The appellants herein imported in 1898 certain sugars, on which duty was assessed under the provisions of paragraph 209 of the tariff act of 1897 (Act July 24, 1897, 30 Stat. 168, c. 11 [U. S. Comp. St. 1901, p. 1647]). The pertinent portion of said paragraph is as follows:
“209. Sugars not above number sixteen Dutch standard in color, tank bottoms, sirups of cane juice, melada, concentrated melada, concrete and concentrated molasses, testing by the polariscope not above seventy five degrees, ninety-five one-hundredths of one cent per pound, and for every additional degree shown by the polariscopic test, thirty-five one-thousandths of one cent per pound additional, and fractions of a degree in proportion. * * *”
The importers protested, the board of general appraisers overruled their protests, and the importers have appealed to this court.
A considerable portion of the record and the greater portion of the briefs are devoted to a discussion of expert scientific testimony concerning the polariscope and the comparative accuracy of various tests made therewith. The polariscope referred to in said act is an instrument so adjusted that, when a ray of polarized light passes through a tube filled with a certain solution of sugar, the scale indicates the percentage of pure sugar. At the time of the passage of the act in question the polariscopic test of the value of sugar had been in commercial use for some 20 years. The tariff act of 1883 provided for *328its use by the government in ascertaining duty, and in the Treasury Regulations then promulgated the methods previously in commercial use were specifically adopted. It was referred to in the tariff act of 1894, and down to the passage of the act in question the results obtained by 'this method were the only ones used in connection with the sale or importation of sugar. But under the sugar bounty act of Oct. 1, 1890 (Act Oct. 1, 1890, 26 Stat. 583, c. 1244, § 231), which provided for a bounty on domestic sugar under rules and regulations to be approved by the Treasury Department, a new method of applying the polariscopic test was adopted, and this method continued to be used during the two years that said act remained in force. Upon the passage of the act here in question, the Treasury Department promulgated regulations for the use of the polariscopic test, which, for the purposes of this discussion, may be assumed to be the same as those applied under the act of 1890. The result of these later modified tests is to oblige the importer to pay higher duties than those enacted under the earlier tests. Against the payment of these higher duties the importers have protested, and the question of the validity of said regulations is the one presented herein.
The distinction between the earlier and later method may be stated with sufficient accuracy for the purpose of this discussion as follows: Under the old, or as it has been termed the “commercial,” system, the actual readings of the scale on the eyepiece of the polariscope were taken as showing the actual value of the sugar for sale or duty purposes; that is, the test was one made by reading by the eye. ' In the case of disagreement in the tests between the parties in interest, a settlement was reached by a third or compromise test. Under the later Treasury method, the test is not absolutely determined by the eye, but the regulations provide that the reading must be corrected by certain arbitrary additions as prescribed in a table prepared by the officers of the government, and that duty must be paid upon such increased value as shown by such corrections or additions.
Counsel for the importers contend that the term “testing by the polariscope” had a well-settled commercial meaning at the time of the passage of the act of 1897, that this innovation upon the previous practice is not justified, and that the use of these regulations prior to the passage of the act of 1897, under the bounty act of 1890, for two years, for the benefit of domestic producers only, and which was discontinued in 1894, does not suffice to give a different commercial meaning to said term. They show that no such method is practiced anywhere else in the civilized world, except in British Guiana, and that as a result of these so-called corrections or additions nearly one-third of the pure refined sugars brought to the port of New York show the absurd and impossible result under said test of being more than chemically pure. The testimony of one of the witnesses for the government (Ernest W. Chapman, government polariscopis’t) on this point is as follows:
“Q. Since the regulations of October 17, 1897, went into effect, what results have you found in respect to the testing of refined sugars? A. We found that the tests run 99.7 to 100.2 or 100.3. Q. How many cases have you found, in testing foreign sugars where the tests run over 100? A. About 30 per cénit 100 or over. Q. Does the record show that they tested over 100? *329A. No, sir. Q. What does the record show ? A. 100. Q. Why do you put on record that the test was 100? A. I didn’t think it would look well to report them. Q. What is your experience with regard to testing the same character of sugars under the previous regulations? A. Seldom we had a sugar tested over 100. Q. Did you ever have any that tested over 100? A. Not that I recall. Q. What was the highest average? A. About 99.8, possibly.”
In this connection it may be noted that the Treasury Regulations of 1897 provide as follows:
“A variation of .02 degrees from the established values may be allowed for errors of observation, temperature, etc.”
Counsel for the United States defend these regulations chiefly on the ground that there is a variation in the reading of the polariscope, according to. variations in temperature at the place where the sugar is tested, and that the corrections and additions provided for by the regulations merely consist in an addition of .3 per cent, for each 10 degrees Centigrade of temperature above that at which the polariscope is standardized, and that in this way the actual amount of pure sucrose in each sample is more accurately determined than was the case under the old eye test. Despite the vehement protests and expert testimony on the part of plaintiffs to the contrary, I think this contention of the government is sustained by the preponderance of proof. It is not proved, however, in the cases where pure sugar tests over 100 per cent., in which case, as has been shown, duty is assessed only upon 100 per cent.
While the corrections and additions do not necessarily furnish an accurate test in any case, an eye test at 17.5 degrees Centigrade does furnish an accurate test. On this point counsel for the United States says:
“We are all agreed that if the polariscope be standardized at 17.5 degrees O., and the solution be made up and the polarization [test] be made in a room at the same temperature, the reading of the polariscope (having first properly regulated it) will give us the true polariscopic test (saccharine value) of the sugar.”
The position of the government herein is shown by the following quotation from the Treasury Regulations:
“79. The expression ‘testing * * * degrees by the polariscope,’ occurring in the act, is construed to mean the percentage of pure sucrose contained in the sugar as ascertained by polarmetric estimation.”
The question presented is whether the Treasury Department was justified in so interpreting the statute of 1897 as to provide that “testing by the polariscope” should mean a correction of such tests by figures, which, under varying temperatures, would show an inaccurate, but more nearly accurate, percentage of the amount of sucrose in a sample than was shown by the tests which, except under the bounty act, had universally prevailed in this country for upward of 20 years. I think the end sought does not justify the means used. The well-settled meaning which “testing with the polariscope” had at the time of the passage of the act could not thus be violently changed by mathematical “corrections and additions,” without evidence that Congress intended such a change. Especially must this be so when, *330as in this case, the use of the table not only does not necessarily produce actually accurate results, but leads to the statement of chemical impossibilities.
The point has not been made on the argument, and it is perhaps irrelevant, but no reason is perceived why, if the government desires to secure uniformity and accuracy, it should not take its tests, or provide that they should be taken, at the temperature at which its instruments are standardized.
The decision of the board of general appraisers is reversed.