[1] This is a suit for infringement of patent No. 578,868, issued to Hans Goldschmidt on March 16, 189?. and now owned by the plaintiff. The patent relates to a process of producing metals and alloys. The defendant is engaged in manufacturing ferro-vanadium, and it is claimed that the process which it employs infringes that of the patent. In order to understand the scope *82o£ the patent, as well as the defenses and the conclusions which I have reached, it is necessary that the history of the invention and the prior art be to some extent reviewed.

In order to fit metals for industrial use, it is necessary to remove from them the elements, or impurities with which they are generally combined in their natural state. Some metals may be industrially used with carbon present, and those have been purified or reduced for many years by means of a coke fire; but by this method a part of the carbon of the coke combines with the metal, and it was. unavailable for producing what are known as “refractory metals,” such as chromium, manganese, and vanadium, because they, in order to be useful for industrial purposes, must be produced free from carbon. It had been discovered some time before the patent in suit was applied for that metallic compounds containing oxygen, sulphur, or chlorine, such as the refractory metals before mentioned, could be reduced and the pure metal obtained hy mixing the metallic compound in a finely ground state with pulverized aluminium, and heating the mixture, which was placed in a crucible, to such a temperature that a chemical reaction took place therein, creating a very high temperature, said to be nearly 3,000° centigrade, which is far in excess of that necessary to start the reaction. When this high temperature is reached the mixture is melted, the impurities are separated from the metal with which they were combined and attach themselves to the aluminium, leaving the pure metal, which, being heavier than the molten aluminium slag, sinks to the bottom of the crucible, and, when cooled, solidifies.

This reaction or means of reducing such metals was discovered by Greene & Wahl prior to July, 1892, when they applied for a patent therefor in the United States (which was subsequently issued on January 3, 1893, as No. 489,303), and by Vautin prior to April 26, 1894, when he applied for a British patent (which was subsequently issued on April 26, 1895, as No. 8,306). It also appears to have been described in other patents and publications, in various forms, at other times prior to the date of the application for the patent in suit. Shortly after the Vautin patent was applied for, the fact that he had discovered this reaction was brought to the attention of the patentee, Goldschmidt (an eminent German chemist and metallurgist), who had theretofore been retained by the Krupp Steel Works of Germany to devise a means to manufacture, on a large scale, chromium and manganese free from carbon. He thereupon went to London, met Vautin, had the reaction explained to him, and saw one or more experiments. The outcome of this was that he and Vautin, in December, 1894, entered into an agreement for utilizing the latter’s discovery and mutually reaping some financial benefit therefrom. The specific method employed by Vautin for initiating the reaction was to place the mixture before mentioned in a suitable crucible and heat the same to the reaction temperature, either by placing the crucible in a coke furnace or by means of a blow pipe projected against the exterior of the crucible. In either case the method of bringing about the reaction temperature was by external heating. The same method was employed by Greene & Wahl. That method, however, according to the patentee, who was experimenting and who was necessarily financially interested in the success of *83what Vautin assumed to have discovered, and, as I think the evidence demonstrates, presented certain difficulties which rendered the process useless, or practically so, for manufacturing on a commercial scale. It was to overcome these difficulties and make the Vautin discovery a commercial possibility that Goldschmidt began the experiments which resulted in the patent in suit. He discovered that the reaction could be produced by heating a small portion of the mixture of the metallic compound and aluminium internally, and that this method largely, if not entirely, solved the difficulties which he sought to remedy. It is thus apparent that Goldschmidt did not discover, nor does he assume to have done so, the aluminium reaction, but only a method by which the reaction could be initiated. °

But such method, the plaintiff claims, is what has made the discovery of Greene '& Wahl and Vautin feasible in a commercial. sense. Goldschmidt’s invention was patented in Great Britain by Vautin in 1896 (the application having been filed in July, 1896, and the patent granted in June, 1897, as No. 16,685). Through an arrangement between Vautin and Goldschmidt it was taken out in the former’s name. That patent is substantially the same as the patent in suit. It was construed and held valid by Mr. Justice Warrington in Thermit, Ltd., v. Weldite, Ltd., 24 Patent Design and Trade-Mark Cases, 441. I quite agree with Mr. Justice Warrington, that the essence of Goldschmidt’s invention, as disclosed in the specification of the patent, was the method of initiating the reaction in a small portion of the mixture, as distinguished from the whole (as was substantially the case when the crucible was placed in a furnace), by bringing the means to produce the reaction directly in contact with the mixture, or, in other words, by internal heating. While it is true that the patentee states that he overcomes the deficiencies, which he mentions, “by causing the reaction to set in, not throughout the whole mass at a time, but only at one point or place of the mass,” it is equally true that, of the three methods mentioned in the specification by which this is done, two are incapable of operating in any way other than inside the crucible. One of the latter is not only stated to be the preferred method, but in practice is the one almost exclusively used. The other method is described as “by means of a blow flame directed against that portion” (a small portion of the mixture).

