Full opinion text
OPINION MacMAHON, District Judge. Plaintiff, John J. Ritter, New York University Professor Emeritus in organic chemistry, sues for infringement of his United States patent, No. 2,573,673, entitled “Method of Producing N-Mono Substituted Imidic Compounds,” issued October 30, 1951 on a continuation in part application filed May 27, 1950 and an earlier application filed January 27, 1945. Defendant, Rohm & Haas Company, a manufacturer of chemicals, denies infringement and counterclaims for declaratory judgment of noninfringement, invalidity, and unenforcéability. Only claims 1, 4-7, 13, and 16-18 of the Ritter patent are before the court for adjudication. In essence, all of them claim discovery of a single step in a process for producing N-mono hydrocarbon substituted imino compounds by reacting three components — any nitrile, certain cationoid compounds, and certain ethylenic substances — under substantially anhydrous conditions. It is important to note that the patent teaches, but does not claim, a second step: that “the products of * * * [the first step] constitute intermediates * * * which will further react readily with water * * * forming various amidic compounds. In the case of water, the resultant products are N-mono substituted amides.” Defendant manufactures four N-mono substituted amides which it converts to amines. Each of its accused processes reacts hydrogen cyanide (a nitrile), sulfuric acid (a cationoid), and an olefin (an ethylenic substance). There is no question that all of those reactants fall within the class of compounds claimed by plaintiff’s patent. The parties are at odds, however, on (1) whether defendant’s processes use water as a fourth component of a single step reaction, and (2) whether defendant’s processes produce an imino compound. Denying infringement, defendant claims (1) that its accused processes are not conducted under substantially anhydrous conditions because they all use water as a fourth reactive component in a single step reaction, and (2) that its processes produce, not an imino compound, but a protonated amide, which it then reacts with water to produce an N-mono substituted amide. Asserting infringement, plaintiff claims (1) that defendant’s processes are conducted under substantially anhydrous conditions because no water, as such, is present and, in any event, water is not a reactive fourth component, and (2) that defendant’s processes produce an imino compound which defendant then reacts with water to produce an N-mono substituted amide. As we said above, Ritter teaches a first step (producing the imino compound) and then a second step (reacting the imino compound with water to produce an N-mono substituted amide). Ritter originally tried to patent both steps. The Patent Examiner rejected this application as unpatentable over the prior art, particularly the Gresham patent. In order to distinguish his patent from the prior art, Ritter gave up the second step and restricted his claims to the first step, the one-step, three-component reaction “under substantially anhydrous conditions.” Defendant claims that the quoted phrase impales plaintiff on the horns of a dilemma. If the phrase is stretched to include as much water as defendant uses, Ritter’s patent (1) reads on the prior art, or (2) covers a different reaction which produces a different product which Ritter never taught or claimed — the protonated amide. If it is restricted to exclude that much water, defendant’s processes do not infringe. The Issues Defendant has also leveled a barrage of other defenses. Defendant contends that the patent is invalid because of anticipation, obviousness, fatally indefinite claims, and misrepresentations to the Patent Office. Defendant asserts that the patent is unenforceable against it because of laches. Moreover, in arguing non-infringement, defendant invokes the doctrine of file wrapper estoppel. We now consider these issues seriatim. A. Validity (1) The I. G. Farben French Patent The first prior art reference cited by defendant is I. G. Parben’s French patent, No. 902,342 (PX 2), which bears a “delivre” date of December 4, 1944. Ritter concedes that this patent discloses his invention, but asserts that it does not invalidate his patent because (1) Farben’s process was not “patented” within the meaning of 35 U.S.C. § 102(a) until after January 27, 1945, the date Ritter filed his application, and (2) even if Farben’s process was “patented” on December 4, 1944, Ritter actually invented his process before that date. (a) Date of the Farben Patent The court tried the issue regarding the date of the Farben patent before all others and found from the bench that I. G. Farben’s process was “patented” within the meaning of 35 U.S.C. § 102(a) on December 4, 1944 (Tr. 151-54). There is no dispute that, in France, all the rights of a patentee accrue to him on the “delivre” date of his patent (Tr. 36-40, 72-73). On that date, the Minister of Industry grants the application by signing a decree. Soon afterward the Patent Office notifies the applicant. As of the “delivre” date, the patentee acquires a monopoly right to exclude others, and he can sue for infringement. However, unlike American patents, French patents are not published on the same day they are granted. Here, the fact that the Farben application had been granted was not published in the Bulletin Officiel de la Propriete Industrielle (the “BOPI”) until July 6, 1945, and the text of the patent was not published in printed form until August 28, 1945. Plaintiff relies on In re Ekenstam, 256 F.2d 321 (C.C.P.A.1958), which held that a foreign invention is not “patented” within the meaning of 35 U.S.C. § 102(a) until the patent is available to the public. Ekenstam explicitly contradicted General Electric Co. v. Hygrade Sylvania Corp., 61 F.Supp. 476, 529 (S.D.N.Y.1944). Both cases involved Belgian patents. General Electric stated in a terse dictum that a Belgian invention is “patented” as soon as the ministerial decree issues, rather than when the patent is published. General Electric relied on Sirocco Engineering Co. v. B. F. Sturtevant Co., 220 F. 137 (2 Cir. 1914), cert. denied, 238 U.S. 636, 35 S.Ct. 939, 59 L.Ed. 1500 (1915), which involved two French patents dated 1890 and 1896. However, there are several crucial differences between the Belgian system and the old French system involved in Sirocco (Tr. 80-82; In re Ekenstam, supra, 256 F.2d at 324). Therefore, General Electric is not entitled to much weight. We shall assume, arguendo, that Ekenstam is correct and that the word “patented” in 35 U.S.C. § 102(a) connotes “available to the public.” The Second Circuit has never said otherwise, for in Sirocco it was undisputed that the 1890 and 1896 patents there involved were available to the public. Thus, the question before us is whether Farben’s French patent was available to the public on December 4, 1944. Plaintiff contends that it was not available until July 6, 1945, when the BOPI published a short paragraph listing I. G. Farben’s name, the date of its application, and the title of its patent (PX 5, p. 355). The BOPI is simply a periodical list of patent applications which have been granted. It is undisputed that the French practice from 1903 to 1964 was to type the patentee’s name, the date of its application, and the title of its patent on a file card which was available to the public in the French Patent Office even before the patent was granted (Tr. 49, 73-75). However, plaintiff’s expert, Dr. Soubbotitch, testified that during World War II the file cards were not kept and the Patent Office was closed (Tr. 49). This was flatly contradicted by defendant’s expert, Mr. Miller, who testified that “the Vichy Government functioned very, very well,” that the file cards were at all times kept quite regularly (Tr. 105-08), and that the I. G. Farben patent could not have been given its number unless a file card had been kept for it (Tr. 117). Neither witness was in France during World War II. Having observed both witnesses and their manner of testifying, the court was more impressed with the demeanor of Mr. Miller and with his answers, which were frank and responsive. The court found him the more credible witness. Accordingly, we find that in 1944 and 1945 the file cards were kept regularly for each and every patent in the French Patent Office and were available to the public even before the patents were granted. There were “watchers” in the French Patent Office who would search the card file for inventions which might interest their employers (Tr. 97, 106-07). The title on every card precisely expressed the subject, matter (Tr. 97-99). Once the “watcher,” or anybody else, found a card that interested him, he could, between 1903 and 1955, obtain a copy of the patent as soon as it was granted, i. e., on the “delivre” date. Miller testified that all that was necessary to obtain a copy of the patent was the payment of a five franc fee fixed by the decree of 1903 (Tr. 74, 82-83 and 101-02 referring to PX 9 for Id., p. 2, item 6). Soubbotitch conceded that five francs was all that was necessary after the BOPI date (Tr. 64). Nevertheless, he claimed that, between the “delivre” date and the BOPI date, no member of the public could obtain access to the patent unless he obtained permission from the Minister of Industry (Tr. 44, 53, 56-57). In support of his testimony, Soubbotitch relied on: (1) an order of 1955 (not in evidence) which provided that patents be kept secret until the BOPI date (Tr. 50-53, 74-75), and (2) a decree of 1966 (not in evidence) which provides that, between the “delivre” date and the BOPI date, no member of the public can obtain access to the patent unless he receives permission from the Minister of Industry (Tr. 62-64). Soubbotitch claimed that both provisions are declaratory of the previous law (Tr. 51-53, 56-57, 64-65). Miller disagreed (Tr. 74-75). We find that neither provision is declaratory of the previous law. There is a gap of eleven years between these two provisions, and during that time, in 1964, the French discontinued the card file system (Tr. 85). It is clear that the French law underwent successive adjustments. We are convinced that Soubbotitch was trying to divine the 1944 law from the state of the law in 1966. He is contradicted by the Langer Bulletin, “Foreign Digest Foreign Patents,” 4th ed., p. 107, which was published in 1951 (Tr. 47-48). Defendant’s position is also confirmed by the Patent Office Board of Appeals, which is bound by the Ekenstam decision. Ex parte Gruschwitz and Fritz, 138 U.S.P.Q. 505, 510 (1961), discusses Ekenstam and explicitly reaffirms the Board’s longstanding practice of considering a French invention “patented” on its “delivre” date. In any event, the dispute between Miller and Soubbotitch is not really substantial, for Soubbotitch admitted that he did not know what criteria, if any, governed the exercise of the discretion of the Minister whose permission he claimed was required (Tr. 45). Even if we were to assume that permission was required, for all we know it was a mere ritual and in no way prevented the public from seeing the patents. We found Miller the better qualified of the two experts. He is a practitioner of patent law in France. Soubbotitch is associated with a law firm in New York, and, although an expert in French law generally, he has no particular background in patent law (Tr. 31-33). His knowledge of French patent practice is based almost solely on what he has read and heard. There is some indication in the evidence that no third person saw this particular French patent, but that is immaterial (PX 6, last paragraph). The controlling question under the Ekenstam, case is not whether anyone in fact saw the patent but whether the patent was available to the public. Accordingly, we find that French Patent No. 902,342 was available to the public and was “patented” within the meaning of 35 U.S.C. § 102(a) on December 4, 1944. This finding raises the question of whether Ritter is entitled to a pre-filing invention date before December 4, 1944. (b) Pre-Filing Invention Date Ritter is entitled to a prefiling invention date if he conceived and reduced his invention to practice before the key date, December 4, 1944. The burden of proof required to establish a pre-filing date of invention rests on the plaintiff, and it is a heavy one; the court “must be persuaded with a certainty which is seldom demanded elsewhere; quite as absolute as in a criminal case * * Moreover, a pre- filing invention date may not be established by the inventor’s “uncorroborated testimony” alone. (i) Conception “The conception of the invention consists in the complete performance of the mental part of the inventive act. All that remains to be accomplished in order to perfect the act or instrument belongs to the department of construction, not invention. It is therefore the formation in the mind of the inventor of a definite and permanent idea of the complete and operative invention as it is thereafter to be applied in practice that constitutes an available conception within the meaning of the patent law. A priority of conception is established when the invention is made sufficiently plain to enable those skilled in the art to understand it.” Plaintiff contends that he conceived his invention on November 6, 1944, when, after a series of earlier experiments and observations of the reaction of his three components, he expressed his idea of what had occurred in specific formulas and equations recorded in his notebook. Defendant contends that there is no corroboration for Ritter’s claimed conception and that, at best, the formulas and equations set forth in his notebook are mere speculations. As proof of conception, Ritter testified that commencing on October 19, 1944 he conducted a series of experiments reacting the three components of his patent and formed an N-mono alkyl imino compound (Tr. 159-163, 238, 257). He offered his notebook in corroboration. The authenticity of the notebook is conceded (Pre-Trial Order, p. 2, par. 4). Its genuineness is attested by its appearance, its chronological order spanning a period of twelve years, from 1933 to May 15, 1945, and its notation of a great variety of experiments covering 300 pages, most of which are unrelated to the claimed invention. Defendant failed to impeach the notebook in any way, and there is not the slightest suggestion in the evidence of fabrication. Defendant asserts that Ritter’s notebook is of no evidentiary value, regardless of how persuasive it may be, because it is a “self serving” document and not the “independent corroboration” the law demands. We note at the outset that this is different from defendant’s legal contention at the pre-trial conference (January 27, 1967 Tr. 13-14). There is no question that Ritter’s oral testimony, standing alone, is insufficient to prove conception. The notebook