Page:Southern Historical Society Papers volume 08.djvu/231

Rh The cavalry, under General Stuart, consisted of two brigades under Hampton and Fitzhugh Lee. Hampton was left at Richmond, and Fitzhugh Lee's brigade, consisting of the First, Third, Fourth, Fifth and Ninth Virginia cavalry, accompanied the army on the Manassas campaign. The total of Stuart's force July 20th was 4,035, of which Colonel Taylor estimates that Fitzhugh Lee had 2,500. This estimate is no doubt nearly correct.

The artillery taken consisted of twenty batteries (and possibly a few more). There were the four companies constituting the Washington artillery, viz: Squiers', Richardson's, Miller's and Eshleman's; the five under Colonel S. D. Lee, viz: Eubank's, Parker's, Rhett's, Jordan's and Taylor's; three attached to Hood's division, viz: Reilly's, Bachman's and Garden's, and the following: Dixie artillery, Striblings', Maurin's, Leake's, Rodger's, Brown's, Grimes' and Anderson's batteries. This list, I think, is incomplete, and I hope some one who has the knowledge will make it correct. Colonel Taylor puts the strength of this artillery at 2,500, which seems to me an over-estimate, as artillery companies in the Confederate army were far more frequently under than over one hundred men.

In the foregoing roster of troops there may be some errors as to the assignment of a few of the regiments, for so many changes were made during the summer, that in the absence of full official reports it is sometimes hard to follow them. No reports of Anderson's division, for instance, are published, and, in consequence, I am not fully certain of the organization of Armistead's brigade. But the gross numbers will not be effected by such errors.

To sum up the entire force at General Lee's disposal between August 16 and September 2, 1862, was— In round numbers, 54,000. This I believe to be an outside estimate of the Confederate strength.

As was seen in my former letter, General Pope had 45,000 men at the time of the battle of Cedar Run, even after deducting nearly 3,000, which he claims as an error in Banks' report, but