WOOLLEY, Circuit Judge.
The bill charges infringement of Letters Patent No. 1,071,364, issued August 26, 1913, to James Ramsey Speer and William L. Forster, assignors of the plaintiff, for “Alloy of Iron.” The defenses are invalidity and non-infringement. The District Court found for the defendant on both issues and dismissed the bill. 236 Fed. 756.” The plaintiff took this appeal.
[706]*706[ 1 ] The case has to do with an alloy of iron, which, in addition to iron, contains seven ingredients in small and varying proportions. They are carbon, silicon, chromium, nickel, phosphorus, manganese and sulfur. All these ingredients are found both in the product manufactured by the plaintiff under the patent and described to the trade as “Adamite” and in the alleged infringing product of the defendant made and marketed under the name of “Phoenix Metal.”
The patent contains two claims, the second prescribing ingredients in precise proportions, the first, in proportions within given ranges. The first claim only is in issue. The patentees say:
“We claim — 1. As a new article of manufacture, an alloy comprising essentially silicon .10% to 2.00% ; chromium .5% to 1.50% ; nickel .25% to 1.00% ; sulfur, not exceeding .05%; phosjihorus, not exceeding .12% ; manganese, not exceeding .45%; total carbon 1.25% to 8.50%; and iron approximately sufficient to complete the 100%.’’
Alloys containing the ingredients of the claim appeared in the manufacture of iron and steel long before the patent. These several ingredients were known to possess definite characteristics or properties, which, when brought together in alloy, bore upon one another and upon the ultimate product in well defined ways. The patentees discovered no new chemical properties or functions in them. Variations in their proportions produced variations in their chemical effect upon one another, and in the chemical and physical characteristics of the product. Thus when a given property, as hardness, toughness, strength, was desired, the founder obtained it by employing in alloy the particular ingredients which the metallurgist, had taught him would produce that property and by pursuing methods of heating which he had largely taught himself. The learning of the metallurgical engineer and the skill of the founder, though separate and distinct, were inseparably combined in producing the result.
Men of these classes in vast numbers had been working together many years before the patent, and by thousands of tests and experiments and by manifold achievements, had made the metallurgy of iron and steel a broad and highly developed field. A claim to monopoly of any part of that field by one entering it at this late day can be sustained only by clear proof of discovery of something there not before found, of an invention of something not before there. Such is the claim of the patentees.
The thing which the patentees claim to have invented is an alloy, which is-“a new article of marufacture” — in a sense, a new metal. Its base is iron, but the metal ,.s neither cast iron nor steel, though possessing characteristics of both. It resembles cast iron in its lack of the quality of elongation, in its resistance to abrasion and in its high carbon content; it resembles st.eel in its tensile strength, toughness, capacity to be worked or iorged, and in the form of its carbon content. It is claimed that it bridges the gap between irons and steels —having chemical characteristics more nearly ¿approaching the former, with physical properties equal to if not excelling the highest grades of the latter. It has been termed a metallurgical paradox. The name given it is — “Adamite.”
[707]*707The validity of the patent depends upon the ability of the plaintiff to establish two facts: First, that Adamite is a new metal; and, second, that the formula of the claim will produce it. The first is the basic fact, for the formula of the claim, both as to ingredients and proportions, is not altogether new. When we speak of Adamite as a new metal, we mean new in the sense of being different from, not merely a variation of, an old metal, as .steel differs from cast: iron.
We shall not discuss in this opinion the highly technical subject-matter of the patent the full significance of which is difficult to grasp without elaborate presentation, but shall content ourselves with a statement of the conclusions upon which, after close study, we base our decision.
[2] The striking thing or central characteristic in the chemistry of the patent alloy is its carbon content, both as to quantity and form. In quantity it lies well within the range of cast iron, but in form, it is that of steel. Its form is mainly combined carbon. That term, as used in the metallurgy of iron and steel, means carbon in union with some one or more metallic constituents in the iron alloy, for instance, iron carbide or the double carbide of iron and chromium. The method of obtaining very high carbon in combined form, by the given proportions of carbon combining and metal hardening ingredients of the patent alloy, in conjunction with the skill of the founder, is a matter of less concern just here than the fact that in the alloy of the patent, high carbon combined is chemically obtained, and that it contributes to and in a measure produces characteristics not before shown in airy metal.
