This is an appeal from a judgment of the District Court for the Western District of New York (Burke, J., presiding), sustaining the validity of Patent No. 2,398,382, granted to the plaintiff, Lyon, on April 16, 1946, upon an application filed November 17, 1942. Of the nine claims all but 6 and 7 are in suit; the judgment held all seven valid, and the defendant concedes infringement. Judge Burke, in a long and careful opinion, 119 F.Supp. 42, has stated the evidence in such detail that we may, as we shall, assume familiarity with it in our discussion of the issues involved.
The defendant first challenges the sufficiency of the specifications to support the claims. Throughout the specifications the “coating” to be vaporized is described only as a “suitable coating * * such as metallic fluoride, metallic sulphide, oxide or the like” (page 1, col. 2, lines 26-28), or “mixtures” of these (page 2, col. 1, lines 26-28). The first five claims speak of “a stable, water insoluble, evaporated inorganic salt,” and of the last two claims in suit number eight is for a “metallic fluoride” and number nine for “magnesium fluoride.” The defendant complains that “suitable coatings” is too vague a description, even when accompanied, as the phrase was, by *531the specific substances mentioned. Considering the kind of coatings that the art had been using, an acquaintance with which the applicant might impute to anyone practising the process, this objection appears to us unduly captious and hostile to the underlying canons of patent interpretation. In any event, Finding 44 finds facts that dispose of the objection. Indeed, it cannot at any rate apply to claims 8 and 9 where the “inorganic salt” is described as “metallic fluoride,” and “magnesium fluoride.” As for the other claims in suit we can see no reason to suppose that when the “suitable coating” of the specifications is limited to a “stable, water insoluble, evaporated inorganic salt,” anyone seeking to use the process would be left in doubt. The claims always measure and limit the scope of the monopoly; and the defendant does not, and could not properly, suggest that anything that Lyon actually invented he did not disclose when the limitations of the claims are imputed to the description in the specifications. Therefore we are not disposed to draw any distinction between claims 8 and 9 and the other five in suit, though, if we did, it would at most only result in a modification of the judgment not important in this action. As to the objection that the claims confuse or combine process and product, it is so patently untenable that we shall not discuss it.
The important questions are whether the invention had been disclosed in any earlier patent, or had been publicly used, before Lyon filed his application on November 17, 1942; and whether his contribution will support a patent. The process was in two steps: first, to heat the “optical surface” in a vacuum until “adsorbed water and grease have been evaporated from the surface,” (page two, col. 1, lines 47, 48); and second, to vaporize an “inorganic salt” within the vacuum, meanwhile keeping the “optical surface” heated; “baking of the optical element, evacuation of the chamber and vaporization of the coating material are continued until a suitable layer of the desired thickness * * * has been deposited,” (page 2, col. 1, lines 64-68). None of the earlier patents disclosed this sequence, and the nearest were as follows. In November, 1925, Fink and Beers filed the application that issued on December 10, 1929, as Patent No. 1,738,-991. It was for a process of coating the inner surfaces of the vacuum chamber of a thermos bottle with vaporized metal, such as magnesium. The purpose of such a coating is to maintain the temperature in the chamber by excluding the heat from the air, or the heat from the contents of the bottle. In April 1929 Winkler and two others applied in Germany for a patent that issued in this country as Patent No. 1,982,774 on December 4, 1934. This was for mirrors made by condensing in a vacuum vaporized metals — e. g. an alloy of aluminum and silver — upon a polished base. (“The adhering capacity of the mirror coatings on metallic substrata is quite considerably increased by keeping the polished metallic substrata hot during the deposition of the mirror coatings.”) There were three earlier patents for coating the upper part of the inside of an electric bulb, so that none of the light should be reflected upwards. There were also patents for mirrors, designed of course to reflect the light; and therefore the opposite of the patented coatings, whose purpose is to transmit through the “optical surface” as much as possible of the light falling upon it. The defendant indeed insists, — relying for this on line 12, col. 1, page 1, of the specifications in suit, —that “mirrors” are among those “optical surfaces” or “elements” that the invention covers. However, regardless of any significance of that word where it appears, there cannot be the least doubt that the invention is only for a “light-transmitting optical” element; indeed the coating itself is spoken of as “a light-transmitting evaporated film” (page 1, col. 1, lines 47, 48). The explanation for the inclusion of “mirrors” may be that, as the reflecting coating of a mirror is usually on the rear surface of a transparent plate, it is desirable to have as little light as possible reflected from the front of the plate, so that all would reach the reflecting surface. The pur*532pose of all these patents so widely diverged from that of Lyon that under § 100(b) they would not serve as anticipations', even if the. sequence of steps had more closely anticipated his process.
