This is the first in a series of posts planned for the next several weeks that address aspects of the Red River campaign of 1864. Rather than a chronological narrative about the campaign, I shall be examining different topics and decision points.
How many Confederates were at the battle of Mansfield (also known as Sabine Crossroads) in northwestern Louisiana on April 8, 1864?
Seems like a simple question, but its not. A starting point is Richard Taylor’s memoirs, Destruction and Reconstruction (1879). In command of the Confederates that day, Taylor wrote that he had 5,300 infantry, 500 cavalry and 3,000 cavalry for a total of 8,800 men. Some authors writing about the campaign have stopped there. But what was Taylor including in that number and what was he overlooking?
Brett wrote a great post years ago — Counting Heads: Looking At Civil War Troop Numbers for Wargaming — that talked about the difference between ‘present for duty’ (PFD) and ‘effectives’. Effectives did not include officers and other personnel not on the firing line, so it was a smaller number than PFD. I believe Taylor was using effectives as it was common for Confederate commanders to do so and Taylor seems to me to have a habit of downplaying his force strength. I believe PFD is a much more revealing statistic, especially when comparing the two sides, which is typically the purpose of counting force strengths. A mathematical formula can be used to convert effectives to an approximation of PFD: for infantry/artillery PFD = Effectives x 1.065/0.93; for cavalry PFD = Effectives x1.065/0.85.1 Applying this formula on Taylor’s numbers equals approximately 6,070 infantry, 570 artillery and 3,760 cavalry for a total of 10,400 PFD.
Steven Newton, in his book Lost for the Cause: The Confederate Army in 1864, estimates Taylor’s infantry at around 6,900 and I found Confederate references in the Official Records that during March 1864 Taylor’s infantry was about 7,000.2 About 300 had been lost at Fort De Russy in March.3 This suggests that when Taylor wrote his memoirs he was undercounting his infantry by a couple hundred and that the true PFD was around 11,000.
So are we done? No. I have identified other men Taylor wasn’t counting.
Right before the battle the Confederate Governor of Louisiana, Henry Allen, organized ten companies of state cavalry in Shreveport and led them to Mansfield. Allen claimed that these companies numbered “eight hundred officers and men” and “shared in the hard fought battles of Mansfield and Pleasant Hill, and acquitted themselves gallantly as good and efficient troops”.4 Added to the above gives around 11,800 PFD (since Allen includes officers, I am not going to convert it as I did the numbers above).
Gov. Allen also issued a proclamation on March 16th that read: “The movements expected to be made by the enemy in this quarter will require the service in the field of every son of Louisiana. …. Every person in Louisiana liable to militia duty will hold himself in readiness at a minutes call for the defense of the State“.5 Locals responded to the Governor’s call by joining the ranks — “old men shouldered their muskets and came to our assistance” wrote one Texan soldier.6 On April 18th Gov. Allen issued a second proclamation which read “The Governor takes this method to thank the militia officers and men who have so promptly responded to his call.”7 How many did so? Hard to say. In addition, there is the mystery of the paroled troops waiting to be exchanged. At Vicksburg the 3rd, 17th, 26th, 27th, 29th, 31st Louisiana were surrendered and paroled. Camps were set up for them to be reorganized while they waited for the exchange to go through.8 Later in 1864 they would form a brigade under Brigadier General Allen Thomas.9 But at the time of the battle of Mansfield they were supposed be out of action due to their paroles. Yet there is evidence that men from these units were present at the battle. How many? Again hard to say. One author wrote that “there may have been from several hundred to several thousand of them” with a footnote that “Diaries and letters in the [Mansfield State Historic] site’s collection indicate that large numbers of parolees fought in the battle.”10 I will assume several hundred for both the citizens called out by the Governor and the parolees , even though it could be a lot more, and thus conservatively put the total at about 12,500 PFD.
One more thing: In his book Taylor also mentions 4,400 “muskets” (another way of saying effectives) in the Missouri and Arkansas divisions who were marching to join him. Using Taylor’s number and the formula above, the Missouri and Arkansas troops would be about 5,040 present for duty. (Newton in Lost for the Cause puts this force at 5,051.) Should they be counted as part of the Confederate force at Mansfield? According to after action reports, the Arkansas division arrived at 3:30 pm while “The battle of Mansfield was then progressing, but Major-General Taylor not deeming it necessary to order us into the fight”.11 The Missouri division only arrived around 6 pm at which point the battle was won; it went into camp instead of taking part in the pursuit.12 I feel it is important to note the proximity and availability of these troops to Taylor.
Not counted are troops further away from Taylor that day, such as Woods’ and Parson’s cavalry brigades on their way from Texas or Hawthorn’s infantry brigade coming from Arkansas, Harrison’s Louisiana cavalry brigade on the north side of the river, or a few miscellaneous other units in Shreveport.
In conclusion, there seems to have been approximately 12,500 confederates at Mansfield with another 5,000 close at hand. I feel that this level of analysis gives a different flavor to the situation than is usually the case.
- Formula from Newton, Lost for the Cause ↩
- Official Records Series1 (OR1) – Volume 34, Part II, p. 1042 and 1043 ↩
- http://www.nps.gov/hps/abpp/battles/la017.htm ↩
- Annual Message of Governor Henry Watkins Allen, to the Legislature of the State of Louisiana. January 1865 ↩
- OR1-34-II p.1047 ↩
- Joseph Palmer Blessington, The campaigns of Walker’s Texas division, 1875 ↩
- OR1-34-III p.778 ↩
- OR1-34-II p.953 ↩
- ORI-41-III p.967 ↩
- Gary Joiner, One Damn Blunder From Beginning to End, page 96 ↩
- OR1-34-I p.604 ↩
- OR1-34-I p.602 ↩