Aetna Steel Products Corporation, Appellant, v. Southwest Products Company, a Corporation, Appellee, 282 F.2d 323 (9th Cir. 1960)

Miketta & Glenny, Los Angeles, Cal., for appellant.

Lyon & Lyon, Charles G. Lyon, Frank E. Mauritz, Los Angeles, Cal., for appellee.

Before HAMLIN and JERTBERG, Circuit Judges, and LINDBERG, District Judge.

LINDBERG, District Judge.

This is an appeal from a judgment and decree of the District Court for the Southern District of California, Central Division. The cause of action is one for alleged infringement of Potter patents numbered 2,626,841 and 2,724,172 for a bearing and its method of manufacture. Plaintiff-appellee, hereafter referred to as plaintiff or plaintiff-appellee, is a California corporation and owns the patents. Defendant-appellant, hereafter referred to as defendant as well as defendant-appellant, is a New York corporation and owns and operates Kahr Bearing (as a division) in Burbank, California. Jurisdiction of the trial court was under 28 U.S.C. § 1338, and the Patent Laws, Title 35 U.S.C. Jurisdiction to review the judgment and decree is conferred by 28 U.S.C. §§ 1291 and 1292 (4).

The bearings involved are known as self-aligning bearings and consist of a ball (truncated and containing a bore) and an outer race in which the ball element is rotatably held. Self-aligning bearings have great importance where it is difficult to obtain exact parallelism between a shaft and its housing bore axis, or where the housing may be deflected under load or while in use. These bearings, both those of the appellee and of the appellant, while not so limited, are manufactured for incorporation in aircraft. Because the fuselage of an airplane must be somewhat flexible during flight it is necessary that the bearings supporting work transmitting rods and levers must be self-aligning; that is, must allow this flexing and angle changing without binding.

Prior to the appearance of appellee's bearings the aircraft industry used primarily the Prentiss bearing, the Heim bearing and the Messerschmidt bearing. Of these, the Messerschmidt bearing was probably most suited to the need but the Messerschmidt bearing had an outer race which was machined and required a key or slot for the insertion of the bearing ball. This key or slot left an inherent weakness in the race and in some cases proved unsatisfactory. Also, being machined, there was too much tolerance between the ball element and the race. There was clearly a long-felt want within the industry for a simple, two-piece bearing, strong and durable.

Potter, what was familiar with the need, reasoned that it would be possible to swage, press or coin a cylindrical metal race of ductible metal around a highly polished ball of hard steel in one operation. The resultant race would be continuous without any slots. The ball element therein would be permanently locked within the race. If the proper tolerance could be achieved and the ball were truncated and contained a bore for receiving a work-transmitting shaft the need within the industry would be met.

With little or almost no capital Potter and others made certain experiments and found out that this could be done. It was also determined that while the pressing operation froze the race upon the ball, applying pressure to the median portion of the formed race would thereafter free the ball.

The bearing thus developed was immediately accepted in the airplane industry. Potter, believing that he had invented something new and useful, very soon made application for a patent. This was back in 1945. The first application covered both the completed article and the method of manufacture. Division between article and method was ordered by the Patent Office in accordance with 37 C.F.R. § 1.142. (See also 35 U.S.C.A. § 121 of present law and Rule 1.141 et seq. Rules of Practice U. S. Patent Office, 35 U.S.C.Appendix.) The applicant then cancelled those claims relating to method without prejudice.

The first application was filed as serial No. 606,678 on July 23, 1945 and matured into Patent No. 2,626,841 on January 27, 1953. Having cancelled method claims without prejudice, divisional application for the method claims was filed August 8, 1947 as serial No. 767,496. This application was ultimately turned down by the Patent Office and appeal therefrom while initiated was never completed. The Patent Office deemed the application abandoned June 11, 1951. A petition to revive was considered and granted for the purpose of perfecting an appeal. But because applicant failed to file a brief the appeal was dismissed October 1, 1951 by Order of the Board of Appeals.

