This is an appeal from the district court’s judgment invalidating plaintiff’s U. S. Reissue Patent 27,366 and dismissing plaintiff’s infringement suit. We affirm.

The patent in question was issued to plaintiff on July 13, 1970 and reissued on May 16, 1972. It covers a process for making a resilient foam product from a frothed-up wet latex compound. The novelty of the process involves drying the foam without its undergoing a phase inversion known as gelling. ^[1] The resulting product is useful as a resilient backing for such things as shoe innersoles and carpets.

The patented process was invented by Edwin Dunn while he was employed by plaintiff’s predecessor in interest. His process converts a wet, air-frothed latex compound into a finished foam product without gelling the wet foam. The invention is applicable to a wide range of polymers. Dunn relied on surface tension to maintain the foam until it had dried to a cellular solid. Prior processes utilized gelling to strengthen the wall structure of the foam during the drying step. It was usual in the foam coating industry to add a gelling agent to assure that gelling would take place so that foam could withstand the rigors of drying. Dunn added sufficient surfactants to the foam to suppress gellation and strengthen the wet foam cell walls so that they could be maintained during drying without gellation. Dunn’s no-gel process demonstrates substantial advantages; it avoids a delicate formula balancing problem, provides significant cost-saving opportunities and creates fewer production problems. It has been described by the plaintiff as an enormous commercial success.

Plaintiff filed suit on October 29, 1975 claiming that defendants were infringing its product. Jurisdiction and venue in the district court were under 28 U.S.C. §§ 1338, 1400.

Defendants contend that an earlier process developed by Delford R. Ashley was identical to and therefore invalidated the Dunn patent. Pursuant to Fed.R.Civ.P. 42(b), the district court ordered a separate trial on the single issue of the validity of the Dunn patent in view of the Ashley process, specifically limiting the issue to compliance with the requirements of 35 U.S.C. § 102. The question on which the district court centered during the trial was[*1059] whether the earlier Ashley process was a gel process as plaintiff contended or a no-gel process as defendant contended. It found that the Ashley process was a no-gel process and concluded that it anticipated and therefore invalidated the Dunn patent. It is from that determination that plaintiff now appeals.

Although the factual questions before the district court were complex and technical, most of the controlling legal principles are well established. A patent is presumed valid. That presumption is based on the acknowledged experience and expertise of the United States Patent Office and recognition that patent approval is a species of administrative determination supported by evidence. Parker v. Motorola, Inc., 524 F.2d 518, 521 (5th Cir. 1975), cert. denied, 425 U.S. 975, 96 S.Ct. 2175, 48 L.Ed.2d 799 (1976). The presumption of validity, however, is not conclusive. It is weakened when pertinent art has not been considered by the Patent Office. Gaddis v. Calgon Corporation, 506 F.2d 880, 885 (5th Cir. 1975). The burden on one who would invalidate a patent is a heavy one. It has been described variously as one of proof “beyond a reasonable doubt,” Zachos v. Sherwin-Williams Co., 177 F.2d 762, 763 (5th Cir. 1949), and is one “by clear and convincing evidence,” Kiva Corp. v. Baker Oil Tools, Inc., 412 F.2d 546 (5th Cir.), cert. denied, 396 U.S. 927, 90 S.Ct. 262, 24 L.Ed.2d 226 (1969). In Bird Provision Co. v. Owens Country Sausage, Inc., 568 F.2d 369 (5th Cir. 1978), we recognized that as between the two standards “ ‘the authorities are in a morass of conflict.’ ” Id. at 373 (quoting Stamicarbon, N. V. v. Escambia Chemical Corp., 430 F.2d 920, 924 (5th Cir.), cert. denied, 400 U.S. 944, 91 S.Ct. 245, 27 L.Ed.2d 248 (1970). However the burden is characterized, it is clear that significantly more than a mere preponderance of the evidence is required and that doubts are to be resolved against the party who attacks the patent. ^[2]

In this case we are dealing with a patented process. A process is a particular method for achieving a particular result. “It is an act, or series of acts, performed upon the subject matter to be transformed and reduced to a different state or thing .. . the process requires that certain things should be done with certain substances and in a certain order.” Cochrane v. Deener, 94 U.S. 780, 788, 24 L.Ed. 139 (1876). Under 35 U.S.C. § 102, only a novel invention is entitled to a patent. Where all the elements of a patented process are found in a single example of prior art, the required novelty is lacking whether the earlier process was patented or not. Mohasco Industries, Inc. v. E. T. Barwick Mills, Inc., 221 F.Supp. 191, 195 (N.D.Ga.1963), aff’d mem., 340 F.2d 319 (5th Cir.), cert. denied, 382 U.S. 847, 86 S.Ct. 61, 15 L.Ed.2d 86 (1968). A process may be anticipated only by a process. W. D. Haden & Co. v. Mathieson Alkali Works, 122 F.2d 650, 651 (5th Cir. 1941). A process said to anticipate a patented process is evaluated according to what it fairly teaches one skilled in the art. Although disclosures of the prior art may fall short of “complete anticipation,” anticipation may be found where such complete anticipation only requires that one of ordinary skill in the art exercise that skill to complete the work. Deep Welding, Inc. v. Sciaky Bros., Inc., 417 F.2d 1227, 1234 (7th Cir.), cert. denied, 397 U.S. 1037, 90 S.Ct. 1354, 25 L.Ed.2d 648 (1969). See also Laitram Corporation v. Deepsouth Packing Co., 443 F.2d 928, 933 (5th Cir. 1971). To constitute anticipation the process claimed to have predated the patented process must have been complete, reduced to practice, and successfully performed. Coffin v. Og [*1060] den, 18 Wall 120, 85 U.S. 120, 124, 21 L.Ed. 821 (1873); Stamicarbon, N. V. v. Escambia Chemical Corp., 430 F.2d 920, 927 (5th Cir.), cert. denied, 400 U.S. 944, 91 S.Ct. 245, 27 L.Ed.2d 248 (1970).

