The engineer officers of the corps being occupied on other parts of the line, no assistance or advice was asked or received, except I altered a portion of their rifle-pits so as to contract the line and not be forced to fire over our own men. The first day of the siege, before the enemy had closed in, I dug two lines of pits for the pickets, about 80 and 30 yards from the fort-the second line to rally in if driven from the first.

On the 20th instant, at my request, the Seventeenth Michigan made a sortie in the night, driving the enemy's pickets and burning a house which they occupied. The rebel sharpshooters were very annoying, causing casualties in the fort every day. I stuck brush in the parapet, along the interior crest, so as to screen us from sight, and enable us to look out without being seen. I also covered, the embrasures with bags and barrels, so arranged as to see out without being seen.

On the 21st, the enemy had a parallel about 300 yards off, half enveloping the northwest bastion. General Ferrero had now taken up quarters in the fort, in a small bomb-proof, built for telegraphic operations.

On the morning of 23d, the attack on the enemy's parallel was made. This attack I strongly opposed.

On the 25th, a battery was discovered on the other bank on the Holston, 150 feet above us (six guns), commanding and having a view of all in the fort. They also had on west front an embrasure battery of six 12-pounders and one 20-pounder Parrott; on north front embrasure battery of two 20-pounder Parrotts, same two 3-inch guns, two other two-gun batteries, caliber unknown (probably

3-inch). These varied from 700 to 1,500 yards in distance from

us-the one across the river 2,500 yards. Every man in the fort had his place assigned him, and ate and slept at his place, so, on an alarm, they only rose and crouched by the parapet. At night 1 man in 4 was awake, 2 officers and 2 non-commissioned officers, besides the regular guard on picket. On an alarm, each man then up woke the three sleeping near him; thus the garrison was at once ready for an attack. I made an embrasure in such manner that, by taking out a few shovels of earth, I could train a gun on the northwest bastion, sweeping its ditch and parapet. The parapet there was strengthened. The whole fort was well fitted with traverses to protect our men, as the enemy had a reverse and enfilading fire on each front. In front of the northwest bastion I made an abatis, concealed from the enemy by a small rise of ground, and inside of the abatis a little entanglement of telegraph wire. We worked night and day, but still at many places we went out and in the fort over the parapet and through the shallow ditch. The work was now known as Fort Sanders, and was very weak, and should have fallen by the ordinary chances of warfare; but the garrison were picked men. We had many alarms and exchanged shots from time to time with the enemy.

About 10 p.m., November 28, the enemy captured most of the outer line of pickets, and drove the others into the fort. Skirmishing and firing continued for two hours; at the end of which time we had not a picket 20 yards out from the fort, and the enemy had secured the crest of the ridge which the work was on, beneath which they could mass troops night or day, within 80 yards of the work, without our knowledge. In spite of the opposite opinion held by most, I prepared for an attack at daybreak.