuart in theOfficial Records (quoted above) both stress the need to screen Ewell and the army's right, a fairly explicit summation of Stuart's intended duties. Stuart himself gives us a fairly detailed glimpse of what he expected of a force charged with establishing such a screen in his very detailed orders to Robertson on June 24th, just before he left the army.59   Second, we have the dispositions of the various cavalry brigades themselves on June 24th, positions which are highly suggestive of Lee's desires, especially when viewed in light of the potential Union threats each brigade faced.

        Simply put, Lee and Stuart established a 360-degree cordon around the Army of Northern Virginia, with cavalry arrayed in the four cardinal directions. While we may not have any single order explicitly stating Lee's desires on this issue, the actual deployments are too obvious to miss. Antoine Lasalle or Joachim Murat, to name two of the finest cavalrymen of an earlier era, would have instantly grasped Lee's intentions from those deployments.

    Since Lee was going north, and looking to slip into Pennsylvania if the chance presented itself, let us go there first. Ewell's corps led the way, and with it on the leading edge was Albert Jenkins' brigade. Jenkins' command was new to the army, fairly weak in strength at less than 1,300 men, and of dubious quality. But these potential problems were offset by the fact that Ewell was ahead of the Army of the Potomac, and faced only hurriedly-raised Union militia-Jenkins' command was not likely to be tested against the veterans of Hooker's army. Certainly Ewell and his division commanders did not view their cavalry support as first-rate. A week prior to the entry of Maj. Gen. Robert Rodes' division of Rebel infantry into Pennsylvania, he had witnessed the results of Jenkins' first foray to and across the Mason Dixon Line. On the 13th, in engagements first at Berryville and then at Martinsburg in the lower Shenandoah Valley, Jenkins' men proved both clumsy and slow, alerting Union outposts prematurely and allowing the temporary escape of the Union garrisons.60   By the 17th, Jenkins had captured Chambersburg, Pennsylvania, and secured a considerable amount of supplies-mostly foodstuffs-from the surrounding countryside. However, that morning, rumors of a large unidentified Federal force were enough to scare Jenkins into ordering a retreat without a shot fired, and his command retired into Maryland in some disarray. Rodes was even more disgusted when he later discovered that the "force" that had caused Jenkins' abrupt departure from Chambersburg was nothing more than a thirteen-man Union patrol, their numbers no doubt exaggerated by excitable civilians.61   Nor, unfortunately, did the situation improve once Ewell was recalled from in front of Harrisburg by Lee's urgent order. On the 29th, Ewell issued orders for his division commanders to retreat from their exposed positions and retire to Cashtown Gap, but Jenkins did not receive those orders until a day later, which at least one historian felt seriously effected the course of events on the I st of July.62

    In fact, the events of June 30th through the evening of July Sit graphically illustrate how overcommitted Ewell's available mounted arm really was. One regiment, the 17th Virginia Cavalry, had accompanied Early to York, and retired with that division after Ewell's recall order reached
him .63 Thus Early had at least one regiment of cavalry with him on the way to Gettysburg that morning, while the rest of Jenkins' command was kept occupied screening Ewell's rear from a perfunctory (but still potentially dangerous) pursuit launched by the Union commander of Harrisburg, Maj. Gen. Darius N. Couch .64   Fortunately for all concerned, Couch broke off pursuit on the 30th after only 5 or 6 miles, and no real threat to Ewell's rear materialized. None the less, even that minimal effort required the attention of at least some Rebel cavalry, and Jenkins' men were spread very thin .65  In addition to Jenkins screening his rear and the 17th Virginia with Early, Ewell did have one other force capable of proper cavalry duties: E. J. White's 35th Virginia Battalion of Cavalry, otherwise known as White's Comanches. White's men had marched north with the corps from Culpeper in early June, doing good service along the Potomac-including a daring feat at Point of Rocks in Maryland on the 17th-and now joined Ewell again for his entry into Pennsylvania. 66   While White's men proved to be good troops, and led Early's division's advance to and through Gettysburg on June 26th, their numbers-less than 300 men-were too few to handle the scouting duties of an entire corps spread between Carlisle and York.67   As things stood on the evening of the 30th, Ewell had 1,536 mounted men-the 1,274 men in Jenkins' brigade and the 262 men of White's Comanches-to screen the movement of his entire corps and protect its rear as it broke contact along the Sesquehanna River .68   Of course, according Lee's original plan, Stuart's cavalry would have joined Ewell by the 29th or so, adding three of the best cavalry brigades in the army to the mix. This would have provided more than adequate mounted forces for these tasks.

