The Commissioner of Patents appeals from an order of the District Court directing him to issue a patent to Appellee on two claims of a patent application.1 Appellee, a West German corporation, is the assignee of the application, entitled 'Thiophenylpyridyl Amine, Chlorothiophenylpyridyl Amine, Their Salts and Preparation.' Claim 1 is generic, consisting of an ammonia-derivative compound selected from a group of compounds described by a general formula. Claim 2, a specific ammonia-derivative of Claim 1, is a specie identified in the chemical nomenclature as 1-azaphenothiazine.

Admitting that the Patent Office is taking a new and 'revolutionary approach,'2 the Commissioner raises four issues here: (1) whether under 35 U.S.C. 103 (1964) structural obviousness alone is an absolute bar to the grant of a patent; (2) whether 'advance in the art' is a fourth condition of patentability required by section 103 and Graham v. John Deere Co., 383 U.S. 1, 86 S. Ct. 684, 15 L. Ed. 2d 545 (1966); (3) whether evidence of comparative tests between the claimed compound and its isomers was insufficient to satisfy the burden of proof; and (4) whether the District Court erred in basing its decision on secondary evidence and not following the 'best evidence' rule.

The factual background can be stated briefly. Appellee's application consisted of six claims including generic, specific, and process claims. The Examiner rejected five of the claims in light of four prior art references, and the last claim was cancelled. The Examiner also found that a 'rule 132' affidavit,3 submitted to traverse the prior art references, was insufficient. The Board of Appeals reversed as to the process claims, which still stand,4 but affirmed the Examiner as to the generic and specific claims, the same claims in issue on this appeal, finding them unpatentable for failure to meet the non-obviousness requirement of section 103. The Board further agreed with the Examiner that the affidavit was ineffective to overcome the finding of obviousness, because no showing was made as to the pharmacological properties of the claimed compounds themselves in the affidavit.5

Apparently finding structural obviousness, the District Court determined that the existence of novel beneficial properties in the claimed compounds served to overcome the finding of obviousness due to structural similarities. In order to meet the prior art references, Appellee introduced evidence showing novel vermicidal properties of the claimed compounds not disclosed or suggested in the prior art. Evidence was also introduced showing the enhanced effectiveness of the compounds as intermediates in the preparation of patented sedatives and antihistamines.

(1)

The third statutory criteria for the grant of a patent, 'nonobviousness,'6 ] is found in section 103. '(A) codification of judicial precedents embracing the Hotchkiss7 condition' of invention, Graham v. John Deere Co., 383 U.S. 1, 17, 86 S. Ct. 684, 693 (1966), section 103 precludes the grant of a patent

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 a person having ordinary skill in the art to which said subject matter pertains.

In the Graham case the Supreme Court, for the first time, devoted itself to an extensive discussion of the concept of 'non-obviousness' as codified in section 103. Some further clarification seems in order in what appears to a blurred area of patent law.

The Commissioner asserts that under the standard on non-obviousness, structural obviousness of the claimed compound from the prior art reference is an absolute bar to a patent. While conceding that the language of section 103, 'subject matter as a whole,' refers not only to the chemical structure or formula, but can include properties of the structure or compound and the results it can accomplish, the Commissioner constructs the following argument:

It should be quite clear that the antecedent of 'subject matter as a whole' in Section 103 of 35 U.S.C. is 'subject matter sought to be patented', which expression means obviously the claimed subject matter, as distinguished from subject matter which is merely disclosed in applicant's specifications but not included in the claims.

Commissioner's Reply Brief, p. 2. But cf. General Tire & Rubber Co. v. Brenner, 127 U.S.App.D.C. 102, 381 F.2d 270 (1967).

Initially we have some doubt as to the question of structural obviousness. The Examiner found that the concept of the claimed compound was either clearly taught or obvious8 from the prior art which disclosed the corresponding oxygen analog of the claimed compound;9 that the concept of varying the ring nitrogen position and ring isomers to create the claimed compound were taught by name and structure;10 and that the methods of preparation were clearly suggested by the prior art. The Board of Appeals agreed. Apparently this was also the finding of the District Court, although the opinion is not entirely clear on this score.11 Since a Patent Office determination of non-obviousness is entitled to great weight, and District Court findings to the same effect will not be set aside unless clearly erroneous,12 we are satisfied as to the structural obviousness of the claimed compound and the prior art references; indeed, the parties do not seem to challenge this. We note, however, that this is not a case where a new property was discovered in an old compound nor one involving a simplistic change in a chemical formula. Appellee's evidence discloses that substantial time and effort were devoted to experimentation in order to develop a new compound with novel and unobvious properties and utility.

