per cent. in his strength as compared with May 31.
These are the only two divisions whose returns near the date of the battle have been found, so far as I know.
To sum up — Stuart's cavalry was increased by 3,000 after May 31, but like the Federal cavalry had been seriously lessened by severe marching and fighting.
If the Federal cavalry could only muster 12,000 out of 16,000 on July 1, Stuart could not have had over 10,000 or 11,000 out of 13,300.
But of Stuart's seven brigades three (Robertson's, Jones's and Imboden's) were not present at Gettysburg, having been engaged (like French's Federal division at Frederick, which is not included in Meade's numbers) in protecting communications, guarding supplies, &c., in the rear.
So Stuart had 6,000 or 7,000 cavalry at Gettysburg.
The Confederate infantry and artillery numbered 64,159 less the small losses in the battles about Winchester, and the far greater losses from the exhaustion of a march of two hundred miles. These losses ha