On March 3, 1919, Hans Vogt, Josef Engl and Joseph Massolle filed a joint German application for letters patent for a sound-reproducing apparatus. On April 4, 1921, Vogt, Engl, and Massolle made a joint application for letters patent of the United States for a sound-reproducing apparatus. Appellees American Tri-Ergon Corporation and Tri-Ergon Holding A. G. are the assignees in interest of all rights under the application of Vogt, Engl and Massolle for letters patent of the United States. For convenience, this application and the parties in interest will be referred to hereafter as Vogt.
One claim of the Vogt application was allowed by the Patent Office. It reads as follows:
“5. In an apparatus for the synchronous recording and reproduction of optical acoustical operations, the combination of a record in the form of a film in motion, a constant source of light radiating upon said film, a photoelectric cell acted upon by said constant source of light for the reproduction and a condenser circuit fed with un-damped high frequencies with which said photoelectric cell is connected up in parallel so that the electrical oscillations are produced in rythm [sic] with the photographic records.”
*939The following four claims were rejected by the Patent Office as lacking in invention :
“1. In sound reproducing apparatus, the combination of a photoelectric cell having an alkali cathode, means for illuminating the cell with acoustically-modulated light, and a circuit connected with said cell in which flows a correspondingly modulated current.
“2. In sound reproducing apparatus, the combination of a photoelectric cell having an alkali cathode, a constant source of light, means for uniformly progressing a film between said source and cell, said film carrying a photographic sound record of varying opacity, so that said cell is illuminated by acoustically-modulated light, which has passed through the film, and a circuit connected with said cell in which flows a correspondingly modulated current.
“3. In sound reproducing apparatus, the combination of a photoelectric cell having an alkali cathode, means for illuminating the cell with acoustically-modulated light, an amplifier, a circuit connecting said cell with said amplifier, a sound reproducing device, and a circuit connecting said amplifier with said sound reproducing device.
“4. In sound reproducing apparatus, the combination of a photoelectric cell having an alkali cathode, means for illuminating the cell with acoustically-modulated light, an amplifier, a circuit connecting said cell with said amplifier, and a circuit connected with said amplifier, in which flows amplified currents which vary correspondingly with the variations of said acoustically-modulated light.”
Thereafter, a suit was brought in the District Court, upon the four rejected claims, under R.S. § 4915, 35 U.S.C.A. § 63, and a decree 1 was entered on December 9, 1931, in which it was adjudged that claims 1 to 4, inclusive, defined a meritorious invention, which was novel and patentable; and authorized the Commissioner of Patents to issue letters patent therefor to appellee American Tri-Ergon Corporation.
In the meantime, in August, 1920, appellant Delmar Whitson filed an application for letters patent of the United States for “Method of and Apparatus for Reproducing Sound.” In April, 1929, Whit-son made, in his application, claims identical with claims 1 to 4, inclusive, in the Vogt application. Appellant Whitson Photophone Corporation is the assignee in interest of all rights under the Delmar Whitson application; and both, for convenience, will be referred to hereafter as Whitson. Also, in the meantime, in December, 1922, appellant Albert Radtke filed an application for letters patent of the United States for “Method of and Means for Optically Recording and Reproducing Sound;” and nine years later— on the same day as the date of entry of the decree authorizing issuance of patent on the Vogt application — Radtke made, in his application, claims identical with claims 1 to 4, inclusive, of the Vogt application. Appellant Radtke Patents Corporation is the assignee in interest of all rights under the Albert Radtke application; and both, for convenience, will be referred to hereafter as Radtke. In January, 1932, instead of granting a patent as authorized by the decree entered during the preceding month, the Commissioner of Patents instituted an interference proceeding between Vogt, Radtke and Whitson. On July 18, 1936, the Examiner awarded priority to Radtke. On March 17, 1937, the Board of Appeals affirmed the award of the Examiner. On January 22, 1938, the Board of Appeals, by its decision, following a rehearing, declined to change its decision of March 17, 1937. A patent covering the four counts was issued to Radtke April 19, 1938.
In May, 1938, appellees American TriErgon Corporation and Tri-Ergon Holding A. G. brought the present suit in the District Court under R.S. § 4915. On November 20, 1939, that court filed its decree,2 awarding priority to Vogt as of that date, and authorizing the Commissioner of Patents to issue letters patent to American Tri-Ergon Corporation embodying the four claims. It was stipulated by all parties: “4. That prior to 1919 Germany was and ever since has been a ‘convention’ country with respect to R.S. § 4887, 35 U.S.C.A. § 32, and its citizens since prior to 1919 have been entitled to the privileges of said statute. 5. That the privileges accorded by the Nolan Act of March 3, 1921, 41 St.L. 1313, 35 U.S.C.A. § 80 et seq., were on or about October 21, 1921, extended to citizens of Germany.”
