This suit for patent infringement was brought by National Dairy Products Corporation under Patent No. 2,919,990 (Podlesak) which was issued January 5, 1960 on an application filed June 22, 1955 by H. G. Podlesak, G. H. Kraft, and R. E. Miller. These men were employees of Kraft Foods, a division of National Dairy. The patent covers a method for packaging cheese. Five defendants are cheese packagers who use the method charged to infringe. Hays-sen Manufacturing Company, also a defendant, supplied machines for use in practicing the method. Separate suits filed against the individual defendants were consolidated by the district court on the question of validity. That issue alone was tried and resulted in a finding of invalidity. In a written opinion embodying his findings and conclusions, the district judge held that although Podle-sak’s invention was not anticipated by the prior art, it was nonetheless not patentable because the evidence submitted did not satisfy the test of nonobviousness set forth in section 103 of the Patent Act of 1952, 35 U.S.C. § 103.

The invention relates to a method for packaging separate units of cheese in a package filled with a preservative gas atmosphere which is substantially free of mold-inducing air. The method is practiced by placing multiple units of cheese in a moving tube of plastic packaging material which is sealed at its forward end and constantly flooded with preservative gas. The progressive collapsing of the packaging tube causes the gas to counter-flow or back-flush around the units and out the tube’s open end, thereby “scrubbing” the surfaces of the multiple units and sweeping out the enclosed air. This method results in the units traveling into an increasing concentration of preservative gases, usually nitrogen or carbon dioxide. As the tube is progressively collapsed over the individual units and the wrapper is cross-sealed and cut into separate packages, the cheese is enveloped by an *889atmosphere of the preservative gas which is substantially free of air.

The essential facts relating to the issue of validity are incorporated in an able and exhaustive opinion written by the district judge. No attack in this appeal is made with respect to those findings.1 Plaintiffs urge the incorrectness of the judge’s conclusion that the method of the patent in suit as a whole was obvious to a person having ordinary skill in the art at the time the invention was made in 1954. Likewise, our only disagreement with the district judge’s decision centers on his resolution of the issue of obviousness.

In arriving at his conclusion of obviousness,2 the district judge followed the teachings of Graham v. John Deere Co. of Kansas City, Mo., 383 U.S. 1, 86 S.Ct. 684, 15 L.Ed.2d 545 (1966). First, he determined “the scope and content of the prior art,” citing Campbell patent No. 2,546,721, Maxfield patent No. 2,160,637, and Ingle-Nawrocki3 patent No. 2,753,268, as epitomizing that art. Campbell ’721 teaches a method for packaging multiple units contained in a continuous tube of wrapping material in individual, closely-fitting wrappers, the keeping qualities of which are enhanced by exhausting the air from the wrapper tube by means of subjecting the units to a partial vacuum prior to severance and sealing. Maxfield ’367 teaches the making, filling, and sealing of containers holding individual units, the keeping qualities of which are enhanced by introducing gas into the package which is only open at one end, thereby displacing to some degree the air in the package. Ingle-Nawrocki ’268 teaches a method for “placing the cheese * * * in an open, flexible, * * * heat-sealable wrapper, flushing the wrapper and the surface of the cheese with a stream of carbon dioxide gas, thereafter applying a vacuum to said wrapper to exhaust substantially all the carbon dioxide therefrom, and heat-sealing the wrapper * * * while under said vacuum.” 4 (Emphasis added.)

Second, in accordance with John Deere, the district judge ascertained the “differences between the prior art and the claims at issue.” Viewing the Swift-Pauly work as “epitomizing the prior art,” he noted the following differences between that work and the Podlesak method: (1) the efficiency of the gas flushing of Swift-Pauly was impeded by imperfect seals and structural impediments to the back flow of gas; (2) the Swift-Pauly work, embodying atmos--pheric exhaustion with gas flushing as only an “incidental act,” was addressed to a “basically different packaging concept from that' of Podlesak.” 5

Third, again following John Deere, the district judge resolved “the level of ordinary skill in the pertinent art.” After enumerating the aspects of the Swift*890Pauly operation which caused it to differ from Podlesak in concept, function, and result, he determined that modifications 6 of the Swift-Pauly method to reach Podlesak “were simple functional expedients within the skill of any mechanic seeking to improve the force of the flow of atmosphere in the particular environment.”

We agree with both the district judge’s description of the prior art and his enumeration of the differences between that art and the Podlesak method. We disagree, however, that “any mechanic” skilled in the art, even after viewing the Swift-Pauly operation in all its aspects, would readily recognize the modifications in that operation necessary to achieve Podlesak, thereby rendering the latter method obvious.