Before the article application was granted, but after April 1, 1952, the date on which the two claims of Patent 2,626,841, the article patent, was allowed, another divisional application for the method claims was filed December 16, 1952 which ultimately matured into Patent No. 2,724,172. (Hereafter the two patents will be referred to by their last three digits: '841 and '172.)

Very early in the history of this controversy one of Potter's associates became associated with the company that is now the defendant in this action. The accused bearings manufactured by the defendant, it may be noted, are so close in similarity to those of appellee that they are interchangeable in application and this fact is made known to the industry.

This action was filed October 17, 1956. Following discovery and pretrial proceedings a pretrial order was entered and thereafter defendant moved for summary judgment. The motion for summary judgment was heard and ruled upon by another judge of the district. Judge Hall commented at the beginning of the trial that he had noticed that the motion had been denied, that he expected the case to be somewhat complicated, and indicated to counsel that he would not entertain a new motion at that time. The case immediately proceeded to trial upon its merits.

It was and is the appellant's position that neither of appellee's patents is valid but that even if they were valid the accused bearings do not infringe.

After a trial of twelve days in which experts testified for both sides the trial court, in an oral decision, held Claims 1 and 2 of Patent '841 valid and infringed, and also held Claims 1, 2, 3, 4 and 6 of Patent '172 valid and infringed.¹ 

Upon appeal appellant condenses it specifications of error into five questions for this court as follows:

1. As a matter of law, did the trial court apply the correct criteria and standard of invention * * *?

2. Did the trial court commit error in failing to find that the Potter Patents had been issued in violation of statutory requirements that no new matter be added * * *?

3. Did the court make conflicting and erroneous findings of fact?

4. Does defendant infringe * * *?

5. Did the trial court err in denying defendant's motion to strike certain testimony [and in failing to receive other testimony] * * *?

Appellant assigns numerous specifications of error with respect to findings of fact and conclusions of law, details of which need not be set forth in this opinion in view of the conclusion reached by this court.

The issues thus raised involve both validity and infringement. We treat first the question of validity as raised by specifications of error included in questions one and two noted above, and in order to better understand the problem it is necessary to examine in detail the file wrappers, Exhibits A, B and C, including the chronology of application 606,678 — that file wrapper which is Exhibit A and which ultimately matured into Patent '841. Schriber-Schroth Co. v. Cleveland Trust Co., 311 U.S. 211, 312 U.S. 654, 61 S. Ct. 235, 85 L. Ed. 132.

July 23, 1945 the Patent Office received and filed said application, duly supported by oath. In brief, it discloses that the primary object of the invention is the production of a spherical, self-aligning bearing which is simple and inexpensive in construction and in which a ball element is freely journaled and securely held without lost motion. The application teaches forming a bearing race over a bearing ball in closely and evenly engaged position, whereby the contact surfaces conform and produce a tight connection. The race is then mechanically expanded in such manner as to produce sufficient bearing clearance and freedom without any undesirable looseness. Illustration is furnished of the manner in which pressure is applied to the periphery of the bearing race at its median portion to elongate the latter toward the two ends of the race so as to produce clearance and eliminate the tight connection.

The bearing race is preferably made out of ductile material such as brass or other metal which can readily be formed by pressing or swaging and later elongated.

The bearing ball is first made of relatively hard metal. The bearing race is made from a cylindrical blank of ductile material. After forming the ductile blank around a ball and into tight engagement the assembly is subjected to sufficient contractive pressure by rolling contact of one or more pressure rolls "by which sufficient force is applied diametrally inwardly to expand and elongate the bearing race toward both of its ends,"²  to relatively release the bearing ball and race. The application teaches that a similar expanding or elongating effect can be produced upon the bearing race after the crown remaining following the forming operation has been machined off by applying the pressure rollers against the cylindrical outer surface of the race. "It is also contemplated that other means may be employed for elongating the bearing race."³  Following the applicant's description of the principles of operation of his invention, he made the following claims:

1. A self aligning bearing, comprising, a bearing ball having a spherical bearing surface and an axially disposed work piece receiving bore, and a bearing race having a raceway therein corresponding with the spherical surface of said bearing ball to freely journal the bearing ball in said raceway, said bearing race being of single piece construction.

