after stating the facts as above, delivered the opinion of the court.

In the view we have of the merits of this case, it becomes unnecessary to pass upon the definiteness or sufficiency of the protest and claim filed witli t he collector. Were the hollow billets imported by appellee steel or iron? They were imported from Saudviken, Sweden. The invoice described them as “hollow steed billets,” and they were so entered at the collector's office. They are described in the record as metal billets, tubular in form, about 18 inches in length, having a diameter of 3j inches, the wall of the cylinder having a thickness of about: three-eighths of an inch. The Shelby Steel-Tube Company, for whom these billets were imported, manufactured from them cold-drawn weldless or seamless steel tubes, which are used in the .manufacture of bicycles or boiler tubes, tubes for surgical uses, and other tubes where strength and lightness is desirable with small bulk. The articles thus made are advertised and sold as steel tubes. These billets were the first of the kind imported, and were in form unknown to the trade of this country at the passage of the McKinley act of 3890. The contention of the appellee is that it is immaterial whether they were sold and bought as steel billets, or that the tubes made from them are sold as steel tubes, or that the material is such as in trade is known as “steel.” Elis contention is that; “steel,” within the meaning of the tariff act, is defined by the act itself, and that these billets are not “steel,” within the definition of paragraph 150 of the McKinley act. In Twine Co. v. Worthington, 141 U. S. 468-471, 12 Sup. Ct. 55, the question involved was the rate of duty upon an article defined In the tariff as “gilling twine.” Mr. Justice lírown said in that case:

“It is a cardinal rule of this court that, in fixing the classification of goods for the payment of duties, the name or designation of the goods is to ho understood in its known commercial sense, and that their denomination in the market when the law was passed will control their classification, without regard to their scientific designation, the material of which they may be made, or the use to which they may he applied.”

For this proposition ho cites a number of authorities.

The case at bar is altogether different, and not within the principle so clearly stated in the paragraph just cited. “Grilling twine” was not defined in the act. Where the act, undertakes particularly and definitely to define what is meant by an article upon which, a specific duty is levied, such definition is at least very persuasive in ascertaining the intent of the lawmakers, fiuth. fit. Const. § 327; End. Interp. fit. par. 365. A manifest distinction exists between definite interpretation clauses which are special and those which are general. The provisions defining the legislative meaning of a particular word used in the act containing the interpretation clause may well he regarded as a part of the law itself, and construed accordingly. Suth. St. Const. § 231, and case so cited. Undoubtedly cases may arise, as observed by Lord Denman, in which, interpretation clauses will rather embarrass the court than afford assistance, inasmuch as the interpretation danse must itself he interpreted, and may itself *500become matter of controversy. Nutter v. Board of Health, 4 Q. B. Div. 375; Reg. v. Justices, 7 Adol. & E. 480.

