Westinghouse Electric Mfg. Co. v. Montgomery Electric Light & Power Co.

TOWNSEND, Circuit Judge. The following statement may conduce to a clearer understanding of the Stanley patent and the questions herein involved: The invention relates to a system of electrical distribution and regulation especially adapted to incandescent light plants. The problem which confronted the inventor, Stanley, was presented by defects in existing systems which resulted in variation in the luminosity of individual lamps upon variations of pressure due to variations in the number of lights in use. Stanley discovered that the cause of this variation was a lack of harmonious action between the dynamo and the converter. He had to deal with the antagonism between leakage at the dynamo and self-regulation of secondary pressure in the transformer. Increasing the length of the primary wire decreased leakage, but diminished self-regulation. Decrease in the length of the primary wire improved regulation, but increased leakage. Stanley further discovered that these antagonistic relations could be co-ordinated to each other by the use of a certain length of wire on the primary of the con — 891 — verter, and he told the public in his patent that the wire should be “of such length that reacting self-inductively upon its own magnetic circuit, the average counter-potential so produced approximately equals the potential applied to the primary circuit.” Of this statement, Judge Coxe says in his opinion, supra: “In other words, if wire be wound on the primary coil until the ammeter practically shows no current when the secondary current is open, the result will be the invention of the Stanley patent.” In addition to this disclosure, Stanley further stated in the succeeding portion of the same paragraph the method employed by him in practice for accomplishing his invention, which method involves the so-called “C2R rule.” The portion of the specifications of the patent in suit containing the above statements and the statement of the C2R rule is as follows: “In the construction of the coils P and S the following principles are to be observed: The first thing to be determined is the length of the primary wire. This should be of such length, that reacting self-inductively upon its own magnetic circuit the average counter-potential so produced approximately equals the potential applied to the primary circuit. When so constructed an ammeter will practically show no current when the secondary circuit is open. To obtain these results in practice I use the following method: I first choose the percentage of efficiency to be obtained. Then having selected a type of magnetic circuit affording as great magnetic conductivity as possible I apply such a length of primary conductor that acting self-inductively upon its core the difference of the counter-potential and applied potential multiplied by the current in the converter shall equal the predetermined loss of energy inevitable in conversion and vary the length of wire until the desired results are attained.” The claims in suit are as follows: “1. In a system of electrical distribution, and in combination, an alternating-current dynamo and converters electrically connected with the main-lino conductors in multiple arc and organized to transform the current in the main conductors into currents of less potential and greater quantity in the secondaries, each converter made with a primary coil containing such length of wire exposed to magneto-electric induction that when operated by the dynamo with which it is to bo used with its secondary circuit open the electrical pressure and counter-pressure in its primary circuit shall be equal with incandescent lamps or other translating devices in the secondary circuits, substantially as and for the purposes set forth.” “3. In a system of electrical distribution and in combination, an alternating-current dynamo and converters organized to transform the current generated by the dynamo into currents of less potential and greater quantity at or near the points of consumption electrically connected with the main-line conductors in multiple are and having their primary circuits constantly closed, each converter adapted to the dynamo operating the system by making its primary coil of such length that when supplied with its full proportionate share .of the entire normal electro-motive force of the machine its secondary circuit being open the electrical pressure and counter-pressure in its primary circuit shall be approximately equal with the translating devices in the secondary circuits of the converter to he cut out of the circuit when not in use without the introduction of any resistance in the place of them substantially as and for the purposes set forth.” We are satisfied upon all the evidence that the length of wire given by the Stanley C2R rule does not substantially differ from, and is not inconsistent with, the disclosure of the discovery of the principle of co-ordination between the generator and the primary. — 892 — For reasons to be hereafter stated, it is unnecessary in the disposition of this case to discuss at length the utility or accuracy of the disclosure as to length of wire, as compared with the particular rule or method used in