Clark v. Wright Aeronautical Corp.

L. HAND, Circuit Judge. Sarazin, a French citizen, and his exclusive licensee, Societc Francaise Tlispano-Suiza, a French corporation, the original plaintiffs, began this action on September 21, 1938. Judge Bright tried the case in April, 1942, and on May 29, 1943, Leo T. Crowley, then Alien Property Custodian, by the usual “vesting” order, transferred to himself all right, title and interest to the patents in suit, including the cause of action at bar. James E. Markham, Alien Property Custodian, succeeded Crowley; and the Attorney General has succeeded Markham. Judge Bright dismissed the complaint by judgment entered May 29, 1944, and Markham appealed on August 17, 1944. We shall disregard the formal plaintiff and the licensee, and speak of the case as though Sarazin, the patentee, were the only plaintiff. Thus the action becomes the usual one for infringement of two patents, issued to Sarazin: the first, Patent No. 2,097,227, on May 4, 1937, of which claims 11, 12, 13, 14, 15 and 16 are in suit; and the second, Reissue Patent No. 20,773, issued on June 28, 1938 (the original having been issued on May 4, 1937), of which claims 5, 6, 7, 8, 9, 10, 12, 13, 14, 15 and 16 are in suit. Judge Bright’s opinion is reported as Sarazin v. Wright Aeronautical Corp., D. C., 54 F.Supp. 244; he has stated the issues and facts so fully that it is unnecessary for us to repeat them. In what we say we shall assume a familiarity with his opinion. Sarazin filed his original application out of which both patents arose, on July 20, 1931; he had already, on December 19, 1930, filed a French application for what is now Patent No. 2,079,227; and on June 30, 1931, he filed another French application for what is now Reissue Patent No. 20,773. In his application in the United States he combined these two French applications; but on August 2, 1932, the Patent Office ordered the application to be divided into two parts, corresponding to the original two French applications. The first question to be considered is of the scope of the original joint application in the United States. The two separate forms disclosed — the “Link Type” and the “Baby Carriage Type” — are sufficiently described in page 245 of the opinion, taken in conjunction with Finding's 4, 5 and 6 which we quote in full in the margin.1 We may [962] proceed at once therefore to a consideration of whether the disclosures contain any germ of the notion that friction should be eliminated from the “damping” devices disclosed; and whether they, are effective, if it is not. Although in the original application Sarazin disclosed and described three figures, by which friction could be added in the “Link Type,” he now relies particularly upon the fact that these are stated only permissively- — they “may be” used. It has become a habit of the scriveners of patent applications to use the permissive form, presumably lest they should too much limit the invention; but in doing so they risk impaling themselves upon tlie other horn of the dilemma, because the specification must contain adequate instructions for the practice of the invention, and permission may easily become imprecision. For exa'mple, in this very application, the provision for friction devices was immediately followed by the phrase: “Said masses may be likewise radially movable, so that their centers of gravity may move towards or away from that of the shaft”: an inevitable result of all “bifilar” pendulums in action. Yet “bifilar” pendulums are Sarazin’s especial reliance.- We do not of course mean that an applicant may not present alternative embodiments of his invention, when the essential elements are plainly stated; but there is not the slightest justification for arguing that, because friction devices were permissively added, they are to be treated as in antithesis to frictionless devices. Frictionless dampers were not suggested in the text, and what turned out to be destructive friction was inevitable in all the figures. Reading the ■ specifications alone, no one would imagine that the absence of friction was a condition of every effective damper; and this conclusion is confirmed when we turn to the original claims. Claim one, as originally filed, was as follows: “Means -adapted to reduce the torsional oscillations of ■ crank-shafts, characterized by the combination of interconnected masses adapted to move in. angular relationship relatively to the shaft with which said masses are connected and around which constitute a fly-wheel having negative inertia.” Claim 2 was more specific, and required the connection of the “masses” with the shaft to be “links being themselves hinged to branches keyed upon said shaft.” Read in abstractu, it is true that these two claims are consistent with a frictionless damper; but it is horn-book law that no claims are to be read in abstractu. The next five claims all contained provisions for added friction means to the action of the “masses”; they read only upon figures 2, 3 and 4 of the application and of the “Link Type” patent. The next five claims read upon figures 5, 6 and 7, of the application; they were for the “Baby Carriage Type.” Claim 8 was as follows: “Means