Rebellion Record: a Diary of American Events: Documents and Narratives, Volume 10.
Frank Moore, Ed.

I had, myself, pointed out some of the defects of the works to the Engineers having charge of them, and I had urged the necessity of having another bridge further up the stream. The fact is, in my opinion, the position was susceptible of being made very strong, but in order to enable a small force to hold it against a large attacking force the works ought to have been entirely enclosed, and with a deep ditch on the outside, so that an attacking column could have had its progress checked. But the works were so constructed as to afford no obstacle in themselves to an attacking enemy, and only furnished a temporary protection to our troops. An attacking force could walk over the rifle trenches without difficulty, and even the works in which the guns were posted could be readily passed over when once reached. On the south side of the river were two hills immediately in the rear of our works, one crowned with a redoubt, constructed by the enemy, which had been remodeled and turned; the other was crowned with sunken pits for guns. In the first I. found Graham's battery, and in the latter Dance's battery, both of Brown's battalion. Besides these works, were two pits for guns in the flat on the right of the railroad, constructed for, posting guns, for the purpose of enfilading the east side of the railroad embankment on the north of the river. These pits, which were not occupied, had attached to them a short rifle trench, and further to the right was another rifle trench, covering the point at which the enemy had had a pontoon bridge. This presents the state of things as I found them, and I must here state that the defence of this position had not been entrusted to me. I had merely been called upon to furnish a detail for picket duty; alternating with both the other divisions of the corps for some time, and latterly with Johnson's only, I hurried to the spot myself, and ordered my command to follow, because I regarded my brigade in danger, and I doubted not I was but anticipating the order which would have been given as soon as the facts reached General Lee and Lieutenant-General Ewell.

I carried no artillery with me, because none was at my disposal. As soon as I had ascertained the condition of things in front and in the works, I rode back across the river to see if my other brigades were coming up, and communicated with General Lee, who had taken his position on the hill on which Graham's guns were posted. Shortly after I reached this point our skirmishers commenced falling back, and the enemy commenced advancing more rapidly, and I sent back to hurry up my brigades. The enemy, having gotten possession of the range of hills in front of our position, now planted a battery of artillery on a prominent point in front and opened, no artillery having been previously displayed by him. The guns were replied to by Dance and Graham, but with little or no effect, as the distance was too great. The enemy's skirmishers, in very heavy line, continued to advance until ours from the front and flanks were compelled to retire into the works, and the enemy's, on the right, advanced to the river bank, about half a mile below the bridge. About this time General Lee ordered one of Dance's guns to be sent to the pits on the right of the railroad, but, before the order was executed, the enemy's sharpshooters had advanced so close that General Lee countermanded the order, as he thought the guns might be disabled by having the horses shot down. About four o'clock General Hays arrived and took command of his brigade; and in a short time after the advance of my column, Hoke's brigade, under Colonel Godwin, arrived, and I sent Colonel Godwin, with the brigade, across the river to report to General Hays. and occupy that part of the trenches which Hays' brigade could not occupy. This plan met with the approval of General Lee, and he directed me to send no more troops across the river, but retain the other brigades on the south side. I sent Gordon's brigade to occupy Jamieson's hill to the right, and the river bank in front of it, and formed Pegram's brigade in rear, out of range of shells, sending the Thirty-first Virginia regiment from it to occupy the rifle-trenches at the gun-pits, on the right of the railroad. About this time the enemy opened another battery in front of our left on the road from the direction of Warrenton, and very shortly afterwards another battery was opened on the right from the edge of a woods. The fire from these batteries crossed and, in a great measure, enfiladed our position, and rendered the bridge quite unsafe. The battery on the hill, in front, also continued to fire, and the fire from all of them was continued until near dusk. The fire from Dance and Graham's batteries was stopped by order of General Lee, I believe, as it was manifestly producing little or no effect, and resulted in a mere waste of ammunition. Green's battery, however, continued to fire as well as it could. During all this time the wind was blowing very hard towards the enemy, so that it was impossible to hear the report of the guns, even at a very short distance. I had remained with General Lee at his request, who, in the latter part of the afternoon, had taken his position on the hill occupied by Dance's battery. About dark the artillery fire ceased, and some movements of the enemy took place, which we could not well distinguish. In a short time, however, some firing of musketry at and in front of the rifle-trenches was observed from the flashes of the guns, it being impossible to hear the report by reason of the wind, though the distance was but short. After this firing had continued for some minutes it slackened somewhat, and, not hearing from it, we were of opinion that it was from and at the enemy's skirmishers, and General Lee, expressing the opinion that the movement by the enemy on this part of the line was intended merely as a reconnoissance or feint, and that it was too late for the enemy to attempt anything serious that night, concluded to retire.

It was then nearly or quite dark, and while