practice, or the alleged discrepancies between them, or to what extent the former is to be imported into the other and made a part of it; As we understand it, Stanley told the world- in the earlier part of the paragraph that, in order, to obtain automatic regulation, the wire must be given a length which, under certain open circuit conditions, would show practically no current or a zero current, and by this C2R rule he showed the method by which he secured the length of wire essential to the accomplishment of the desired results. Twenty-five pages of defendant’s brief are devoted to a discussion of its construction of the patent in suit. It is unnecessary to discuss all the arguments advanced in support of this construction. The fallacy of these contentions, and the fact that they proceed upon a misconception of complainant’s position and our decisions, is indicated by the following quotations from defendant’s brief: “After stating in his specification that the primary circuit should be of such length that an ammeter will practically show no current when the secondary is open, the patentee says, ‘To obtain these results in practice I use the following method,’ and then the C2R rule is given, clearly indicating that the part of the specification which the complainant now relies upon will not give these results in practice, but that to obtain these results in practice, the 02R length of wire must- be secured. The complainant, however, wishes now to reject the C2R rule and the length of wire given by it.” The first argument of defendant, whereby the rule is made a part of the claims in suit, is stated by counsel in its brief as follows: “This court in its opinion on motion for preliminary injunction in the case at bar set at rest this question (the question whether the rule was part of the claims in suit), holding that in its decision in the Saranac Case it did not mean that the rule was to be regarded as an element of the combination, but ‘that the first and third claims of the patent were for combinations which broadly included his invention.’ The express finding of this court in the Saranac Case that the Stanley patent is limited to the length given by the Stanley C2R rule which this court read into the claims to save them from anticipation, and which therefore is a different length from that given by the length of the claims, stands, and has not been modified in the slightest, and not being so modified is still controlling against the complainant.” In support of this argument counsel for defendant refers to the statement in our opinion on the motion for preliminary injunction that we were not convinced that in the Wagner transformer used by defendant there was the length of primary wire required by the Stanley patent, and that infringement was not so clearly established as to justify a preliminary injunction. Counsel, therefore, claims that this statement shows that “this court did not agree with the lower court that this rule which it applied was the Stanley rule or some other method which would give the length of wire which this court sustained the patent for,” etc. There is no foundation in fact or in our opinion for this assertion. Both of the judges who sat in said cause were of the opinion that the construction adopted by the court below was the correct one. But wé reversed the order of the court below with great hesitation, upon the expert testimony as to the sandwiched coils in defendant’s transformers and the conflicting theories upon other ab — 893 — struse electrical problems, and gave the defendant the benefit of the doubt raised, not as to the construction of the rule or its application, but because, as stated in our opinion: “The experts are in flat contradiction upon the question of infringement, and, in view of the large importance of the controversy, the rights of the parties should be reserved for decision until the final hearing of the cause.” Counsel for defendant quoted from our opinion on the motion for preliminary injunction as follows: “We are not convinced that in the Wagner transformer used by defendant there is the length of the primary wire required by the Stanley patent.” Counsel then says: “This court could not have meant ‘we are not convinced that in the Wagner transformer used by defendant there is such length of primary wire that there is practically no current when the secondary circuit is open,’ for it was expressly admitted, so this court could only have meant that it was not convinced that the Wagner transformer had the length of wire given by the Stanley C2R rule.” On the contrary, this is exactly what the court did mean, namely, that it was not convinced that the length of primary wire used by the defendant corresponded to the length of wire by which the co-ordination of the transformer to the generator was secured, according to Stanley’s statement of his discovery, and that the co-ordination between the generator and the primary was secured by such a length of wire that under certain conditions the current would be practically zero when the secondary was open. Counsel for defendant insists that “the part of the specification immediately preceding the Stanley C2R rule and in the same paragraph with it is imported into the C2R rule and made a part of it,” while in our opinion we say: “The rule was not to be regarded as an element of the combination.” The defendant, quoting certain portions of said decision, but omitting our statement in regard to the rule, says as follows: “The decision of this court was not to the effect that the Stanley invention is for co-ordinating the transformer to the generator with which it is to be used by any length of wire, but that it is for co-ordinating the transformer to the generator by means of ‘the length of the primary wire required by the Stanley patent’; that is, ‘the Stanley length.’ It is evident, therefore, that this court regarded ‘the length of the primary wire required by the Stanley patent,’ or the ‘Stanley length,’ as some peculiar and definite length, and did not regard the Stanley invention as covering any length of wire by which the transformer was co-ordinated to the generator.” A reference to the claims in suit shows that the wire of suitable length was for such a length exposed to magneto-electric induction that when operated by the dynamo on open circuit the electrical pressure and counter-pressure in the primary circuit should be equal. In the prior opinions of this court in the Saranac Case the character and scope of the discovery and invention of the patent were discussed and defined, and the validity of the claims in suit was sustained. In the opinion in this cause on appeal from the order for preliminary injunction we said as follows: “The decision of this court in Westinghouse Co. v. Saranac Lake Co. was tn effect that Stanley contributed to the prior art the discovery that the auto — 894 — matic and constant regulation of the pressure of the alternating current at the secondary terminals depended upon the co-ordination of the transformer to the generator, and the invention that this could be effected by means of a primary wire of suitable length, that the first and third claims of the patent were for combinations which broadly included his invention, and that because, in ord<- r to instruct those skilled in the art how to ascertain the length of the primary wire he had formulated in his patent one rule for doing so, the rule was not to be regarded as an element of the combinations.” In the opinion of Judge Uacombe in the Orange County Case, he said, referring to our opinion in the Saranac Case, as follows: “It seems to this court that the Court of Appeals found that the claims of the patent were for a combination of the elements therein set forth, of which one element was a primary coil having a length of wire equal to what would be found to produce the indicated results when applying the Stanley rule.” In this statement we concur. It may, therefore, be considered as determined by the prior decisions in this circuit that as Stanley was the first to discover and suggest that the difficulties encountered in attempting to secure self-regulation were due to an improper length of wire in the primary coil, and to disclose a method and means for determining the proper length of wire, thereby proportioning the energy absorbed by the generator to the energy consumed, he was entitled to cover broadly any transformer having substantially such length, irrespective of the method by which such length was ascertained. It follows from this conclusion that the Stanley C2R rule, which states the method of securing the proper length is not to be regarded as an element of the combination of means of the claims in suit, which include such wire, and that the question of infringement depends on whether the primary of defendant’s transformer has substantially such a length of wire as would have been obtained if its length had been determined by the Stanley C2R rule. Upon further examination of the prior opinions bearing on the questions raised herein, and exhaustive consideration of the briefs and arguments of counsel, we remain of the opinion that the claims in suit are for combinations which include the invention of a transformer which produces the automatic regulation of the Stanley patent by the use of a length of wire substantially the same as that which would result from following the method stated by him in his general rule applicable to all conditions, and that a transformer having such a length of wire is an infringement of the claims in suit, irrespective of the process or rule or method by which the result may have been obtained. That the defendant, in support of its argument, misstates the position of complainant, appears from the following statements in the respective briefs: Defendant. Referring to the Hopkinson equation, counsel says: “The complainant, seeing the force of the situation, now says that it is not necessary to have in a transformer the corresponding length of wire given by the 02R rule.” Complainant. “I admit, as emphatically and as broadly as words can make it, that no transformer infringes which does not have on its primary a length substantially such as it would be if such length had. been determined by the Stanley C2R rule.” — 895 — In the disposition of the questions herein we have proceeded upon complainant’s admission, quoted above, and for that reason have found it unnecessary to pass upon defendant’s claim founded on its statement of complainant’s