Opinion
No. 3194.
June 29, 1925.
Appeal from the District Court of the United States for the District of Delaware; Hugh M. Morris, Judge.
Suit in equity by the Peirce-Smith Converter Company against the United Verde Copper Company. Decree for complainant, and defendant appeals. Affirmed.
For opinion below, see 298 F. 763.
See, also, 293 F. 108.
Charles Neave, Edward L. Blackman, and Maxwell Barus, all of New York City, for appellant.
William H. Davis, Livingston Gifford, Merton W. Sage, and John F. Neary, all of New York City, for appellee.
Before BUFFINGTON and WOOLLEY, Circuit Judges, and BODINE, District Judge.
This is a suit for infringement of Letters Patent No. 943,280 granted to Elias A.C. Smith in 1909 and afterward assigned to the plaintiff. Claims 1, 2 and 3 are in issue. The District Court found these claims valid and infringed and awarded an injunction and accounting. Peirce-Smith Converter Co. v. United Verde Copper Co. (D.C.) 293 F. 108; Id. (D.C.) 298 F. 763. The case is here on the respondent's appeal.
The invention of the Smith patent is a process for bessemerizing copper matte in a converter having a non-corrodible lining. As no one can have an approximate understanding of the purpose and position of the invention without some knowledge of the art in which it stands, we shall describe the art very briefly and state one of its problems.
Copper matte is the product of smelting copper ore in a furnace and consists almost entirely of a mixture of iron sulphide and copper sulphide. It requires further treatment to break up and remove the iron sulphide (FeS) and then convert the remaining copper sulphide (Cu2S), which is called "white metal," to metallic copper. This treatment is known as bessemerizing or "converting" copper matte. Copper matte, therefore, is in a sense an "intermediate."
Bessemerizing is carried on in a large vessel — a converter — into which the molten matte and a quantity of silicious flux are placed and involves blowing air in large volumes through the molten matte so as to burn out or oxidize the sulphur and iron. The oxygen of the air unites with the sulphur in the copper sulphide and in the iron sulphide of the matte to form oxidized sulphur which is a gas (SO2) and readily goes off in that form. But the oxygen of the air also unites with the iron in the iron sulphide of the matte, thus forming oxidized iron (FeO) which is not a gas and is more troublesome to get rid of. In order to remove it, a silicious or so-called "acid" flux is used, the silica (Si2O2) of which unites with the oxidized iron (FeO) to form a liquid ferrous silicate slag that can from time to time be "skimmed" off the surface of the bath or can be poured off by tipping the converter. In this way the iron is extracted.
That part of the silicious flux which is not silica forms no chemical combination with the converter content. It is chemically inert. It absorbs heat, is melted and such matte-forming constituents as it may happen to have, such as copper, iron, gold or silver, are added to the matte. Its earthy constituents, such as alumina, lime and magnesia, are carried off with the slag.
The bessemerizing operation proceeds in two stages — the first, the slagging stage, during which the silicious flux is added and the iron and part of the sulphur are eliminated from the matte, leaving copper sulphide "white metal"; and the second, or finishing stage, in which by further blowing in of air without flux the elimination of the residue of sulphur from the matte is completed, leaving "blister" copper. The claims in issue of the patent in suit deal only with the slagging stage.
The converter in which these chemical or metallurgical reactions take place is a large vessel, and, in order to provide the requisite structural strength and to admit of the requisite tipping operation, it is made of steel. But it was known that the slag of the bath when in contact with the steel speedily attacks and destroys it. So, from the very beginning, a lining was placed inside of the shell. Three types of such lining have been known and used. As their place in the art has an important bearing on both the issues of validity and infringement in this suit, we shall not describe them here, but shall speak of them later and in the order in which they appeared in the art. We shall now only indicate that in each type a lining of magnesite brick, which is incorrodible in the sense of not being as readily attacked by the molten contents of the converter as other brick, is laid against the inner wall of the steel shell.
