Full opinion text
OPINION McLEAN, District Judge. This is a patent suit which was begun on March 5, 1965. The complaint, which has been several times amended, in its present form contains three counts. In the first, plaintiff SCM Corporation (SCM) seeks a judgment declaring that each of three United States patents owned by defendant Radio Corporation of America (RCA), Patents Nos. 3,052,-539 (539), 3,052,540 (540), and 2,922,-883 (883) is invalid on various grounds and that SCM has not infringed them. In the second, SCM seeks recovery of the royalties which it paid to RCA during the time that SCM was a licensee under the 539 patent. SCM claims that RCA induced it to enter into the license agreement by fraud. In the third count, SCM charges that RCA procured the 539 patent by practicing fraud upon the Patent Office and that consequently, in asserting rights under that patent, RCA has violated Section 2 of the Sherman Act and has caused damage to SCM. RCA has counterclaimed for a judgment determining that each of these patents is valid and that SCM has infringed each of them. RCA seeks an injunction against infringement and an accounting. A second counterclaim, in which RCA charged SCM with violating the antitrust laws in a variety of ways, was dismissed by this court on SCM’s motion, SCM Corporation v. Radio Corporation of America, 276 F.Supp. 373 (S.D.N.Y. 1967). The order of dismissal was affirmed by the Court of Appeals, SCM Corporation v. Radio Corporation of America, 407 F.2d 166 (2d Cir. 1969), cert. denied, 395 U.S. 943, 89 S.Ct. 2014, 23 L.Ed.2d 461 (1969). This opinion will first consider SCM’s first count and RCA’s counterclaim, i. e., the issues of validity and infringement of the three patents. It will deal with each patent separately. It will first treat the 539 patent, the most important of the three, to which the bulk of the evidence was directed. The 539 Patent This patent is entitled “Electrostatic Printing.” It was issued on September 4, 1962 to RCA as the assignee of the applicant, Harold G. Greig, a scientist employed by RCA. The original application for the patent was filed on September 29, 1951. This application was later abandoned and a continuation-in-part application was filed on October 1, 1953. In the approximately nine years which intervened between the filing of the continuation-in-part and the issuance of the patent, there were many proceedings in the Patent Office, culminating in an action instituted by RCA against the Commissioner of Patents in the United States District Court for the District of Columbia. Some of these proceedings have an important bearing upon the issues to be decided here. The patent relates to “office copying,” i. e., the reproduction of documents. It contains twenty-two claims. Fifteen of these are process claims, i. e., claims devoted to the process or method of making copies of the documents to be reproduced. The remaining seven are product or “paper” claims, i. e., claims devoted to the “recording element” on which the copies are formed. This element consists of a specially coated paper. RCA’s trade name for the process is Electrofax. On June 17, 1963 (effective as of March 1, 1963) RCA granted to SCM a license under claims 1-13, 20 and 21 of the patent. These are all process claims. The agreement licensed SCM to make, use and sell “Eleetrofax Office Copiers in which, or in the operation of which, such Licensed Patent is used.” The license agreement afforded SCM the right to terminate it after one year upon 30 days’ notice. On March 4, 1965, SCM availed itself of this privilege and terminated the license. It began this action on the following day. SCM contends that the 539 patent is invalid under 35 U.S.C. § 102, § 103 and § 112, that is to say, that it is anticipated by prior art, obvious to one having ordinary skill in the art, and inadequately described in the specification. SCM asserts that each of the 22 claims of the patent is invalid under each of these statutory provisions. Furthermore, SCM contends that each claim is invalid and unenforceable because RCA procured the patent by “fraudulent conduct in the United States Patent Office and the United States District Court for the District of Columbia.” Finally, SCM contends that each claim is unenforceable because RCA failed to grant a royalty-free license under the patent to the United States Government and because RCA has misused the patent by separately licensing and charging royalties to the manufacturers of copy machines and the manufacturers of copy paper. The charges of fraud and the contention of invalidity under Section 112 because of inadequate description of the invention are closely related. More testimony was devoted to them than to any other subject. Understanding of the case will be enhanced by considering these contentions first. To that end it is necessary to set forth a chronological summary of the facts established by the evidence. Events Prior to the Filing of the Original Application on September 29,1951 The story begins in the early 1940s. On October 6, 1942, United States Patent No. 2,297,691 was issued to Chester F. Carlson on “Electrophotography.” This was the basic patent on a process for copying documents which subsequently became well known as “Xerography” or the “Xerox process.” The Haloid Company acquired the rights to this invention. In 1948 it employed Battelle Memorial Institute, a research organization in Columbus, Ohio, to do further research on it. Over a period of two years, from June 1948 to June 1950, Battelle engaged in extensive experimental and development work on this subject. It rendered quarterly reports to Haloid upon its activities and conclusions. RCA maintained research laboratories in Princeton, New Jersey, which were staffed with a sizable number of scientists and engineers working in a variety of fields. One section was known as the Acoustical and Electromechanical Research Laboratory. For many years, from 1942 to 1965, its head was Charles J. Young. Harold G. Greig, a chemist, was one of his group. Young heard about Xerox in 1948. It interested him at once. One of the earliest documents in this case consists of some handwritten notes by Young made on February 4, 1949 and “Witnessed and understood” by Greig on February 7, 1949, in which Young jotted down questions “regarding extent of patent coverage obtained by Carlson, et al. on Xerography.” Young visited Battelle and Haloid in October 1950. He was shown the work going on there in Xerography. His interest increased. Greig, presumably at Young’s direction, made copies and abstracts from time to time of portions of Battelle’s quarterly reports to Haloid to which he was granted access. On September 25,1951, RCA took a license from Haloid under the patents relating to Xerography, including the five Carlson patents and four others issued to other persons. It is apparent that Young and Greig were eager to devise a document copying process which, without infringing Carlson’s patent, would afford RCA an opportunity to compete with Xerography in what promised to be a lucrative field. To comprehend their problem and to appreciate what they achieved, it is necessary to understand the Xerox process and the respects in which Greig’s invention differed from it. This is a highly technical subject, one upon which the parties have expended many hundreds of pages of expert testimony in an effort to inform the court. What follows is a