The merchandise involved in this case consists of 10 carloads of lead residue imported from Canada and entered at the port of Philadelphia in July and August of 1963. It was assessed with duty as lead dross under paragraph 392 of the Tariff Act of 1930, as modified by the Torquay Protocol to the General Agreement on Tariffs and Trade, 86 Treas. Dec. 121, T.D. 52739, at V/16 cents per pound on the lead content. It is claimed to be flue dust dutiable under paragraph 391 of said tariff act, as modified, supra, at % cent per pound on the lead content.
The pertinent provisions of said tariff act, as modified, are as follows:
391 Lead-bearing ores, flue dust, and mattes of all kinds- per lb. on lead content
392 Lead bullion or base bullion, lead in pigs and bars, lead dross, reclaimed lead, scrap lead, antimonial lead, antimonial scrap lead, type metal, Babbitt metal, solder, all alloys or combinations of lead not specially provided for_ 1%60 per lb. on lead content
At the trial, plaintiff called the following witnesses:
Morris B. Kaufman, president of Toronto Refiners & Smelters Company of Toronto, Canada, who testified that he had been with the company for 40 years and that its business is the smelting, refining, and manufacture of lead and lead articles, particularly the manufacture of lead for the purpose of manufacturing batteries;
*183William A. McElroy, sampler and weigher in the employ of Alfred J. Cooke, who had been in that occupation for 20 years;
Alfred J. Cooke, who has been in the business of analyzing, sampling, and weighing nonferrous metals, residues, and scrap for 35 years;
William Toner, analytical chemist, in the employ of Alfred J. Cooke; and
Milton S. Hess, purchasing agent of Bers & Company, Inc., of Philadelphia.
In addition, plaintiff introduced into evidence photographs showing conditions in the yard of Toronto Refiners & Smelters Company (exhibits 1 through 6); purchase orders and settlement sheets between Toronto Refiners and Bers & Company (exhibits 7, 8-A, and 8-B), and the assay reports of Alfred J. Cooke (exhibit 0).
Defendant called as its witnesses:
John Augustine, customs supervisor sampler, who had been a sampler in August and September 1963;
James W. Beckwith, customs inspector at Philadelphia, who had been a customs sampler in the summer of 1963;
Joseph R. Piccinino, commodity specialist, who had been a customs examiner or examiner’s aide in 1963;
Arthur L. Grill, chief chemist of the Philadelphia U.S. Customs Laboratory; and
Alvis C. Mansfield, who has held the position of analytical chemist in the U.S. Customs Laboratory at Philadelphia for over 20 years.
Defendant placed in evidence the samples which were the basis of the Customs Laboratory reports (exhibits B through J); A-5 forms with instructions for sampling the merchandise (collective exhibits K and L); and Customs Laboratory Record Cards (exhibit M).
The principal issue in this case is whether the imported merchandise is flue dust. According to the testimony, lead residues include flue dust, scrap material, battery lead, lead drosses, solder, Babbitt, and type metal, in fact, everything left after refining lead. Flue dust is a type of lead residue consisting of particles in the form of fines removed from the smoke which occurs in the course of smelting. It is collected on baghouse filters or by a scrubber and water operation. Mr. Kaufman testified that in his opinion and in the opinion of smelters, flue dust is the lowest grade of merchandise in the smelting field. It is unhealthy and poisonous and is difficult to market.
The merchandise involved herein originated in the course of the production of tetraethyl lead by the Ethyl Corporation of Canada. It was purchased in May 1962 and May 1963 by Toronto Refiners & Smelters Company of Toronto, Ontario, and was resold to Bers & Company, Inc., of Philadelphia in June and July of 1963.
*184The Ethyl Corporation collected flue dust by a scrubber operation and stored it in water in tanks in the ground. The instant merchandise was wet when removed from the tanks, but, by the time of arrival in Toronto, most of the water had leaked out. At the Toronto plant, it was stored out of doors in piles in a cement yard near the railroad tracks. According to the evidence, such merchandise hardens in the sun and tends to get lumpy and chunky.