That the billets in question have all the physical characteristics of steel in strength, elasticity, and homogeneity of character is abundantly shown by the evidence. In chemical composition this metal also responds to approved tests for steel. The difficulty is to draw a distinction between certain grades of malleable iron and grades of mild or soft steel. Iron and steel shade into each other, and the known chemical and physical tests furnish no absolute guide by which we may always determine just when iron ceases to be iron or steel ceases to be steel. In Greenwood on Iron and Steel, a work regarded by all the experts who have been examined in this case as of very high authority, it is said of malleable or wrought iron, that it was “formerly described as iron in the lowest degree of carburization; but, with the advance which has happened in late years in the manufacture of steel, all attempts to frame a definition of ‘malleable iron’ upon a chemical basis have been futile, since in its lowest per cent, of carbon, comparative freedom from such impurities as silicon, sulphur, phosphorus, etc., occurring so largely in pig iron, it is rivalled and even excelled by the mild steels produced by the Siemens and the Bessemer processes. Definitions based upon its mechanical qualities are also equally unsuccessful, for'the superior qualities of malleability, tensile strength, ductility, and welding, which, until a comparatively recent date, were considered to be the special attributes of malleable iron, are all possessed in an equal number or superior degree by the mild steels now produced in such large quantities, and with the utmost uniformity and regularity, by the processes above mentioned.” “Steel” he defines to be “a compound of pure iron, with small percentages, ranging usually from .1 to 1.25 per cent, of carbon, existing not as graphite, but either as combined or dissolved carbon, the latter view now receiving influential support.” “All other elements, although several are invariably present in greater or less proportion, must still be regarded as impurities in the steel, notwithstanding that it may be advantageous to introduce some of them to impart special qualities to the metal, or to neutralize the effect of the presence of other of them.” His own definition of “wrought iron,” as well as of “steel,” he bases upon the mode of production. Thus, he says malleable or wrought iron “would embrace the commercial varieties obtained either as the result of the decarburization, and more or less complete separation of several of the impurities of pig iron during the process of puddling, or as the product of the direct treatment of certain^ores in the Catalan bloomery, Siemens rotary, or other furnace, in which a semi-fused product is obtained, possessing the malleability of wrought iron.” The term “steel,” “embracing also what is known as ‘ingot iron,”’ “would be reserved to distinguish such varieties of iron as are delivered in a state of fusion, allowing of the metal being cast, at once into a malleable ingot from the furnace, crucible, or other vessel in which it is produced.” Greenw. Iron & Steel (4th Ed.) 203.

This definition of “steel” is supported by some of the experts examined by the importer. It is evident that congress has adopted a *501definition of “steel” which is based in part upon the mode of production. That definition is found in paragraph 150 of the McKinley act, and is in these words:

“All metal produced from iron or its ores, which is cast and malleable, of whatever description or form, without regard to the percentage of carbon contained therein, whether produced by cementation, or converted, cast, or made from iron or its ores, by the crucible, Bessemer, Clapp-Grifiiths, pneumatic. Thomas-GilChrist, basic, Siemens-Martin, or open hearth process, or by the equivalent of either, or by a combination of two or more of the processes, of their equivalents, or by any fusion or other process which produces from iron or its ores a metal either granular or fibrous in structure, which is east and malleable, excepting what is known as ‘malleable-iron castings,’ shall he classed and denominated as ‘steel.’ ” 26 Stat. 577.

The affirmative provisions of the definition which must exist to justify classification for dutiable purposes are these: (1) That it shall be a metal produced from iron or its ores: (2) ’That it shall he a east metal (3) That it shall be malleable. (4) If these conditions exist, and it is not what was known to the trade and commerce of the country at the time of the adoption of the act as “malleable iron eastings,” then the statute requires that it shall he classified as “steel,” irrespective of the process by which it was made, or the per cent, of carbon contained, or whether it he granular or fibrous. That the metal in question is a compound of iron “produced from iron or its ores” is not controverted. The contents, other than iron, found by chemical analysis, are:

Combined carbon.....................................................20

Silicon ..............................................................02.‘>

Manganeso ..........................................................081

Phosphorus ..........................................................045

Sulphur .............................................................002

That it is granular in structure is practically undisputed. That it is malleable is conceded on all sides. The controverted question is as to whether it has been “cast.” The fact of casting does not make it steel, it being entirely possible to make steel without casting. We do not understand that this proposition is controverted by any of the experts or by any of the authoritative writers upon the manufacture of steel. The capability of being “cast” is doubtless a test of some value in determining the value of the metal and the extent of the impurities contained. The definition of “cast,” as given by Mr. Webster, is: “To form into a particular shape, by pouring liquids into a mold.” The Century Dictionary defines it as “that which is formed by founding; anything shaped in or as if in a mold.” The only direct evidence delivered by any witness, from actual knowledge of the method of production adopted by the Sandviken manufacturers of these billets, was that of Mr. Belcher, through whom the appellee bought this metal and made his importation. Mr. Belcher was a witness for the importer, and testified that he had visited Sandviken, and has seen billets of this kind made. That witness said:

