delivered the opinion of the court:
Paragraphs 16 and 17 of the emergency tariff act of May 27, 1921, read as follows:
16. Cotton having a staple of one and three-eighths inches or more in length, 7 cents per pound.
17. Manufactures of which cotton of the kind provided for in paragraph 16 is the component material of chief value, 7 cents per pound, in addition to the rates of duty imposed thereon by existing law. .
The merchandise involved in this suit consists of certain cotton cloth represented by several different protests and several entries. The merchandise was assessed for duty under the appropriate paragraph of the tariff act of 1913 as to which there is no issue, and in addition thereto a duty of 7 cents per pound, under paragraph 17, of the emergency tariff act, was levied. Appellants claim that the merchandise is not of the kind provided for in paragraph 17, since the cotton from which the cloth was made did not have a staple of 1% inches or more in length. The appeal to this court is taken from a judgment of the Board of General Appraisers overruling appellants’ protest.
Upon the question as to whether the cloth was made of cotton such as is provided for in paragraph 16, there were .two witnesses for the importers and three witnesses for the Government.
Hugh Byron Gordon, for the importers, testified that he was a chemist with the United States Testing Co.; that he had been with them for four and one-half years; that he, in his business, tested cotton cloth for the purpose of ascertaining the staple of cotton contained therein; that he was a graduate of Miami University, spent four years post graduate work in the University of Illinois, and received a degree of doctor of philosophy in chemistry; that he had considerable experience as a teacher in collegiate work in chemistry including the work in teaching the chemistry of textiles and dyeing; that his studies were reduced to writing in the form of a paper published in September or October (probably previous year, 1923); that his article on this question was published in the Textile World, a monthly magazine published in the interest of the textile industry; that it has a large circulation in the United States. Extracts from his testimony are as follows:
Q. Go on and tell the remainder of your experience, Doctor.- — A. Now, in this study, as I said, I had tried to find out what other people do in order to tell the length of staple of raw cotton from which the manufactured fabrics are made, and I inquired from a number of cotton men and other cotton experts, and they disagreed as to what the change in length of staple was, if any, in the manu*555facture of cotton. Some said it did not change, some said it was shortened, the average of fiber in the sample was shortened by manufacturing. Others said it was actually lengthened. Obviously when cotton experts disagree it is necessary to go into the subject and find out. So I wrote to three different mills and asked them to send me samples of raw cotton which they used in the mill which were of different staple lengths. I asked I think for 1)4 inch, 1^- and 1% cotton, and I asked them also to send me in addition to the raw cotton, yarn made from that same type of cotton, and cloth manufactured from that same yarn and cotton.
Q. That is in the course of your experience?- — -A. These were the samples I wanted to study on. From all three mills I obtained samples of raw cotton and yarn and cloth. I did not get any 1%-mch staple, but I got 1^-inch staple from all three. In order to make sure that the mills were giving the correct estimate of the length of the cotton, the raw cotton, I took these samples of raw cotton to the appraiser’s office, of the United States Government, and the gentleman there was kind enough to assist me in determining this, and several men there measured the staple length of that raw cotton for me. I also measured it myself. We did not always agree exactly with the statement given by the factory, but we agreed, I should say, as closely as it is reasonable to expect that two staplers will agree.
In addition to these samples from three mills I obtained samples of different lengths of cotton, raw cotton from the Department of Agriculture, which were looked upon as standards by the Department of Agriculture, and got the staple length of these not only from the Department of Agriculture, but I measured them myself also and agreed very well with the Department of Agriculture. Then I measured the fiber length of these raw cotton samples, and also the samples of yarn and fabric manufactured from the raw cotton at the different mills. Of course with the samples from the Department of Agriculture I had no corresponding material.