The issue raised by the Commissioner is the extent to which this structural obviousness operates in the determination of the non-obviousness of the subject matter as a whole under section 103. Structural obviousness generally arises because the claimed compound is-- in the chemists' jargon-- a homolog, analog, or isomer of the prior art references.13 In fact, the enormous number of known organic compounds gives rise to a situation in which absolutely unique and unknown groupings of atoms in a completely new chemical compound is a rare occurrence. In light of this datum, this Court and the Court of Customs and Patent Appeals for some years now have been of the view that patentability is not determined on the basis of structural obviousness alone. Obvious molecular modification coupled with a showing of novel properties14 or superiority of known properties15 can establish patentability.

Perhaps the best expression of the view that obvious but novel compounds should be considered in conjunction with their properties is Judge Rich's excellent opinion in Application of Papesch, 315 F.2d 381, 50 CCPA 1084 (1963). After discussing the various authorities and concluding that the statute required a deconcluding of obviousness and patentability after consideration both of the chemical structure and the pharmaceutical and biological properties of the claimed compound, Judge Rich stated:

From the standpoint of patent law, a compound and all its properties are inseparable; they are one and the same thing. The graphic formulae, the chemical nomenclature, the systems of classification and study such as the concepts of homology, isomerism, etc., are mere symbols by which compounds can be identified, classified, and compared. But a formula is not a compound and while it may serve in a claim to identify what is being patented, as the metes and bounds of a deed identify a plot of land, the thing that is patented is not the formula but the compound identified by it. And the patentability of the thing does not depend on the similarity of its formula to that of another compound but of the similarity of the former compound to the latter. There is no basis in law for ignoring any property in making such a comparison. An assumed similarity based on a comparison of formulae must give way to evidence that the assumption is erroneous.

Id. at 391. The sound reasoning of the Papesch case has been unanimously and repeatedly reaffirmed by the Court of Customs and Patent Appeals.16 Doubtless we must agree with admonitions against deciding questions of chemical obviousness on the basis of structure alone. In response to the arguments of the Commissioner and those who assert that if a compound is structurally obvious, a showing of novel properties should not make it any less obvious,17 several points should be noted.

First, it would appear that the Commissioner's position is hardly supported, but on the contrary is substantially undermined, by Graham v. John Deere Co., 383 U.S. 1, 86 S. Ct. 684 (1966), although he seems to place great reliance on this case. Graham was the first authoritative opinion dealing with the question of obviousness under section 103. It is clear from that unanimous decision that hard and fast rules of general applicability are impossible to achieve under section 103. Rather, the section 'lends itself to several basic factual inquiries,' enumerated by the Supreme Court:

Under 103, the scope and content of the prior art are to be determined; differences between the prior art and the claims at issue are to be ascertained; and the level of ordinary skill in the pertinent art resolved. Against this background, the obviousness or nonobviousness of the subject matter is determined.

Id. at 17, 86 S.Ct at 694. It seems rather elementary to us that the 'background' referred to for determination of obviousness clearly is not limited to the issue of structural similarity. Position isomerism, as involved here, focuses on close structural similarity which is to be taken into consideration with all other relevant facts in measuring the standard of obviousness under section 103. E.g., Application of Wiechert, 370 F.2d 927, 933, 54 CCPA 957 (1967); Application of Mehta, 347 F.2d 859, 52 CCPA 1615 (1965). This, we think, is the teaching of Graham.

While agreeing with this basic factual approach of Graham, the Commissioner assetts that where structural obviousness is assumed, 'inquiry by the court into the issue as to whether the claimed subject matter has unexpected beneficial properties is prohibited under the Graham case.' .he then argues that this prohibition is derived from the fact that the issue of chemical properties is, at best, a 'secondary consideration' under Graham. It is true that the Court in Graham mentions certain secondary considerations such as 'commercial success, long felt but unsolved needs, failure of others, etc.,' 383 U.S. at 17, 86 S.Ct. at 694. But it is equally true that no court, including the Supreme Court, has ever considered chemical properties to be 'secondary' in the determination of obviousness. See Note, 112 U.PA.L.REV. 1139 (1964).18

A second factor, to which we have already alluded, is the rare occurrence in finding a completely new compound whose molecular arrangements cannot be predicted, at least on paper. If we are to accept a determination of structural obviousness which cannot be rebutted by evidence of chemical, pharmaceutical, or biological properties, few pharmaceutical compounds will be patented, and the incentive aspect of the patent system, which is basic, will be lost from an important area.19 The fact that 'monopoly' power exists in the patent system does not mean, ipso facto, that it should be treated as an evil in the antitrust sense. While such exclusivity commands scrutiny, it must be recognized that this temporary 'monopoly' is a deliberate creature of the Constitution and one long implemented by Congress to promote technology and encourage progress.