Aside from the chemistry of its composition, Adamite has characteristics of structure and performance which distinguish it scientifically and commercially from other metals. It has the quality of intense hardness and toughness (steel characteristics) coupled with an extraor - dinary ability to withstand abrasion, especially under heat (cast-iron characteristic). These are procured without the use of the expensive iugredients, chromium and nickel, in large quantities, by depending primarily upon high carbon, the cheapest hardening element — and using it in excess of what was considered permissible in a product having steel characteristics. Its tensile strength is more than twice that of ordinary soft steel. While possessing the stiffness of cast iron with its casting characteristic, it has the hardness and toughness of steel with the ability to be forged, rolled and worked like steel.
It is testified that when made into rolls it does from two to four times the work of other rolls in use when it came on the market and lasts from two to four limes as long, thereby increasing output and decreasing cost of production and operation.
The fact that Adamite possesses these qualities and that they are qualities of merit, is evidenced by the tribute paid by the art in very substantial royalties and in a steadily increasing consumption at prices markedly higher than those asked for rolls o C other metals.
While the defendant has met the case with much evidence that the art contained a vast number of alloys of varying degrees of hardness, toughness and strength, produced by the energetic agents of carbon. [708]*708chromium and nickel free from patent restraint, it has not shown that these characteristics or properties existed in the measure, in the combination, and in the peculiar relation in which they are found in Adamite. It has not shown that Adamite does not possess the characteristics claimed for it, or that any product possessed them before. Jn fact, there is evidence in abundance that there is such a product as Adamite; there is no evidence that there is no such product.
The alloy being novel, the next question is whether it was the result of an. inventive conception. It was manifestly not an obvious combination. While its proportions were reached after many experiments, they were the result to which the experiments were directed. Somewhere in their progress lay a conception of the thing achieved. If the thing is otherwise a patentable invention it does not fail for lack of a definite point at which the conception appeared.
Aside from the question of patentable invention, the validity of the patent is challenged on the ground that its disclosure is insufficient. In this we find for the patent. The charge of insufficiency of disclosure is based doubtless upon the lack of disclosed mechanical detail, made necessary by the fact that the skill of the founder in pursuing foundry practice is called upon to supplement the metallurgical disclosure of the patent. It is, impossible to disclose in a patent the founder’s peculiar skill, and being no part of the invention, it is not required to be made a part of the patent disclosure.
That the disclosure is sufficient when read by one skilled in the art, is evidenced by the fact that when followed with skill, the product of the patent is obtained. When followed by one unskilled or. when something else is obtained, the validity of the patent ceases to be a matter of concern.
The remaining question as to the patent’s validity is whether the disclosed invention had been anticipated by alloys of the same formula in the prior art.
The defendant found in the prior art and produced in evidence for-mulae and analyses of many iron and steel products, which were either within the analysis of the patent alloy or as close to it as is the analysis of the product of the defendant upon which the plaintiff bases its charge of infringement. These analyses appeared in patents, publications, and in the records of steel manufacturers kept in the regular course of their business. They are Hadfield Patents, Nos. 732,365 and 735,666; the Borchers Article; the Totten Patents, No. 872,483; 878,691; 947,263; Isaac G. Johnson Co.; Crucible Steel Co.; use of Mayari ore by Carpenter Steel Co., Maryland Steel Co., Reading Iron Co., and the defendant.
To the voluminous evidence upon the defense of anticipation we have given careful study. We find it unnecessary to discuss the evidence in this opinion for the reason that all of it is fatally deficient in one essential, namely, in failing to show, that the products of the several prior art analyses were Adamite.
While showing products of varying characteristics and recording faithfully the progress of the art of steel making, the alleged anticipating alloys nowhere showed or even suggested when made or pub[709]*709lished that they were Adamite or were metals which possessed in structure or performance the Adamite characteristics of cast iron and steel, or suggested to the art either the existence of such a metal or how to make it. If any one of the alleged anticipating alloys was Adamite, that fact, so far as the record shows, was not known to those who produced it or used it, and not being recognized'as a new product with its distinctive characteristics, its production was purely an accident without profit to the art and without value as an anticipation. We are satisfied, therefore, that Adamite has not been anticipated by alloys, which, while accidentally of the same analyses, were not shown to be the “article of manufacture” of the patent.
