Eastman Kodak Co. v. E. I. DuPont De Nemours & Co.

298 F.Supp. 718 (1969)

EASTMAN KODAK COMPANY
v.
E. I. DuPONT de NEMOURS & COMPANY, Inc.

Civ. A. No. 6038.

United States District Court E. D. Tennessee, N. D.

February 11, 1969.

Francis T. Carr, Douglas G. Brace and Edward J. Handler, New York City, Jackson C. Kramer, Knoxville, Tenn., George W. Petersen, Rochester, N. Y., for plaintiff.

Dexter N. Shaw, Gordon S. Rogers, Philadelphia, Pa., J. W. Baker, Knoxville, Tenn., C. Harold Herr, John E. Dull, Wilmington, Del., for defendant.

MEMORANDUM

ROBERT L. TAYLOR, Chief Judge.

This is a patent case involving novelty yarns. The action was filed under Title 35 U.S.C. Section 146. The trial was protracted, extending over a period of five weeks and two days. Testimony and argument consisted of 4,888 pages, and 360 exhibits were filed.

Plaintiff, Eastman Kodak Company, hereinafter sometimes referred to as Eastman, seeks an order of the Court directing the Commissioner of Patents to issue a patent to plaintiff. Eastman owns a patent application.

Defendant, E. I. DuPont de Nemours & Company, Inc., hereinafter sometimes referred to as DuPont, owns U. S. Patent No. 2,985,995 (hereinafter called '995) which was granted on May 30, 1961 based on application SN68130 filed on November 8, 1960 by William Walter Bunting and Thomas Larsen Nelson, who claim the benefit of the filing date of an earlier application SN752451 filed August 1, 1958, hereinafter sometimes called Bunting parent application.

Plaintiff is the assignee of U. S. Patent Application SN75396 filed on December 12, 1960 by Richard F. Dyer, as well as a predecessor application SN400,544 of Dyer filed on December 28, 1953, hereinafter sometimes called Dyer's parent application. Dyer filed his application in 1953 for a patent on the product and process of his 1952 work. The application disclosed two types of apparatus; one, the Figure 8 machine, which produces intermittently bulked yarn by an oscillating reciprocal guide which governs the rate of yarn fed to the jet; the other worked by variable speed rolls. The application was involved immediately in an interference with a Breen patent on the priority of invention of continuously bulked yarn and subsequently on the priority regarding intermittently bulked yarn. Both interferences were decided adversely to the Dyer application.

Plaintiff claims that Dyer made the invention of the counts (claims) at issue before Bunting and Nelson. This action was preceded by Patent Interference No. 93581 involving the Bunting and Nelson patent and Dyer's application, Serial No. 75,396 filed on December 12, 1960.

Count 1 (Claim 1), which is typical, of the Bunting and Nelson patent reads as follows:

"A compact interlaced multifilament textile yarn essentially consisting of closely adjacent continuous filaments free from ring-like or other filament loops, the filaments being randomly intermingled with adjacent filaments and groups of filaments along the length of the yarn to maintain the unity of the yarn by frictional constraint between the filaments, the yarn having a coherency factor of at least 2.5 when tested in the absence of adhesive and at zero yarn bundle twist by the hook-drop test."

It is to be noted that this count or claim defines the yarn as a compact, interlaced and loop-free yarn which has a coherency so that it can be used as a twist substitute.

The ultimate issue is whether Dyer's 1953 application covers any of the claims in Patent No. '995. That question depends on whether the 1953 application is a constructive reduction to practice of the invention described in the counts in interference. As bearing on what the application disclosed, the Court considers whether Dyer reduced his claimed invention to actual practice in 1952 by making the yarn which is referred to in the record as the Stevens "head-end" yarn.

A pre-trial order was filed, following the pre-trial conference, which contains eighteen so-called factual issues and the same number of legal issues, all of which appear to be subsidiary to the aforementioned issue.

Dyer's 1960 application disclosed the invention, but it was materially different from his 1953 application. Eastman's position before this Court has been that unless the 1953 application discloses the invention, the Court could not validly order a patent issue to Eastman. Before the Patent Office Dyer relied solely on the language of the 1953 application; and without any testimony being given the Board of Patent Interferences found on May 16, 1967 that the 1953 Dyer application failed to disclose any conception of the invention contained in the '995 patent. If this finding is correct, Eastman cannot prevail in this action.

With respect to the issues presented to the Board of Patent Interferences and decided adversely to plaintiff, the burden of proof is upon plaintiff. This burden is a heavy one and requires evidence that "carries thorough conviction ``that the Patent Office decision was in error.'" Morgan v. Daniels, 153 U.S. 120, 14 S.Ct. 772, 38 L.Ed. 667; Aloe Creme Laboratories, Inc. v. Texas Pharmacal Co., 335 F.2d 72 (5 Cir.); TVA v. Monsanto Chemical Co., 383 F. 2d 973 (5 Cir.).

As to issues neither presented nor decided by the Board of Patent Interferences, plaintiff is only required to prove its case by a preponderance of evidence. TVA v. Monsanto Chemical Co., supra.

Theories of Plaintiff:

In 1952, Dyer prepared at the request of Management a yarn which he called an intermittently lofted yarn which was sent to J. P. Stevens Company, Inc. in North Carolina, hereinafter called Stevens, with which Eastman had a joint development program for preparation of novelty fabrics of different types. The purpose of the cooperative program was to find profitable lines of yarn development for Eastman and to give Stevens a source of new and better yarns that it could use for introducing new fabrics to the industry. Novelty yarns are those which vary in some characteristic along the length of the yarn with the result that when woven into a fabric unusual and different visual effects are caused.