simplified version which will suffice for present purposes. Both processes make use of the phenomenon of photoconductivity, the effect of light upon discharging, i. e., conducting away, an electrical charge. Reduced to its simplest terms, an “element,” i. e., a metal plate or a piece of paper, is covered with a coating or layer of a photosensitive material and given an electrostatic charge. Light is then shone through the object to be copied if that object is transparent (i. e., a transparency), or reflected from it if it is opaque (i. e., a sheet of paper), on to the photosensitive layer. The light strikes only those portions of the layer which correspond to the blank or white portions of the object to be copied. The dark portions of the object do not transmit or reflect the light. The light discharges the electrostatic charge from those portions of the layer which it strikes, leaving the charge undisturbed on the rest of the layer. The result is a “latent image” on the layer, consisting of electrically charged particles which correspond exactly to the dark portions of the object to be copied. If the object is a document containing printed words, for example, the latent image corresponds to those words, that is to say, it is a copy of them. This latent image is first developed by treating it with a powder, then “fixed” by the use of heat, thereby making a permanent copy of the original document. In the Xerox process as developed at Battelle, the plate was made of metal. It was covered with a coating of selenium. Once the image was formed on this metal plate, it was necessary to transfer it to a piece of paper in order to obtain the desired copy. From the beginning, Young and Greig were seeking a way to form the copy directly upon a piece of paper, thereby eliminating the transfer step. To do so would not only simplify the process, but would have certain other advantages. It would eliminate the need to clean the plate each time of one print before another print could be made. It would also avoid problems of “fatigue,” i. e., wearing out of the plate. Greig hit upon the idea of coating paper with a layer of zinc oxide enclosed or suspended in an “electrically insulating, film-forming vehicle,” usually referred to as a “resin binder.” He believed that the photoconductive properties of zinc oxide would cause an electrostatically charged paper coated with the zinc oxide-resin binder mixture to respond to exposure to a pattern of light in the same manner as the selenium coated metal plate used in the Xerox process. In February 1951, Greig began experimenting with “print tests” of various zinc oxides mixed with various binders. He prepared different combinations, coated them on paper, applied the electrostatic charge and the light, developed the latent image and “fixed” it. Sometimes the result was a good clear copy of the original document; sometimes it was not. He concluded that at least some zinc oxides could be used in this fashion to make a satisfactory print directly on paper, thereby eliminating the cumbersome transfer step of the Xerox process. He determined which zinc oxides would work and which would not by the simple expedient of trying them. Those which made good prints were satisfactory. Those which did not were not. It was Young’s practice to write monthly “progress reposes” of current developments in the research work of his group. These were distributed to other members of RCA’s research staff. The progress report for February 1951 contains the first reference to Greig’s experiments. In it Young stated that “a new photosensitive surface has been found which shows promise in electrostatic printing.” He explained it briefly and concluded by saying, “Work on this new photosensitive surface will be actively continued.” Greig’s researches finally led to conclusions definite enough to enable him to file with RCA’s patent department on April 2, 1951 a formal “disclosure” of his invention. Under RCA’s procedures, such a “disclosure” by the inventor was the first step in the long process of obtaining a patent. Upon the receipt of a disclosure, RCA’s staff of patent lawyers would go to work preparing an application' for a patent to be filed with the Patent Office. In his disclosure Greig stated that the purpose of his invention was “To provide a photosensitive paper suitable for electrostatic printing * * *.” He described the coating of the paper as “a suspension of pure zinc oxide powder in a solution of an insulating binder.” He stated that the coating materials were readily available, relatively cheap, and nontoxic. He said that the requirements for the zinc oxide “appear to be that the material should be chemically pure, or better, in quality and a finely divided powder in form.” He stated that five different zinc oxides had been shown by print tests to possess the desired photosensitive properties, whereas one zinc oxide had been found unsuitable. After pointing out the advantages of his method over the selenium coated metal plate, he concluded with the statement that his electrophotographic paper “may open up an entirely new line of attack on the continuous printing by photoelectrostatic methods.” Greig continued with his print tests on a number of zinc oxides. Two of these figure prominently in this case. One was a zinc oxide made by New Jersey Zinc Company called Florence Green Seal 8 (FGS 8) which made a good print. The other was one made by American Zinc Company called AZO 33 which gave poor results. The Original Application Filed on September 29, 1951 Finally the patent attorneys were ready to file the patent application. Greig signed the application as the inventor. He designated an RCA patent attorney and an RCA patent agent as his agents to prosecute the application, to amend it and to transact all business in the Patent Office connected with it. RCA’s interest in the application was thus fully disclosed from the outset. The application stated in its first sentence that the invention “relates to an improved process of forming images on a carrier base.” It went on to describe the process at some length, and to point out its advantages over previous processes, by which it presumably meant Xerox, although it did not refer to that process by name. The application stated that the “carrier base” is given “a coating of a composition comprising zinc oxide powder of relatively pure grade suspended in a vehicle composed of an electrically insulating, film-forming substance.” At a later point it elaborated upon the phrase “relatively pure grade” to the extent of saying: “It has been found, unexpectedly, that relatively pure zinc oxide, such as that designated C.P. or U.S.P. gives much better results in the present process than ordinary technical grade zinc oxide commonly used heretofore in pigments or in coated papers. Although the reason is not understood, zinc oxide having a very low concentration of impurities, when combined with the type of vehicle described herein, is capable of taking and holding an electrostatic charge in the dark and of being discharged in the presence of light.” Chemical analyses of three satisfactory zinc oxides were set forth. The application let it go at that. No test other than purity was suggested for distinguishing good zinc oxides from bad ones. The application concluded with thirty-three claims, most of which pertained to the process. A few pertained to the product, described as an “image