On August 6th General Ambrose Burnside advanced into East Tennessee. On the 14th Rosecrans advanced, using his cavalry to make demonstrations against the river crossings above the town. Bragg directed General Simon B. Buckner to give up Knoxville and move nearer to Chattanooga. About this time some of the troops which had been with J. E. Johnston in Mississippi were sent to him and General James Longstreet with part of his corps and the reserve artillery from Lee’s army was ordered to Chattanooga.

On September 4th Rosecrans crossed the river at Bridgeport. On the 9th he learned that Bragg had evacuated Chattanooga and was withdrawing towards Dalton. He at once pushed eastward sending McCook’s corps and much of the cavalry towards Alpine and Summerville to cut Bragg off. Crittenden’s Corps was to pass through the city and pursue via the Ringgold road and Thomas’s Corps was to move on Lafayette via Trenton. This disposition separated Rosecrans’s right and left wings by 40 miles and, due to the impassable mountains, no two corps were within supporting distance.

Rosecrans was enamored by his apparent success by again pushing Bragg back many miles without a heavy battle. He believed Bragg was rapidly withdrawing again and concentrating some where around Dalton, Georgia. If kept on the run, who knew how far back toward Atlanta Bragg might be driven.²

The Army of the Cumberland’s three corps were widely separated. Bragg had finally decided that he must make a stand somewhere. Recognizing the large gaps separating the three Union corps he decided to try to isolated and crush Thomas’s center corps advancing toward Lafayette from Bridgeport. With Bragg’s army outnumbering Thomas’s isolated corps by more than two to one, it appeared the army of Tennessee would finally give Rosecrans a long over due defeat.³

Bragg’s planned maneuver was for Maj. Gen. Thomas Hindman’s division to push south from his position at Lee and Gordon’s Mill, up the Chickamauga Creek to fall on Negley’s left flank. At the same time Negley’s front would be assailed directly by the division of Maj. Gen. Patrick Cleburne coming rapidly over Pigeon Mountain by way of Dug Gap to fall on the marching Federals leading brigade.⁴

Map 1

By the morning of the 10th the divisions of Hindman, Cleburne and Breckinridge were in position. Hindman soon became over cautious fearing his own flank may be exposed. D. H. Hill, Cleburne’s commander, also hesitated for reasons that are still unclear. The net effect was that Bragg’s planned attack never went forward. This hesitation gave Thomas the chance, over the next day or so, to realize his exposed position and pull back avoiding a well laid but poorly executed trap.⁵

What may have happened had this plan gone forward is amazing to consider. Bragg had succeeded in concentrating his whole army within a few miles of Thomas’s corps while the two Union flank corps were each about 20-25 miles away, easily 2 days march and out of supporting distance. Had the attack gone off as ordered by Bragg some 20-30,000 of his 55,000 troops would have fallen on Thomas’s 20,000 by surprise and in front and flank. What’s more, Thomas’s unsuspecting column was not concentrated but stretched out in road march order, each division and brigade marching some distance behind the next leading brigade.

On September 9th Negley's division moved across the mountain and took up a position in McLemore's Cove, near Rodgers' farm, throwing out his skirmishers as far as Bailey's Cross-Roads. Baird's division moved across Lookout Mountain to the support of Negley. Negley’s skirmishers saw the Confederate cavalry in front, drawn up in line; citizens reported a heavy force concentrated in his front at Dug Gap, consisting of infantry, cavalry and artillery. ⁶

Reynolds’s and Brannan’s divisions were ordered to march toward dug gap on the afternoon of the 9th, the march to be made on the 10th. Negley’s and Baird’s to continue toward La fayette on the 10th. Thomas’s whole corps was ordered to make a “general pursuit” of Bragg toward La Fayette on the 10th.⁷ Its clear that Rosecrans had no idea that Bragg had concentrated at La Fayette until the evening of the 11th or perhaps noon of the 12th. Up to that time, Rosecrans believed Bragg’s army was roughly centered around Dalton and moving toward Rome. This means Bragg’s force was 20-30 miles nearer than Rosecrans believed and this made the army of the Cumberland dangerously exposed on the evening of the 9th September.⁸