*940The question mainly urged by the parties for our determination is that of priority. The decision of the District Court, from which this appeal is taken, is chiefly concerned, also, with that question. Except for a conclusion of law and decree that Vogt and his associates are “the true, first, original and joint inventors of the invention * * * ” and that each of said claims 1 to 4, inclusive, defines a meritorious invention, novel and patentable as of March 3, 1919, and except for two findings, which will be hereafter set out, the court relies, largely, upon the findings and conclusions which appear in its earlier decision of 1931. In its opinion, the court said: “This Court further finds, upon a careful examination of Justice Adkins’ decision, that Justice Adkins specifically held Vogt’s effective date to be March 3, 1919, and further finds that in order for Justice Adkins to have authorized the Patent Office to issue a patent to Vogt, he had to find that the Vogt patent disclosed the four counts. This Court further holds that while Justice Adkins’ decision does not, as to this Court, involve the principles either of res adjudicata, or of stare decisis, it is a considered finding in a controverted case that Vogt disclosed the four counts at issue, and should control this Court, unless this Court should be perfectly clear that there was error in the prior decision, or that there was, in this case, additional evidence offered which might have induced a different decision in the prior case if it had been there offered. The Court finds neither the error nor the evidence.” (Italics supplied)
However, the first and most important question which was presented for decision to the District Court, which was presented on this appeal by amicus curiae, and which we must now decide, is whether the four claims in issue do define an invention, patentable as of March 3, 1919. Whether or not Vogt is entitled to the earlier date and whatever the merits of the respective contentions concerning priority, it is necessary, first, that a sufficient showing be made of invention over the prior art as it existed on March 3, 1919, which is the earliest filing date claimed by any of the parties. This is true in spite of the fact that no appeal was taken from the 1931 decision. As the patent therein authorized to be issued to Vogt has never been issued, and as the present suit arises again under R.S. § 4915, the question of patentability, thus presented to us for the first time, is the question of primary importance.3 We note, in passing, Vogt’s contention that this is not properly before us although — it admits — the issue might properly have been presented under R.S. § 4915, if fair notice had been given of the grounds of such a defense; if convincing evidence had been presented in addition to that considered by Justice Adkins; if opportunity had been given to rebut such evidence by cross-examination or additional testimony; and if the issue had been passed upon by the District Court. Vogt cites no authority for this claimed limitation upon Section 4915, or upon Hill v. Wooster. We know of no such authority. Moreover, a large quantity of additional evidence was presented by all parties; severe cross-exam*941ination and thorough rebuttal occurred; the District Court considered the issue; and passed upon it. “In our capacity as guardian of the public interests, * * * we must consider the question.” 4
We note, also, that at this point the principle of Abbott v. Coe5 becomes aplicable, i. e., that the “judicial function is exhausted when there is found to be a rational basis for the conclusions approved by the administrative body.” Availing itself of its equipment for deciding the intricate and technical questions which it is pre-eminently qualified to solve, the Patent Office decided in 1927 that no invention was involved in the claims here in controversy and refused to issue a patent. If, therefore, there was a rational basis for that decision, it should be sustained.
The following excerpts from the opinion of the Board of Appeals of the Patent Office, rendered in 1927,6 succinctly state the issues, and the reasons which, in our opinion, support the conclusion that there was no invention:
“What appellants are claiming is a sound reproducing apparatus including a photoelectric cell. Specifically the subject matter appealed relates to the reproduction of sound from a photographically recorded film record which may also include optical representations as well as sound records. It is a purpose of the claimed construction to reproduce the sound in synchronism with the optical effects although the claims are directed only to sound reproduction. In the specific embodiment of reproducing apparatus disclosed, a photoelectric cell having an alkali cathode is employed. The cell is illuminated by acoustically modulated light thereby producing in an electric circuit connected with the cell a correspondingly modulated current which may be heard in a telephone. The current passes through an amplifier before it reaches the telephone.
“Prior to appellants’ entry into the field it had been proposed to illuminate a selenium cell with acoustically modulated light and to utilize the current so produced to operate a sound reproducer. This is disclosed in the Fritts patent. What appellants have done, therefore, is to substitute for the selenium cell of the Fritts patent a photoelectric cell and to amplify the resulting current. As showing a photoelectric cell employed in a reproducing apparatus the examiner cites the British Polyphos patent. It is the view of the examiner that it did not involve invention to make the noted substitution in the Fritts patent under the teachings of the British patent. In the British patent it is proposed to use the photoelectric cell in the reproduction of graphic signs, pictures and the like. The specification of the British patent points out the superiority of the photoelectric cell over the selenium cell in the reproduction of pictures owing to the absence or substantial reduction of time lag in the photoelectric cell. The examiner holds that the reproduction of pictures by means of the photoelectric cell is quite analogous to the reproduction of sound by such a cell and accordingly holds that it would be uninventive to substitute in the Fritts apparatus the photoelectric cell of the British patent.
sji % jf:
“ * * * The examiner urges that even in the absence of the disclosure of the British patent it did not involve invention to employ a photoelectric cell in the reproducing apparatus of the Fritts patent in view of the known characteristics of that form of cell. However that may be we are of the opinion that the appealed claims are lacking in invention over the combination of the Fritts and British patents.
“Claims 3 and 4 include in addition to the general organization specified in claims 1 and 2 an amplifier in circuit with the photoelectric cell and the sound reproducing device. The examiner, in the statement on appeal, does not discuss claims 3 and 4 separately and only states with respect to current amplification that since the current from the photoelectric cell is almost infinitesimal it must be amplified before available for use. Appellants comment on the failure of the examiner to cite a reference disclosing amplification of the current derived from a photoelectric cell and urge this as an additional reason why claims 3 and 4 should be allowed. It may be that the failure of the examiner to cite a reference showing current amplification was due to his knowledge that it is common practice *942to amplify weak electric currents in all sorts of situations and systems and we might take judicial notice of that fact. However, in order that the record may be complete we call attention to the patent to Hull, No. 1,313,187, granted Aug. 12, 1919 on an application filed Nov. 8, 1915 as showing the amplification of an electric current derived from a photoelectric cell.” (Italics supplied.)
It was this Patent Office decision which the District Court rejected in its decision of December, 1931. In the present case the issues of invention and patentability were retried and a voluminous new record was made. Nevertheless, as already indicated, the District Court confined its second decision, in large part, to the question of priority and made only two findings which bear upon the question of invention, as follows :
“6. * * * a lengthy trial was had before Justice Adkins, and extensive findings of fact and conclusions of law were filed by Justice Adkins and a final decree was entered to the effect that the four claims here in controversy were patentable, * * *.