carrier.” On March 14, 1952, the Patent Office Examiner rejected all the claims. The Office Action cited the Carlson 691 patent and rejected certain claims on the strength of that patent, together with others. On August 29, 1952, RCA filed its response to this action. It withdrew three claims, made slight modifications in another, and argued at length that Greig’s invention was not anticipated by Carlson’s because (1) Carlson did not mention zinc oxide, and (2) Carlson’s process required a metal backing plate and the transfer of the image formed on that plate to the ultimate paper. On April 30, 1953, the Examiner again rejected all but three of the claims. RCA did not pursue the argument further at that time. It eventually decided to abandon the application and to file a continuation-in-part, which it did on October 1, 1953. The 1953 Photoconductivity Measurements On February 4, 1953, Young made another of his longhand memoranda under the heading, “Investigations Needed.” One of the items he jotted down without elaboration read “Electrostatic Orientation of ZnO in Paper Coating.” This may have been the genesis of the experimentation which culminated in the photoconductivity measurements made in the summer of 1953 to which so much testimony and argument have been devoted in this case. RCA’s patent attorneys believed that it would strengthen the patent application if a criterion could be devised for distinguishing those zinc oxides which were satisfactory for use in Greig’s process from those which were not. Merely specifying that the zinc oxide must be pure was not sufficiently precise. Although, as a practical matter, one could always tell a good zinc oxide from a bad one by making a print with it, as Greig had been doing for more than two years, this trial and error method would not do for patent specification purposes. Consequently, the patent attorneys consulted another group of RCA scientists at Princeton, the Solid State Physics Section. Henry B. DeVore, a member of that section, was assigned to help. He undertook to measure the photo-conductivity of zinc oxides by a rather elaborate testing apparatus which was apparently his own creation. The nature of DeVore’s tests and the apparatus he used in making them, another very technical subject, is not in dispute. It may be briefly summarized in a simplified fashion as follows. DeVore compressed dry zinc oxide into pellets. He attached two electrodes to each pellet. One electrode was connected to a source of electric power and the other to a resistor. He shone light upon the pellet. The light affected the current flowing across the pellet, which in turn affected the voltage across the resistor. DeVore measured the change in voltage. He then computed by a complicated formula the “surface photoconductivity” of the pellet, i. e., the difference between the conductivity of the surface of the pellet in the light and its conductivity in the dark. The fact of cardinal importance here is that the light which DeVore used was “chopped,” i. e., it was interrupted, not continuous or steady. He accomplished this by shining the light through a rotating slotted disc which interrupted the light 23.5 times every second. This produced alternating flashes of light and darkness upon the surface of the pellet. In making these tests, DeVore subjected the pellets to different wavelengths of light throughout the spectrum, from 6,000 A to 3,600 A. He found that their sensitivity to light was highest at a wavelength of about 3,900 A, which is ultraviolet light. The result of the measurements was expressed in a number. This number, at first blush, is formidable, to say the least. The number most frequently mentioned in this case is “10 — 9 ohm — /square/ watt/cm2.” Fortunately, the most important part of this expression, for our purposes, is 10 — 9, which is merely an exponential way of expressing a tiny fraction which could be expressed with equal accuracy as .000000001, or 1 billionth of something. In this case, the “something” is the “ohm —1.” This is a unit of electrical conductivity. The rest of the expression quoted above, i. e., “/square/watt/cm.2” refers to various other elements in the computation which remain constant throughout. Thus, “watt/cm.2” refers to the intensity of the light, and “square” to the ratio between the length of the electrodes and the width of the space between them. DeVore made up pellets of twelve different zinc oxides. He measured the surface photoconductivity of each and set forth the results in fractions of a negative ohm per square per watt per square cm. (“ohm — Vsquare/watt/cm.2”). The results differed with the different zinc oxides. Thus, the numerical value for one was 10 — 7, for another 10 — 10, etc. At the same time, Greig made print tests with these zinc oxides. As usual, he could tell by looking at the prints which zinc oxides were satisfactory and which were not. By comparing the results of the print tests with the results of his electrical measurements, DeVore was able to establish a correlation between them. He discovered that those zinc oxides for which he had found a numerical value higher than 10- made good prints and those for which his numerical value was lower than 10-9 made bad ones. He therefore concluded that 10-9 was the criterion which separated the good from the bad. The Continuation-in-Part Filed on October 1, 1953 Armed with DeVore’s measurements, RCA filed its continuation-in-part application. This differed in material respects from the original application. The introductory portion stated that “the photoeonductive substance should have a photoeonductive surface conductivity higher than a certain minimum.” A few pages later it set forth a method of making photoconductivity measurements. This passage described in substance what DeVore had done, but it did not say that the light used in the measurements was interrupted, or “chopped.” The description ended with the following statement: “It has been found that, to be useful in the present invention, the substance selected should have a surface conductivity of at least about 10-9 ohm-1/ square/watt/cm.2 when exposed to some wavelength within the range of about 3800-7000 A. “Having established the threshold value of photoconductivity needed in the process, it is possible to test any photoeonductive substance otherwise suitable from the standpoint of stability, compatibility, dark resistivity, color, etc., in order to determine whether it can be used.” The application went on to say that a photoeonductive substance found most suitable was zinc oxide, which could be either white or pink. The application stated: “Of the various white zinc oxides commercially available not all have a surface photoconductivity greater than about 10-9/ohm-1/square/watt/cm.2 at a wavelength of about 3900 A which is about the wavelength of peak photo-sensitivity for this material, but many pure grades of ZnO made by a dry process have been found suitable.” The application set forth twenty-seven claims, of which fifteen were process claims, nine were product claims relating to the recording element, and three related to apparatus for electrostatic printing. The nine claims relating to the recording element specifically referred to the “about 10-9” criterion. The others did not. This application fared no better than its predecessor. On March 29, 1954, the Examiner rejected all the claims. This produced an amendment filed by RCA