2. A self aligning bearing, comprising, a steel bearing ball having a curved bearing surface, and a single piece bearing race of malleable material having an inner raceway corresponding in curvature with the curved surface of said ball and in which said ball is rotatively retained.

3. The method of forming a self aligning bearing having a bearing ball and a ductile bearing race, said ball and race being formed with corresponding curved surfaces there-between, comprising, assembling said ball in an annular blank having an inner cylindrical surface substantially corresponding in diameter with that of the bearing ball and having opposite end portions, compressing said end portions inwardly to conform and produce a binding engagement with the curved surface of said ball, and finally compressing the median portion inwardly between the ends of the bearing race to elongate the bearing race longitudinally towards its opposite ends whereby the bearing surfaces between the bearing ball and bearing race are released to permit free rotation between the ball and raceway.

4. The method of forming a self aligning bearing having a bearing ball of hard material and a bearing race of ductile material in which said ball is retained and freely journaled, comprising, assembling said ball in an annular bearing race blank having an inner cylindrical surface corresponding in diameter with the diameter of the ball and having oppositely disposed radially converging end surfaces, compressing said end portions inwardly with said inner cylindrical surface of the bearing blank formed closely against the surface of said ball while said radially converging end surfaces assume normal position to the axis of the bearing race, whereby the bearing race tightly grips the surface of the ball, and finally applying sufficient inward pressure against the perimeter of said bearing race to expand and elongate the bearing race and free the bearing surfaces between the ball and bearing race.

December 4, 1946 the Patent Office required division between Claims 1 and 2, which relate to the bearing itself, and 3 and 4, which relate to the method of manufacture.

May 15, 1947 the Patent Office received word from the applicant requesting a cancellation of Claims 3 and 4 without prejudice and requesting the addition of the following claim:

5. A self contained and self aligning bearing, comprising, a bearing ball having a smoothly finished perimeter and an axial work piece receiving passage therein, and an annular bearing race, said race comprising a ductile collar having an inner annular raceway formed inwardly at both longitudinal ends thereof and conforming closely and freely over the curved surface of the bearing ball, whereby the ball and bearing race are movably retained together.

June 22, 1948 Claims 1, 2 and 5 were rejected as being fully met by Chambers (referring to Chambers, et al., Patent No. 2,382,773) and by Scoville (referring to Scoville, Patent No. 1,266,061).

December 8, 1948 the Patent Office received communication from the applicant cancelling Claims 1, 2 and 5 and adding the following:

6. A self aligning bearing, comprising a bearing ball having a spherical bearing surface and an axially disposed work piece receiving bore, a malleable single piece annular bearing race having an inner cylindrical surface corresponding in diameter with the diameter of said ball, means for compressing the outer end portions of said race in close complimentary engagement over the spherical surface of said ball, and means applicable around the median portion between the ends of said face for expanding the race longitudinally to produce free rolling contact between the ball and race.

November 1, 1949 Claim 6 was rejected both on the ground that as written it embraced both an element and a means of manufacture, and also upon Chambers in view of Heim (referring to Heim, Patent No. 2,476,728). The communication from the Patent Office also recited: "The patent to Chambers discloses the full equivalent of applicant's structure. Further, no invention would be involved to expand the race longitudinally to produce a free rolling contact surface as taught by Heim."

