Wright Aeronautical Corp. v. General Motors Corp. 166 F.2d 636

SPARKS, Circuit Judge.

Plaintiffs-appellants charged appellee with infringement of United States Patent No. 2,103,643 to Salomon, issued December 28, 1937, on an application filed March 31, 1933. It was alleged, and the court found, that the patent, at the times in issue, was owned by a Swiss corporation, hereinafter referred to as Redynam, and that Wright Aeronautical Corporation, hereinafter referred to as Wright, is the exclusive licensee for the aircraft engine field, and has been since the patent was issued.

The defendant-appellee, General Motors Corporation of Delaware, through its Allison Engineering Division, of Indianapolis, referred to hereinafter as Allison, manufactures radial, air-cooled aircraft engines including the engine alleged to infringe.

The issues raised were validity and non-infringement as to the claims sued on, namely, Nos. 3, 4, 5, 6, 7, 8, 10, 11, 13 and 18. After finding the facts specially and rendering its conclusions of law thereon, the court held there was no infringement as to any claim, but it did not pass upon the question of validity. From that ruling this appeal is prosecuted.

The patent is alleged by plaintiffs to be based upon an application filed in France on April 7, 1932, now French patent No. 748,-909. According to the present specifications, the patent here in issue is applicable to various types of machinery, stationary as well as movable, and allows the replacement of the usual fly wheels adapted to regularize *637the movement of machine shafts of such machinery. Its application in the airplane engine field only is here involved. The question is whether the bifilar type of vibration damper used by appellee infringes the claims in suit which the District Court found were limited to a damper of the unifilar type.

The court found the following facts, all of which are supported by substantial evidence. Allison is, and has been for many years, a well-known manufacturer of aircraft engines of the in-line, liquid-cooled, “V”-type, similar in those respects to automobile engines which have for a long time been manufactured by Cadillac and other General Motors Divisions. These Allison engines, incorporating the centrifugal vibration dampers which are here alleged to infringe the Salomon patent, were manufactured by Allison beginning as early as January 1937, and have been sold from shortly after that date to the present time, in ever increasing quantities. There has been no material change in the design, or construction of the accused Allison vibration dampers since they were first put out by Allison.

These Allison engines were sold to the British Purchasing Commission before the United States entered the war; but as regards the issues here, those engines are like all the other engines manufactured by Allision, beginning in January 1937, including those sold to the United States Army for use in its current fighter planes such as “P-38,” “P-51,” and others.

The invention of the patent in suit is characterized by a roller movable about a stub shaft, the presence of which so limits the mass of the roller that it is incapable of accomplishing the results accomplished by defendant’s bifilar damper.

All licenses granted by plaintiffs under the Salomon patent here in suit, included other patents not here in suit.

The claims in suit were all inserted in the Salomon application after the accused Allison damper was being manufactured and sold.

Proceedings which took place and actions which were taken by or on behalf of plaintiffs, before the Commissioner of Patents and by way of communications made to defendant and the industry generally, preclude plaintiffs from maintaining that the patent in suit is of such scope as to cover a bifilar vibration damper or balancer.

Defendant Allison’s construction is best shown in the physical device which is an actual damper with its attached crankshaft, taken out of an Allison engine just as it would be actually installed in a modern fighter airplane.

Both the construction and the operation of the accused Allison damper are different from any disclosure of the Salomon patent. The accused Allison damper comprises six separate centrifugal pendulum balancers, each one of which consists of a large external damping weight which is suspended by two small pins from the flange or web attached to the crankshaft. Each of the pins passes through a hole in the weight and also through a hole in the flange. The use of those two pins, for that purpose, causes it to be known as a “bifilar suspension” of the damping weight. Its operation is dependent upon both of the pins being present. If one pin were removed, the construction would be inoperative. It is a fact, however, that the pins do not themselves perform any useful damping function in the device. Their weight is too small to have any effect on the damper; and, aside from that, taken alone they would be tuned to the wrong frequency and hence could not, by themselves, perform any useful damping function in the defendant’s construction. Their weight is not even taken into account in the calculation.

