Page:Southern Historical Society Papers volume 32.djvu/237

 Battle of Cedar Creek. 225

don. This book has been commended through the Confederate Veteran as " the most important record for the student of correct history of the battle of Cedar Creek." Stated concisely, the argu- ment of this " record" is that Early lost this battle, and by not fol- lowing Gordon's advice. I would be untrue to convictions derived from witnessing that calamitous field, confirmed by reading all that the record contains of it, if I did not challenge the statement that General Gordon's Reminiscences is ''correct history" of Cedar Creek. I can but wish that the task had been taken up by some one better qualified by station and ability to give weight to the truth of which I testify; concerning a battle that General Gordon states " no other save Gettysburg has provoked such conflicting and varied comment."

Any account of Cedar Creek calls for a statement of numbers of the two forces. Here there is no little conflict. Figures have probably been handled in partizan spirit by both sides. But the record affords all the data requisite for approximate accuracy, which is my aim. The statement of the Union strength has been care- fully, and presumably faithfully, compiled in Livermore's Civil War Numbers and Losses, and it is here quoted:

Sixth and iQth Corps " effectives," Eighth Corps, .....

Kitching's Division, ....

Total infantry and artillery,.

Deduct regulars detached, Deduct losses October [3th,.

Actual infantry and artillery .... Effective cavalry, ......

Total all, 30,829

For palpable error Livermore's Confederate table is rejected, and the following is taken from the record:

Early 's effective infantry and artillery, September 3oth return, 6,291. From this Gordon's Division is omitted. Its September loth return was 2,961. Deduct Winchester and Fisher's Hill