on September 7, 1954, which cancelled three claims, amended seven others, and argued the merits of the claims at some length. A further amendment followed on August 8, 1955. This added four claims, three of which contained the “about 10-9” language. The Patent Office was adamant. On December 2,1955, the Examiner rejected all claims. The Examiner added additional reasons for his rejection by a supplemental letter on March 7, 1956. RCA’s response to this action was an amendment filed on September 4, 1956, supported by various affidavits. These constitute the principal basis of SCM’s charge of fraud. Before discussing the amendment and the affidavits, we must first consider certain activities of the RCA research staff which preceded the filing of this amendment. RCA’s Measurements and Tests in 1955 and 1956 In May 1955 DeVore had occasion to reconsider his 1953 photoconductivity measurements. In that month he measured, apparently on the same apparatus which he had used in 1953, using chopped light, pellets made of two new zinc oxides which he had not measured in 1953, New Jersey Zinc Company’s C.E. 8099-5-3 and C.E. 8099-3. He also measured new pellets of FGS 8, other pellets of which he had measured in 1953. By this time this zinc oxide, because of its particularly favorable photoconductive properties, had been adopted by RCA as its standard zinc oxide for use in its work on Electrofax coatings. DeVore found that his results on FGS 8 in May 1955 differed from the results which he had obtained on that zinc oxide in 1953. This led him to observe in his notebook under date of May 25, 1955: “All curves of surface conductivity in N.B. P-3129 [DeVore’s 1953 notebook] should be raised one order of magnitude.” Translated into layman’s language, De-Vore was saying in this entry that his 1953 photoconductivity values were wrong and that they should have been approximately ten times higher. He based this conclusion upon the fact that his measurement of FGS 8 in 1955 yielded a photoconductivity value of 3.45 x 10~7, whereas in 1953 he had found it to be 3.5 x 10-8, a difference of almost exactly ten times. In another notebook entry made on the same day, May 25, 1955, DeVore said, apropos of the measurements which he had just completed: “[T]he peak surface photoconductivity turned out to be just about 1 order of magnitude higher than the value computed two years ago. The improved energy measurement accounts for most of the difference, and sample to sample variations take care of the rest.” The reference to “improved energy measurement” relates to the calibration of the equipment, which DeVore concluded had not been precisely accurate in 1953. The phrase “sample to sample variations,” apparently refers to the fact that both in 1953 and 1955 DeVore measured several samples or pellets of each zinc oxide. The results were not exactly the same for each sample. Although the reference to “sample to sample variations” may be thought to qualify to some extent the flat statement that the 1953 measurements should be raised by one order of magnitude, nevertheless the net result of the two entries, made on the same day, leaves no room for doubt that in 1955 DeVore believed that his measurements in 1953 had been ten times too low. He accounted for the error on the theory that his equipment had not been properly adjusted in 1953. DeVore’s notebook indicates that he discussed his 1955 results with Greig and also with E. C. Giaimo, another member of Young’s group. Nevertheless, RCA seems to have done nothing about them at the moment. In the summer of 1956, however, which was after the patent application had been again rejected by the Patent Office, a flurry of activity took place. James A. Amick, a Ph.D. in physical chemistry, was employed in RCA’s Materials Research Laboratory. RCA assigned him, Giaimo, and a third scientist, Spicer, to review DeVore’s work and to make additional photoconductivity measurements of their own. On June 14, 1956, Young wrote a memorandum of his conference with Amick, Giaimo and Hill. The latter was an RCA patent agent who up to that date had handled the patent application. The memorandum stated: “To strengthen our position on the basic Electrofax paper application the tests of photoconductivity made originally by DeVore have to be repeated, the measurement procedure established and absolute numbers obtained. Hill will then prepare a revised application.” Amick first checked DeVore’s equipment as it had been recalibrated by De-Vore in 1955. He found it to be now correctly adjusted. On June 6, 1956, Amick checked DeVore’s calculations upon the basis of DeVore’s data. He found that DeVore’s computations were correct. It followed that the error in the measurements in 1953, if there was one, must have been due to faulty data rather than to incorrect arithmetic. Shortly thereafter, Amick, Spicer and Giaimo made new photoconductivity measurements on two zinc oxides, the familiar FGS 8 and AZO 33. They used DeVore's equipment and employed chopped light. The photoconductivity value which they obtained for FGS 8 was 9 x 10~7. The value for AZO 33 was 5.8 x 10-9. These results corresponded roughly with DeVore’s 1955 results on FGS 8, although the result obtained by Amiek’s group on this zinc oxide was even higher than DeVore’s result in 1955. This confirmed DeVore’s opinion that his 1953 measurements were at least ten times too low. Giaimo also employed a different set of apparatus to make measurements on FGS 8 and AZO 33 by the use of light that was continuous, i. e., steady, rather than chopped. This produced photo-conductivity values for FGS 8 and AZO 33 approximately 100 times higher than the values obtained by chopped light in 1956. Young noted these developments in his progress reports for June and July 1956. The June report revealed that the 1956 measurements by chopped light yielded higher values than DeVore had obtained by that method in 1953. Young suggested as a possible explanation that “photo-conductors made under identical conditions are known to vary by an order of magnitude or so from batch to batch.” The July report noted that the use of steady light yielded results 100 times higher than those obtained by chopped light in 1956. At this point an RCA patent attorney named Greenspan succeeded Hill as the lawyer in charge of the patent application. Greenspan prepared four affidavits, two of which were signed by Amick, one by Greig and one by DeVore. He filed these in the Patent Office in support of an amended application filed on September 4,1956. The DeVore affidavit is the most important for our purposes. In it DeVore described the apparatus and methods he had employed in making his photoconductivity measurements. He clearly stated that he had used chopped light. He then set forth in tabular form the “values of maximum surface photoconductivity” which he had obtained for each zinc oxide. He did not say when he had obtained them. The table listed fourteen zinc oxides. Twelve of them, including FGS 8, were the ones which DeVore had measured in 1953. The other two were the ones which DeVore had measured for the first time in 1955, New Jersey Zinc C.E. 8099-5-3 and C.E. 8099-3. The affidavit did not state that De-Vore had measured FGS 8 again in 1955. It did not include the photoconductivity value for FGS 8 which DeVore had obtained in 1955. The values set forth in the table for all the zinc oxides, except the two New Jersey Zinc Company ones which DeVore had first measured in 1955, were the values which DeVore had obtained in 1953. They were not corrected by increasing them by one order of magnitude. Greig’s affidavit described his method of making print tests. He attached copies of prints which he had made with a number of zinc oxides, including the fourteen which DeVore had measured. Amick’s principal affidavit, as far as pertinent for present purposes, recapitulated the data in DeVore's affidavit and in Greig’s. It set forth in parallel columns for each of the fourteen zinc oxides the photoconductivity value reported in DeVore's affidavit and the result of the print test stated in Greig’s affidavit. It then concluded that those zinc oxides which DeVore had found to possess a photoconductivity value of greater than 10~9 made good prints, and conversely, those with a photoconductivity value of less than 10-9 made poor prints. Three observations may be made about this affidavit. 