delivered the opinion of the court:
The appellant imported by the tank steamer Warwick, by official gauge, 1,140,105 gallons of blackstrap molasses, and entered the same at New Orleans, La., December 16,1922. The collector classified the same for duty under paragraph 502 of the tariff act of 1922 as a molasses not imported to be commercially used for the extraction of sugar, or for human consumption, and as testing above 52 per cent of total sugars, at one-sixth of 1 cent plus one-sixth of 1 cent, plus 20 per cent of one-sixth of 1 cent per gallon. The importer protested, which protest was duly overruled and an appeal perfected to the Board of General Appraisers, where, after hearing, judgment was entered for the Government, and against the protestant. Prom that judgment the importer appeals.
The appellant contends: First, that the molasses imported was not sampled by the Government samplers, as provided by the Treasury regulations in force at the time of importation; second, that such regulations are mandatory, and unless complied with the Government test is a nullity; third, that, even if this Government test may be considered not as mandatory, the manner in which it was taken caused incorrrect results; and, fourth, that the importer did make tests in conformity with the Treasury regulations in force at that time,, and that the results of the tests made by the importer are correct and should be taken for duty purposes.
The Government contends: First, that the sampling of the molasses in question was done in conformity with the Treasury regulations in force at that time; second, that, even if such sampling was not done in accordance with the Treasury regulations, such regulations are not mandatory, but directory only; and, third, that the test of sugar content made by the Government is accurate and should be taken lor duty purposes.
Paragraph 502 of the tariff act of 1922 is as follows:
502. Molasses and sugar simps, not specially provided for, testing not above 48 per centum total sugars, twenty-five one-hundredths óf 1 cent per gallon; testing above 48 per centum total sugars, two hundred and seventy-five one-thousandths of 1 cent additional for each per centum of total sugars and fractions of a per centum in proportion; molasses testing not above 52 per centum total sugars not imported to be commercially used for the extraction of sugar, or for human consumption, one-sixth of 1 cent per gallon; testing above 52 and not above 56 per centum total sugars not imported to be commercially used for the extraction of sugar, or for human consumption, one-sixth of 1 cent additional for each per centum of total sugars and fractions of a per centum in proportion.
*434The first question necessarily involved is the nature of the Treasury regulations as to the sampling of blackstrap, molasses in force at the time of this importation.
T. D. 39350 (42 Treas. Dec. 268), a general regulation issued December 9, 1922, by the Secretary of the Treasury to provide for certain contingencies created by the prior passage of the tariff act of 1922, and in force at the time of importation of the molasses in question, is as follows:
The regulations regarding the weighing, taring, sampling, classification, and polarization of imported sugars and molasses covered by articles 943 to 1009, Customs Regulations of 1908, extended by article 532 of the Customs Regulations of 1915, are hereby supplemented by the following regulations relating to the sampling and testing of molasses and sugar sirups not specially provided for, covered by paragraph 502 of the tariff act of 1922. These regulations were prepared with the assistance of the Bureau of Standards, Department of Commerce
Examiners should exercise great care in sampling, and, unless there shall appear some good and sufficient reason therefor, all the samples from an individual shipment shall be divided into not more than two composite samples, each composite representing one-half of the shipment. Whenever practicable, a single composite sample shall be made to represent the shipment. Two duplicate samples from each composite shall be sent to the laboratory for test. Examiners and laboratory chiefs should use every precaution to have molasses sample tests completed before the samples are affected by fermentation or other causes.
Article 532 of the Customs Regulations of 1915 provided:
The regulations governing the weighing, taring, sampling, classification, and polarization of imported sugar, contained in the Customs Regulations of 1908, as amended, are continued until otherwise directed.
The applicable portion of article 958 of the Customs Regulations of 1908 is as follows:
Molasses discharged from tank vessels shall be sampled as it is pumped from the tanks, a sample of uniform quantity being drawn at either regular intervals of approximately fifteen minutes or for every 5,000 gallons discharged
The contention is made by the Government here that T. D. 39350, above quoted, was the only customs regulation in force at the time of importation which controlled the sampling of the molasses in question. We think the provisions of article 958, above quoted, were in full force and effect at the time of the importation here and that a reasonable construction of the various regulations above quoted leads to the conclusion that T. D. 39350 was intended to and does apply to that portion of paragraph 502, tariff act of 1922, relating to molasses and sugar sirups not specially provided for, and that the portion of article 958 above quoted was the applicable regulation for the sampling of the importation in this case. In other words, at the time the molasses in question was imported and sampled the Treasury regulations required that the molasses should be sampled as pumped from the tanks, and that in so doing a sample *435of uniform quality should be drawn at either regular intervals of approximately 15 minutes or for every 5,000 gallons discharged.