In the first type of lining, used for a long period just prior to the invention in suit, there was a thick coating of silicious, or so-called "acid," material placed on the face of the magnesite brick and held there by plastered clay. The molten contents of the converter came into contact with the silicious coating which, together with additional silicious material thrown in through the converter mouth during the operation, furnished the silica to unite with the liberated iron of the matte in forming the fluid slag. The acid coating, as intended, was readily consumed during the slagging stage of the converting operation — carried on at 2200° to 2350° F. — and had to be renewed at much expense and delay after converter operations of only a few hours and after a copper production of only about fifty tons. Taking their name from the acid coating of the magnesite lining, converters of this type were known as "acid" converters. Except for certain basic converters alleged as anticipations and basic converters used experimentally, acid converters constituted the entire prior art of the patent in suit. It had long been realized by workers in the copper smelting art that an acid converter was wasteful and costly in operation and that if a true non-corrodible basic converter could be made (that is, one lined only with magnesite brick and where all the flux would be supplied through the mouth of the converter) and if such a converter could be made to stand up and endure, great economies would be effected and increased production obtained. But the trouble was that an unprotected basic lining yielded quickly to the attack of the molten mass within. Almost continuous attempts for nearly thirty years had been made to bessemerize copper matte in basic lined converters, yet entirely without success. Smith, the patentee, in collaboration with Peirce, began work on the matter in 1901. At first they regarded the problem as mechanical rather than chemical and overcame certain mechanical difficulties incident to basic-lined converters by improvements described in the patent awarded them in December, 1909 (No. 943,346). But they found that notwithstanding these mechanical improvements the basic lining yielded to attack by the slag very much as before. Smith then directed his attention to the metallurgical aspect of the problem and in the course of time made discoveries from which he conceived the invention of the patent in suit.
Smith did not try to revolutionize or even to change the prevailing bessemerizing process. He took the process as he found it and sought to practice it in a way that would permit the use of an unprotected basic lining in a copper converter and save it from immediate destruction. In doing this he was confronted by certain fixed factors. He knew that one of the purposes of the converter operation is to eliminate the iron from the matte; that iron must be eliminated and run off in the form of slag; that in order to get the iron in that form, silica must be employed to combine with the iron; that, as the use of pure silica is not economical, the requisite silica for that purpose is obtained from some cheap and available material known to contain it; that silica in such material is always combined with other ingredients, some of which may be value-bearing and most of which are not value-bearing; and that this combination material, called a "flux" and which is used only for its silica content, contains silica in varying proportions. He found that the varying proportions of silica in the flux affected in varying ways the character and the temperature of the ferrous silicate slag in the converter and he discovered that to insure an effective operation — both chemical and mechanical — a thin and fluent slag must be obtained, and that a thin and fluent slag of a given chemical composition could be produced either at high temperature or at low temperature, and that when produced at high temperature it would promptly attack and destroy the basic lining, and when produced at low temperature it would not substantially attack or destroy the basic lining. This was his capital discovery. It was a phenomenon then unknown and still unexplained. It showed clearly that not the slag alone but a slag above a certain temperature was the thing which all along was the enemy of the basic lining. Yet this naked discovery did not solve the difficulty nor, standing by itself, did it amount to invention. Miami Copper Co. v. Minerals Separation, 244 F. 752, 755, 157 C.C.A. 200. Smith grasped the principle of his discovery and sought to develop it into invention by finding and disclosing a medium or means which would bring it into practical action. Morton v. New York Eye Infirmary, 5 Blatchf. 116, Fed. Cas. No. 9,865. And this he did by so proportioning the amount and composition of the silicious flux to the volume of the air blast admitted during the smelting blow as to produce a thin and fluent slag at the low temperature which he had discovered would not substantially attack the lining.