Intending to create a novelty yarn for the development program, Dyer constructed a relatively simple apparatus which consists of passing a yarn through an air jet which agitates the filaments of the yarn in such a way that the filaments may or may not be permanently rearranged. By means of a moving yarn guide Dyer increased or decreased the speed at which the yarn entered the jet. The process which he used produced a yarn which had alternating segments of lofted and non-lofted yarn. Lofting imparts a high density of loops that are formed by filaments bending back on themselves and protruding from the main yarn bundle with a resultant increase in yarn size, or bulk, Eastman contends that the non-lofted portions are the loop-free, compact, coherent yarn of the '995 patent and that the manufacture of that yarn in 1952 was an actual reduction to practice of the invention.

Plaintiff says that Dyer's 1953 application disclosed three yarns, the intermittently lofted yarn, the lofted segments, and the non-lofted segments. Dyer claims that the language is sufficiently clear that a person skilled in the art could learn how to make the yarn of the '995 patent by reading his application. He says that it defines the essential process variables which are the following: passing the yarn through an air jet, maintaining equal rates of yarn feed to the jet and yarn take-up from the jet, and maintenance of positive tension on the yarn. Eastman contends that by using the essential process variables and a variable speed roll apparatus described briefly in the 1953 application, one skilled in the art could make an unlofted yarn extending indefinitely without insertion of the lofted segments which were characteristic of the yarn Dyer sent to Stevens.

Finally, plaintiff says that DuPont has marked with the '995 patent number intermittently lofted yarns, so-called "slub" yarns, which are similar in nature to Dyer's intermittently lofted yarns. Eastman contends that DuPont should not be allowed to take the inconsistent position that Dyer's intermittently lofted yarn does not come within the '995 patent claims.

Theories of Defendant:

Defendant contends that Claim 1 of the Bunting patent requires that there be a compact, interlaced multifilament yarn essentially consisting of "closely adjacent continuous filaments free from ring-like or other filament loops," and that the quoted language excluded any yarn which has loops, whether those loops be continuous or whether they be in a yarn with alternate sections of unlooped yarn.

The Patent Office noted that loopy and intermittently lofted yarns were part of the prior patented art and cited the Breen Patent No. 2783609, which at column 9, line 68, reads as follows:

"* * * Intermittent impulsing of the multifilament being processed can be used to produce a novelty yarn having alternating smooth lengths and bulked regions produced according to the described process."

That same Breen patent in an earlier interference was awarded priority over Dyer's 1953 application with respect to intermittently lofted yarns. The Patent Office considered that the 1953 Dyer application claimed an intermittently lofted yarn which by definition contained loops and could not fall within the Bunting and Nelson claims which specified freedom from loops.

As further proof that the Patent Office was correct in its several holdings that the Dyer application was not a constructive reduction to practice of the '995 yarn, defendant says that Dyer does not sufficiently disclose a process by which one skilled in the art could make the product of the invention.

Defendant contends that finding a twist substitute was a long-existing problem in the textile industry and in 1956 Bunting and Nelson were assigned to the task of solving the problem and after about two years of work, in May 1958, made the invention at issue. Such invention provides a yarn that has the handling properties of a true twist yarn and can be used to replace twisted yarns in textile operations in the manufacture of standard types of fabrics. The Bunting and Nelson invention is used very extensively at all seven textile fibers plants of the defendant and the amount of yarn of the invention that has been sold by defendant runs into millions of pounds. Defendant contends that if plaintiff had any recognition of the invention or its utility in 1952 or 1953, it would not have waited until after defendant was on the market before making claim to the invention.

Defendant further contends that the Stevens yarn does not establish that the process used by Dyer in 1952 and described in his application produces a yarn falling within the claims of the '995 patent. Defendant says that the Stevens yarn is not proved to be the yarn which Dyer made in 1952, and that, even if it is proven, one cannot accurately determine the characteristics of a yarn that was woven in 1952 and remained in the fabric until removed for testing in 1968. Defendant contends that the unbulked portions of the Stevens yarn were excluded from the '995 claims because the segments were loopy and otherwise not compact and were not coherent to the minimum extent required by the '995 patent.

Defendant likewise contends that marking certain slub yarns with the '995 patent number was an unintentional mistake made by persons who did not know the limitations in the patent and that the mismarking has been corrected.

Pertinent Issues:

I. Does the Dyer application, Serial No. 400,544, filed December 28, 1953, have sufficient disclosure to establish conception and a constructive reduction to practice of the invention of any of the counts of Interference No. 93,581?

(a) Does the Stevens yarn and testimony as to its manufacture in 1952 establish an actual reduction to practice of any of the interference counts prior to May, 1958, which is the earliest date claimed by Bunting and Nelson for conception and reduction to practice?

(b) Does the 1953 application contain sufficient disclosure to teach a man skilled in the art the product of the '995 patent and a method of producing it?

(1) Was there disclosed a conception and appreciation of the yarn of the '995 patent?

(2) Was there sufficient disclosure of the jets and process variables?

(3) Was the disclosure of process and apparatus so precise that a man skilled in the art, if he followed the teachings of the application, would necessarily or inherently produce the yarn of the '995 patent?

II. Was the association of U. S. Patent No. 2,985,995 on DuPont's commercially marketed intermittently lofted yarns erroneously made by defendant without any intent to deceive and by persons without knowledge of the limitations of the '995 patent?

The Bunting and Nelson Discovery:

Twisting the yarn bundle has traditionally been the method used to make a yarn compact and to protect it during weaving. The twisting process, however, is an extra step in yarn manufacture, is costly, and is slow in relation to the modern spinning speeds of synthetic continuous filament yarns. Testimony showed that the need for a twist substitute was a well-recognized and long-standing need in the yarn producing industry.