contention as to the C3R rule length of wire. In view of the fact that anticipation and lack of patentable novelty are not res adjudicata as to this defendant, the further defense is here pressed that the Stanley rule, so far as defendant follows it, was anticipated by Kennedy and Hopkinson. This contention is based on the fact that the Stanley C2R rule did not appear in Stanley’s original specifications of November 36,1885. It is claimed by complainant, however, that the Stanley discovery and invention as construed by this court was disclosed in said original specifications, and was then embodied in a transformer. And the date of the Stanley invention was found to be in 1885 by Judge Coxe in the Saranac Case, and by the Commissioner of Patents in interference proceedings of Stanley against Slattery. The Kennedy patent and articles were fully considered and disposed of by the court below and by this court in the Saranac Case, and need not be reconsidered. But, so far as concerns the Hopkinson equation, whether the Stanley invention be given the date of 1885 or 1888 is immaterial, for the following reasons: Tike the other alleged anticipations relied upon in the Saranac Case, it fails to state that one “nuf determine the proper length of the primary coil by connecting the transformer in circuit with the dynamo with which it is to be used, and then winding on wire until the loss indicated by xhe formula C3R, with the secondary circuit open, equals a certain loss of energy.” It appears from defendant’s testimony that they did not use the Hopkinson formula until 1891 or 1893, some years after Stanley had explained why and how his rule disclosed a practicable method for securing self-regulation. The Hopkinson equation is not mentioned or discussed by7 complainant’s expert, and we are therefore left without any means of determining the accuracy of Hopkinson’s statement that “for practical purposes these equations are really sufficient.” But upon the testimony of defendant’s experts it is shown to have no bearing on the issues herein. Thus, defendant’s expert, Nipher, says: “It was these equations which gave the public — those skilled in the art— the information which they needed concerning the nature of the problem, and the work of Rankin Kennedy and Zipernowski and Deri furnished the -practical application of these principles, at the same time .giving speeifle and sufficient instructions concerning the exact method of procedure, in order to realize the conditions in practice.” But this court has already considered and disposed of these alleged anticipations, and has held that they failed to disclose the invention in suit. Again, as defendant’s expert, Nipher, says: “The length of the wire wound on in this method of design [Tlopkmson’s] is entirely different from that which would be obtained by following the Stanley rule.” That is, the defendant relies on the disclosure by Hopkinson both as a defense against infringement, on the same ground on which it claims — 896 — noninfringement by its own transformer, and also as an anticipation. It claims, on the one hand, that upon its construction of the Stanley rule, which admittedly makes it indefinite and misleading and gives an impracticable length of wire, the defendant does not infringe, because its construction, in which wire of a different length is used, follows the practicable method disclosed by Hopkinson, which gives the proper length of wire. And, on the other hand, it claims that its method anticipates the Stanley rule because it follows the prior Hopkinson equation. The defendant, therefore, is in this dilemma: If the Hopkinson equation, as Nipher, defendant’s expert, asserts gives a length of wire different from that specified by Stanley, then, as we are of the opinion, as hereafter shown, that the defendant uses the Stanley length, the contention that it uses the length of the earlier Hopkinson equation is not sustained. And, if the Hopkinson equation could be so construed as to give the Stanley length of wire, then the admission that it merely gave the information by means of which Rankin Kennedy, etc., furnished a later practical application of these principles, as stated above by defendant’s expert, then Hopkinson is relegated one step further back than the post-art publications already disposed of in the earlier opinions of the court. We find nothing-new in this record which indicates that any of the so-called prior art adversely affects the status of the patent in suit. The single question to be disposed of on this appeal is that of infringement. Infringement is denied, on the following grounds: “Defendant uses a different length of wire than that prescribed by the patent, by reason of using less iron in the core.” In support of this proposition defendant states that the length of wire used in its transformer has a length in the primary coil over 50 per cent, greater than the length in the complainant’s Great Barrington transformers and contends that if the Great Barrington transformers have not the C2R length then the Hopkinson length antedates the C2R length whatever the latter may be. For reasons stated above' the alleged priority of the Hopkinson length equation is immaterial. The defense that as defendant uses a core of