as claimed in claim 1, wherein the mobile masses move within a hollow member integral with the shaft so that projecting bosses (preferably fitted with rolls) mounted terminally on said masses are constantly applied by centrifugal effect against hollows provided in said hollow member and that the difference between the radius of each of said hollows and that of the projecting boss appertaining thereto should constitute the radius by which said mass moves, said radius being as s.hort as may be desired.” Each of the other four claims incorporated all the elements of claims 1 and 8, thus including the words in parentheses. Taking the application as a whole, it is apparent that Sarazin, not only had no notion that it was essential to eliminate friction from his damping devices, but that he presupposed its existence. Not only did he actually add friction devices in five of his “Link Type” claims; but in all his “Baby Carriage” claims he expressly declared that rollers were not necessary, which would result in a sliding contact between the boss and the arcuate “hollows” of the member 19. Claims 1 and 2 — the only ones which by any possibility could be made to serve — must be read upon their disclosure and the disclosure was of de [963] vices which in operation discovered an amount of friction which turned out to be prohibitive; as Sarazin himself learned and very frankly declared, in two French patents which he took out later. He filed his application for the first of these on May 22, 1934, in which he spoke as follows: “Contrivances of this kind are known which are constructed in the following manner: the movable masses are attached by means of small connecting rods, which are themselves articulated on the one hand to the shaft and on the other band to the movable mass, and which are able to oscillate freely around the attachment pivot. The centrifugal forces acting upon the movable masses during the rotation of the shaft occasion in the articulations strong pressures upon the surfaces sliding upon one another, so that the frictions produced at these places prevent more or less the free oscillation of the masses and thereby the anticipated damping effect.” Figure 1 was in substance the same as figure 1 of the original disclosure of the patents in suit; and of it he said: “The vibration damper illustrated in Fig. 1 has, keyed on the shaft 1, a support 2, to which the movable masses 3 are attached by mentis of small connecting rods 4. Each of these small connecting rods 4 has two bearings 5, one of which supports a pivot 6 that is integral with a support 2, and the other a pivot 7, that is integral with the mass 3. The surface of pivots 6 and 7 slides on the interior of the bushings of the bearings 5. The frictions will oppose thereto a resistance to the pendular movement of the movable masses. The centrifugal forces acting upon masses 3 are such that the forces transmitted by the small connecting rods are a multiple of the weight of the moving masses; the surface pressures produced upon the bushings 5 will be sufficient to brake the pendular movement, so that the damping of the torsion vibrations of the shaft 1 cannot be effected.” He filed his application for the second French patent on January 7, 1935, and in this he said that the invention related “to a contrivance for reducing vibrations by means of masses connected in a movable manner to the vibrating part, and it consists in the fact that the connection between the mass and the vibrating part is effected by means of at least one rolling body, rolling on the one hand on a race that is connected to the vibrating part and on the other on a race that is connected to the mass.” “In the known contrivances” on the other hand “the centrifugal forces which act during the rotation of the shaft bring about, in the articulations, high pressures of the surfaces sliding upon one another and, by reason thereof, forces of friction which disturb or even prevent the pendular movement of the masses. The damping action which is sought is no longer obtained at all.” A similar, though less explicit passage also occurred in Sarazin’s German patent which was published on August 12, 1941. To this Sarazin’s attorneys now answer that the language may have been that of his French attorneys, not his own, because he did not sign the application. This evasion we cannot accept: the notion that attorneys should have volunteered so to discredit their client’s earlier patents is important only for its hardihood. Possibly, it would be unfair to use these admissions as evidence that the patents in suit were absolutely inoperative; but they are the strongest possible support for Judge Bright’s findings 8 and 10 which we quote in the margin,2 and which we accept. In conclusion we hold that the patents in suit contained no suggestion of avoiding friction, and that Sarazin at the time had not the least intimation that it was important to do so; but that he labored under the common mistake of the art that friction might [964] be an advantage. There was the amplest support in the record for the 7th Finding that “For a centrifugal pendulum damper to work efficiently, it should be as nearly as possible frictionless, i. e. substantially