A third salient point comes from the Commissioner himself. The Patent Office Rules, particularly Rule 132, permit an applicant to meet an initial rejection with proof of unobviousness. A traditional method has always been to permit evidence of novel properties or utility. Indeed, prior to the Graham case, this was the Commissioner's consistent position. E.g., Brief for Commissioner, Application of Higgins, 369 F.2d 414 (C.C.P.A.1966). We find nothing in Graham which calls for the radical shift in the law proposed.

Fourth, the patent statutes as a whole would seem to favor the Papesch standard. The section 101 criteria-- utility-- is largely dependent on the question of what the compound will do, and its properties are of prime consideration here. See Brenner v. Manson, 383 U.S. 519, 86 S. Ct. 1033, 16 L. Ed. 2d 69 (1966).

The Commissioner, in addition to reliance on Graham, refers to a recent decision of this court to support his position, Hays v. Brenner, 123 U.S.App.D.C. 96, 357 F.2d 287 (1966). He views Hays as holding that section 103 is an absolute bar to a patent where obviousness is assumed. It is true that some of the language in Hays affords a colorable basis for his contention, but it is also clear that the holding was not directed to the question of structural obviousness alone but to obviousness of the subject matter as a whole, a completely different situation. In Hays we made no effort to determine what the claimed subject matter was in the case of a chemical compound. The plain language of section 103 makes it clear that obviousness of the subject matter as a whole is not limited to a consideration of structure; and this is consistent with the common sense of the situation.

In rejecting the broad proposition advanced by the Commissioner we should make it clear that structural obviousness is certainly a large factor to be considered in the determination of obviousness of the subject matter as a whole. Whether it is called an inference of obviousness, Application of Mills, 281 F.2d 218, 223 (C.C.P.A. 1960), a prima facie showing of obviousness, Application of Rosselet, 347 F.2d 847, 850, 52 CCPA 1533 (1965), or a presumption of obviousness, Application of Henze, 181 F.2d 196, 201, 37 CCPA 1009 (1950), is, we believe, immaterial.20 The crucial factor is not the definition attached to the relationship nor the legal phraseology used to describe deductions reached from that definition, but rather the proximity of the relationship in terms of non-obviousness of the subject matter as a whole.

The final assertion from the Commissioner's sweeping contentions is interesting but does not compel a contrary result. He views the claim as defining by name and diagram a mental concept of a chemical compound which, to the ordinary chemist, would convey both the structure and characteristics of a family of compounds. The Commissioner would have us adopt the following logic: Since section 103 refers to 'subject matter as a whole,' the antecedent of which is 'subject matter sought to be patented,' these obviously refer only to the claimed compound; hence, if that claim does not specifically refer to the particular property which demonstrates non-obviousness, it does not distinctly claim the invention as required in section 112, paragraph 2, but instead claims that compound for all uses and properties, thus overclaiming and foreclosing the public from using the compound in an unrelated, but unobvious, field.21 Cf. Application of Kirk, 376 F.2d 936, 942 (C.C.P.A.1967). The Commissioner relies upon some out-of-context statements from Graham, and also on the companion case, Calmar, Inc. v. Cook Chem. Co., 383 U.S. 1, 26, 86 S. Ct. 684, 15 L. Ed. 2d 545 (1966). In the latter case, however, the parties agreed upon certain limitations in the claims. In light of existing law it cannot be doubted that the claim itself must not be divorced in a vacuum from the specifications and descriptions accompanying it. See, e.g., Schering Corp. v. Gilbert, 153 F.2d 428, 432 (2d Cir. 1946).

(2)

The Commissioner's second claim of error concerns some language in the District Court's opinion to the effect that 'advance in the art' is not a condition for patentability.22 The Commissioner refers us to several unrelated cases in which certain language suggests the contrary,23 and to Graham. Appellee responds by pointing to the exclusion of 'advancement' from the three legislatively-created statutory conditions of patentability, 35 U.S.C. 101-103 (1964).

From the introductory sentence in Graham it is clear that the Supreme Court was well aware of the distinction between statutory conditions of patentability and constitutional standards.24 This is perhaps best illustrated by the following excerpt:

Innovation, advancement, and things which add to the sum of useful knowledge are ingerent requisites in a patent system which by constitutional command must 'promote the Progress of * * * useful Arts.' This is the standard expressed in the Constitution and it may not be ignored.

