226 Southern Historical Society Papers.
losses, 505, leaving for Gordon's Division at Cedar Creek, 2,405. For Kershaw's Division there is no September report. Returns August 30th, 3,445. Losses: Humphrey's Brigade, at Berry ville, Septembr 3d, 148; Bryan, 30; Connor, October I3th, at Cedar Creek Crossing, 182. These deducted leave for Kershavv 3,085. Early's total infantry and artillery at Cedar Creek, 12,780. Early's cavalry, two divisions under Lomax and Rosser, is not enumerated in the record. Battles and Leaders gives it at 2,900; or a total of 15,680. But such was the condition of our cavalry that it was almost a negligible quantity, and Lomax, with the largest division, never got under fire.
Judgment is claimed against General Early on the ground that he should have made his advance continuous, after the morning vic- tory. The claim is founded upon the contention that this was feas- ible and was caused to halt by General Early. I maintain that such assumption is not warranted by facts. After planning and ordering the assault General Early committed it to the division commanders. It was especially entrusted to Generals Gordon and Kershaw, who led their commands upon separate points, for simultaneous assault, and acted, for the time, independently of each other and General Early; who, after seeing Kershaw's assault launched, posted himself with Wharton's division and the artillery at the pike crossing, until it should be uncovered by Kershaw. Conducted by his division commanders the attack on the Federal left, Crook's Eighth corps, was brilliantly successful. It was so continued until the advance halted itself, for cause, to- wit: It spent its force, and encountered a vastly superior array of the enemy in a strong defensive position. Still, at this juncture, General Early, by his report did order a fur- ther advance, by Gordon and Kershaw; which being considered impracticable by them was not made. Then, he states, he did not deem it prudent to press further, and, therefore, determined to be content with trying to hold the advantages gained. Receipt of such order has been denied. But admitting this, what does it matter ? If continuous advance was not impracticable, why did it halt ? Why did not these division commanders make it continuous, while it was in their hands? No actual order to halt them has been specified. It cannot be contended that any order to advance continuously was necessary. Continuous advance as long as practicable was covered in the original order of attack.
It is true that General Gordon's "War Reminiscences" says 11 orders from headquarters put an end in the early morning to con-