* * * * *
“11. The invention upon which Plaintiffs seek the grant of a patent under said Vogt et al. U. S. application is defined by the four counts or claims recited on page 1 of the opinion of this Court in this suit. Said invention as more fully set forth in the opinion and findings of Justice Adkins in the prior suit, Equity No. 47,904 aforesaid, involves the discovery that a photoelectric cell of the alkali cathode type in conpmction with appropriate immediate ciraüt connections therefor could be used in sound reproduction apparatus in such manner that' the resulting reproduced sou/nds correspond substantially with the sounds originally produced or recorded upon a film record ” (Italics supplied)
On this appeal, it becomes necessary, therefore, for us to look to Justice Adkins’ findings, and to the voluminous. new record7 — much of which of course was not available to him — to determine, first, the condition of the prior art as it existed in 1919, and, second, whether the findings made in 1931 correctly stated the prior art. In doing so we shall keep in mind the fact that the interpretations placed upon the four claims, by the three parties to the present appeal, vary in considerable degree. Moreover, in their adverse contentions concerning priority, each challenges, as well, positions taken by one or the other concerning the condition of the prior art, and as to what disclosures in the four claims constitute invention. It is well to discover, therefore, as clearly as may be done, the contentions now made by each of them, and to check their contentions against the findings and the record.
Whitson’s claim of invention is the broadest of all: “ * * * merely the use of a photo-electric cell to reproduce sound by acoustically modulated light in any operative circuit of prior art.” Radtke’s present position is stated in its brief as follows: (1) “There were four counts in the interference. The Patent Office found that Radtke was prior as to all four, but the Radtke patent later by amendment limited counts 1 and 2 to an amplifier combination (its claims 30 and 31) so now, in the exact language, the patent contains only the third and fourth counts, as claims 32 and 33 respectively (3R. p. 1804).” (Italics supplied.) (2) “ * * * that he was pri- or to all parties as to all counts; but that the first two counts are so broad in their wording as probably to be unpatentable in view of prior art in addition to that before Justice Adkins. Radtke, in good faith, by amendment limited counts 1 and 2 in his patent to the wording of his corresponding claims 30 and 31 which contain additional limitations over counts 1 and 2. Counts 3 and 4 are patentable only when based on a disclosure of an operative circuit connecting the photo-electric cell with the amplifier tube, examples of which operative circuit were disclosed by Radtke but not by TriErgon or. Whitson.” (Italics supplied.) Vogt’s present position is more difficult to determine. It incorporates into its brief Justice Adkins’ findings numbered 4, 5, 6, 7, 9, 10, 11, 12, 13, 14 and 15; and apparently adopts them to indicate the nature and extent of its discoveries and alleged invention. Again, in its brief it says: “ * * * it appears by comparison of these claims with the actual contribution to the art which Vogt made, as summarized in Mr. Justice Adkins’ findings, that the claims are no broader than his actual *943achievement, and appropriately distinguish from the prior art as also shown in these Findings, in respect to the novel combinaition of elements by which Vogt was able for the first time in the history of the •world to successfully reproduce sound from a film record or acoustically-modulated light.” (Italics supplied.) It denies, however, that it claims “the result of obtaining sound reproduction from a film record.” Again, it says: “The Findings of Mr. Justice Adkins sufficiently support his Decree that the subject matter in suit, is patentable, and the present record falls far short of containing any additional or conflicting evidence which would justify this Court in upsetting his decision.” Vogt also adopts, in its brief, Justice Goldsborough’s finding No. 11, in the present case, which reads in part as follows: “11. * * * Said invention as more fully set forth in the opinion and findings of Justice Adkins in the prior suit, Equity No. 47,904 aforesaid, involves the discovery that a photoelectric cell of the alkali cathode type in conjunction with appropriate immediate circuit connections therefor could be used in sound reproduction apparatus in such manner that the resulting reproduced sounds correspond substantially with the sounds originally produced or recorded upon a film record.” When we examine Justice Adkins’ findings, we find that up to a certain point they are in accord with those of the Patent Office decision of 1927. Thus, he found (No. 6) that for many years prior to 1919 the selenium cell had been often suggested as a light sensitive means for translating acoustically modulated light rays into electric current variations for sound reproduction; that from time to time prior to 1919 various patents and publications had issued, directed to apparatus and methods of reproducing sound by projecting light through a variably blackened strip on to light sensitive cells of the selenium type. In other respects they are in sharp disagreement and, in the light of the present record, of which in part, of course, Justice Adkins did not have the benefit, are obviously incorrect. Some of these latter incorrect findings are among those upon which Vogt apparently now relies.
I Thus, in finding No. 7, Justice Adkins stated: “At the time of plaintiffs’ invention here involved there had never been commercial use of photoelectric cells, and all known actual uses of the same had been confined to laboratory photometric measurements, i. e. measurements of light emitted by lamps and the measurement of star light. At the time of plaintiffs’ invention there was so little knowledge of photoelectric cells, and so little use therefor that they could not be purchased on the market as an article of commerce, and it was necessary for research workers desiring to experiment with these cells to have the necessary equipment to make and to painstakingly make and test out the same, or else have them made in especially well equipped laboratories.” But the record clearly reveals that the contrary was true. As a matter of fact, the photoelectric effect was discovered by Hertz in 1887,8 if not earlier.9 In 1888, Hallwacks conducted a series of fundamental experiments upon the subject; in 1889, Elster and Geitel published the first of a long series of papers on the subject; in 1890, they described what was the forerunner of the modern photoelectric cell, viz., a glass bulb containing alkali metals, which was evacuated and then sealed; in 1892, they used a photoelectric device as a sensitive element in the first photoelectric photometer, used for measuring the ultra-violet radiation from the sun; in the same year (1892) they constructed the first gas-filled cell, containing an alkali metal deposited in the presence of hydrogen.10 A cell for photometric purposes was devised as early as 1904 j11 the opening of the modern epoch in photo-electricity is attributed to Einstein in 1905, on account of discoveries for which he received the Nobel Prize.12 By 1912, prac*944tical cells were being produced.13 In the Bulletin of the Bureau of Standards, June 17, 1918,14 the photoelectric cell is described in detail. The writer therein stated that he had investigated at least a dozen photoelectric cells and the language of the article indicates that — contrary to the court’s finding — such cells could be and were purchased on the market.15
Of particular significance, as indicating the basis for the 1931 decision, are Justice Adkins’ findings numbered 8, 9 and 10. No. 8 purports to set out the essential requirements of any light responsive cell in the successful reproduction of sound; No. 9 purports to set out the respects in which selenium cells are deficient for this purpose, specifically:
“(a) The current response to light in selenium cells, as known prior to the invention here in issue, was very erratic under different conditions and circumstances, of use.