is the only corroborative evidence introduced by plaintiff; if it is legally insufficient to establish corroboration, Ritter cannot prove an invention date prior to the French patent, and his patent is invalid. We have found no case in this circuit expressly dealing with the nature of the corroborative evidence required. Decisions elsewhere are split. Some rigorously demand “independent” evidence; any documents prepared by, or under the supervision of, the inventor are “self serving” and cannot provide corroboration. Others stress that it is the oral testimony of the inventor which must be corroborated, and this can be done by any means, including the use of diaries, notebooks, and records prepared by the inventor, as long as this evidence is “clear and convincing.” We are disinclined to rely on quantitative rules of evidence in our search for truth. Their appeal of simplicity is outweighed by the vice of blindness. Rather, we look to the purpose behind the rule requiring corroboration for guidance in its application. The manifest purpose of the rule is to prevent fraud. When the validity of a patent turns on the exact date a certain event occurred, or discovery was made, there is an inherent risk of perjury if after-the-fact oral testimony by the most interested party, the alleged inventor, can carry the invention date back beyond the filing date. In light of that purpose, it would be absurd to hold that Ritter’s notebook is not sufficient corroboration, but that “independent” evidence of his conception date is required. The issue here is the exact date when, in late 1944, almost twenty-three years ago, Ritter conceived his invention. His notebook, a document of uncontested authenticity, is a contemporaneous record of his thoughts and actions. It is hard to imagine what more reliable corroborative evidence could be found. Indeed, we note that it was similar documentary evidence that Judge Learned Hand held sufficient to establish a pre-filing invention date in United Shoe Machinery Corp. v. Brooklyn Wood Heel Corp., 72 F.2d 263 (2 Cir. 1935). To assert that the “independent” recollection of some of Ritter’s former students about what transpired during the few crucial weeks in 1944 would be more probative than the immutable notebook flies in the face of common experience. Memories are fallible, particularly in trying to recall the precise date of long forgotten events whose importance is only subsequently created by the Byzantine nuances of litigation. To rule out Ritter’s notebook on the ground that it is “self serving” is to exalt labels over reason. There is not the slightest hint that the notebook is a fabrication or that any of the entries are not genuine expressions of Ritter’s thoughts and deeds on the date entered. Of course, if the notebook is probative, it is “self serving” in the sense that it aids Ritter’s cause, but all evidence suffers from such corrupting limitations. . We hold, therefore, that Ritter’s notebook is admissible as corroborative evidence and turn to a detailed examination of its contents. On November 6, 1944, at page 245 of his notebook, Ritter wrote: “Assembling these facts, and harmonizing them with general considerations from the nature of the reaction system, I now believe the following to have occurred: Ritter testified that the formula at the right of the third equation, is the formula for an N-mono hydrocarbon substituted imino compound, and that it is the same formula appearing in more generalized form in his patent. Defendant’s expert, Dr. Mark, agreed (Tr. 301-02), and we so find. Plaintiff contends that the foregoing evidence proves conception of his invention before the key date. Defendant argues, however, that Ritter’s testimony and his notebook show only speculation. It asserts that the notebook shows that Ritter did not have the vaguest notion of his invention until years later, when he filed his amended application for his patent. Before the key date, Ritter never named any of'the structures on page 245 except the final amide (see Tr. 667). One should not assume that at that time he gave no more significance to the imino at the end of the third equation than he gave to the structures at the end of the first and second equations. As we shall see, the imino compound was a tangible “intermediate;” the structures at the end of the first and second equations were intangible “transients.” (See Tr. 636-37.) At page 245, Ritter used the word “believe.” Ritter testified on his pretrial deposition that “in other places in the notes” he wrote down what he thought might happen, sometimes what did happen, and sometimes his interpretations (Tr. 650). As late as January 6, 1945, he wrote “speculations on reaction mechanism or dynamics” (PX 12, p. 276). Defendant stresses the testimony of its experts that Ritter’s equations were speculations and that other possible, and more likely, intermediates may also be speculated (Tr. 275-280, 321-332, 509, 971-73; PX 16; DX AA, AB and AC). Defendant further points to the fact that the only product Ritter analytically identified was the amide and that he never isolated, purified, or identified his postulated imino by accepted chemical analyses and tests (Tr. 237-39, 251, 643-44). Finally, defendant cites the testimony of its own expert that, without such tests, no one could determine the actual structure of the intermediate (Tr. 282). 'When expert chemists, like plaintiff and defendant’s experts, use the words “speculation” or “belief,” they do not mean that their formulas and equations are wild, irrational, unintelligible guesses. Rather, it is perfectly plain that when skilled chemists, like Ritter, first experiment with a chemical reaction and then write specific structural formulas of starting compounds, reaction mechanisms and resulting products, they are applying well-known laws, concepts and symbols of chemistry along with background knowledge to give rational physical explanations of observed facts (Tr. 180; 565-69; 669; 671; 673). Plaintiff’s expert Meyers had no difficulty either in understanding or explaining Ritter’s formulas and equations (Tr. 565-69) or in recognizing the intermediate as an N-mono hydrocarbon substituted imino compound (605-08). Defendant’s expert Mark conceded that he understood Ritter’s formulas and equations, that they were expressed in accordance with the laws of organic chemistry, and that Ritter’s postulated imino compound was one possible intermediate at the stage of the reaction where it is shown in his equations (Tr. 271-73; 282). Thus, there can be no question that Ritter’s formulas and equations made his invention sufficiently plain to enable those skilled in the art to understand it (Tr. 605-08). It is, therefore, of no moment that the words “imide,” “imidic,” and “imino” do not appear anywhere in Ritter’s notebook, nor that he never used any such terms until they first appeared in his amended application of 1951 (Tr. 676). As Ritter testified, although he did not name any imino compound, he knew that the reaction shown in the notebook involved an imino compound, and he clearly expressed the imino by structural formula (Tr. 676-77). We agree. Words are unnecessary to express an idea. Symbols, formulas, and equations are the non-verbal language of organic chemists and convey meaning better than words to those skilled in the art. In the modern world of science, symbolic communication rules vast domains of knowledge, experience, and