These consolidated appeals involve two patents, in the field of bonded magnets, owned by Minnesota Mining and Manufacturing Company (3M). Walter S. Blume and the Electrodyne Company (defendants) appeal from judgments holding that Blume infringed 3M's U. S. Patent No. 2,999,275 (275 patent) and thereby also breached an agreement not to compete (described by the district court as the "noncompete agreement"). 3M appeals from a judgment holding its U. S. Patent No. 3,235,675 (675 patent) invalid for obviousness.1 We affirm the judgments.

Magnets may be divided into two broad classes, sintered (or cast), on the one hand, and bonded, on the other. Sintered magnets are formed by firing magnetic materials (e.g., aluminum, nickel and iron) at high temperatures into a single coherent body. Sintered magnets have high magnetic energy but tend to be hard, brittle, and thus extremely difficult to work or machine. Bonded magnets were first developed in the 1930's by mixing small particles of magnetic material with a binder material such as plastic or rubber. The resulting magnet could be easily worked and machined. By 1954 bonded magnets were commercially available, but their magnetic energy was low.

While bonded magnets were being developed, a group of scientists with the Philips Company were studying the qualities of a newly-discovered class of hard magnetic materials composed of an iron oxide sintered with barium, strontium or lead. In the early 1950's members of the Philips group secured a patent (the 778 patent) and published an article (the Philips article) describing a process in which this sintered material (hereinafter "barium ferrite") would be ground to fine particles, many as small as a single crystal. These particles would be placed in a mobile condition in a non-magnetic binder and subjected to a magnetic field which would orient the particles in substantially the same direction. The particles would be further oriented when they were again sintered into a coherent, dense body with exceptionally high magnetic energy.

Blume was aware of both existing bonded magnet technology and the then-recent research on the qualities of barium ferrite. During the mid-1950's he engaged in research which led to the development of a high-energy bonded magnet. In 1958 Blume filed a patent application which became the 275 patent when issued in 1961. This patent covered a process for mixing plate-like particles of barium ferrite with a non-magnetic binder, and orienting the magnetic particles by means of mechanical forces exerted on the material by rolling or extruding processes. Orientation of the plate-like particles produced a magnet with twice the magnetic energy of earlier bonded magnets. Although it was well known in the art to orient particles in a non-magnetic matrix by exposing the material to a magnetic field, the mechanical orientation taught by the 275 patent process proved to be the first commercially practicable method of orienting the particles. The high energy bonded magnet created by the 275 process became a great commercial success.

In an attempt to secure a patent which covered the product produced by the 275 process, Blume in 1962 added claims 8-10 to an existing patent application, which became the 675 patent when issued in 1966.2 These claims provide as follows:

8. A permanent magnet material comprising a dispersion of particles of a permanent magnet material in a non-magnetic matrix, a substantial portion of said particles having two substantially parallel opposed faces the distance between which is no greater than the dimension across said faces.

9. A permanent magnet material comprising a dispersion of small bodies of a permanent magnet material in a non-magnetic binder, said particles being in the form of right cylinders and having a length to width ratio of no more than about one.10. A permanent magnet material comprising a dispersion of small discs of permanent magnet material in a non-magnetic molded binder, said discs having opposite faces lying in parallel planes and having a thickness no greater than the width of said faces.

These claims included within their description the plate-like particles of barium ferrite which, when oriented by the 275 process, produced a high-energy bonded magnet. However, the 675 patent product is not limited to the use of barium ferrite material, but rather encompasses any magnetic material which could be given the specified shape. Furthermore, although the advantages of orientation are discussed in the file wrapper, the 675 product claims are not limited to oriented particles of the specified shape. The file wrapper indicates that even unoriented particles of the specified shape will produce a better bonded magnet than would particles of random shape. A specific method for obtaining particles of the requisite shape does not form part of the claimed invention. Thus the scope of the 675 patent includes, but is broader than, the product of the 275 patent process.

