Union Carbide & Carbon Corp. v. Graver Tank & Mfg. Co., Inc.

MAJOR, Chief Judge. This is an appeal from a judgment, entered December 10, 1951, by Judge Charles A. Dewey, sitting by assignment in the United States District Court for the Northern District of Indiana, Hammond Division, adjudging defendants in contempt of court for violating an injunction issued by Judge Luther M. Swygert July 10, 1950. The proceeding which led to the entry of the judgment in controversy was an aftermath of litigation involving United States Patent No. 2,043,960, relating to a process or method of electric welding' and to compositions for use in such process. In the original suit, sixteen process and seven composition claims were relied upon. All of the process claims, as well as composition claims 24, 26 and 27, were held invalid. Composition claims 18, 20, 22 and 23 were held valid and infringed. Linde Air Products Co. v. Graver Tank & Mfg. Co., D. C., 86 F.Supp. 191. This holding of validity and infringement as to these four claims was, on appeal, affirmed by this court. 7 Cir., 167 F.2d 531. At the same time, this court held valid all process claims and composition claims 24, 26 and 27. On certiorari, the Supreme Court affirmed Judge Swygert in all particulars and reversed this court insofar as its decision was inconsistent with that of Judge Swygert. 336 U.S. 271, 69 S.Ct. 535, 93 L.Ed. 672 (afterward referred to as the first decision of the Supreme Court). The Supreme Court granted rehearing, limited to the question of in *105 .fringement of the four valid composition claims, and again sustained Judge Swygert’s finding of infringement as to defendants’ composition 660, by application of the doctrine of equivalents. 339 U.S. 605, 70 S.Ct. 854, 94 L.Ed. 1097 (afterward referred to as the second decision of the Supreme Court). Thereupon, plaintiff disclaimed all the process claims and composition claims which had been held invalid. Subsequent to the second decision of the Supreme Court and on July 10, 1950, Judge Swygert, pursuant to the mandate of that court, issued a permanent injunction enjoining defendants “From making, using, advertising or selling Lincolnweld 660 flux or any other welding composition or flux covered by any of Claims 18, 20, 22, and 23 of said Patent No. 2,043,960, or offering so to do, aiding or abetting or in any way inducing infringement of said Claims * Of the four claims held valid 'by Judge Swygert and sustained by the Supreme Court, for the infringement of which defendants have been held in contempt by Judge Dewey, plaintiff says, and we think correctly, that claim 23 is typical. It reads: “23. A finely-divided unbonded fusible composition for use in electric welding employing a bare metal electrode as a source of weld metal, said composition being substantially free from substances capable of evolving a detrimental quantity of gas under welding conditions and containing calcium fluoride and a major proportion of one or more alkaline earth metal silicates." (Italics here and subsequently in this opinion ours, unless otherwise noted.) Defendants continued the manufacture and sale of 660 flux until July 10, 1950. In the meantime and during the course of the litigation, defendants commenced the manufacture and sale of its composition fluxes numbered 770, 780 and 760 (sometimes referred to as the “700 series” Lincoln fluxes). Plaintiff’s contempt motion, filed in December 1950, named only the “700 series” fluxes. Early in 1951, defendants introduced another composition, 840. Thereupon, the court gave plaintiff leave to amend its motion for contempt so as to include this new number. The controversy here revolves around the effect given to the requirement of the claims that the composition contain “a major proportion of one or more alkaline earth metal silicates.” In the former litigation, the controversy was confined to the words, “one or more alkaline earth metal silicates,”, while instantly it is more specifically directed to the words, “a major proportion.” Defendants’ 660, adjudged to infringe, consisted of 95% or more of manganese silicate, while plaintiff’s Unionmelt 20 consisted of slightly less than 95% of alkaline earth metal silicates. The Supreme Court in its second decision held that manganese silicate was the equivalent of an alkaline earth metal silicate and, applying the doctrine of equivalents, found infringement. In the compositions of both plaintiff and defendants there before the court, the ingredients were component parts and had been fully prefused, with complete chemical reaction. That this dual action was taught by the patent there can be no doubt. Whether it was required is a question, as will subsequently appear, of much importance in the instant case. Plaintiff states that the appeal involves a single contested issue, that is, whether the trial court was clearly in error in finding that the defendants violated the writ of permanent injunction by the manufacture, sale and use of the accused fluxes. Obviously, that is the ultimate issue, but its resolution is dependent upon numerous subsidiary issues raised by the defendants. They contend, (1) that none of the accused fluxes contain “a major proportion of alkaline earth metal silicates,” as called for by the claims; (2) the specifications demand that “the chemical reaction between the components of the welding composition must be completed before it is used in welding,” i. e., that silicates be formed therefrom; (3) broader claims were rejected and the limitations to “a major proportion” and to “silicates” as distinguished from their component ingredients introduced to