Thus on the 10th, Thomas’s whole corps was to push forward from Bailey’s Cross Roads with Negley and Baird and with Reynolds and Brannan, to Steven’s Farm at the eastern base of Steven’s Gap. The distance from Easley’s Farm, Reynolds and Brannan’s camp on the 9th, was about 12 miles or a good days march for a division-sized force. From Bailey’s Cross Roads to La Fayette it was also about 12 miles distant. Thomas’s corps would be stretched out over twenty miles during the best part of the day. His twenty thousand infantry, 68 guns and perhaps 70-80 wagons and ambulances would make a continuous column of 10 miles or more. This means that the lead divisions could not be supported by the two following divisions for at least three or four hours if attacked.⁹

Had Hindman and Hill’s troops moved forward as ordered, they would have struck Negley’s and Baird’s Divisions by around noon, or a little before, as they approached Dug Gap from the west. The Union troops would have had barely enough time to form some kind of rough line before coming in to close action with the rebel forces. With the defenders caught between attacks in front and from their extreme left and by surprise, its probable there would have been some confusion and perhaps a little panic. Thomas’s four divisions each numbered some 5,000 infantry; Negley’s and Baird’s divisions together mustered not more than 10,200 muskets that day.⁷

Map 2

The following combat model uses the historic hit/1000 index as found in Livermore for the battle of Chickamauga. The purpose of this counter factual situation is not to say that any event could have or should have happened. The purpose is to understand just how devastating the dysfunctional command structure in the Army of Tennessee and among the Confederate high command may have been.⁷

Using the Livermore historic hit/1000 index as the basis we attempt to calculate combat effects for the counter factual battle of McLemore Cove, September 10, 1863.¹⁰

Orders of battle:

Orders of battle

Confederate Forces engaged
Hill 9000
Hindman 4500
Stewart 5000
Infantry 18500

First round Negley & Baird v Hill, Hindman & Stewart
Using hit/1000 for the battle of Chickamauga.
Union hit/1000 292 (Livermore, p105,106)
Confed hit/1000 172
5700+4500=10200 * 292/1000 = 2978 k&w (Confed)
18500 * 172/1000 = 3182 k&w (Union)

The effects of surprise are taken to increase the Vulnerability of the Union force and decrease for the confederates. Following Dupuy, surprise effects may have three levels; Minor, Substantial, Complete. ¹¹

SUPRISE EFFECTS

Applying each value for minor surprise gives the following:

2978 k&w (Confed) * 0.8 = 2382 k&w (Confed)
3182 k&w (Union) * 1.3 = 4137 k&w (Union)

We also apply a factor for missing troops of 0.08 for the Union & 0.02 for the Confederate, loser & winner respectively.

10200 * 0.08 = 816 * 1.3 = 1061
18500 * 0.02 = 370 * 0.8 = 296

Giving total losses:

Union engaged 10,200, 4137 k&w, 1061 mia tot 5198 (54%)
Confed engaged 18,500, 2382 k&w, 296 mia, tot 2678 (15%) 3.6

It is apparent that the effects of this engagement would have been devastating to Thomas’s corps. The effects for Substantial or Complete surprise would be horrendous.

We find the remaining forces available

18500 – 2678 = 15822 Confed remaining
10200 – 5198 = 5002 Union remaining

Second round Brannan & Reynolds v attackers & reinforced by Liddell’s division.

10300 = * 292/1000 = 3007k&w (Confed)
15822 + 3175 = 18997 * 172/1000 = 3267 k&w (Union)

Although the first round fight would have occurred over several hours (4-5?), we continue with minor surprise effects vs. Brannan & Reynolds. This is due to the hasty defense and the shock of the defeat of a large part of the corps. Applying each value for minor surprise gives the following:

3007 k&w (Confed) * 0.8 = 2406 k&w (Confed)

3267 k&w (Union) * 1.3 = 4247 k&w (Union)

We also apply a factor for missing troops of 0.08 for the Union & 0.02 for the Confederate, loser & winner respectively.