1. The data contained in the DeVore affidavit would better have supported a conclusion that the dividing line between good and bad zinc oxides was 10~8 rather than 10-9. As far as the zinc oxides which produced good prints were concerned, 10~8 and 10~9 were equally correct criteria, for all the zinc oxides which produced good prints had a photoconductivity value, according to DeVore, of higher than 10-8, and therefore necessarily higher than 10-9. The converse, however, was not entirely true. There was one zinc oxide, New Jersey Special No. 3, which produced a poor print but was found by DeVore to have a photo-conductivity value higher than 10~9, but lower than 10~8. Hence the fact that this zinc oxide gave an unsatisfactory print would be consistent with a 10-8 criterion but not with a 10~9. 2. The photoconductivity value given for AZO 33 was that which DeVore had found in 1953, 1.5 x 10-11. This was consistent with the results of Greig’s print test, which showed that the print on this zinc oxide was poor. In 1956, however, Amick had found a photoconductivity value for this zinc oxide of 5.8 x 10-9 which is higher than 10-9. Judged by the 10-9 standard, therefore, AZO 33 should have made a good print. The affidavit did not include Amick’s 1956 photoconductivity finding on this zinc oxide. 5.8 x 10~9 is lower than 10-8. Again, this indicates that 10-8 was a better criterion. 3. Finally, it should be noted that in 1961 Amick wrote a chapter for a book eventually published in 1965. The chapter was entitled “A Review of Electrofax Behavior.” In it he set forth precisely the same data as was contained in DeVore’s and Greig’s affidavits in 1956. This time Amick drew the conclusion that: “Oxides whose maximum surface conductivity lies above about 10"8 ohms-Vsquare/watt/cm2 are satisfactory for Electrofax use.” Amick was in charge of assembling the data to support the amended application. Greenspan was the one who wrote it up in the affidavits. Greenspan testified that Amick told him that “they had repeated these tests and some others and had gotten data that was somewhat different * * He testified that Amick asked, “Does this spoil our patent claims?” or words to that effect, to which Greenspan replied, “Don’t worry about that * * * I would take care of that. And you, Dr. Amick, take care of the technical side of it.” Greenspan further testified that he “did not want to see the details of it,” that he asked Amick to “evaluate the data” and to give him “representative values.” He said that this is what Amick did and that this is what he, Greenspan, presented to the Patent Office. Amick testified that he made the decision as to what photoconductivity measurements were to be reported to the Patent Office in these affidavits. More specifically, it was he who decided to report DeVore’s uncorrected 1953 results rather than DeVore’s 1955 results on FGS 8 or Amick and Giaimo’s 1956 results. He testified in substance that after discussing the situation with De-Vore, he concluded that measurement within an order of magnitude was “the precision we could expect.” RCA’s Subsequent Dealings with the Patent Office On the strength of these affidavits, Greenspan argued in support of RCA’s amended application that the 10-9 criterion which RCA had previously put forth in its continuation-in-part application filed on October 1, 1953, was a valid dividing line for distinguishing satisfactory zinc oxides from unsatisfactory ones. He referred to it as “applicant’s ti. e., Greig’s] discovery” and pointed out that no such criterion was revealed in the prior art. The result of all this argument and affidavit-drawing might well have been discouraging. On July 25, 1957, the Examiner again rejected all the claims. RCA, however, was undaunted. On January 22, 1958, it filed a further response to the Patent Office Action. It withdrew a few claims but urged reconsideration of the others. This perseverance was at last rewarded, at least partially. On July 17, 1958, the Examiner allowed the process claims but finally rejected the product claims. Additional argument on the part of RCA as to the latter was unavailing. In January 1959, RCA appealed to the Board of Appeals of the Patent Office. In its brief on appeal, RCA argued the importance of the 10-9 criterion. On September 30, 1960, the Board of Appeals affirmed the Examiner’s decision. In October 1960, RCA began an action against the Commissioner of Patents in the United States District Court for the District of Columbia, seeking a judgment compelling the Patent Office to grant a patent on the product claims. At the conclusion of the trial of this action on January 25, 1962, the District Court ruled that RCA had not shown that the Patent Office was in error in rejecting the claims. The court recommended, however, that RCA endeavor to amend the claims in an effort to overcome the Patent Office’s objection. In accordance with the court’s suggestion, RCA cancelled certain of the claims in issue and amended the others. The amendment did not change the 10-9 language. The Patent Office accepted the. amendment and at long last the patent issued on September 4, 1962. Of the twenty-two claims in the patent as finally issued, all but one, the first, included the phrase “surface photoconductivity higher than 10-9, etc.” The specification referred to this criterion and described the test by which the photoconductivity of a particular zinc oxide could be determined. The test so described was essentially DeVore’s procedure, with the notable exception that the patent did not mention that the measurements were to be made with the use of chopped light. The Alleged Fraud on the Patent Office The most grave charge of fraud on the part of RCA in dealing with the Patent Office is based upon the affidavits which RCA filed with the Patent Office in September 1956 and the argument made by RCA to the Examiner on the strength of those affidavits. As we have seen, in the continuation-in-part application filed in October 1953, RCA had taken the position, as a result of DeVore’s 1953 measurements, that zinc oxides having a photoconductivity value of at least “about 10~9 ohms-1/square/ watt/cm.2” were satisfactory for use in Greig’s process and that those with a lower photoconductivity value were not. The affidavits filed in September 1956 were intended to support that position. Amick’s affidavit stated that conclusion. The affidavits omitted data which bore upon the validity of this conclusion. They