Assistant Attorney General McReynolds and Solicitor General Hoyt for petitioner.
[197 U.S. 135, 138] Messrs. Albert Comstock, William R. Sears, Aldis B. Browne, Howard T. Walden, and Page, McCutcheon, & Knight for respondents.
Mr. Justice Holmes delivered the opinion of the court:
,Whitridge, White, & Co., the respondents, on June 18, 1900, imported from India certain gunnies, invoiced in rupees. The invoice contained a certificate from the American consui, dated April 19, 1900, that the exchange value of the rupee at that date was 32 cents, estimated in United States gold dollars. For the purpose of ascertaining the ad valorem duties under the act of July 24, 1897 (30 Stat. at L. 151, chap. 11, U. S. Comp. Stat. 1901, p. 1663), schedule J., clause 341, in July, 1900, the collector of the port of Baltimore estimated the value of the merchandise at the date of the consular certificate by converting the invoice value into dollars, taking the rupees at 32 cents. The importers entered protest and the collector reliquidated the entry, taking the rupee at 20.7 cents. The Secretary of the Treasury, on June 6, 1901, wrote that satisfactory evidence had been produced to him that the value of the rupee was 32 cents at the date of the consul's certificate, and directed a reliquidation at that rate. The collector of the port reliquidated accordingly on June 12, 1901. The importers (respondents) protested, and the matter was submitted to the board of general appraisers in New York. Act of June 10, 1890 (26 Stat. at L. 137, chap. 407, 14, U. S. Comp. Stat. 1901, p. 1931). The board found that the exchange value of the rupee at the date of certification was 32 cents, but that the metal value was 20.7 cents, as estimated by the Director of the Mint and proclaimed by the Secretary of the Treasury for the quarter year beginning April 1, 1900, and ruled that the latter rate should have been taken, and directed a reliquidation on that footing. The collector appealed to the circuit court and then to the circuit court of appeals, both of which sustained the board of appraisers. 129 Fed. 33. The United States then obtained a writ of [197 U.S. 135, 141] certiorari from this court. The question is whether the Secretary of the Treasury had power to order reliquidation at the rate of 32 cents.
There is, to be sure, a preliminary question as to the conclusiveness of the Secretary's action under the statute. Technically it does not appear that his decision was not based on a finding as to the metal value of the rupee; that is to say, as to the value on April 19, 1900, in fractions of a gold dollar, of the silver contained in the coin. If the decision were based on such a finding we may assume that it would not be open to review. United States v. Klingenberg, 153 U.S. 93 , 38 L. ed. 647, 14 Sup. Ct. Rep. 790. But the greater part, at least, of the argument was made on a different assumption, which, in view of our conclusion, we shall adopt. We do so the more readily because, upon the public and well-known facts, it is not to be supposed that the imagined finding as to the value of silver was made, and the policy of the Treasury Department to adopt the exchange value of rupees was well-known and publicly declared. It would not be consistent with the honor of the government to take the exchange value and then to cover itself from correction, if it was wrong, by suggesting that it had gone upon a different ground, when that ground could not have been taken by any one knowing the prices of the time. There is another argument for the conclusiveness of the Secretary's action which is so closely connected with the merits that we shall not separate it from our general discussion of the act.
The power of the Secretary depends on the construction of the act of August 27, 1894 (28 Stat. at L. 509, 552, chap. 349, 25, U. S. Comp. Stat. 1901, p. 2375).
On the other side we start with the consideration that, to an ad valorem tax, it must be an object to ascertain the true value of the thing taxed at the time as of which it is taxed, and that the invoice price is referred to only to that end. The history of the statutes shows a series of continually closer approximations to it, and to our mind helps the contention of the government, not that of the other side. The statutes began by fixing the rates for specified coins absolutely. Then in 1873, they provided in the language of the first part of 25, quoted above, for an annual estimate by the Director of the Mint, and a proclamation. Act of March 3, 1873 (17 Stat. at L. 602, chap. 268, Rev. Stat. 3564, U. S. Comp. Stat. 1901, p. 2428). In 1890 the estimate was required to be quarterly, instead of for the year. Act of October 1, 1890 (26 Stat. at L. 567, 624, [197 U.S. 135, 143] chap. 1244, 52). Finally, on August 27, 1894, the statute received its present form, with the proviso from which the Secretary derives his clearest grant of power. The general purpose of this proviso undeniably is to secure a closer approximation still. In construing it we must bear this obvious purpose in mind. While no doubt the grammatical and logical scope of a proviso is confined to the subjectmatter of the principal clause, we cannot forget that in practice no such limit is observed, and when, as here, we are dealing with an addition made in new circumstances to a form of words adopted many years before, the general purpose is a more important aid to the meaning than any rule which grammar or formal logic may lay down. Georgia R. & Bkg. Co. v. Smith, 128 U.S. 174, 181 , 32 S. L. ed. 377, 380, 9 Sup. Ct. Rep. 47.