We are of opinion that Adamite is what it is said to be, namely, “a new article of manufacture,” different from but similar to both cast iron and steel, that its production involves invention, and that claim 1 of the patent is valid.
We now approach the question of infringement. The patentees admit that as metallurgy is not an exact science, any use of the patent-formula within the wide range of each element will not necessarily produce Adamite (otherwise they are forced to admit anticipation), and that Adamite is produced only when one skilled in the art follows the formula with a proper regard to the correlation of its several elements. This being admitted, then certainly a person may pursue the formula and not get Adamite. Then he does not infringe. Or he may pursue the formula and get Adamite. Then he does infringe. The principal question upon the issue of infringement, therefore, is, not whether the defendant’s product was of the analysis of the patent, but whether the defendant’s product was the product of the patent, namely Adamite. This being the question, it may be inquired into along three lines:
1. Whether the formula of the defendant’s alleged infringing alloy "was the patent formula;
2. Whether it was outside the patent formula because of its higher percentage of sulfur; and
3. Whether the product of the defendant, either within or without the formula of the patent, was Adamite.
The analyses of the defendant’s metal, offered as evidence of infringement, were not literally within the formula of the patent. They were within the formula as to percentages of all ingredients except that of sulfur. As to percentages of that ingredient they were uniformly and radically higher. The defendant maintains that this difference saves it from infringement; the plaintiff charges that the defendant used an excess of sulfur as a supplement to a low percentage of chromium (though such low percentage was within the chromium proportion of the patent); that this sulfur excess was the equivalent of any deficiency in the chromium of the patent alloy, and constituted infringement. Upon this point there was sharp controversy among the experts. But whatever may be the solution of this metallurgical question, the thing which we want to know is — what did the defendant get from what it used. If it got Adamite, then we must decide the question of equivalents in order to determine whether it got [710]*710it in the way the patent taught. If it did not get Adamite, it is immaterial whether an excess of sulfur was or was not an equivalent of, any part of the patent percentage of chromium, for then it did not infringe.
The parties to this litigation are rival manufacturers of rolls. Th'e essentials of a good roll sought by all roll manufacturers are hardness, resistance to abrasion, strength to withstand prodigious strains, and durability. Each had striven independently and in its own way to make improvements in these qualities. Each had succeeded. They had worked, as it now appears, on diametrically opposite lines. The plaintiff, through the patentees, employed chromium and chromium alone as an element to overcome the natural inclination of carbon to crystallize and to fix and hold it in combined form (omitting for the moment the influence of the element of silicon). The patentees did not employ sulfur for that purpose. They defined in their patent what elements were “essential” to produce that chemical result. Sulfur was not one of them. In fact they considered sulfur an impurity and repudiated it as an element or chemical to produce any desired result. They recognized it as a necessary evil and arranged to overcome it. They dealt with sulfur in iron because it was there and because it was impossible to eliminate it economically. They regarded it as a nuisance and an element not to be used but to 'be gotten rid of so far as could be done. Therefore, in their formula, the patentees placed sulfur at a low maximum (.05%), and called for its reduction to almost the vanishing point.
In making its rolls, the defendant pursued just the opposite practice. It recognized a merit in sulfur. It considered that (in its business at least) the merit of sulfur offset its objectionable quality and proceeded to make use of it. It did not reduce the normal sulfur content of its raw material, but, long before the patent (March, 1911), it actually increased sulfur above the normal content of the ore by deliberately adding iron pyrites to its alloy. It used a percentage of sulfur that was from two to four times the permissible maximum of the patent proportion. It did. this to obtain an effect which was not then widely known sulfur would produce. That effect was double: first, it hardened and made smooth the surface of the roll, and second, (whether the defendant knew it or not,) it acted to hold the carbon in combined form not merely on the surface but throughout the section, and produced the result (which the defendant did know) of hardening the whole metal, the very quality which later the patentees sought, and obtained perhaps in a greater degree, by the use of chromium. If sulfur in the proportion used by the defendant did hold carbon in combined form and did thereby harden the metal of the section, the defendant achieved the result in a way different from that disclosed by the patentees/ for the patentees’ way was by the use of chromium.