Q. What was the result of your experiments along that line? — A. Briefly that the average length of fiber in the manufactured goods, taking precautions to disregard too short fibers, fibers which were too short to come under the staple measurement, and disregarding also any extremely long fibers which might be in the sample, then the average length of the fibers is essentially the same as the staple length of the raw cotton in manufactured goods, talcing the same precautions in regard to the measurement of the fibers myself, and disregarding the exceedingly short ones, which in many eases are below the average length of the fibers, I found that they were the same, ‘practically the same, the only difference being that the fibers from the manufactured product were perhaps two or three hundredths of an inch shorter than the raw cotton. The difference was so small that it is generally lost sight of in a staple measurement for raw cotton. (Italics ours.)
Q. Did you make at the request of Morse Brothers an analysis of several samples sent to you by them? — A. Yes.
Q. And making the analysis, did you take into consideration the experiments you had made which you have just described, and all of your information gained professionally? — A. Yes.
Q. Did you carefully and correctly analyze the samples? — -A. Yes.
Q. Did you make a certificate of your analysis or test to Morse Brothers?— A. Yes.
Q. Are these the samples of cloth which you analyzed? — A. Yes, those are the ones.
He then identified the exhibits and in each instance stated positively that they were made from cotton, the staple length of which *556was under 1% inches. His answers to questions regarding one of the exhibits are as follows:
Q. Have you got a memorandum or record made by you to refresh your recollection? — A. I have.
Q. You may consult it. — A. 695, the warp I found average fiber length one and thirty-hundredths inches, which indicates a staple length of raw cotton from which it was manufactured of one and five-sixteenths inches.
In explaining how he determined the staple of the cotton from the samples of cloth examined he testified as follows:
Q. Will you tell us exactly what you did in order to determine the length of the staple? — A. Why, yes. I took these samples and removed any sizing material from them by soaking them in boiling soapy watei for a considerable while, and then rinsing them out and drying them. I then withdrew certain pieces of yarn from each sample, and untwisted the yarns so that I could pull out the fibers without any danger of breaking them, and pulled out wisps of these fibers in that way, and spread them out on a black surface, and with a rule measured the length of each of the fibers. Now, in doing this, as I mentioned before in describing my previous study, I did not consider short fibers, say anything under one and one-tenth inch. I did not consider that at all, and also if a fiber was over one and seven-tenths inches I did not consider it. I think I found three or four in the three hundred fibers measured; I think it was three or four over one and seven-tenths inches. * * *
Q. Now, in calculating the length of the staples as you testified jrou did, did you make any allowance for the shrinkage or breakage in the course of manufacture? — A. The staple, you understand, is the average of the raw cotton, average usable raw cotton. I estimated that from the measurements made from the fibers. As I said I measured the fiber lengths here, and the average was as stated, and then I found making an allowance for any change in the manufacture this length, average length of fiber indicated a raw staple, a staple of raw cotton of the length which I stated on direct examination. You will recall in one case the average fiber length was one and thirty-one hundredths inches, which 1 stated indicates is made from raw cotton of a fiber staple, of a staple length of one and five-sixteenths inches. Now, one and five-sixteenths inches is a good deal longer than one and thirty-hundredths, so I made an allowance in the interpretation of my result.
Q. The allowance you made, was that a matter of guesswork or approximation or actual findings made by you as a result of experiments? — A. The results of experiments, the result of this study which, as I say, was published last fall in which I had examined quite a number of manufactured goods.
Q. In your experiments, did you make any concession or make any allowance as to a coarse yarn and finer yarn? — A. Yes; you mean in my study of the variation of lengths?
Q. Yes. — A. Yes; I made that allowance.
Q. And what did you find? — A. I found that, now I may get the limits of fineness a little wrong, but I think the finest yarn I examined was from a sample of cotton, 122; I meant that the finest was 122 and that the coarsest I examined was 60, and the results were so close together in the two cases that any change, any difference in length due to the difference in the fineness of the yarn spun was below the difference in length that people will normally consider the same staple length. Of course, you are aware that there is some variation in lots as coarse as one and five-sixteenths or one and five-eighths.
*557J. V. Sutcliffe, witness for importers, after stating bis qualifications, testified, tbat be examined the exhibits involved in this case and found all the exhibits to be of a staple under 1% inches; and that, in determining the length of the staple from which the goods were made, he made allowance in his measurement of the fibers in the cloth, based upon his understanding of the manufacture of cotton.