“(b) The current response to light in selenium cells involves serious and variable ‘lag’ and ‘persistence’ characteristics which greatly distort sound modulated currents.
“(c) There is no ‘linear’ or ‘straight line’ relationship between the selenium cell current and the impinging light.
“(d) Selenium cells have fatigue characteristics which distort sound modulated currents.”
Then follows finding No. 10, which reads as follows:
“10. Prior to the time of plaintiffs’ invention herein involved it had not been determined that photoelectric cells possessed any of the above stated essential characteristics for satisfying the requirements of sound reproduction. Plaintiffs by their work in 1918 and 1919 were the first to investigate and demonstrate that photoelectric cells possessed such characteristics as to meet each of the above stated requirements.”
Apparently Vogt is now willing to take full credit for everything which appears in this latter finding.
The present record reveals, however, that Vogt made no such extravagant claims in the original application, and that the characteristics of a photoelectric cell, credit for discovery of which was thus given to him, were well known to the art before he and his associates began their work.16 What Vogt claimed by way of invention was merely the substitution- of the photoelectric cell for the selenium cell in a combination which he then described. This *945is clearly revealed by the amendment of the Vogt application filed in 1925 in which the following language appears: “It is, of course, old and well known to illuminate a selenium cell with acoustically-modulated light and utilize the current thereby produced to operate a loud speaker or other sound reproducer. * * * It is. of course, true that the photoelectric cell per se was well known prior to applicants’ effective date, namely, the date of filing of the corresponding German application, March 3, 1919. So far, however, as is shown by the prior art cited by the Examiner and so far as we are familiar with the prior art, applicants are the first to see the advantage of substituting the photoelectric cell for the selenium cell in a combination in which current is to be reproduced by the light sensitive cell in correspondence with rapid changes of radiant energy, such as those corresponding to varying sound vibrations.” (Italics supplied.) Ballard, Vogt’s witness, testified as to the absence of lag in a photoelectric cell, as well as concerning its other advantages over the selenium cell. Josef Engl, one of the Vogt co-applicants, testified: “We knew at the time March, 1919, and before that, during the year 1918, that we needed for our purposes photoelectric cells. The photoelectric cells per se were a known instrument. It was known at that time how to manufacture the cells. Knowledge of that manufacturing could be gained by looking up all old publications.” (Italics supplied) Presumably, therefore, the characteristics of the cell, as described by Vogt’s witness Ballard, were known to Vogt before March, 1919; because Engl’s statement concerning the knowledge of the co-applicants was unqualified.
Again, as indicating the basis for the 1931 decision, is Justice Adkins’ finding No. 31: “The British patent to Duddell, No. 24,546 of 1902 (Defts.Ex.Y) mentions that photoelectric apparatus other than the selenium cell may be used, but such statement does not disclose that what is known as the alkali cathode photoelectric cell or any cell using equivalent phenomena could be used, * * But the art, as revealed by the present record, was well acquainted with the alkali cathode photoelectric cell long before 1919. All that Vogt had to do was to read Duddell in connection with that knowledge, which as Engl testified “could be gained by looking up all old publications.”
Specifically, in that connection, an article by Kunz and Stebbins17 described the alkali cathode photoelectric cell in minute detail, and the relative advantages to be obtained by using each of the four alkali metals: sodium, potassium, rubidium and csesium; in cells containing either helium, argon, or neon gases. Even as early as 1892, as previously indicated, Elster and Geitel had described the alkali cathode photoelectric cell. In 1907, Bergwitz described a very large photoelectric cell (about 4 inches in diameter) with a potassium cathode, which he selected for experimental purposes. This, incidentally, was published in German (Physikalische Zeitschrift) where it was peculiarly available to Vogt. There is no reason to sup*946pose that Duddell was not familiar with some of these studies; certainly there is no escape from the conclusion that by 1919 Vogt, reading Duddell on Kunz and Steb-bins and the other references, was fully informed as to the subject about which it is here proposed to give him credit for discovery. Radtke’s witnesses testified that Kunz supplied photoelectric cells with alkali cathodes to the American Development and Operating Company in 1916 and 1917, and that Radtke also procured such cells from Kunz in 1917. In its brief Radtke says that in terms of the 1919 art the photoelectric cell having an alkali cathode was old. Whitson, the third party in this suit, called as a witness Benjamin Chromy, who testified that: “Lots of them [photoelectric cells] were made in 1917” and, further, that there was little fundamental difference between photoelectric cells made in 1932, and those described in the Bureau of Standards article published in 1918. In its brief Whitson contends that he was reading and studying articles dealing with photoelectricity as early as 1906, and in 1915, was proposing to use a photoelectric cell for sound reproduction.