avoid rejection; (4) the limitations were relied on both in the Patent Office and in the courts to distinguish the claimed flux from the prior art, and (5) the distinction expressed *106 in these claim limitations and the specifications has become res adjudicata in this case. Other -contested issues raised by the defendants may or may not become material, dependent upon a decision of those which we have stated. The prior electric welding art, the nature of the Jones invention, the reason for sustaining the validity of the four composition claims now in dispute, as well as that for invalidating three composition and all process claims, are shown in detail in the opinion and findings of Judge Swygert, by the opinion of this court and by the opinions of the Supreme Court, and we need indulge in no detailed repetition. It may be noted, and we think this is without dispute, that plaintiff’s flux, with its “alkaline earth metal silicates,” as well as defendants’ 660 flux, held to infringe by use of its manganese silicate, adjudicated to be an equivalent, are both completely reacted and, when fused, consist of chips of homogenous, glassy material described by the patent as “Characterized by vitreous fracture” (i. e., broken glass). On the other hand, the accused fluxes consist of non-homogenous granules composed of particles of the different components held together by a binder but visibly retaining their individual characters. This difference is due to the fact that the chemical reaction between the components of plaintiff’s welding composition are complete before use, while in the accused compositions the chemical reaction is not complete before use. True, there is a dispute; in fact, defendants contend that there is no chemical reaction between the components of the accused fluxes, while plaintiff contends that at least partial reaction takes place. Defendants contend that' they do nothing more than heat the raw materials in a gas oven for the purpose of driving out water vapor and other gases, and not for the purpose or with the result of producing chemical reaction among the components. We think it is not in dispute but that the chemical reaction is essential to produce the “silicates” called for in the claims. 1 Defendants’ process by which their compositions are produced requires heating for short time to a maximum temperature of from 1400° F. to 1800° F., while plaintiff’s process by which its fluxes are produced requires heating for a longer period at a temperature of 2800° F. This difference in the temperature accounts for the fact that plaintiff’s fluxes are completely chemically reacted prior to use in the welding process while those of the defendants experience no such chemical reaction or, at any rate, only partial reaction. More specifically, plaintiff’s product when manufactured and sold is completely reacted from a chemical standpoint, while defendants’ product is not. At most, and this is doubtful, it has been only partially reacted, and the completion of the reaction takes place, if at all, at the time of and in connection with the welding process. Plaintiff devotes more energy than necessary, we think, in emphasizing the importance of its invention. Its merits have been sufficiently extolled by the courts before which it has been litigated and it requires no further testimonial. And we think some of the confusion in the case arises from the failure to distinguish between plaintiff’s composition as described in the four claims in suit and the process employed in the utilization of its composition. It must be kept in mind that all of its process claims have been declared invalid in view of the prior art. As Judge Swygert stated, D.C., 86 F.Supp. 191, 195, “The mechanical operational steps involved are identical witht those used by Robinoff [No. 1,782,316]. Indeed, the patentees who started with the teachings of Robinoff merely sought an improved welding composition or flux to be used with his operational method.” In fact, it appears undisputed but that both plaintiff and the defendants use the process or operation taught by Robinoff, as they each have a right to do. Plaintiff, naturally and rightly, relies heavily upon the two decisions of the Su *107 preme Court, the first of which, 336 U.S. 271, 69 S.Ct. 535, 93 L.Ed. 672, held that Rule 52 (a) of the Federal Rules of Civil Procedure, 28 U.S.C., is controlling and that reversal is permissible only where the findings of the trial court are “clearly erroneous,” and the court sustained the findings of the District Court both upon the issues of validity and invalidity of the claims in suit. The court pointed out, referring to the District Judge, 336 U.S. at page 274, 69 S.Ct. at page 537, “He concluded that what was really invented' was that which was claimed and bounded by the composition claims Nos. 18, 20, 22, and 23.” The court also pointed out, 336 U.S. at page 277, 69 S.Ct. at page 538, “We have frequently held that it is the claim which measures the grant to the patentee.” (Citing cases.) We need not stop to inquire to what extent, if any, the Supreme Court has modified or retracted its view as to the applicability of Rule 52(a) in patent litigation (but see Great Atlantic & Pacific Tea Co. v. Supermarket Equipment Corp., 340 U.S. 147, 71 S.Ct. 127, 95 L.Ed. 162). This is so for the reason that we think the issue for decision is legal rather than factual and, furthermore, we are not so much concerned with the findings made by Judge Dewey as we are with those which he failed to make and on one important issue (subsequently discussed) which he refused to make. In the second Supreme Court