Practically the whole controversy here arises from the work of Government Sampler Joebert, who had charge of the sampling from the time pumping began until 7 o’clock on the following morning.
The pipe through which the pumping was done led from the ship to the importer’s tanks on shore and through a small shelter house on the dock. Within this house a cock had been placed in this pipe, and from this cock or spigot all samples were drawn by both Government and importer samplers.
The testimony as to the method followed by Sampler Joebert, during the time he was assigned for sampling is substantially as follows:
EL C. Huber, factory superintendent for the importer, testified that he was present when the sampling was done, from the time pumping was started until about 9 o’clock on the morning of December 13, and that the pumping started on the evening of December 11; that Government Sampler Joebert took his first sample about an hour and a half after pumping was started, consisting of about a gallon to 2 gallons; that Joebert took only one sample, and that no other Government sample was taken until 10^2 hours afterward. John Johnson, an employee of the importer, testified that he began sampling at the time pumping began, and took samples every half an hour thereafter until 7 o’clock the next morning; that he saw a Government sampler take a sample about an hour or an hour and a half after pumping began, which he took by putting his can under- . neath the open spigot and filling it about three-quarters full, or a quantity of about a gallon and three-quarters, or 2 gallons, and that he did not see the Government sampler any further before the witness left at 7 o’clock the next morning; that he knew Sampler Joebert, and the witness Huber was with witness during the time he was on duty.
John Burst, roundsman and night watchman for the importer, testified that he went on duty at 5 o’clock on the afternoon of December 11, and was present when the 'Warwick began to pump; that tne Government sampler arrived in an automobile at about a quarter past 10 and that said sampler then went with the witness Huber to the levee with his molasses can, and left thereafter at about 10 minutes past 11 p. m.; that the said sampler came back and left about 10 minutes past 12 a. m. and did not return; that the witness remained on duty until 6 o’clock on the morning of December 12, and did not see any other Government sampler after 12.10 a. m.
The Government sampler, Robert J. Joebert, testified that he began to sample the cargo of the Warwick at 6.30 p. m. December 12, and that he remained on duty until 7 a. m. the next morning *436and was then relieved by Sampler Paul. He had no recollection of how many samples he took, but related, in answer to various inquiries, what his usual method of procedure was in such cases. lie stated that he took samples throughout the night which ho regularly marked, but was uncertain how he marked them. He was uncertain in all particulars about this particular importation and afterwards admitted that his records were erroneous. In fact, his testimony was so uncertain as to call for the following observation from General Appraiser McClelland:
This witness is uncertain in every way. He don’t know whether the ship arrived on the 11th or the 12th, who was there, or the importer or anything.
His testimony is entirely unsatisfactory, uncertain, and of little probative value. Government Sampler Paul, who was to relieve Sampler Joebert, arrived at work at about 7 o’clock on the morning of December'12. He found a can of a capacity of about a gallon and a quarter to a gallon and a half, about one-quarter full of molasses, which presumably had been left by Sampler Joebert. Joebert was not on the work when he arrived, and Paul did not see him on that occasion, and does not know whether the molasses contained in the can was molasses which had been drawn by Joebert in accordance with the Treasury regulations. The proof further shows that from 7 o’clock on the morning of December 12, until the entire discharge of the cargo on the morning of December 14 the sampling was done by Government samplers Paul, Demarest, blacker, and Spriggins in a regular way and substantially in accordance with the Treasury regulations. Some attempt is made to impeach this by the importer by an attempt to prove that there was an approximate interval of an hour and a half between the time each Government sampler ceased work and his successor began. This, however, is controverted, and the importer, in our opinion, has failed to establish, by a preponderance of the evidence, the claim it makes in this respect.
The proof further shows that the samples taken by Sampler Paul were commingled with those taken by Sampler Joebert, and a test made of these samples as commingled, the said samples representing a total of 435,000 gallons of molasses.'
From a consideration of this testimony it is impossible to avoid the conclusion that the weight of the testimony shows that Government Sampler Joebert did not take his samples in accordance with either the letter or the spirit of the Treasury regulations pertaining thereto.
