This patent appeal is another illustration of the absurdity of requiring the decision of such cases to be made by judges whose knowledge of the relevant technology derives primarily, or even solely, from explanations by counsel1 and who, unlike the judges of the Court of Customs and Patent Appeals, do not have access to a scientifically knowledgeable staff.

The case is unusual in several respects. The chemical patent in suit, Heiss Patent No. 3,102,875, of which defendant Jefferson Chemical Company, Inc. (Jefferson) is the assignee, was pending in the Patent Office for more than ten years-- from August 7, 1953 to September 3, 1963.2 During the period there had issued, after a mere five years of processing, Price Patent No. 2,866,774,3 of which plaintiff The General Tire & Rubber Company (General) is the assignee, for what at least for present purposes may be considered, cf. p. 1287 & n. 12 infra, to be the very invention here contended to have been made by Heiss. The claims allowed and held to be valid were copied, except in one particular hereinafter noted, by Heiss' solicitors from the Price patent, in the first instance so as to provoke an interference; as will later appear, there is no convincing evidence that Heiss was aware of the invention now asserted until the Price patent was published.4 Jefferson says that all this is immaterial, except perhaps for Heiss' unawareness of what he had done which it feebly challenges, and responds with a tuquoque, namely, that General, which now claims the Heiss patent was invalid for anticipation and obviousness, has collected huge royalties on the Price patent, now shortly to expire, to which many of its arguments of invalidity would be equally applicable.5 With this background, we briefly state the case, referring to Judge Carter's opinion in the District Court for the Southern District of New York, 363 F. Supp. 871 (S.D.N.Y.1973), for an ampler development.

The action was brought by General for a declaration of the invalidity of certain claims of the Heiss patent on various grounds, of which we need consider only 102(e) (anticipation), 102(f) (patentee not the inventor), 103 (obviousness), and 112 (inadequate disclosure). Jefferson counterclaimed for a judgment of validity and infringement. General conceded infringement if the challenged claims were valid. Judge Carter held that they were and directed an injunction and an accounting.6

Heiss' alleged invention concerns polyurethanes. Such products are made by mixing a polyol, i.e., an alcohol with molecules containing more than one hydroxyl group,7 with an isocyanate, an organic compound containing an NCO group (nitrogen, carbon and oxygen) in its molecule.8 To insure the presence of enough NCO groups for the complete reaction, an excess of isocyanate is added. The polyurethanes claimed in the Heiss patent-- both by direct claims to the products and by claims to the processes of their manufacture-- are extensively used for foam matresses, seat cushions, insulation, adhesives and protective coatings. The parties agree that we may take as representative Claim 3 of the Heiss patent which reads:

(1) a condensate of propylene oxide and a polyhydric alcohol having from 3 to 4 hydroxyl groups, and (2) an organic polyisocyanate, the amount of (2) being in excess of that theoretically required to react with the reactive hydrogens of said condensate.

The foundation of the polyurethane industry was laid in Germany in the 1930's and 1940's as a result of the researches of Dr. Otto Bayer, who taught cross linkage by use of diisocyanates so as to improve the durability of polyurethane products; see fns. 7 and 8. An extensive article entitled 'Industrial Application of the 'Diisocyanates" appeared in a German chemical journal in 1947. However, Dr. Bayer's alcohol components were esters, the result of reacting an alcohol with an organic acid, and, as found by the district court, 363 F. Supp. at 878, 'polyester urethanes were subject to destruction by hydrolysis.'9 Continued work on the problem by Dr. Bayer and his associates resulted in the Windemuth patent No. 2,948,691, which was applied for on May 6, 1952 and was granted on August 9, 1960.10 The Windemuth patent involved the mixing of isocyanates with ethers, the result of reacting an alcohol with an alkylene oxide, rather than with esters, as Bayer had previously done. The specifications disclosed that 'the reaction of polyglycolethers and aromatic or aliphatic monoisocyanates-- depending on the molecular Weight of the polyglycolether and the nature and quantity of the isocyanate-- gives rise to products which are either insoluble in water, however, soluble in organic solvents or which dissolve in cold water but not in hot water.' They stated also that 'Products of considerable importance are obtained by reacting polyfunctional polyglycolethers and polyfunctional isocyanates in quantities larger than those required for saturating the hydroxyl groups present.' Claim 2 of the Windemuth patent was for:

2. A process for making polymers which comprises reacting a polyalkylene ether of a saturated aliphatic hydrocarbon polyol, said polyalkylene ether having at least two hydroxyl groups per molecule and a molecular weight of at least about 500, and an organic isocyanate containing only hydrocarbon radicals and NCO.