In estimating the amount and composition of the flux he was without doubt dealing with variables, yet with variables which were susceptible of fairly accurate ascertainment. The flux, en masse, affected directly, though temporarily, the cooling of the molten bath. The composition of the flux also directly affected the temperature of the bath. What he did was to vary its composition so that it should contain less silica and correspondingly more inert heat-absorbing non-silicious material and thereby control the temperature of the bath and correspondingly lower the temperature of the slag. The customary way of controlling the temperature of a furnace is by increasing and decreasing the fuel and the air blast, but in operations of this kind the air blast, either for mechanical or economical reasons, is maintained steady and may be regarded as a fixed factor, impractical or unprofitable to vary for the purpose of regulating heat. So Smith left the air blast alone and sought to regulate the temperature of the converter bath, and thereby the character of the slag, through the media of amount and composition of the flux which is fed into the molten bath; that is, he sought to control the temperature of the bath by a co-ordination of the elements of fixed air volume and variable amount and composition of the flux. He also found that when the silica content of the flux is decreased over much and the content of inert material is correspondingly increased, the bath tends to become cold or "freeze," and that this difficulty can be corrected by again proportioning the amount and composition of the flux by increasing the silica content and decreasing the inert material. And so throughout an operation the temperature of the bath and the quality of the slag can be kept below the danger point, whereby an unprotected basic lining — the second type of lining — can be safely and economically used. This constitutes his invention. He disclosed it in the patent in suit. A typical claim reads as follows:
"I. The method of bessemerizing copper matte in a converter having a non-corrodible (non-acid) lining with the employment of an acid-flux, which consists in forming a molten bath of matte in converter of such volume as to retain its fluidity as against losses by radiation during the blow, and so proportioning the amount and composition of the flux to the volume of air blast admitted that at the termination of the blow there will result a thin and fluent slag at a temperature insufficient to substantially attack the lining, substantially as described."
The description in the specification covers the temperature and gives illustrative proportions of the amount and composition of the flux. The invention is addressed to ores of different characters and to available fluxes of variable silica content. In view of the inherently variable problems, the claims and specifications are, we think, definite enough to comply with the requirements of section 4888, R.S. (Comp. St. § 9432).
Did this amount to invention? The respondent says it did not; and, moreover, if it did, the respondent says that Smith is not entitled to a patent because it had already been conceived and practiced by others. The anticipations relied upon are, inter alia, the patent granted Baggaley and Allen in 1904 (No. 766,654) and the Baggaley and Heywood operation at the plant of the Pittsburgh and Montana Copper Company at Butte, Montana, in 1905 and 1906. The Baggaley and Allen patent, now expired, covered a method of recovering values from ores by dissolving in molten baths. It was not addressed to the problem of saving basic linings in copper converters, nor, apparently, was that problem in the minds of the inventors, although Baggaley had recognized the advantage of a basic converter and the importance of keeping its magnesite lining cool and tried to do it by the use of water jackets. True, the method of the Baggaley and Allen patent for recovering values included the molten bath of matte, blowing air into it, adding ore relatively high in silica, fluxing the iron of the bath and "replenishing the bath with material relatively low in silica and high in matte-forming compounds," the latter elements being at times, possibly, within the proportions and the composition of the flux of the patent in suit. The effect of a flux "low in silica" was inevitably to lower the temperature of the bath and correspondingly to improve the quality of the slag. But if it ever protected the basic lining (which was by no means proved) it was accidental. The patentees did not know it; or, knowing it, they did not tell it to the art. It is only in the light of what Smith afterward discovered, and did, that the disclosures of the Baggaley and Allen patent have any application to Smith's problem. But the important thing is that, if, in the light of Smith, the Baggaley and Allen method can be used to protect a basic lining, Baggaley and Allen did not disclose how it can be done, that is, they did not suggest even remotely the idea of Smith so to proportion "the amount and composition of the flux to the volume of air * * * that * * * there would result a thin and fluent slag at a temperature insufficient to substantially attack the lining."