It is apparent, therefore, not only that Justice Adkins attributed to the photoelectric cell — as of 1919 — a mystery which in fact did not exist, at least so far as concerns the knowledge of the prior art, but that these findings have been, in effect, repudiated by each of the parties to the present appeal or by their witnesses. In this connection, as well as in our consideration of subsequent findings and contentions, it is important to bear in mind that, as Vogt points out in its brief, “the electrical art of sound communication has for many years been in the hands of trained engineers who are experts and scientists (e. g., the technical staffs of the telephone and radio companies) and they like inventors are fully chargeable with knowledge of the prior art, as the same may appear by expert testimony and by reference to prior scientific publications.” It is with knowledge of the prior art as understood by such men that we are concerned.18 “It is upon the difficulty, obscurity, or even novelty of the art concerned that depends the kind and degree of skill or knowledge which must be possessed by those who assume to apply it.”19 Paraphrasing the language of the Supreme Court in' Carnegie Steel Co. v. Cambria Iron Co.,20 it may be said that the knowledge of the prior art is not limited to the knowledge of lawyers or of the public generally, but is that of the experts and scientists, the trained engineers who have been working for many years in the electrical art of sound communication. If expert testimony and prior scientific publications reveal such knowledge, sufficient to teach such men or to suggest to them what is here claimed to be invention, then the fact that such claims may seem remarkable, or new, or profound, to lawyers or to the public generally, is not sufficient to constitute invention.
II As previously pointed out, Vogt’s theory of invention, as it appears from the amendment of 1925, was based upon the “advantage of substituting the photoelectric cell for the selenium cell in a combination * * * ” as there set forth. Whitson concedes that Haines, British Patent No. 18,-057, issued in 1906; Fritts, No. 1,203,190, issued in 1916; and Reis, No; 1,607,480, issued in 1926; disclose every element of claims 1 and 2 except that a selenium cell is utilized instead of a photoelectric cell; consequently, that nothing new is involved in these two claims except substitution of the latter cell for the former. Whitson contends, nevertheless, that making the substitution constituted invention and that he was prior to Vogt. Radtke contends that counts 1 and 2 involve merely the substitution of a photoelectric cell for a selenium cell and are met by the prior art; hence that they do not disclose invention; to the contrary that the Radtke patent “by amendment limited counts 1 and 2 to an amplifier combination (its claims 30 and 31) so now, in the exact language, the patent contains only the third and fourth counts * * *. The reason why the patent does not contain the first two counts in their exact language is because Radtke considered that they were invalid as worded and he did not permit them in his patent. New prior art not before Justice Adkins was discovered.” *947(Italics supplied.) Amicus curiae, in its brief, contends that the photoelectric cell was an obvious substitute.21
Justice Adkins found that: “35. The patents and publications (Defendant’s Exhibits A to J, inclusive, and Exhibits U to Z, inclusive) taken either singly or collectively, as of March 1919 and prior thereto, fail to teach or to suggest to one skilled in the art that the photoelectric cell was a suitable instrumentality for use in the reproduction of sound, or as a substitute for the selenium cell.” (Italics supplied.) The italicized portion of the finding, standing alone, is clearly erroneous in light of the present record. Thus, in 1916, Jacob Kunz and Joel Stebbins, in an article published in the Physical Review, stated:22 “J. G. Kemp and W. F. Schulz have shown that it is possible and advantageous to replace the selenium in stellar photometry by the photoelectric cell.” They cited, in support of this statement, articles by Kemp and Schulz which were published in 1913. Moreover, the error of the italicized portion of finding No. 35 is revealed by Justice Adkins’ own finding No. 17. There— in discounting and minimizing the disclosure of the British patent to Polyphos,23 No. 3701 of 1908 — he said: “17. * * * *948This patent suggests the substitution of the photoelectric cell for the selenium cell in picture transmission, but fails to disclose any circuit arrangements for making such a substitution possible, * * It is apparent, therefore, that the substitution of the photoelectric cell for the selenium cell for some purposes had been suggested at least as early as 1908, and that substitution had been made, actually and advantageously, at least as early as 1913, six years before the earliest filing date claimed by any of the three applicants.
It is important to note, as regards Justice Adkins’ finding No. 17, that the reason there given for concluding that the Polyphos patent was not an effective disclosure, is not because it fails to suggest the substitution of one cell for the other, but because (1) it makes no reference to sound reproduction, and (2) it fails to disclose circuit arrangements. But it is not necessary ,in order to defeat á claim of invention that a complete disclosure of the prior art should have been made in one reference.24 Mention will be made hereafter of the prior art as it concerns circuit arrangements. For the present it suffices to say that it will not do to say in one finding, generally, that the prior art did not suggest a substitution, then in another to admit that a particular reference did suggest such a substitution, but did not disclose circuits sufficient to make substitution possible for the particular purpose of the present claim. It is not necessary, in order to prove that the prior art taught or suggested the substitution of one well known element for another, that such teaching be contained in a single reference or even that substitution for the identical purpose had been proposed.25 A skilled mechanic would know that if a hammer could be used to drive a nail it could also be used to drive a tack; or that if a wheel could be used on an automobile it could also be used on a truck. That substitution of the photoelectric cell for the selenium cell had been suggested for one purpose and actually successfully accomplished for another, would be enough to put even a routineer on notice as to the probability that additional successful substitutions could also be made.26 Language of Judge Learned. Hand, in the Ruben Condenser Co. case,27 is peculiarly pertinent in its application to the present case: “But all new combinations are not patentable combinations. Especially in chemical and electrical experiments happy solutions may be reached by testing out variants reached merely by permutations of old elements. Possibly the patent law should protect such industrial achievements, but it does not; the work of the well-equipped laboratory which by trial and error checks off a series of formulas, may be the proper path of industrial advance, as the Germans found it to be before the Great War; but it does not demand what we call ‘invention.’ Something more personal to the inventor, something which better measures his imaginative powers, is required.”
Ill If we may judge from one of his findings, Justice Adkins based his award of a patent in 1931, in part at least, upon the *949theory that Vogt was the first to overcome theretofore insurmountable difficulties in amplifying photoelectric cell currents for any practical uses. This appears from his finding No. 14.28 Without further explanation it might be assumed that Vogt, by incorporating this finding into its brief, is now willing to take credit for this discovery. The present record reveals that Vogt did not make the discovery and has never claimed to have done so.