It will he noted that the words used are “against that portion,” not “against a portion of the crucible.” When the ordinary meaning of “against” is considered with the context, the conclusion is inevitable that the patentee meant that the flame of the blow pipe should be projected directly against the portion to be ignited, not against that portion indirectly through the crucible. The heating by external means of a small portion, as distinguished from substantially the whole, would have made no advance in the art, because undoubtedly scientists used interchangeably a blow pipe with a furnace, for experimental purposes at least. Vautin and others, including Goldschmidt in his subsequent experiments, had done so. But Vautin and others admit the novelty and utility of Goldschmidt’s method. Unquestionably, no one before Goldschmidt had discovered that the reaction could be started by internal heating.

*84[2] Viewing the invention in this light, the question arises whether the claim covers and protects it. The defendant contends that what the patentee claimed was the heating initially of a small portion of the mixture, rather than substantially the whole, and that to so construe the claim as to make internal heating an element also-, is to violate the settled rule that an element which is not present cannot be read into a claim (McCarty v. Lehigh Valley R. R. Co., 160 U. S. 110, 16 Sup. Ct. 240, 40 L. Ed. 358). The plaintiff, on the other hand, contends that internal heating'is actually an element of the claims. The question thus presented is important, as the decision as to whether the patent involves invention, as well as whether it exhibits novelty, may depend upon its solution. While it is true that" the claim refers to the heating of a small portion of the mixture, and does not, in so many words, refer to the doing so internally, it does limit the whole process therein mentioned by the words “as set forth.” The patentee specifically claimed the “heating a small portion of that mixture to initiate the reaction of said portion, as set forth.” The use of the words “as set forth” requires that the claim be construed in connection with the specification, and that there be imported into the former the particulars of the latter. Seymour v. Osborne, 11 Wall. 516, 547, 20 L. Ed. 33; Consolidated Roller-Mill Co. v. Walker, 138 U. S. 124, 133, 11 Sup. Ct. 292, 34 L. Ed. 920; Westinghouse v. Boyden Power Brake Co., 170 U. S. 537, 558, 18 Sup. Ct. 707, 42 L. Ed. 1136. The rfieans “set forth” in the specification for heating such portion is, as before found, to be applied internally.

The words “heating of a small portion” are themselves susceptible of three meanings, viz.: (1) The application of the heating means directly to the portion to be heated; or (2) indirectly through the crucible; or (3) both; but the way described in the specification is the first. Construing them, then, in connection with the specification, it follows that they must be held to mean heating directly, and therefore internally. In addition, every consideration calls for the adoption of that meaning. It is not conceivable that the patentee, knowing as he did that a small portion could be heated externally, and that such a method did not solve all of the difficulties which he was endeavoring to overcome, while the heating internally did, should have used those words in any other sense than that of direct heating.

[3] Also to construe them differently might invalidate the patent. In such cases, where more than one interpretation is possible, it is the general rule that the construction which “sustains and vitalizes” the patent should be adopted. National Hollow B. B. Co. v. Interchangeable B. B. Co., 106 Fed. 693, 715, 45 C. C. A. 544 (C. C. A. 8th Cir.), and cases there cited. The claim, I think, must therefore be construed to cover and be limited to the process of internally heating a small portion of the mixture to be reduced. This was also the construction given by Mr. Justice Warrington to similar claims of the British patent. Viewing the claim in this light the defendant’s contention—that Vautin exhibited to Goldschmidt the means of bringing about the re,action by a blow pipe directed against a portion of the crucible, and thus that Goldschmidt invented nothing—becomes immaterial, because, even if he did (which, considering the burden of proof, I would have *85great difficulty in finding), the initiation of the reaction in only a portion of the mixture, as distinguished from substantially the whole, is only a part of the invention of the patent. For the same reason, the articles of Roberts-Austin and Winkler become immaterial on the question of novelty.

Under the circumstances of this case I cannot conceive that I would be justified in holding that the patent did not involve invention. It may have been obvious that the heating of a small portion externally rather than the whole would prevent a too violent reaction; but that change alone would not have made manufacturing on a commercial scale feasible. As before shown, Goldschmidt did more than make that substitution. Professor Chandler very clearly and satisfactorily points out the reasons why internal heating was not obvious. In "addition, when it is considered that the aluminium reaction had been known for some time, that eminent scientists, such as Greene and Wahl in this country, and others abroad, struggled to make it a commercial success and failed, and that Goldschmidt has succeeded in doing so, it is clear, I think, that what Goldschmidt did was invention. Flis was the last step which “turned failure into a success,” and as such entitled him to a patent therefor. The Barbed Wire Patent, 143 U. S. 275 , 283, 12 Sup. Ct. 443, 36 L. Ed. 154; Cimiotti Unhairing Co. v. American Fur Refining Co., 198 U. S. 399, 25 Sup. Ct. 697, 49 L. Ed. 1100.