Coming to the Baggaley and Heywood operation at the Pittsmont Mine, these experienced engineers had the problem of saving the basic lining of a very large converter which they were trying to operate and with which they were having many troubles. They attempted to solve it by endeavoring to keep the bricks cool from the outside; that is, they attacked the problem externally instead of internally. When Heywood arrived at the mine, however, he conceived the idea of keeping the converter lining cool by keeping the bath cool and this he attempted by adding cool materials to the bath. In this he approached Smith's idea. He told the men "to add cool ore or to turn it down or add cool slag or whatever cool material was on the floor"; he also told them to put into the converter "from time to time all the cold silica ore it would stand without freezing up or getting a slag so thick it could not be poured off." He said nothing about the chemical composition of the flux, although it is true that the silica content of the flux which was used at that plant was known. He dealt only with amounts and then only to the extent that the result would not be to freeze the bath and stiffen the slag. His procedure was such as a householder follows when he puts ashes — sometimes coal — on the firebed of an overheated domestic furnace, in quantity sufficient to cool it off yet not sufficient to put it out. If in this operation Heywood's workmen at any time hit upon the amount and composition of the flux required by the Smith process and attained its result, it was purely accidental and was without profit to the art and without value as an anticipation. Pittsburgh Iron Steel Co. v. Seaman-Sleeth Co., 248 F. 705, 709, 160 C.C.A. 605. Certainly it taught the art nothing about the composition of flux in producing a slag which would not destroy a basic lining.
Being of opinion that the process of the Smith patent was not anticipated, we come to the question whether the process amounts to invention.
Smith did something. The thing he did was based on observation and discovery. He observed that the magnesite lining of a copper converter was at times attacked and destroyed by the slag and at other times it was not attacked by the slag. Therefore he knew that, for some reason then unknown, slag would at one time attack the lining and at another time would not attack it. Smith moved from these observed facts to the discovery that the slag attacked the lining when at high temperature and that it did not attack it when at relatively low temperature. Having made this discovery he was the first to show the art the difference between a vicious slag and an innocent slag and he proceeded to embody this discovery in an invention which avoided one and obtained the other. And this he did by regulating the temperature of the slag in a way that has proved entirely practicable. Here was inventive discovery which involved the intelligent comprehension of relations not before recognized, although actually existing, followed by the conception of how they can be practically utilized. Eck v. Kutz (C.C.) 132 F. 758, 779. Smith's process, later disclosed in his patent, was not a temperature process. The problem was a temperature problem, and its solution was by a method to control slag temperature. Regarding the air blast as a fixed factor Smith gave the art variable factors of the amount and composition of the flux to be applied to the fixed factor of air blast as a way to produce slag at a temperature which will not attack the magnesite lining.
Before the invention of the patent in suit slag was regarded generally as a vicious element in the process of converting copper. It was not known that there was good slag and bad slag, for until Smith appeared it was not known that slag had two qualities under different conditions. Before that time slag was regarded as an element with one bad habit of attacking and destroying acid and basic linings. Smith was the first to show that slag had two habits, or that it was of different kinds, innocent and vicious, and he was the first to show how one could be obtained and the other avoided, and, in consequence, how a basic lining could be preserved through greatly increased length of operation.
We think this involves patentable novelty. While dealing with old elements in the sense of old ingredients, it involves new results from their new proportions, embracing new qualities of utility. Bethlehem Steel Co. v. Churchward International Steel Co. (C.C.A.) 268 F. 361, 364. Moreover, were the question of invention closer than it seems to be, the position which the invention promptly took in the art would aid in deciding it, for it is recognized that the fact that an art has long presented a problem and that the process of a patent has solved it, as well as the fact that the process has gone into large general use and has produced new and economical results, speak for its inventive character. Thropp Sons Co. v. De Laski Thropp Circular Woven Tire Co., 226 F. 941, 947, 141 C.C.A. 545; Globe Knitting Works v. Segal, 248 F. 495, 498, 160 C.C.A. 505.