lesser weight and therefore a length of wire different from that used by complainant, it does .not use the length of wire of the Stanley rule, is immaterial because, as the Court of Appeals has held: “Tbe amount of'wire for a given character of current supply cannot'be stated in feet and inches because it is, to some extent, dependent upon other things, such as the quality of iron employed in the core, the quality of copper used in the coils, the shape of the transformer, and the way the coils are applied.” ___ The curve sheets illustrating tests of complainant’s transformers and of defendant’s transformers in the Saranac Case and in this case show that in each the currents are substantially the same, and that the current falls with an increase in the number of windings until it reaches the point where the current is practically zero, this being the point where by the application of the C2R rule the winding should cease. And, while different lengths of wire were used on certain of the com — 897 — plainant’s Westinghouse transformers, and on the infringing Saranac transformers, and on defendant’s transformers, yet in each case the lengths of wire cease at the same point or just beyond the knee of the curve. Again, it appears from a comparison of such of complainant’s and defendant’s transformers as have substantially the same length of wire that the respective transformers are substantially alike in material, size, and shape of cores. In fact, the assertion of complainant’s expert, Waterman, that the defendant’s transformers are in their essential particulars like those held to infringe in the Saranac Case does not seem to be seriously disputed. The construction of defendant’s transformers in the Saranac Case is stated by Judge Coxe in his opinion therein, where, referring to defendant’s plant} he says: “Each converter is made with a primary coil containing such length oí wire exposed to magneto-electric induction that when operated by, the dynamo with which it is to be used, with its secondary circuit open, the electrical pressure and counter-pressure in its primary circuit are approximately equal.” A further defense is that: “Defendant’s transformers have their coils sandwiched, which not only gives a different length of wire from the patent, but does not -require such transformers to be adapted to the dynamo, except as to pressure, and therefore they do not infringe.” Great stress is laid by counsel for defendant on this feature of sandwiching. As explained by the expert for defendant, it consists in the “subdivision of the coils into sections to avoid excessive electro-motive ■force between adjacent parts.” By this method of construction the primary and secondary coils are brought close together and the amount of leakage is reduced. The result of sandwiching the coils is to require a greater length of wire. No reference is made in the Stanley patent to sandwiching, and its practical importance in constructing transformers was not at first appreciated by the complainant company, and it did not adopt this construction until the year 1892. ■The plausible argument is made that as Stanley used a large core to prevent leakage, while the defendant produces the same resulf by sandwiching its coils, necessarily using in so doing different lengths of wire, that thereby the necessity of correlating the transformer to the dynamo as to “frequency and electrical pressure,” as in the Stanley construction, is dispensed with, and that infringement is thus avoided. But it appears from the testimony of defendant’s experts that this sandwiching and its desirability to prevent leakage was well known long prior to the invention in suit; in fact, that it was a common expedient for that purpose. We fail to find any suggestion that this arrangement dispenses with the necessity of “the co-ordination of the transformer to the generator” by the use of a primary wire of suitable length. And we do not understand that in these circumstances a mere difference of length of wire by reason of sandwiching is any more material upon the question of infringement than differences in other parts of the apparatus. We again repeat, but with our italics, what this court said in the Saranac Case: — 898 — “The amount of wire for a given character of current-supply cannot be stated in feet and inches because it is, to some extent, dependent upon other things, such as the quality of iron employed in the core, the quality of copper used in the coils, the shape of the transformer, and the ioay the coils are applied.” The whole argument of defendant in support of noninfringement proceeds upon the theory that the Stanley rule is misleading and impracticable. Thus, Prof. Gray, one of defendant’s experts, having testified as to his construction of the rule (a point to be hereafter discussed), says as follows: • “X-Q. 29. And, in your opinion, no transformer will have ,what you term the Stanley length of wire unless it has on its primary a length which will give these impossible results. Is that right? “A. I believe the instructions in the patent lead to such an impossible conclusion. “X-Q. 30. So that, when you say that the Wagner transformers which are involved in this controversy do not have the Stanley length of wire, you would