delivered the opinion of the court:
This is an appeal from the judgment of the United States Customs-Court, First Division, one judge dissenting. No question of classification is involved and the actual issue is a novel one upon which, until this case, no court has been called to pass, so far as we can determine.
The merchandise consists of certain species of liquids which, in common parlance, are frequently referred to as “Chinese wines,” or “Chinese medicinal wines.” It was imported in March 1934, and classified by the collector of customs at the port of San Francisco,. Calif., under paragraph 24 of the Tariff Act of 1930, which paragraph reads as follows:
Par. 24. Chemical elements, and chemical and medicinal compounds, preparations, mixtures, and salts, distilled or essential oils, expressed or extracted oils, animal oils and greases, ethers and esters, flavoring and other extracts, and' natural or synthetic fruit flavors, fruit esters, oils and essences, all the foregoing, and their combinations when containing alcohol, and all articles consisting of vegetable or mineral objects immersed or placed in, or saturated with, alcohol except perfumery and spirit varnishes, and all alcoholic compounds not specially provided for, if containing 20 per centum of alcohol or less, 20 cents per pound' and 25 per centum ad valorem; containing more than 20 per centum and not. more than 50 per centum of alcohol, 40 cents per pound and 25 per centum ad valorem; containing more than 50 per centum of alcohol, 80 cents per pound and1 25 per centum ad valorem.
The correctness of the classification is conceded by both parties.
The duty assessment was at 80 cents per pound and 25 per centum ad valorem, the wines having been found to contain more than 50 per centum of alcohol by volume.
It is the contention of importer that the alcoholic content should be determined by weight and protest was made accordingly. If so determined, it appears that the merchandise would fall within that provision of the paragraph reading—
containing more than 20 per centum and not more than 50 per centum of alcohol,, 40 cents per pound and 25 per centum ad valorem.
*270The proceeding is, in effect, a suit against the Government by way of protest, as the statute provides, to recover the difference between the two rates. The protest was overruled by the majority of the trial court and the importer appealed here.
The sole issue is succinctly stated in the brief on behalf of appellant as follows:
Should the percentage of the alcoholic content of merchandise which is normally dutiable under the provisions of paragraph 24, Tariff Act of 1930, be determined by volume or should it be determined by weight?
It will be noted that paragraph 24, supra, is silent as to the method or standard to be used in determining the alcoholic content of the commodities classifiable thereunder for tariff purposes. It makes no provision for determination by either weight or volume. No regulation of the Treasury Department with respect to the matter seems to have been in existence at the time of importation, and it is pointed out that paragraph 24, supra, does not contain any provision for regulations.
In September 1934, T. D. 47361, 66 Treas. Dec. 260, was promulgated by the Acting Commissioner of Customs, but it is of no consequence in this case.
It is pointed out by counsel for. appellant that the requirement of paragraph 24, supra, is that the articles therein provided for in their complete state shall be assessed with duty per pound — that is, by weight. Also, it is pointed out that paragraph 24, supra, covers—
* * * all articles consisting of vegetable or mineral objects immersed or placed in, or saturated with, alcohol * * *.
One of the witnesses called by the importer testified that in determining the percentage of alcohol in the articles last above described he would have to do it by weight, saying “I couldn’t determine it by volume, because I couldn’t get the volume of the irregular sized material.” This statement was not contradicted in the record, and the only comment found in the Government’s brief concerning it is:
Mr. McElligott’s testimony that he would have to determine the alcoholic content of vegetable substances enumerated under paragraph 24, supra, by volume [weight] is beside the point, as the merchandise involved is medicinal preparations and not vegetable substances.
. From the foregoing facts it is argued on behalf of appellant that the paragraph should be construed to mean that the alcoholic content of all the articles therein provided for should be determined by weight, it being argued, in effect, first, that it is something of an anomaly to determine by volume the amount of one ingredient of an article where the entire article is to be assessed upon a weight basis; and, second, that since in some articles the alcoholic content must be determined by weight the rule of noscitur a sociis applies with respect to the *271others. Also, the doctrine of expressio unius est exclusio alterius is invoked.