March 7, 1950 the Patent Office received cancellation of Claim 6 and a request for the addition of the following claim:

7. A self-aligning bearing construction involving inner and outer bearing members, said inner bearing member comprising a bearing ball having a spherical bearing surface and an axially disposed bore for receiving a shaft, a non-ferrous malleable metal single piece outer bearing race member having a spherical socket corresponding in shape to the spherical inner bearing member and having parallel radial end walls, said outer race member being stressed such that the metal adjacent the inner peripheral surface area is compressed and the metal adjacent the outer peripheral surface is under a stress tension to form an unstretchable peripheral area which, when subjected to a rolling pressure, will cause the metal adjacent the inner peripheral surface to expand the ends of said outer bearing member in a direction away from the axis of the self-aligning bearing.

The above claim is reproduced as it appears after certain minor changes were made in the wording at the request of the Patent Office.

January 5, 1951 the Patent Office rejected Claim 7 for the reason that it wasn't sufficiently definite to describe a finished article and for the further reason that a method limitation, i. e., the stressed condition of the metal, is an improper identification of a finished article. The office also considered the claim, as then understood, to be unpatentable over either Claus (referring to Claus, Patent No. 1,664,189) or Chambers. The examiner contended that "each of the above references disclose a self-aligning bearing comprising a single piece outer member having a spherical inner bearing surface and an inner bearing ball member having an outer spherical surface corresponding in shape to the inner bearing surface of the outer member and being freely journaled therein."

July 6, 1951 the Patent Office received a communication from the applicant in response to the above wherein it was observed that claim seven and a new claim were not directed to a finished article but to a combination of elements which are instrumental in producing the finished article. It was urged that it was conceivable that such an article could be furnished to a fabricator who would have the facilities at hand for applying the rolling pressure to the specially stressed outer bearing member.

With this last communication Claims 8, 9 and 10 were added. Claim 8 reads:

8. A self aligning bearing construction involving inner and outer bearing members, said inner bearing member comprising a bearing ball having a spherical bearing surface, a malleable single piece outer bearing race member having a spherical socket corresponding in shape to the spherical inner bearing member, said outer race member being stressed such that the metal adjacent the inner peripheral surface area is compressed and the metal adjacent the outer peripheral surface is tensioned.

Claims 9 and 10 were later rejected and no purpose would be served in reproducing them here.

April 1, 1952 Claims 7 and 8 were allowed and became Claims 1 and 2 of the patent. This action was declared final.

July 22, 1952 formal Notice of Allowance was mailed by the Patent Office to the applicant.

If we study the application most carefully up to March 7, 1950 we understand an attempt to patent a spherical self-aligning bearing in which a ball element is freely journaled and securely held without lost motion in a single piece, continuous bearing race. The ball was to be first made of relatively hard metal. The bearing race was preferably made of ductile material such as brass or other metal which can be readily formed by pressing or swaging and later elongated. The application contemplated forming the bearing race over the bearing ball in closely and evenly engaged position by means of forcing suitable dies toward each other upon the assembly under tremendous pressure, whereby the contact surfaces conform and produce a tight connection. The race was then to be mechanically expanded and elongated to produce sufficient clearance.

The application, up to said date, teaches that this mechanical expansion in a longitudinal direction is accomplished by subjecting the assembly to contractive pressure by rolling contact of one or more pressure rolls by which means force is applied inwardly upon the median portion of the race to expand and elongate the bearing race towards its two ends. Following the forming operation the race is left with a pronounced crown. The specifications teach that the expanding or elongating effect can be produced after the resultant crown has been machined off. (In the finished article the crown is always machined off, leaving the outer surface of the race smooth and cylindrical.)

Claims 3 and 4 of the original application both speak of elongating the race toward its ends by applying force to the median portion of the race.

Claim 6, added December 8, 1948, speaks of expanding the race longitudinally.

