1,000 men for one year would give 200 men to the service; this being repeated each year for three years would in the aggregate amount to 600 men furnished. On the other hand, a town which furnished 1,000 three-years" men would give to the service during the first year 200 men, leaving 800 men to serve the second year. Having acquired more efficiency and discipline than one-year's men, they would be ordinarily brought in closer contact with the enemy and incur risks that would increase the per centum of loss over the first year, which, if estimated at 25 per cent., would amount to a loss of 200 during the second year, leaving 600 men in service at the end of the second year.

This calculation, of course, excludes individual instances of severe loss in action, which cannot be estimated according to any settled principle, and which would equally affect His Excellency's hypothesis and mine. It will thus be seen that the Government receives from the one town the same number of years of actual service that it receives from the other, which is the point of His Excellency's argument, and is exactly what the method adopted by the War Department gives to each as its credit.

The errors of His Excellency's statistics are not more objectionable than his reasoning as to the amount of credits to be allowed. He is of opinion that giving credit for the "term of enlistment" is a violation of the act, which requires the President to give credit for the "term of service."

It has always been the custom of the Army, and I think, generally speaking, the usage of Congress, to treat the two phrases as identical in their signification, the rule being that all soldiers serve out their term of enlistment, unless they are killed in action, die, or are discharged for disability, in which cases their term of enlistment expires, as well as their term of service. Deserters do not desert out of the service, but are required to make good the time lost by desertion. He proposes to determine the number of days or months of service actually rendered and give credit for that amount of service, and require new men to make up the deficiency. He admits that it is impossible to arrive at the truth as to this amount, but proposes to assume that it is a given number of days, months, or years, and then adopt that supposition or guess as a basis for determining the number of credits, and the localities to whicg the objections that may be urged to this plan are-

First. It is confessedly incorrect as to numbers, and there is no claim set up that the credit would be apportioned to the sub- districts entitled to receive it, other than that each locality would receive its pro rata amount regardless of what it has hitherto furnished.

Second. Localities whose loss was the most severe on account of the gallantry of the troops they furnished would have the least credit, because His Excellency proposes to charge to their account the time lost by their men from being killed or discharged before the expiration of service. Thus, the locality which puts into service to-day a regiment of 1,000 men for three years" service, which regiment goes into action to-morrow and loses 600 men, would be entitled to a credit for one day's service for 600 men and a credit for the residue (400 men) so long as they remained in actual service, and the loss of 600 men would have to be made up by new men from the same locality, because they had failed to that extent in the amount of actual service rendered, while other localities whose troops were so fortunate as to escape would be relieved from furnishing men to make up the loss, their quota still being full.