Rut the defendant did not long rely upon sulfur alone for the hardening of its metal; because it was well known that chromium would overcome the opposite effect of silicon and hold carbon in combined form and harden metal. Therefore, as a combined chemical and commercial expedient, the defendant began as early as September, 1911, [711]*711(six months after resorting to sulfur and two years before the Ad-amite patent,) to use an iron ore known as Mayari ore, because, unlike any other iron ore, it contained in its natural state the alloying elements of chromium and nickel. It used this ore with these two chemical elements in order to obtain in the metal of its rolls the very hardening effect which chromium was then known to produce, atid the toughening effect which nickel was known to produce. By hundreds of assays of metals made through two years before the patent, the defendant showed that, in the manufacture of its rolls, it used the chromium and nickel content in Mayari ore in varj/ing proportions, (sometimes within the proportions afterwards given in the patent,) together .with sulfur in much higher proportions than the sulfur proportion of the patent formula.
The natural chromium of Mayari ore was used to supplement the high sulfur of its previous practice, and the two became the agents employed by the defendant long before the patent to do the thing which those two ingredients were 1 hen known to do, namely, hold carbon combined and thereby harden metal. It is claimed by the defendant that it began this practice of roll metal making Jong before the patent and consistently pursued it down to the bringing of this suit. It is not disputed that the defendant began it before the patent and jirac-ticed it beyond the date of the patent in a manner which metallurgi-cally and commercially was a distinct advance in the art.
Having shown that for their purpose sulfur is bad, certainly the patentees cannot complain of its use by another who finds that for his purpose it is good. The defendant says in effect:
“Sulfur is bad tor some purposes, but I have found that it combines carbon and hardens metal. Therefore, for my purpose sulfur is good. I will use it in large quantities. Chromium does the same thing. Everybody knows that. I will use chromium too. I will vary the proportions of one or the other or of both of these ingredients in my alloy until I get the hardness I want. Nobody told me to do this. Certainly the patentees did not.”
As we have said, the defendant had used chromium and sulfur in varying proportions to combine carbon and procure hardness for some years before the patentees obtained the patent whereby carbon is combined and hardness obtained essentially by chromium and in no degree by sulfur. Their licensee, now complains that in continuing its early chromium-sulfur practice beyond the date of the patent, in different proportions, the defendant employed an equivalent of the alloy of the patent, and infringed. The charge arose in this way: After using Mayari ore for three years, with nickel and chromium running on an average of about .30% and .50% respectively, and in the relation of about three parts nickel to five parts chromium, with sulfur at about .16% the defendant, on August 17, 1914, raised the percentages of nickel and chromium to about .65% and .90% respectively, and lowered sulfur to about .13% on the average; and' again, in 1915, the defendant lowered sulfur to about .12% and used nickel and chromium on an approximate parallel of about .80%. In this-the plaintiff charges the defendant with infringement. It supports its claim of infringement by the argument, though the defendant’s increase of chro[712]*712mium and nickel was not in the relative proportions of these two elements as they are taught b}r the patent formula, (two parts of chromium to one of nickel,) that the substantial amount of sulfur made up for the low amount of chromium, (though itself within the patent range,) and they together produced the result which chromium would produce alone if used in the larger proportion of the patent formula, and that therefore the use of sulfur in connection with chromium was a fair equivalent of the latter element in the larger proportion of the patent.
This contention is without force as an argument of infringement, unless it is accompanied with evidence that the product of such an equivalent was Adamite. If the product was something else, whether as good or not so good, it did not infringe even if the formula used were the equivalent of that of the patent. The grant of Tetters Patent to Speer and Forster did not deprive the defendant of its right to continue its long established use of high sulfur, or its almost equally long use of chromium and sulfur to harden roll metal by combining carbon; nor did it restrict the defendant to the precise proportions in which it had before used those two elements; nor did it prevent it using them in higher proportions, even under the plaintiff’s theory of equivalency, when the resultant product is not Adamite. It is then pertinent to inquire — What was the defendant’s alleged infringing product? The plaintiff says it was Adamite, and proceeds to prove it in two ways, first, by analyses of the alloy, and, second, by its performance.