3M subsequently acquired the rights in the two patents and entered into the noncompete agreement with Blume. The agreement provided that Blume would work for 3M for one year and would refrain from any participation in the magnet business for a five-year period beginning on the day Blume was last employed by 3M. Since Blume's last day of employment with 3M was on September 30, 1968, the noncompete agreement was due to expire on September 30, 1973.

In May 1971 Blume and 3M provisionally agreed to a modification of the noncompete agreement which would have allowed Blume to enter the sintered magnet business in exchange for an extension of the noncompete agreement in the bonded magnet business to May 1, 1976.3 Blume, however decided not to enter the sintered magnet business, and therefore the May 1971 modification never took effect.

Blume then sought permission to reenter the bonded magnet business. After much discussion and correspondence, which the district court's opinion relates in full detail,4 3M sent a letter-amendment dated June 26, 1972. This letter recited the noncompete agreement, released Blume to make sintered magnets and test equipment, and in the crucial clause, provided as follows:

This letter, which is a substitute for the letter of May 3, 1971, shall serve to release you to

B. Establish or acquire facilities for making and/or selling Matrix-bonded Permanent Magnets, with the express understanding that in making such Matrix-bonded Permanent Magnets you will not, prior to May 1, 1976 infringe any unexpired patent in your name which 3M obtained with its purchase of the Magnetic Division of Leyman Corporation, especially your U. S. Patent No. 2,999,275.

Blume agreed to the proposed amendment on July 5, 1972. He subsequently formed the Electrodyne Company and contracted for construction of a manufacturing facility. By June of 1975 Blume had begun to manufacture high-energy bonded magnets by a "trade secret process" which Blume contended did not infringe the 275 patent process. On February 6, 1976, 3M filed its complaint, which charged Blume with infringement of both the 275 and 675 patents, and breach of the amended noncompete agreement.

Blume denied infringement and breach of contract and alleged affirmative defenses based on contractual and estoppel grounds. Blume asserted that the June 1972 letter-amendment immediately released him to make high-energy bonded magnets provided only that he not employ the 275 patent process prior to May 1, 1976. After that date, Blume contended, the letter-amendment granted him a license to practice the 275 patent.5

On Blume's motion, the district court bifurcated the trial and first tried the issues raised by Blume's affirmative defenses. The district court concluded that 3M was not estopped from bringing suit and that the amended noncompete agreement neither immediately released Blume to make high-energy bonded magnets nor granted Blume a license to practice the 275 patent after May 1, 1976, and Blume appealed.

Following the second portion of the bifurcated trial, the district court held that Blume's trade secret process infringed the 275 patent and thereby breached the amended noncompete agreement. Blume appeals from this holding. The district court further held that the 675 patent was invalid as obvious and therefore was not infringed, a holding which 3M appeals. All three appeals were consolidated and are now before this Court.

The district court held that the noncompete agreement, as amended by letter of June 26, 1972, neither immediately released Blume from the 675 patent nor granted Blume a license to practice 3M patents after May 1, 1976, the date the noncompete agreement expired. Blume now protests that the bifurcation of the trial, which Blume requested, in retrospect served only to obscure the issues. He therefore seeks on appeal "to present at one time the whole story" by means of a sixty-page appellate brief, which devotes 49 pages to a highly argumentative "Statement of the Case."

The arguments which Blume addresses to this Court evidence a misapprehension of our role. Fact-finding is entrusted to the district court, and its findings are reversible only for clear error. Fed.R.Civ.P. 52. This Court is not to reconsider the whole evidence de novo, and this is all the more true where, as here, the district court's findings depend upon assessments of credibility. United States v. Aluminum Co. of America, 148 F.2d 416, 433 (2d Cir. 1945).