meaning, and we must recognize that reality. That Ritter’s formula for his intermediate would convey his idea to skilled chemists is conclusively demonstrated by the fact that defendant’s employee, Dr. Bortnick, working independently with the reactions of the same three components and long before any prospect of this litigation, postulated an imino compound using the same formula as Ritter (PX 28, p. 3). The focus of the inquiry on the issue of conception is the idea formed in the mind of the inventor and not on whether his idea is right or wrong. It is, therefore, immaterial on the issue of conception that others skilled in the art might, with equal logic, conceive other possible mechanisms and intermediates. True, the postulated imino compound was not isolated or identified by chemical tests of the type urged by defendant, but, as we shall see, such tests are unnecessary in the circumstances. The question to be answered is: Did Ritter have a definite and permanent idea of the invention expressed in his claims? Manifestly, Ritter’s formulas and equations of November 6 show not some vague ruminations of what might, or might not, have happened, but a decisive, unequivocal, definite, permanent, and clear expression of a completed and concrete idea. He had made up his mind without doubt or ambiguity and expressed his thoughts in specific formulas and equations. The mental part of his invention was over and done with, then and there, and he was committed to the starting materials, the mechanism of the reaction, and the intermediate product involved here. There is not the slightest evidence that Ritter ever wavered or changed his mind before the filing date. The January 6 notation “speculations on reaction mechanism” in no way indicates that Ritter’s conviction was wavering or that he had any second thoughts about the concrete idea expressed earlier on November 6, at page 245 of his notebook, where he shows his invention in its entirety, as expressed in the claims of his patent. The later “speculations” of January 6, at page 276, are merely an even more elaborate explanation of the same completed basic concept expressed earlier. As Ritter testified, he was merely studying the reaction on January 6, still learning about it, and he has been learning about it ever since (Tr. 649). Surely continued study of the theoretical underpinnings of his process did not negative the finality of his original conception. We hold, therefore, that Ritter did conceive his invention before the key date. We turn to the question of whether he reduced his invention to practice. (ii) Reduction to Practice What constitutes a reduction to practice depends on the particular facts of' each case. Ritter’s invention was not upon a product but upon a step in a process. The issue is, therefore, governed by the rule that “[a] process is reduced to practice when it is successfully performed.” Reduction to practice, thus, requires proof that the process was actually carried out, that it worked to produce its intended product, and that the product was useful. “Products are useful if they serve as starting materials or intermediates in producing other materials or articles which are directly useful.” The intended result of the performance of Ritter’s process was to produce an N-mono hydrocarbon substituted imino compound useful as an intermediate to make an amide. Contrary to plaintiff’s contention, the mere performance of the process recited in the claims is not enough. To establish a reduction to practice, it must also be shown that the specified imino compound was actually produced. Ritter’s right to a reduction to practice, therefore, depends upon proof (1) that he did perform his process on or before the key date, December 4, 1944; (2) that performance resulted in the production of an N-mono hydrocarbon substituted imino compound; and (3) that the imino was useful as an intermediate to produce a useful amide. Plaintiff claims that prior to December 4, 1944 he conducted a series of experiments reacting the three components of his patented process and produced an N-mono hydrocarbon substituted imino compound which he either hydrolyzed to an amide immediately or stored and hydrolyzed to an amide later. Defendant contends that Ritter’s experiments, prior to the key date, did not establish that the reaction of the three components of his patent actually produced an imino compound; that Ritter was only speculating its existence because he never isolated, purified, analyzed or identified any intermediate as an imino compound by standard chemical tests; that his tests failed to exclude the possibility of other theoretical intermediates; that his laboratory runs did not employ the equal molecular proportions required by the patent claims; that he did not establish any utility in either the intermediate iminos or in the ultimate amides; and that the tests conducted prior to the key date were limited to a relatively few compounds and did not encompass the broad generic classes claimed in the patent. We first consider whether Ritter did perform his process, how it operated, and what it produced. Ritter testified, and his notebook shows, that before the key date (i. e., before page 259 in his notebook), he did run a series of tests reacting the three components of his patented process —five different nitriles, six different ole-fins, and sulfuric acid (Tr. 158-163, 171, 197-98, 205-08; PX 12, pp. 237-259). Those tests proved that five important assertions, which later appeared in his patent, were not just theoretical conceptions but operative facts. Thus, the tests demonstrated that: (1) the three components reacted readily at room pressure and temperature (Tr. 159; PX 12, pp. 247, 251, 253-54) and with 92% sulfuric acid (PX 12, pp. 252-53), as well as with 98% sulfuric acid (Tr. 159-163, 186-87, 222-24; PX 12, pp. 251, 253); (2) the reaction resulted in a tangible intermediate product in the form of a viscous liquid (PX 12, pp. 251, 254) from which a crystalline deposit eventually formed (Tr. 187, 241; PX 12, pp. 256, 270); (3) the intermediate product was stable because after it was formed the reaction could be interrupted or stopped and the intermediate could be left standing overnight or over a weekend (Tr. 176, 188-190, 242-44, 648-49; PX 12, pp. 238-240, 254, 256); (4) the intermediate, in either its liquid or crystal form, would readily hydrolyze to the same end product, whether hydrolyzed at once (PX 12, pp. 237, 251-52) or left standing and hydrolyzed later (Tr. 244, 649; PX 12, pp. 238-240, 254, 256) ; and (5) the end product resulting from the hydrolysis of the intermediate was identified as an kimono substituted amide (Tr. 187, 236-37, 272; PX 12, pp. 243-46). Defendant offers no evidence to refute either the tests or their observed results. We find, therefore, that before the key date Ritter did practice his invention by laboratory runs of his process and demonstrated that the three components of his patent did react under the condition specified in his patent and did produce an intermediate product which could be hydrolyzed immediately, or separated, stored and hydrolyzed later, to form an amide. The question is thus reduced to whether the intermediate product was an N-mono hydrocarbon substituted imino compound or some other substance. It is clear that Ritter did identify the “ultimate” product as an N-mono substituted amide by analysis and determination of its molecular weight (Tr. 