I learned in the afternoon of November 27, by General Thomas's despatch of 8 A. M., already quoted, that A. J. Smith's troops were not, as I had supposed, already in Nashville, but that some of them were expected there that day, and would come forward to join me at once.

In the morning of November 28, at 8:45, I reported my withdrawal to the north side of the river, saying:

"My troops and material are all on the north side of Duck River. The withdrawal was completed at daylight this morning without serious difficulty. c.o.x holds the ford in front of Columbia, and Ruger the railroad bridge, which I partially destroyed. Stanley is going into position a short distance in rear of c.o.x. I think I can now stop Hood's advance by any line near this, and meet in time any distant movement to turn my position. I regret extremely the necessity of withdrawing from Columbia, but believe it was absolute. I will explain fully in time. Reinforcements will have to march from Spring Hill or Thompson's Station. Supplies should be sent to Thompson's Station."

After withdrawing to the north bank of Duck River I telegraphed on the morning of November 28:

"I am in doubt whether it is advisable, with reference to future operations, to hold this position or to retire to some point from which we can move offensively. Of course we cannot recross the river here. I could easily have held the bridge-head at the railroad, but it would have been useless, as we could not possibly advance from that point. Please give me your views and wishes."

This was answered by General Thomas at "8 P. M.," the answer being received by me next morning, November 29.

It is thus seen that up to the morning of November 28 I was still hoping for reinforcements on the line of Duck River, and thought I could stop Hood's advance by any line near the Columbia and Franklin pike, which I then held, as well as meet in good time any distant movement to turn my position. Accordingly, at 9:10 A. M. that day I telegraphed to General Thomas:

"I have all the fords above and below this place well watched and guarded as far as possible. Wilson is operating with his main force on my left. The enemy does not appear to have moved in that direction yet to any considerable distance. I will probably be able to give you pretty full information this evening. Do you not think the infantry at the distant crossings below here should now be withdrawn and cavalry subst.i.tuted? I do not think we can prevent the crossing of even the enemy's cavalry, because the places are so numerous. I think the best we can do is to hold the crossings near us and watch the distant ones."

But I learned soon after noon of the same day that our cavalry found the fords so numerous that they could hardly watch them all, much less guard any of them securely; and a little later I learned that the enemy's cavalry had forced a crossing at some point only a few miles above, between Huey's Mill and the Lewisburg-Franklin pike. At 2:30 P. M. I telegraphed General Thomas:

"The enemy was crossing in force a short distance this side of the Lewisburg pike at noon to-day, and had driven our cavalry back across the river on the pike at the same time. The force is reported to be infantry, but I do not regard it as being probable. Wilson has gone with his main force to learn the facts, and drive the enemy back, if possible."

DELAYING THE RETREAT FROM DUCK RIVER

In the appendix to General Thomas's report the date of the above despatch is given as "3:30 A. M." It was answered by General Thomas at "10:30 P. M." and his answer was received by me November 29 (no hour mentioned in the records). The Department of the Ohio records say that I sent it at "2:30 P. M." The appendix to my report mentions the date "November 29," but does not give the hour. My official report, as published, also says this information was received "about 2 A. M. on the 29th"; but this is evidently a clerical error: Clearly the report should read, "about 2 P. M. on the 28th."

But our cavalry was unable to drive that of the enemy back, and hence Hood was free to lay his pontoon bridge and cross his infantry and artillery at any point above Columbia. We had not been able to hold even the crossings near us.

The same day, November 28, at 4 P. M., I telegraphed:

"If Hood advances on the Lewisburg and Franklin pike, where do you propose to fight him? I have all the force that is necessary here, and General Smith's troops should be placed with reference to the proposed point of concentration."

And again, at 6 P. M.:

"The enemy's cavalry in force has crossed the river on the Lewisburg pike, and is now in possession of Rally Hill.

