- Atlanta untenable -- losses during the siege Compared with those of Sherman, and with those of Johnston from Dalton to Atlanta.
Having stated that our position at Atlanta was untenable, I shall now undertake to give proof of the correctness of my opinion, by demonstrating in what manner Sherman might have captured the city in less than one-third of the time he actually devoted to that end, notwithstanding the idle assertion of General Johnston that he could have held Atlanta “forever” ; also to demonstrate that had I ventured to remain longer than fifteen days within the trenches around the city, Sherman could have finally forced me to surrender, or have put my Army to rout in its attempt to escape; and, lastly, that he could have attained his object with one-fourth the loss he sustained during a siege of forty-six days, provided he had availed himself of the natural advantages afforded him. In order to render these operations clear to the mind of the reader, I invite his attention to a map, page 167. The Federal commander chose well his crossing of the Chattahoochee, as he approached Atlanta, and the move of McPherson and Schofield upon the Augusta road was ably conceived and executed. Thomas, however, should not have formed line of battle along the lower part of Peach Tree creek  with a view to cross the creek, as he endangered the safety of his Corps when he placed it in the angle, formed by this stream and the Chattahoochee, and thus isolated himself from Schofield and McPherson. His right should have rested in the vicinity of, and have covered, the ford nearest the mouth of Peach Tree creek, with a line of skirmishers extending to the Chattahoochee, and batteries in position at proper intervals in rear of this line, and likewise on the north side of the river, so as to thoroughly command the approaches to the railway bridge. His right being established at this ford, his left should have been thrown back north of Decatur, and his entire line strongly entrenched. From this position of perfect safety, he could have made constant demonstrations against the city, whilst McPherson and Schofield destroyed the road to Augusta. At the same time, by use of the batteries near the mouth of Peach Tree creek, and of those on the north bank of the river, near the railway bridge, he could have easily thrown a division across the creek, and have established a strong tete-de-pont on the south bank of the river. I would thus have been forced to form line of battle facing Peach Tree, with no possible chance of successfully assaulting the enemy at any point. His right and rear would have been covered by a deep and muddy stream; his front protected not only by breastworks, but also by one of the branches of Peach Tree creek. I could not have attacked either his left or McPherson and Schofield, without marching out of Atlanta, and exposing our left flank to Thomas. I would have been compelled to abide quietly the destruction of one of our principal roads, without the ability to strike a blow in its defence. After my loss of the Augusta road, McPherson and Schofield should have marched by the right flank down Peach Tree, in rear of Thomas's line, until their right rested on the Chattahoochee, and then have halted. General Sherman having his Army thus massed, and well in hand, in rear of Peach Tree creek, should have thrown across the Chattahoochee a sufficient number of pontoon bridges to allow the easy and rapid  passage of his troops; have sent two corps of Thomas's Army across and down the Chattahoochee, on the northwest side, to a favorable crossing just below Camp creek or one of the deep ravines or creeks, heading in the direction of East Point, and running toward the river upon the southwest side; have laid pontoons, crossed over, and strongly entrenched. Whilst this move was in progress, the main body should have made heavy demonstrations along the line of Peach Tree to the Augusta road, which diversion would have held my Army in position on the north side of Atlanta. The two corps below Camp creek having their line and out-posts established sufficiently in advance to allow full space for the massing of the Army, Thomas should have left the division in the tete-de-pont protected by ditch and abatis, whilst another division, with dismounted cavalry, occupied his position on Peach Tree creek with cavalry on their left, and a few batteries to support their line; then have marched, at dusk, with the remainder of his Army to join the two corps below Camp creek, followed by Schofield and McPherson. The transportation of the Federal Army having been previously parked