    To the west Lee sent his single largest brigade, the mixed command of John Imboden. Imboden's men were another command only recently with the Army of Northern Virginia, also on loan from Sam Jones' Department of Southwest Virginia. Imboden's troops were another unit that was more used to raiding and guerrilla action than the tedious but critical work of outposting for an army on the move. Imboden's men were also fairly green. The 18th Virginia Cavalry had only been raised in December of 1862, and though large numbers of those men had seen prior service with the 62nd Partisan Rangers, they were unfamiliar with each other as a un it.69   The 62nd Virginia Mounted Infantry was the infantry component of the 62nd Partisan Rangers, who, when the 18th Virginia Cavalry was formed, became the 62nd Virginia Infantry. Subsequently, this regiment was mounted in time for the Gettysburg Campaign, but, given their previous service on foot, had obviously not had time to learn much about the special skills cavalry operations require.70   Fortunately, however, any threat from the west was remote, as it was deep in the Pennsylvania mountain country, with the nearest Union troops in West Virginia. Imboden's primary responsibility was to gather supplies, destroy as much of the Baltimore & Ohio Railroad as he could, and range deep into the mountain country, to provide an early warning should a threat develop .71

    Lee himself had a number of doubts about Imboden's men in a traditional cavalry role. As late as July 9th, he called them unsteady, and expressed a desire to see them dismounted and converted to infantry.72   Of course, Imboden's fine performance on the retreat to Williamsport dispelled much of this worry. Even so, that duty was much different than the screening and outposting that Lee expected of Stuart, and further illustrates why Lee sent this command to the place of remotest threat to his army. By July I st, they were ordered to Chambersburg to replace Pickett's division, who had been detailed to defend the army's main supply trains, and it was not until midday on July 3rd that Imboden's brigade actually reached the field at Gettysburg .73

    Covering the northern and western approaches to his army, Lee had committed two of his seven available mounted brigades, and as we have seen, those two were the least effective of the cavalry forces he had and so were facing the least significant threats. Covering the southern and eastern arcs, however, was a different matter. Hooker's Union army remained southeast of Lee, and could approach either the Army of Northern Virginia's flank or rear at will.

    Likely the most misunderstood mission of the campaign was that jointly held by Brig. Gens. Beverly H. Robertson and William E. "Grumble" Jones. Together, these two brigades-with a combined strength of about 2,700 men-were left to screen the rear of Lee's army and to protect the vast quantity of supplies that were heading southwards into Virginia as the campaign progressed. These two men have been the most likely scapegoats in the postwar accounts for not replacing Stuart, but a close examination of their own duties and actions shows that to have done so would only have meant neglecting their own critical mission .74

    These two commands were also of mixed quality. Robertson commanded two regiments, the 4th and 5th North Carolina, both of whom were inexperienced. These two regiments were formed in North Carolina in the fall of 1862, but did not join the Army of Northern Virginia until June of 1863: Brandy Station was their first significant action.75  Robertson himself was not held in high esteem by most of the army's officers. He was considered a good instructor but deficient in actual combat situations. Blackford's opinion of him was probably the common one: General Robertson was an excellent man in camp to train troops, but in the field, in the presence of the enemy, he lost all self-possession, and was perfectly unreliable."76

    Jones' men, on the other hand were considered very good troops, and Jones himself was by some accounts the "best outpost officer in the army."77  Certainly, Jones  was expected to bolster Robertson in case of trouble. It is also likely that Stuart left these two men behind because of strained personal relations with each of them arising from their previous service together with the Army of Northern Virginia. No doubt Lee felt that leaving them behind, on what amounted to an independent mission, would create the least difficulties in day-today operations. Certainly this force, especially with the inclusion of Jones and his fine brigade, gave Lee a competent cavalry force to his rear.