decision, 339 U.S. 605, 70 S.Ct. 854, 94 L.Ed. 1097, notwithstanding the court’s sweeping denunciation of one who practices a fraud upon a patent and the application of the doctrine of equivalents to the facts before it, there is little, if anything, in the opinion which affords an answer to our instant problem. We note, however, that the court again stressed the weight to be given a finding by the District Court on the issue before it and stated, 339 U.S. at page 609, 70 S.Ct. at page 857, “A finding of equivalence is a determination of fact.” Plaintiff goes so far as to suggest that defendants’ argument here is little more than a rehashing of that made to the Supreme Court. We do not so understand. There, the court held that the substitution by the defendants in their flux 660 of manganese silicate was the equivalent of the “earth metal silicates” called for in the claims. It should be pointed out, however, and this is important, that there was no contention in that case that the patentee was estopped by its previous representations and activities from claiming that manganese silicate was the equivalent of the earth metal silicate called for in the claims. In any event, we do not understand that the Supreme Court by its decision intended to retralct or modify the doctrine announced in Exhibit Supply Co. v. Ace Patents Corp., 315 U.S. 126, 62 S.Ct. 513, 86 L.Ed. 736. There, the court had before it a patent on a gaming device commonly referred to as a pin ball game. This court, 7 Cir., 119 F.2d 349, had held the patent valid and infringed. The claim called for a certain conductor means “imbedded in” the table. Previously a claim which required that the conductor means be “carried by” the table had been disallowed in the Patent Office. The Supreme Court, as in its second decision in the instant litigation, took the case solely on the issue of infringement, found non-infringement and reversed this 'court. In doing so the court stated 315 U.S. at page 136, 62 S.Ct. at page 518: “Whatever may be the appropriate scope and application of the doctrine of equivalents, where a claim is allowed without a restrictive amendment, it has long been settled that recourse may not be had to that doctrine to recapture claims which the patentee has surrendered by amendment.” The court further stated: “By the amendment, he recognized and emphasized the difference between the two phrases and proclaimed his abandonment of all that is embraced in that difference. [Citing cases.] The difference which he thus disclaimed must be regarded as material, and since the amendment operates as a disclaimer of that difference it must be strictly construed against him.” In Schriber-Schroth Co. v. Cleveland Trust Co., 311 U.S. 211, 217, 312 U.S. 654, 61 S.Ct. 235, 238, 85 L.Ed. 132, the court stated: “But the particular invention to which the patentee has made claim in conformity to the statute is not always to be *108 ascertained from an inspection of the specifications and claims of the patent alone. Where the patentee in the course of his application in the patent office has, by amendment,. cancelled or surrendered claims, those which are allowed are to be read in the light of those abandoned and an abandoned claim cannot be revived and restored to the patent by reading it by construction into the claims which are allowed.” These explicit pronouncements 'by the Supreme Court have been followed by this court in Dixie Cup Co. v. Paper Container Mfg. Co., 7 Cir., 169 F.2d 645, and Ric-Wil Co. v. E. B. Kaiser Co., 7 Cir., 187 F.2d 1. Judge Dewey in his conclusions of lav/ stated that the accused fluxes infringe because they do “substantially the same thing, operate in substantially the same way, and produce the same result as the composition described in the claims,” and it is fair to assume that Judge Dewey’s decision resulted in the main from the application of this principle. There is no doubt from the cases but that this test is often applied in determining the issue of infringement. The question arises, however, as to the validity of its application in a case where the doctrine of estoppel is properly invoked. While we find no case where this question has been discussed, we think it obvious that there are instances where both doctrines cannot be given effect because of their inconsistency. It appears plain that the principle relied upon by Judge Dewey could not have been given effect in the Exhibit Supply Co. case because it would have met head-on with the estoppel principle which the court applied. In fact, its application would have required a different result in that case. While we shall later refer to the findings made by Judge Dewey, we think it pertinent to make an observation or two concerning his able and well-written opinion. He states, “The Jones et al. invention is so broad and teaches- so many new and additional features regarding electric welding that it seems to me the patent itself transcends in importance the compositions which are set out in the claims of- the patent, - and that' the claims are in truth subordinate to the invention shown by the patent.” At another point he states, “The defendants in .relying upon the opinion of their counsel, it seems to me, should have known that the patent was much broader than the claims standing alone might indicate.” These statements appear to indicate that in his view the scope of plaintiff’s invention was not to be measured by its claims. If so, we think the view was erroneous. See quote, supra, from 336 U.S. 271, 277, 69 S.Ct. 535. Also, the opinion, in response to- defendants’ 'contention with reference to estoppel, stated, “It