It appears that in the instance of spirits, wines and other beverages, provided for in schedule 8 (paragraphs 801 to 815, inclusive) of the Tariff Act of 1930, the alcoholic content is determined by volume, this being expressly provided by the statute. See paragraph 811, Tariff Act of 1930, and in connection therewith 26 USCA, § 1150 (a), •§ 1158 (b), § 1158 (c) and § 1158 (d). Also in prohibition matters (27 USCA § 4) and in pure food and drug matters (21 USCA § 3 and Regulation 25 (a) for enforcement of Pure Food and Drug Act, 1930) the volume standard is provided.
One of appellant’s arguments is that where Congress intended that the alcoholic content should be determined by volume, it made •express provision therefor and since it provided no standard in paragraph 24, it is urged that in view of certain testimony hereinafter discussed and in the light of the wording already alluded to, it is a legitimate conclusion that it was the intention of Congress in the instance of the so-called medicinal wines to apply the weight standard.
The Government states in its brief that it “does not claim that percentages of alcohol cannot be determined by weight, nor does the Government dispute the accuracy of the weight method,” but it earnestly contends that the “volume method is the method to be followed in determining the percentage of alcoholic content of the involved merchandise.”
This contention is based largely upon what the Government claims is shown to have been a long-continued administrative practice, namely, that from as far back as the 1913 tariff act, at least, the alcoholic content of merchandise of the kind at bar has been determined by the volume standard.
The decision of the majority of the trial court was based largely, if not wholly, upon what it found to have been that administrative practice. The dissenting judge was of the opinion that such practice had not been “clearly proved” with respect to the “alcoholic medicinal substances of paragraph 24 * * *.”
Certain of appellant’s assignments of error are leveled at the .admission of testimony by witnesses for the Government respecting the practice of the Prohibition Bureau, the Internal Revenue Bureau and the Food and Drug Administration respecting the practice in or by those agencies in determining for their purposes the alcoholic content of distilled spirits and of “Chinese medicinal wines.” Objections to this line of testimony were made at the time the questions were asked .and in instances where such objections were overruled exceptions were duly taken and noted of record.
So far as the testimony relating to the practice in determining the alcoholic content of distilled spirits is concerned, we regard it as wholly *272immaterial. The statute fixes the volume standard, and testimony that that practice was followed in cases involving such neither adds to-nor detracts from the sum total of human knowledge upon that point. Even were it held to be error it would fall within the classification of harmless error and would present no reason for reversal and remand of the case for a new trial.
A somewhat different situation exists with respect to the testimony regarding the practice with reference to determining the alcoholic content of merchandise such as that at bar by the Internal Revenue-Bureau for its purposes, the Food and Drug Administration for its-purposes, and the Prohibition Bureau (during the time of Federal prohibition) for its purposes, although the statutes likewise fix the volume standard in those instances. Had this case been tried before a jury the admission of such evidence might have been prejudicial to appellant’s case, unless most careful instructions relative to it had been given by the court. However, it was not tried before a jury but before a discriminating court, and from a careful reading of the majority opinion we do not conclude that any particular weight was given it by that tribunal.
A pertinent excerpt from the majority opinion reads:
* * * four witnesses who testified on behalf of the defendant [the Government], each of whom was employed by the Government as a chemist, testified that over a period extending back at least seventeen years in the administration of the customs and revenue laws of the United States the alcoholic content of goods of the character of those at bar was determined by volume, and it is apparent that the determination of alcoholic content of liquid medicinal preparations containing alcohol dutiable under paragraph 24 and its predecessors has, during such-time, been uniformly made on the basis of volume. * * *.