Prior to March 7, 1950, at which time Claim 7 was added, all of the teaching, including the drawings, is to the effect that a soft collar is first pressed or formed around a hard ball and thereafter expanded longitudinally so as to free the ball. The expansion contemplated up to this point is always longitudinal. Up to this point the applicant realized no success in his effort to patent. This is understandable, for even a layman appreciates that if pressure is applied to the meridian portion of a ductile collar encompassing a hard ball, the ductile metal will spread, forcing the ends of the collar away from the meridian. This phenomenon would take place irrespective of whether or not any peculiar stresses were locked therein.

Webster's New Collegiate Dictionary defines ductile as "Capable of being drawn out or hammered thin, — said especially of metals, as gold." The applicant himself suggests "brass or other metal which can readily be formed by pressing or swaging and later elongated."

Clearly gold or lead would be inappropriate for purposes of a bearing race in view of the intended application of these particular bearings, but in view of all that the application teaches up until March 7, 1950, those metals could be utilized and would fulfill each step then contemplated; for up until that date a spreading of the metal near the meridian of the race in a direction parallel with the axis was all that was contemplated.

An understanding of the prior patents within the art relied upon by the Patent Office in rejecting all claims prior to March 7, 1950 will serve to highlight the lack of invention in the teaching of the application up to that point. Two of the patents thus relied upon are Chambers and Heim.

Chambers et al., No. 2,382,773, patented August 14, 1945, application January 2, 1943, relates to improvements in self-aligning bearings and a method of making the same. The ball element of Chambers' bearing is identical with the ball element in Potter's bearing. Chambers' race element is not cylindrical as is Potter's, but rather like a rod with a flattened disk at one end through which a hole is machined corresponding in diameter with the ball element. Upon each plane of this resulting ring a circular V-groove is cut completely encircling the hole, leaving a continuous lip encircling the hole both on the top and on the bottom of the ring. The specifications contemplate forcing suitable dies toward each other and into the circular V-grooves while the ball element is positioned within the hole. The two dies are semi-spherical on the inside and as they are forced together they force the continuous deformable lips over simultaneously in one operation whereby the bearing surface of the race is formed complimentary to the bearing surface of the ball element. The grooves are thereafter filled with a soft metal. Chambers, et al., does not contemplate the formation of a tight engagement and thereafter a freeing step.

Heim No. 2,476,728 relates to the mechanical treatment of bearings to case their action preparatory for use and teaches a simple and practical art for remedying excessive tightness of bearings. Heim states, "it may be noted that in many bearings, which term is used in a broad sense throughout this application, the construction or method of manufacture is such that when completed the inner member is so tightly gripped by the parts within which it relatively swings or rotates that it is ill-prepared to be put into practical use." He then discusses the problem of freeing such bearings while maintaining the uniformity of contact between the contact surfaces.

His invention consists of a frame which securely holds the ball by means of a pin through its bore. Another part of the frame supports an air hammer with an adjustable stroke. The head of the hammer is shaped to fit the race upon which it will strike. While the hammer is striking the race is swung throughout its movement. Because the stroke of the hammer is limited, no one spot of the race will receive more force than necessary or desired, since after the first stroke or two the metal of the race will be deformed to the point where the hammer will no longer reach that point. Since the ball is much harder than the race the deformation will always occur in the race.

Heim, of course, teaches much more than the mere striking of a race with an air hammer but what it teaches certainly embraces that solution.

There was, moreover, much testimony in the trial that freeing bearings by hammering upon the race is very old in the art. This hammering, of course, produces a mechanical expansion of the race and that is precisely what Potter teaches prior to March 7, 1950.

A large number of prior patents, some not considered by the Patent Office, relating to the bearing art were brought to the court's attention and discussed throughout the trial. Exhibit "D" is a looseleaf notebook consisting of over twenty such patents.⁴  We will consider the more pertinent ones.