on the north side of the Chattahoochee, General Sherman could with entire safety have massed his Army in the space of one night on the southeast side of the river, below this creek, as the two divisions left on Peach Tree had a secure place of refuge in the tete-de-pont in the event I had moved out in that direction. These preparations completed, he should on the morning of the morrow have ordered the divisions and cavalry, along this stream, to make demonstrations against the city whilst Thomas pushed forward in the direction of East Point — changing front forward on his left — and formed line of battle with his left flank resting as high up on Camp creek1 as it would afford protection against its being turned, and his right extending to or across the West Point Railway; have instructed Schofield  and McPherson to move rapidly, as they had done upon Decatur and the Augusta road, to deploy on Thomas's right along the south bank of South river and east side of Shoal creek, with their right thrown back southeast of Decatur,2 and to entrench the whole line. Such would have been the position of the Federal Army within twenty-four hours after it left Peach Tree creek, and within ten days after its first crossing of the Chattahoochee, subsequent to the operations about Kennesaw Mountain, provided it had moved with the rapidity usual with the Confederate Armies. Even had forty-eight hours been required to perfect this movement, Sherman could have so manceuvred as to have held the main body of my troops on Peach Tree until he was willing I should become apprised of his real purpose. In other words, he could without difficulty have entrenched south of Atlanta, before I could have received the necessary information to warrant a change of position from the north to the south side of the city. Moreover, had I divined at an early instant his contemplated move, his position in rear of Peach Tree, and that of the two corps on Camp creek would — by demonstrations on the north and south sides of the city, with an Army double our own — have rendered it an easy matter to him to gain possession of Atlanta, in spite of every effort on my part. General Sherman knew as well as I did, that every available man in the Confederacy had been sent either to General Lee, in Virginia, or to General Johnston, in the mountains; that, consequently, he had nothing to fear from the direction of Macon, and that one division would have sufficed to protect his rear, south of the city. When Grant marched round Pemberton at Vicksburg, and placed his rear in front of General Johnston, commanding an Army of twenty-five or thirty thousand men at Jackson, Mississippi, he executed successfully not only one of the boldest, but one of the grandest movements  of the war. It will rank with one of the many similar moves of the immortal Jackson, and receive the tribute due to the talent and boldness which planned and achieved it. It was, however, fortunate for General Grant that a “Stonewall” was not at Jackson, Mississippi. No especial daring on the part of General Sherman would have been required to carry out the operations I have designated, since he had no enemy to fear in his rear. General Grant was reported, at this period, to have said that the Confederacy was but a shell. As I have just remarked, I could not have received in time sufficiently reliable information to justify a change from the north to the south side of Atlanta, and to attack the Federals before they had thoroughly entrenched; it would have been equally impossible to assault later, with hope of success, his line, protected in front by works and abatis, on the left by Camp creek, and on the right by being thrown back and entrenched southeast of Decatur. This position of the enemy would have necessitated the immediate abandonment of Atlanta or have shut up our Army in the pocket, or cul de sac, formed by the Chattahoochee river and Peach Tree creek, and finally have forced us to surrender. Had I attempted to extricate the Army, it would have been almost impossible to have pierced the enemy's works south, and utterly impossible, by reason of the proximity of the Federals, to have laid pontoons and crossed Peach Tree creek — as I would have done when Sherman was at the distance of Jonesboroa, but from which I was hindered by the presence of the prisoners at Andersonville. By reference to the map (page 167) it will be perceived that Sherman had simply to advance his right flank, in order to form a junction with the troops, near Decatur, and thus completely hem in our Army. This plan for the speedy capture of Atlanta could have been executed