236-37, 243-46). Identification of the ultimate amide, however, does not establish that the intermediate product was an imino compound, for Ritter admitted that many products hydrolyze to an amide (Tr. 672-73). Nonetheless, Ritter’s identification of the ultimate amide did limit the class of possible intermediates to those which hydrolyze to an amide and clearly warranted the deduction that an imino was a likely intermediate since, as even Bortnick admitted, “a hydrolysis step is required to go from a structure of the imino type to the amide structure” (Tr. 954). Dr. Mark testified that in 1944 Ritter could have isolated the intermediate compound, washed it and performed a series of standard chemical tests which would have revealed its properties (molecular weight, boiling point, etc.) and have enabled him to determine the chemical structure of the intermediate conclusively (Tr. 321-28, 334). Without such test, according to Mark, it was impossible to exclude other compounds as the intermediate of the reaction (Tr. 271-282, 288-292, 302, 321-23, 334). Dr. Meyers testified that such tests were, and still are, impossible to perform because the imino intermediate cannot be isolated in a purified form (Tr. 756-763, 820-21, 874). Scientific studies agree (Tr. 306-OS). We find support for Meyers’ view in the facts that Mark, who was retained by defendant seven months before the trial, never conducted such tests on this crucial issue, although he said they could be done in two weeks (Tr. 328, 333). Moreover, there is no suggestion that Bortnick, or anyone else, ever isolated or identified Ritter’s intermediate by such tests. We conclude, therefore, that Meyers is right. In any event, there is no dispute that Ritter did not isolate and identify the intermediate compound by such tests. Indeed, he admitted that he was not interested in performing any tests for that purpose because he was only interested in getting the ultimate product, the amide (Tr. 189, 671). We turn, then, to whether he did correctly identify his intermediate as an imino compound by other means. Mark testified that Ritter’s imino compound was only one of many possible compounds which might be produced by the reaction, but he gave the formulas for only four others, which he said were equally speculative possibilities (Tr. 275, 278-282; PX 16). Ritter and Meyers testified, in substance, that no other intermediates were probable or rational, that the starting materials and products similar to the resulting products were well-known compounds, that the steps carried out were of a type generally known, that the reaction could therefore be predicted with reasonable accuracy, and that modern analytical equipment, not available in 1944, has proven Ritter right (Ritter, Tr. 668-673; Meyers, Tr. 587). In order to determine whether the intermediate was an imino compound or some other substance, we must appraise the rational likelihood of the other possibilities postulated by Mark. We begin by noting Mark’s failure to give us a mechanism or other explanation of why any of his four other postulates are rational or likely. In the absence of such an explanation, they are all highly suspect at best. We add to this the fact that Ritter did give a mechanism showing the formation of his imino, and Mark’s concession that the imino postulated by Ritter conforms to the laws of organic chemistry and is therefore rational and possible. We turn, then, to the question of whether Mark’s other four postulates are equally rational and possible. We deal first with formulas III, IV and V. According to a principle of organic chemistry, molecules tend to combine in the easiest and most direct way and to produce the simplest product. It was for this reason that Dr. Mark conceded that III, IV and V were less likely than II, which he preferred, because it is a direct addition product (Tr. 329). It is plain that IV and V are not direct addition products but very complex cyclic compounds having huge molecular weights. They have never been reported in the literature. Dr. Mark could not even give them a name (Tr. 279-281). It seems to us, therefore, that if IV and V exist at all, which we doubt, they represent a far more indirect and complicated reaction mechanism and intermediate product than would seem probable, especially in the absence of any rational explanation by Mark. We, therefore, reject them. Dr. Mark testified that all of his postulated intermediates would hydrolyze to an amide (Tr. 276-77). Meyers testified on cross-examination that III would hydrolyze to an amide but only under certain conditions, but that “it would not give an amide sitting in water under the conditions that the Ritter intermediate is put into water” (Tr. 769). The issue here is not whether formula III will ever hydrolyze to an amide, but whether it will hydrolyze to an amide under the conditions of Ritter’s experiments. Meyers’ specific answer, therefore, is more pertinent and persuasive than Mark’s general assertion upon which defendant has chosen to rest, for despite an opportunity to contest Meyers’ specific testimony by its later witness, Dr. Bortnick, defendant failed to contradict Meyers on this point. Instead, it compounded Mark’s generalization by Bortnick’s testimony that III would convert to a salt in water, not that it was unstable in water under Ritter’s conditions (Tr. 930-31). When we add this to the fact that Mark preferred II to III, we conclude that on balance III was, and is, a highly irrational and improbable intermediate. Further evidence that Mark’s formulas III, IV and V are not rational possibilities is found in the prior teaching of Whit-more and Hanztsch. Meyers testified, without contradiction, that Professor Whitmore taught in 1932 that olefins combine with sulfuric acid to create a carbonium ion: and Hanztsch taught in 1931 that the reaction of a nitrile with sulfuric acid produces a nitrilium salt in an ionized form: Meyers drew the mechanism of Hanztsch’s reaction (PX 19) and testified that Ritter conceived that the carbonium ion could be substituted for Hanztsch’s hydrogen ion. Meyers explained that the carbonium ion performs exactly the same function as Hanztseh's hydrogen ion, that is, it remains intact and is singly bonded to the nitrogen atom. That this is so is demonstrated not only by Ritter’s equations but also by all four compounds postulated by Mark in plaintiff’s Exhibit 16, for in each of them the carbonium ion remains intact and is singly bonded to the nitrogen atom. Since the carbonium ion remains intact, it seems most likely that Ritter’s reaction proceeds along the same mechanism as Hanztsch’s reaction and produces the same generic product, a nitrilium salt (Tr. 568). We are convinced from this array of evidence that Ritter’s experiments in 1944 did not produce III, IV and V, but instead produced either I or II. Mark testified that II was well known in the literature (Tr. 326). Meyers testified that I and II were the same thing. The basis for his opinion was that, although I and II appear to have a different structure, actually II, the nitrilium salt, is a resonance hybrid of I, the imino. Referring to the formulas shown in the margin, he said that “to an organic chemist, these are identical structures * * *. Any chemist * * could have drawn any one of these three structures * * * and any organic chemist would have known what that meant if these structures were put in the patent” (Tr. 572). This