It is clear from the record that the material produced in the course of the operations of the Ethyl Corporation was flue dust and that it was flue dust at the time that it arrived at the yard of Toronto B,e-finers & Smelters Company. Some of that material remained in that yard over a year and some of it several months. During that time, tons of all kinds of metal used in the operation of the smelter, including battery plates, battery grids, broken batteries, and scrap, passed through the yard. While every effort was made to keep such material out of the flue dust pile, it appears that some pieces did fall on it and were not removed. Under the circumstances, it would be almost impossible to keep all extraneous matter out.
Mr. Kaufman testified that the maintenance of uncontaminated stockpiles was essential to the proper economic operation of the smelter. Inventories must be kept properly. Materials cannot be mixed and smelted. The market value of battery lead scrap was three to five times greater than that of flue dust. The seller would be penalized if it shipped flue dust with battery plates mixed in it and would not want to ship merchandise that cost 5 or 10 cents a pound with merchandise for which it was getting only 2 cents a pound. Clearly, it was to the interest of the seller to keep the flue dust as free from other materials as possible.
According to Mr. Kaufman, the trade tolerates a certain amount of contaminants in flue dust. He said the maximum was 25 to 30 pounds in a carload weighing 150,000 pounds, or 0.02 percent. Such contaminants might be an odd piece of a battery grid, a piece of the top of a battery with the cables on, an odd piece of metal, or a piece of round rubber. Mr. Kaufman said that in the instant case, he inspected the merchandise as it was loaded onto the railroad cars for shipment to Bers and found that there was no more than the maximum of 25 to 30 pounds of contaminants present.
Both Mr. McElroy and Mr. Cooke stated that it was not unusual to find a small percentage of contaminants in residues such as those involved herein. Mr. McElroy said he meant traces, a few pieces of any kind of material, about one-tenth of one percent or less. He had supervised the taking of samples when the instant merchandise was being unloaded and did not see any unusual amount of contaminants. Mr. *185Cooke considered anything less than one percent a small amount of contaminants and said he had found all sorts of things in cars — iron, monkey wrenches, items that inadvertently are thrown into a car, even three or four bushels of corn.
Different samples were drawn by the parties and different conclusions reached as to what the merchandise was at the time of importation.
Plaintiff’s samples were drawn under the supervision of Mr. Mc-Elroy who testified that as each car is unloaded every fifth bucket is placed in a separate pile. That pile is mixed, coned, flattened, and quartered again and again until it is reduced to approximately 25 pounds and any lumpy material is broken down to uniform size. The sample is weighed, dried, and the percentage of moisture determined. The material is then run through a micropulverizer and screened through a GO mesh screen. The sample is mixed and quartered again and worked down to a laboratory sample about the size of a pay envelope.
The prepared samples herein were analyzed by Mr. Toner by the gravimetric method of lead dichromate, the result of his analyses being found in the reports in collective exhibit 9. These reports designate the merchandise as lead flue dust, give the moisture percentage found by Mr.McElroy, and the lead content on two bases, “Lead on Dry” and “Lead on Original.” According to Mr. Toner, flue dust is a highly oxidized material containing lead, and is gray or yellow in color.
The lead content found by these assays was the basis of the price settlement between Toronto Refiners and Bers, Toronto Refiners being paid only for the lead content. While the purchase orders (exhibits 7 and 8-A), referred to lead residues, the settlement records, except as to the carload covered by exhibit 7, designated the merchandise as flue dust. Payment was based on the “Lead on Original” percentage in the assays, less 1.50, less smelting and other charges.
The Government’s contention that this merchandise is not flue dust is based on Mr. Mansfield’s analyses of samples taken by Government samplers. These samples were obtained by digging down 6 inches in each railroad car and scooping out some material. One of the samplers, Mr. Augustine, testified that the material was gray and metallic; that it had a hard crust and little lumps in it, and that it looked uniform to him.
Mr. Mansfield testified that in physical appearance one sample was dark gray and the others light gray, and that there were some yellow particles 'and some pieces of metal in them. He said they contained some metallics and some stone-looking material and that the metallics and fines were separated. Both portions were weighed and the percentages *186determined. The metallics were pieces of metal, such as metal that had solidified, or pieces of 'battery grids. One sample had chips of a broken rubber casing from a storage battery.