is unnecessary to discuss these questions as I am satisfied the defendants’ fluxes are composed of substantially the same ingredients as those set out in the claims of the plaintiff’s patent.” We doubt the validity of this reasoning. The fact, if such it be, that the accused fluxes are composed of substantially the same ingredients is no reason in itself, under the authorities we have (heretofore discussed, to deny the application of the doctrine of estoppel. Thus, it appears to us that the result reached by Judge Dewey rests upon the erroneous premise, (1) the scope of the invention was not confined to the patent claims; (2) even though defendants had not overstepped the bounds of the claims they had come “toó close to the patent”; (3) the accused fluxes are “substantially the same” as the patented flux and infringe because of the sameness “in their use and results.” As already noted, the decision in this case must depend largely upon the meaning attributed to the words “a major proportion of one or more alkaline earth metal silicates,” contained in each of the four claims. Defendants construe this language to call for a composition containing something more, than 50% of alkaline earth metal silicates or their equivalents, While plaintiff argues that the language calls for nothing more than a percent which constitutes an important part of the flux. Dictionary definitions of the words “major” and “proportion” are relied upon by the parties as sustaining their respective contentions'. _We think they are of little help because definitions are available which support either view. *109 It is our judgment that defendants’ interpretation of the controversial language must be accepted. The ordinary and usual connotation attributable to such words is hardly susceptible to any other conclusion. But more important is the history attached to these claims and, more particularly, the ■controverted language from the time the claims emerged from the Patent Office until the commencement of the instant litigation. Even though the specifications do not define “a major proportion,” it would-seem that its teachings have been forcibly illustrated by plaintiff’s performance. Under the protection of the patent, and particularly these claims, it commenced and during the years has continued the manufacture and sale of its fluxes, which admittedly contain 94% of alkaline earth metal silicates or their equivalents. It is interesting and we think also pertinent to note that so far as we are aware plaintiff had never, until the commencement of this contempt proceeding, contended or raised any question hut that this controverted language required a composition containing at least more than 50% of the designated silicates. We suspect, however, that plaintiff’s failure previously to advance its present definition of “a major proportion” was due to the realization that such a concession would place in serious question the validity of the claims. If “a major proportion” means or can be interpreted to mean any percent — 10, 20 or 40 — their validity would be open to serious challenge because of their indefiniteness and uncertainty. As the Supreme Court said, first decision, 336 U.S. at page 277, 69 S.Ct. at page 538, “The statute makes provision for specification separately from the claims and requires that the latter ‘shall particularly point out and distinctly claim the part, improvement, or combination which he claims as his invention or discovery.’ ” (Title 35 U.S.C.A. § 33.) These observations, of course, are not intended to reflect upon the validity of the claims, but they are such that we should hesitate to now accede to an interpretation which, if it had been made at the time their validity was under attack, would in all probability have resulted in a contrary decision on that issue. The file wrapper "history, we think, militates strongly against plaintiff’s instant position. Perhaps no good purpose can be served and certainly it would unduly prolong this opinion to review this history in detail. On numerous occasions claims general in their nature were rejected by the Examiner, and it was only after amended claims were submitted which were specific in calling for “a major proportion” that the present claims were allowed. In arguing in support of claims 32 and 33, which eventually became claims 22 and 23 (two of those in suit), the patentee stated: “Judging from applicants’ knowledge and the prior art cited, the applicants were the first to discover that the compositions disclosed in the present application, e. g., composi- ’ tions consisting principally of alkaline earth metal silicate in a finely divided unbonded state, are useful as electric welding compositions in which a bare metal electrode is employed as the source of weld metal.” In the same argument, in distinguishing Severance (No. 429,777), the applicant stated: “Therefore, Severance does not disclose nor suggest an electric welding composition in which an alkaline earth metal silicate or its equivalent constitutes the major portion of the composition. Even if the content of the pumice-stone and the dolomite or silicate are considered together, such ingredients still would constitute but a minor proportion of the entire compound.” This argument was made "by the applicant even though Severance showed a silicate content of between 20 and 33%, whileh was distinguished on the basis that it constituted “but a minor proportion of the entire compound.” Other claims were rejected by the Examiner, with the statement: “Further he does not teach that an alkaline earth metal silicate and/or an alkaline earth metal fluoride in