On the other hand, it appears from the testimony, not contradicted, (hat the importer took samples of the molasses every half hour from the time pumping began until pumping was concluded, and that these samples thus taken were preserved and afterwards analyzed and the sugar content computed by the chemists of the importer, substantially in accordance with the customs regulations in force at that time.
*437The reports of the appraiser and collector in this case recite that the sampling was done in accordance with the "Government rules and regulations.” While the ordinary presumption of correctness attaches to this return, it is nevertheless subj ect to impeachment by ■evidence of its incorrectness. In T. D. 28249 (G. A. 6620) General Appraiser Somerville held in a similar case that where imported merchandise is dutiable by weight, the action of Government weighers would not be subject to review by the Board of General Appraisers so long as the weighers acted strictly within their statutory duty and proceeded on no wrong principle in performing the duty imposed on them by law, but held further that this principle does not apply where such irregularities occur in the weighing of goods as probably would lead to erroneous results, and in such case the board would undertake to correct the weights returned by the Government weighers in accordance with the facts as shown by the evidence. The decision in that case is well considered and we think states the law properly as applicable here.
We come now to a consideration of the effect of the aforesaid failure of Government Sampler Joebert to conform to the Treasury regulations. The courts have recognized two general groups, or classes, of laws and regulations made thereunder. The distinction between these classes is recognized in United States v. Dominici (78 Fed. 334). There the importation was shooks of American manufacture, reimported under paragraph 493 of the act of October 1, 1890, providing for the free entry, under general regulations to be prescribed by the Secretary of the Treasury, of American products. The court held such regulations, reasonably made,- to be mandatory, but distinguished them from regulations issued under the general power of the Secretary of the Treasüry to regulate the administrative details of customhouse business. In French v. Edwards (80 U. S. (13 Wall.) 506), the Supreme Court, speaking through Mr. Justice Field, said:
There are undoubtedly many statutory requisitions intended for the guide of officers in the conduct of business devolved upon them which do not limit their power or render its exercise in disregard of the requisitions ineffectual. Such generally are regulations designed to secure order, system, and dispatch in proceedings, and by a disregard of which the rights of parties interested can not be injuriously affected. Provisions of this character are not usually regarded as mandatory unless accompanied by negative words importing that the acts required shall not be done in any other manner or time than that designated. But when the requisitions prescribed are intended for the protection of the citizen, and to prevent a sacrifice of his property, and by a disregard of which his rights might be and generally would be injuriously affected, they are not ■directory but mandatory. They must be followed or the acts done will be invalid. The power of the officer in all such cases is limited by the manner and conditions prescribed for its exercise.
While the ease just cited applies to statutory enactments, it is equally applicable here. The law is well settled that the reasonable *438regulations of the Secretary of the Treasury, when he is authorized by statute to establish the same, are entitled to and have the force of law. — Gump v. United States (3 Ct. Cust. Appls. 137 [139]; T. D. 32384).
The current of authority also seems to tend toward the conclusion that a failure by responsible Government officials to comply with such regulations, in making tests for duty purposes, when properly shown, will invalidate the tests made by them. Origét v.- Hedden (155 U. S. 228) was a case where seven packages of imported goods were ordered by the collector sent to the public store and only one was examined. The importer claimed that the statute required all such packages to be examined and that a failure to do so invalidated the test. The court held that the collector, although ordering the seven packages to store, had not ordered the examination of all of them, but said:
* * * we are not inclined to deny that it might happen where the collector had given specific direction for the examination of more than 1 package out of 10, and the importer had relied on the direction,' the omission to examine the number of packages directed might under some circumstances be availed of by him as constituting a want of the examination to which he was entitled.
In 1905 certain raw sugars were imported at New Orleans and the importer protested against the legality of the methods used by the customs authorities in sampling the same. It appeared that the cans used by the samplers were much too small, with loose covers and unfit for the purpose; that the covers of the cans were left open and the sugar samples left on the wharf exposed to the sun, which dried them out; that an insufficient number of samples were taken by the Government samplers, not fairly representative of the various parts of the shipment, and of only about 7 per cent to 10 per cent of the total number of bags of sugar. It further appeared that the importers had made correct tests based on the samples of at least 50 per cent of the bags. The board sustained the protest and ordered reliquidation on the basis of importers’ tests. — T. D. 26628 (10 Treas. Dec. 102).