It may be, as defendant contends, that the preferred method of ignition mentioned in the patent, and now commonly referred to as the “Cherry method,” is the thing which made the use of the Vautin reaction possible for manufacturing on a commercial scale. But, as Goldschmidt invented that method, and as it is covered by the-claim of the patent, I fail to see how the fact that it was the cause of the success can in any way detract from the merit of the patent. In addition, the fact that the defendant uses internal rather than external heating, at the risk of infringement, is significant as to the superiority of the former process.

It remains to consider the question of infringement and one other question respecting the validity of the patent. The specification of the patent, I think, undoubtedly contemplated the placing of the mixture in a crucible and then directly or internally igniting a small portion thereof, and, as the reaction thus started is self-propagating, that it would quickly extend throughout the balance of the material in the crucible, and that thereafter, if it was desired to reduce more material than that contained in the crucible at the beginning, the mixture could he added as the reaction was taking place.

There is a conflict in the testimony as to the exact method pursued by the defendant. The plaintiff’s witnesses claim that the defendant places a heated crucible fthe temperature of which, the defendant’s witnesses state, is from 800 to 900° centigrade, but not sufficient, however, to start a reaction in the mixture) in a chamber; that there is then spread over the bottom of the crucible, from one to two inches of the mixture to be reduced; that there is then placed on the latter the ignition powder, composed of the same or equivalent substances *86as those mentioned in the patent, wrapped in paper; that the paper is then lighted, which in turn ignites the powder, and the reaction is thus started in the mixture which is on the bottom of the crucible; that then the additional mixture to be reduced is gradually fed, through mechanism located at the top of the chamber, into the crucible. The defendant admits that such is its process, except that its witnesses say that no reduction mixture is first placed in the crucible, and that the paper in which the ignition powder is contained takes fire from the heat of the crucible, thus igniting the powder, which, in a molten state, spreads over the bottom of the crucible, and that thereafter the reduction mixture is gradually fed into the crucible in the manner before mentioned.

I do not think that it makes any material difference which is the correct description. I am inclined to believe, however, that the plaintiff’s version is correct, especially in view of the fact that the defendant studiously limited its testimony in this respect to the period from August 1 to December 31, 1910, during which time it claims that metals were reduced for commercial purposes only twice. If its process had been the same at other times, it is difficult to understand, when the thoroughness with which its case was tried is considered, why no effort was made to show that fact, but that, on the other hand, care was evidently taken to limit the description to the period of time before mentioned, and this notwithstanding that on rebuttal one of the plaintiff’s witnesses (who was unavailable as a witness.when its direct testimony was taken) testified that during the period of time he was with the company, namely, from August, 1909, to July, 1914, the process followed was that which had been previously described by Mr. Beck.

[4, 5] The defendant contends that it does not infringe for two reasons: (1) Because neither the metallic compound nor the metallic aluminium which it uses is “finely pulverized,” as the claim of the patent provides; and (2) that the patent covers the treatment of a mass of material, while its process does not treat a mass, but quite the reverse. It is further urged that, as the patent does not define the precise meaning of the expression “finely pulverized,” it is invalid for want of sufficient description. As to 'the latter contention it must be observed that, as the patent was .for an improvement in the prior art, it did not attempt to describe (except incidentally so as to describe the improvement) the character or degree of fineness of the mixture in which the reaction of the prior art was to take place. The patent being addressed to those skilled in the art, and being merely for an improvement on the then existing art, into which improvement the character of the mixture did not enter, and having sufficiently described the improvement, all the requirements of the law were satisfied. Loom Co. v. Higgins, 105 U. S. 580, 585, 26 L. Ed. 1177. The claim, however, contains as an element that the metallic compound, as well as the. aluminium, shall be in a “finely pulverized” state. Undoubtedly, therefore, defendant can escape infringement if it does not use those materials in that state. To. ascertain exactly what is meant by the expression “finely pulverized,” resort must be had to the prior art, because it was in respect to that art that these words were used in the patent, and, of course, such art must be that of metallurgy.