It is true that the foregoing excerpt contains the words “and revenue-laws” but, taken as a whole, we think the majority really gave weight only to the evidence respecting the practice as to “liquid medicinal preparations * * * dutiable under paragraph 24 and its predecessors * *
It may be said that the testimony in the case was taken before a single judge sitting in California on circuit. It was taken at different sessions, Cline, J., presiding at the first session and Evans, J., at the second. Neither of these judges was a member of the First Division which finally passed upon the case. We have examined the record with care. In many instances objections were sustained and, upon the whole, we do not feel that the admission of any of the testimony which was admitted proved prejudicial to appellant’s case and constituted reversible error.
In this somewhat peculiar jurisdiction it not infrequently happens, that testimony is taken before a single judge on circuit who does not. participate in the final determination of the case. Such judge may, not improperly in our opinion, admit testimony, even though it borders *273upon the doubtful, and leave it to the full division to winnow the wheat from the chaff, and so long as it does not appear that illegitimate evidence influenced the decision no harm results and no reversible error is committed. We may add that, upon the whole, we regard the rulings of the single judges before whom the evidence in this case was taken as being quite fair. They obviously sought to keep it within legitimate bounds.
When the testimony is sifted and the part relating solely to the practice in determining the alcoholic content of these so-called medicinal wines for customs purposes is separated from that relating to the practice for other purposes, the former is found to be rather meager, and it is so mingled with the latter as to render the task of separation difficult. However, it does appear that some of the witnesses for the •Government, who were employed as examiners for customs purposes, testified that they determined and reported alcoholic content by volume, and the testimony carried the practice back over a long period •of years. We do not set forth this testimony in detail here because, under the view which we take of the case, it is not deemed necessary.
No question of classification is involved here and, as we view the issue, the standard to be taken presents a question of law solely. It is presumed that the amount of the alcoholic content found and reported was correct if the volume standard should have been applied, and this appellant does not question, but the purely legal question of whether that was the correct standard must be determined without any particular reference to the rule relating to presumption of correctness.
We are of the opinion that the evidence adduced is not sufficient to establish that the long-continued and uniform practice at the ports •of the United States where the bulk of this merchandise is imported has been to determine the content by volume, but, even were it established, the question, we think, is still open in this case to a judicial determination of the proper construction of the statute itself. The Government cites the cases of United States v. Basket Importing Co., 13 Ct. Cust. Appls. 98, T. D. 40941, and Furida, Inc. v. United States, 22 C. C. P. A. (Customs) 321, T. D. 47361, in support of its contention that the courts—
and especially this Court, in passing upon close questions of construction in tariff legislation, have stressed the importance of uniformity of decisions and settled and well-understood administrative practice.
That we stressed those matters in the decisions cited and that we have done so in other cases is quite true, but it must be noted that what we have particularly emphasized has been the desirability of uniformity in judicial decisions. However, long-continued administrative practice may aid in the interpretation- of tariff statutes, but it does not preclude the courts from construing or interpreting a statute *274contrary to such, practice when an issue as to its interpretation is presented for the first time.
In numerous decisions of the Supreme Court of the United States, the rule has been laid down- that for administrative practice to be even persuasive in a case, it must be proved to have been known, long continued, uniform and general. The Dollar Savings Bank v. United States, 86 U. S. 227; Merritt v. Cameron, 137 U. S. 542; United States et al. v. Missouri Pacific Railroad Co., 278 U. S. 269.
The Government’s brief cites a long list of cases in which it is asserted that the alcoholic content of merchandise of the kind here involved was determined by volume. The first of these is Wing Yee Chong & Co. v. United States, 11 Ct. Cust. Appls. 329, T. D. 39142. Careful examination of the opinion in that case fails specifically to disclose that it was so determined, although it is the fair inference that such was the case. The merchandise was classified under paragraph 16 of the 1913 tariff act, predecessor of paragraph 24 of the Tariff Act of 1930, supra, and assessed with duty “per pound.” In addition to the duty per pound the collector assessed an additional duty of $2.20 per proof gallon under the provisions of section 600 of the 1918 internal-revenue act, evidently upon the theory that the merchandise contained distilled spirits, although it was classified as stated. In determining the alcoholic content of the merchandise as distilled spirits naturally the volume standard would have been used as required by statute, and it is not unreasonable to suppose that it was used in testing the amount for the purposes of paragraph 24, but this is only an inference. No question as to what standard was proper was there raised.