In seeking to establish infringement by proof that the analysis of Phoenix Metal (the .defendant’s product) was within the range of the patent analysis, the plaintiff fell into the identical fault in which the defendant fell in its effort to prove anticipation by analyses. There the plaintiff stoutly complained that the defendant treated the case as purely one of analyses; that it utterly ignored the physical characteristics of'the products, and thus failing to show that the products of those analyses were Adamite, failed to show anticipation. In this we have found the plaintiff was right. Now the plaintiff, in its endeavor to prove infringement, pursued precisely the same fatal course by treating that issue as one of analyses, utterly ignoring the physical characteristics of the alleged infringing products, and likewise failing to show that the products were Adamite.
The physical characteristics of true Adamite are many and pronounced. This we must believe, for the patentees say so. The range of analysis, however, is broad and uncertain, with corresponding uncertainties in the resultant product. Within a given range, plaintiff’s expert testified that he would “expect” “approximately” the product of Adamite, but it was also testified that within the lower and upper ranges of the claim, the lines of demarcation between Adamite and steel on the one hand and between Adamite and cast iron on the other, are hard to determine, because in one range the metal “shades off” into steel arid in the other it “shades off” into cast iron. Hence it is certain that some products within the broad range of the analysis of the patent will not be Adamite; while others may be Adamite not [713]*713with fixed hut with many variations of characteristics. The central question of infringement, therefore, is, whether the product of the defendant is Adamite within any of its variations.
For proof that Phoenix Metal is Adamite, the plaintiff relies, as we have said, first, upon approximate similarity of analyses, and second, upon approximate similarity of performance. The plaintiff has shown by its own witness that identity of the two metals cannot be established by proof even of identity of analyses; and it is very certain that similarity of performance does not establish identity of the metals. Phoenix Metal tnay perform as well as Adamite, yet be an altogether different metal. Evidence of similarity of performance, however, has probative value; this we have considered.
The plaintiff’s proof of performance goes to the lasting qualities of the two metals, due to similarity in their great strength. Yet, even when endeavoring to prove identity of metals by similarity of strength, the plaintiff showed a dissimilarity in the metals by showing a dissimilarity in their strength. It proved that the tensile strength of Adamite is 70,000 to 100,000 pounds per square inch, while that of the Phoenix Metal is 45,000 to 55,000 pounds, and that of ordinary soft steel is 30,000 to 40,000 pounds. From this it appears that the tensile strength of the defendant’s low chromium and high sulfur hardened product is but 50% greater than that of soft steel, while the tensile strength claimed for Adamite is about 200% greater. This radical difference in tensile strength indicates a radical difference either in the quantity or form of carbon or in both, and that the two metals are not the same.
But proof of identity of the two1 metals was not restricted by the nature of the metals to .evidence of similarity of chemical analyses and performance. It might have extended to their physical characteristics. These are such as to make them readily distinguishable, and it is to such that we naturally look in our effort to determine whether the two metals are the same or are different. Recognizing this, the defendant did not rely alone upon the difference between the analysis of its alloy and that of the patent in defending the charge of infringement, but proceeded to show by evidence of physical facts that Phoenix Metal is not Adamite. On the other hand the plaintiff produced no evidence of physical facts in support of its charge that Phoenix Metal is Adamite.
Stated briefly, the proofs are these: The patentees say in their specification:
“We have not found it feasible or possible to produce our alloy (Adamite) commercially in a cupola, or like furnace. * * * Also the constituent analyses as given for our alloy cannot he ohtamed from the aupóla or like furnace in the case of castings of larger section. In attempting to make an alloy in a cupola the percentages of silicon, sulfur, phosphorus, we have found too high to got uniform results. We have found that any such percentages of silicon, sulfur and phosphorus as are contained in the metal, the tests of which are reported in the Iron Age, as above stated, would he destructive to the remarkable results which we have obtained and described herein.”
The plaintiff made no attempt to prove that Phoenix Metal is made in. the mariner disclosed by the patent for the manufacture of Adam-ite, or that Adamite with high sulfur in alloy can be made in cupolas, [714]*714notwithstanding the statement of the specification to the contrary. The defendant, however, proved its method of manufacture, and showed that Phoenix Metal rolls,- — “castings of larger section,” — with their high sulfur content, are and have always been made in cupolas and are made in precisely the way the patentees state in their specification it is impossible to make Adamite. If Adamite cannot be made in cupolas, then certainly Phoeni x Metal made in cupolas is not Adamite.