Birchwood No. 1,050,422, patented January 14, 1913, describes a ball and socket joint formed in the following manner. A cylindrical recess in the end of a rod has a flared slot cut into and through it, leaving two halves, somewhat cut away, facing each other. The ball element, which consists of a ball supported on a pedestal, is inserted between the two halves which are then pressed around the ball. This operation causes the flared slot to assume a parallel position through which the pedestal is free to move. The only similarity between Birchwood and Potter is that the ball element in each acts as a die member in the forming operation and remains a part of the finished bearing. Tight engagement and subsequent liberation is not contemplated.

Porter No. 1,123,796, patented January 5, 1915, consists of a ball and socket joint made by placing the ball element into a cylindrical element bored in the end of a rod, the side walls of which are then turned or swaged over against the ball, locking it therein. Again, the only similarity is the fact that the ball element acts as a die in the forming operation.

Erickson No. 1,481,000, patented January 5, 1924, discloses a ball container utilized in ball bearings. The container, which serves to keep the balls in place between the bearing surfaces, is made of sheet metal and is stamped around the balls so as to lock them therein. Again, the balls act as dies but there the similarity ends.

Fiegel No. 1,693,748, patented December 4, 1928, discloses a ball and socket joint remarkably similar to the self-aligning bearings involved here. Fiegel begins with a tube which is flared out at one end. The inner surface is roughened and lined with babbitt. The babbitt is machined so that it conforms in diameter with the ball and the babbitt in the end away from the flare of the tube is machined into a semi-spherical seat. The ball element, which is a truncated sphere with a bore through it, as is Potter's, is placed in said seat and the entire assembly is forced down a narrowing die, flare end last. As the assembly is forced down the narrowing die the flared end is forced inwardly, which in turn forces the babbitt above the equator of the ball to be forced against and around the ball, locking it permanently therein. No liberation step is contemplated.

Hoern No. 1,798,738, patented March 31, 1931, relates to a ball and socket joint such as are commonly used in the heads of tappet adjusting screws. Each ball is utilized as a tool to form the socket in which it permanently remains. The invention contemplates forcing the ball into the end of a solid rod while the rod is held in a die which controls the flow of the metal of the rod into any desired shape and also confines it. More than one-half of the ball is forced into the rod and the metal of the rod, including that portion above the equator of the ball, is forced into binding engagement with the ball. An important feature of this invention is that the ball while frozen in this manner may be worked upon. The ball is then liberated from this tight grip of the socket by applying appropriate pressure to the sides of the head in which the ball is held.

Paulus No. 2,252,351, patented August 12, 1941, relates to the method of producing power transmitting units utilizing ball and socket joints. Paulus recognized that "it has heretofore been proposed to shape one of the elements about the companion element." His invention seeks to eliminate or avoid many of the difficulties encountered when this method is employed. One of these difficulties is the fact that the connection is too tight when first made. He teaches striking the ball element while the companion element is positioned upon a special anvil. He calls this an adhesion-breaking step.

Townsend No. 2,335,710, patented November 30, 1943, relates to a bullet core stripping machine. For the purpose of salvaging the steel cores without damaging them the outer jacket is stretched during the cutting operation by means of being rolled. This could be considered an equivalent of the stretching contemplated by Potter as indicated by his claims during the first five years of his application.

Taylor No. 2,382,349, patented August 14, 1945, but applied for August 23, 1943, teaches a method of making a race for a self-aligning bearing such as employed in the Messerschmidt bearing. An outer ring member substantially the same as is used by plaintiff is coined or press-formed about a hard steel blank, spherical and truncated but lacking the shaft receiving bore. The blank is then turned ninety degrees and forced out of the race, the interior surface of which now conforms to the blank in shape. Forcing it out in this way produces the key slot typical of the Messerschmidt bearing race. While the idea of using the ball element as a die in the forming operation is completely lacking in Taylor, the forming operation is essentially identical with the operation taught by Potter. After the initial forming operation similarity between the two patents is lacking.

We note the foregoing patents in the bearing art for the purpose of demonstrating the appropriateness of the action of the Patent Office in rejecting Claims 1, 2, 5 and 6 of the article application, as well as for the proposition that the patent as issued lacks patentability.