In 1956, Bunting and his colleague Nelson began experiments to develop a twist substitute yarn which would eliminate the costly twisting procedure. Their work was primarily directed to the use of an air jet which would impart twist alternately in a clockwise then in counterclockwise direction. During the course of their work, in May, 1958, they ran a yarn through their jet which had bundles of both black and white filaments, the purpose being to observe the effectiveness of the air jet as a twister. They noted that the yarn treated by the air stream in the jet had relatively closely spaced gray portions, a phenomenon never observed prior thereto. Examination of the gray areas showed that black and white filaments had become interlaced. The yarn as a whole was loop free, was compact and was coherent due to the filament interlace or entanglement. The interlace induced coherency, i. e., the characteristic of a yarn holding its filaments in a compact bundle, was immediately recognized by Bunting and Nelson as a substitute for the coherency imparted by yarn twisting.

Late in 1958, Bunting filed his parent application which described in its claims the yarn of his invention as compact and sufficiently interlaced to provide the yarn at zero twist with the handling properties of a true twist yarn of at least one-half turn per inch. Bunting's continuation application filed in 1960 defined the minimum degree of coherency as at least 2.5 when measured by a hook-drop test. The hook-drop test essentially measures the density of points of intense filament entanglement. Claim 1 of Bunting's '995 patent (Count 1 in the interference) limits the claims to a continuous filament yarn which is (1) compact, (2) free from ringlike or other filament loops, (3) interlaced, and (4) coherent to at least the extent of a 2.5 coherency factor.

The jet devices described in the '995 patent separate and mingle filaments by either single or opposing air streams which impinge on the yarn at close to right angles to the threadline. By contrast the jets described by Dyer in his applications involve passing the thread down a venturi tube in which the air moves generally in a parallel relation to the threadline. Venturi jets work on a vacuum principle caused by air expanding from a narrow passageway called the throat.

Actual Reduction to Practice

Eastman contends that in 1952 Dyer made an intermittently bulked yarn which corresponds at least in its nonlofted, unbulked segments to the yarn described by the '995 patent. The process is said to have been the same disclosed in the 1953 application with the apparatus illustrated in Figure 8. Plaintiff relies on the yarn and proof of the process to establish a reduction to practice of the counts of the interference.

A good example of the meaning of "reduction to practice" is given in the case of Corona Cord Tire Co. v. Dovan Chemical Corp., 276 U.S. 358, 48 S.Ct. 380, 72 L.Ed. 610 (1928):

"* * * A process is reduced to practice when it is successfully performed. A machine is reduced to practice when it is assembled, adjusted and used. A manufacture is reduced to practice when it is completely manufactured. A composition of matter is reduced to practice when it is completely composed. * * *" p. 383, 48 S.Ct. p. 387.

Defendant argues that Eastman cannot rely on the Stevens yarn to show that Dyer made the yarn of the '995 patent. The objections are that plaintiff has not proved that the Stevens yarn was yarn made by Dyer and that no valid tests can be run on a yarn woven into a fabric sixteen years before its removal. Because of the disposition of the second objection, the first is not discussed.

Extensive inter partes tests were conducted on the Stevens yarn to determine if the unbulked portions had the characteristics described in the '995 patent. Mr. Faw and Mr. Hoskins, engineers employed by Eastman, testified that the yarn was carefully taken from the fabric and twist removed in preparation for the hook-drop test. The hook-drop test required zero twist yarn and care was taken to back twist substantially all the twist which was in the yarn. The hook-drop test itself as set out in the '995 patent requires that a yarn strand be clamped and allowed to hang vertically with a specified weight attached to the lower portion. A weighted hook is inserted through the middle of the yarn bundle and lowered until the hook stops, the distance descended measured, the hook reinserted and the process repeated. The inter partes test results as reported by Mr. Hoskins were that the unbulked segments had a coherency factor of 7.6, more than the minimum required by the '995 claims. The 7.6 result excluded some questionable readings and was calculated in the manner approved by defendant's experts. Both plaintiff and defendant assembled montage microphotographs of lengths of the Stevens yarn. Faw and Hoskins concluded from studying the montages: first, the montages showed that their methods for back-twisting for the hook-drop test were successful; and secondly, they confirmed that the unbulked portions had significant filament interlace or entanglement.

Mr. Faw stated that valid hook-drop tests could be run on a yarn which had been in a fabric as long as the Stevens yarn. He further stated that the stresses of weaving had probably tended to decrease the filament entanglement that was originally imparted to the yarn by the air jets.

Mr. Faw testified that the unbulked portions of the Stevens yarn were essentially free of ringlike or other filament loops and were compact. He further testified that even under no tension the yarn remained reasonably compact, i. e., the filaments hung together except at points of sharp bends or selvedge points caused from residing in the fabric. Mr. Faw concluded that the unbulked segments were the compact, loop-free, coherent yarn of the Bunting and Nelson patent.

Defendant's experts, Dr. Norton and Dr. Backer, disagreed with that conclusion. Dr. Backer, an expert in yarn study and consultant to DuPont, maintained that a valid hook-drop determination could not and was not made on the Stevens yarn. Mr. Faw recognized that the Stevens yarn had in some characteristics changed from the yarn which Dyer described in his 1952 report, plaintiff's exhibit 25. For instance, the length of the bulked segments had apparently increased about eight inches and the unbulked about one inch. Considering the increased length, crimp added to the yarn fibers in weaving, twisting, and long residence in a fabric, Dr. Backer concluded that the characteristics of the yarn when made could not be determined with any assurance.

