expenditures must average the correct proportion of the totals available, due regard being had to the character of the demands arising. Every effort is made to avoid obligations which will result in the overexpenditure of any given appropriation. Consequently, when it is seen that the monthly demands on such an appropriation are running ahead of the allotted sums, costs which would otherwise be charged to this appropriation are located under other appropriations which, as has been explained in preceding paragraphs, are also chargeable with such expenditures. Attention is largely concentrated on keeping accurate records showing the available balances of the sources of funds (i. e., the appropriations), rather than on clearly showing the exact objects for which sums are expended. It is true that the objects of expenditure are faithfully recorded through the cost accounting system, but their importance is diminished owing to the interest in the appropriation accounting.

The acounting instructions, in treating of this matter at some length, conclude as follows:

Appropriation charges show the source or legal authorization for the expenditure of the various sums of money devoted to output and maintenance. Cost charges show the disposition of the money, the final resting place of values. The two classifications are independent and cannot be made identical without impairing the legality of one or the correctness and administrative value of the other. The cost of an item of output cannot be correctly determined except by taking into consideration the elements thereof contributed by numerous appropriations.

Now the bureaus concerned, being mainly interested in the charges to their appropriations, do not follow with equal attention the question of costs under the various expenditure heads, especially since those accumulated costs affect a variety of appropriations with many of which a given bureau is not concerned. A glance at the diagram gives some idea of the extent to which this diffusion of sources is carried. It does not tell the whole story, however, as it does not bring out the facts relating to individual pieces of work, or jobs. Not only is it true that several appropriations have charged to them jobs of similar nature, but also it is necessary that the cost of any single job be made up of charges from a variety of appropriations. The resulting complexity is obvious.

Imagine an industrial plant which, in additon to the usual plan of exhibiting its receipts and expenditures in all the detail necessary to give a clear and comprehensive understanding of

the business conducted, including a complete cost system, also endeavored to show how every element of income was distributed, tracing each sum received in such a way as to indicate the final disposition of every sum throughout the multiplication of bookkeeping details. Such a plant might then be in position to answer a question like this: "You have spent $10,000 for repairs to buildings. Where did you get the money?" The answer would be: "The sum spent for repairs to buildings was obtained thus: Sales to Wm. Smith & Co., $43.10, interest on investments $7.42, merchandise account $8421, royalties account $3.02, and so on through a list of receipts. Such a system would be well-nigh impossible and perfectly useless. It would, however, be closely similar to our naval appropriation cost accounting system, where actually more attention is given to following the status and distribution of the sources of supply (the appropriations) than to the objects of expenditure.

To improve conditions the first step would be to have all estimates submitted under the same headings as those used for the final expenditures themselves. In this connection the following is quoted from the conference on accounting, June, 1914:

It will not be possible to put into effect an entirely suitable and satisfactory system until the methods and classifications used in submitting estimates to Congress and the wording of appropriation bills are correspondingly modified and revised so that all estimating, appropriating accounting instructions are harmonized and so that identical terms, classifications and meanings will exist throughout.

However, even were estimates submitted under the logical headings provided by the accounting system, if these estimates after approval were embodied into appropriations, as the present estimates are, the situation would be only slightly improved. There would still remain the fact that several distinct sources of supply of funds have to be entered on the treasury books, and that complications would exist due to the necessity of tracing each source to its final disposition, as well as analyzing each expenditure head, to show whether two or more appropriations were involved. The real cure for our troubles is to let estimates remain estimates and not to convert them into separate and distinct appropriations at all, except that the total of all estimates would make the total of the Naval Appropriation Act.

Is any advantage, in the way of strict accountability or accurate information to Congress, gained by the present system of

many appropriations? To answer intelligently, let us examine the relations existing between the Navy Department and the Treasury Department and Congress, in regard to these appropriations.

The Treasury Department is charged with the duty of determining whether or not any given voucher covering expenditures of funds by the Navy Department is a proper voucher, and if all legal requirements have been complied with. If found correct in that particular the treasury then proceeds to charge up the voucher to the appropriation indicated on its face, provided of course such appropriation is properly chargeable with such an expenditure. There are thus two distinct questions involved in the auditing of a voucher: First, is it in legal form? Second, is it chargeable to the appropriation named? The treasury is the only authority on the first question, and the Navy Department must conform absolutely to its rulings. As regards the second question, the treasury is necessarily at a loss, for, as has been shown, many appropriations are often chargeable with a given expenditure. The best the treasury can do is to satisfy itself that at least the appropriation inscribed on the voucher is one of those to which such an expenditure may be charged. It therefore rests with the Navy Department to keep the record straight in regard to appropriations, and that is what actually takes place. It is true that there is a large class of vouchers each of which by its nature could be charged to only one appropriation, and the treasury can judge of itself in such a case whether or not the correct appropriation has been inscribed. But as there is also a great number of vouchers for which the treasury has to rely on the navy for the correct appropriation classification, the effect is practically to leave the whole decision in the hands of the navy. Consequently, no advantage in the way of accountability is gained by maintaining the system of many appropriations, considering only the relations existing between the navy and treasury.