For the Government, Harold Fielding testified that he was a textile analyst in the customs service and had been so employed for about four years, and that he analyzed textile fibers in the Fall River Textile School, where he was a student through an arrangement with the Federal Board; that under the directions of the analyst in charge, Mr. Gassmann, he determined the staple in the exhibits in controversy and found the staple to be over 1% inches. He stated that he pursued no recognized method, but followed out the instructions of the analyst in charge; that he did not know whether the method was the commercial method or not; that it was not the method ¿hat he had previously used at Fall River; that he never stapled raw cotton; that it was not his intention in making his analysis to state what the staple was in the raw cotton from which the textiles were manufactured; that it was merely to determine the length of the fibers in the sample he had; that in determining the staple of the fiber of the manufactured goods he ignored the short fibers and considered only the very long fibers, and the medium sized fibers; that it was practically impossible to count fibers, and that he did not count them.
Henry A. Gassmann, witness for the Government, testified that he was analyst in charge for the Government and had been for twenty and one-half years; that he supervised the work of the analyst Sutcliffe, in analyzing the exhibits in controversy, and that he, Gassmann, was consulted in cases requiring his decision; that he did nothing with the analysis himself, except to verify the report when Sutcliffe showed it to him; that the short fibers were rejected in the estimate of staple length for the reason that they may have been the broken ends of long fibers; and that “we only count those that are long and disregard the short fibers”. * * * “We are only taking the longest possible fibers and conclude from that what the original staple was”; that he only regards what he finds in the cloth; that the operation pursued is the only method for analytical purposes to determine the length of the fiber, and that this method was the commercial method.
Francis Cottave, witness for the Government, testified that he was a textile analyst and had been such for fourteen years, in the service of the Government; that the cloth exhibits were made from .cotton over 1 % inches in length; that 25 or 30 per centum of the staples *558in the yarn that he analyzed were 1% inches or over, and that 75 per centum were less than 1% inches.
It will be seen from a perusal of the testimony that the staple length of unmanufactured cotton could be told with reasonable accuracy by applying methods, the correctness of which methods are not seriously contested. It will also be observed that when it comes to stapling in print, or telling from the manufactured cloth what the staple of the raw cotton was that went into the cloth, it is quite another matter, and that there is no definite, fixed rule governing his kind of test or analysis.
We have examined all of the authorities on the subject that are .obtainable and we find very little on the question. The Government's publications on the subject of stapling cotton only apply to 'the stapling of raw cotton. The Textile World of the issue of October 13, 1923, published an article by H. B. Gordon (presumably the witness, Hugh Byron Gordon), upon the subject of cotton fiber length and stapling length, which is the only published authority we have found on this question. This authority is in accord entirely with the testimony of the witness, Gordon, as it appears in the record. Figures, tables, diagrams, and detailed explanations of methods in the article tend to support the conclusions arrived at in the testimony of the witness.
It is apparent that the Board of General Appraisers were laboring under a misapprehension of the intention of Congress in the enactment of paragraph 17. Its opinion indicates that, even if the long fibers preponderated in the cotton before manufacture, it would have to be shown that the long fibers constituted the component material of chief value of the article. We can not agree with this view of •the case.
From an examination of the authorities it is clear that there are commercial methods of determining the staple of raw cotton, which are recognized as fairly accurate and sufficient for commercial purposes, without the necessity of an actual count of the long and short fibers in order to determine which preponderate. The Commercial Classification of American Cotton, Department of Agriculture, Circular No. 276: Standards for Cotton'Classification in the United States and Abroad, Agricultural Economics, United States Department of Agriculture, Nos. 92 and 82: Bureau of Markets, United States Department of Agriculture, No. 41, p. 12; Grace v. United States, T. D. 40377, G. A. 8846.