Appellants contend that the court erroneously (1) limited the scope of the patent to the phases of the unifilar type; (2) found that the bifilar, utilized by defendant, is not equivalent to the one patented; held (3) that the presence of a stub shaft in the patented device is a necessary element of the invention; (4) that defendant employs only the construction of the prior art; (5) that the claims were so broadened in the Patent Office, after defendant’s accused device appeared, as to bar this action for infringement; (6) that the arguments addressed to the Patent Office and certain written communications addressed to defendant and others, by Rubissow, the authorized agent of the owners and licensees *638of the patent, precluded appellants from claiming that the accused device infringes.

At all times here involved it was well known that torsional vibrations in the form of successive twisting and untwisting effects in the crankshaft of a multiple-cyl-indered internal combustion engine are caused by successive explosions in the cylinders and that by reason thereof flexional vibrations due to “sagging” occur in the shaft. The .parties agree that the twisting and untwisting follow each other in rapid succession; that, in a modern aircraft engine such as defendant’s, there may be as many as 200 to 300 torsional oscillations per second, and that, if the oscillations are severe enough and repeated often enough, they may shatter the crankshaft. Consequently, in order to meet and counteract these dangers, torsional vibration “dampers” or “balancers” are attached to shafts of internal combustion engines, usually on the fly-wheel, which serve to reduce oscillation and vibration. So far as known, the modern airplane engine could not operate at presently required speeds in the absence of an efficiency device such as defendant employs.

It further seems uncontradicted that the art has long known three species of balan-cers or dampers. The three species referred to may be designated as friction, spring, and centrifugal pendulum dampers. The difference in achievements of each of the three is largely one of degree.

Friction dampers or balancers are discussed fully in the opinion of Clark v. Wright Aeronautical Corp., 2 Cir., 162 F. 2d 960, in which the court made clear the deficiency in this sort of balancer arising from friction which greatly detracts from the ultimate efficiency of such device, and held that friction devices did not anticipate later comparatively frictionless balancers and that their suggestion and adoption fell far short of furnishing the solution to the problems arising under modern high speeds.

In the second type of damper, the relatively free movement between the shaft and the masses of the balancer is resisted by springs which transmit to the weights the disturbing forces due to torsional vibrations of the shaft and transmit back to the shaft reaction forces. Such a device has a fixed natural frequency of vibration of its own and accordingly can eliminate only vibrations which have that same fixed frequency, and cannot eliminate, at all speeds, vibrations which vary in frequency with the changing speed of the engine. Spring bal-ancers have had only limited commercial use.

The third type of damper is that of the centrifugal pendulum damper or balancer, such as that prescribed by Salomon and those manufactured by Wright or Allison.

A centrifugal pendulum balancer consists of a weight suspended from and carried by a revolving shaft in such a way that it moves about a center eccentric to the shaft, so that when the shaft is rotated and goes into twisting vibration, and the weight swings out of its mid-position (in which it lies farthest away from the shaft), as a result of the vibration of the shaft, centrifugal force moves the weight back toward mid-position, just as gravity moves the ordinary clock pendulum back toward mid-position. It involves little friction and no springs. While a clock pendulum is periodic in its movement, i. e., makes each complete swing back and forth in the same time, a pendulum balancer is not periodic; rather, it is aperiodic, because it does not always make a complete swing back and forth in the same period of time. This is necessarily so because the period of swing depends on how fast the shaft is rotated. If it is rotating rapidly the balancer must swing back and forth in a much shorter period than when the shaft is rotating slowly. Salomon and some of his predecessors disclose such devices and both plaintiffs and defendant avail themselves of their use, but in a different manner.