As regards the relations between the Navy Department and Congress, we find that Congress approves certain estimates, and that in passing the naval bill these estimates are solidified into separate and distinct appropriations. If emergencies later arise. which require the expenditure of sums greater than those estimated in certain cases, it becomes necessary to obtain deficiency appropriations from Congress. At the same time other estimates

may be found greater than needed, but no method is available of using overestimates to offset underestimates, except that of shifting charges, to which reference has been made. This method is not to be commended. How much more sensible it would be to definitely recognize that estimates are in fact only estimates, and to enact a naval appropriation bill providing for a total amount made up of the sum of the estimates which have been approved by both Houses of Congress, but not making separate and distinct appropriations in the case of each estimate.

Such a bill should be composed of general estimates and specific estimates. The first would consist of the regular working amounts which have been called the leading appropriations, and also the supplemental amounts for similar purposes; the second would be those sums allowed by Congress for specific projects, such as for public works, and also for amounts allowed for increase of the navy. The second class of estimates would in reality be allotments, the amounts not to be exceeded; while the first class would be expended as nearly as practicable in accordance with the estimates, with proper explanations of all variations. The general estimates, should, however, be made up to cover the objects of expenditure as laid down by the accounting system, accompanied by such text as would be necessary to substantiate the requests; the present indefinite and misleading wording should of course be abandoned.

A good example of what is meant is found in the current appropriation "Pay of the Navy." Here we find a group of separate estimates for a variety of purposes for which the appropriation is to be used, and then the clause," and the money herein specifically appropriated for 'Pay of the Navy' shall be disbursed and accounted for in accordance with existing law as Pay of the Navy' and for that purpose shall constitute one fund." The fact is here recognized that the various estimates are based on probabilities, but that so long as the total is not exceeded, variations from these estimates are permitted.

Another example, not quite so clear, is found under Ordnance and Ordnance Stores" of the current act, where we find: "Provided further, That ordnance materials procured under the various ordnance appropriations shall hereafter be available for issue, to meet the general needs of the naval service, under the appropriation from which procured." The provision has the virtual effect of closing all the ordnance appropriations into

one appropriation, so far as the purchase and use of material is concerned.

An appropriation act passed in the way described above would then be made up of detailed estimates, in which the purpose of each sum would be definitely described. Certain sums would be fixed allotments, not to be exceeded. The others would be simply estimates. The total amount appropriated would then be entered on the treasury books as one sum.

In expending this sum the Treasury Department would continue to exercise its function of passing on the legality of the vouchers submitted. It would make no attempt to classify expenditures under the various estimates. On the other hand, the Navy Department would report to Congress all expenditures made aaginst the corresponding estimates, and would give suitable explanations of all variations made. The estimates and expenditures being made under the same descriptive headings, a complete and satisfactory report would be feasible. Congress would have as much and even more control over the moneys expended than at present, since the best that can now be done is to show on the one hand the expenditures made from the many appropriations, and on the other hand to describe in entirely separate tables the purposes for which money was used.

Another feature should be the cumulation of appropriations from one year to the next, so that there should always be on the treasury books but one appropriation running for current needs. The ordinary practice is to maintain the sums of each fiscal year separate from other fiscal years, and this enormously increases the complexity of accounting. In each appropriation act the balances remaining from the previous year should be re-appropriated for the ensuing year, and closed in to the new appropriation. This has already been done in the current appropriation bill, so far as the appropriations for the preceding year are concerned; each distinct appropriation balance being carried forward and credited to its successor. This principle should be carried further, and all balances should be closed into the one fund.

It should be evident that the plan proposed differs entirely from the method sometimes designated as that of "lump appropriations," in that detailed estimates are made and finally reported up, which is not true of the usual lump appropriation.

To make clear the plan proposed, the amounts found in the current appropriation act have been rearranged and distributed in