Mr. Mansfield analyzed the samples by the volumetric molybdate method and recorded the results on the cards comprising exhibit M. On the back of these cards are notations describing the physical appearance of the samples and giving the percentages of metallics which ranged from 4.56 to 8.89 percent. Mr. Mansfield testified that these samples did not fall into a pattern of anything he was familiar with. Two resembled lead dross and the others looked like a mixture of powder and broken pieces of battery grids and chips of rubber casing. He concluded that the merchandise was not flue dust on the ground that it contained metallic pieces and also because of its color. He said he had never analyzed flue dust which contained large lumps of metallic lead. He had tested flue dust from Mexico and said that it was a light color, much lighter than the samples herein. In his experience flue dust which is a light gray powder is exceptional. The witness stated, however, that the chips of battery casing or battery grids in the material could be contaminants. They were foreign to the sample he received and he did not know whether they were seeded throughout the car. In his opinion the material herein was a type of lead dross because the physical appearance of several of the samples was identical to that of a lead dross. He said it was a lead bearing residue but was not flue dust. In his view, it came within the definition of lead dross in the Summary of Tariff Information, 1929, page 892:
Lead dross is an intermediate product, skimmed from molten lead or residual from other processes and is, like scrap lead, a source of secondary lead production.
Mr. Mansfield said that lead drosses could be obtained from skimming or from cleaning a lead product and that in his experience there were other lead bearing residues which were considered lead drosses. He said that lead dross normally ranges from a gray to a grayish-yellow, depending on the lead oxide content, and contains a large percentage of metallics, from 30 to 80 percent. It does not usually contain either flue dust or battery grids. He said that scrap lead samples could have almost anything in them but he did not regard the merchandise herein as scrap lead.
Mr. Cooke, recalled in rebuttal, testified that he handles 5 or 10 flue dust tests a week and that some are light yellow, some light gray, and some dark. He had seen flue dust the color of the Government samples, exhibits B through J. In his opinion, the merchandise herein was a flue dust because it was a highly oxidized material. It was not a lead dross because lead drosses have a very high percentage of metallics. He would *187hesitate to call anything a dross that had less than 30 percent metal-lics. While flue dust does not usually have as high as 5 to 8 percent metallics in it, he has run into 2% to 3 percent.
Mr. Cooke also testified that the method of sampling used here was standard procedure on most residues of this type. Where cars of lead residues contain lumps of metal representing 30, 40, or 50 percent of the car, those portions have to be sampled separately and a composite sample made eventually. In his opinion, grab samples were not representative because the sampler tries to get a little bit of everything he sees regardless of its relationship to the load. He stated further that a sample from the top of a car might be altogether different from the material on the bottom. Mr. Grill also testified that where the merchandise is not homogeneous, it would be necessary to take several samples throughout the car, blend them, and reduce them to a representative sample.
Since defendant’s samples in this case were taken merely by scooping out some material from about 6 inches down in the railroad cars, they are not as representative as plaintiff’s samples, particularly as to metallic pieces or other foreign matter, since it is unlikely that such pieces would be uniformly distributed throughout the car. Larger samples and samples of an entire shipment have been held to be more representative. Geo. Sail Metals Co. Phila. v. United, States, 14 Cust. Ct. 265, Abstract 50236; Generad Aniline Works, Inc. v. United States, 6 Cust. Ct. 511, Abstract 45218; Philipp Bros., Inc. v. United States, 30 Cust. Ct. 216, C.D. 1523; Consolidated Cork Corp. et al. v. United States, 54 Cust. Ct. 83, C.D. 2512. Since the Government’s samples here were unrepresentative, Mr. Mansfield’s conclusion that the merchandise was not flue dust because of the percentage of metallic pieces in the samples is of little probative value.
Furthermore, the evidence refutes Mr. Mansfield’s finding that the merchandise is lead dross. It was not obtained from skimming molten lead or cleaning a lead product. Both Mr. Mansfield and Mr. Cooke said that lead dross contains a large percentage of metallics, from 30 to 80 percent. Mr. Cooke testified that he would hesitate to call anything a dross which had less than 30 percent metallics. The merchandise here, using Mr. Mansfield’s figures, contained no more than 9 percent metallics, and according to plaintiff’s witness contained much less.