The patentees claim for Adamite the qualities of cheap tool steel. Upon specific analyses they say:
“The results obtained in these cases were materials which were hard and resistant to the action of abrá sion under heat, and susceptible of being forged. * * *
“We believe we have made a discovery of a new composition of metals, which in its primary or cast st¡te, resembles cast iron in its action undercutting tools in the lathe, under the acetylene torch and in its failure to show elasticity or elongation or reduction of area over ordinary cast iron; whereas in its secondary or worked state ivs ultimate strength, its forging and wearing quallies class it with a steel product. * * *
“Our product is suitable for the greatest variety of uses. It may be used for castings, both large and small, where wear and resistance to heat are important factors. It may,also be used for tools, such as hammers, axes, picks, shovels, saws.”
The plaintiff produced no proof that Phoenix Metal is a tool steel metal suitable for the many uses or for any of the uses for which Adamite is stated by the pateni.ee to be suitable; that its action under cutting tools or under the acetylene torch is like that of Adamite; that like Adamite it lacks elasticity, elongation or reduction in area over ordinary cast iron; and most important of all, the plaintiff produced not the slightest proof '.hat Phoenix 'Metal is forgeable. The forgeability of Adamite is one of its distinctive features. Had Phoenix Metal possessed that quality, it could easily have been proved. That Phoenix Metal does not possess it is to be inferred not alone from the plaintiff’s silence but from the large sulfur content of the metal, the red shortness incident to such content, and the consequent difficulty or impracticability to forge metal with that property.
It was testified, .that while both metals show unusual strength, the tensile strength of Adamite is nearly if not quite double that of Phoenix Metal. There is no evidence that the alleged infringing rolls from which turnings were' taken and analyses made were of Adamite strength.
Rolls of the two metals are used at times in different stands for different purposes, according to differences in their characteristics. Rolls of Phoenix Metal are frequently used for finishing purposes, while Adamite rolls are" not usually so employed, because they are much rougher, in fact, rougher even than sand iron rolls. When so employed, Adamite rolls wear rough sooner than Phoenix Metal rolls. Rolls of Phoenix Metal finish up smoother and hold their surface better than rolls made of Adamite.
In cutting rolls made of the two metals, different lathes are required. In roughing off Phoenix Metal rolls a very high grade and high speed steel tool is required until the skin is removed, then they may be finished [715]*715by a common carbon steel tool, but in roughing off Adamite rolls the high grade high speed tool has to be used throughout.
The rolls from which turnings were taken and analyses made were not subjected to these distinguishing tests.
Testimony as to physical characteristics of the two metals was given by rolling mill men who had operated rolls made of both metals. That the two metals are readily distinguishable was testified by all. The general trend of their testimony is shown by the testimony of one witness given in the margin.1
In determining the issue of infringement, as raised and tried in this case, we are restricted to and controlled by the testimony produced in this case. How such an issue may be determined upon other evidence we venture no opinion. But from the testimony in this case, as we read it, it clearly appears, that Phoenix. Metal and Adamite, while having an approximate similarity of analysis and performance, possess physical characteristics which distinguish one from the other; that these characteristics are readily susceptible of proof; that the defendant produced testimony that the metal of its rolls possessed the distinctive physical characteristics of Phoenix Metal, and that the plaintiff produced no testimony to the contrary. While charging infringement because of the approach of the analysis of Phoenix Meta! to the analysis of Adamite, the plaintiff did not prove that the rolls from which turnings were taken and the analyses upon which it relies were made, had any of the physical characteristics of Adamite, in fact, the plaintiff failed to prove that Phoenix Metal is Adamite, and therefore failed to prove infringement.
The burden of proving infringement rests upon the party charging-[716]*716it. If the peculiar subject matter of the patent makes such proof difficult, that is its misfortune; it cannot operate against the party charged. When infringement is charged it must be proved. In this case, we think it has not been proved. Therefore, we find with the trial court upon this issue, and direct that
The decree below be affirmed.
BUFFINGTON, Circuit Judge concurred in the finding of the majority of the court that the patent in suit was valid, but dissented from the finding of the majority on the question of infringement.