6. The process of claim 2 wherein the reaction mixture contains an excess of said isocyanate over that required to react with all of said hydroxyl groups.11

The district court correctly characterized the advances made by the Windemuth patent as follows:

Its great advance over prior teaching was in the making of polyether (as opposed to polyester) based urethanes. Ethers are sensitive to hydrolysis but are not destroyed by water, as are esters. Moreover, here the diisocyanate ingredient necessary for cross linkage is not required in such large quantities so that the resultant product is less expensive to produce.

363 F.Supp. at 878. Admittedly Claim 6 of Windemuth includes Heiss' claims; the only relevant difference is that Heiss copying from the Price patent, singled out as a condensant propylene oxide, whereas Windemuth used for the condensate the generic term 'polyalkylene ether', which also included ethylene oxide as a condensant.

The specifications in Heiss' application, which were substantially the same as in the patent as issued, describe the invention as relating 'to a new series of resins, compositions of matter containing same and to the method of preparing such products.' They list a great variety of 'suitable hydroxy compounds for use in the production of organic compounds containing urethane groups,' among which polypropylene glycols are only one, and an equally large array of suitable isocyanates. All claims were rejected as unpatentable over various cited references, not including the Windemuth patent which had not yet issued. Three amendments met the same fate.

While a fourth amendment was pending, Patent No. 2,866,774 was issued on December 30, 1958, to Dr. Charles C. Price. This described an invention relating 'to rubbery reaction products of organic di-isocyanates and high molecular weight polyglycols which have unusual physical characteristics.' The specifications referred to the difficulties that had been encountered with polyesters due to their tendency to hydrolysis, the preferability of polyethers from this standpoint, and certain problems that had been encountered with them. Price said he had found 'that polymeric materials built with alkylene oxide units having at least three carbon atoms, such as propylene oxide units, are far superior to those built with ethylene oxide units because the resultant rubbery product has improved water resistance and has less tendency to crystallize.' Reading of the Price patent by Heiss' solicitor led to a fifth and then a sixth amendment of the Heiss application in which certain claims of the price patent were copied in order to provoke an interference. After proceedings which are set forth in the district court's opinion, 363 F. Supp. at 876-877, and are here unnecessary to detail, the Board of Interference decided in favor of Price. This, however, was not on the ground of prior conception and reduction to practice but on the basis that his patent was dependent on a molecular weight for the condensate of at least 600 whereas the Heiss claims had no such limitation. Thereafter, through the mysterious processes of the Patent Office, the Heiss patent issued on the claims limited to propylene oxide essentially as copied from Price, but without limitation concerning molecular weights12 and with Windemuth cited as prior art.

The arguments take us back to a territory we reconnoitered not long ago in Carter-Wallace, Inc. v. Otte, 474 F.2d 529 (2 Cir. 1972), cert. denied, 412 U.S. 929, 93 S. Ct. 2753, 37 L. Ed. 2d 156 (1973). Indeed, General contends that the Carter-Wallace decision, where we held the patent for meprobamate ('Miltown') was invalid because the application did not adequately disclose the effectiveness of Miltown as a tranquilizer and because its utility as an anti-convulsant was obvious, is controlling in its favor. Putting aside for a moment the question of obviousness in the instant case, we would find it somewhat bizarre that a 17-year monopoly can be gained by copying claims from a patent issued to another half way during the ten-year life of an application in the Patent Office, when nothing in the specifications, then or even later, disclosed that Heiss had any idea that use of propylene oxide to condense the polyol would yield superior results. However, we prefer to place decision on the closely related grounds that the Heiss patent is invalid under 103 (without reliance on Carter-Wallace) and, to the degree necessary to supplement this, 102(f).