By pulling and manipulating the cotton a representative sample" is obtained, which, by applying the rule, while under certain conditions of manipulation, will show the staple length, but it is conceded' that in contested matters where results as near perfect as is ^possible are desired, an actual count of the threads of the pull is *559necessary. Circular No. 41, supra. If fibers 1% inch.es and over in length preponderate over the fibers less than 1 % inches, then it can be said that the cotton has a staple of 1 % inches and over.
By paragraph 16 Congress meant to levy duty of 7 cents a pound on ail raw cotton having a staple of 1 % inches or more in length.
In the enactment of paragraph 17 Congress intended that if the component material of chief value of the importation was of cotton, and that if that cotton, before it was made into the cloth, had a staple length of 1 % inches or more, then the cloth should bear the additional duty of 7 cents per pound. Wilkes-Barre Lace Mfg. Co. v. United States, 11 Ct. Cust. Appls. 519.
If we understand the position of the Board of General Appraisers, they would have paragraph 17 read as if it said that the component material of chief value must be cotton, all of the fibers of which measured 1% inches or more, or that fibers 1% inches or more must constitute chief value of the merchandise. This is an erroneous interpretation of the intent of Congress. The Congress, no doubt, anticipated some importations which might consist of wool, or silk, or rubber, together with cotton, and if cotton was of chief value and fell within the provisions of paragraph 16 (fibers 1% inches or more in length preponderating), it should be subject to the additional duty. The misinterpretation of the meaning of paragraph 17 by the board, no doubt, influenced it in its finding that the presumption of correctness attaching to the collector’s action had not been overcome. The following statement from the board’s opinion will be noted:
Therefore, even if the testimony did establish the affirmative of what the importers tried to prove, namely, that there was a greater number of short fibers than of the long fibers used in the manufacture of the cloth under consideration, which, however, was not proven, it would still be necessary to show that the quantity of cotton staple of 1% inches or more in length therein did not in fact constitute the component material of chief value in the article. On this point no testimony whatever was introduced or even offered by the importers. (Italics ours.)
It, therefore, may be contended that the board’s decision rests upon the fact that they found from the evidence that the importers had not proved that the short fibers, in the raw cotton, preponderated over the long fibers. In view of the finding of the collector that the merchandise was composed in chief value of the kind of cotton specified in paragraph 16, we think it was incumbent upon the importers to show, since the goods were wholly of cotton, that the fibers in the raw cotton 1% inches and over did not preponderate over the shorter fibers.
We think the fair weight of the evidence is with the importers on this question. Their testimony not only shows that the short fibers found in the cloth preponderate, but that the short fibers in the cotton from which the cloth was manufactured preponderated.
*560The witness Gordon testified that he had taken samples of cotton of a known staple and measured their length and determined the staple of the cotton, and also had taken cloth made from that kind of cotton and had determined from these experiments that, after the fibers had been manufactured, they were practically the same in length as they were before manufacture, the only difference being that the fibers from the manufactured products were perhaps two or three hundredths of an inch shorter than from the raw cotton, and that the difference was so small that it was generally lost sight of in a staple measurement of raw cotton; that he made allowances for shrinkage, and, from his experiments, stated positively the raw cotton from which the cloth was manufactured had a staple of less than 1% inches. This testimony is nowhere rebutted or refuted in the record.
It is suggested that in the process of manufacture the fibers may shrink in length and that, even if they measured less than 1 % inches in the cloth, this would not indicate that they were from raw cotton of fibers shorter than 1% inches. Many elements could enter into the consideration as to whether or not the fibers in the cloth were shorter than they were before they were manufactured. Recognized authorities agree that the process of manufacturing cotton yarn breaks many fibers. Methods of Determining Length of Cotton Staples and Illustrations of Their Application, by H. A. Cobb, Ph. D. The Government’s witnesses disregarded all the short fibers in their estimate. (No actual count was made.)
It is clear then that they discarded many of the short fibers which were in the original cotton and which were not broken by manufacture. This alone imports inaccuracy.