"Wilson is trying to get on to the Franklin road ahead of them. He thinks the enemy may swing around in behind him and me, and strike Spring Hill, and wants Hammond's brigade to halt there. Please give it orders if you know where it is. Also, I think it would be well to send A. J. Smith's force to that place."

In the night of November 28-9, about 2 A. M., I received the report of the cavalry commander, conveying the information given him by prisoners that the enemy had commenced to bridge the river near Huey's Mill, and urging the necessity of immediate retreat to Franklin.(13) The staff officer who handed me the despatch called my attention especially to the words urging immediate action, and I considered the subject quite a long time. But there did not seem to me to be any necessity for such haste. The enemy could not accomplish much before morning. It would then be early enough to decide what must be done. Besides, it was not yet certain that Hood was attempting to cross his infantry at Huey's Mill. The vigorous action of his cavalry might be intended only to induce me to fall back, and thus give him the use of the crossing at Columbia, and of the turnpike from that place, for the movement of his infantry, artillery, and trains.

In the morning, November 29, I sent a brigade of infantry toward Huey's Mill to reconnoiter and report the enemy's movements. At the same time Stanley was ordered to Spring Hill, with two divisions of his corps, to occupy and intrench a good position commanding the roads at that place and protecting the trains and reserve artillery which had been ordered to be parked there. Ruger's division of the Twenty-third Corps, except one regiment, was ordered to follow Stanley. The army was ready to occupy Spring Hill in full force, and in ample time to meet any possible movement of the enemy either on that place or, by the Lewisburg pike, on Franklin.

In my orders to Ruger, dated 8 A. M., directing him to move at once to Spring Hill, he was ordered to leave one regiment to guard the river until dark and then join him at Spring Hill. It was then intended, in any event, to hold Spring Hill until the morning of November 30. At the same time Ruger was directed to order his troops guarding the river below to march at once for Franklin.

DELAYING THE RETREAT FROM DUCK RIVER

But very soon after these orders were issued-that is, soon after 8 A. M.-a courier from Franklin brought me the two following despatches from General Thomas:

"Franklin, November 28, 1864.

"(By telegraph from Nashville, 9 P. M.) "To Major-General Schofield: "If you are confident you can hold your present position, I wish you to do so until I can get General Smith here. After his arrival we can withdraw gradually and invite Hood across Duck River, and fall upon him with our whole force, or wait until Wilson can organize his entire cavalry force, and then withdraw from your present position. Should Hood then cross river, we can surely ruin him. You may have fords at Centreville, Bean's [Beard's] Ferry, Gordon's Ferry, and Williamsport thoroughly obstructed by filling up all the roads leading from them with trees, and then replace your infantry by cavalry. Send an intelligent staff officer to see that the work is properly done. As soon as relieved, concentrate your infantry; the cavalry will be able to r.e.t.a.r.d, if not prevent, Hood's crossing, after the roads are thoroughly obstructed, if they do their duty. The road leading from Centreville to Nashville should be thoroughly obstructed. I am not sure but it would be a good plan to invite Hood across Duck River if we can get him to move toward Clarksville. Is there no convenience for unloading beyond Thompson's Station?

"Geo. H. Thomas, Major-General, Commanding."(14)

The published records give this despatch as having been sent at "8 P. M." The Department of the c.u.mberland records say that it was telegraphed in cipher to Franklin at 9 P. M., and there deciphered and sent by courier to my position near Columbia. The records do not show the hour of receipt by me; but my reply to General Thomas of 8:30 A. M., November 29, and my orders to Ruger of 8 and 8:45 A. M., and to Stanley before and after 8 A. M., and my despatch to Wilson of 8:15 A. M., fix the time of the receipt by me of this despatch from General Thomas at a few minutes after 8 A. M., November 29.

The other despatch was as follows:

"(U. S. Military Telegraph.) "Franklin, Tenn., November 28, 1864.