As has recently been written,

Perhaps no area of patent law is more uncertain than that concerning the application of the section 103 test of 'nonobviousness' contained in the Patent Act of 1952 to chemical compounds which are conceded to meet the other two patentability requirements of 'novelty' and 'usefulness.'

Note, Standards of Obviousness and the Patentability of Chemical Compounds, 87 Harv.L.Rev. 607 (1974). Conflicting views have been expressed on the question whether a compound which is structurally obvious on the basis of the prior art is patentable as a composition of matter because it is discovered to have an unexpected useful property. In re Papesch, 315 F.2d 381, 50 CCPA 1084 (1963), and Commissioner of Patents v. Deutsche Gold-und-silber-Scheideanstalt Vormals Roessler, 130 U.S.App.D.C. 95, 397 F.2d 656 (1968) (Burger, J.), are the leading authorities answering that question in the affirmative. Mansanto Co. v. Rohm & Haas Co., 312 F. Supp. 778 (E.D.Pa.1970), aff'd on other grounds, 456 F.2d 592 (3 Cir.), cert. denied, 407 U.S. 934, 92 S. Ct. 2463, 32 L.Ed.id 817 (1972), and even more forcefully the district court's decision in Carter-Wallace, supra, 341 F. Supp. 1303 (E.D.N.Y.), argue that in such cases only a use patent should be available. Here, as in CarterWallace, we find it unnecessary to choose between the two polar views. Assuming without deciding that some variation of Papesch and Deutsche Gold-und-Silber-Scheideanstalt represents the correct view,13 as we did in Carter-Wallace, we think Heiss' claim must fail.

Before going further, it is desirable to state and justify our conclusion that the Windemuth patent is prior art. Although Jefferson claims the district court left this issue open, Judge Carter's opinion clearly held Windemuth to be prior art, over a contrary contention by Jefferson which was based on Heiss' allegedly having conceived his own invention and allegedly having reduced to practice the teaching of the Windemuth patent before the Windemuth application was filed,14 363 F.Supp. at 878 n. 12. The ground for the court's ruling was that Heiss had made no such claim before the Patent Office. We are not at all certain we agree with that reasoning, for which no authority was cited; General's attempt to support it by analogy with the doctrine of file-wrapper estoppel, Leggett v. Avery, 101 U.S. 256, 25 L. Ed. 865 (1880); Goodyear Dental Vulcanite Co. v. Davis, 102 U.S. 222, 228, 26 L. Ed. 149 (1880); Keith v. Charles E. Hires Co., 116 F.2d 46 (2 Cir. 1940), is unpersuasive. But Judge Carter's conclusion was correct. The Rules of Practice of the Patent Office, 35 U.S.C.App. I, 37 C.F.R. 1.131(b), provide that in order to avoid a United States patent application cited as prior art, 102(e), 'The showing of facts shall be such, in character and weight, as to establish reduction to practice prior to the effective date of the reference, or conception of the invention prior to the effective date of the reference coupled with due diligence from said date to a subsequent reduction to practice or to the filing of the Application.' The Rule accords with the decision in Marconi Wireless Telegraph Co. v. United States, 320 U.S. 1, 35, 63 S. Ct. 1393, 1410, 87 L. Ed. 1731 (1943), in the context of an anticipation defense to an infringement action, that, in order for the alleged prior art to antedate the invention at issue on the strength of the conception date of the prior art, the artisan must have 'proceed(ed) with diligence to reduce it to practice,' cf. 35 U.S.C. 102(g). See also Scharmann v. Kassel, 179 F.2d 991, 996, 37 CCPA 903 (1950); Riche v. Permutit Co., 47 F. Supp. 275, 278 (D.Del.1942), aff'd per curiam, 135 F.2d 922 (3 Cir. 1943) (both standing for similar rule of law in context of interference proceedings). Jefferson made no showing, nor did it apparently appreciate the need to attempt a showing, of diligence. Although Heiss claimed a conception date of November 14, 1951,15 and although the date of Windemuth's U.S. application in May 6, 1952, Heiss does not assert reduction of the claimed invention to practice until September 1952.16