Immediately upon the granting of the patent, the invention took a high place in the art. Smith made his invention in the early part of 1909 and introduced it commercially during the same year. At that time the production of copper by conversion from copper matte was not less than 1,000,000,000 pounds annually in the United States and Mexico. Aside from the amounts produced by Smith in his research work and by Baggaley and Heywood in their labors at the Pittsmont Mine, every pound was treated in acid-lined converters. This had been the universal practice since the introduction of copper converting as the result of Mahnes' work in 1880. So rapidly did the art take up the invention that in the year 1912 more than 85 per cent. of the copper converted in the United States and Mexico was converted in basic-lined converters by companies licensed under the patent in suit. Today every pound of copper converted throughout the world is converted in basic-lined converters. When placed in the scale, these matters weigh heavily in favor of invention. The claims in suit are valid.
Because of the metallurgy of the case we at first found the issue of infringement more difficult to decide than the issue of validity. We therefore called for, and have been aided by, a second argument on infringement.
Greatly compressed, the story of this issue is as follows: The Anaconda Copper Company, with plants at Anaconda and Great Falls, was one of the largest licensees under the patent in suit. Wheeler and Krejci were then in its employ at the Great Falls Smelter. In 1906 they attempted to convert copper in a basic-lined converter and failed. In 1910, Wheeler went to Garfield and Salt Lake to observe the basic converters there in operation under the Smith process. At Garfield, Wheeler inspected the operations, fully discussed them with Smith, and had in hand the reports of others who had inspected the operations and the report of Mathewson, who was in charge of the Anaconda plant. In 1911, Wheeler and Krejci started basic converting operations of their own and, later, were awarded a patent for their process. Negotiations with the plaintiff for a license under the Smith patent, begun by the defendant in 1909, fell through, and in March, 1911, the defendant abandoned the use of acid-lined converters and installed basic-lined converters and operated them under the process of the Wheeler and Krejci patent No. 1,068,470.
Taylor, smelter superintendent of the defendant's plant, in obedience to directions, visited the Garfield smelter and the Anaconda smelter for the purpose of observing the working of the Smith process, and, sometime after his return, the defendant operated basic-lined converters on the basis of what Taylor had learned there. Many difficulties ensued. In 1913, Taylor asked a Mr. McGregor, who was on his way to the copper region, to look around and send him information about operating basic-lined converters. Under date of December 3, 1913, McGregor wrote Taylor a letter in which he described in detail the method of handling the converters at Great Falls (under the Smith process), including the method of putting a magnetite coating on the magnesite lining. This letter told the defendant two things: One, how a magnetite coating was initially placed upon a magnesite lining, and, the other, how otherwise the Smith process was practiced.
In his patent Smith said nothing about a magnetite coating upon a magnesite lining — the third type of converter appearing in the art — and as it came up for the first time on the issue of infringement we must pause to explain, very briefly, its metallurgical reactions.
In the bessemerizing process, which we have described, the iron of the matte is first oxidized to ferrous oxide (FeO), and, if there be silica present to unite with each particle of ferrous oxide as it is generated during the blow, a ferrous silica slag is formed. But if there be no silica present, or insufficient silica present, to unite with all the ferrous oxide that is formed during the blow, then the free ferrous oxide (FeO) is further oxidized by the blast to magnetic oxide of iron (Fe3O4), or magnetite, as it is called. Once formed, the magnetite will not unite with silica or be chemically reacted upon by the slag.