    Just prior to his departing the army, Stuart gave detailed orders to Robertson, as senior of the two brigadiers, describing the mission envisioned for these men. Given the controversy these instructions caused, they are also worth quoting in full:

    The first significant charge lodged against Robertson is that he failed to keep Lee informed of the Union army's advance across the Potomac and into Maryland.80   Certainly Lee was aware of Union activity in crossing the Potomac. He was alerted to the erection of the Union pontoon bridge at Edwards Ferry, near Leesburg, by the 23rd, since Lee mentions it in his own letter to President Jefferson Davis of that date, and further makes it clear he expects the Union army to cross momentarily.81   It can hardly be said then, that Lee was uninformed of the Union movements north. Professor Edwin Coddington neatly summarizes the question that virtually every historian and participant has since asked: Why did Robertson and Jones remain stationary in the lower Valley for five days, and why did Lee permit this seeming inaction?82

    For permit it Lee did. Robertson did not publish much by way of explanation of his actions during the last week of June, but did rather hotly defend himself in Battles and Leaders, where he insisted that he was in daily communication with Lee during the entire period, and fulfilled all that Lee expected of him. As proof he cited the fact that Lee offered not a single word of criticism in his [Lee's] official report.83  Mosby, in his own analysis of the actions of Stuart's cavalry, speculates that orders must have passed between Lee and Robertson that subsequently countermanded Stuart's instructions of the 24th, to explain this apparent approval on Lee's part. 84  Far from being out of contact for days, then, Robertson was instead apparently right where Lee wished him to be.

    The reason for that is likely found in what is perhaps the most under-discussed aspect of the campaign: Lee's need for supplies. While Lee's own final reports on the campaign do not address the issue in detail, virtually all his orders and messages to subordinates during the campaign show a great deal of emphasis on gathering supplies. The problem was that by 1863 it was increasingly difficult for Lee's army to subsist in Virginia, ravaged as it was by war, and supplies from other regions could only come via the increasingly unreliable railroads.85   In short, Lee was being forced to make serious compromises with his ability to maneuver and fight in order to serve a strained logistical system. Nothing more aptly illustrates that need than the detachment of Longstreet and two divisions of infantry to North Carolina that spring to help gather supplies, a mission Lee deemed vital enough to risk fighting the battle of Chancellorsville without Longstreet's men present. 86

        It was critical, then, for the Army of Northern Virginia to leave central Virginia for two reasons. First, to give local agriculture some chance to recover and harvest for the coming winter, and second, to gather as much as possible from elsewhere so as to offset many of the expected shortages. Hence the desire to move the Army of Northern Virginia north, into Pennsylvania if possible, with its fertile fields and well-stocked farms. The message traffic between Lee and his subordinates reveals this concern in detail. Virtually every order from Lee contains some discussion of supplies, and on June 21 st, Lee issued General Orders Number 72, which discussed the methods of taking supplies in detail to all commands.87   By the end of June, the results of this emphasis could be seen quite clearly, as large herds of animals and many thousands of pounds of foodstuffs were heading south. For example, Professor Coddington mentions that as many as 26,000 cattle and 22,000 sheep were taken out of Pennsylvania in June. All sent south down the valley and collected near Mount Jackson, Virginia. 88  Even if exaggerated, the numbers were still significant: Lee was sending vast herds back to Virginia that represented an investment in the army's future survival.

    It is all the more curious, then, that historians-even Coddington, himself-discount the need for Jones and Robertson to remain in Virginia. Coddington remarks: ". . Lee's fears for his communications were not as all absorbing as they ordinarily would have been, and should the Union army got athwart them, the effect would have been inconvenient rather than fatal."89   While this statement was true enough in that Lee was not receiving a steady supply flow from Virginia, it discounts the critical flow to Virginia, which, when viewed in light of the rough winter just passed, certainly loomed significant in Lee's thoughts. From that perspective, the retention of these two cavalry brigades near Berryville makes perfect sense: they were positioned to screen that supply flow up the length of the Shenandoah Valley.

    After all, the departure of the Army of the Potomac from Virginia did not entirely remove that threat to Lee's supply lines. A substantial Union force remained to menace those supplies so assiduously gathered up north, in the form of the 10,000 Union troops under Maj. Gen. William H. French. Positioned at Harpers Ferry, these troops remained in place even after the Army of the Potomac had moved on. While French's troops were ordered to leave Harpers Ferry for Frederick, Maryland, on the evening of the 28th, the care with which French removed stores and withdrew from that exposed place meant that there were as many as 3,000 Federals on Maryland Heights as late as July 2nd.90  Certainly French's men did not represent a major threat to Lee's rear, a fact Lee readily acknowledged with the casual way he simply bypassed Harpers Ferry on this campaign. But French's troops were still a force that needed watching, and Robertson's combined force of 2,700 cavalrymen was the only force in the right location to provide that masking. Did Lee's communications with Robertson between the 26th and the 30th of June carry orders to that effect? All we can do is speculate, since no messages seem to have survived, but it is a reasonable speculation that they did.