"(By telegraph from Nashville. 9:30 P. M.) "To Major-General Schofield: "Your despatch of 3:30 [2:30] P. M. just received. If Wilson cannot succeed in driving back the enemy, should it prove true that he has crossed the river, you will necessarily have to make preparation to take up a new position at Franklin, behind Harpeth, [while] immediately, if it become necessary, to fall back.

"(Signed) Geo. H. Thomas, Major-General, Commanding."

The records of the Department of the c.u.mberland merely state that this despatch was sent in "cipher." The appendix to my report gives the hour "9:30 P. M." The appendix to General Thomas's report fixes it at "10:30 P. M." The despatch from General Thomas to General Halleck of 10 P. M., November 28, forwarding my despatch of "8:45 A. M.," indicates that at 10 P. M. Thomas had not received my report of "2:30 P. M." Hence "10:30 P. M.," as given by General Thomas, must be the correct hour of the above despatch. It was answered by me, together with the preceding telegram, at 8:30 A.M., November 29; and was probably received by me at the same time as the previous despatch,-very soon after 8 A. M.,-as indicated by my despatch to Wilson of 8:15 A. M.

I thus learned, a short time after eight o'clock on the morning of the 29th, that A. J. Smith had not yet arrived at Nashville, and that the position behind the Harpeth River at Franklin was that to which I must retire when compelled to fall back.

DELAYING THE RETREAT FROM DUCK RIVER

(Another despatch from Thomas, dated November 28, 10 A. M., appears in the records, in which he said: "... General Smith will certainly be here in three days... ." But when that despatch reached my headquarters in the field, the cipher-operator had left his post and gone to Franklin. Hence the despatch could not be read by me in time to be of any service. The records do not show when I received it.)

I was then confronted with the grave question, How long might it be possible to hold Hood back, and thus gain time for Thomas to get up his reinforcements? By holding on to the crossing of Duck River at Columbia until dark that night, and thus preventing Hood from using the turnpike for the movement of his artillery and trains until the next day, we would practically gain twenty-four hours; for he could not move them readily over his mud road from Huey's Mill. To do this, I must not only head Hood off at Spring Hill, but defeat any attempt he might make to dislodge me from the north bank of Duck River.

Early on November 29, I sent the following brief despatch in reply to both of those which had been received a few minutes before from General Thomas:

"The enemy's cavalry has crossed in force on the Lewisburg pike, and General Wilson reports the infantry crossing above Huey's Mill, about five miles from this place. I have sent an infantry reconnaissance to learn the facts. If it proves true, I will act according to your instructions received this morning. Please send orders to General Cooper,(15) via Johnsonville. It may be doubtful whether my messenger from here will reach him."

The appendix to General Thomas's report says that I sent this despatch at "8:30 A. M." The appendix to my report says "8:20 A.M." This despatch was evidently in answer to those from General Thomas of 8 P. M. and 10:30 P. M., November 28, as indicated by my orders to Stanley and Ruger, and my despatch of 8:15 A. M. to Wilson.

Soon after 10 A. M., November 29, the first report from the brigade sent toward Huey's Mill showed that the enemy's infantry was crossing the river at that place. That report is not found in the records, and I do not recollect its words. But it did not produce the impression upon my mind that Hood's movement was so rapid or energetic as to prevent me from doing what seemed of such vital importance. Therefore I decided not to yield my position unless compelled by force to do so. While considering this question I had detained one of Stanley's two divisions (Kimball's), and had suspended the orders for Ruger's division to march to Spring Hill. When the decision was reached, I put Kimball's and Wood's divisions in position between Duck River and Rutherford's Creek, and Ruger's north of that creek, to resist any attempt the enemy might make upon our position. I then sent the following to Stanley at Spring Hill:

"Near Columbia, Tenn., November 29, 1864, 10:45 A. M.

"Major-General Stanley, Commanding Fourth Army Corps.