The art has known for years both unifilar and bifilar pendulum balancers. A clock pendulum is said to be unifilar because it is suspended by one thread from one point. In this art it means any construction in which the bob weight swings or oscillates about a single point. The bifilar differs in that it corresponds in some respects to a pendulum in which each of two points of the swinging weight is suspended by a thread. In a unifilar pendulum, the unattached end of the one thread or cord travels through an arc while in the bifilar *639pendulum construction, such as the lawn swing, the platform does not tilt during the swinging but remains on even keel. Defendant’s balancer is a bifilar pendulum device and the balancer weight has the same sort of swinging movement as the platform of the lawn swing except that in actual operation the whole balancer is rotating with the engine crankshaft at the same time it is swinging. The devices of the prior Chenard and Walcker French patent No. 632,017, and the Sarazin patents are likewise bifilar.

It was largely upon this feature of mechanical contrivances that the issues turned in the District Court. We well know that the center of gravity is a point about which a body or mass is in balance. If the mass is pushed in line with the center of gravity, it will move in a straight line and not rotate ; but if it is pushed out of line with its center of gravity it will tend to rotate and, when it moves, it tries to go in a straight line. This is because of centrifugal force, which may become so powerful as to cause a fly-wheel to burst into pieces and fly in all directions. A pendulum balancer brings-the pendulum weight back to its mid-position just as force of gravity brings the pendulum back to its mid-position in any centrifugal pendulum. The device of the patent is carried on a revolving shaft and is-so constructed that centrifugal force necessarily holds it in mid-position in which its-weight lies farthest from the shaft. When the weight is set in motion by some disturbing force, it swings inwardly toward the center of the shaft and then the centrifugal force causes it to swing back.

Claims 3, 6, 7, 13 and 18,1 are copied in the margin, and are fairly representative of all claims in suit.

Salomon entitled his first application, on March 31, 1933, as an “Improvement in Devices Adapted to Eliminate Oscillation.” He stated in his application that he had invented certain new and useful improvements in devices adapted to eliminate or to *640reduce speed oscillations, vibrations and jerks (our emphasis). The stressed words were cancelled and in paragraph 5 of the present application appears the following:

“My invention provides the elimination of * * * oscillations, of vibrations and jerks through the mere play of transformations of kinetic energy * * * exerted on >:< * * 'filtering masses’ without * * * pivotal connections, springs, brakes or stops of any kind, so as to reduce to a minimum the friction arising.”

Here is a clearly expressed intention on the part of the patentee not to reduce to a minimum, but to eliminate oscillations, vibrations and jerks by reducing to a minimum, not by eliminating, the friction thus arising. It is under these circumstances that appellants urge that Salomon was the first to disclose that all oscillations could be eliminated by eliminating all friction. This suggested remedy may be a correct solution. However, this record discloses that no one, including plaintiffs and defendant, has ever been able to eliminate all friction, and for that reason, no doubt, no one has ever been able to eliminate all oscillations, vibrations and jerks. Certainly, neither Salomon nor Allison has done so.

The original application, as well as the present one, informs us of the intricacy and lack of efficiency of springs and fluids for such purpose because of their frictional action, which the applicant states was well known to the art. True, he does refer to “reduction” as well as elimination of friction in paragraphs 5 and 9 of the specification; however, the title in the original application, as well as in the patent, remains the same.

It must be remembered that an invention, in order to be patentable, must consist of more than an idea, or a mere thought. It must have been reduced to practice, and this Salomon has not done. According to the evidence here, those versed in this art, including the patentee, were well aware, at the time the original application was filed in this case, that springs and fluids should be eliminated, because the use of either would cause friction.

Salomon in both his original and his present application, stated that all these things could be accomplished by his unifilar disclosure, and he said nothing whatever with respect to a bifilar pendulum. However, Salomon and appellants urge that the patent disclosures cover a bifilar pendulum. We here place in juxtaposition Salomon’s Figure 4 and one of defendant’s damper weights as illustrative of their disclosures:

One of Six Aliison Damper Weights Taken from Deft’s £xh. HH.