statement was not even averted to in extensive cross-examination; nor was Meyers directly challenged on this point by other witnesses. Bortnick, in his direct testimony, did not refute Meyers’ testimony that I and II are resonance hybrids. Rather, first he evaded, stating that II was a nitrilium salt and I an imino compound (Tr. 948). Meyers, of course, would agree, but the issue was whether they were resonance hybrids. Later, in response to the court’s questions, he tended to agree with Meyers (Tr. 997-99). Mark drew II as -N fp C and identified it as a nitrilium salt. Mark testified that Ritter, at page 245 of his notebook, drew his imino in its ionized + form, -N = C (Tr. 316-17). Asked on cross-examination whether II was a resonance hybrid of the ionized form of I, Mark replied, “I don’t think so” (Tr. 319). He gave no explanation whatever for his opinion. However, he did agree with Meyers that resonance hybrids were “twins” (Tr. 318). Meyers, in Exhibit 19, gave a somewhat different formula for a nitrilium salt as: Bortniek corroborated Meyers’ formula for the nitrilium salt. “[T]hese compounds have been shown as N triple bond C with a plus charge on nitrogen, they have been shown as N double [bond] C with a plus charge on carbon * * *. [I]t has been represented both ways and it is understood by chemists to mean that it is neither of those but some intermediate structure” (Tr. 998-99). Bortnick drew a nitrilium salt as (DX AA). Bortniek testified that the three dots on top of the double bond (-N = ' CH) “are meant to show the same kind of phenomenon,” namely a resonance hybrid (Tr. 999). We find that Dr. Mark’s formula for the nitrilium salt is technically inaccurate, that nitrilium salts are resonance hybrids, and that the only formulas which correctly reflect this phenomenon are those of Dr. Meyers and Dr. Bortniek. We further find that an organic chemist would understand that a nitrilium salt is chemically identical with Ritter’s imino sulfate. Ritter testified that his experiments showing that his intermediate hydrolyzed to an amide, taken together with the well-known nature of the antecedent reagents and the sound, published “body of fact and theory bearing on the situation,” notably “the monumental work of Arthur Hanztsch bearing on the behavior of nitriles in strong acids,” led him to the “only reasonable conclusion that one could draw with respect to the structure of the compound,” viz., “the unequivocal conclusion”. that it was an N-mono hydrocarbon substituted imino sulfate (Deposition, Tr. 668-673). Meyers agreed (Tr. 559-569). Neither Mark nor Bortniek were asked about the background teaching of Whitmore or Hanztseh. Nevertheless, the soundness of Ritter’s conclusion is reinforced by the fact that Bortniek, himself, fully aware of the phenomenon of resonance hybrids and before any prospect of this litigation arose and with no motive to obfuscate, formulated the product of the reaction precisely as did Ritter (PX 28). So, too, did other scientists working independently (Tr. 306). That Ritter’s postulation of the product was sound is further buttressed by Bortniek who recognized that an “hydrolysis step is required to go from the structure of the imino type to the amide structure” (Tr. 954) and that an imino compound “is generally supposed to be an intermediate in all hydrolysis reactions of nitriles and salts of this type” (Tr. 958). It stands uncontradicted that the N-mono hydrocarbon substituted imino compound is referred to in the scientific literature as the Ritter intermediate (Tr. 193). Other hydrolysis reactions of nitriles were known in 1944. The evidence is clear that the starting materials and products similar to Ritter’s ultimate products were well-known compounds, the steps carried out were of a type which were generally known in the art, and had been previously performed on similar materials, e. g., by Hanztsch. In such circumstances, it was not necessary for Ritter to isolate, purify and identify his intermediate by chemical tests and analysis of the kind urged by defendant in order to reduce his invention to practice. “The reactions to be obtained could, therefore, be predicted with a reasonable assurance of accuracy and under such circumstances it is not necessary that the proof of identity of the products be as exhaustive as if entirely new substances or procedures were involved.” Reiners v. Mehltretter, 236 F.2d 418, 421, 32 CCPA 1019 (1956); Guinot v. Hull, 204 F.2d 281, 40 CCPA 982 (1953). Finally, test techniques which postdate Ritter’s invention tend to establish that his postulation was indeed accurate and that he did produce his intermediate on or prior to the key date. Plaintiff contends that the imino compound can be conclusively identified today by the infrared and ultraviolet spectra and the nuclear magnetic resonator. Meyers testified that “we have taken the infrared spectrum of the imino compound, and there is no doubt that it has a very prominent peak in the infrared corresponding to the carbon nitrogen double bond and this was confirmed by many others besides myself” (Tr. 578-582, 587). Meyers’ testimony on this point was in no way impeached on cross-examination (Tr. 763-64). Defendant introduced selected excerpts of Meyers’ deposition into evidence. There Meyers testified that he could not distinguish the infrared spectrum of the imino from that of a protonated amide since, according to Meyers, they both have an N double bond C (Tr. 823). There is nothing in the record, however, from which we can determine whether he was then referring to the product of the reaction where the concentration of the sulfuric acid was 88% or 96%, and without such evidence his deposition testimony is meaningless. In any event, there has not been the slightest suggestion from any witness called by the defendant that a reaction using 96% to 100% sulfuric acid ever produced a protonated amide. Ritter said it would not (Tr. 653), and, significantly, Mark, who postulated “many” possible compounds, never gave a protonated amide in any of his examples. Further, Glikmans et al. have conducted independent research with the Ritter reaction, and they claim that they have isolated the Ritter intermediate and identified it as an imino compound (PX 18). There is earlier literature to the contrary, but the Glikmans study is the most recent (1966) and employs the most modern instruments for determining the presence of organic bonds. We conclude, therefore, that these post-invention date tests do corroborate the fact that Ritter did produce an N-mono hydrocarbon substituted imino compound prior to December 4, 1944. The use of such post-invention techniques as an aid to the court in resolving the existence or non-existence of a physical phenomenon of the type involved here has been expressly approved in this circuit. Helene Curtis Industries, Inc. v. Sales Affiliates, Inc., 233 F.2d 148, 154-155 (2 Cir.), cert. denied, 352 U.S. 879, 77 S.Ct. 1011 (1956). Accordingly, we find that Ritter did in fact perform his process and produce an N-mono hydrocarbon substituted imino compound by reacting the three components of his patent under its specified conditions before December 4, 1944. The mere fact that Ritter’s process successfully produced the N-mono hydrocarbon substituted imino compound does not establish reduction to practice. Before the key date Ritter had to demonstrate that this imino had utility. Reiners v. Mehltretter, supra, 236 