omitted the results of DeVore’s 1955 measurement of FGS 8 and of the measurements made by Amick, Giaimo and Spicer in 1956 of FGS 8 and AZO 33, all of which tended to contradict DeVore’s 1953 findings. They omitted all reference to DeVore’s opinion, expressed in his notebook in 1955, that his 1953 results should be corrected by raising them by one order of magnitude. The affidavits set forth DeVore’s 1953 results without any such correction. There is some evidence to the effect that photoconductivity measurements made by chopped light can be expected to vary by approximately one order of magnitude from one batch to another of the same zinc oxide, even though each batch is measured in exactly the same way. Young expressed that opinion in his June 1956 progress report. His opinion has found some unexpected corroboration years later in comments made in the initial report of the expert who testified for SCM at the trial. This variation is apparently what Amick had in mind when he testified that a measurement within one order of magnitude was “the precision we could expect.” It is tolerably clear that Amick justified the omission of the data which pointed to 10-8 as the proper criterion by persuading himself that, inasmuch as 10“8 was only one order of magnitude higher than 10-9, it was accurate enough to continue to put forward 10-9 as the correct dividing line. I am not convinced by this argument that the failure to reveal the results of the more recent chopped light measurements was proper. In the first place, if such a variation in results is to be expected, that fact in itself was important and should have been revealed. There was no hint of it in the affidavits or in Greenspan’s arguments to the Examiner. In the second place, the fact that some variation may exist about a given norm does not change the norm. The evidence establishes that the correct norm was at least ICh8 rather than 10~9. Amick himself expressed that opinion in his 1965 book, even on the basis of DeVore’s uncorrected 1953 data. If some variation is to be expected, and if DeVore’s 1953 data is increased by approximately one order of magnitude, as DeVore recommended, then the correct result could be nearer 10~7 than HH, about 100 times higher than 10-9. On the basis of the evidence I find that it was inaccurate to report 10-9, without change or qualification, as the correct standard. Did RCA, that is to . say, the RCA scientists and lawyers who had a hand in this affair and whose knowledge binds RCA, know that it was inaccurate? I cannot escape the conclusion that they did. Amick knew all the facts. Young, as the head of the project, must have known them. In all likelihood Greenspan knew them as well. In any event, he should have known them. It was his duty to inform himself about them. He could not avoid responsibility by trying not to “see the details.” Charles Pfizer & Co. v. Federal Trade Commission, 401 F.2d 574 (6th Cir. 1968), cert. denied, 394 U.S. 920, 89 S.Ct. 1195, 22 L.Ed.2d 453 (1969). I appreciate that there are times when the conclusions to be drawn from scientific experiments come down to a matter of professional judgment or scientific debate. See Armour & Co. v. Wilson & Co., 274 F.2d 143 (7th Cir. 1960); Ritter v. Rohm & Haas Company, 271 F.Supp. 313, 343 (S.D.N.Y.1967). In my opinion this is not such a case. I find that RCA knew that the unqualified assertion of 10~9 as the correct standard was inaccurate. I find that RCA continued to assert that standard to the Patent Office because RCA feared that otherwise its chance of obtaining a patent would be impaired. In somewhat incoherent testimony Greenspan as much as admitted that he was afraid that any changes in the results of the 1953 measurements or any concession that variations in result from batch to batch could be. expected to occur would cast doubt upon the contention made by RCA to the Patent Office that it had discovered a practicable scientific method for separating satisfactory zinc oxides from unsatisfactory ones. Greenspan testified : “The problem I had was that we had gone into the Patent Office and we said there are two classes of zinc oxide that are separated by the value of surface photoconductivity. * * * So, if the test itself was not reproducible, or if it didn’t really divide zinc oxides into two parts, then right at the outset it wasn’t a worthwhile thing to go after. * -X- * * * * “Now, I knew of the criterion. It was my problem to present to the Patent Office reasons why I thought this limitation in the claim distinguished the claim over the prior art and whether he spoiled it as related to that problem that I had: that is, would it make it more difficult for me to argue patentability with this limitation.” It may be unduly harsh to characterize RCA’s conduct here as fraud. Conduct which has been so labeled in the decisions has, by and large, been more reprehensible than this. But at the least, it was conduct which was lacking in candor. It was intentional nondisclosure of relevant data which might have affected the outcome of the patent application. It has been held in several cases that the party asserting invalidity of a patent on the ground of fraud has the burden of proving, not only the fraud, but also the fact that the fraud was material, i. e., that if the Examiner had known the true facts, he would not have authorized the issuance of the patent. Charles Pfizer & Co. v. Federal Trade Commission, supra; Corning Glass Works v. Anchor Hocking Glass Corp., 253 F.Supp. 461 (D.Del.1966), rev’d on other grounds, 374 F.2d 473 (3d Cir. 1967), cert. denied, 389 U.S. 826, 88 S.Ct. 65, 19 L.Ed.2d 80 (1967); Baldwin-Lima-Hamilton Corp. v. Tatnall Measuring Systems Co., 169 F.Supp. 1 (E.D. Pa.1958), aff’d, 268 F.2d 395 (3d Cir. 1959), cert. denied, 361 U.S. 894, 80 S.Ct. 198, 4 L.Ed.2d 151 (1959). In at least one Supreme Court decision, misleading affidavits submitted by the applicant which “were not the basis for it [the granting of the patent] or essentially material to its issue” were held insufficient to overcome the presumption of validity which attaches to a patent. Corona Cord Tire Company v. Dovan Chemical Corp., 276 U.S. 358, 374, 48 S.Ct. 380, 384, 72 L.Ed. 610 (1928). The issue of materiality is an issue of ultimate fact to be determined by the court, as any other fact, on the basis of all the evidence. Although there is language in Hazel-Atlas Glass Co. v. Hartford-Empire Co., 322 U.S. 238, 247, 64 S.Ct. 997, 88 L.Ed. 1250 (1944), which is susceptible of the construction that materiality is to be presumed, I cannot reconcile this language, if taken literally, with the other decisions. The facts in Hazel-Atlas were unusual and somewhat extreme. I do not regard the Court’s statement that the patentee was “in no position now to dispute” the effectiveness of the misrepresentation as establishing a rule for all cases. Proving materiality is by no means an easy task. In Charles Pfizer & Co. it was established by calling the Examiner himself who testified that he would not have granted the patent had he known the facts which defendants failed to disclose to him. But this is an unusual procedure, and one which is not normally possible in view of the policy of the Patent Office against permitting its Examiners to testify in private litigation. In the absence of such direct testimony, the court must fall back upon the Patent Office record, the “file wrapper” of the