The district court determined, and we agree, that the disputed contract clauses were ambiguous. The court therefore allowed both parties to submit an abundance of parol evidence to aid in the construction of the agreement. This evidence, which included both testimony by the parties and written materials, is analyzed in careful detail in the district court's opinion.6 We have reviewed the evidence and the arguments of Blume and conclude that the district court's ultimate determination regarding the intent of the parties and the meaning of the noncompete agreement was not clearly erroneous.

Blume contends that from June 1975 to September 1976, and from June 1978 to September 1978, he practiced a trade secret process (TSP) which produced a high-energy bonded magnet without infringement of the 275 patent process. Unlike the 275 patent process, which mixes unoriented magnetic particles with a binder and then orients the particles by means of rolling or extrusion, the TSP allegedly achieves the necessary orientation of particles during the mixing step, through use of a specially modified Banbury mixer. The district court rejected this allegation, holding that the TSP was the equivalent of the 275 patent process, Graver Tank & Mfg. Co. v. Linde Air Products, 339 U.S. 605, 609, 70 S. Ct. 854, 856, 94 L. Ed. 1097 (1950), and therefore infringed that patent.7

Blume contends that this determination must be vacated and the issue remanded because the district court failed to review a deposition by Walter Blume given under protective order and filed by 3M with the district court in camera immediately before the second trial. Blume asserts that had the district court reviewed this deposition the nature of the modifications to the Banbury mixer would have been made clear and the district court would have reached a contrary conclusion on the issue of infringement.

Blume concedes that the deposition he now contends the district court should have reviewed was never offered into evidence. Since the deposition was never offered into evidence, the district court was not required to review it.8 Rommel-McFerran Co. v. Local Union No. 369, 361 F.2d 658, 662 (6th Cir. 1966); see Processteel v. Mosley Machinery, 421 F.2d 1074, 1076 (6th Cir. 1970). A finding of equivalence is a determination of fact made after consideration of the patent, the prior art, and the particular circumstances of the case. Graver Tank & Mfg. Co., supra 339 U.S. at 609, 70 S. Ct. at 856. On the evidence properly before it, the district court's determination of equivalence was not clearly erroneous.

There are three essential elements of patent validity: novelty, utility and nonobviousness. 35 U.S.C. §§ 101-103. 35 U.S.C. § 103 provides:

§ 103. Conditions for patentability; nonobvious subject matter.

A patent may not be obtained though the invention is not identically disclosed or described as set forth in section 102 of this title, if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains.

The purpose of this section is to distinguish true invention from a mere change of detail which may produce novelty but is not reflective of invention. See Frantz Mfg. Co. v. Phenix Mfg. Co., 457 F.2d 314, 327 (7th Cir. 1972). An examination of the subject matter sought to be patented requires a consideration of the scope and content of the prior art, the differences between the prior art and the claims at issue, and the level of ordinary skill. Sakraida v. Ag Pro, Inc., 425 U.S. 273, 280, 96 S. Ct. 1532, 1536, 47 L. Ed. 2d 784 (1976).

The district court held the 675 patent invalid for obviousness in light of the prior art embodied by the 778 patent and the Philips article.9 The 675 patent, the district court noted, spoke of a dispersion of "particles", "small bodies" or "small discs" of a permanent magnet material in a non-magnetic binder. These particles, furthermore, were of a specified shape, namely "Discs" or "right cylinders" having opposite faces lying in parallel planes and having a thickness no greater than the width of the faces. Since a dispersion of magnetic particles in a non-magnetic binder was well known in the art, the district court determined that the sole patentable invention, if any, disclosed by the 675 patent concerned the particular shape of the magnetic particles. The 778 patent, however, had previously described a process whereby barium ferrite was reduced to small particle size, many of the particles being in the form of single crystals. And the Philips article had described these particles as plate-like, with preferential crystal growth along the basal planes so that the large dimensions of the crystal were in the basal planes and the small dimensions were normal to them. These disclosures, the district court concluded, were virtually identical to the description of particle shape in the 675 patent. Since both bonded magnet technology in general and the shape of these particles were revealed by prior art, the district court held that the 675 patent would be obvious to one of ordinary skill in the prior pertinent art.