“It is first made from the iron ore by what is called the ‘direct method,’— ‘Sandviken direct process.’ Charcoal is used as a fnel. It is then decarbonized, and rim into molds. There is an intermediate process, now, Mr. Greenwood. The process is not the subject of patents. It is a secret process; it is a privilege that I had to see it. I must object to telling the intermediate *502process. I will tell you the final one. It is taken from these molds to a-rolling mill. It is rolled or squeezed into pieces about ten feet long. Q. Circular? A. Just in the same form as you see it, and as I have described, — , cylindrical. It is then taken to a kind of table, where there are saws at sixteen or eighteen inches-apart, and it is pushed against these saws, you know, just as quick as we can say it, and it is sawed into this shape, lengths, sixteen' or eighteen inches apart.”

This uncontradicted evidence would seem to conclusively establish that these billets were composed of a metal which had/ at one stage of the manufacture, been cast. .The statutory definition of “steel” does not limit it to a metal which has been cast as a result of but one stage in its production. If as a first result the iron in the ore is reduced to a metallic sponge or pasty mass, and- in ihat condition delivered from the furnace, and then/ by a second operation, melted and “run into molds,” it is clearly a “cast steel,” within the meaning of the act. The definition furnishes no authority for an arbitrary limitation to a metal cast at once upon delivery from the reduction furnace.

The learned counsel for appellee has very strenuously urged ihat steel produced by a “direct process” is never cast, and that, when his witness Belcher said that this metal had been “run into molds” at one stage of its production, he did not mean that it had been “cast.” He has urged that by the direct processes the iron is reduced from its ores to a spongy ball or mass, and in that condition hammered into a bloom, or by other pressure squeezed into a mold or form, and ihat Belcher’s statement that the Sandviken manufacturer produced the metal here involved by direct reduction of the ores, using charcoal as a fuel, is the controlling fact stated by him, and that which he further said is to be understood in the light of the theory that the “direct processes” are all inconsistent with casting. This theory is, to some extent, supported by one of the expert witnesses who testified for the importer, who expresses the opinion that these billets were never cast, but produced as wrought iron as at the bloomery furnaces of this country. Mr. Greenwood describes steel as produced in three ways: First, directly from certain pure iron ores; second, by the decarburization of malleable iron; or, third, by the decarburization of pig iron. Greenw. Iron & Steel, par. 677. The same author says, at paragraph 678, that '“the direct reduction of iron ores for the production of steel embraces the reduction in the Catalan forge, in the Siemens rotator, by the Chenot process, etc., in each of which processes rich ores of iron, such as the purer oxides, are heated along with charcoal or carbonaceous matters, and thereby either steel or a hard steely iron is produced.” By “steely iron” we understand to be described a grade of mild or soft steel which approaches so closely in physical and chemical characteristics the metal known as “malleable wrought iron” as to be difficult to distinguish the one from the other, or a grade of wrought iron so nearly approaching steel as to be almost undistinguishable. Mr. H. De B. Parsons, an expert of marked intelligence, declares the term “steely iron” to be unscientific,- and one which 'should not be used. He says that the point when -a metal ceases to be wrought iron and reaches a grade known *503as steel is difficult of definition, in that “they fade away, one class to the other, as daylight to darkness.” But he gives it as his opinion that, when the physical and chemical qualities are given, the metal should always be classed as the one or the other, and that only when doubt exists is the term “steely iron” employed. The method of production he denominates the “Sandviken direct process.” “It is made,” he says, “by what is known as the ‘direct process,’ straight from the ore.” There seems, from Mr. Belcher’s meager description of the method of manufacture adopted by tbe Sandviken people, to be three stages or steps in the process. The first is the reduction of the ores, charcoal being used as a fuel. Whether this reduction operates to liquefy the iron contents of the ores, or to reduce it simply to a spongy or pasty ball, he does not say. The' third and last stage he describes as occurring after the metal is taken from the molds into which it has been poured. “It is taken,” he says, “from these molds to a rolling mill; it is rolled or squeezed into pieces about ten feet long,” of a cylindrical character. These pieces are then sawed into lengths of eighteen inches, and the billet as exported is complete. The second or intermediate process, he says, is a secret process, and he declines to explain it. J t is, however, very (dear that (his intermediate process was one by which the metal was decarbonized, and is a step which occurred before the metal was “run into molds.” If it be true, as this witness deposes, that, at one stage of the production of this Handviken metal, the metal was “run into molds,” then we clearly have a. “cast steel.”

