The statistical tables here given are in many cases simply approximate.
This is the case in all similar works, but it is not always acknowledged.
Nothing is more plausible to the eye than a compact and well-arranged series of tables; and they may look just as imposing when they are not worth the paper on which they are written.
All that has been previously said about the difficulty of detailed accuracy applies in the highest force to these figures.
In some cases, for instance, men missing after a battle were promptly dropped from the regimental rolls, in other cases they were allowed to remain indefinitely; in some cases recruits were taken up on the list as soon as the names were forwarded, in other cases the men deserted on the way and never appeared on any regimental list at all. In most regiments there was a mingling of men enlisted at different dates, and subject to discharge at various times; in some cases three months men and three years men were combined; in some cases reduced regiments were consolidated.
There were cases in which, on the return of regiments, there was no mustering officer at hand to muster them out; they were then dismissed to their homes, with orders to come back on a certain day and meet the officer.
Some of these men did not take the pains to return for the mere technicality, and thus have never been legally mustered out to this day,—perhaps stand recorded as deserters.
In other cases men were transferred, while on detached service, from one regiment to another, were perhaps dropped from the rolls of one and never taken up on the rolls of the other, being thus left free to drop out of the ranks altogether if they wished.
Adjutants were ill or were killed, and their notes and memoranda were taken up and carried on, after a fashion, by some one wholly new to the work.
There are at this
moment, probably, thousands of pension applications awaiting action and likely never to be acted upon, merely from complications like these; and there were very few regiments whose tabular statements are not embarrassed by just such difficulties.
The statistics of recruits, in particular, are sometimes so confusing that they have to be ignored altogether, thus making the aggregate of some regiments appear lower than it actually was. It must be distinctly understood, therefore, that the following tables represent in many cases only approximate figures.
It is now the general tendency of military statisticians to class the mortally wounded with the killed, not with the wounded; and yet it is impossible to establish any but an arbitrary limit for this classification, since a man may be mortally wounded and yet live for many weeks or even months.
In the same way it is impossible to discriminate positively between those mortally wounded and those dying in prison; or even to determine whether a soldier—as in the case of Stonewall Jackson
—was killed accidentally or by the enemy.
In all cases the figures here given replace entirely those given in a circular of inquiry, in pamphlet form, printed and distributed from this office in 1891, and based on the comparatively scanty information then attainable.
None of the present tables are taken from those in the printed regimental histories, though much use has been made of these; but they are all based on manuscript name-lists prepared for the purpose and founded (1) on printed State records, (2) on original muster rolls and monthly returns, (3) on corrections received from the United States War Department, usually through the office of the Massachusetts Adjutant-General
, but sometimes directly.
The ‘number on regimental rolls’ is designed to include every individual who actually served with the regiment, but not re-enlistments in the same regiment.
‘Casualties by engagements’ includes men killed in action or mortally wounded, and also those missing in action, probably killed.
No attempt has been made to give statistics as to the number of wounded in action, because the method of recording these differed so greatly in different regiments—some officers recording very slight wounds and others ignoring all but serious ones—that there exists no solid basis of comparison.
The prison list, which follows later, is mainly prepared from several unofficial manuscript books on file at the Adjutant-General
's Office, purporting to give lists of Massachusetts
officers and soldiers who died
either in Confederate prisons or in the camps of paroled prisoners.
These books give also the dates of their supposed deaths, the whole being made up largely from the affidavits of returning comrades.
The precise authorship of these books is now unknown, but they were probably compiled under the general direction of Col. Gardner Tufts
, State Relief Agent
, who was in the habit of sending out men to meet returned prisoners and to obtain information as to those left behind.
This information was necessarily hearsay evidence, and in many cases may have proved incorrect, or only approximate.
It was easy to err in regard to the middle initial of a soldier's name, or to forget whether he had served in the 2d Infantry or 2d Cavalry; especially if he had, as sometimes happened to a prisoner, exchanged his own tattered uniform for that of a dead soldier of some other regiment, possibly from some other State.
The actual fact of death was also sometimes a matter of mere surmise; if the witness said that a certain soldier was ‘dying when he left,’ this man's name would naturally go upon the rolls, although his disease might in twenty-four hours have taken a favorable turn, so that the patient actually recovered.
The only possible remedy for these errors has been in comparing the lists, as carefully as possible, with the names and records to be found on the original rolls or in the best regimental histories, and this has been done.
Names not thus confirmed are marked with a star, and any additional information has been appended in footnotes.
It has been thought better to print even doubtful cases,—the doubt being indicated,—than to omit any that may possibly be authentic.