383 U.S. at 6, 86 S. Ct. at 688. Appellee and the District Court are correct in stating that advancement is not one of the three conditions of patentability under the statutes. But this overlooks the clear, and at least implied, discussion in Graham of the constitutional standard. The conditions effectuating this standard are for congressional determination, but the standard itself is to be implemented by the Commissioner and the courts.25 Thus, the semantic correctness of the District Court opinion should not be misconstrued. Cf. Baenitz v. Ladd, 124 U.S.App.D.C. 237, 240, 363 F.2d 969, 972 (1966).

The requirement of advancement is an illusory one, focusing on something incapable of precise definition. Other than an implication that advancement was the equivalent of 'Progress' as used in the constitutional phrase,26 the Court in Graham made no effort to define what constitutes an advancement. Similarly, the opinion does not tell how to determine whether an advancement has been made, although it does state more than once that advancement in the constitutional sense is tied in with the statutory condition of nonobviousness.27 The Commissioner's position is best summed up in the response of the Solicitor at oral argument in this court when he was asked whether advance in the art is a requirement additional to the three specified in the statute, or is related to the statutory requirement of non-obviousness. He replied that in most situations the mere fact that the difference between the prior art and that which is described and claimed in any particular application or patent would per se be an advance in the art, but he also could envision situations in which that would not necessarily be true.

The Commissioner has conceded in this reply and elsewhere that a determination of unobviousness in most cases, but not necessarily all, is equivalent to an advance in the art. Appellee asserts that the congressional creation of the third statutory condition, non-obviousness, accomplished the constitutional purpose of the patent provision, and incorporates the concept of advance in the art. Indeed, Graham draws a close parallel between the determination of non-obviousness and advancement to the extent that the inquiry into non-obviousness 'comports with the constitutional structures,' 383 U.S. at 17, 86 S. Ct. 684. Similarly, it is quite clear that the pivotal condition discussed in Graham, non-obviousness, which was added to the patent statute in 1952, strongly equates itself with the judicial condition of 'invention,' previously so long used. Both formulations, invention and non-obviousness, 'place emphasis on the pertinent art existing at the time the invention was made and both are implicitly tied to advances in that art,' id. at 14, 86 S.Ct. at 692.

But, while the major emphasis may be placed on the condition of non-obviousness, we do not find that such a determination, ipso facto, is also an ultimate determination on the question of advancement. The other two statutory requisites, novelty and utility, have always been thought to be separate and independent conditions. If advance in the useful art is the necessary standard to sustain patentability, the more logical view is that the three independent statutory conditions are not separate from advancement but are the very indicia of it, in particular the condition of non-obviousness. See Note, 54 GEO.L.J. 1320, 1339 (1966).

Under this analysis we have some difficulty in perceiving how a claimed compound which is new, useful, and unobvious can still not constitute an advance in the art.28 The Commissioner's position is that even if the three conditions are fulfilled, a step backward in the art is not patentable. But the relevant art does not go 'backward' when the three conditions are fulfilled. The Commissioner's position appears to be a contradiction in terms.

If, as is asserted by amicus, the Commissioner is really contending that before any claimed subject matter is patentable, it must 'advance the art' in the sense that it is demonstrated to be better than that which preceded it in terms of cost, performance, or otherwise, we must reject such a view.29 We do not equate the constitutional standard of advancement with a requirement that the claimed subject matter be 'better' than the prior art.30 The Constitution uses the word 'Progress', and progress is an advancement step by step to new, useful, and non-obvious subject matter:

Progress is most effectively promoted by protecting those who enrich the art as well as those who improve it. Even though their inventions are not as good as what really exists, such inventors are not being rewarded for standing still or for retrogressing, but for having invented something. The system is not concerned with the individual inventor's progress but only with what is happening to technology.

Rich, Principles of Patentability, 28 GEO.WASH.L.REV. 390, 402 (1960).

This is our understanding of the standard of advancement as interpreted in Graham. It is correct that Graham admonished the Commissioner to 'tighten up' on applications, and mildly rebuked him for leniency in applying the conditions of section 103. 383 U.S. at 18, 86 S. Ct. 684. But this was not the creation of a new requirement of patentability31 nor was it a mandate for the Patent Office to enforce the statutory conditions and constitutional standard with the zeal manifest in the Commissioner's position here.