A leading case on the subject is American Sugar Refining Co. v United States (1 Ct. Cust. Appls. 228; T. D. 31273). There sugar was imported and Treasury regulations were in force providing for the taking of suitable samples and for a polariscope test of the same; detailed instructions were contained in these regulations as to the method to be used in making such test. Tests were made as provided by these regulations, but not being satisfactory to the Government officials, another test, called the settlement test, was used, whereupon the importer protested. The court, speaking through Barber, Judge, said in part:
We think upon a careful examination of these regulations that it must be held that they are general as applying to all sugars and general as applying to all *439importers, and that they convey fair notice to the importers that all sugars subject to duty will be classified as a result of the tests made by the experts in the collector’s office'appointed for the purpose, and that no. tests for classification will be accepted by the department other than such official tests. * * *
The contention of the Government that its prescribed tests are more accurate than the settlement tests, which was upheld in the other branch of this case, and the fact that so much care has been taken in prescribing these regulations for the testing of sugars, to the end that accurate results shall be obtained, all lead to the conclusion that in this case the official tests more nearly represent the true quality of the sugars for classification purposes than the settlement tests; but whether this be so or not we see no reason for holding that, at their option alone, the officers connected with the collection of the customs duties shall be permitted to go contrary to the general regulations they have themselves prescribed. To hold otherwise, we think, would have a tendency to produce great confusion and uncertainty in the classification of sugars, because, if not limited to its own tests, the Government may accept or compel the acceptance of any private tests, settlement or otherwise. If it is found that the prescribed regulations are not adequate to protect either the Government or .the importers, it would seem that the same power still resides in the Secretary of the Treasury to make general regulations for the protection of both that was vested in him when the regulations in question were promulgated.
Reliquidation was ordered on the basis of the official polariscope test.
A somewhat similar case was American Sugar Refining Co. v. United States (3 Ct. Cust. Appls. 69; T. D. 32352). There the question was as to the tare to be allowed on certain sugar bags. The Treasury regulations in force provided that actual tare should be allowed. The collector estimated the tare. The court said:
It would seem clear that these regulations, taken as a whole, contemplate that actual tare shall be ascertained, and prescribe the method of ascertaining what it should be. The only regulations brought to our attention made by the Secretary of the Treasury which point out the method of ascertaining the real tare are those which we have quoted. When these methods have been applied, it would seem to be opposed to the very terms of the statute to permit a mere guess or estimate as to what tare should be allowed to be substituted for the result.
Reliquidation was ordered on the basis of the actual tare as found by the weigher.
Applying these principles to the case at bar, it is obvious that the sampling done by Sampler Joebert was invalid and illegal and can have no approval as an official act. The Government can not be permitted to make regulations mandatory upon the importer, and on its part disregard them.
But one inquiry remains, namely, what is the most accurate sampling of the molasses imported here, as the same appears from the record, and what will most nearly approach the results which would have followed if the Treasury regulations had been complied with? The Government is interested only in one thing — the collection of its duties according to the letter of the statute. The *440importer has no rights here except to demand that he pay no more than the statutory duty on the material he imported.
We hare seen that the sampling done by Sampler Joebert was not done in conformity with the Treasury regulations. Furthermore, it was not done in such a way as to be fairly representative of the molasses pumped during his watch. His samples having been commingled with those of Sampler Paid, and one test being made of the whole, invalidated this test of 435,000 gallons. Government tests were made separately of the samples taken by samplers Demarest, blacker, and Spriggins, which were, respectively, on retest, 55.76 on 362,500 gallons, 52.46 on 410,000 gallons, and 51.59 on 53,355 gallons. During the whole of the work the importer took samples each half hour which were tested by the Treasury polariscope test. While this test was not strictly in conformity with the Treasury regulations, in that a sample was not taken each fifteen minutes, or of each 5,000 gallons, and while the polariscope used by the importer was an instrument inferior to that used by the Government analysts, we are nevertheless constrained .to believe that such “test will more nearly represent the true sugar content of the molasses in question during the time that samplers Joebert and Paul sampled the first 435,000 gallons. The importer’s test, as shown by the evidence, was 51.36. As to the balance of the importation, the Government tests should be taken.
We therefore conclude that the judgment of the Board of General Appraisers should be, and is hereby, reversed and remanded, with directions to reliquidate the entry on the basis of the importer’s polari-scope test of 51.36 on the molasses reported as sampled by Joebert and Paul, 55.76 on that sampled by Demarest, 52.46 on that sampled by Hacker, and 51.59 on that sampled by Spriggins.
The quantities reported by the several samplers do not aggregate the same amount as that reported by the official gauger. • They •should be used as the comparative basis of the computation. Reversed and remanded.