Doctor Backer made exhaustive observations of the Stevens yarn during the inter partes tests. Of all the approximately 100 points of hook stoppage which he examined with a microscope he observed only five stopping points at which there was not a factor other than filament interlace to which he ascribed primary responsibility for the stoppage. Even at those five points he would not attribute stoppage to interlace, but said the character of the yarn suggested that other factors which are known but not easily observable were the most likely causes. He listed seventeen factors other than interlace which the science recognized or he observed and which were capable of influencing hook stoppage. He made observations of 110 such instances of interference, the most frequent intruding factors being the following: foreign white filaments entangled in the yarn bundle, 46 times; filament loops, 26 times; severe bends, 15 times; and residual twist, 9 times. Mr. Faw found some errors in the graphic presentation of these factors in defendant's exhibit 96, but Dr. Backer said only three of those errors out of approximately 100 were significant, and that the three changes were so few in relation to all his observations that his conclusions would not be affected. He concluded that virtually none of the hook-drop readings made in the inter partes test were reliable indicators of filament interlace and therefore could not establish the minimum coherency levels required in the '995 claims.

Doctors Norton and Backer stated that the montages of the Stevens yarn did not show jet interlaced filaments. Dr. Norton, who had some experience with air jets, indicated he only looked for the distinct braiding patterns exhibited by yarns made by the Bunting process even though he admitted a yarn could possibly satisfy the '995 coherency requirements without having the braiding pattern. Dr. Backer on the other hand did not limit his search to those braided structures. He said he saw absolutely no yarn entanglement or interlace which was attributable to air jet treatment; he further stated that at the inter partes test he saw no such interlace in any of his microscopic examinations of the unbulked portions of the Stevens yarn.

The montages did indicate considerable intermingling of filaments in the unbulked portion of the Stevens yarn. Dr. Backer testified that the filament entanglement observed in those areas was only such filament migration as is typical of any continuous filament yarn like the Stevens yarn which has been twisted, run over guides in processing, woven into a fabric, removed and back-twisted. DuPont ran a non-jet treated yarn through the same processing that the Stevens yarn's history included and made a montage, defendant's exhibit 122. Dr. Backer stated that the yarn in defendant's exhibit 122 showed the same kinds of filament disarrangement and the same type points of entanglement as the Stevens yarn even though the exhibit 122 yarn had never been jet treated.

Plaintiff failed to prove by a preponderance of the evidence either that the unbulked segments of the Stevens yarn when made was coherent to the extent required by the counts of the interference or that any coherency was imparted by the air jet. The Court must therefore conclude, with proper deference and respect for the competency and conscientiousness of Mr. Faw and Mr. Hoskins, that plaintiff cannot rely on the manufacture of that yarn to establish a reduction to practice of the invention here in issue.

Dyer's Disclosure to a Man Skilled in the Art

Plaintiff relies on the teachings of the Dyer 1953 application to establish a conception and constructive reduction to practice of the invention of the '995 patent. To constitute a constructive reduction to practice an application's disclosure must be sufficient to teach a man skilled in the art the invention and how to practice it. The Telephone Cases, 126 U.S. 1, 535, 8 S.Ct. 778, 31 L.Ed. 863 (1887); E. I. DuPont de Nemours & Co. v. Ladd, 117 U.S.App.D.C. 246, 328 F.2d 547 (1964); Spero v. Ringold, 377 F.2d 652, 656-660 (Cust. & Pat.App. 1967); Travis v. Baker, 137 F.2d 109, 30 C.C.P.A. 1271 (1943). The specification should be explicit enough that the invention can be practiced without the exercise of inventive faculties or extensive experimentation. Application of Tansel, 253 F.2d 241, 45 C.C.P.A. 834 (1958); Travis v. Baker, supra; Thompson v. Dicke, 110 F.2d 98, 27 C. C.P.A. 931 (1940); Brand v. Thomas, 96 F.2d 301, 25 C.C.P.A. 1053 (1938).

Does the 1953 Application Disclose the '995 Yarn?

We turn first to the question of whether the 1953 application describes to one skilled in the art the invention, i. e., the yarn of the interference counts.

In the Patent Office, the Primary Examiner for Interference 93581 held that the 1953 application disclosure did not include the yarn disclosed in the counts of the interference. In affirming the decision, the Board of Appeals in their opinion of June 20, 1963, at page 3, stated:

"We have carefully considered each of the references to the specification which appellant urges suggests a yarn having the absence of loops but with filamentary entanglements and support the view of the Examiner that the specification lacks any basis or suggestion for such a product. Each reference seems to be concerned either with the tangling of the filaments to form loops or the forwarding of the yarn without entanglements. We find no expressed recognition in the specification that a yarn could be fabricated from continuous filaments and utilized for further textile manipulation similar to a twisted yarn, but without twisting, by entangling of the filaments free of looping." (D. Ex. 30)

The Board of Interference Examiners confirmed the prior dispositions of the matter. The Board summarized its holdings at page 5 of its Decision of May 16, 1967, as follows:

"We have independently read, examined and studied the complete, original disclosure found within the four corners of application Ser. No. 400,544, including the written description, drawings and original claims, without finding any disclosure, explicit or inherent, or even any suggestion of the structure of a yarn that will support the emphasized language quoted above from the respective counts [meaning ``* * * closely adjacent continuous filaments free from ringlike or other filament loops, the filaments being randomly intermingled with adjacent filaments and groups of filaments along the length of the yarn * * *']"

In the same decision the Board held (pp. 11-12):

"We are convinced that Dyer disclosed in his parent application no conception of the invention here in issue. In our view he summed up the total of his conception, and appreciation at the date of filing his parent case, at page 2 of his application where he stated:

______ I have found by maintaining the yarn passing through the jet under a certain degree of tension, that the separation of filaments and resultant fluffing up of the yarn can be avoided.