    Finally, there is one other reason that Lee could not easily redeploy Robertson's men into the void on his right: timing. It is clear that Lee had to anticipate that Stuart would spend several days out of contact with him even under the best circumstances. Stuart's departure on June 25th meant that Lee could not anticipate hearing from him until the 28th or 29th at a minimum, meaning that there would be a delay of that long even before Lee could start to anticipate problems. In fact, Lee did send orders to Robertson on the 30th to join the army, suggesting that Lee only gave Stuart twenty-four hours of leeway before acting to cover the gap in his cavalry screen that now yawned between Lee's men and the Federals.91

    Certainly by the 29th, Lee had begun to express concern about Stuart, and it is significant that on July 1st, when Ewell's aide, Maj. G. Campbell Brown, was sent to Lee to report the arrival of the Confederate Second Corps on the field, Lee's first request of Brown was to know if anything had been heard from Stuart.92   Nesbitt suggests that this concern was more personal than military, given the close relationship between the commanding general and his cavalry commander, but this suggestion is hard to credit. Lee was a professional soldier and a proven commander, and that he would not be concerned about the prolonged absence of three brigades of cavalry in the midst of such a tactically fluid situation, despite the potential ramifications for the rest of his army, belies belief.93   While there may have been a personal note in Lee's concern, certainly much of the worry he exhibited over Stuart during this period was framed by his operational context.

    Even more time was now consumed by moving Jones' and Robertson's commands into place from Virginia to Pennsylvania, and these brigades did not arrive at Greencastle, Pennsylvania, until late on July 1 st.94   From Greencastle they rode into position on the army's right flank by the 2nd, and the 3rd of July found Jones' brigade near Fairfield, where they repulsed a Union cavalry probe.9  Hence, even after Lee called these troopers to him, it took approximately three more days for these two units to move into position and take up the duties Lee required.

    It is this time lag that is the crux of the problem. Lee could not simply, with a wave of his hand, shift Robertson and Jones from their positions in northern Virginia to the army's exposed right flank. Instead, he needed a great deal of lead time to get those units in place. Worse yet, what Lee really needed was accurate intelligence of the Union movements in the days prior to the battle, and it was simply not enough to slide a cavalry force into place at the last minute. Rather, it takes time for a cavalry screen to establish itself effectively, usually a minimum of a day or two to learn the country and begin to assemble a fragmentary picture of enemy movements. Any replacement force for Stuart would have needed to be in place and establishing that screen by June 28th or 29th at the very least-the time frame Lee expected Stuart to have returned to the army. Essentially, for Lee to effectively replace Stuart with Robertson's force, he would have to have sent orders to that effect as early as June 26th, simply based on the distances involved and the normal inertia generated by any Nineteenth Century communications system. In effect, Lee would have had to have anticipated Stuart's failure almost from the minute Stuart left, raising the obvious question then, that if indeed failure was anticipated, why let Stuart go at all? The answer, of course, is that Lee did not anticipate that failure-instead he relied on Stuart's judgment that the cavalry commander could complete his ride successfully and be in place when Lee needed him. Once Stuart was overdue, Lee had no easy answers: even if he hadn't regarded Robertson's mission in Virginia as vital, he was still confronted with the time delay problem that left his army uncovered to enemy probes from the right, and blind to Union movements for several days.