pure form or in just any proportion of a combination is efficacious.” Further claims were submitted, including claims 11 and 13 (claims 18 and 20 now in issue). The applicant stated of these claims, “These characteristics are, briefly, a high content of alkaline earth metal silicate, and substantial absence of uncombined iron oxide.” *110 Thus, these four claims were allowed on applicant’s representation that the alkaline earth metal silicate was of “high content,” that it was the “principal” ingredient of the combination and that it was of “a major proportion,” in.contrast with “a minor proportion” as disclosed by Severance. The representations made at that time are in marked contrast to plaintiff’s present position, wherein it is asserted in effect that the proportion of earth metal silicates is immaterial and that the test is whether they are sufficient to obtain the desired result. Plaintiff calls attention to the fact that claim 24 was allowed, whioh called for “a fluxing material for electric welding comprising metallic silicate and calcium fluoride,” and appears to argue from this that “a major proportion” was. not a controlling consideration upon which allowance of claims was obtained. It must be noted, however, that this broad claim was held invalid 'by Judge Swygert (affirmed by the Supreme Court) and, as already suggested, we think the claims before us would have met the same fate had they 'been couched in such broad language, that is, if they had failed to designate the proportion of silicate or if the designation employed, i. e., “a major proportion,” had been interpreted as plaintiff now urges. Judge Dewey did not determine, in fact refused to make a finding, as to the proportion of silicate contained in the accused fluxes. It would have been well near impossible to do so. Rarely have we studied a record which is so replete with contradictions and with opinions, some of which are based on assumptions admitted not to exist. Judge Dewey stated: “It would be a little difficult from the testimony of the experts to come to a positive decision as to whether defendants’ composition did contain a major proportion of silicate as the experts are somewhat divided on the question. All of the experts for the plaintiff and at least some of the experts for the defendants admit that there are some silicates of a kind referred to in the claims of the patent appearing in defendants’ product in some appreciable degree, but differ as to the amount.” This statement, so we think, begs the issue. How can a composition cilaim which requires “a major proportion” be infringed by a composition which contains the critical element only in an “appreciable degree”? There is no dispute, however, but that all of the 'four accused fluxes contained silicates in amounts far less than 50%. Both the plaintiff and the defendants have submitted charts showing in color the component elements of the aocused fluxes. (770 is not shown on the Charts because, as we understand, defendants have 'ceased its manufacture and sale.) The percentage of silicates of alkaline earth metals and equivalents contained in the accused fluxes, as shown by plaintiff’s chart, are 840, 35.6%; 760, 24%; 780, 41%. These percentages are quite favorable to the plaintiff, perhaps more so than the record justifies, for the reason that they include elements which plaintiff designates as equivalents and which may or may not be; in any event, they 'have not been so adjudicated. Plaintiff’s chart also shows what it denominates as “incipient” silicates (i. e., components heated to free them of gas and moisture) in the following amounts: 840, 57.4%; 760, 29.2 to 64%. No “incipient” silicates are shown for 780. “Incipient” silicates, according to plaintiff’s theory, are components which embrace the potentiality of developing silicate when chemically reacted. The word “incipient,” so far as we know, is without significance in chemistry and is a stranger to the patent specifications and to t'he proceedings in the Patent Office. It is plaintiff’s theory, however, as we understand, that what it denominates “incipient” silicates are the equivalent of the silicates called for in the claims, and thus plaintiff reaches the conclusion that the. aocused fluxes contain “a major proportion” of silicates. The frailty of this argument resides in the fact that it ignores the teachings of the patent as well as the basis upon which the claims were adjudged to be valid. This argument ignores the difference between the complete chemical reaction attained at high temperature taught by the patent and employed by the plaintiff in producing a com *111 position with “a major proportion” of silicates, and defendants’ method of heating at low temperature to free its composition of gas and moisture but not sufficient to produce silicates in any definable amount. The patent specification states, “The chemical reactions between the components of the welding composition must be completed before it is used in welding. Failure in this regard most surely invites porosity.” And it must be remembered that a chemical reaction between two properties has the effect of producing a totally different property (see footnote 1). Again, plaintiff by its commercial flux illustrates the teaching of its patent with a composition which has been subjected to a complete chemical reaction of its elements prior to use in welding. This is also true of defendants’ 660 flux, adjudged to infringe. 