*87The experts, unfortunately, disagree as to exactly what the meaning of these words were in the prior art. It would serve no useful purpose, I think, to review their testimony. The violence of the reaction which Goldschmidt was seeking to overcome was largely dependent upon the degree of the fineness of the mixture, and it would be unwarranted to assume that he intended to have the mixture of a greater degree of fineness than was necessary to start and continue the reaction. The degree of fineness was also effected by the character of the metals to be reduced. Vautin and others used granulated aluminium. When these facts are considered, it would seem to be unreasonable to assume, as defendant’s experts contend, that the words “finely pulverized” should be construed to mean what is commonly known as “dust.” Upon the whole I am constrained to find that the mixture which the defendant uses, samples of which are in evidence, is finely pulverized in the sense in which those words were used in the prior art, and hence in the patent.

On the remaining question regarding infringement the defendant’s theory is that the patent deals with a mass of mixture already in the crucible before the reaction is to be started, and that its process does not infringe because it does not deal with a mass in that sense. It is quite apparent, however, that the defendant has simply reversed the method outlined by Goldschmidt for initiating the reaction. The latter first placed the material to be reduced in the crucible and started the reaction on the top by an ignitor placed there, adding fresh material if necessary. The defendant places the ignitor in the bottom of the crucible and causes the mixture to be fed in from the top, gradually, of course, because, if all of the material were dumped in a mass on the ignitor, the latter would be smothered, and the reaction in the material would not take place.

[6] In determining whether this difference avoids infringement, two preliminary questions must, I think, be solved, namely: Whether Goldschmidt’s invention is to be considered merely as a variation of Vautin’s, or whether it is to be considered as a broad invention in itself; and, second, whether there is any substantial difference, in result or mode, between the two processes, because if there is not a mere reversal of form will not avoid infringement. Vautin, as well as others, had discovered the aluminium reaction. He and they adopted the conventional way of starting it. But unquestionably this method had its disadvantages, which were real and substantial. Goldschmidt discovered the internal ignition of a small portion, and thus the means of making the reaction commercially useful. True, he, providing a different method of initiating the reaction, made in one sense merely an improvement on Vautin; but in another sense he made a great step in advance. In the latter sense he is not to be considered as one who has made a mere improvement on an existing process, and thus entitled to a monopoly only in respect to the particular form which he described. The law affords him a range of equivalents commensurate with the degree of his invention. Paper Bag Patent Case, 210 U. S. 405, 415, 28 Sup. Ct. 148, 52 L. Ed. 1122.

I am unable to find any satisfactory evidence that the defendant’s process produces any different or better results than that described in *88the patent, or that it is more suitable to the production of ferrovanadium. As before found, the defendant mixes “finely pulverized” metallic compound with “finely pulverized” metallic aluminium. It undoubtedly heats “a small portion of that mixture to initiate the reaction of said portion,” and heats such portion internally, even using the preferred form of ignitor described in the patent, and that which has made the patented process a success.. The reaction, thus begun, transfers itself to the remaining parts of the mixture as the same are fed into the crucible, thus causing “a continuation of the process by the heat developed by said initial action.” The claim provides that, after heating the small portion, such portion shall not be brought “out of contact with the remaining main portion,” and that the “reaction then transfer itself to said remaining main portion.” The necessity of those elements is due to the fact that the reaction is self-propagating by contact alone, as no combustible gases are formed by the reaction. They were not a part of Goldschmidt’s improvement, hut were necessary elements of the reaction itself. In order that the reaction will continue from any part in which it has started to any other part, the latter part must be brought in contact with the former, and after the reaction has started must not be brought out of contact with it.

It makes no difference in either the mode of operation or result, so far as the evidence shows, when the portions are brought into contact with each other, provided that they are in that relation soon enough for reaction to continue. According to the process described in the patent, the mixture in the bottom of the crucible, when the reaction is started, is in one sense out of contact with that on the top where the reaction starts. The above expressions or elements of the claim must be construed in the light of these facts. There are two ways of continuing the reaction, once started, namely, by placing all of the material, in the first instance, in a crucible so that the reaction will spread throughout all of that material, as Goldschmidt’s patent described, or as the defendant does, and as the specification of the patent in a manner points out, by continually adding to the material in which the reaction has been hegun other material in which.it is desired that it should be continued. In either case the part in which it is desired that the reaction shall continue is, in the sense of the patent, not out of contact with that in which it has been initiated. In the defendant’s process the mixture in which the reaction first takes place is not any more out of contact with the remaining portion in the hopper than is the material in the bottom of the crucible, when the process described in the patent is used, out of contact with that wherein the reaction starts. Thus there is, I think, no difference in principle whatever between the two methods, and, consequently, that the defendant has used the essence of the invention, varying the process only in form. Hence I must hold that it has infringed the patent.

Before this case was argued, and, indéed, before the taking of the proofs was completed, the patent expired. Consequently the plaintiff is entitled only to a decree for an accounting. This it may have, with costs.