Birchwood, Porter, Erickson, Fiegel, Hoern and Chambers, to name a few, teach the utilization of the ball element as a die member in the forming operation.

Chambers, Fiegel and others taught that one element of such bearings should be shaped about its companion member. Paulus in 1941 recognized that this was old, as did Heim.

The need for liberation when this method of forming was utilized was certainly recognized by both Paulus and Heim, as both teach a solution to that very problem. Hoern also recognized the fact of tight connection, utilized it, then teaches that striking the outer member releases the tight engagement. Paulus and Heim treat this tightness as an undesirable result and teach a solution, while Hoern treats it as a desirable step to be utilized and thereafter teaches a solution.

Thus it is that all of the features of Potter's patent, '841, so far as disclosed prior to the March 7, 1950 amendment to the application, are old in the art. This in itself does not preclude invention but we are committed to the rule announced by the Supreme Court in Great Atlantic & Pacific Tea Co. v. Supermarket Equipment Corp., 340 U.S. 147, 152, 71 S. Ct. 127, 130, 95 L. Ed. 162:

"Courts should scrutinize combination patent claims with a care proportioned to the difficulty and improbability of finding invention in an assembly of old elements. The function of a patent is to add to the sum of useful knowledge. Patents cannot be sustained when, on the contrary, their effect is to subtract from former resources freely available to skilled artisans."

See also Moist Cold Refrigerator Co. v. Lou Johnson Co., 9 Cir., 249 F.2d 246; Bergman v. Aluminum Lock Shingle, etc., 9 Cir., 251 F.2d 801; Rohr Aircraft Corp. v. Rubber Teck, Inc., 9 Cir., 266 F.2d 613.

Any skilled artisan in the bearing art could take Taylor's patent, add to it the concept of utilizing the ball element rather than a blank in the forming operation — a concept suggested repeatedly in the prior art — and, if the end result were a bearing frozen in tight engagement, liberate it as taught by Heim, Hoern and others, and he would produce a bearing in all respects identical with that of plaintiff. But by so subtracting from the prior art he could not claim this aggregation as a monopoly. Lincoln Engineering Co. of Illinois v. Stewart-Warner Corp., 303 U.S. 545, 58 S. Ct. 662, 82 L. Ed. 1008.

It is only when we view the stresses and strains described in the March 7, 1950 amendment do we recognize something that has not been so described in the prior art patents, and we now examine that aspect of the case relating to question 2 of appellant's six proposed questions: Does the amendment adding Claim 7 to the application constitute new matter?

By adding Claim 7 to the application on March 7, 1950 the applicant informs the Patent Office that the bearing race member is stressed such that the metal adjacent to inner surface is compressed and the metal adjacent the outer surface is under a stress tension forming an an unstretchable peripheral area which, when subject to a rolling pressure, causes the ends of the race to be expanded away from the axis of the bearing member.

This claim, it should be pointed out, does at least two significant things. First, it limits the metal to be used in the race to those metals which are ductile, yet capable of locking therein the compressive and tension stresses which the claim, for the first time, speaks of; and, secondly, this claim calls for a bearing race which will expand in a direction away from the axis, radially rather than longitudinally. This latter concept, however, is not only new but at variance with all the teaching of the application and specifications prior thereto.

This claim was added almost five years after the application was filed. Also, the evidence is clear to the effect that the bearings manufactured by plaintiff have been manufactured in the same manner continually since prior to the application date, and this is so whether the process is explained in accordance with the first teaching or as latterly incorporating the concept of certain stresses and strains locked in the race.