make the same statement concerning every transformer that was ever built? “A. When I say that the defendant’s transformers do not contain the Stanley length of wire, I mean practically that the length of wire used in defendant’s transformers could not Have been obtained by the Stanley rule, because I consider the rule to be entirely an unworkable one. “X-Q. 31. Being an entirely unworkable rule, in your opinion, of course, no transformer that was ever built could have what you are construing to be the Stanley length of wire? “A. I consider that no transformer that was ever built, and successful, could have been designed by this rule.” The substantial and difficult proposition presented by the argument of defendant on the question of infringement, therefore, is that the commercial transformers do not have the Stanley C2R length of wire; or, in other words, that the patent is invalid for lack of invention. We have carefully studied the argument in defendant’s appendix on this point, and we think each proposition there advanced is met by the discussion in complainant’s brief, supported by the facts proved or admitted on the record. That Stanley made and disclosed an invention of a practical, working device must be assumed; that upon defendant’s construction of the patent a transformer built according to its instructions would be a useless one is admitted. The decision as to infringement depends upon the construction to be given to the phrase “loss of energy inevitable in conversion,” occurring in the following passage in the specifications: “To obtain these results in practice 1 use the following method: I first choose the percentage of efficiency to be obtained. Then having selected a type of magnetic circuit affording as great magnetic conductivity as possible I apply such a length of primary conductor that acting self-induetively upon its core the difference of the counter-potential and applied potential multiplied by the current in the converter shall equal the predetermined loss of energy inevitable in conversion and vary the length of primary wire until the desired results are attained.” This language is construed by the counsel and experts, respectively, as follows: — 899 — Complainant. ■'T contend that this phrase means the copper loss in tlie primary before tlie secondary is applied to the core; or, as electricians say, the primary open-circuit copper loss." Defendant. “I understand the passage referred to to give instructions to wind on wire •until the copper losses in the primary equal the total loss inevitable in conversion. I was careful to point out: in my direct testimony that I considered tlie inventor to believe that there was no iron loss, and also that there was a probability that the energy lost in the secondary was considered to be part of the energy converted and was therefore not included in the loss appearing in the primary. * * * "X-Q. 25. You uso the term Toss inevitable in conversion’ to include what? “A. I consider that the losses inevitable in conversion include the copper losses in both coils, and also the iron loss.” Defendant’s construction is supported by expressions in Stanley’s original specification and by indications that he knew very little about iron losses in the core of the transformers, and it is claimed that he thought the copper losses were the same for all loads. Prof. Nipher, expert for defendant, says on this point as follows: “The loss of energy inevitable in conversion he thought was a definite loss, constant for all loads, and the same as for no load,, and could therefore be determined by finding the C2R loss in tlie primary coil with tlie secondary circuit open. The Stanley rule was based on this fundamental misconception of the action of the transformer.” And it is claimed that because of his ignorance in these regards he made an impracticable and misleading rule, which, if followed, upon one interpretation would give a length of wire which would be “ridiculously small,” or upon another interpretation “enormously large.” The arguments in support of these assertions are elaborately discussed in the opinions of defendant’s experts. It is admitted that the language used is capable of either of the constructions claimed. The construction contended for by defendant assumes that the meaning of the term “predetermined loss of energy inevitable in conversion” means the “total losses in the transformer,” including therein “the full load primary and secondary copper loss plus the iron loss, which is the same at no load as it is at full load.” Tlie result of this construction is to make the patentee say that after having selected in advance a percentage of efficiency, which, it is agreed, is the ratio between the input and output at full.load, say, for example, in a 1,000 watt transformer, an efficiency of 95 per cent., which would mean a loss of 50 watts, that then the C2R rule means that the constructor is to wind on wire on the primary until the copper loss thereon shall equal the total copper and iron losses of energy inevitable in conversion. This construction is explained by one of defendant’s experts as follows: — 900 — ' “The patentee had a theory that there were no iron losses, that the iron core was a frietionless vehicle, and that