On the other hand Cartwright and Turner took put three earlier patents that were much nearer to Lyon’s, for they disclosed coating an “optical surface” with a film of “inorganic salt.” The first — No. 2,207,656 — was .applied for in December, 1938 and issued on July 8, 1940. It disclosed a process of depositing a coating of non-reflecting substances — -“metallic-fluorides” — upon a lens or the like by vaporizing it in a vacuum “in the manner now generally well known in the silvering of mirrors and the like” (page 2, col. 1, lines 21-23). The second — -No. 2,281,-474 — was applied for in March 1939 and issued on April 28, 1942. It disclosed a method of superimposing at least two coatings of different “indices of refraction”; and presupposed the vaporization of “inorganic substances” in a vacuum. The third — No. 2,281,475 — was applied for in August 1939 and issued on the same day as the second: it was for a process (and for its product) of coating a lens with similar substances by vaporizing them in a vacuum and thereafter “baking” at from 350° to '500° C. in order to increase their resistance to detachment — “ruggedness,”—a practise that became known in the art as “post-baking.” The specifications of none of these patents suggested that the “optical surface” should itself be kept hot while it was being coated; the nearest approach to this being in the third patent which did declare it desirable to bake the “surface” at between 400° to 450° C. before vaporization (page .1, col. 1, lines 41-45), — a step that has come to be known as “preheating,” and that is, as we have said, the first step of the patented process. Thus, although Lyon’s advance lay only in keeping the “optical surface” itself heated while it was being coated, this had not appeared in any patent or printed publication on November 17, 1942. Before considering whether the patent may depend upon it, we will consider whether Lyon’s process had been “in public use or on sale” under 35 U.S.C.A. § 102(b); or whether “the invention was made * * * by another” before Lyon, under § 102(g).
Cartwright was one of the most experienced experts in this field, and had been active in experiment even before the year 1938. The military services, and particularly the Navy, were particularly anxious to find a practical way of coating binoculars,' periscopes and other “light-transmitting” glass surfaces, which should make the film at once tenacious and viable. Indeed, some of the testimony imputes to the services a scarcely understandable estimate of the importance of such a coating. The trouble had been that all the earlier coatings, though their proper composition was known, could be readily scratched or even rubbed off. One condition of .success was well understood: before being coated the glass surface should be made as clean as possible: all water or grease must be taken off. An accepted way to do this was Lyon’s first step: heat the glass in a vacuum. Again, a familiar, perhaps the only practical, way was to coat the glass after it had been cleaned by vaporizing the “inorganic salt” in the same vacuum in which the glass had been “preheated.” Finally, it was well understood that to “post-bake” added much to the' tenacity of the bond and the “ruggedness” of the coating. However, for some reason, apparently still not quite understood, the art never took Lyon’s second step, now generally accepted as the way to secure the optimum bond of the coating to the glass.
Cartwright did indeed fully disclose even this in a letter, written on November 30, 1939, to one, Wilson, of the “Research Corporation,” a company engaged! in the exploitation of patents, in whose employ he then was, though his chief calling was that of instructor in physics in the Massachusetts Institute of Technology. This letter described the “development made during the last few months on ‘invisible glass’ ”; and the third and ninth experiments which it de*533scribes we quote in the margin.1 Not only does the letter prove beyond cavil that Cartwright did conceive that to coat the glass in a vacuum, while heated, made the coating “very much harder,” but it also shows not only that he had practiced it, but that he considered the possibility of “getting protection on the method”: i. e. having it patented. Furthermore, on the day before, — November 29th, — he had told one, Hewlett, a physicist of the General Electric Company, of coating the glass while it was hot, and on December 4th Hewlett wrote to his own company describing that process, and adding that “the ruggedness of these films is very striking (see the two samples he gave me). They may be washed in tap water and even scratched with a knife blade without producing any scratches.” Moreover, neither does the letter, nor any testimony, intimate that Cartwright imposed any promise of secrecy on Hewlett, although Hewlett might be regarded as in some sense a competitor. Hewlett experimented with the information he had got, and made a number of coatings in the summer of 1940, but finally decided to abandon the process, because he concluded that those coatings deposited on heated glass scratched more easily than those on unheated. Since his user was not only experimental, but in addition was abandoned, we disregard it under § 102(g); as we also do the work of one, Begg, an employee of the Eastman Kodak Company, which that company attempted to resuscitate later and to use as an anticipation of Lyon. But Cartwright’s activities we may not so summarily dispose of, and to them we turn.