*641All claims in issue except 13 and 18 are applicable to figure 4, of which claims plaintiff says claim 6 is typical:2 The numbers on Figure 4 refer to the following elements mentioned: 1 — shaft; 6 — cylindrical stubshaft, around which is arranged a torus-shaped recess, or guideway, limited by 6 and 7; 8 — a hollow roller or damping mass, with a very small play having a solid axis 10, around which is another torus-shaped guideway wherein a ball, 9, which is another damping mass, may move with a very small radial play. Six of the devices, as shown by Figure 4, encircle the engine shaft, as shown by Figure 8. He says this compound roller and ball system, which he refers to as a double pendulum, is equivalent to a bifilar pendulum. However, he adds that the use of element 9 is optional.

In the Allison device the letters refer to the following elements: A — damping weight or mass; BB — pins or rollers; C — flange; DD — holes in and through A and C. This weight does not move in any guideway or recess. It is pivotally suspended, by means of two pins or rollers BB from a circular web or flange C. The pins pass through holes DD in the flange and through corresponding holes in the damping weight. As in the patent, six of Allison’s devices, as shown by his Exhibit HH, likewise encircle the engine shaft. It is not denied that this constitutes the well known bifilar suspension, which had been known to this art for many years prior to any disclosures which Salomon ever made. He admitted this in the application, filed January 14, 1939, for his United States Patent No. 2,316,288, wherein, to corroborate his statements to that effect, he refers to the French bifilar patent, issued to Chenard and Walcker, No. 632,017 in 1927.

It would seem that the charge of infringement here made is largely if not entirely based on appellants’ characterization of one of Allison’s pins or rollers BB, as a damping weight and as equivalent to Salomon’s internal damping masses 8 and 9 in Figure 4. The District Court, however, found that the operation of Allison’s device is dependent upon both pins being present and would be inoperative if one were removed, and further that those pins do not perform any useful damping function in the device.

It is conceded by appellants’ expert that every figure in this patent in which any physical construction is shown embodies a “stub shaft.” See Fig. 4, Nos. 6 and 10. Stub shafts were not shown in most figures of Salomon’s corresponding French patent; however, as the District Court found, they are characteristic of the disclosures here. They are not used by Allison. The supposed advantage of them is to prevent the balls or rollers from dropping across from one side of the guideway to the other when the fly-wheel slows down to a point where centrifugal force is no longer effective to hold the damping weights in their intended effective places. To accomplish this, however, is to cause friction and to limit the damper weight to less than half the diameter of the cavity in which it is contained, because the stub shaft is always in the center. The uncontradicted evidence shows that a balancing weight of adequate mass to be of use in Allison’s engine could not be installed in Salomon’s structure, because of the presence of stub shafts.

It seems to be true that Sarazin, in 1931, was the first to disclose a bifilar pendulum to dampen torsional variations of a shaft, being upon hinges or moving on rollers. However, he confessed in his later patents that his earlier patent was a failure because it generated, not some friction, but too much friction. Clark, Attorney General, v. Wright Aeronautical Corporation, 2 Cir., 162 F.2d 960, 966. The court there said: “Even though in the case at bar we were to concede that the finally successful damper did contain those elements which Sarazin first brought together, it would not follow that he (Sarazin) deserv*642ed a patent, for he coupled them with another element which undid their combined value.” So is it with Salomon, in the case now before us. Appellants urge with much seriousness that Salomon in this disclosure was the first to discard what had previously been thought essential friction, whereas those skilled in the art taught that some friction was necessary, and was required in order to accomplish the desired result. A study of this very interesting subject convinces us that if any appreciable number of those skilled in the art ever taught the necessity of some friction in order to eliminate all oscillations (and we find none who do), it was done on the theory that some friction was unavoidable, and for that reason it should be reduced to a minimum. That is precisely the teaching of Salomon in this patent, yet, like Sarazin, he adds the use of other elements such as solid, granular and fluid masses each of which the art, including Salomon, then well knew causes friction. He also uses stub shafts which the court also found produce friction, and that finding is substantially supported by evidence. He discloses and describes both high and low friction and he fails to differentiate between them.

We think the patent discloses no bifilar pendulum, but it discloses two separate unifilar pendulums, one inside the other, which Salomon refers to as double pendulum. Bifilar does not mean two pendulums. It means two cords or filaments by means of which each bobweight, or dampening weight, swings. Salomon does not disclose such a construction, nor does he claim to do so. Moreover, he makes the use of ball 9, in figure 4, optional, which if omitted would completely destroy any basis for contention that such construction contains a bifilar pendulum.