        When viewed as a whole, then, it can be seen that the Confederate cavalry dispositions were carefully made, Lee matched forces of varying quality to the several missions he needed to accomplish, especially with an eye towards deploying the most combat-ready commands towards the greatest potential threats. Jenkins and Imboden-with 1,500 and 2,000 men, respectively, and considered the least effective cavalrymen availablewere sent on missions to the north and west that would likely bring them in contact only with hurriedly raised Union militia, or where, in Imboden's case, deep in the western mountains, a serious Union threat to Lee's flank was remote. Robertson and Jones were both stronger, at almost 2,700 sabers, and better trained for the mission at hand, and so were sent to cover Lee's rear with its vital supply connection for the army's future. While a job with greater potential threat than that to be faced by Jenkins or Imboden, the Virginia flank was not to be the Union Army's most likely avenue of operations as long as Lee was in Pennsylvania. Of course, this left the task with the greatest threat to Stuart himself, and he naturally reserved both the strongest and best Cavalry forces-the 4,500 men of Fitzhugh Lee's, Chambliss', and Hampton's commands-for the critical duty of screening Lee's eastern flank directly facing the 94,000 men of the Union Army of the Potomac. Should any one of these forces fail to fulfill it's mission, replacing it with another command would not be an easy task, since all of the other mounted troops were in places remote from each other and fulfilling missions that needed doing. All the worse for Lee, then, when it proved to be Stuart who was not in place as expected, thus leaving the most critical task of all unfulfilled.

    Ultimately, Stuart's ride around the Army of the Potomac was an error in judgment, given that it cost Lee Stuart's services for the most critical week of the campaign. While Stuart clearly had the authority to make the movement, based on discretionary orders from Lee, final responsibility for that communications lapse can reside with no one else but Stuart. However, it is inaccurate to simply dismiss Stuart's ride as the product of wounded pride, as has been charged by historians who are quick to point to the public criticism of the young major general after Brandy Station.96   This explanation ignores the fact that there were valid military reasons for making the movement, and that Lee and Longstreet, as the immediately effected parties, both thought the ride a good idea. Of course, their approval was based on the assumption that Stuart would be out of contact only three to four days, and be ready to resume his right flank screening duties by the end of June at the latest. Therefore, while it proved to be a poor choice tactically, it was also a decision that was rooted more in military realities than in Stuart's pride.

    However, most discussion of the ride has presented it as an all-or-nothing case. It is quite possible that a lesser force could have accomplished all the positive military results that Lee and Stuart contemplated, and yet still left a cavalry screening force between Lee and the Union army. Longstreet's note of the 22nd of June, quoted in full above, suggests just such a possibility, though it has generally been interpreted otherwise. Longstreet assumed that Hampton would be left behind to take charge in Stuart's absence, and while this has generally been assumed to mean that Longstreet meant simply that Hampton's command would replace Jones or Robertson in remaining in Virginia, it does suggest an alternative: dividing the raiding column to keep at least one brigade behind as a hedge against disaster.

    In moving behind the Union army, Stuart could not hope to win a passage through by force. Clearly, any cavalry column operating in the Union rear could only hope to succeed by relying on stealth and speed. The larger the column, however, the less stealthy and speedy it could be. More troopers meant more forage to capture, more time consumed in feeding breaks, and larger areas needed for those halts. A larger column required more road space, and would be more difficult to maneuver around obstacles. All of these things add up to a solid argument that the column should have been smaller, not larger, to increase those two advantages.

    Given that Stuart was going to be in the middle of the Union army, and that ultimately the Federals deployed two cavalry divisions against him anyway (with about 9,000 Union troops-overwhelming odds against the Rebel raiding column in any case), the difference in numbers between two brigades and three were not significant anyway. In either case, Stuart could expect to be decisively outnumbered. Stuart could well have left Hampton behind, in addition to Jones and Robertson, specifically to cover the Confederate right flank and remain in contact with Lee and Longstreet in case his raiding column ran into trouble.97

    Planning for contingencies should be expected in military operations, where nothing ever seems to go as planned. The tactical circumstances of the ride, however, were all based on a best-case situation of completing the movement within 4 days, about the length of time it would take for the Union army to catch up with Lee's advance. A number of clues indicate that this was Lee's expected window: Major McClellan's own reminder that Lee knew how much time the ride would take in best case circumstances, the amount of time consumed by Hill and Longstreet's corps crossing Maryland, and Lee's own growing concern for Stuart's whereabouts after June 29th all point to the same conclusion. Lee expected to have his entire force unified in Pennsylvania by the end of June, and ready to deal with whatever new circumstances the campaign produced at that point. However, on Stuart's part, keeping to that schedule would demand that almost nothing go wrong for the duration of the ride, and as we have seen, things started to go wrong at the very beginning. It is this failure to provide Lee with alternatives, or to turn back at a point when he could still reach Lee's intended flank in a timely manner, that is perhaps Stuart's defining mistake of his ride.