By the fall of 1939 he had already found that his work on non-reflective coatings took up so much of his time that he employed a student named Clark to help him, and Clark kept with him until May, 1940. Before he left, Cartwright had enjoined upon him the strictest secrecy which Clark observed, not only while he was in Cartwright’s employ, but, also, after he was employed by another company, the National Research Corporation. Cartwright himself continued to work for the Institute until August, 1940, when he went into the employ of the Navy for that month, during which, although he was engaged in making non-reflecting lenses, he did not use the patented process; nor does he appear to have told anyone about it. He gave as his reason for not using it that he had no apparatus that would produce a high enough vacuum. On the other hand, al*534ready while he was at the Institute, he made and sold coated lenses, as a private venture; and, after he left the Navy and went into the service of the Corning Glass Works at Corning, New York, he again took up the same business with the help of his wife. It resulted in a substantial income, at times amounting to $3,000 a year, and many, though not all, of the lenses were coated while hot.
Were these activities either a “public use” or “sale” under § 102(b), or was Cartwright a prior inventor under § 102 (g) ? True, the fact that the process could not be deduced from the lenses sold would not alone prevent their sale from being a “public use.” 2 It is indeed difficult to see why this should be so, for such sales could not in fact enlighten the art, especially in the case of a process, which the product does not ordinarily disclose, unlike a machine which may disclose its concealed mechanism, if it be dismantled. The law does not however appear to have made any such distinction. However, although such a product will be a “public use,” it will be so only on condition that it is neither experimental, secret, nor abandoned; and we think that Cartwright abandoned his discovery as soon as it emerged from the stage of experiment. We do not rely upon the secrecy that he enjoined upon Clark except in so far as it goes to confirm the continued lack of finality of his activities. His disclosure to Hewlett is hardly consistent with unqualified secrecy, although it too is not inconsistent with the continued tentative character of what he was doing. But we agree with what Judge Burke said in his opinion [119 F.Supp. 48]. “Whatever experiments he had conducted * * * and whatever occasional use he had made of it in his commercial work had convinced him that it was inferior in results to his own patented post-baking method, and that the results obtained were not worth the effort, and did not warrant bringing it to light in the art. * * * His testimony at the trial leaves no doubt about the fact that he is still uncertain whether one 'could not get as hard a film if you left out the actual heating in the vacuum.’ ” The defendant replies that it is not necessary that a third party who puts the invention in “public use or on sale” must understand the value and importance of his discovery. That is indeed true; a man may conceive a process, put it in “tangible form,” test it out and fail to exploit it, because he does not see how valuable it is. That was not, however, all that Cartwright did. He did indeed test it out to his satisfaction; and when he had done so, he concluded it would not do what he was after. It did not produce a more “rugged” film and he gave it up. It was in effect an abandonment; it did more than fail to advance the art; it put the process among those efforts that are proved useless. All the reasons that have made the courts refuse to treat experimental users as anticipations,3 apply even more convincingly; it is not alone that such activities are not evidence of anticipation, they are evidence against. We do not forget that in Corona Cord Tire Co. v. Dovan Chemical Corp., 276 U.S. 358, 383, 48 S.Ct. 380, 387, 72 L.Ed. 610, the Supreme Court said of a process that it is “reduced to practice when it is successfully performed”; but it should be noted that the Court also declared that “the tests of Kratz” were not “abandoned laboratory experiments.” We conclude therefore with Judge Burke that Cartwright neither put the process to “public use,” nor was he its prior inventor.