As early as 1907, Bergwitz, writing in German (Physikalische Zeitschrift), described successful experiments which he had conducted for amplifying currents from a photoelectric cathode cell by connecting it with a sensitive telephone. Langmuir disclosed amplification of photoelectric cell currents as early as 1914.29 By 1916, Wold,30 and by 1917, Kunz,31 disclosed amplification of photoelectric currents by means of the thermionic amplifier or audion; as did the Bulletin of the Bureau of Standards in 1918.32 In neither his German nor his United States application did Vogt claim invention on this ground. In the original United States application no reference was made to amplifiers. The German application of March 3, 1919, recognized that devices for amplification were known to the art; as is shown by the following language: “All that remains then to be done is to amplify, if necessary, the variations of the current in the cell with a constant auxiliary voltage being applied. Amplifier devices of this sort have been used for some time in telephone arts. * * * These stronger current variations, than those of the photoelectric cell can be amplified, if desired, as above pointed out. In this manner energy sources are attained so that an adequate sound intensity is producible even in comparatively large rooms or halls.” and by the following language: “ * * * these currents being supplied to the ear with the use of well-known amplifier and reproducing apparatus.” (Italics supplied) In an amendment of the Vogt application filed in 1925 the following language appears: “It will also be noted that the last line of Claim 4 of the German application refers to the use of well-known amplifier apparatus.” (Italics supplied) It is true that Vogt amended its United States application to refer to amplifiers, but it does not rely upon this to show invention. Instead, it expressly disclaims it.33 Ballard, Vogt’s *950witness, testified that suitable amplifiers were generally known and available before 1919.34 Radtke testified that the thermionic amplifier could be purchased in 1915 and 1916, and was available as a means of increasing photoelectric currents sufficiently to make them of practical value.35 Chromy, Whitson’s witness, gave similar testimony. Whitson contends in his brief that Kunz, in his article published in 1917, demonstrated a connection of an amplifier to a photoelectric cell, and that this article constituted a part of the prior art. It will be seen, therefore, from the present record, that all parties and their witnesses repudiated this fundamental finding of Justice Adkins.
IV As previously noted, Vogt incorporated into its brief Justice Adkins’ finding No. 15, which reads as follows: “15. Plaintiffs [Vogt] were the first to devise suitable electrical connections between photoelectric cells and amplifiers for the amplification of acoustically modulated photoelectric cell currents. The circuit described in the specification of plaintiffs’ application and shown in Fig. 3 of. the drawings thereof constitute one of the suitable circuits devised by them for connecting a photoelectric cell as used for sound reproduction to an amplifier.” (Italics supplied.) In this instance again, without explanation, the inclusion of this finding in its brief with approving reference thereto might seem to indicate that Vogt intended to take credit for discovery of circuits and other suitable electrical connections. But Vogt has never claimed more than that such suitable connections were disclosed. To the extent, therefore, that Justice Adkins relied upon supposed discovery by Vogt of suitable circuits and connections, it is obvious that his determination of invention was without basis, Radtke says that Justice Adkins was misled into making this finding. Vogt now interprets the finding as meaning merely that “prior to Vogt, no one had connected a photoelectric cell to a circuit in which distortion-free sound modulated currents would be set up and that Vogt’s disclosed circuit was suitable for connecting the cell, to an amplifier."
In Vogt’s German application of March 3, 1919, this feature, which is vested with so. much uncertainty and mystery36 in the Adkins finding, was actually described, simply and casually, in terms of the prior-art: “The bundle of light rays emanating from the luminous source is rhythmically-changed in its intensity by the film travel-ling past it, and is made to impinge upon a. photo-electric cell (alkali hydride with a rare-gas filling). All that remains then to-*951be done is to amplify, if necessary, the variations of the current in the cell with a constant auxiliary voltage being applied. Amplifier devices of this sort have been used for some time in telephone arts.” (Italics supplied.) And, again, in the claims which conclude the application, appears the language, “these currents being supplied to the ear with the use of well-known amplifier and reproducing apparatus.” 37 (Italics supplied) Neither Radtke nor Whitson contends that invention lay in the devising or discovering of suitable circuits 38 per se. Here again, therefore, for the fourth time we have a fundamental finding by Justice Adkins repudiated by all parties.
V At this point we come to a consideration of Radtke’s theory of invention : That it was he who first disclosed an amplifier combination, i. e., an operative circuit connecting the photoelectric cell with the amplifier tube. Radtke relies heavily upon the decision of the Examiner of Interferences, but its claim of invention can derive little support from that source. It must be remembered that the Examiner’s opinion was written upon the assumption that the question of invention had been settled by the 1931 decision of the District Court — adversely to the former decision of the Patent Office. The Examiner’s language, consequently, was concerned with priority and sufficiency of disclosure, not with invention. Priority and sufficiency of disclosure are matters which, in the present case at least, are secondary to that of invention. If the subject which the parties seek to disclose is not an invention, it makes no difference whether it is well or poorly described, or whose disclosure came first. If, in this year 1941, two contesting applicants should claim to have invented the wheelbarrow, the fact that one described it in the most elaborate and comprehensive manner, while the other failed completely in his effort to describe it, would not entitle the former to receive a patent.39 Radtke’s argument— based upon the premise that it was inventive to disclose an amplifier combination— is carried to considerable length, also, in an effort to prove that Vogt failed to disclose this feature. Thus, the argument is made that Vogt did not teach how to use the condenser and the high frequency oscillations; and that the Patent Office found the Vogt disclosure deficient because it did not teach that the size of the condenser should be such as to accomplish tuning of the photoelectric circuit to the high frequency wave. But Radtke admits that “both condensers and high frequency oscillations were well known in 1919, as all parties agree.” Radtke argues that when on the trial Tri-Ergon demonstrated what it considered was the Vogt disclosure, “a bypass condenser was placed across the telephones, which bypassed the ‘high’ frequency currents and without which there would have been no ‘action in the circuit’, * * *. The Vogt application does not mention telephones nor, of course, any bypass condenser.” Radtke follows this statement immediately by an admission: “Vogt says that in telephone circuits and in radio circuits it was not uncommon practice to use bypass condensers, wMch is true.” (Italics supplied.) But, Radtke continues, “it does not follow that a bypass condenser was also known to be useful in a photo-electric system containing the sound controlled electron emissions, ‘high’ frequency currents, and probably some *952mixed (beat) currents resulting from their interaction.” (Italics supplied.) Radtke argues further that in the interference proceedings Vogt constructed apparatus which contained a detector; that the Patent Office found that there was no justification for the use of a detector; that there is in fact no necessity for detectors in photoelectric cell circuits; and that there is no detector in the system of the Radtke application, which is in almost universal use in motion picture .heatres and studios. Again it is argued that Vogt failed to teach how to connect amplifiers and sound reproducers. Radtke argues: (1) “There is no doubt that audio amplifier tube circuits were well known before Vogt and that the art knew how to connect them in telephone and radio circuits;” (2) but Vogt is describing a photoelectric circuit. which, he says, required inventive genius to connect to an audio amplifier; (3) Justice Adkins found for Vogt on this point; (4) consequently, the failure to teach how to make the connection is fatal.