In accordance with another feature ______ by maintaining the yarn passing through the jet under considerably less tension, coils, ringlike loops and whorls are formed along the yarn surface.

"Nothing in either statement provides any basis for the count language discussed in the foregoing."

The burden of proof was in the interference proceeding in the Patent Office to establish its case by a preponderance of the evidence. As previously indicated, plaintiff's burden on all the issues considered and passed on in the Patent Office is much greater in this action. The evidence must be such as to carry "thorough conviction." Morgan v. Daniels, 153 U.S. 120, 14 S.Ct. 772, 38 L.Ed. 657 (1894); Heston et al. v. Kuhlke, et al., 179 F.2d 222 (C.A.6, 1950); Toledo Scale Corp. v. Westinghouse Electric Corp., 351 F.2d 173, 176 (C.A.6, 1965).

A strong presumption that the Patent Office was correct still obtains where the evidence before the Court is not the same as that before the Patent Office. O'Donnell v. United Shoe Machinery Corp., 2 F.Supp. 178 (D.C.1933) and Powell v. McNamara, 5 F.Supp. 628 (D.C.1933).

The testimony on behalf of Eastman was that the 1953 Dyer application disclosed three separate yarns, including the unbulked segments which plaintiff urges is the yarn of the '995 patent. Disregarding for discussion the question of whether an unbulked segment in an intermittently bulked yarn can technically be termed a "yarn," Eastman did show that the described unbulked segments were loop-free, continuous filament yarn. A close issue on the testimony was whether the other characteristic of the Bunting yarn was described, i. e., whether the yarn was entangled or interlaced to a significant degree.

To establish that a new yarn, an entangled yarn, was disclosed in Dyer's 1953 application (plaintiff's exhibit 1) plaintiff relies on the following quoted portions of that document, at page 2, line 11:

"* * * I have found by maintaining the yarn passing through the jet under a certain degree of tension, that the separation of filaments and resultant fluffing up of the yarn can be avoided; * * *"

at page 4, line 8:

"* * * the very high velocity of the air caused an appreciable separation and subsequent tangling or loopiness in the yarn * * *,"

at page 4, line 14:

"* * * if a small amount of tension is maintained in the yarn strand subsequent to its exit from the air jet * * * the loopiness can be eliminated * * * I may angle the treated yarn away from the exhaust jet to impart improved tangling to [sic] loopiness;"

at page 11, line 24:

"However, if the yarn is put under tension * * * such tension prevents the action of the air in jet 51 from affecting the surface of the yarn to impart loops and whorls therein;"

and the entirety of claim 13.

Mr. Hoskins and Mr. Faw testified that taking the application as a whole, a yarn such as the '995 yarn is disclosed.

For DuPont, Drs. Norton and Backer testified that the application did not disclose to one skilled in the art a new yarn which was compact and significantly entangled. Dr. Norton interprets the quoted portions of the application to teach only the prevention of loop formation. That interpretation is consistent with Dyer's own testimony at page 109 of the Trial Transcript: We "knew how to make bulk yarn, and this apparatus operated to interrupt the normal operation so that we made a normal yarn or nonbulked yarn at certain intervals." Dr. Norton read claim 13 as being primarily directed to motivating yarn and attempting to prevent any change in yarn structure. He said that claim 13 should not be read as disclosing an entangled yarn as does claim 1 of Dyer's 1960 application. The 1960 claim specifically says the purpose of filament separation is to get interlocking whereas the intent of claim 13 to Dr. Norton is that any filaments separated from the yarn bundle should be rejoined without any filament rearrangement.

The statements of Mr. Dyer and Mr. Faw tend to indicate that in 1953 they did not intend that the 1953 application disclose a new yarn in the unbulked segments. Mr. Faw, who was Dyer's supervisor in 1952 and 1953, read Dyer's reports and his application; yet he admits that in 1952 he was not aware of the entanglement in the unbulked portions (see Transcript page 1276). Although both Mr. Faw and Mr. Dyer testified that they early became aware of the entanglement, only Dyer specifically testified that this was before the filing of the application; of course even if they were aware of the entanglement before the filing, that would not aid a person skilled in the art who had to rely solely on the disclosure of the application. Proof of an alleged inventor's conception and reduction to practice requires full corroboration by other than the inventor's own self-serving testimony or records. Tidewater Patent Development Co. v. Gillette Company, 273 F.2d 936, 940 (C.A.4, 1959); American Mach. & F. Co. v. Liggett & Myers Tobacco Co., 172 F.Supp. 12, 16, D.C., aff'd per curiam 272 F.2d 451 (C.A.3, 1959). At page 12, line 18, of the application, Mr. Dyer refers to the unbulked sections as "normal." His explanation that the term meant only normal in appearance is consistent with the facts, but the term clearly obscures any claimed disclosure of a new, entangled yarn. Eastman did not market an entangled yarn, unbulked throughout its length, until after DuPont placed its '995 yarn on the market.

Plaintiff failed to show by a preponderance of the evidence, including testimony which was not before the Patent Office, that Dyer's 1953 application disclosed the entangled coherent yarn of the invention. A fortiori Eastman failed to show by a preponderance of the evidence that Dyer disclosed a twist-substitute yarn.

Was There Sufficient Disclosure of Jets and Process Variables?

Since the first Dyer application did not describe a new yarn product in the unbulked segments, the issue of disclosure of process variables and apparatus bears chiefly on the issue of inherency.

Mr. Hoskins testified on behalf of plaintiff that reading the 1953 application in light of the state of technology at the filing date disclosed to a man skilled in the art both apparatus and process variables sufficient to produce the '995 interlaced yarn. As proof of his conclusions he constructed two machines which he said were described in the application.