It is wholly immaterial, under the statutory definition of “steel,” whether the metal be cast when first delivered from the furnace, or cast after a second operation, by which, after removal from the furnace as a pasty or spongy mass, it was melted and decarbonized. The “secret process” which is described as an intermediate process by Belcher, and which he declines to divulge, probably consisted in the means used to decarbonize the metal before running it into molds. Certain it is that the metal was rendered so plastic as to enable it to be run into molds, for the direct evidence of the witness is that it was “run into molds,” before the final process by which it was rolled into the shape we find it in commerce. Under this positive evidence it is impossible to assume that the witness meant, by running into molds, that the metal as a spongy mass was either hammered into a bloom or squeezed into a form. Neither is this evidence in conflict with the other statement of the witness that the metal was produced from the ore direct. It is true that a steel is made direct from the ores without casting by some of the direct methods, but it is not true that all the “direct” methods exclude casting. A rude form of steel making is that of the Catalan furnace. By the Catalan method, the iron in a spongy mass, as described by Greenwood, is withdrawn from the furnace, “to be shingled under the steel hammer for the expulsion of the slag and extraneous matter, together with the consolidation of the mass by welding together the spongy granular mass into a more solid bloom.” Wrought iron is made in the same way in bloomery furnaces. Another “direct” method is described by Greenwood as the “Chenot process.” This involves two *504operations, “in the first of which,” says Mr. Greenwood, “a metallic sponge is obtained, and in the second operation this sponge is melted in crucible along with carbonaceous matters.” The spongy mass, when thus decarbonized and reduced to a plastic condition, is cast into ingots. Greenw. Iron & Steel, pars. 685, 686. Another direct method described by the same author is called the “Siemens direct process.” By that method the metal is produced from the ores direct and, when reduced to a mass of spongy consistence, “is shingled under the hammer, or pressed in squeezers or other apparatus,” in the same manner as by the Catalan process or bloomery methods, before mentioned. “But,” says Mr. Greenwood, “the process, as now applied to the manufacture of steel, is generally used only as a preliminary stage in the production of steel in the open-hearth steel-melting furnaces, to which furnace the balls of metallic sponge, or the shingled blooms from the same, are at once transferred from the rotator for fusion with the other materials of the ordinary charge of the steel-melting furnace.” Paragraph 678. In determining whether these billets are composed of a cast steel, nothing material can be predicated of the absence of cast marks, or of their hollow cylindrical form, for the only direct evidence in the record shows that this form was the result of a rolling-mill process, subsequent to the removal of the metal from molds into which it had been run. The great weight of expert evidence contained in this record, based upon physical structure and characteristics and chemical composition, is in harmony with the opinion of the general board of appraisers. The opinion of expert Gray, so much relied upon by the' appellee, that this metal shows too little manganese to have permitted rolling if made by any of the processes he names, and too little silicon to have been made by a crucible process, is not supported by the other expert evidence in the record, much of which comes from witnesses of the highest intelligence, wholly disinterested. Phosphorus, sulphur, and silicon seem to be regarded as impurities in steel. Manganese is an antidote which serves to minimize the evil effects of their presence. The fact that these impurities were present in very small quantities rendered the use of but very little manganese necessary. Thus the weight of expert opinion and the only direct evidence in the record concur in establishing the fact that these billets, at one stage of their manufacture, have been cast Having all thé other requisites of steel, we are constrained to reverse the judgment of the circuit court, and sustain the classification of the general board of appraisers. The cause will be remanded, with directions to enter judgment in accordance with this opinion.