F.2d at 421. “The inventor need prove only one practical use” to reduce his invention to practice. Conner v. Joris, 241 F.2d 944, 947, 44 COPA 772 (1957). “But in order to establish utility of a product it is not necessary to show that it can immediately and without change perform a useful function. Products are useful if they serve as starting materials or intermediates in producing other materials or articles which are directly useful.” Reiners v. Mehltretter, supra, 236 F.2d at 421-422. N-mono hydrocarbon substituted imino compounds serve as intermediates in producing N-mono substituted amides. Several uses of N-mono substituted amides are described in Ritter’s initial application, filed with the Patent Office on January 27, 1945 (DX H, p. 6). Those uses, and others, appear in Ritter’s patent (PX 1, col. 4, lines 71-75; col. 5, lines 1-16). The first use given is that: “Some of the compounds herein described, for example, N-oetyl acetamide and N-isobornyl acetamide are useful substitutes for camphor in the plasticizing of nitrocellulose.” N-octyl acetamide is produced by the patent’s first four examples. Before the key date, Ritter performed the same process as Example I of the patent (PX 12, pp. 240, 246). Then he wrote, “I believe that these amides might have industrial value along several lines, as plasticisers, etc.” He drew the formulas for camphor and for N-octyl acetamide and noted, “Similar molecular architectures?” (PX 12, p. 246.) On November 22, 1944, at page 255 of his notebook, Ritter ran the same process again and then proceeded to demonstrate that N-octyl acetamide is a useful substitute for camphor in the plasticizing of nitrocellulose: “This was made for skin casting with NC dope, in collaboration with Mr. C. Snead. Skin-cast seems better (at least softer, more plastic) than camphor skin made at same time. Both were heated 7 hrs on steam radiator and the ‘Camplex’ sheet still showed up good as new. Will make 300 g to knead a block for skin shaving at Joe Davis plant, Arlington, N. J. Promised this next week.” Thus, before the key date, Ritter demonstrated that an N-mono substituted amide was directly useful. Accordingly, his imino intermediate was useful and so was his process. Defendant also claims that some of Ritter’s experiments did not employ the equal molecular amounts required by the patent claims. The statement is true insofar as it refers to experiments conducted before he conceived his invention, at page 245 of his notebook, on November 6, 1944. There is no question that at page 245 Ritter does express his formula in molecular proportions, as defendant’s expert Mark conceded (Tr. 545-46). Moreover, there can be no question that the notebook experiments, from November 6, 1944 through December 4, 1944 and afterward, employ molecular proportions, except where Ritter makes clear that he is deliberately experimenting with an excess; for example, PX 12, page 253. The mere fact that Ritter continued to experiment after he made his discovery of the need for molecular proportions in no way negatives invention. Moreover, the patent teaches that excesses can be used but usually are not because of economic reasons (PX 1, col. 2, lines 40 to 54). In the light of the evidence, this phase of defendant’s argument is wholly untenable. Finally, defendant contends that Ritter’s tests were limited to a relatively few compounds and did not encompass the broad generic classes of reaction components claimed in the patent. There can be no question that Ritter’s experiments were limited to a relatively few compounds and that each of the components of the patented process comprise numerous compounds running into the thousands (Tr. 771-73). Nonetheless, the evidence is clear that, insofar as pertinent to this reaction, all of the members of a class of each of the components have common properties and molecular structures which are critical to the reaction (Tr. 583-87). Against this, defendant has not named a single' member of any of the three classes of compounds that was, or is, inoperative. Once Ritter conceived the generalized imino formula, a skilled chemist would have been satisfied that all known nitrile radicals and all known olefin radicals would work in their respective “R” positions, because nitriles and olefins were well known to be homologous series (Tr. 583-86), that is, they have common qualities. When chemical compounds have a common quality rendering each useful in the process patented, tests showing the operativeness of every member of that class are not required. We, therefore, conclude that plaintiff did conceive his invention and reduce it to practice before December 4, 1944 and that, therefore, French Patent No. 902,342 does not anticipate Ritter. We also conclude, for the same reasons, that Ritter’s invention was, and is, operative. We turn, then, to the other prior art references. (2) Other Prior Art A patented process must be “new and useful.” 35 U.S.C. §§ 101, 102. Anticipation is a strictly technical defense. “Sections 101 and 102 prevent patentability only where the invention was ‘identically disclosed’ by the prior art.” The mere fact that a process is “new and useful” does not mean that it is patentable; “[a] patent may not be obtained * * * if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to persons having ordinary skill in the art to which the subject matter pertains.” 35 U.S.C. § 103. In order to resolve the ultimate legal question, the court must make three preliminary factual determinations: (1) the scope and content of the prior art; (2) the differences between the claims and the prior art; and (3) the level of ordinary skill in the pertinent art at the time the invention was made. In making these factual determinations, the court must guard against stepping into use of hindsight and resist the temptation to read into the prior art the teachings of the invention in issue. In order to filter out the hindsight which permeates much of the testimony, we must undertake the task of mining the facts respecting the state of the prior art as of 1944 from the texts of the references themselves. We shall be forced to go into much technical detail, for the art prior to Ritter was not clear and generalized, but obscure, confused, and fragmented. (a) The Wieland and Dorrer Article Defendant’s second prior art reference is an abstruse scholarly paper by Heinrich Wieland and Eugen Dorrer of the Bavarian Academy of Science, published in 1930 (DX I, Tab 8). Parts of this paper contain uncanny hunches. Parts leave blatant contradictions unexplained and would lead a reader down many dead ends. Other parts are so plainly wrong that a skilled chemist would not find them heuristic at all. Wieland and Dorrer were exploring Friedel-Crafts chemistry, which involves using aluminum chloride (AICI3) and hydrogen chloride (HC1) in a variety of reactions (Tr. 455-56; Roberts-Caserio, Modern Organic Chemistry 548 (1967)). The experiments involved here were a continuation of two of their earlier papers (see DX I, Tab 8, n. 1), which had established that enols, in the presence of aluminum chloride and hydrogen chloride, combine with hydrocyanic acid (HCN) to form aldimines. Now they were trying to demonstrate that non-aromatic olefins can combine with hydrocyanic acid (which is a nitrile) under the same conditions. They used atmospheric pressure and cool (freezing) temperature, as Ritter often does. Wieland and Dorrer ran this reaction