Another case cited is that of United States v. John Aquino, Inc., 21 C. C. P. A. (Customs) 26, T. D. 46365. There the merchandise was classified under paragraph 804 of the Tariff Act of 1922, predecessor of paragraph 804 of the Tariff Act of 1930, duty being assessed at $1.25 per gallon. Presumably, the volume standard was applied as provided by the statute. We held the merchandise classifiable under paragraph 24 of the act, precedessor of paragraph 24 of the 1930 act, supra, and it is supposed that in the liquidation under our judgment the alcoholic content as determined by that standard was taken for duty purposes. There again no question as to what standard should be used was raised or considered.
Another of our cases referred to is that of United States v. Julius Wile Sons & Co., 22 C. C. P. A. (Customs) 267, T. D. 47327. Our decision there turned on the failure of proof as to commercial designation of the term “sauces,” and the question here involved was never presented or thought of during consideration of the case, although it should be said that the opinion indicates that the volume standard was applied in determining the alcoholic content of the merchandise.
*275A detailed review of the other cases cited is deemed unnecessary. All were cited as bearing upon administrative practice. In none of them was there a judicial determination or consideration of the issue now before us.
Testimony was introduced by both parties bearing upon ordinary commercial and trade practices, independent of any governmental purpose, in ascertaining the alcoholic content of medicinal compounds. Perhaps the weight of such testimony is' to the effect that the volume standard is most generally used, but that the weight standard is sometimes applied is established. Pertinent parts of the United States Pharmacopoeia were also introduced in evidence and are alluded to in the arguments. These also show both weight and volume standards.
We are somewhat impressed with the testimony respecting the accuracy, or rather the uniformity, or lack of uniformity, which results from application of the different standards. That uniformity results from use of the weight standard is stated by all the witnesses testifying upon that point, whether called by the Government or the importer, but the preponderance of the evidence seems to be that it does not result from use of the volume standard. The explanation given by the witnesses is that volume is affected by changes in terns perature while weight is not. That the first of these effects was recognized by Congress is clear because it took temperature into consideration in legislatively defining “Proof spirits” and “Gallon.” Sections 1158 (c) and 1158 (d) USCA, respectively, defined those terms as follows:
(c) Proof spirits. Proof spirits shall be held to be that alcoholic liquor which contains one-half its volume of alcohol of a specific gravity of seven thousand nine hundred and thirty-nine ten-thousandths (.7939) at sixty degrees Fahrenheit. (R. S. § 3249.)
(d) Gallon. In all sales of spirits a gallon shall be held to be a gallon of proof spirit, according to the standard prescribed in the preceding subsection, set forth and declared for the inspection and gauging of spirits throughout the United States. (R. S. § 3250.)
By fixing a definite temperature at which distilled liquors, etc., are tested the statute doubtless assures uniformity as to them, but if they, or the articles falling under paragraph 24, supra, be tested by volume at different temperatures, it seems clear from the testimony that different volumes of alcohol would be found even in the same-article. There is no positive testimony in this record as to whether the merchandise involved was tested under uniform temperatures or not, but certain expressions indicate that such was not the case.
We do not concur in the thought expressed on behalf of appellant, that use of the volume standard leads to absurd or ridiculous results, but we are of the opinion that the weight standard assures greater *276uniformity than tibie volume standard particularly where the latter is applied under varying degrees of temperature.
Our conclusion in the case rests principally upon our view that, since the complete articles covered by paragraph 24, supra, are assessable upon the weight basis and since in some of them the alcoholic content is determinable, under the undisputed testimony given, by weight only, it seems more in harmony with reason and common sense, Congress not having provided to the contrary, that the weight standard should be applied in determining the alcoholic content of all the articles falling within the paragraph.
It is so adjudged. Consequently, the judgment of the United States Customs Court must be and the same is reversed, and the cause is remanded for further proceedings consistent with this opinion.