three days' collision were ascertained by Colonel John B. Bachelder, who has devoted years to it; has walked many times over every part of the field; saw in many instances the dead of the two armies while lying as they fell; talked with the wounded on the field and in hospitals; examined the head marks of the graves; has since examined many reports in manuscript, as well as those that have been published, both Federal and Confederate. It would be difficult for General Longstreet, from his own personal knowledge of the field of battle, to detect in the map a single error, though, of course, it may not be wholly free from them.
Now, as to the official reports to which General Longstreet refers, and on which he relies to prove that I failed to do what he erroneously says I was ordered to do. General Anderson, he says, states in his report that "a strong fire was poured upon our right flank, which had become detached from our left." This does not show that my brigade uncovered McLaws' left any more than it does that he uncovered my right. I have stated that my right was uncovered when I recalled my brigade, and this has been confirmed by others who were present at the time. Then follows an extract which, according to General Longstreet, is from General Lee's report, as follows: "But having become separated from McLaws, Wilcox's and Wright's brigades advanced with great gallantry, breaking successive lines of the enemy's infantry, and compelling him to abandon much of his artillery. Wilcox reached the foot and Wright gained the crest of the ridge itself, driving the enemy down the opposite side, but having become separated from McLaws, and gone beyond the other two brigades of the division, they were attacked in front and on both flanks, and compelled to retire, being unable to bring off any of the captured artillery. McLaws' left also fell back." * * *
General Longstreet does not claim that General Lee wrote from personal knowledge; he knows that he did not so write. But he and myself, and in fact every officer of the Army of Northern Virginia who served under him, know that his official reports are marvelously accurate. I do not of my own knowledge know of inaccuracies in any one of them except that of Gettysburg, and in that they are in unimportant details. In that report he refers to four of Anderson's brigades—Posey's being one—advancing and taking part in the battle late in the afternoon of July 2d. There were but three of those brigades that were engaged, and Posey's was not one of them. General Lee refers to Wilcox's and Wright's