The first demonstration model was the so-called Figure 8 machine which comprised a yarn feed from a supply package passing over a "high speed roll;" thence to an air jet; thence from the jet to a "low speed roll;" and finally to a take-up package which is shown as a bobbin on a twister machine. The rolls are mounted on a common shaft so that the speed differential is achieved by the high speed roll having a larger diameter than the low speed roll. In such situation where there was no interruption of travel of the yarn strand, the larger feed roll caused more yarn to be fed into the jet than was being taken out at the other end. This is called overfeed and resulted in the filaments comprising the yarn strand being mixed and looped, that is, bulked by the air jet. A reciprocating guide 66 operated by a cam was provided, which was designed to pull the yarn strand out of its normal path of travel. The effect of this was to overcome excess speed of the feed roll, causing the effective rate of feed of the yarn to the jet to be the same as that at which the yarn was being removed. Thus, yarn overfeed was eliminated, thereby permitting the air jet to whip the filaments about and entangle them, but not permitting any loops or whorls to be imparted to the yarn. By properly arranging the cam, the extent to which the yarn was pulled from its normal travel and the length of the respective sections of yarn could be controlled within practical space limitations.

After the description of the operation of the Figure 8 apparatus, Dyer described in his 1953 application, at page 13, the following configuration:

"Obviously a like result could be obtained by driving the high speed rolls 52 and 53 from a separate and variable speed source not shown while the slow speed rolls 54 and 55 are driven at a constant speed."

Mr. Hoskins stated that from this teaching in light of the 1953 technology he was able to construct the "variable speed roll machine" which he demonstrated to the Court. Eastman relied on this machine configuration to show that Dyer disclosed a method by which indefinitely long lengths of unbulked, entangled yarn could be produced without a bulked region.

Although the testimony conflicts, we will assume without deciding that the yarn produced by the demonstration machines was the compact, coherent yarn of the invention. Defendant fundamentally objects to the demonstration machines on grounds that the jets and process variables used by Mr. Hoskins are not taught in the application.

Mr. Dyer and Mr. Faw testified that the 1953 application described the three process variables essential to making the '995 patent yarn. Those variables were said to be: (1) treatment in a fluid jet to separate and intermingle the filaments; (2) zero net overfeed of yarn to the jet, which is the same as saying equal rates of yarn speed into and away from the jet; and (3) controlled tension on the yarn. They stated that the amount of tension on the yarn is described, particularly in claim 13, as that minimum tension necessary to prevent the formation of loops in the normal lofting process. Mr. Faw pointed out that the '995 patent teaches interlacing over a broad range of various yarn tension, yarn speeds, and air pressures; but both he and Mr. Dyer admit that to interlace one must choose a proper jet, a proper yarn, proper yarn speed, proper yarn finish, and proper tension. They both testified that after the essential variables are disclosed, other variables may be adjusted without use of inventive skills, but by the application of skills and understanding common to those skilled in the art of making lofted yarn in 1952.

Doctor Norton testified on behalf of DuPont that the 1953 application did not take into account enough variables and did not teach how to select proper jet and process variables to produce an interlaced yarn. The Bunting patent teaches that variables such as yarn finish, filament denier, filament count, and yarn composition will affect and may even prevent yarn interlace. Dr. Norton stated that the teaching of minimum tension as set out in Claim 13 and elsewhere in the application (e. g. page 2, paragraph 1) was directed to preventing separation of filaments and loop formation. He found no indication that minimum tension was intended to allow entangled filaments.

Concerning sufficiency of disclosure of jet designs, Mr. Dyer and Mr. Faw testified the proportional dimension parameters, or limits, were set out in the 1953 application; although specific size measurements were not given, they stated that those measurements were determined by beginning with the diameter of the yarn and applying the proportional dimensions. Defendant contrasted this disclosure with the detailed sizes given in Dyer's 1960 application. The jet used on the variable speed roll machine had its nozzle backed out from the orifice opening, thus allowing increased air flow. That technique is taught in Dyer's 1960 application; but Dr. Norton testified that it was not taught in the earlier application and that the air flow thus generated was more than the maximum that the 1953 application provided for.

The Figure 8 demonstration machine varied from the schematic diagram in the 1953 application in that Figure 8 showed a loop or "U" shaped yarn path when the reciprocal guide was pulled back, whereas the demonstration model had a yarn guide configuration which caused a straight line yarn pull out. Dr. Norton observed that the "U" shaped path would cause variations in tension on the yarn with a constant speed cam. Varying tension is consistent with the variation in degree of bulkiness which Dr. Backer observed in the bulky segments of the Stevens yarn.

Doctor Norton described ex parte experiments that he carried out for DuPont in which he used jets like those described by Dyer. He found that by increasing the air pressure and air flow to extents not taught in the 1953 application, he was able to interlace yarn.

The Court finds that by using the apparatus and process variables described by Dyer, it is possible to make the yarn of the '995 patent although successful production requires going beyond the teaching in levels of air pressure and air flow. Plaintiff did not show by a preponderance of the evidence and the application teaches a man skilled in the art how to make the yarn of the interference counts. In the first place, by admission of the plaintiff's witnesses, one must choose a proper set of process variables to get interlace. Before a choice of variables could be made, one would have to know what product he is trying to make; since the 1953 application did not disclose a compact loop-free yarn with significant filament entanglement, there would be no basis for a selection of the proper combination of process variables. Secondly, the teaching as to process variables and jet design is not sufficiently definite to disclose how to make an interlaced yarn without extensive experimentation.