The undisputed evidence discloses that neither Salomon nor any of his agents in procuring this patent, or in the trial of this cause in the District Court, ever expressed an appreciative word for a bifilar pendulum. Plaintiffs’ expert witness testified that the patent did not cover the use of a bifilar construction. Salomon was not present at that trial. However, his agent, Rubisow, Who was also the agent and promoter for plaintiff Redynam in practically all the countries of the world except France, testified at the trial and engineered the prosecution of this patent application to a successful termination in the Patent Office. There he was confronted with the Carter British patent to which he responded: “Sal-omon is not interested in the bifilar suspension. He has a much more simple solution than the ‘bifilar’ for the reduction of 'high frequency oscillations. Therefore, he quite willingly leaves the bifilar suspension to others; that is a useless and injurious complication.” To the Commissioner he also stated: “Salomon, in order to reduce the friction as much as possible, has his multiple masses roll themselves on the guideway without any intermediary parts such as rollers, cylinders or articulations.”

To the citation of Sarazin he said: “Sarazin never has his movable masses themselves roll on a guideway without any rollers, which is one of the characteristics of the Salomon movable masses.” This statement was made to the Commissioner on April 6, 1937, and just two days later, by supplemental amendment, Salomon for the first time, presented any of the claims in suit. He then presented claims 3, 4 and 5 of the patent, in an admitted attempt to put in claims somewhat broader than any theretofore allowed, but still on the basis of what Rubissow had said to the Commissioner and within two days after his submission. Nothing that Rubissow said to the Commissioner was ever retracted or corrected. All other claims in suit were inserted still later, and none were presented or allowed except upon the basis of, and after the making of, those representations to the Commissioner. The record shows that both Salomon and the plaintiffs’ officers had reprimanded Rubissow for what he had there said, but neither of them made any effort to correct it. From this it would seem that the patentee and his agent thus made an interpretation of his disclosures and claims which the patentee will not be permitted to ignore. Dwight & Lloyd Sintering Co. Inc. v. Greenawalt, 2 Cir., 27 F.2d 823; Atkins v. Gordon, 7 Cir., 86 F.2d 595; and Thomas v. Simmons Co., 7 Cir., 126 F.2d 743. By this action it seems clear to us that the patentee and his *643representative took advantage of the ambiguous wording to insert the claims against bifilar constructions, suspended on rollers, in order to secure the claims over the Sarazin reference, just because he told the Commissioner that the patent or the claims did not cover them.

Substantially the same representations were disseminated in thousands of letters throughout the industry by Rubissow acting as Salomon’s patent promoter and as agent for the Swiss plaintiff and its alleged French or Luxemburg predecessors. These circulars were widely circulated as a part of Rubissow’s promotion plans in attempting to sell licenses under the Salomon patents, for which he received a 35% commission on all royalties collected. After the examination of Rubissow by the court, it indicated very strongly that Rubissow was not greatly concerned about the truth of any statement he made to any customer of the Salomon patent. The record further discloses that Salomon was at all times supplied with copies of the material put out by Rubissow. The only answer to such contention is that Allison could not have been greatly misled because it was already in production.

It is quite true that Allison was already in production “in ever increasing quantities,” in 1937, which went not only to the British war effort but also to the United States war effort in the form of production for its Army Air Force in some of its leading fighter planes. By reason of this fact, if the patent were valid and infringed, the United States Government would be subject to suit in the Court of Claims as a contributory infringer. This fact alone, of course, is no more determinative of this issue than it would be determinative in the case of any other contributory infringer, but it is entitled to consideration, with all the other evidence, in determining the credibility and the motive of the witnesses participating therein.