10300 * 0.08 = 824 * 1.3 = 1071 mia
18997 * 0.02 = 304 * 0.8 = 296 mia

Giving total losses:

Union engaged 10,300, 4247 k&w, 1071 mia tot 5318 (52%)

Confed engaged 18,997, 2406 k&w, 296 mia, tot 2702 (14%) 3.7

Remaing after two rounds:

2-rds; 18500+3175 – 2678-2702 = 16295 Confed remaining (75%)
10300+10200 - 5198–5318 = 9984 Union remaining (51%)

We may also add in the troops under Bragg not engaged, Cheatham’s division of about 7455 infantry and Preston’s Division of about 4510 men. Total effective 27,960

Thomas, defeated and facing nearly three times his surviving force, would probably be forced to make a rapid retreat back over the mountains and into Tennessee.

Map 3

This would leave Rosecrans with his two wings separated by over 40 miles, nearly 4 days march, and each wing being about half Bragg’s whole force. The question then becomes could Rosecrans hold Chattanooga and could McCook’s corps safely withdraw and avoid being crushed by Bragg?

At the battle of Chickamauga, the main engagement began about 10am and continued until about 6pm, some eight hours. By 2:30pm however, the first clash between Longstreet’s forces and the Union right wing divisions of Sheridan, Davis, Wood, and Brannan was over. These first three units were routed and Brannan hung on after wheeling back about a mile. Later, the divisions of Negley and Van Cleve were caught up in the chaos and routed. At McLemore’s Cove on September 10 1863 Sunrise was at about 5:30 am and sunset about 6:15. Twilight began at about 4:45am and ended around 7pm. Thus, there appears to have been ample time during daylight to complete both rounds as shown in our combat model.¹²

Sun Table

At Chickamuaga, both sides had lost some 28% of the total force. A large portion of Rosecrans’s army was able to withdraw in reasonably good order in to Chattanooga. In our McLemore’s Cove model, this appears problematic. First, Rosecrans’s remaining force is split; meaning, that Crittenden’s XXI corps would be at risk even in Chattanooga. McCook’s XX corps would be even more exposed and would have had to make a long and hasty retreat to avoid being trapped and suffering a fate like Thomas.

Map 4

Thus, it seems likely that the army of the Cumberland would have needed to make a major withdrawal, perhaps all the way back to Murfreesboro. This opens an entirely new realm of contemplation for what the long-term effects may be. The history shows that from the date of the battle of Chickamuaga to the fall of Atlanta, September 20 1863 to September 1 1864, was about 11-1/2 months. Also that the Union advance from Murfreesboro to Chickamauga took about 2-1/2 months. We may then project that had the army of the Cumberland been forced to pull back into Tennessee around mid-September 1863, it may have easily taken until December 1864 or January 1865 to complete the conquest of Atlanta. How this may have affected the final out come is hard to say but the effect on the 1864 election and the long term out come would likely be unfavorable to the Union prospect.

Another question is how such an outcome would have effected Grant’s position. With the situation in Tennessee and Chattanooga stable by the end of 1863, Grant was available to move to Virginia by March of 1864. Had the Army of the Cumberland been forced back to Murfreesboro in September, it may have taken at least until May or June 1864 for Grant to have been available to go to Virginia. Would Grant have been placed in command to lead the advance from Murfreesboro to Chattanooga? Would Grant have been sent to Virginia and Sherman sent to Murfreesboro? Again, while it’s difficult to evaluate these questions, the basic effect would seem most likely to delay any chance of a Union victory. Perhaps a few months delay would be all that was needed to make a negotiated political settlement a real possibility?

The purpose of this model is not to say that any event could have or should have happened. The real conclusion is to understand just how devastating petty bickering and jealousy among the Confederate high command was. We can find a number of other cases where lack of cooperation among confederate generals had strong negative effects on operations. Not withstanding these effects, the Confederate forces put up a tough fight for four years.