Inherency

The foregoing discussion of the inadequacy of disclosure of jets and process variables established that the disclosure is not so precise that one skilled in the art would inherently or necessarily produce the yarn of the invention if he followed the application's teaching. Inherency would allow Dyer to get credit for inventing the yarn of the '995 patent without disclosing the new yarn itself.

To claim an invention as inherent in a process described, it is not enough to show that by manipulating the disclosed process variables it may be possible to make the product of an invention. The product must necessarily or inevitably be created. Kropa v. Robie, 187 F.2d 150, 38 CCPA 858 (1951); In re Draeger, 150 F.2d 572, 32 CCPA 1217 (1945); Brand v. Thomas, 96 F.2d 301, 25 CCPA 1053 (1938).

Since a proper selection of numerous process variables is required to produce the interlaced yarn, and failure to properly account for one might prevent yarn entanglement, plaintiff may not rely on the doctrine of inherency on the basis of the 1953 disclosure.

Lack of Conception

Failure to disclose the invention in the application prevents issuance of a patent upon that filing and is significant with relation to the questions of conception and appreciation of utility. When an alleged inventor files a first application which lacks a disclosure of the invention and subsequently files a later application which does disclose the invention, it may be concluded that the alleged inventor had not conceived or reduced the invention to practice at the time the first application was filed. In Clark v. Camras, 204 F.2d 273, at 280, 40 CCPA 963 (1953), the Court stated:

"* * * The involved invention solved an apparently serious problem confronting the art, yet we are unable to find in either of these Camras patents any hint of the problem here solved, or how to solve it. If the combined erase-record heads disclosed in these two Camras patents (upon which heads he seeks to predicate priority) had embodied the invention of the issue counts, it is inconceivable to us that Camras would not have disclosed it in the applications for these patents had he in fact conceived the involved invention and reduced it to practice at that time. Vreeland v. Miller, 112 F.2d 829, 27 C.C.P.A., Patents, 1284. * * * Indeed, after careful study of this extensive record, we can only conclude that Camras' whole case is an attempt to establish nunc pro tunc that exhibits 4, 12 and 17 [U.S.Patent No. 2,418,542] embody the invention of the issue counts."

The preponderance of the evidence fails to show that Dyer had any conception of a yarn which was a twist substitute, an essential element of the Bunting invention. Plaintiff's counsel in the course of the trial conceded that Dyer did not appreciate that the unbulked portions of his yarn could be used as a twist substitute until he saw that DuPont had so used it. The evidence with respect to the 1952 yarn confirms the soundness of the concession.

Dyer made a report dated October 10, 1952, which is filed as plaintiff's Exhibit 25. He stated in that report that at the September 16 meeting of the T.E.C. and J. P. Stevens' development group some interest was expressed in the possibility of making an intermittently bulked yarn. "This yarn would be normal yarn with bulked sections at appropriate intervals, and would produce a novel patterned effect in a fabric." He stated further that experiments had been carried out to determine methods of making intermittently bulked yarn and that a small sample had been shipped to J. P. Stevens. A sketch of the apparatus that was used to produce the samples of the intermittently bulked yarn was attached to the report and designated as Figure I. He reported that the yarn was made up of 24 inch bulked segments alternating with 24 inch normal segments. A photograph attached to Exhibit 25 as Figure II shows the loops and bulky nature of the yarn in its alternate segments. The report states that the yarn had 4.4 turns per inch and was therefore a twisted yarn.

As pointed out by defendant in its brief, the essence of the invention is that the yarn, by reason of the entanglement of filaments, has the handling properties of true twist yarn and is therefore a twist substitute. Since the 1952 yarn was twisted, it follows that Dyer did not have in mind a twist substitute when he made the 1952 yarn.

Dyer stated that anyone seeing the terms "normal yarn" and "normal sections" in plaintiff's exhibit 25 would not know there was any entanglement present. Faw agreed. There was no reference to the need for a twist substitute made in Exhibit 25.

Dyer knew there was a need in the industry for a substitute for the slow and expensive twisting operation and he knew of prior attempts in the industry to find a twist substitute. If he conceived a compact loop-free interlaced yarn of the invention in making his 1952 yarn, it strains credulity to believe that he would have waited until after he had analyzed the yarn marketed by DuPont under '995 before bringing the matter to the attention of Management.

Eastman strongly contends that the twist substitute function of the yarn was not an essential element of the Bunting invention and that Dyer is entitled to an award of priority of invention if plaintiff shows that Dyer conceived and reduced to practice a yarn which was compact, loop-free, and coherent to a substantial degree by reason of filament entanglement. Plaintiff relies on the following cases to establish that Dyer would only have to appreciate the utility of such a yarn as ordinary textile yarn rather than the utility as a twist substitute which Bunting found. Swan v. Thompson, 80 F.2d 374, 23 CCPA 806, (1936); Weis v. Woodman, 65 F.2d 274, 20 CCPA 1211 (1933); Foss v. Oglesby, 127 F.2d 312, 29 CCPA 1005 (1942). The principle established in those decisions is that an inventor is entitled to all the uses of his invention, and a subsequent inventor may not deprive him of the benefit of his invention by finding a new function or utility.

Plaintiff's reliance on these decisions is misplaced because Dyer did not conceive and reduce to actual or constructive practice a yarn with the same physical characteristics of the '995 yarn regardless of whether he appreciated the special utility. In each of those cases the product of the first claimant either expressly disclosed substantially the same product of the subsequent claimant or the process of the first claimant inherently produced the product. As previously indicated, Dyer did not in the 1952 yarn make a compact yarn, substantially coherent by reason of jet induced filament entanglement and Dyer did not disclose such a yarn expressly or inherently in his 1953 application.