the losses were all in the copper, and he thought that when the wire was wound on until the copper loss was equal to the predetermined loss that he would have a transformer which, when completed, would deliver the power put into it, less the copper loss on open circuit. He thought he had taken into account all the losses in the transformer.” ■ • As stated above, there is evidence outside the patent indicating that the patentee in common with the general public was ignorant in regard to iron losses in the core at this period. One of the experts for defendant says as follows: “If the patentee’s ideas as above recited had been correct, if there had been no iron loss, and if the copper loss had been constant throughout the entire working-limits of the converter, this Stanley rule would not have been absurd any more than it would have been absurd to build a buggy- with the friction on one axle equal to the friction on four axles, if the other three axles had no friction. In the Pope letter of December 10, 1886, he has stated that this iron core ‘should have such capacity as to act as a frietionless vehicle to transform all the lines of force from the primary to the secondary circuit, without loss in transmission.’ ” The forcible arguments in support of complainant’s construction, as stated above, accord with the fundamental principles of interpretation applied to this patent, and enforced by its inventive results, and with the reasoning which would naturally be attached to the language used. These arguments may be summarized as follows: 1. It is settled-by the decisions of this court that the patentee contributed the discovery of the dependence of regulation of pressure upon the co-ordination of generator and transformer, and the invention of a means by which this could be effected, namefy, the primary wire of suitable length, and disclosed a method for determining that length with mathematical exactness by a rule applicable to all conditions, including the winding on of wire until the loss indicated by the C2R, with the secondary circuit open, shall equal a certain loss of energy. 2. The defendant’s construction would nullify the invention by making the patentee direct either the use of a wire so short that it would burn up the transformer, or so long that it would extend for hundreds of miles. This construction, as shown above, would further make the patentee say that' the copper loss on the primary with the secondary open ■ should^ be equal to all loss of energy with the secondary applied, although the whole context indicates that the patentee was dealing with open circuit conditions. 3. Counsel for defendant in the Saranac Case assumed that complainant’s construction was the correct one. In their brief they say: “There can be no question that ‘the predetermined loss of energy inevitable in conversion,’ here referred to, is the wire or resistance loss in the primary coil at no load, and that the rule deals merely with the primary coil, and can have nothing to do with the total loss in the converter when the secondary coil had been adapted thereto. * * * It is merely the wire or copper loss in the primary coil when there is no secondary coil, or with the secondary circuit open.” 4. It is the duty of the court to adopt such a construction of a meritorious patent as shall sustain rather than invalidate it — 901 — As this court said in Dixon-Woods Co. v. Pfeifer, 55 Fed. 390, 395, 5 C. C. A. 153: “The patentee told the trade of which he was a member by what mechanical means breakage of glass in the process of annealing coaid be saved; in other words, how to anneal glass better and more economically. His patent described clearly enough the ways in which bars and tho operative mechanism should be constructed and operated, and the glass should be conveyed through the leer. lie did not know, or he did not tell, why the new method would produce better results. He simply told how to construct a machine which carried the glass through the leer on a level, and saved much breakage; but he ought not to lose the statutory benefits which would certainly belong to him if he had seen and described the philosophy of his machine accurately.” The patentee stated a rule to be applied with reference to the only losses which at the time of the invention were capable of ascertainment by the then known devices of the art. It now appears that by the use of later appliances other losses, of which the patentee was ignorant, may be measured and ascertained. But we do not understand that this ignorance of the patentee, or the knowledge subsequently gained from the prior art, affects the status of the invention in suit. The fact that the patentee did not fully understand the principle upon which his invention operated, or that his instructions were capable of a construction which would render the patent impracticable and defeat its purposes, should not deprive him of the benefit of a meritorious invention, provided it appears, as is found in this case, that the patent sufficiently disclosed to those skilled in the art the cause of previous defects and a new and useful discovery and invention, by means of which they might be successfully overcome. He is not to be deprived of the benefit of his invention because lie