Some authorities, in years past, contended that carding and combing of the cotton in order to prepare it for yarn making removed the short fibers. This is disputed by the best known authority on this subject, EL A. Cobb, Ph. D. In his publication Methods of Determining Length of Cotton Staples and Illustrations of Their Application, at page 62 he says:
For a long time it has been believed and taught that the operations of the cotton mill tend to remove the shorter fibers, and that it is mainly the longer ones that are spun. I do not find that there has ever been any definite proof that this is true. In fact, my investigations, so far as they have been carried, show the contrary, namely, that under good management and in ordinary circumstances there is comparatively little variation in the composition of the cotton coming from the various machines; that is to say, the stock everywhere in the mill, up to the spinning frame, presents about the same variation in the length of its component fibers.
It is urged by some that, if the fiber is stretched while wet, it may be extended in length. We find no definite proof of this fact, but, even if so, it maybe due to the fact that the spirality of the fiber is tiras smoothed out, not necessarily increasing the length of the fiber, *561but increasing its measuring qualities. From a consideration of all tliese facts; that is, that the manufacturing may break the fibers, thus increasing the number of short fibers in the manufactured “pull,” that manipulation may stretch the fibers, that washing in hot water may shrink the fibers, or have the opposite effect, and from many other considerations it is at once apparent that the fibers in the manufactured cloth might be longer or shorter than they would be in the cotton bale. The true situation can be told, with reasonable accuracy, only by those who have experimented exhaustively in trying to ascertain the difference in length, if any existed.
The only witness in this case and the only authority we find on the question tell us, after long and faithful experimentation, that there is very little difference in the length of the fiber before manufacture and after manufacture. No other witness and no other authority discovered, disputes this contention.
The Government’s witnesses made no attempt to ascertain if there was a difference in the length of the staple before manufacture and after manufacture. In estimating the staple lengths they took into consideration only the very long fibers in the cloth. They rejected all short fibers and also stated that they were only concerned with determining the length of the fiber in the cloth. Obviously this method was not only incorrect but very unfair. They admitted that they did not count the fibers, and it is apparent that where they speak of their “count” they mean “estimate.”
The second witness for the Government had nothing to do with the estimate or with the analysis, but directed the first witness how to do it. The testimony of the third witness is decidedly in favor of the importers, since he stated positively that 75 per centum of the fibers in the cloth were less than 1% inches in length. While this does not prove conclusively that 75 per centum of the fibers in the raw cotton were less than 1% inches in length, it certainly can not be said to support the position of the Government.
While there is no recognized method of determining the staple of cotton after it is manufactured into cloth, Congress must have contemplated that some means could be devised for determining this fact. The testimony of the witnesses for the importers discloses a method used by them, which seems to be a reasonably fair one, and, while not capable of producing absolute accuracy, it would seem to be reasonably calculated to produce the result desired by Congress The method pursued by the Government can not be regarded as a fair effort to determine the staple in the cotton before it is manufactured, much less is it to be relied upon in determining staple from the cloth.
*562■The board stated that the importers had not overcome the presumption of correctness attaching to the finding of the collector. We do not know what method of determining the staple the collector used — if any at all — but if the importers had introduced no testimony, the presumption of the correctness of the collector’s finding would not have been overcome and would have been conclusive, under the circumstances of this case. The importers having assumed the burden of overcoming the presumption and having introduced testimony making a prima facie case controverting the presumed facts, then the presumption attaching to the collector’s finding falls. Then if the evidence supporting the importers’ contention outweighs the evidence supporting the collector’s position, the importers are entitled to a decision in their favor. In weighing the evidence, the presumption of correctness attaching to the finding of the collector is not to be regarded as having evidential value, and can not be weighed against the evidence of the party challenging the correctness of his finding. United States v. Edson Keith, & Co., 5 Ct. Cust. Appls. 82; United States v. Bloomingdale Bros. & Co., 10 Ct. Cust. Appls. 149.
We think the board’s decision is contrary to the clear weight of the evidence, and also misinterprets the intent of Congress in the enactment of the paragraphs in controversy, and its judgment is reversed.