without friction. A centrifugal pendulum damper which operates with substantial friction will not deliver counter impulses which are equal in force to and 180° out of phase with the torsional impulses applied to the shaft, both requisites for efficient damping of the torsional vibrations.” With these three findings in mind, we proceed to a consideration of the art, when Sarazin filed his original application on July 20, 1931. In 1911, one, Frahm, a German, obtained a patent for damping the “resonance” vibrations which arise in bodies subject to periodic impacts. It is not necessary to go into the details except to say that he did this by means of an auxiliary body which was caused to vibrate by the vibrations of the main body “in resonance” with the vibrations of that body, and that this was done by a unifilar pendulum, tuned to the vibrations which were to be corrected. In the same year the well-known Siemens Company obtained a German patent (No. 272,575) which can best be understood in its Austrian equivalent, filed two years later (No. 72,007). This device was to cancel the responses of power machines receiving periodic increments of driving power; it made use of the principle, “well known in the case of the double pendulum, that a movable mass whose driving force fluctuates rhythmically does not follow these fluctuations if there is attached to it, so as to be capable of swinging freely, a second mass whose natural oscillations occur in the same rhythm with the oscillations of the driving force, i. e. are in resonance” with it. The Siemens damper consisted .of a curved glass tube with bulbs at either end, fixed to “the rotating part of the machine” near its periphery. In one of the two forms disclosed a second straight tube also connected the bulbs; and in both forms the curved tube contained an optional valve to control the flow of liquid through it. The 'tube and bulbs were partly filled with liquid, which the centrifugal force of rotation drove into the curved tube and filled it, also partially emptying both bulbs, so long as the rotation remained uniform. When the uniformity of rotation was interrupted by an impulse, received by, for example, a shaft, the inertia of the liquid caused it to flow from one bulb to the other, thus tending to check the added increment of rotation. If the return of the liquid was timed so that it would fill both bulbs equally by the time a second impulse arrived — that is, if the damper was “tuned” to what Sarazin chooses to call the “order” of the impulses — the result would be in part to cancel the impulses at all “frequencies” including those which correspond to the “natural frequencies” of, again for example, a shaft. This completely realized Sarazin’s object, although obviously the means employed were toto coelo different. The Siemens disclosure prescribed that “impulses proceeding from” the oscillation of the liquid in the tubes should “exactly counteract the impulses of the driving force and keep in equilibrium with it at every moment.” To accomplish this end “means are further provided for adjusting exactly the duration of oscillation of the columns of liquid and for making it variable within certain limits” (i. e. the valve in the curved tube). The defendant’s expert testified that it was possible by manipulation of the valve to fix the periodicity — “frequency”—of this flow and reflow of the liquid. It is clear that the liquid would respond to each impulse, and the defendant’s expert also testified that the amplitude of the oscillation has no relation to the frequency “within reasonable limits”; and that at all speeds each impulse would provoke its cancelling response, just as in Sarazin’s disclosure. The 30th Finding accepted as true that part of Siemens’s first claim which disclosed “that the natural oscillation proceeds in rhythm with the fluctuations of the power supplied to, or derived from, the machine.” It is certainly far from being “clearly erroneous.” In 1927 one, Chenard, filed an application which resulted in a French patent, No. 632,017, for a damper upon a shaft actuated by successive impulses of driving force. Figure 3 is the closest to Sarazin’s disclosure, and alone need concern us. It and its accompanying text disclose a forked member, made part of the lower end of the [965] crank arm, and having holes at each side through which axles were passed. The forked member embraced the two ends of a “mass” whose motion neutralized the irregularities in the motion of the shaft due to the successive impulses. In each end of the “mass” was a hole larger than the axle, and each axle passed loosely through one of the holes, and suspended the mass by both ends. Irregularities in the movement of the crank arm were damped by the inertia of the “mass” measured in amplitude by the looseness of its suspension. The patent speaks of the moving “masses” — including Figure 3 — as “aperiodic”; that is, as only returning to, and not passing, the position from which they were dislodged by the change in rotation of the shaft. We will so assume; that is, we will assume that the friction of the metal of the holes in the mass, riding over the metal of the axles would be so