The foregoing combat model uses the historic hit/1000 index as found in Livermore for the battle of Chickamauga. In that battle both forces numbered well over 50,000 men. Since the forces in our model are much smaller, some adjustment may be of value. Smaller forces tend to not have the staying power for prolonged heavy combat. Using the historic values for our model with much smaller forces may exaggerate the losses on both sides. After examining a number of civil war battles with forces of similar size to those in our model we adopt new values as shown below to recalculate the results for the McLemore Cove 10-Sept-63 model.

RECALCULATION WITH NEW FIGURESFOR HIT/1000 & MIA (Author)

14th Corps Gen. Geo. H. Thomas
Negley’s Div. 5700 effective infantry
Baird’s Div. 4500
Brannan Div. 5000
Reynolds Div. 5300
total approx. 20500 effective infantry
Total 68 guns

Hill 9000
Hindman 4500
Stewart 5000
Infantry 18500

First round Negley & Baird v Hill, Hindman & Stewart

Using hit/1000 for the battle of Chickamauga.

Union hit/1000 292(Livermore, p105,106)
Confed hit/1000 172

Author’s values derived for average mid size forces

Average Mid size force

5700+4500=10200*173/1000 = 1765 k&w (Confed)

18500 * 128/1000 = 2368 k&w (Union)

The effects of surprise are taken to increase the Vulnerability of the Union force and decrease for the confederates. Following Dupuy, surprise effects may have three levels; Minor, Substantial, Complete.

SUPRISE EFFECTS

Applying each value for minor surprise gives the following:

1765 k&w (Confed) * 0.8 = 1412 k&w (Confed)

2368 k&w (Union) * 1.3 = 3078 k&w (Union)

We also apply a factor for missing troops of 0.08 for the Union & 0.02 for the Confederate, loser & winner respectively.

10200 * 0.08 = 816 * 1.3 = 1061

18500 * 0.02 = 370 * 0.8 = 296

Giving total losses:

Union engaged 10,200, 3078 k&w, 1061 mia tot 4139 (41%)

Confed engaged 18,500, 1412 k&w, 296 mia, tot 1708 (9%) 4.55

It is apparent that the effects of this engagement would have been devastating to Thomas’s corps. The effects for Substantial or Complete surprise would be horrendous.

18500 – 1708 = 16792 Confed remaining

10200 – 4139 = 6061 Union remaing

Second round Brannan & Reynolds v attackers & reinforced by Liddell’s division.

10300 = * 173/1000 = 1782k&w (Confed)

16792 + 3175 = 19967 * 128/1000 = 2556k&w (Union)

Although the first round fight would have occurred over several hours (4-5?), we continue with minor surprise effects vs. Brannan & Reynolds. This is due to the hasty defense and the shock of the defeat of a large part of the corps. Applying each value for minor surprise gives the following:

1782 k&w (Confed) * 0.8 = 1426 k&w (Confed)

2556 k&w (Union) * 1.3 = 3323 k&w (Union)

We also apply a factor for missing troops of 0.08 for the Union & 0.02 for the Confederate, loser & winner respectively.

10300 * 0.08 = 824 * 1.3 = 1071 mia

19967 * 0.02 = 399 * 0.8 = 319 mia

Giving total losses:

Union engaged 10,300, 3323 k&w, 1071 mia tot 4394 (43%)

Confed engaged 19,967, 1426 k&w, 319 mia, tot 1745 9(%) 4.7

Remaing after two rounds:

10200 – 4139 = 6061 Union remaining

2nd 19967 – 1745 = 18222 Confed remaining

10300 – 4394 = 5906 Union remaining

2-rds; 18500+3175 – 1708-1745 = 18222 Confed remaining (84%)

10300+10200 - 4139–4394 = 11967 Union remaining (58%)

We may also add in the troops under Bragg not engaged, Cheatham’s division of about 7455 infantry and Preston’s Division of about 4510 men. Total effective 30,187.

The effects of this recalculation are not dramatic but reduce Union losses slightly. Confederate losses are also reduced somewhat. The overall difference in losses appears to be slightly more favorable to the rebels. Thus, either model confirms the attack may have resulted in a significant tactical defeat for Thomas, and thereby a major strategic crisis for Rosecrans and the Union high command.