John M. Schofield, Forty-six years in the Army

Our cavalry gave us accurate information that the enemy was advancing on the 21st, when Cox, with Wagner in support, was ordered to interpose between the enemy's cavalry and Columbia; while Stanley, with two divisions of the Fourth Corps, marched from Pulaski to that place, and our cavalry moved on the enemy's right to cover the turnpike and railroad. The whole army was in position at Columbia, November 24, and began to intrench. Hood's infantry did not appear in sight until the 26th. Cox had a brush with the enemy's cavalry, which had driven in one of our cavalry brigades. That action was magnified at the time, and afterward, into evidence of a race between our troops and the enemy for the possession of Columbia. In fact, Ruger's troops at Columbia were quite capable of holding that place against Forrest and Hood's infantry was not within a day's march of either Cox or Stanley until after both had reached Columbia.

We held our intrenched position in front of Columbia until the evening of November 27, inviting an attack, and hoping that Thomas would arrive with, or send, reinforcements in time to assume the offensive from Columbia; but reinforcements did not come, and the enemy did not attack. It became evident that Hood's intention was not to attack that position, but to turn it by crossing Duck River above; hence the army was moved to the north bank of the river in the night of the 27th. It was still hoped that the line of Duck River might be held until reinforcements could arrive. General Thomas was very urgent that this should be done, if possible, as the arrival of General A. J. Smith's corps from Missouri had been expected daily for some time, when General Thomas intended, as it was understood, to come to the front in person with that corps and all the other troops he could assemble in his department, take command, and move against the enemy.

About that time was disclosed one of those contrivances