troops commence firing rapidly in the formation of skirmishers (not to attract our fire) from Redoubts 2,3, and 4 directly to the woods to their front and our left. This was the movement upon our flank which annoyed us so much, came near several time in the afternoon attaining our rear, and resulted in the desperate fighting which raged so continuously all the afternoon up the wooded ravine from C to D and generally through the woods thereabouts. I then returned to General Hooker and reported the results of my reconnaissance, which amounted to the following conclusions:

1st. That the ravine at C was impracticable for artillery designated to be put into position on the enemy's plateau at this point, and that it would be difficult to get a battery into position on our side of the ravine in consequence of the large number of trees we would be obliged to fell to unmask it; besides, I had serious doubts whether the battery so placed would prove sufficient, in consequence of the crest of the ravine toward the enemy being higher than the one on our side.

2nd. That our troops had better be passed to the front either up the road or around through the woods, instead of through the entanglement.

3rd. That the open space constituting the enemy's position extended apparently (from my point of view) considerably beyond his right redoubt, thus affording a probable chance of getting at this right or rear. (Redoubt No. 1. I could not see.)

The last conclusion or supposition I communicated to the brigadier- general commanding on his arrival. At about 2 p.m. the directed me to pass around to the enemy's right and see what chances existed of turning his position. I retreated the Hampton road to E, passed Averell's cavalry at F (his pickets were some distance in advance),and went far enough up the road on Allen's estate to derive the conclusion that the opening where the enemy's works were and that on Allen's estate were either continuous or very near together, and that a movement around them might be decisive. This conclusion proves to have been partly erroneous (see sketch), the two openings being separated by woods more than a mile across, and two small streams, branches of the Achershape, and connected by a good straight road, G H (see sketch), unobstructed, passing through the woods and crossing the two streams upon mill-dams in perfect order. This conclusion I immediately reported to the general.

At about 4.30 p.m., by the general's order, I returned to Allen's estate to hasten General Emory's proposed movement upon the enemy's right and rear, and overtook him moving forward toward G with three regiments of infantry, one battery, and a detachment of cavalry. Arrived at G, we discovered the mill-pond K, crossed on the dam by the road G H. After considerable deliberation, General Emory decided than this force was inadequate to attempt the movement along G H upon the enemy's right and rear, he at that time being of course ignorant of the fact that the road struck the enemy's right at Redoubt No. 1, and there passed to his (the enemy's) rear at a point nearer Williamsburg than was the enemy's center at Fort Magruder. General Emory's force was undoubtedly too weak to attempt cutting off the enemy's retreat. A careful examination of the enemy's position intensifies the regret that the pressing and repeated demands by the brigadier-general commanding, made immediately on his arrival at Hooker's position and subsequently, for

re-enforcements had not been promptly complied with, in order that this his desire to turn the enemy's position by diverting Kearny's division around by Allen's estate might be carried out. Had these conditions been fulfilled, the enemy must have countermanded his flank