One of those letters was received by the appellee, General Motors. It was written February 7, 1938, more than four years after this complaint was filed. It covers four printed record pages and contains the following language:

“I beg to inform you that I am the exclusive representative for the negotiation throughout England, America and other countries, of the license rights regarding the * * * Salomon patents. * * *

“These patents fully control the market of dynamic dampers of any kind using free masses, rolling * * * substantially frictionless * * * or the like, acting under the influence of restoring forces provided by centrifugal forces. These dynamic dampers * * * eliminate all vibrations, jerks and oscillations * * *.

“ * * * If you are interested by the ‘bifilar suspension’ we would draw your attention to the fact that this form in itself cannot be validly protected being anticipated by the Chenard Walcker patents published in 1927 (No. 604,040 France and No. 458,463 Germany) which patents are expired.

“The licenses which we grant give full and valid protection in accordance with the Salomon patents.” (Our emphasis.) It is quite evident that the Chenard-Walcker French patent just referred to is erroneously numbered No. 604,040 and should be No. 632,017. The latter fully corresponds in all other respects with Rubissow’s description, and is relied upon by appellee.

We think that none of the claims in issue read upon defendant’s bifilar construction either verbally or substantively.

From the evidence it clearly appears that no testifying expert ever saw a construction of Salomon’s disclosures under this patent alone. It is also clear that Salo-mon does not recommend the use of a bifilar construction. It would seem equally clear that appellants have no desire to use a bifilar system. Their apparent intention is to continue to use the unifilar construction, and prevent their competitors from using the bifilar, which is in no sense an equivalent to the unifilar. Appellee is not here seeking a patent on the bifilar; it is merely defending against a charge of infringement of the unifilar.

Judgment affirmed.

LINDLEY, District Judge

(dissenting in part).

*644Defendant mounts on a disc affixed to the crankshaft six bifilar pendulum bal-ancers, the multiple number being so provided and designed as to eliminate vibrations of different frequencies. Each one of these pendulum rolls around with the disc and the crankshaft and, at the same time, oscillates to and fro according to the laws of centrifugal force. For each balancer two holes about 2 inches apart are drilled in the disc and two corresponding holes of oversize and larger than those in the disc are drilled through the balancer. In each of the two holes on each disc is placed a steel roller and on these two rollers the balancer’s two over-size holes are hung, thus affording opportunity for oscillation of the balancer on the two suspension points of the pendulum. Thus I think it is undisputed that each unit constitutes a bifilar pendulum balancer, which rolls and oscillates freely without substantial friction. I think it perfectly obvious from examination of this device that it works according to the same physical law, in the same mechanical manner and produces the same result as the pendulum rolling weight disclosed in the Salomon patent. The result is a substantially frictionless centrifugal pendulum balancer. It includes pendulum masses which roll on rollers along arcuate guideways eccentric to the axis of rotation of the main shaft, relatively free of friction. This damper, as Salomon’s, balances its vibratory thrusts against those of the disturbing impulses. Each is capable of being tuned to an order and automatically remains tuned to that order at all speeds of rotation of the crankshaft. Each is substantially frictionless. Defendant, I think, employs Salomon’s means for damping oscillation in a rotatable shaft, “comprising a part fixed to and rotatable with such shaft; said part being provided with at least one arcuate guideway the center of curvature of which is eccentric to the axis of rotation of said shaft and a structurally integral damping mass freely movable substantially frictionlessly along said gu'ideway, in the direction of and in response to shaft vibrations and in response to centrifugal force” in the words of Claim 8. The device seems to me equally applicable to the language of the other claims in suit.