with only four different olefins: cyclohexene, asymmetric diphenyl ethylene, stilbene, and styrene. They were trying to demonstrate that the four olefins would react the same way. Instead, each one reacted differently. (i) Wieland and Dorrer’s Reaction with Cyclohexene This reaction came closest to the process which Ritter patented. Wieland and Dorrer reacted cyclohexene with HCN and HC1 in the presence of the catalyst A1C13 and the solvent benzene. The resulting “batch” was left standing for six hours. Then an excess of ice was added. Then one of the products was extracted and distilled (DX I, Tab 8, pp. 7-8). Chemical analysis showed that each molecule contained 7 carbon atoms, 13 hydrogen, 1 oxygen, and 1 nitrogen (DX I, Tab 8, p. 7). Wieland and Dorrer, at page 2, formula VII, formulated this compound as: This is a species of the N-mono substituted amides, the genus taught, but not claimed, by Ritter’s patent (Tr. 518, 881). Their general formula is: Wieland and Dorrer called this compound “N-formyl cyclohexyl amine.” Since this label is an alternate way to express “N-mono eyclohexyl substituted formamide” (May 12, 1967) Tr. 34-37; see PX 12, pp. 254-55), a skilled chemist would have understood that Wieland and Dorrer’s product was an N-mono substituted amide (Tr. 518, 622, 881). Wieland and Dorrer postulated that this amide had been preceded by cyclohexyl formimido (hydro) chloride, which is a species of the genus which Ritter later called the “N-mono hydrocarbon substituted imino compounds” (Tr. 520). They were unable to isolate this imino intermediate. They apparently thought that the isolation was prevented only because of the Aids (DX I, Tab 8, p. 3). (ii) Wieland and Dorrer’s Reaction with Asymmetric Diphenyl Ethylene The same reaction was run, except that asymmetric diphenyl ethylene was used as the olefin. This is a conjugated olefin. Hence, it is not encompassed by Ritter’s claims However, it presumably would work in the Ritter process. In Wieland and Dorrer’s process, after the ice hydrolysis a product was formed. Wieland and Dorrer analytically demonstrated that this product was alpha, alpha-diphenyl propionitríle (DX I, Tab 8, p. 2, formula IV). This is a nitrile. Wieland and Dorrer, because of their previous work with enols, had expected to get chloroaldimine (DX I, Tab 8, p. 2, formula II). On page 2 they speculated that formula IV had been preceded by chloroaldimine and that chloroaldimine had been preceded by formimide chloride (no formula given). Neither chloroaldimine nor formimide chloride is an N-mono hydrocarbon substituted imino, compound. Chloroaldimine is a C-mono chloro-hydrocarbon substituted imino compound (Tr. 607). Formimide chloride is an unsubstituted imino compound (Tr. 606). Nevertheless, both are fairly close to Ritter’s intermediate. Formimide chloride should have been formed in this reaction, and it should have been followed by some kind of substituted imino compound (Tr. 622, 881). However, we are convinced that no imino of any kind was formed in the reaction as Wieland and Dorrer performed it. Although this part of Wieland and Dorrer speculates almost the same intermediate as Ritter, the end product of this experiment is completely different from the amide taught by Ritter. Wieland and Dorrer probably felt that their imino intermediates were formed before the ice hydrolysis, although they do not clearly say so. Therefore, they are saying that in this particular experiment the hydrolysis rebuilds the carbon-nitrogen double bond back again to a triple bond! Ritter, Hanztsch, and all the other evidence in this case teach that hydrolysis breaks the carbon-nitrogen double bond down further into a single bond. We are convinced that Wieland and Dorrer’s experiment with asymmetric diphenyl ethylene, which indisputably produced a nitrile, did not go through any imino intermediate, although it should have. This was simply one of those experiments that “go haywire.” Nevertheless, Wieland and Dorrer “correctly” speculated the iminos which should have occurred. Another fascinating aspect of their speculation is that they realized what Gresham later failed to realize, namely, that the acid was entering the reaction (DX I, Tab 8, p. 2). This was hard to realize because the negative ion of the acid was attaching and then splitting off. However, both of these remarkable speculations about the iminos and the acid were based on knowledge which was peculiar to chlorine chemistry, and a skilled chemist would have seen that they were. There is no evidence that anyone before 1944 could have applied this knowledge to speculations about sulfuric acid reactions. (iii) Wieland and Dorrer’s Reaction with Stilbene Stilbene is the “trivial” name for the symmetric isomer of asymmetric diphenylethylene; it is less reactive because its double bond is in a less exposed position (DX I, Tab 8, p. 5). Like its isomer, it is conjugated and is not encompassed by Ritter’s claims. In Ritter’s process, stilbene presumably would combine with the nitrile (HCN). In Wieland and Dorrer’s process, stilbene failed to combine with the nitrile. It simply dimerized, i. e., combined with its own molecules. (iv) Wieland and Dorrer’s Reaction with Styrene Styrene is another conjugated olefin. However, it worked very well in Ritter’s process, and Ritter claimed it specially in claims 11, 12, and 15. In Wieland and Dorrer’s process, the styrene failed to combine with the nitrile. It simply polymerized, i. e., combined with its own molecules. (b) The Legal Effect of Wieland and Dorrer Upon Ritter’s Patent (i) Anticipation The crucial issue under Wieland and Dorrer is obviousness. However, we shall first deal with the issue of anticipation. This issue was raised obliquely by some of defense counsel’s questions, which implied that Wieland and Dorrer’s HC1 + AICI3 directly anticipates Ritter’s H2SO4 because both are cationoid compounds. Anticipation, unlike obviousness, looks only to complete units and therefore considers references separately, one by one. It is hornbook patent law that if a reference would have infringed, if later, it anticipates, if earlier. Wieland and Dorrer’s process, if later, would not have infringed the Ritter patent. Their HCN would read on Ritter’s claims (“a nitrile,” “hydrogen cyanide”). Two of their four olefins would read on Ritter’s claims. However, their HC1 + AICI3 would not read on Ritter’s claims: “a cationoid compound from the group consisting of sulfuric acid, hydrocarbon sulfonic acids and alkyl sulfuric acids.” (Claims 1, 2, and 3.) Moreover, all of Wieland and Dorrer’s imino compounds are imino chlorides. None of them would read on Ritter’s claims. The “N-mono hydrocarbon substituted imino compounds” in Ritter’s claims are clearly limited to N-mono hydrocarbon substituted imino bisulfates and sulfonates, because Ritter limited his cationoid compounds to acids whose negative ions are either OSO3H or SO3. Defendant argues (May 12, 1967 Tr. 40-42) that if Wieland and Dorrer had been later than Ritter, and if Ritter had sued them for infringement, Ritter would have won on the doctrine of equivalents. However, the doctrine of file wrapper estoppel prevents resort to those “equivalents” which were given up in the patent application proceedings. Ritter’s original application tried to claim “a cationoid compound” (DX H, p. 15), and Wieland and Dorrer’s HC1 + AICI3 would have read on that. However, Ritter gave up all but the subge