patent, for such light as it may cast upon the problem. In this instance, as is frequently the case, the light that the file wrapper casts is dim. After a painstaking study of the difficult language of the Examiner’s “actions,” I have found no convincing evidence, one way or the other, as to what the Examiner would have done had he been informed in September 1956 that photoconductivity measurements by the DeVore method were somewhat inexact and that in any event, 10~1 was a more accurate minimum measurement for good zinc oxides than 10”9. Conceivably, he would have discarded the DeVore procedure as worthless and would have denied the patent on the ground that RCA had failed to provide a reliable scientific method for distinguishing good zinc oxides from bad. On the other hand, for all that appears, he might have concluded that a specification of at least 10-8, even allowing for some variation, was sufficiently definite. In the final analysis, what position the Examiner would have taken had RCA been candid with him remains a matter of speculation. Under these circumstances, I am compelled to conclude that SCM has failed to carry the burden of proving that RCA’s nondisclosure was material in a “but for” sense, i. e., that the patent would not have issued if RCA had revealed all the relevant information. I must conclude therefore that SCM has not proved that RCA procured the patent by fraud and has not shown the patent to be invalid on that ground. This conclusion, however, does not dispose of the matter. We still have to deal with the doctrine of “unclean hands.” The line between fraud which will invalidate a patent and “unclean hands” which will bar its enforcement is a shadowy one. The decisions generally are not clear on it. Some courts phrase the results in terms of invalidity, others speak of unenforceability. This is a case in which the problem must be faced. There is broad language in several decisions to the effect that a patent is affected with a public interest, that complete candor and full disclosure is required of a patent applicant in his dealings with the Patent Office, and that any inequitable conduct on the part of the applicant in obtaining a patent will be sufficient to dissuade a court of equity from rendering him its aid in enforcing the patent against infringers. Hazel-Atlas Glass Co. v. Hartford-Empire Co., 322 U.S. 238, 64 S.Ct. 997, 88 L.Ed. 1250 (1944); Kingsland v. Dorsey, 338 U.S. 318, 70 S.Ct. 123, 94 L.Ed. 123 (1949); Precision Instrument Manufacturing Co. v. Automotive Maintenance Machinery Co., 324 U.S. 806, 65 S.Ct. 993, 89 L.Ed. 1381 (1945); Keystone Driller Co. v. General Excavator Co., 290 U.S. 240, 54 S.Ct. 146, 78 L.Ed. 293 (1933); Beckman Instruments, Inc. v. Chemtronics, Inc., 428 F.2d 555 (5th Cir. 1970). These decisions do not appear to require a finding that the inequitable conduct had a “but for” effect on the granting of the patent as a prerequisite to a court’s refusal to enforce it. Judge Wright was concerned with this problem in Corning Glass Works v. Anchor Hocking Glass Corp., supra. Having found that the party challenging the patent had failed to establish its invalidity on the ground of fraud because he had failed to prove that the patent would not have issued if the true facts had been revealed, he went on to say: “Even though misrepresentations made to the Patent Office are not legally material to the issuance of a patent, nevertheless, this Court, being a court of equity, can and should refuse to enforce the patent if the Court finds the patentee made intentional misrepresentations to the patent examiner, i. e., if the patentee came into the court with unclean hands. The proceeding before the patent examiner is ex parte and an examiner has no way, in many cases, to ascertain the truthfulness of the representations made to him. Necessarily he must rely on the good faith of the applicant. Absolute honesty and good faith disclosure is necessary.” 253 F.Supp. at 470. I agree with this statement. It fits the present case. No one can tell with certainty what would have happened if RCA had dealt fairly with the Patent Office. But the fact remains that RCA did withhold relevant facts. Which side in this litigation is to suffer from this conduct? It is appropriate that it should be RCA who suffers. Any other rule would fail adequately to discourage conduct of this sort merely because of the circumstance, which must be present in many cases, that it turns out to be impracticable to ascertain what the Examiner, who did not know the true facts, would have done if he had known them. The evidence here justifies the conclusion that this court should not enforce a patent obtained under these circumstances. I so hold. SCM’s Other Charges of Fraud In view of the conclusion I have reached on the 1956 affidavits, it is unnecessary to dwell upon SCM’s other accusations. Mention may be made of the only one which involved any misstatement of fact. This consisted of a series of references in Greenspan’s written arguments to the Patent Office, one made in September 1956 and another in January 1959, to “the applicant,” Greig, as the discoverer of the photoconductivity measurement technique. Of course, Greig did not devise this test, and hence, speaking literally, these statements were not true. But they could hardly have been intended to be taken literally, and in any event, they could not have misled the Examiner. In the first place the Examiner, accustomed to dealing with RCA and other large corporations, must have understood that RCA operated in the production of patents on what might be called an assembly-line basis. One man had an idea, another man tested it, others were called upon to make still more tests or measurements, and a patent lawyer, himself a full-time RCA employee, wrote up the application and the affidavits and handled all the dealings with the Patent Office. At the very outset of a patent prosecution, the inventor assigned all his rights to RCA and designated an RCA attorney as his agent to represent him in prosecuting the application. This was not a case of a lone inventor struggling to do by himself all the inventing and all the testing and all the paper work. It was a large-scale cooperative enterprise. In the second place, it was abundantly clear from the affidavits that it was DeVore, not Greig, who had made the photoconductivity measurements. Although Greenspan’s choice of words may have been unfortunate, there is no sound basis here for invalidating the patent or holding it unenforceable. The same can be said of the other statements or omissions of which SCM complains. I have considered each but believe it unnecessary to discuss any of them in detail. They are at most highly debatable. They are not sufficiently convincing evidence of misrepresentation or nondisclosure. The conclusion which I have reached as to the unenforceability of this patent is based solely upon the nondisclosure of the photo-conductivity results in 1955 and 1956. Section 112 Each of the twenty-two claims in the 539 patent, except the first, is limited to a process or a product which employs a zinc oxide “having a surface photo-conductivity higher than 10-9 ohms-1/ square/watt/cm.2 at about 3900 A.” The specification states: “In order to determine whether or not a particular zinc oxide is suitable for use in the present invention, a test of its photoconductive properties may be made as follows.” The specification