3M contends that the 778 patent and the Philips article did not disclose discrete plate-like particles; it insists that the particles were spherical until fused together into the final sintered product. However, three expert witnesses, whom the district court found credible, testified that the Philips article disclosed individual or discrete particles which, although they grew larger and became more oriented during the final sintering, were plate-like from the start. Under § 103, the scope and content of the prior art is a factual matter to be determined by the district court. Graham v. John Deere Co., 383 U.S. 1, 17, 86 S. Ct. 684, 693, 15 L. Ed. 2d 545 (1966). Where, as here, the district court chooses to credit the testimony of three expert witnesses as to what the prior art discloses, we cannot say the finding of the district court was clearly erroneous. Fed.R.Civ.P. 52. Accord, Norfin, Inc. v. IBM Corp., 625 F.2d 357, 364 (10th Cir. 1980).

3M next contends that the 778 patent and the Philips publication are deficient as prior art because they lack a disclosure enabling one skilled in the art to make the discrete plate-like magnetic particles, citing Seymour v. Osborne, 78 U.S. (11 Wall.) 516, 555, 20 L. Ed. 33 (1870). Assuming, without deciding,10 that the enabling disclosure doctrine is applicable in this context, we find that the prior art contained an adequate disclosure. The district court heard testimony from four expert witnesses to the effect that the 778 patent and the Philips article taught that plate-like particles of barium ferrite could be produced by sintering barium carbonate and iron oxide, then grinding and milling the resulting barium ferrite down to a specified size. Again, the district court's decision to credit these expert witnesses and to find as a matter of fact that the prior art contained a sufficient enabling disclosure is not clearly erroneous.

3M characterizes the testimony of these expert witnesses as "naked opinion evidence" and contends that it submitted "scientific fact evidence" to the contrary which must prevail as a matter of law, citing Lovas v. General Motors, 212 F.2d 805, 808 (6th Cir. 1954). But Lovas, which held that testimonial evidence "positively contradicted by the physical facts" lacks probative value, is inapplicable here. The "scientific fact evidence" introduced by 3M was a test performed in 3M facilities by 3M employees, without notice to Blume, which purported to show that when one skilled in the art follows the teaching of the 778 patent, most of the resulting particles of barium ferrite are not plate-like. In response to 3M's post-trial motion under Fed.R.Civ.P. 52(b) for amendment of findings and judgment, the district court specifically declined to give the test conclusive weight:

We reject 3M's contentions that its (test results) testimony requires that the Court amend its findings and conclusions. In connection with what 3M says about the failure of its test for trial to produce plate-like particles by following the teachings of the Philips prior art, it is noteworthy that in that test at least some of the particles produced were plate-like. Also, there was no indication of the amounts of ingredients used in 3M's test for trial .... Defendants argue the total amount is critical for the success of the process and to change the total amount of ingredients, even though the proper proportions are maintained, will require an entirely different process if the same end product is to be obtained.11

As noted by In Re Michalek, 162 F.2d 229, 232, 34 Cust & Pat.App. 1124 (1947), "it is not a difficult matter to carry out a process in such fashion that it will not be successful and, therefore, the failures of experimenters who have no interest in succeeding, should not be accorded great weight." An experiment by an interested party which shows that it may be possible to operate within the disclosure of a prior art patent without obtaining the disclosed product does not overcome the presumption that the process, if followed by one skilled in the art, will produce the product. Application of Weber, 405 F.2d 1403, 1407 (Cust & Pat.App.1969). Thus the testimony of the four expert witnesses was not, by means of 3M's test, "positively contradicted by the physical facts," and the district court's determination to credit the expert witnesses reflects neither a mistake of law nor a clearly erroneous view of the facts.

For the reasons stated above, the judgment of the district court is AFFIRMED.