It is not seriously contested by Jefferson that, with the Windemuth patent as prior art, the use of propylene rather than ethylene oxide as the condensant was structurally obvious 'to a person having ordinary skill in the art to which said subject matter pertains,' 35 U.S.C. 103. General goes even further and contends with much force that such use not merely was structurally obvious but was actually described in Windemuth since that patent refers to 'alkylene oxides,' a term including propylene as well as ethylene oxide-- these being the two most prominent alkylene oxides.17 But even if we were to reject that argument in light of the stress that Windemuth placed on ethylene oxide, whose formula is C(2)H(4)O, it is still true that propylene oxide, whose formula is C(3)H(6)O, is the next higher homolog. If any structural change is obvious to one skilled in the art, a substitution of the next higher homolog would seem to be. In re Hass, 141 F.2d 122, 31 CCPA 895 (1944), In re Coes, 173 F.2d 1012, 36 CCPA 1067 (1949), and In re Henze, 181 F.2d 196, 37 CCPA 1009 (1950), although sharply criticized in various other respects over the years and indeed no longer good law in some, see, e.g., In re Stemniski, 444 F.2d 581, 58 CCPA 1410 (1971), tend strongly in that direction as to end products, and we see no reason for taking a different view with respect to as crucial a component as that here involved.

However, under the Papesch/Deutsche Gold-und-Silber-Scheideanstalt test, this is not fatal to the validity of the Heiss patent if the substitution of propylene for ethylene oxide developed a property unexpected to one with ordinary skill in the art and if all other conditions were met. It is true, as the district judge held, that Windemuth and his associates had not recognized the superior water resistant effects from using propylene oxide as the condensant, and both Jefferson and the district judge make much of a speech by Dr. Price in February 1961, How Chemists Create a New Product, emphasizing this and extolling his own discovery of the superiority of propylene oxide for that purpose. Nevertheless we would have considerable doubt of the correctness of the district court's conclusion of lack of obviousness even under the Papesch/Deutsche Gold-und-Silber-Scheideanstalt test, with the limitations we would deem necessary, see fn. 13 supra, and even if Heiss had himself recognized the advantage of the propylene oxide.

The criterion is not as the district judge said, 363 F. Supp. at 880, whether 'an ordinary journeyman would have understood that propylene oxide had these unexpected qualities in 1950 and 1951.' The proper test was suggested in Indiana General Corp. v. Krystinel Corp., 421 F.2d 1023, 1030-1031, (2 Cir.), cert. denied, 398 U.S. 928, 90 S.ct. 1820, 26 L. Ed. 2d 91 (1970), refusing to accord patentability to an improved ferrite, where Judge Medina observed that 'experimentation is of the essence of the art involved in this lawsuit' and that 'the practitioners of this art are, and of necessity must be highly educated, sophisticated persons who generally have at their disposal laboratory facilities and staffs of competent assistants.'18 As we noted in Carter-Wallace, Inc., supra, 474 F.2d at 543, citing Eli Lilly & Co. v. Generix Drug Sales, Inc., 460 F.2d 1096, 1103 (5 Cir. 1972), this, of course, is not to say that the development of an unexpected property is to be regarded as obvious simply because it is the result of patient work in the laboratory. But with Windemuth's patent as prior art, a try at the next higher homolog of ethylene oxide would seem to constitute experimentation which is 'on the level of ordinary skill in the art,' 421 F.2d at 1031.19 See also Note, supra, 87 Harv. L.Rev. at 626-28.