with an insignificant loss, as it would have been achieved mainly by manoeuvre. In view of the impaired morale of the Army at the close of the Dalton-Atlanta campaign; the numerical inferiority of our  forces; the fact, previously mentioned, that all available troops in the Confederacy, east of the Mississippi, had been sent either to the Army of Northern Virginia or to the Army of Tennessee, with the exception of small forces guarding the seaboard; in view of the proximity of the Chattahoochee river, which flows within five miles of Atlanta, along the foot of the general slope from the mountains of Georgia to the plains, forming with Peach Tree creek a complete cul de sac, in which Atlanta is situated; the advantages to be derived from Camp or any other creek in that vicinity, or from the deep ravines running to the river from the southwest side, behind which the Federal Army could have been rapidly deployed forward into line, with no enemy to fear from the rear; in which position Sherman could have readily supplied his Army from the nearest point on the railway north of the river, and have quietly awaited my surrender; in view, then, of the above enumerated sources of weakness and danger, I do not hesitate to challenge the military world to refute my assertions, not only as to the feasibility of the plan I have demonstrated, but also as to the untenability of Atlanta. How long, I venture to inquire, is it probable that Sherman, after the capture of Jonesboroa, would have tarried before occupying the identical position I have designated? The extraordinary haste I made to evacuate Atlanta, after the Federals gained possession of Jonesboroa, on the Macon road, fifteen miles below the line from Camp creek to and along South river and Shoal creek, is proof of the great dread I entertained of a speedy occupation of this line. In lieu thereof, Sherman, during or immediately after the destruction of the Augusta road, threw Thomas across Peach Tree creek, into the cul de sac aforementioned, separated him from McPherson and Schofield, and subjected him to an assault by the main body of our Army, which should have resulted in the rout and capture of the greater portion of his Army. This move was, moreover, unnecessary, as it was impossible for him to invest Atlanta, approaching it from the north. He therefore consumed  forty-six days in the achievement of results which might have been accomplished within a fortnight. I have, from Dr. A. J. Foard, medical director of the Army, a statement of the total number of killed and wounded from July 18th to September 1st; in other words, from the day I assumed command to the evacuation of Atlanta. As I have already asserted, the number of men wounded in an Army which is standing its ground, and fighting, or is advancing, and driving the enemy, should not constitute an actual loss, and should only in part be comprised in the sum total of losses, since almost all the slightly wounded, proud of their scars, soon return to ranks. Therefore the only correct method of ascertaining the entire loss, during a siege or a campaign, is to deduct the number of effectives at the close of the siege, or campaign, from the number of effectives at the beginning of operations, after adding to the total strength the number of reinforcements, and deducting therefrom the number of troops permanently detached. This calculation should embrace the sick, and those who die from natural causes; losses under this head are, however, rarely of consequence during active operations. It is, as a generality, only when troops are lying in bivouac, or in quarters, that the ravages of disease are to be feared. Therefore, the fairness of this method can in no degree be affected by this minor fact. Upon this broad and equitable basis, I herewith submit the official return of Colonel A. P. Mason, assistant adjutant general of the Army of Tennessee, showing its strength at different periods whilst under my command. As I desire to compare my strength and losses with Sherman's and Johnston's, I present, at the same time, Dr. Foard's official report of the killed arid wounded; General Sherman's returns, showing his effective strength and estimate of losses; and the official statement of General Johnston's adjutant general, exhibiting the strength of the Army of Tennessee at different periods, during the campaign from Dalton to Atlanta. 
On recaiitulating the entire losses of each Army during the entire campiaign, from May to Sefitember, inclusive, we have, in the Union Army, as per table appended:4
Killed 4,423 Wounded 22,822 Missing 4,442 Aggregate loss 31,687.