This magnetite is, however, very soluble in a ferrous silicate slag, just as salt is soluble in water. So the magnetite as it is formed goes into solution in any slag that may be present. The hotter the slag is, the more magnetite it will dissolve, just as the hotter the water, the more salt it will dissolve. Consequently, if, after the slag present has dissolved all the magnetite it will hold, more magnetite is formed, it will crystallize out of the solution upon the relatively cool walls of the converter. When thus crystallized on the walls the magnetite forms a coating upon the magnesite lining and the thickness of the magnetite coating so deposited is subject to control by two factors — the amount of silica and the temperature. By increasing either or both factors, some or all of the magnetite coating may be dissolved off the walls. By decreasing either or both factors, magnetite may be deposited. This magnetite reaction was known to the art and advantage of its protective quality was taken by some of Smith's licensees, notably the Anaconda Copper Company, and also by Wheeler and Krejci, who incorporated it in their patented process, and finally by the defendant who employed it as an element which it says exonerates it from the charge of infringement.
Reduced to narrow terms, the defendant's practice was this: When starting a converter newly lined with magnesite brick — basic lining — it blew the first charge without any silica at all. Thereby the iron in the bath was converted into magnetic oxide of iron and was liberated. The iron thus liberated immediately deposited itself on the walls of the magnesite brick in the form of a coating. The blow was continued until the coating had reached the thickness desired. Then the Smith process was taken up and thereafter practiced throughout the operation. Did this amount to infringement?
We shall not express our views on the validity of the Wheeler and Krejci patent. We are concerned only with what the defendant did and whether it infringed the Smith patent. There would be no doubt of infringement but for the defendant's use of a magnetite coating upon a magnesite lining. About such a coating, however, the Smith patent is silent. While Smith may have known as much about magnetite as other workers in the art, he disclosed little knowledge of it in respect to the operation to which he directed the invention of his patent, and the little he knew inclined him to avoid it. Certain it is that his invention was not directed to protecting the magnesite lining with a magnetite coating. Yet it is equally certain that in the practice of his patented process such a magnetite coating at different temperatures and bath compositions develops and is used generally by his licensees as a temperature indicator and by some of them as a means to protect the magnesite lining. The defendant also learned of the development of magnetite in the slagging stage of bessemerizing copper and learned from Smith's licensees that it might be advantageously used as a protection to the magnesite brick when placed between the brick and the assailing slag. The defendant, like those from whom it gained the information, first heated the converter without the use of silica and thus started the operation by a process which was not Smith's. Having once obtained the magnetite coating on the magnesite lining, it then shifted to the Smith process and practiced it faithfully. But the defendant says that, in doing this, it was protecting the newly established magnetite coating, not the underlying magnesite lining, to which alone Smith's invention is directed. Yet it appears by the evidence that from time to time the magnetite coating will dissolve under temperatures and slag conditions which affect magnetite before they affect magnesite. In consequence, there are times when the magnetite coating will disappear and the magnesite lining will stand bare and exposed to the slag. During these times Smith's process comes into full play and performs its function of protecting the exposed magnesite bricks. Thus the defendant used Smith's process initially to protect the newly developed magnetite coating and finally to protect the costly magnesite lining when the coating dissolved.
Some one other than Smith may have first conceived the advantage of a magnetite coating on a magnesite lining, that is, the advantage of having a magnetite coating stand between the vulnerable lining and the vicious slag. It is possible (assuming that the Smith process is not directed to the magnetite coating) that the defendant may not have infringed the patent if it had used the process only to protect the magnetite coating which was standing on guard, as it were, against attacks by the slag upon the lining. In other words, a defense in such case may conceivably be based on the distinction between mediate and immediate protection. (On this we express no opinion.) But the defendant did more than that. While using the process of the patent to protect the coating, it always used the process to protect the lining when in the variations of heat and flux content the coating sloughed off and the lining became exposed. Here was the thing to which Smith's process is directed, and here, when the occasions arose, as they repeatedly did, the defendant used Smith's process for the purpose and in the way he had taught the art. We find infringement.
Coming finally to the defense of laches, we cannot hold the action is barred on that ground. Although the plaintiff delayed four or five years in bringing suit, the law of Window Glass Machine Co. et al. v. Pittsburgh Plate Glass Co. (C.C.A.) 284 F. 645, announced on the peculiar facts of that case, is not applicable to the very different facts of this case.
The decree below is affirmed.