great as to exhaust the momentum of return at or before the mass reached its first position. On that assumption there was no need of “tuning” the movements of the “mass” to the sequence of the impulses, nor was anything of the kind described. Had the axles been loose rollers, however, as they are in Chilton’s device, the “mass” would have been “periodic,” and in that event “tuning” would have been necessary. The art had not yet learned that the absence of friction was a desideration; but, as has already appeared from the references, and, as also appeared in the testimony, “tuning,” when desired, was well understood when the art was dealing with a pendulum. Finding 42 to that effect has support in the evidence. In August, 1929, one, Carter, filed an application which resulted in British Patent No. 337,466 of 1930. This too was for damping or checking the torsional variations of a crank shaft, which came from the recurrent impulses of an internal combustion engine. This Carter did by a “mass” extending below the crank shaft, so that in rotation it wo-Id act as a counterpoise to the impulses received by the crank pin from strokes of the cylinders. This “mass” contained two circular cavities, one on either side of a line drawn vertically from the crank pin through the crank shaft; and in each cavity Carter placed one or more discs of a diameter less than the cavity which held them. Variations in the speed of the shaft caused by the successive impulses were taken up by the rotation of the discs in the cavities; but the damping was principally to be effected by filling the cavity, so far as possible, with some heavy liquid. Each disc “can slide radially and circum-ferentially relative to the crankshaft journals and can roll or completely rotate about on axis (Í. e. the axis of the disc or the axis of the cavity) eccentric to the crank shaft journals in the cavity” (page 4, lines 86-92). This was a “periodic” or pendulum movement, but it was damped by the liquid, and of course it was not a bifilar pendulum, as Chenard’s “mass” would have been, had the axles, 17, freely rotated in the holes in the lugs of the member 18. About two years earlier a German patent had been granted to one, Salmson, for a damper built upon somewhat the same principle as Carter’s. The “mass” was kidney-shaped and held loosely in a cavity in the extension of the crank shaft. A channel ran through it to make easier the flow of the oil which filled the cavity. Except for the presence of the oil the damper would not have been markedly different from Sarazin’s “Baby Carriage Type” with the rollers omitted, and, as we have seen, in all the original claims rollers were not necessary, but only “preferable.” This patent added little if anything to Carter. From the foregoing it is apparently true that Sarazin was the first to disclose a bifi-lar pendulum to dampen torsional variations of a shaft, hung upon hinges or moving on rollers. Had this occupied the field and driven out all others; had it been, as it stood, a great step forward, something might be said for it as a patentable invention. It was nothing of the kind. As we have seen, Sarazin himself confessed in his later patents that it was substantially a failure, because it generated too much friction. So far as appears, no machine has ever been made according to his disclosure. It is true that the Ranger Aircraft Company took out a license, and manufactures the “Baby Carriage Type”; but it tries, so far as possible, to eliminate friction by equipping the rollers with ball bearings, the direct opposite of eliminating rollers al [966] together and allowing the bosses to contact with the “arcuate paths.” In any event the license is slight evidence of recognition by the art. Licenses are repeatedly the result of a desire to avoid litigation; and in the case at bar the minimum royalty— $5000 for twelve years — although substantial, may well have been no more than the price of peace. How much above the minimum one half per cent upon the price of the engine would be likely to amount to, the record does not show. It was only when the art by experience came to realize that such devices will not accomplish their purpose unless they are freed from friction as much as is mechanically possible, that success arrived.- When that was grasped the rest followed; and perhaps ,Sarazin might himself have succeeded, if he had seasonably followed his two later French patents with United States patents. In dealing with the issue of invention, we have tried, so far as possible, to rely upon objective factors in preference to our a priori judgment, drawn from what seems to -our untutored experience to be within the range of a person skilled in the art. Instead of trying ourselves to mirror his capacities, we look to the length of time during which the incentive existed to contrive the invention, to the number of unsuccessful efforts that were made in that period, to the density — so to speak — of those efforts at about the time when the invention was. made, to whether success came independently to several inventors at about the same time, and to the extent to which after the invention appeared, it supplanted what had gone before.3 These usually are hard questions to answer; but when they can be answered, they form a substantial basis for inference. The test is always whether exceptional ingenuity was necessary to make the new combination out of the old elements; and the rationale is that what has for long escaped the quest of competent experimenters in the field, spurred on by hope of gain, and has been able to push out earlier contrivances, demanded talent out of the common. 