Therefore we at length come to the question whether Lyon’s contribution, his added step, was enough to support a patent. It certainly would have done so twenty or thirty years ago; indeed it conforms to the accepted standards of *535that time. The most competent workers in the field had for at least ten years been seeking a hardy, tenacious coating to prevent reflection; there had been a number of attempts, none satisfactory; meanwhile nothing in the implementary arts had been lacking to put the advance into operation; when it appeared, it supplanted the existing practice and occupied substantially the whole field. We do not see how any combination of evidence could more completely demonstrate that, simple as it was, the change had not been “obvious * * * to a person having ordinary skill in the art” — § 103. On the other hand it must be owned that, had the case come up for decision within twenty, or perhaps, twenty-five, years before the Act of 1952 went into effect on January 1, 1953, it is almost certain that the claims would have been held invalid. The Courts of Appeal have very generally found in the recent opinions of the Supreme Court a disposition to insist upon a stricter test of invention than it used to apply — indefinite it is true, but indubitably stricter than that defined in § 103 4 Indeed, some of the justices themselves have taken the same view.5 The Act describes itself as a codification of existing law, as it certainly is in the sense that the structure of the system remains unchanged. Moreover those decisions that have passed upon it have uniformly referred to it as a codification, although so far as we have found none of them has held that § 103 did not change the standard of invention.6 And so the question arises whether we should construe § 103 as restoring the law to what it was when the Court announced the definition of invention, now expressly embodied in § 103, or whether we should assume that no change whatever was intended. To decide that question it seems desirable to look briefly backward.
From 1793, when the second patent act was passed,7 until the Act of 1952, the only statutory standard for invention was that the discovery should be “new and useful”; and indeed the Act of 1952 itself repeats this same test in § 101. Congress did not try to define it but left it to the courts to develop by precedent. So far as we can find, it was not until 1850 that the Supreme Court made any such attempt; so that, although the disclosure must be “new,” it was so, provided it had not been published or in public use and was original. However, in Hotchkiss v. Greenwood, 11 How. 248, 267, 13 L.Ed. 683, the Court imposed an authoritative gloss upon the word, which it put in the following words: “unless more ingenuity and skill in applying the old method” were necessary “than were possessed by an ordinary mechanic acquainted with the business, there was an absence of that degree of skill and ingenuity which constitute essential elements of every invention. In other words, the improvement is the work of a skillful mechanic, not that of the inventor.” The instruction to the jury, the *536exception to which the court overruled, had been in substantially the same words, 11 How. at page 253, 13 L.Ed. 683: if “no other ingenuity or skill” be “necessary to construct the knob than that of an ordinary mechanic acquainted with the business, the patent is void.” Thereafter this became the standard rubric and was applied in many cases. The variants were numberless; and “invention” became perhaps the most baffling concept in the -whole catalogue of judicial efforts to provide postulates for indefinitely varying occasions. However, the Court never formally abjured it; nor has it ever substituted any other definite test. Even Cuno Engineering Corp. v. Automatic Devices Corporation, 314 U.S. 84, 62 S.Ct. 37, 86 L.Ed. 58, — which in expression probably was the furthest departure — recognized the continued authority of Hotchkiss v. Greenwood, supra; 314 U.S. at page 90, 62 S.Ct. at page 40: “if an improvement is to obtain the privileged position of a patent more ingenuity must be involved than the work of a mechanic skilled in the art.” Again 314 U.S. at page 91, 62 S.Ct. at page 40: “The principle of the Hotchkiss case applies to the adaptation or combination of old or well known devices for new uses.”
We assume that the usual presumption against retroactivity, would have applied to § 103, had the purpose not been so explicit to make the Act retroactive; but § 4(a) leaves no possibility of doubt as to that: it is to “apply to unexpired patents granted prior to such date” January 1, 1953 “except as otherwise provided”; and all that was “otherwise provided” was in § 4(e): “Nothing contained in Title 35, as enacted by section 1 hereof, shall operate to nullify any judicial finding prior to the effective date of this Act on the validity of any patent by a court of competent jurisdiction.” 66 Stat. 815.® This interpretation is, moreover, confirmed by the report of the House as to § 103, which we quote in the margin.8 9 Therefore, we can see no escape from accepting the text as it reads, except as “at a later time * * * some criteria * * * may be worked out.” It only remains to ask whether, if the change is to apply retroactively, the Act is pro tanto unconstitutional. Arguendo we will accept it as possible that the decisions of courts may give such an absoluteness to the words of a statute as to make a deliberate substitution of another locution the same as though the precedents had been incorporated verbatim into the statute; that, if that were true in the case at bar, the Act' would in effect grant new monopoly; and, finally, that the grant of a new monopoly retrospectively would be invalid.10 We may accept that hypothesis because the facts do not fulfill the promises. In the first place § 103 only restores the original gloss, substantially in ipsissimis verbis; which has never been overruled; but on the contrary for seventy or eighty years had continued to be regarded as authoritative. Moreover — and this is the important consideration — although it may have ceased in practice to be followed, and had come to enjoy no more than lip service, there never has been the slightest intimation of any definite substitute ; nothing more than an unexpressed and unacknowledged misgiving about the increased facility with which *537patents were being granted. Such judicial attitudes are indeed the stuff of which much of the law is made; but we cannot agree that, however controlling upon the lower courts, they are a warrant for that solid assurance, the disappointment of which will make a statute invalid. Courts again and again shift their position ; and, although they are apt to do so under cover of nice distinctions, they impose the risk of anticipating the changes upon those who may have acted upon the faith of the original. Certainly a legislature, whose will the courts have undertaken to proliferate, must be free to reinstate the courts’ initial interpretation, even though it may have been obscured by a series of later comments whose upshot is at best hazy.