Vogt replies that no difficulty was involved in connecting (coupling or attaching) the disclosed circuit to known amplifiers, in light of the prior art; and that Radtke added nothing of inventive .significance by disclosing his amplifier combination. William C. Ballard, Professor of Electrical Engineering at Cornell University, called as a witness for Vogt, and who qualified as an expert upon a record of years of academic training and professional experience, testified that amplifiers and circuits of the kind involved in the present case were generally known to thos,e skilled in the art prior to 1919 ;40 that the *953meaning of an oscillating circuit was well known before 1919, as was understanding of how to feed undamped high-frequency-currents into such an oscillating circuit; that “there is not the slightest difficulty in arranging such a circuit from knowledge of the prior art, and to set such a circuit into operation;” that “Once having transformed the light impulses into variations of modulated high frequency current, as shown on Plaintiff’s Suit Exhibit 11, the rest of the problem was easy, because we had numerous cases where amplifying tubes and circuits and loud speakers were available. You could go out and buy them in the open market, complete, without any experimentation to be done, and put them next to the coil that had this modulated high frequency current and convert it back into sound again.” (Italics supplied); that the adjustment or tuning of the circuit was described by DeForest in 1912, that it was well within the knowledge of one skilled in the art as of 1919 and earlier; and that he agreed heartily with the statements of Justice Cardozo, made in the case of Radio Corp. of America v. Radio Engineering Labs., Inc.,41 concerning DeForest’s 1912 disclosure, which were read to him in the form set out in the margin.42 Dr. Henry Crew, Emeritus Professor of Physics at Northwestern University, called as a witness by Radtke, on cross-examination testified that during the first World War amplifiers for electrical currents were in use in radio circuits and in telephone circuits; that DeForest demonstrated their use long before the War, and that many ways of connecting up circuits for that purpose were known to those skilled in the art.43 On this issue Whitson joins Vogt in contending that no invention lay in the circuits or in the connecting of them.44 Chromy, Whitson’s witness, testifying concerning the 1917 article by Kunz,45 stated that the article descibed a circuit properly connected and further: “Assuming, of course, that we have the sound track and the reproducing device, such as head phones or a loud speaker, the sound track *954could be reproduced through this circuit and the response or reproduction, I believe, would be faithful and understandable.” Especially damaging to Radtke’s theory of invention and claim of disclosure is that portion of Justice Goldsborough’s decision which reads as follows: “The patent application of March 27, 1922, undoubtedly disclosed the counts at issue, but it is to be noted that on the 26th day of May, 1921, when Radtke sent his application to his attorney in Washington, the so-called British abridgments, which contained a description of the circuit described by Radtke in his application, had been in a scientific library in Chicago, frequented by Radtke, since May 17.”
In this connection, Vogt relies' particularly upon findings numbered 13, 14 and 15, made by the District Court in the present case, which read as follows:
“13. The testimony herein of plaintiffs’ expert Prof. Ballard and the accompanying plaintiffs’ exhibits showing the state of the prior art, established with thorough conviction that the disclosure of the Vogt, et al. original U. S. application in issue was sufficient to enable those skilled in the art or science to which it relates to make and use the apparatus defined in the four claims in issue.
“14. The criticisms made from time to time in the Patent Office interference proceedings, of the Vogt, et al. U. S. application, all relate to things which were theretofore well known and understood by those skilled in the art or science to which the application relates, and the use of which in conjunction with the disclosure of said application was obvious to those skilled in said art o'r science. .
“15. Defendants herein offered no proofs tending substantially to support the criticisms of the disclosure of the Vogt, et al. U. S. application which had been previously made in the Patent Office interference proceedings, or tending substantially to refute the plaintiffs’ evidence herein establishing the sufficiency of such disclosure.”
In our view these findings are supported by the evidence, and prove that Radtke invented nothing. However, we do not reach the conclusion therefrom that Vogt invented anything either. Carefully read, it will be seen that these three findings could have been equally well made concerning a textbook description of any well known device. As already pointed out, the fact that a claimant sufficiently discloses something so that those skilled in the art could make and use the apparatus, is far from enough to show that he invented the apparatus.
As to every factor of supposed novelty, therefore, and as to every claim of means, combination, and connection, upon which invention depends, one or the other of the three applicants not only concedes but contends that it was taught by the prior art. We are satisfied that, individually and collectively, they have successfully challenged every contention and disproved every claim. The fact that all three present parties were working, at remote points of the compass, toward the same objective; and that numerous others for many years had been experimenting and writing toward that same objective, is suggestive in itself, that what is here involved is not invention46 but, rather, the natural step by step development of the art;47 until at the end the opportunity for invention had passed,48 and when, later, there became a generally recognized dem*955and49 for a film record type of sound reproduction, no more was required than to take the final, obvious step50 of substituting the photoelectric cell, with its well known advantages, for the selenium cell with its well known inadequacies.