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 deserved a patent, for he coupled them with another element which undid their combined value. Nobody can say that, if the art had realized that friction must be ruthlessly outlawed, it would have taken unusual powers to make Chenard, for example, into an effective damper; certainly it was far nearer to Chilton structurally than either of the patents in suit. It is true that courts have repeatedly said that success is not a reliable test of invention, and nothing can be truer, when taken without scrutiny and out of its setting; but failure is an almost infallible test of non-invention, and for all practical purposes Sarazin’s disclosures were failures, and indeed confessed failures. In the view we take it is not necessary to consider those claims which were added in the Patent Office after Sarazin learned of the Chilton machine. The inevitable logomachy between the ingenious and indefatigable solicitor and the badgered examiner is here irrelevant; as it almost always is. We will not go into the question of what time elapsed between the original application and the submission of the claims now in suit; nor will we rely upon any estoppel. The new claims we hold invalid, read as the plaintiff asks us to read them: that is, to cover dampers like the defendant’s dampers. When so read, they demand for their support a disclosure, not only not even intimated in the specifications, and unknown to the art at the time but an interpretation of it actually repudiated by the very inventor him'self. We hold invalid all the claims in suit. Judgment affirmed. Footnotes [1] “4. The object of both patents in suit is to reduce torsional vibrations in the crankshafts of engines such as internal combustion engines. A certain number of torsional imiralses are implied to an engine shaft during each revolution, depending upon the number of cylinders and their rate of firing per revolution. The ‘frequency’ of these disturbing impulses, i.e. the number per second, accordingly is equal to the number of impulses per revolution multiplied by the number of revolutions per second. The frequency of the disturbing impulses accordingly varies in proportion with the engine speed, and somewhere within the operating range of engine speed the frequency at which these impulses are being applied may bo equal to the ‘natural frequency’ of the shaft, i.e. the frequency to which the shaft most readily responds. When the torsional impulses are applied at a rate approximating the natural frequency of the shaft, a condition of resonance will occur which may cause serious damage to the engine unless the torsional impulses are counteracted. “5. The damper construction shown in the ’227 patent (known as the ‘link type’) consists of a plurality of pendular masses connected near their ends to links by means of axles journaled in the masses and in the links by friction bearings; the links in turn being connected by similar axles and bearings to a central member fixed on the engine crankshaft. The patent also teaches the addition of special friction-increasing washers or sleeves at the axles to damp the angular movement of the pendular masses. “6. The damper construction shown in the Reissue patent (the ‘baby carriage' type) also consists of a plurality of pen-dular masses which are loosely connected to a collar on the engine shaft to hold them in place. In one form the masses have at each end bosses on which rollers are mounted by axles journaled in friction bearings, these rollers engaging ar-cuate tracks in a retaining member se [962] emed to the shaft to rotate therewith. In a modified form the rollers are mounted by similar journaled axles on the said retaining member and the masses have tlie arcuate tracks engaging the rollers. It is taught that the bosses may bo used to engage the arcuate tracks, the rollers being omitted.” [2] “8. The axles of all of those !?arazin dampers produce high friction under the pressure of centrifugal force and none of said damper construed ions is of efficient use because of their excessive friction under the load of centrifugal forces on the masses, which will prevent efficient operation in damping the torsional vibrations. “10. Karazfii’s rollers with axles (his ‘baby carriage’ type) were not originally represented to he friction-reducing devices but only to enable a shorter pendulum length to be attained. The movement of Ms rollers on the tracks will not eliminate or reduce the friction of the axles in their bearings and these ¡Sarazin devices including the rollers with their axles will produce fiielion too great for efficient operation aiid neutralizaiion.” [3] Safety Car Heating & Lighting Co. v. General Electric Co., 2 Cir., 155 F.2d 937, 939.