The trial court considered the Ashley process in the light of these principles. That process was claimed by the defendants to be an exact anticipation of the Dunn process. The evidentiary conflict was largely a battle of experts. Ashley testified that his process was the same as the process disclosed and claimed in the Dunn patent. Ashley’s testimony was supported by defendant’s expert witness, Thomas H. Rogers. To compare the patented method and prior methods, Mr. Rogers performed each of five tests described in the Dunn patent. He testified that in each instance the results showed the processes to be identical. Mr. Rogers also testified that every ingredient in the Ashley formula was identified in the Dunn patent as a suitable ingredient. It was Rogers’ opinion that the Ashley process was the same as the process described and claimed in the Dunn patent.

Plaintiff’s expert contradicted Rogers’ testimony and offered his own theories and tests to demonstrate that the Ashley process was different from the Dunn process.

The trial judge credited Mr. Rogers. He found his credentials as an expert in the field to be impeccable and his testimony to be clear, convincing and unwaivering. He was less impressed with both the credentials and testimony of plaintiff’s expert. Plaintiff complains strenuously that in this treatment of expert testimony, the trial judge committed grievous error. Its counsel argues that the tests recited in its own patent are meaningless in the present context, ^[3] that the patent was misrepresented by Mr. Rogers, that his testimony is untrue and that his conclusions are incorrect. The competency and qualifications required of expert witnesses is a matter committed to the broad discretion of the trial judge. Perkins v. Volkswagen of America, Inc., 596 F.2d 681 (5th Cir. 1979); Barnes v. General Motors, 547 F.2d 275 (5th Cir. 1977). When it sits without a jury, the trial court’s discretion is also the sole determinant of the weight to be given expert testimony. Pittman v. Gilmore, 556 F.2d 1259 (5th Cir. 1977). These principles apply with peculiar force in patent litigation. We said in Bird Provision Co. v. Owens Country Sausage, Inc., 568 F.2d 369, 372 (5th Cir. 1978),

The ultimate question of patent validity is one of law. That conclusion of law, however, must be based on the results of several basic factual inquiries. . . . Findings of facts in patent cases, no less than in other areas of law, are tested on appeal under the strictures of the “clearly erroneous” review standard of Fed.R.Civ.P. 52(a).... Indeed, because patent cases so frequently involve conflicts in the evidence, especially in expert testimony, they seem particularly suited for the review limitations imposed by Rule 52(a).

See also Graver Tank & Mfg. Co. v. Linde Air Products Co., 336 U.S. 271, 69 S.Ct. 535, 93 L.Ed. 672 (1949), aff’d on rehearing, 339 U.S. 605, 70 S.Ct. 854, 94 L.Ed. 1097 (1950). It was stated even more pointedly in Ziegler v. Phillips Petroleum Co., 483 F.2d 858, 861 (5th Cir. 1973), cert. denied, 414 U.S. 1079, 94 S.Ct. 597, 38 L.Ed.2d 485 (1974), as follows:

Because the scientific underpinnings of chemical patents are so complex, trial courts must depend heavily on expert witnesses for both explanation and evaluation of the patent disclosures and the accused infringing operations. Similarly, appellate courts must rely heavily upon the findings and credibility choices made by trial courts. Otherwise, we would be thrust into the position of reconsidering credibility and resolving conflicts in expert testimony, a position that mistakes this Court’s function.

[*1061] The evidence in the record before us was in sharp conflict. On paper the credentials of the experts for both parties were impressive. On the critical fact issues, there was clear support for each contending position. Under such circumstances we cannot find the trial judge’s findings to be clearly erroneous.

Plaintiff argues that the trial court erred in making a finding under 35 U.S.C. § 103 involving obviousness when the trial had been limited by a pretrial order to 35 U.S.C. § 102. The point is well taken and we view the trial court’s Finding 30 as inappropriate. It is clear to us, however, that its other findings were sufficient to establish the conclusion of the patent’s invalidity under Section 102. •

The trial court’s conclusions resulted in invalidation of all eight of the patent’s claims, ^[4] including some which involve materials not utilized in the Ashley process. The affected claims of the Dunn patent involved use of specific starting materials. The use of different starting materials in an identical process, however, does not militate against the court’s findings of invalidity under Section 102.

Notwithstanding the difficult and complicated technical issues, this case ultimately devolved into a fairly simple fact case. Our appellate function is clear. Arguments that the testimony in favor of the plaintiff is more persuasive than the testimony in favor of the defendant, are directed to the wrong forum. Because the findings of the district court are not clearly erroneous, its judgment must be affirmed.

AFFIRMED.