This being true, it is not necessary to decide the question of priority so ably presented on this appeal. We may assume, hypothetically, that either the Patent Office or the District Court correctly decided the question of priority, as between such disclosures-^ may have been made by each of the parties.51 But as none of those discoveries constituted invention, the question of priority is immaterial. Each of the parties, it is true, apparently assumed that the issue of invention had been settled and that only the question of priority would be considered. It was, perhaps, because of that assumption that each — relying upon some phase of the history of sound reproduction which seemed most favorable to it — at the same time joined so wholeheartedly with the others in demonstrating that none invented anything. As previously indicated, this court cannot be guided by the assumption upon which the parties seem to have relied. Instead, it is required to determine the question of invention from the record.
That record convincingly establishes that (1) the District Court incorrectly decided, in 1931, the issue of invention then presented to it; (2) the lower court in the present case incorrectly relied upon its earlier decision; (3) as the Patent Office correctly determined in 1927, none of the four claims in suit described an invention; and (4) no one of the parties is entitled to enjoy the monopoly which it desires.
The motion of the Commissioner of Patents, that the appeal be dismissed as to him, is denied;52 the decree of the District Court is reversed, and the cause is remanded to the District Court with instructions to dismiss the complaint.
Reversed.
I think the decision in this appeal should be upon the question of the correctness of the decree below on the issue of priority and the ancillary issue of disclosure, and that the court should not pass upon the question of patentability.
Under the pleadings the only issues before the trial court were the one of priority as between Whitson, Radtke and Vogt (American Tri-Ergon Corporation), and the further ancillary issue as to whether Vogt’s application disclosed what was defined in the claims. It is true that there are statements in' the conclusions of law and decree which, if read without reference to the state of the pleadings, seem to indicate that the trial court did pass upon the question of patentability of the apparatus described in the claims. But, in view of the absence of any issue of pat-entability under the pleadings, those statements must be disregarded, as surplusage. Not only was there no issue raised below by either Whitson, Radtke or Vogt as to patentability, but also the Commissioner of Patents made no contention against pat-entability. On the contrary, he sought to be dismissed from the case, upon the theory that he was neither an “opposing party” within Section 4915 of the Revised Statutes, nor an adverse party. He relied upon Eno v. Coe, 1939, 70 App.D.C. 337, 106 F. 2d 858, and Coe v. Hobart Mfg. Co., 1939, 70 App.D.C. 2, 102 F.2d 270. The Commissioner did not appeal from the decision of Judge Adkins on the question of patentability of the same claims in the case of American Tri-Ergon Corp. v. Robertson, and in the instant case the Commissioner was clearly acquiescing in that decision.
*956It is true that in Hill v. Wooster, 1890, 132 U.S. 693, 10 S.Ct. 228, 33 L.Ed. 502, in a suit brought under Section 4915 to determine a question of priority, the question of patentability of the claims involved was passed upon judicially for the first time on appeal. But in that case the Commissioner of Patents was not a party in the trial court and the public interest, therefore, had been without representation in a judicial proceeding. The theory of the decision in Hill v. Wooster was, I think, that under such circumstances the appellate court had, and should exercise, the power to protect the public against a monopoly in respect of palpably non-patentable claims. But a very different situation exists in respect of the claims in the instant case. For the public interest has twice had representation in judicial proceedings on the question of patentability of these claims — once when in American Tri-Ergon Corp. v. Robertson the Commissioner contested their patenta-bility, and again in the instant case when, although the Commissioner was made a party below, he sought to be dismissed from the case, thus choosing, as the representative of the public interest, not to raise any question as to patentability. I think that the decision of the majority constitutes, under the circumstances, an unwarranted collateral attack upon the decision of the trial court in American TriErgon Corp. v. Robertson, which stands as a valid and unappealed decision of a competent court in this jurisdiction to the effect that the claims are patentable. And in the instant case, where, unlike Hill v. Wooster, the question of patentability is at least debatable, and where the improvement involved has gone into wide commercial use in an important industry, the question of patentability ought not, in my opinion, in fairness to the parties, be decided for the first time on appeal.
I am of the further view that the present decision of the majority on the question of patentability can not properly consider, as it does, scientific references which áre not in the record, and which are not referred to in the briefs.1 A court can, of course, in order to obtain an understanding of the factual aspects of a patent case, take judicial notice of established scientific facts, and it may, in order to inform itself concerning them, consult standard scientific works; but it can not take judicial notice of a disputed fact. The question of patentability discussed in the majority opinion is whether or not it was obvious to substitute for a selenium cell in a sound reproducing apparatus a photoelectric cell of alkali cathode type in conjunction with circuit connections and an amplifier. The determination of that issue is dependent upon the answer to a question in respect of the state of the art so far as concerned knowledge of the properties, uses and advantages of the photoelectric cell of the alkali cathode type and of the selenium cell. More specifically, that question is whether the properties, uses and advantages of a photoelectric cell of the alkali cathode type were so well known in the art at the time that Whitson, Vogt and Radtke were severally working upon the substitution which they now assert is patentable, that it can not be said to have required inventive genius to make the substitution for selenium. The scientific references to which I have referred are cited by the majority as relevant to that question of fact. I think an appellate court can not properly reach an answer to such a question upon the basis of its own examination of data not in the record, and which may be disputable, without first affording opportunity to the parties to be heard in respect of it and an opportunity to supply additional relevant data. The rule in administrative proceedings that “Nothing can be treated as evidence which is not introduced as such” (Brandeis, J., in United States v. Abilene & So. Ry. Co., 1924, 265 U.S. 274, 288, 44 S.Ct. 565, 569, 68 L.Ed. 1016), applies even more strictly in judicial proceedings.