The District Court reached its conclusion of no infringement upon the sole ground, apparently, that Salomon is limited to a unifilar construction and that, inasmuch as defendant employs a bifilar construction, it does not infringe. It was the difference between bifilar and unifilar construction which the court believed to be decisive. But I can not agree that the language of the claims justifies a conclusion that Salomon’s invention was limited to unifilar construction and that, therefore, any device which did not employ unifilar construction could not infringe. It is clear that the use of both unifilar and bifilar pendulums was old and that any engine designer, in deciding what balancer he would build, had the option of choosing either bifilar or unifilar pendulums. This was the skilled workman’s choice, well known to the art. Salomon’s invention lay essentially, not in the choice of one of the two kinds of suspensions, but an entirely different thought, the free rocking substantially frictionless movement, whichever type was used. He was not inventing a bifilar or a unifilar balancer. He was prescribing the essential elements an efficient and successful pendulum balancer should have. Invention could not lie in the choice of the one of the two old forms of suspension, and Salomon did not pretend that his invention lay in any such prescription. It lay entirely in his inventive idea applied to either form of pendulum. This is clear from the contents of the file-wrapper. True it is that Salomon, through his agent Robissow, expressed disbelief in the efficiency of the bifilar construction, but he complained at all times that devices which employ a bifilar construction infringe. “It is,” he said in effect, “an infringement, whatever method of suspension or whatever principle of suspension is involved.”

The trial court found also that Chenard in his French patent of 1927 discloses “a substantially frictionless bifilar pendulum.” I think the Chenard patent omits any teaching of the necessity of the reduction of friction to a minimum and, indeed, *645Judge Bright, in the Sarazin case, in D.C., 54 F.Supp. 244, 247, expressly so found, and the Circuit Court of Appeals, on appeal in Clark v. Wright Aeronautical Corp., supra, approved. In view of those decisions, I shall not elaborate upon their reasoning in this respect except to observe that Judge Bright said, in Sarazin v. Wright, “it depends upon friction for its reactional effect.”

The District Court found that Carter’s British patent of 1930 discloses a construction “substantially like that of the Pratt & Whitney engine put out prior to 1935.” But Carter’s idea was to obtain friction. Indeed, he said the desired effect could be obtained by “solid or fluid friction” or by internal friction. Frictional restraint is a part of every form of embodiment suggested by Carter without any hint of tuning the balancer to suppress any particular kind or order of vibration. The complete specifications clearly show that he was working solely in the art of friction-impeded pendulums, none of which achieved the results of Salomon. Judge Bright said that his patent clearly taught friction as the solution, and this interpretation, I think, was correct. The prior art disclosed both unifilar and bifilar devices, but none of it revealed a substantially frictionless damper, such as Salomon’s or defendant’s.

The trial court found that Salomon’s invention is “characterized” by the presence of a stubshaft which so limits the mass of the roller as to be incapable of accomplishing the results accomplished by defendant’s bifilar damper. I think this is a gratuitous finding beyond the scope of what is essential to the achievments of Salomon’s patent. The so-called stubshaft appears in the figures accompanying the patent, which show a round piece of material in the center of the track around which the roller travels in its free movement. The roller is not in contact with the stubshaft in normal operation and the latter serves only to prevent the roller from falling against the walls of the container whenever the motor stops. There is a small clearance between the central part, the so-called stubshaft, and the periphery of the container and the roller is smaller than that clearance so that there is sufficient leeway. The stubshaft may serve a useful purpose, but it is unnecessary to a complete realization of Salomon’s invention and detracts in no wise from rolling or rocking freely and unrestrictedly on a curved guideway. Defendant’s expert witness testified that it would not require invention to remove the stubshaft, saying that it would depend upon engineering judgment or mechanical skill. The stubshaft is described in Qaim 12 but is omitted from other claims. I think it is not an essential element in Salomon’s set-up. Furthermore, “where a patent contains both a broad and a narrow claim and suit is brought on the broad claim, we cannot construe into it a limitation not therein expressed, but which is expressed in the narrower claim and by which alone one is distinguished from the other. To do so would be making over the contract between the public and the patentee.” National Tube Co. v. Mark, 6 Cir., 216 F. 507, 521.

I reach the question of estoppel. Whatever Rubissow said, I agree, the patentee became responsible for. Though Rubissow testified that he was reprimanded for acting beyond the scope of his authority, he was permitted to appear and act for Salo-mon. Secret limitations upon his authority in representing his principal may not be invoked against third persons with whom he dealt in pursuance of his apparent authority. Indeed, plaintiffs, as I understand their position, do not contend otherwise. So Salomon is bound by whatever Rubissow said in his transactions in behalf of his principal.