then proceeds to describe DeVore’s technique in measuring the photoconductivity of zinc oxide pellets, but it does not specify whether the light to which the pellets are subj'ected is steady or chopped. Since it does not even mention the chopping, obviously it says nothing about a chopping rate. The description concludes with the following passage: “It has been found that, to be useful in the present invention, the zinc oxide selected should have a surface photoconductivity of at least about 10“9 ohmrVsquare / watt/cm.2 when exposed to a wavelength of about 3900 A. “Having established the threshold value of a surface photoconductivityneeded in the process, it is possible to test any zinc oxide otherwise suitable from the standpoint of stability, compatibility, dark resistivity, color, etc., in order to determine whether it can be used.” It, is apparent that RCA’s purpose in so limiting the patent claims and in describing the test by which that limitation can be applied, was to differentiate this invention from prior art patents, notably Patent No. 2,287,348 to Hayden, which spoke in general terms of coating paper with zinc oxide. RCA recognized that it was necessary to make clear that it did not claim all zinc oxides. It claimed only those which possessed a specific minimum photoconductivity as determined by a specific test. 35 U.S.C. § 112 requires that the specification shall contain “a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains * * * to make and use the same * * The section also requires that: “The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.” The question is whether the 539 patent complies with these requirements. SCM contends that it does not because (1) the l(h9 criterion is incorrect, and (2) the description of the measuring procedure is unworkable because it does not specify the type of light to be used. I conclude that these contentions must be sustained. I have already held that the 10~9 number is incorrect and that it should be at least 10~8, with the additional qualification that some variation is to be expected in testing different batches of the same zinc oxide. As they now stand, the claims are too broad. They include zinc oxides which in fact are not satisfactory for the purposes of the patent. It follows that the claims do not distinctly claim the subject matter of the invention. The failure of the specification to identify the type of light to be used is another fatal defect. The evidence is clear that the results obtained with steady light are startlingly different from those obtained with chopped light. Moreover, even as to chopped light, there is evidence to show that the results vary according to the rate of the chopping. It is clear, therefore, that, to be of any value to one desiring to make the test, the description should have specified, not only that the light should be chopped, but also how frequently it should be chopped, i.e., the number of flashes of light per second. No reason is apparent for the omission of this vital information from the patent specification. DeVore’s affidavit clearly stated that he had used light chopped 23.5 times per second. The Patent Office was thus fully advised on this point and no question of fraud or nondisclosure can arise here. I can only surmise that the failure to include this data in the patent specification was due to inadvertence. This explanation is made more probable by the fact that in another Greig patent, No. 3,060,020 issued on October 23, 1962, entitled “Method of Eleetrophotographically Producing a Multicolor Image,” DeVore’s measuring procedure is again described, but this time the description is explicit in stating that: “The light beam projected onto the surface is chopped at about 23.5 c.p.s. by a constant speed rotating disc, pierced to produce equal intervals of light and darkness.” Unfortunately for RCA, the inclusion of this information in the affidavit does not save the patent. It must appear in the patent itself. Kaiser Industries Corp. v. McLouth Steel Corp., 400 F.2d 36 (6th Cir. 1968), cert. denied, 393 U.S. 1119, 89 S.Ct. 992, 22 L.Ed.2d 124 (1969). For the same reason it would seem to make no difference that descriptions of DeVore’s procedure which appeared from time to time in various learned publications mentioned the fact that chopped light was employed. There is a conflict in the opinions of the respective experts as to whether a person familiar with these matters in reading the patent would necessarily understand that the use of steady light could not have been intended. I am not persuaded that he would. In any case, there is no way in which such a knowledgeable reader could have divined the chopping rate, even if he did believe that chopped, rather than steady, light was to be used. Curiously enough, it seems to be the fact that in actual practice, no one pays any attention to this elaborate measuring technique. Certainly Greig never did. He determined which zinc oxides would work and which would not by trying them out, i.e., by making prints with them. Apparently RCA’s licensees do the same, or perhaps by this time the suitable zinc oxides are so well known that there is no need to measure their photoconductivity. SCM alone seems to have concerned itself with DeVore’s procedure, as will appear in the subsequent discussion of SCM's second count. This fact makes this whole discussion seem theoretical and unreal. Section 112, however, makes no allowance for such a situation. It requires a patentee to describe his invention in “full, clear, concise, and exact terms,” in his patent specification and to distinctly claim the subject matter in his patent claims, whether anyone bothers to read them or not. Only by so doing can he secure his monopoly. The court has no discretion. It must enforce the statute. Since it is clear that the specification and the claims do not fulfill the requirement of Section 112, claims 2 to 22, inclusive, are invalid. As to claim 1, the situation is somewhat different. That claim does not refer to the 10-9 criterion. SCM attacks it nevertheless for overclaiming, in violation of Section 112. I conclude that this contention must be sustained. Claim 1 pertains to a “process of forming a visible image.” It states that one of the steps of that process is “providing a recording element comprising a composition comprising a photoconductive zinc oxide suspended in an electrically insulating, film-forming vehicle for said zinc oxide.” All zinc oxides are photoeonductive to a greater or lesser extent. Claim 1 includes all of them, despite the fact that, as the specification emphasizes, not all zinc oxides are useful in this process. Claim 1 is therefore invalid. Corona Cord Tire Company v. Dovan Chemical Corp., supra; Teleflex Incorporated v. American Chain & Cable Co., 273 F.Supp. 573 (S.D.N.Y.1967). The claim appears to have been intentionally drafted in order to obtain a monopoly broader than is justified. Under such circumstances, the claim cannot be upheld by reading it in the light of the specification. Georgia-Pacific Corp. v. United States Plywood Corp., 258 F.2d 124 (2d Cir. 1958), cert. denied, 358 U.S. 884, 79 S.Ct. 124, 3 L.Ed.2d 112 (1958). Section 102 This section presents no difficulty. Section 102 prevents patentability only where the invention was “identically disclosed” by the prior art. Ling-Temco-Vought, Inc. v. Kollsman Instrument Corp., 372 F.2d 263 (2d Cir.1967). That is