2 On the other hand, defendants’ accused fluxes are not completely chemically reacted before use but, admittedly, there is a reaction in some undeterminable amount at situs. Judge Swygert found that plaintiff had developed a welding composition “in which all reactions were entirely completed by prefusing at high temperature.” In fact, it was this complete chemical reaction prior to use by which Judge Swygert distinguished the composition claims from the prior art. D.C., 86 F.Supp. 200. In its brief filed in this court on the original appeal, plaintiff, referring to claim 23 as typical of the four claims now in suit, stated, “That, and every other claim in the patent, is qualified by the unequivocal instructions of the patent specification,” and enumerated five conditions which “must” exist in order to perfect a successful welding composition. The first “must” stated was, “The chemical reactions between the components of the welding composition must be completed before it is used in welding. Failure in this regard most surely invites porosity.” And' following the enumeration of the five “musts,” t)he brief stated, “Such imperative language puts that portion of the specification in the same Category as a ‘disclaimeri of anything which does not conform to those five ‘musts.’ ” Also, when the case was previously here, plaintiff, in response to defendants’ argument that the claims were so broad that they would cover fluxes whether prefused or not, answered, “But the mandatory language of the specifications requires: ‘The chemical reactions between the components of the welding composition must be completed before it is used in welding. Failure in this regard most surely invites porosity.’ ” Plaintiff in its former brief further stated : “Defendants also contended below that, if these claims were construed to cover their flux with its major proportion of manganese silicate, then in that light they were anticipated by Miller. The Trial Court took cognizance of and rejected that contention. The Trial Court quoted from the Miller patent to show that the preferred form of the Miller patent was a coating containing acidic and basic constituents which would react during the welding operation — whereas the patent in suit requires that acidic and basic constituents be reacted before the composition is used for welding.” And again it was stated: “The compositions in suit, without exception, are prefused in accordance with the mandatory direction of the patent that all chemical reactions be completed before use. The clay flux was nothing more than a mechanical mixture of clay, lime and ferromanganese which had been pulverized and heated well short of prefusion and then put into a tumbling barrel so as to intermix the ingredients thoroughly.” Thus, in view of the unequivocal representations heretofore made, we think it too late to contend that complete chemical reaction and fusion before use in welding was not an essential requirement of the invention. It was this characterization by the patentee which distinguished it from the prior art, both in the Patent Office and before the courts. *112 True, Judge Dewey found, “The mater= ials have been completely reacted by the kiln treatment,” but that statement is modified by the further statement, “Moreover such reacted acidic and basic materials are reacted with each other by the time-they come out of the Lincoln kilns to an extent-that the resulting flux contains a substantial amount of silicate.” If the component parts of the accused compositions have been -completely reacted and thereby only contain the percentage of silicates shown, there would be no infringement because they fail to constitute “a major proportion.” We do not understand, however, that plaintiff makes any contention that there has been complete chemical reaction of the components of the accused fluxes. Otherwise, plaintiff’s argument as to “incipient” silicates would be without point. It is forced to the position, contrary to that -heretofore assumed, that complete chemical reaction and fusion- prior to use' is, after all, immaterial and that defendants’ fluxes infringe even though there has been no or only partial chemical reaction and fusion prior to use. And this is on the theory that the -components of the accused fluxes possess the potentiality, upon complete reaction, to produce silicates in “a major proportion.” There is no finding. upon which this theory can be predicated, as Judge Dewey found on this point only that, “There is sufficient silicate in the defendants’ fluxes to accomplish the desired results * * In our opinion, the application of the doctrine of equivalency in this case is to ignore the teachings of the patent, the representation upon which the claims in suit were allowed and, more pointedly perhaps, the representations by which their validity has been sustained in the courts. While the difference between plaintiff’s composition and those accused may not be ' great, it is that difference which distinguished plaintiff’s composition from the pri-' or art and which enabled it to sustain the validity of its grant. It is now estopped from -claiming otherwise. The judgment appealed from is reversed and remanded, with directions that it be vacated. Footnotes [1] Dr. Willard, a professór of chemistry and an important expert- witness for the plaintiff, was asked, “Now, what, in a little more detail, do you mean by a reaction?” Ilo answered, “A reaction in the chemical sense is one in which there has been a chemical combination between two substances to form another substance with totally different properties.” [2] Referring to the 660 flux, Dr. Willard testified, “This is a prefused material. We know that because unless the elements are combined at a high temperature and fused, there is no reaction between them.”