Fortunately all of Rubissow’s statements upon which defendant relies were in writing, and this court’s duty in this respect becomes one of determining from the four corners of the documents just what he conveyed by express statement and inevitable implication.

In effect defendant contends that Rubis-sow effectually limited Salomon’s invention, in the proceeding before the patent office and in communications to defendant, to devices of unifilar character and, for all practical purposes, disclaimed bifilar devices. In doing so, defendant relies upon such express statements as: “Mr. *646Salomon is in no way interested in the principle of twofold suspension, which, in his opinion is uselessly complicated * * * is not interested in bifilar suspension, * * * has a more simple solution * * * for the reduction of high frequency oscillations, * * * leaves the bifilar suspension to others,” as “a useless and injurious complication.” Standing alone these quotations from voluminous documents might well sustain defendant’s contention. But the precise intent of Ru-bissow and its effect in patent law must be deduced from an examination of all he said. In all his writings Rubissow was insisting that Salomon was the first to teach and put into practice the thought “that the use of free or quasi-free suspended masses” successfully reduces to a minimum “oscillation vibration and torsion irregularities in rotary shafts axial or torsional * * * by means * * * of the centrifugal forces acting with” or “applied to said masses” during the motion of the shaft. This, said he repeatedly, was without restraint by springs or connecting rods so that the masses moved freely with a minimum of friction. “Each rolling body,” said he, “rolls * * * on a guiding surface which is solidary with the shaft” and “constitutes a rolling surface for a second body.” This teaching of Salo-mon, Rubissow expressly claimed, was infringed by Sarazin, Reed, Wright, His-pano, Taylor and Chilton, all of whom, the record discloses, employed bifilar suspension devices. Directly following his statement that Salomon preferred a unifilar construction, he added “But Salomon attacks Wright and Hispano because they obtain their suspension by using the fundamental characteristics of the Salomon patents.” It seems perfectly clear that Rubis-sow was continually and constantly insisting that the bifilar devices mentioned infringed Salomon not because they were bifilar in character, but in spite of that fact, and because, in each of the bifilar devices, the designer availed himself of Salomon’s elemental discovery. He pointed out that use of bifilar devices was old in the art. In effect, he said that Salomon did; not object because the builder employed bifilar construction, for he (Salomon) believed that the choice between unifilar suspension and bifilar suspension was one within the option of the skilled designer; that both methods were old, but that if the designer of either a unifilar or bifilar construction employed Salomon’s essential principles, he infringed. He attributed some such language as this to Salomon. “I don’t care which kind of suspension you employ, uni-filar or bifilar, but if you employ my inventive idea, you infringe.” This seems perfectly obvious from the documents in evidence. And I agree that bifilar suspension was old in the art, that anyone was free to employ it and that if the skilled designer saw fit to employ bifilar suspension instead of Salomon’s preferred unifilar idea, he could not escape infringement, if he made use of Salomon’s essential inventive idea as expressed in his claims.

This, I think, is a far cry from estoppel; it possesses none of the essential elements. Nor can it be interpreted as a limitation of the claims so as to include only unifilar devices and to exclude those which employ bifilar suspension. It is merely a statement that Salomon does not consider the choice between the two methods of construction an essential element of his invention.

The District Court found that all claims in suit were inserted in Salomon’s application after the accused Allison damper was being manufactured and sold. However, the court did not find that the substance of these claims was then asserted for the first time. Upon examination of the file it appears that the essence of each claim in suit appears in the claims as they were originally filed; that there was continuity of the same inventive idea from the time of the filing of the original application until the patent finally issued and that there was no material deviation from or expansion in the subject matter. Consequently, interpretation, as well as validity, is to be tested as of the time of the original application. Crown Cork & Seal Co. v. Ferdinand Gutmann Co., 304 U.S. 159, 58 S.Ct. 842, 82 L.Ed. 1265; Ceramic Process Co. v. General Porcelain Enameling & Manufacturing Co., 7 Cir., 129 F.2d 803.

I would reverse the judgment with directions to find the claims valid and infringed and to order an accounting as prayed.