Finally, we cannot agree that this retroactivity should be limited to a recognition of the patent only in the future: that is, that, although the pat-entee may enjoin its infringement after January 1, 1953, he may not recover damages for any earlier infringement or collect any profits of the infringer. As a matter of construction § 4(e) seems to us conclusive; it preserves any “finding” of a court before January 1, 1953; but leaves the court free thereafter to apply the Act as though it had existed when the unexpired patent issued. Were we to deny the patentee the usual remedies, an anomalous situation would arise; for it would follow that, although an infringer would not be liable, quoad the recovery of damages or profits, for taking his chances in disregarding the patent, he would nevertheless have no protection for any investment or other industrial expense that he might have made before the date when the Act took effect. We should either accept as an adequate protection such assurance as infringers might have drawn from the earlier precedents, or we should find them as we do, too amorphous for that purpose. Believing that they will not serve, we hold that the patent should be regarded as valid ab initio for all purposes.
There remains only one. other point of enough importance to demand discussion: i. e. whether what Lyon had done before November 17, 1941, — a year before his filing date — forfeited his right to a patent under § 102(b) of the Act: that is, whether he had put it in “public use or on sale in this country, more than one year prior to the date of application”. Lyon began to work for the Naval Research Laboratory in January, 1941, trying to find an effective coating. By the middle of June, 1941, he was assured that he “had consistent results as far as hardness and adherence of the film was concerned.” He “appeared to be on the right track”; but much testing had still to be done. Nevertheless, “by the first of July” he “was reasonably certain that” he “was getting a hard durable film * * remarkably consistent in the results.” “Many, many tests were made,” which, however, by September, 1941 had convinced him “that we were in a position to try this thing out on some actual simple instruments.” There were a number of old binoculars, which were given to him to coat and it was arranged that these should be introduced surreptitiously, so that the users would not know which had been coated by the new process and which had not. From binoculars he went to other instruments, so that by October first the “Optical Shop of the Naval Gun Factory” had “completed certain experimental work,” and “estimated that $10,-000” was “required to provide for equipment and further research.” The “experimental work” had “developed to such a stage as to place the product in production.” The officer in charge concluded the letter from which we quote by requesting that the Bureau of Ordnance should approve a “project” for $20,000 both for “production equipment in the shop” and for five units for “repair activities afloat.” Such an approval was issued on October 30; but it is to the last degree unlikely that any production should have taken place within three weeks, that was not part of the “further research” deemed necessary on October 1st. *538At any rate it is clear that the defendant, which had the burden, did not prove its defence. Its theory is that, if with Lyon’s consent the Navy exploited the process, it was the same as though he had done so himself; indeed in the end the Navy’s exploitation would be far more profitable to him than any sales that he could personally make. It is not necessary to pass upon the legal validity of that interpretation of § 102(b), because, though we should accept it arguendo, it was not proved, as we have just said.
We agree with Judge Burke’s searching and comprehensive analysis, and the judgment will be affirmed.
Judgment affirmed.
On Petition for Rehearing
The petition for rehearing is denied except as to the point raised by the first two sentences of Point IV, subdivision 1, thereof. As to the point thus excepted from denial further consideration brings us to the conclusion that the patent specifications fail adequately to support claims 1, 2, 4 and 5 which describe the coating claimed therein in terms so broad as to include materials which, so far as the proofs show, produce optical surfaces which are not light-transmitting surfaces.
We hold, therefore, that, without disturbing our affirmance as to claims 3, 8 and 9, the decree below must be reversed as to claims 1, 2, 4 and 5 for lack of adequate disclosure. With leave, however, to the plaintiff, if so advised, to file a petition for rehearing addressed to the holding above stated, containing any pertinent argument not covered in his. earlier briefs.