"General: General Wood's reconnoissance shows a considerable force, at least, on this side of the river. I have halted Kimball's division this side of the creek and put it in position. I will try to hold the enemy until dark, and then draw back. Select a good position at Spring Hill, covering the approaches, and send out parties to reconnoiter on all roads leading east and southeast. Try to communicate with Wilson on the Lewisburg pike. Tell him to cover Franklin and Spring Hill, and try not to let the enemy get between us.

"Very respectfully, "J. M. Schofield, Major-General."

NO SERIOUS DANGER AT SPRING HILL

The situation early in the morning had been a very simple one, free from any embarra.s.sment or unusual danger. If the plan then decided on and ordered had been carried out, three divisions of infantry and nearly all the artillery of the army would have been in position at Spring Hill and well intrenched long before the head of Hood's infantry column, without any artillery, came in sight of that place late in the afternoon. That position would have been secured beyond doubt until the next morning. The other two divisions (c.o.x's and Wood's) would have withdrawn from Duck River and marched to Spring Hill early in the afternoon, before the enemy could seriously interfere with them. Ruger's one regiment, without impedimenta, was directed to march along the railway track to Spring Hill, and thus avoid any interference from the enemy. The army would have marched to Franklin early in the night of the 29th, instead of after midnight as it actually did. That would have given the enemy the afternoon and night in which to lay his pontoons and cross his artillery and trains at Columbia. But that would not have been a serious matter, in view of the situation as it was understood by me up to about 8 A. M. of the 29th; for the information I had received up to that hour justified the belief that both A. J. Smith's troops and those concentrated at Murfreesboro' would meet me at Franklin, or perhaps at Spring Hill, where we would be able to give battle to the enemy on equal terms.

But in view of the information received by me after eight o'clock that morning, and the altered plan decided on soon after ten o'clock, the situation became very materially different. Under this plan the army must be ready to encounter a formidable enemy either in the position then occupied on Duck River, or at some point on the road between that place and Spring Hill. Hence I determined to keep the main body of troops together, and trust to Stanley's one division to hold Spring Hill until the army should reach that point. That is to say, I decided to take the chances of a pitched battle at any point the enemy might select between Duck River and Spring Hill, as well as that of holding the latter place with one division against any hostile force which might reach it before dark.

There was no anxiety in my mind about what might happen at Spring Hill after dark. The danger which actually developed there between dark and midnight-of which I knew nothing until several days afterward-resulted entirely from faulty execution of my orders.

I arrived at Spring Hill at dusk with the head of the main column, having ordered all the troops to follow in close order, and (except Ruger's troops, which I took to Thompson's) to form line on the right of Stanley's division at Spring Hill, covering the pike back toward Columbia. c.o.x's division, being the last, was to form our extreme right. In that contemplated position, if Hood had attacked at any time in the night we would have had decidedly the advantage of him. I had no anxiety on that point. When informed, about midnight, that c.o.x had arrived, I understood that my orders had been exactly executed, and then ordered c.o.x to take the lead and the other divisions to follow, from the right by the rear, in the march to Franklin.

But it happened that only Whitaker's brigade of Kimball's division, to which I gave the orders in person, followed Ruger's. Hence that one brigade was the only force we had in line between Hood's bivouac and the turnpike that night. If that fact had been known to the enemy, the result would have been embarra.s.sing, but not very serious. If the enemy had got possession of a point on the pike, the column from Duck River would have taken the country road a short distance to the west of Spring Hill and Thompson's Station, and marched on it to Franklin. The situation at Spring Hill in the night was not by any means a desperate one. Veteran troops are not so easily cut off in an open country.

NO SERIOUS DANGER AT SPRING HILL

The annotation upon the copy filed in the War Department of the order actually given to the troops on November 29 explains how that mistake occurred. In brief, the draft of an order prepared in writing for another purpose, but not issued, was by some unexplained blunder subst.i.tuted for the oral orders actually dictated to a staff officer. It was an example of how the improvised staff of a volunteer army, like the "non-military agencies of government," may interfere with military operations.