 I here reiterate that it is impossible General Johnston should have turned over to me fifty thousand six hundred and twenty-seven (50,627) effectives on the 18th of July (as shown in Colonel Falconer's report), for the reason that he had this number in full on the 10th of that month. When, according to this same report, we suffered a loss, over and above the killed and wounded, of four thousand and seventy-three (4073) men who abandoned their colors, and went either to their homes or to the enemy just prior to the retreat across the Chattahoochee river, it is not reasonable to assume that no desertions occurred from the 10th of July--the date of his last return — to the 18th, when a change of commanders took place in the face of the enemy, and under extraordinary circumstances. The supposition that many deserted during this interval is but just and natural. I am, therefore, confident that I am over-liberal in the estimate given--forty-eight thousand seven hundred and fifty (48,750) effectives — in my official report of the effective strength of the Army of Tennessee, when I assumed command. However, I will, in this instance grant, for the sake of argument, that my force on the 18th of July was fifty thousand six hundred and twenty-seven (50,627) effectives. On the 20th of September, when stragglers had been gathered up, the effective strength of the Confederate Army, according to Colonel Mason's report, was forty thousand four hundred and three (40,403). This number, subtracted from fifty thousand six hundred and twenty-seven (50,627)--less thirty-one hundred (3100) permanently detached to Macon and Mobile, about the beginning of the siege — shows a loss of seven thousand one hundred and twenty-four (7124), to which should be added two thousand prisoners returned to the ranks by exchange, soon after the fall of Atlanta, and before Colonel Mason made up his return on the 20th of September. These prisoners were overlooked by myself and my chief of staff at the time I made my official report, and increase the total  loss, from all causes, to nine thousand one hundred and twenty-four (9124). Whilst Dr. Foard's report of the killed and wounded is correct, the above estimate is beyond doubt equally accurate,. since I received no reinforcements, during the siege, which were not sent back soon after their arrival, with the exception of about two hundred and fifty men of Gholsen's brigade (which small force I have not taken into account), as the following letter from General Shoupe will indicate:
Although the number of killed and wounded in the Army of Tennessee proper, during the siege, amounted to twelve thousand and twenty-three (12,023), the actual loss was nine thousand one hundred and twenty-four (9124); thus proving that near three thousand wounded returned to the ranks. I shall now sum up the loss of the enemy during that same period. 8 General Sherman reports his loss in killed, wounded, and missing, around Atlanta during July, August and September,. to have been fifteen thousand and thirty-three (15,033). His actual loss during the siege must assuredly have been in excess of this number. In accordance with his recapitulation,9  he had on the 1st of July an Army of one hundred and six thousand and seventy (106,070); on the 1st of August, ninety-one thousand six hundred and seventy-five (91,675); and on the 1st of September, eighty-one thousand seven hundred and fifty-eight (81,758), demonstrating an actual loss of twenty-four thousand three hundred and twelve (24,312) men within two months. This number, less the troops discharged or permanently detached, must be the real loss he sustained. I have not been able to glean from his statements the decrease of his Army from this latter source. I find, however, the following recorded in Shoupe's Diary on the 17th of August:
Enemy's pickets called to ours, and stated that a Kentucky Division, twenty-two hundred (2200) strong, was going out of service, and that neither Old Abe nor Uncle Jeff would get them in service again.Taking his own statements as a basis of calculation, and assuming the correctness of the report by the picket relative to the discharge of twenty-two hundred (2200) Kentuckians thirteen days prior to the fall of Atlanta, his actual losses (provided he did not during the siege receive reinforcements, of which I can find no mention in his Memoirs), prove to have been twenty-four thousand three hundred and twelve (24,312), plus nineteen hundred and two (1902) killed and wounded early in September, minis twenty-two hundred (2200) discharged; showing an actual loss of twenty-four thousand and fourteen (24,014) effectives against my loss of nine thousand one hundred and twenty-four (9124), although every aggressive movement of importance was initiated by the Confederates. On the other hand, and according to my opponent's statement,10 General Sherman had, after Blair's Corps joined him near Rome, a force of one hundred and twelve thousand eight hundred and nineteen (112,819) effectives to oppose General Johnston; and at the close of his victorious march from Dalton to Atlanta, one hundred and six thousand and seventy (106,070) effectives, which subtracted from the total number  one hundred and twelve thousand eight hundred and nineteen (112,819) in the field, at the beginning of the campaign, demonstrates an actual loss of only six thousand seven hundred and forty-nine (6749) against General Johnston's loss of twenty-five thousand (25,000) men. This comparison of losses under opposite modes of handling troops, evinces the truth of the principle for which I contend: that losses are always comparatively small in an Army which drives before it the enemy day after day, as in the instance of the Federal Army during the Dalton-Atlanta campaign; or in an Army which holds its ground, as in the instance of the siege of Atlanta when the Federal loss was greatly in excess of our own, by reason of the enthusiasm and self-reliance of the Northern troops having, in the sudden check given to their sweeping career of victory, been somewhat counteracted by depression, consequent desertion, and the tardy return of absentees.