35 U.S.C. § 132 provides that no amendment shall introduce new matter into the disclosure of the invention. While this section was enacted in 1952, after the amendment here involved, that much of the section which excludes new matter is no more than a codification of the existing law. Applicable law did require that application be made within one year of public use or sale in this country. (See § 31 of repealed Title 35. This provision now appears in § 102 of present Title 35.) As stated in Muncie Gear Works v. Outboard Marine & Manufacturing Co., 315 U.S. 759, 768, 62 S. Ct. 865, 869, 86 L. Ed. 1171:

"The claims in question are invalid if there was public use, or sale of the device which they are claimed to cover, more than two years before the first disclosure thereof to the Patent Office. Cf. [Chicago & N. W.] Railway Co. v. Sayles, 97 U.S. 554, 557, 559, 563-564 [24 L. Ed. 1053]; Schriber-Schroth Co. v. Cleveland Trust Co., supra [305 U.S. 47], at [page] 57 [59 S. Ct. 8, at page 12, 83 L. Ed. 34]. Section 4886 of the Revised Statutes would in terms provide for their invalidity had they been offered by application rather than by amendment; and whatever may be the efficacy of an amendment as a substitute for an application, it surely can effect no more than the application itself."

We are faced then with this trilemma: If the amendment adding Claim 7 to the first application actually presents new matter to the Patent Office and the patent as issued depends upon said new matter, the claims, as just indicated, are invalid, having been in public use for almost five years. If, as claimed by appellee, the alleged new matter is readable upon the specifications, drawings and claims as first filed, it must be held that the same would also be readable upon the prior patents in the art, including those which caused the Patent Office to reject the earlier claims, and if this be the case there is nothing patentable. It is clearly the law that a patent should not be granted for discovery of a result that would flow naturally from the teachings of the prior art. In re Kepler, 132 F.2d 130, 30 CCPA 726; Application of Heinrich, Cust. & Pat. App., 268 F.2d 753; Application of Welter, 255 F.2d 944, 45 CCPA 1034; Application of Libby, 255 F.2d 412, 45 CCPA 944; In re Bourdon, 240 F.2d 358, 44 CCPA 740; and In re Eisenhut, 245 F.2d 481, 44 CCPA 974. Lastly, if as suggested by the trial court at one point,⁵  the alleged new matter — the stresses and strains set up in the amendment — are not part of the patent, what is left is indistinguishable from the claims rejected by the Patent Office. If the applicant was entitled to a patent upon any of the claims rejected, his remedy was by way of appeal from the rejection. See §§ 141 through 146 of Title 35 U.S.C. Unless a rejection has been appealed to the Board of Appeals a district court is powerless to grant a patent, and only then if the Commissioner is made a party defendant and the action filed in the District of Columbia; but that is what the court below would be attempting had it found validity upon some claim other than the claim or claims awarded by the Patent Office. 35 U.S.C. § 145.

As we have just indicated, the stress and strain matter added by the March 7, 1950 amendment is either new to the application — hence unpatentable — or no more than a scientific explanation of what takes place in the manufacture of Potter's originally described article as well as other bearings patented in the art. A scientific explanation cannot be patented in any event, whether originally disclosed in an application or by way of amendment. De Forest Radio Co. v. General Electric Co., 283 U.S. 664, 51 S. Ct. 563, 75 L. Ed. 1339.

Potter Patent No. '172, the method patent, is dependent upon the application for the article patent. The only difference between the claims allowed in Patent '172 and the claims rejected in application 767,496, the first divisional application, is the fact that the claims finally allowed contain language concerning the same compressive and tension stresses heretofore discussed in relation to Patent '841. Again, this reference to stresses and strains was not brought to the attention of the Patent Office until 1950. All that we have said in connection with Patent '841 applies equally to Patent '172. It suffers the same infirmities.

We find, therefore, that both Potter Patents, 2,626,841 and 2,724,172, are invalid. While we feel this conclusion could be reached upon factual considerations we reach it, as indicated throughout, as a matter of law. Both patents being invalid, we do not reach the question of infringement, or consider other specifications of error.

Reversed and remanded with directions to enter judgment for appellant.