The very fact that those in charge of Swift-Pauly operations, all of whom were skilled and experienced in the art,7 did not, after four years of work, develop the Podlesak method is strong evidence that the invention was not obvious. This evidence of nonobviousness is strengthened by the fact that the Swift-Pauly work with gas, embodied in the combination of the Ingle and Campbell teachings, was finally abandoned sometime in 1954. Despite the knowledge and skill of the Swift-Pauly workers, their experiments led them away from the Podlesak method. This movement is dramatically illustrated by the fact that shortly before the Swift-Pauly work with gas was abandoned, its researchers concluded that “CO2 content is of little importance in mold inhibition.” Moreover, not a single person skilled in the art testified that the Podle-sak method was obvious at the time it was invented in 1954. The defendants’ sole testimony in this regard was provided by a patent lawyer who admittedly was not an expert in the cheese packaging field.

The district judge in the course of his opinion said:

The principle of gas flushing, to the extent it was followed in Swift-Pauly and disclosed in the Ingle patent, was practiced unwittingly without the realization that it could be utilized for less cumbersome, faster, less expensive and yet satisfactory packaging method. (Emphasis added.)

But such an unwitting prior use does not evidence obviousness in a later invention; in fact it indicates just the opposite. Tilghman v. Proctor, 102 U.S. 707, 26 L.Ed. 279 (1880). The Swift-Pauly workers were in pursuit of the same end result as that achieved by Podlesak — a mold-free cheese packaging method — yet they did not recognize the value of using gas alone to displace mold-inducing air. Instead, they relied on a vacuum to displace the air, using gas as an incidental aid in creating the vacuum.8

*891 As a primary contention advanced before this court, the defendants again seek to raise the spectre of plaintiff’s “unclean hands” as a bar to any relief. In response to this same claim below, the district judge stated, “Defendants have not met the requisite burden of proof concerning the allegations of fraud in the procurement of the patent.” We agree. Likewise, we are unpersuaded by the defendants’ contention that the plaintiff’s failure to make known to the Patent Office pertinent prior art during the prosecution of its application and the “failure of the Patent Office to cite art found to be factually pertinent” overcomes the presumption of validity. As the district judge found, the “essential subject matter” of the most pertinent of these uncited prior art references, Campbell and Maxfield ’367, were before the examiner through disclosures in the cited prior art.9

As the Supreme Court indicated in John Deere, “Such secondary considerations as commercial success, long felt but unresolved needs, failure of others, * * * might be utilized * * * as indicia of obviousness or nonobviousness, * * ” 383 U.S. at 15, 16, 86 S.Ct. at 694. Applying these considerations to the instant case, we observe that for a substantial period of time before 1954, the cheese industry sought improved methods of packaging to achieve an attractive product with long shelf life that could be produced efficiently at a low cost. Many methods in use by 1954, including air exhaustion, antimycotics, and combinations of vacuum and gas flushing, all required various hand operations on individual units which increased costs and reduced efficiency. In marked contrast to these other methods, the district court found, “Plaintiff, by use of this method, realized substantial savings, faster production and sufficient shelf life of the product to permit elimination of local packaging plants.” We believe that these prior unsuccessful attempts to satisfy this long felt need in the cheese industry are strong indicia that the Podlesak method, successfully achieving aims so long sought, was not obvious. Rex Chainbelt, Inc. v. General Kinematics Corp., 363 F.2d 336, 337 (7th Cir. 1966). In this regard, we cannot accept defendants’ contention that the commercial success of Podlesak was due primarily to improved machinery, improved packaging materials, and plaintiff’s advertising. For the district court expressly found, “Although the commercial success, in part, may have been due to improvements in machinery and promotional activities, these, in turn, were occasioned by the subject matter of the method of the Podlesak patent.” The worth of the Podlesak method is further indicated by its arrogation by many other cheese packagers in competition with Kraft. Such arrogation can hardly comport with a finding of obviousness.

By looking backward, after an invention has been conceived and put to practical test, the inventive-level concept may appear simple and capable of easy realization by those who are skilled in the art. This sort of view has a certain logical allurement in the instant case. Yet it must be remembered that those who combined the Campbell and Maxfield teachings, the Swift-Pauly workers, skilled as they were, did not hit upon the idea which is central to the Podlesak method. Even Podlesak and his coworkers, despite their skill in the art, conceived their idea only after months of experimentation. In these circumstances, what we said in Zegers v. Zegers, Inc., 365 F.2d 156, 159 (7th Cir.), cert. denied, 385 U.S. 948, 87 S.Ct. 320, 17 L.Ed.2d 226 (1966) is *892pertinent, “There is no doubt, * * * that the fact that the solution to a problem is simple, or appears so, when viewed in retrospect, does not mean the solution was obvious when it was made, and that courts must guard against the exercise of hindsight in assessing the obviousness of a given improvement in the art.”

The judgment is reversed.