advices that Bragg was in full retreat, which seemed to him to render haste unnecessary.
No proof exists that Bragg knew, before the battle, of the desperate effort made by Rosecrans to thwart him by massing his forces on the left, but it is not probable that such knowledge would have made him hesitate to offer battle; for, as is shown in the exhortation addressed to his army, he must have been inspired with faith in success by the reinforcements of his command already received and about to arrive. There is a riddle, however, in this connection which the rebel records fail to clear up. While, according to Bragg's report of the battle, a considerable body had joined him some days before, Longstreet's corps was still on the way to him when he decided to attack on the 18th, and could not reach the field in time to support him. How did it happen that he deliberately did not wait for the advent of the flower of Lee's army, sent to him especially to ensure a victory over Rosecrans? Had he fought on the 18th, Longstreet would have been too late for any share in the action. As it was, he arrived only in time to contribute to the rebel success on the second day. No light has ever been thrown on Bragg's motive.
The battle about to be fought forms no exception to the all but general rule that, in the great actions of the Civil War, the losing as well as the winning side claimed that it had to contend against far superior numbers. Yet, as will be seen from the following comparison, the contestants in this struggle were almost evenly matched, there being but a slight numerical superiority on the Union side. According to the tri-monthly return of the Army of the Cumberland of September 10, the Fourteenth Army Corps had then present for duty 22,781 officers and men, the Twentieth 13,156, and the Twenty-first 14,660, making a total of 50,597 infantry and artillery. This aggregate was reduced by about ten per cent. by the disabled and stragglers lost during the continuous and hard marching between the 10th and the 19th, thus leaving (say) about 45,000 effectives,