Defendant claims that it has not been established that a material consisting of flue dust and lead particles is known in the trade and commerce as flue dust and that the commingling Statute (section 508(a), Tariff Act of 1930, as amended) applies.
As previously stated, there is evidence that the trade accepted a certain amount of tolerance for various contaminants. The percentages *188given as acceptable were less than the percentages of metallics found by Mr. Mansfield, but his findings are inconclusive since based on unrepresentative samples. Other evidence indicates that the merchandise did not contain an unusual amount of contaminants. It would appear from the settlement records that the imported merchandise was acceptable to the purchaser as flue dust, a particular kind of lead residue.
The commingling statute applies where the merchandise consists of two distinct commodities, not impurities, subject to different rates of duty, which have been commingled together in shipment, but not where the merchandise is a single tariff entity. United States v. F. W. Myers & Co., Inc., 45 CCPA 48, C.A.D. 671; Consolidated Elevator Co. v. United States, 8 Ct. Cust. Appls. 267, T.D. 37536; United States v. Amendola, 5 Ct. Cust. Appls. 516, T.D. 35156; J. M. McCulloughs Sons Co. v. United States, 65 Treas. Dec. 595, T.D. 46985. Moreover, section 507, which permits allowances for tare and draft in certain circumstances, contemplates that there may exist impurities in imported merchandise, but that their presence shall not affect the duties to be levied on the merchandise, unless the impurities are of such character or in an amount not usually found in such or similar merchandise. Cargill Grain Co., Inc. v. United States, 30 CCPA 78, 88, C.A.D. 219; Socony Vacuum Oil Co., Inc. v. United States, 44 CCPA 83, 89, C.A.D. 641; F. W. Myers and Co., Inc. v. United States, 47 Cust. Ct. 195, C.D. 2302.
In I. C. Harris v. United States, 71 Treas. Dec. 678, T.D. 48936, the merchandise consisted of aluminum borings, containing 58.6 percent aluminum, less than 4 percent copper, and 41.4 percent dirt, oil, moisture, and other refuse. The court held that section 508 applied only where there were two different kind's of merchandise and that moisture, dirt, and other refuse did not constitute merchandise, and that the negligible amount of copper did not bring the importation within the terms of the section.
In. Pacific Vegetable Oil Corp. v. United States, 15 Cust. Ct. 161, C.D. 964, the court cited the Consolidated Elevator Co., case, supra, and stated (p. 164) :
* * * The court there construed the tariff term “screenings” as a commercial product having a distinct value and a particular use and it confined the term “impurity” to such as had no value and was destroyed. Clearly the tariff meaning of the term “screenings” does not embrace all that is removed from a commodity by a screening process, as that term was defined by the witnesses herein. In view of the foregoing decision, refuse, consisting of seven-hundredths of 1 per centum *189of foreign seeds, a negligible quantity, and 4.89 per centum of dirt and chaff would not be considered a commercial product, nor is there any evidence before us that the instant refuse ivas so used.
While the foreign material in the flue dust in the instant case may not have been impurities in the strict sense, and may have had some commercial value, the amount was small and there is no evidence that it was separated out and used as a distinct commercial product.
The evidence establishes that it was of such character and amount as was usually found in such or similar merchandise; therefore, its presence does not affect the classification of the imported merchandise as flue dust.
While the protest claims that the collector found an excessive lead content, no mention of this claim is made in plaintiff’s brief and it seems to have been abandoned.
The evidence does not clearly establish the method 'by which the lead content was determined by the collector. According to worksheets attached to entry 5812, Cooke’s assay was used as the basis. If the method used was different from that employed in reaching a commercial settlement, the evidence does not establish that the collector’s determination was erroneous or that some other figures were correct.
We hold that the merchandise involved herein is properly dutiable under paragraph 391 of the Tariff Act of 1930, as modified, as flue dust, at % cent per pound on the lead content as found by the collector.
To that extent the protest is sustained. Judgment will be rendered accordingly.