We find it unnecessary, however, to decide this question. The record does not present a picture of Heiss and his associates recognizing the deficiencies of ethylene oxide as a condensant and discovering the superior water resistant qualities of propylene, as the result either of diligent work or of a flash of genius, or as even appreciating that a new discovery had been made. The task that Heiss had set for himself was the discovery of resins which, as said in the specifications, 'may be plasticized and rendered more flexible by effecting their preparation in the presence of a non-reactive organic material, and preferably a non-reactive organic liquid which boils at a temperature not lower than 200 degrees C., and preferably not lower than 250 degrees C., at atmospheric pressure.' In addition, as Jefferson itself has pointed out to us, Heiss testified at trial that he was interested in carrying out the reaction in the presence of a liquid plasticizer with useful dielectric properties as a means to 'solidity' or 'gel up' the plasticizer and thereby embody the dielectric properties in a solid composition. Nothing in Heiss' notebooks or in the patent application as filed indicated a preference for propylene oxide. The notebook entry for November 14, 1951, the alleged date of conception,20 in listing the advantages of Heiss' experiments to date, states that ether linkages 'should be more stable than' ester linkages and lists as a further advantage that

3. The use of polyglycols (or other ethylene or propylene oxide condensates) offers unlimited variety in building up molecules of various degrees of linearity or non-linearity.

All this is taught by Windemuth. Although Heiss testified that he and his associates eventually found that using propylene rather than ethylene oxide improved water resistance, there is nothing to indicate that he regarded this as of particular importance and much to show that to the degree he took an interest in water resistance he regarded the water resistance of his products as being imparted by the presence of the plasticizer. When the specifications boast of water resistance, they refer to all the products covered, whether made with propylene or ethylene oxide. Ethylene oxide was used even in one specific example of how to make a product for which water resistance was presumably of importance. Jefferson's expert admitted freely and repeatedly under cross-examination that the specifications did not disclose a distinction with regard to water sensitivity between ethylene oxide and propylene oxide based products. And Heiss himself testified at trial that, at least insofar as he was working with the liquid plasticizer with the useful dielectric properties ('Aroclor'), supra, he had not looked into the relative water sensitivity of propylene oxide and ethylene oxide based urethanes 'because, for example, if either one of these resins was dissolved in something like Aroclor 50 per cent, which is very hydrophobic, I doubt very much that in these types of materials that that would be a big factor because the Aroclor is providing the nonsensitivity to the moisture.'

It is true that a patent is not rendered invalid by the fact that an inventor did not know why the new and beneficial result was obtained. Compare United States v. Adams, 383 U.S. 39, 86 S. Ct. 708, 15 L. Ed. 2d 572 (1966). But just as 'accidental results, not intended and not appreciated, do not constitute anticipation,' Eibel Process Co. v. Minnesota & Ontario Paper Co., 261 U.S. 45, 66, 43 S. Ct. 322, 67 L. Ed. 523 (1923),21 so an unappreciated result of a structurally obvious modification cannot of itself constitute invention. While Heiss may have noted the superior water-resistant properties afforded by using propylene rather than ethylene oxide as a condensant, he revealed no appreciation of the importance of this until publication of the Price patent. We hold that the invention described in Heiss' specifications and claims as they read until 1958 was obvious in the light of Windemuth;22 and that, even if under Papesch/Deutsche Gold-und-Silber-Scheideanstalt the invention described in the claims copied from Price would be patentable because of the disclosure of additional properties in compositions derived from propylene oxide condensates, then as to that claimed invention, Heiss 'did not himself invent the subject matter sought to be patented,' 35 U.S.C. 102(f). Contrast In re Zenitz, 333 F.2d 924, 52 CCPA 746 (1964). We thus have no occasion to consider whether because of the inadequate disclosure in the application until the new claims were copied from Price in 1958, Heiss is not entitled to the benefit of the doctrine of relation back, 35 U.S.C. 120 (cf. id. 112, 132), compare Benz v. Celeste Fur Dyeing & Dressing Corp., 156 F.2d 510 (2 Cir.), cert. denied, 329 U.S. 736, 67 S. Ct. 101, 91 L. Ed. 635 (1946), with Technicon Instruments Corp. v. Coleman Instruments, Inc., 255 F. Supp. 630, 640-641 (N.D.Ill.1966), aff'd, 385 F.2d 391, 393 (7 Cir. 1967), a point not specifically raised by General in this court, nor to consider other attacks which General does mount.

We therefore reverse the judgment of the district court with instructions to enter judgment granting General's complaint and dismissing Jefferson's counterclaim. General's request for attorney fees, 35 U.S.C. 285, is denied.