Even if Dyer had made a compact, substantially entangled yarn, the evidence indicates that it would have been an accidental, unwitting or unappreciated practice. Plaintiff failed to show by a preponderance of the proof that Dyer appreciated that he made a yarn which was new by reason of being compact and substantially coherent because of entanglement. He failed to disclose such a yarn in either his 1953 application or his report on the Stevens yarn, plaintiff's Exhibit 25. His testimony indicates that he thought of the unbulked segments of his yarn as the "normal" supply yarn in which he had inserted bulked areas at intervals (see page 109 of the trial record). He cannot rely solely on his uncorroborated statement that some time before 1953 he realized that the unbulked areas were entangled, particularly since the documentary evidence from that time period is inconsistent with what he now remembers. Therefore, even if Dyer had made the compact, entangled yarn of '995, it would have been an accidental, unwitting and unappreciated practice on which he could not base a claim of priority. Eibel Process Co. v. Minnesota & O. Paper Co., 261 U.S. 45, 43 S.Ct. 322, 67 L.Ed. 523 (1923); National Dairy Products Corp. v. Borden Co., 394 F.2d 887, 890 (7 Cir.); International Nickel Company v. Ford Motor Co., 166 F.Supp. 551 (D.C.1958); Jennings v. Hill, 184 F.2d 187, 38 CCPA 701 (1950).

Defendant's Patent Number Mismarking

Plaintiff claims that defendant represented to the public that continuous multifilament yarn, having loopy portions deliberately inserted therein (slub yarn), was made under patent '995 and is thereby estopped from contending that the Dyer yarn, having interlaced portions free from bulkiness, is not covered by patent '995.

The proof showed that DuPont had included references to the '995 patent on labels for Dacron (polyester yarn) and acetate slub yarns and had associated the patent with the terms Rotoset and Cycloset in technical bulletins. Those two designations were applied to the slub yarns on labels and in technical bulletins.

The mismarking was first discovered in June, 1967 during development of acetate slub yarns by a Mr. Fetterman. He had marked the yarn with labels which contained the '995 patent and Cycloset designations because his slub yarn was highly interlaced, much like the yarns which were made in the plant and which were properly under the '995 patent. Mr. Fetterman had never read the Bunting patent. An engineer at the same plant, Mr. McCutchan, had read the patent; and when he noticed the number's appearance on slub yarn labels, he told Mr. Fetterman about the possible marking error and raised the question with his superiors. A letter from the Patent Division of the Textile Fibers Department directed that the labels drop all reference to the '995 patent. Mr. Taylor of the Patent Division investigated and found the same mistake had been made regarding DACRON slub yarns and ordered those labels be changed. He indicated that programs for patent labeling review had only sporadically been used; and persons in the Products Division, rather than the Patent Division or Legal Department, had decided when to mark a product with a patent number.

Mr. Thompson and Mr. Lee each had some responsibility for the association in the technical service bulletins of the '995 patent with DACRON slub yarns. Someone in the Products Division told Mr. Lee, publisher of the bulletins, to reference the Rotoset designation with the '995 number. Mr. Thompson of the Patent Division of Textile Fibers Department reviewed technical service bulletins associating DACRON slub yarns with the term Rotoset, but he did not know the nature of the DACRON slub yarns.

Doctor Black, Product Manager first for DACRON then nylon, and now Director of the Patents and Research Division, testified that when the bulletins first began to use Cycloset and Rotoset designations and footnoted them with the '995 number, all the yarns produced under those designations were yarns properly under the '995 patent. All bulletins making the reference of '995 to Cycloset and Rotoset have been withdrawn. He indicated that there were very few persons in the DuPont organization who knew the limitations of the '995 patent; and he said that in the future the Patent Division of the Textile Fibers Department would be responsible for patent marking.

Doctor Black testified that production of the mislabeled slub yarns was insignificant in amount. He said that in total pounds produced slub yarns of all types are only six thousandths, or six tenths of one per cent, of total pounds of interlaced yarns other than slub and producer bulk yarns.

The Court is convinced that the error in mismarking was mistakenly and innocently made with no intent to deceive and is not a basis for estoppel in this case.

In summary, the Court holds:

(a) that Dyer's application, Serial No. 400,544 filed December 28, 1953, did not contain sufficient disclosure to establish conception and a constructive reduction to practice of the invention of any of the claims of Interference No. 93,581;

(b) neither the Stevens yarn nor testimony as to its manufacture in 1952 established an actual reduction to practice of the product to which any of the interference counts refer prior to May, 1958—that date being the earliest date claimed by Bunting and Nelson for conception and reduction to practice;

(c) Dyer's 1953 application did not disclose to a man ordinarily skilled in the art a yarn which was compact, loopfree, and substantially coherent by reason of jet-induced filament entanglement;

(d) the 1953 application did not disclose to one ordinarily skilled in the art a process for making the yarn of the interference counts;

(e) in the 1953 application Dyer did not disclose to one skilled in the art process and apparatus which would necessarily or inherently produce the yarn of the interference counts;

(f) Dyer did not establish a conception of the invention covered by the interference counts prior to the conception and reduction to practice of Bunting and Nelson;

(g) the association of U. S. Patent No. 2,985,995 on DuPont's commercially marketed intermittently lofted yarns was erroneously made by defendant without any intent to deceive anyone and without knowledge of the limitation of the '995 patent;

(h) plaintiff is not entitled to an order of the court decreeing that Dyer was the prior inventor of the yarn in issue and directing the Commissioner of Patents to issue to it a patent containing claims corresponding to the claims of defendant's patent '995.

This memorandum shall constitute findings of fact and conclusions of law as provided for in Rule 52 (a).