may have been mistaken in his statement of the reasons why the result was secured, or may have failed to correctly state the theory of their operation. As to the use of the word “predetermined,” we think that the inventor merely meant that the constructor should ascertain beforehand by experiment, as complainant’s expert says and explains, certain “physical properties, with regard to which the construction is to be worked out.” 5. It is admitted that at the time of the Stanley invention there was no means of measuring the loss of energy inevitable in conversion, other than the copper loss on open circuit. Complainant quotes from defendant’s brief in the court below the following statement: “Stanley knew nothing about the loss of energy inevitable in the core. If he did, lie said nothing about it in his patent, and he had no way of measuring the loss of energy inevitable in the core at the time his patent was applied for. The only way in which the loss of energy inevitable in conversion can be measured is by a watt-meter. At the time of Mr. Stanley’s invention and application for patent there were no watt-meters.” Therefore, when Stanley instructed the constructor to wind on wire until the loss of energy was provided for, he must necessarily have referred to that loss which was capable of being ascertained by measurement. Neither the patentee nor the public, at the date of the invention, could have estimated losses other than the copper losses on the open primary. — 902 — 6. That defendant’s construction should not be adopted is further indicated by the following statements in complainant’s brief, which seem to be sufficiently established by the record. Referring to defendant’s construction, counsel for complainant says: “If the phrase means this, then a transformer built according to the Stanley rule will violate every line of the specifications outside the rule. It will have an oversaturated core, whereas the specifications say that an undersaturated core is ‘indispensable.’ It will have a leakage current of enormous and even destructive volume, whereas the specifications say that the leakage current will be' ‘practically zero.’ It will not have anything like the efficiency chosen, although the rule says its object is to produce a transformer having the chosen efficiency. It will not be self-regulating, although the object of the rule is to produce a self-regulating transformer. It will not be a transformer at all, because it will burn up the minute the current is turned on. It will be nothing but a piece of electrical fireworks.” It remains to consider the further argument of defendant that if the length of the primary wire on the core were such that the ammeter would show practically no current, as specified by Stanley, there is nothing in the specifications to show when to stop, because the rule fails to state definitely ho'w much is to be- wound, and that a person might wind an amount of wire which would be so short that it would burn out, or so long that it would stretch to hundreds of miles. But, upon an examination of the curves shown in complainant’s exhibits, it appears that the number of turns when the working point is first reached is the time when you first get the zero result, as stated in the patent. If you continue to wind more wire on, the transformer would not be self-regulating. Therefore the patent may fairly be interpreted as directing the constructor to stop at that point. We think the contention of complainant that the patent does, in substance, say “wind on wire until the current becomes practically zero, and then stop,” is correct because it says “that the primary coil should have such length that an ammeter would practically show no current,” etc., when the secondary circuit is open, and practically says, or may fairly be construed to say, that the wire should be wound on until the desired result is secured. Counsel for defendant asserts that under the instructions of the patent the constructor may continue to wind on such a great number of turns, after reaching the zero point, that the transformer would not be a commercial one, and claims that the length is “indefinite and indeterminate” because “the number of turns may be so great, and in consequence thereof the resistance of the primary coil so high, that the transformer would not be commercial.” We think it has been sufficiently shown that no such construction is justified. In view of the importance of the interests involved, we have given exhaustive consideration to the various intricate technical questions raised, and have endeavored to dispose of them in the light of the invention derived from the expert evidence. It is, of course, possible, in view of the conflicting claims upon the electrical problems and tests involved, that we may have been mistaken in some of attempted statements of fact. But; in any event, we are satisfied that under,the construction of the patent originally adopted by this court in its opinion in the Saranac Case, infringement by defendant.is abundantly es — 903 — tablished, and that the adoption of the view as to anticipation and construction now contended for by the defendant would practically result in a reversal of our former opinions. The decree is affirmed, with costs.