If the proviso were a serarate subsequent act we should note that the case in which the Secretary is authorized to order a reliquidation is not confined in terms to a difference in the value of standard coins in circulation, but exists whenever there is such a difference in the value of the foreign money specified in the invoice. The invoice is required to be made out in the currency of the country of export or the currency actually paid, which may not be coins at all. Act of June 10, 1890 (26 Stat. at L. 131, chap. 407, 2, U. S. Comp. Stat. 1901, p. 1886). It is true that the difference referred to in the proviso is a difference from the proclaimed value, and that the proclaimed value has reference to standard coins. Whether, in view of this fact and of Rev. Stat. 2903 (U. S. Comp. Stat. 1901, p. 1922), the words would cover a difference in value between paper expressed in terms of current coin and current coin, if paper were the currency shown by the invoice or the consul's certificate to be the currency to which the invoice referred, need not be considered. That question did not arise in Cramer v. Arthur, 102 U.S. 612 , 26 L. ed. 259. However that may be, suppose that the currency mentioned in the invoice, although coined, was a token currency having by legislative fiat the value of a fraction of some current coin of universal worth, but itself having no such worth derived from the metal it contained. Such a [197 U.S. 135, 144] token might vary in value much below or above the fraction of the coin by which it purported to be measured. Suppose that the value of the latter coin only had been proclaimed. It would be going far to say that the Secretary could not order a reliquidation upon a variance of more than 10 per cent between the value of the token currency in the invoice and the proclaimed value of the governing coin.
The case last put is the case at bar, except that it is not admitted that the rupee was technically a mere token, and that the value of the rupee itself had been proclaimed, subject to a note-'value of the rupee to be determined by consular certificate.' At that time, although it was not noted until a little later in the year by the Director of the Mint, India was on a gold basis. As the rupee had a legally fixed ratio to another coin also valued by the Director,-the gold pound,-it is plain that the value of the rupee as so much silver and its value as a fraction of a pound might fall apart, and yet both be given by the Director's tables. It would be giving a very literal construction to the body of 25 to say that it forbade the Secretary to take the fraction of the pound rather than the silver bullion as the measure of the value of goods, if the former represented the unit of actual cost. But, supposing that the fraction of the pound was the unit of cost, it seems to us that at least under the proviso, if not under the body of the section, the Secretary could order a reliquidation on the basis of the units actually used. It would be simply a correction in conformity with the truth and the actual meaning of the words of the invoice. The other argument for the conclusiveness of the Secretary's action, to which we referred at the outset, was that, for all that appears, this may have been what happened. The gold which the rupee represents is 1 shilling and 4 pence, or about 32 cents. But, as in this case the exchange value and the value as a fraction of a pound were the same, it does not matter to our decision whether we say that in such circumstances the action of the Secretary was conclusive or say that it was right. [197 U.S. 135, 145] We have shown that, in our opinion, the proviso, if not the body of 25, would have warranted the action of the Secretary if it had been a later independent statute. We are of opinion that it is not to be construed differently because of its form. In addition to the considerations which we have mentioned, we are confirmed in our view by the facts which were known at the time. It is true that the most conspicuous recent event was the fluctuation in the value of silver. But the movement of silver, especially after the repeal of the Sherman act, on November 1, 1893 (28 Stat. at L. 4, chap. 8, U. S. Comp. Stat. 1901, p. 2355), had been downward, and the proviso contemplated at least equally a possible rise in the foreign money with which it dealt. On the other hand, there was before Congress the Herschell report on the coinage of silver in India, of which six thousand copies had been ordered to be printed by a resolve of the Senate, concurred in by the House (28 Stat. at L. Appx. p. 5), and which had been printed in 1893. This report recommended the closing of the mints against the free coinage of silver, and predicted as a consequence the divergence between the intrinsic trinsic value of the rupee and the value of its ratio to the pound as fixed, taken hypothetically as 1 shilling and 4 pence. It contemplated even a raising of the ratio as possible. The report was followed by the closing of the mints in the same year, and the result predicted came to pass. However small may have been the imports from India in 1894, the fact predicted by the Herschell report was one of the most striking incidents in the recent financial history of the world, and we cannot suppose that it was not considered when the proviso was passed. Before the date of this export gold was adopted as the standard, and the ratio of the rupee fixed at 15 to 1, or 1 shilling and 4 pence, in 1899. The exchange value did not change very much, remaining at near the conventional ratio, but the decline in bullion made the divergence referred to more marked. It was objected that some of the facts which we have mentioned were not proved in the case, but they are public facts, and when we are asked to declare that the Secretary [197 U.S. 135, 146] exceeded his powers we have to consider what might have been before his mind.
As we have said, it would be only by a very literal construction of the earlier part of 25 that the collectors would be bound to estimate the value of a cargo invoiced in rupees by the bullion of the rupee when, in the invoice, rupee meant a certain fraction of a pound. But, however that may be, we are of opinion that when the Secretary has satisfactory evidence of that state of facts, under the proviso he is authorized to order a reliquidation in order to make the value in United States currency correspond with the actual value of the goods. It is not necessary to consider any wider problems as to the power of the Secretary. We confine our decision to the particular case.