Our analysis of the District Judge's opinion shows that the statement that advance in the art was not a condition of patentablity meant no more than that the constitutional standard of advancement was demonstrated by the utility of the claimed compound over the prior art through its unobvious anthelmintic properties, regardless of the relative superiority in this respect. This seems to be the essence of the entire paragraph which the Commissioner attacks as erroneous. Surely we could impute no other interpretation to the opinion and we are satisfied that the District Judge did not mean that advance in knowledge was unnecessary to patentability.32

(3)

In an effort to overcome the burden of structural similarity with the prior art, Appellee specified in its complaint in the District Court, as it had done in its application before the Patent Office, that the claimed compound had unexpected utility in two areas. First, evidence was offered showing that the claimed compound was superior to three of its prior art isomers as a vermicide, a utility alleged to be unobvious in the prior references. Second, evidence was offered showing that patented antihistimines and sedatives had greater supeority when prepared from derivatives containing the claimed compound than when prepared under the then-known methodology. The Commissioner makes several claims regarding the sufficiency of this evidence.

(a) Direct Utility as a Vermicide

The first phase of Appellee's evidence consisted of comparative tests between the claimed compound and three prior art isomers, demonstrating the superiority of the claimed compound when utilized as a vermicide. These tests, found by the District Court to be universally recognized showed the following: (1) the claimed compound was equal in anthelmintic activity on ascaris worms33 to two of the prior art isomers and one hundred times more effective than the third isomer; (2) the claimed compound was the least toxic34 when compared with the three isomers; and (3) the claimed compound demonstrated a superior therapeutic value by the following comparative ratios, 1: .45: .70: .01.

This comparative evidence was fully adopted by the District Court and led to that Court's conclusion that the claimed compound would be 'equal or superior as a vermicide' when compared with the prior art isomers. 251 F. Supp. at 628. Furthermore, the District Court found that none of the prior art references relied upon suggested that the azaphenothiazines (the claimed compound and its prior art isomers) would be useful as a vermicide. Thus, the District Court concluded:

When the prior art does not contain the slightest suggestion that the properties of a claimed compound would make it useful for a certain advantageous purpose, then the utility itself is unobvious, and it is unnecessary for an applicant to prove by comparative tests that his compound is more useful for the certain purpose than closely related prior art compounds.

Ibid.

The Commissioner makes two arguments directed at this evidence. First, he argues that the comparative tests did not afford a sufficient basis for the District Court's unequivocal holding that the claimed compound was equal or superior to the prior art isomers when utilized as a vermicide.

As we have previously noted, see note 12 supra, we are governed in our review of the factual findings of the District Court by the 'clearly erroneous' standard of Rule 52 of the Federal Rules of Civil Procedure. On the basis of the evidence produced on comparative tests, we cannot say that the District Court erred in holding that the claimed compound 'was equal or superior as a vermicide to the other three isomers * * *,' 251 F. Supp. at 628. Thus we reject the Commissioner's first argument.

The Commissioner's second argument is directed at the statement of the District Judge just quoted, to the effect that an applicant need not prove that his claimed compound is more useful by comparative testing where the prior art does not suggest that the properties of the claimed compound would be useful for a certain purpose.35

In support of its legal conclusion, the District Court relied upon E. I. DuPont De Nemours & Co. v. Ladd, 117 U.S.App.D.C. 246, 328 F.2d 547 (1964) and Application of De Lajarte, 337 F.2d 870, 52 CCPA 826 (1964). The Commissioner challenges the applicability of these cases. As was expressly noted by this court in DuPont, the issue there involved a question of anticipation under 35 U.S.C. 102, not a question of obviousness under section 103. Moreover, since section 102 was involved, no issue was raised regarding the necessity of making comparative tests involving closely related prior art compound.36 We think that the De Lajarte case, however, supports the conclusions of the District Judge. Although there was some doubt as to whether anticipation or obviousness was involved, 337 F.2d at 873, the court considered both sections and concluded:

In the total absence of evidence in the record to indicate that the amber glass disclosed by Lyle (the prior art reference) would be expected to have desirable electrical insulating properties, we can find no justification for placing the burden on applicant to conduct experiments to determine the insulating properties of the colored glass disclosed by Lyle.

Id. at 874.

In addition to distinguishing these cases, the Commissioner's principal argument is that where an applicant is attempting to demonstrate that his claimed compound possesses novel or unobvious beneficial properties he must show that such properties are not in fact possessed by the prior art reference. Under the ruling of the District Judge no such showing was required of Appellee. The Commissioner relies upon Application of Henze, 181 F.2d 196, 37 CCPA 1009 (1950).

In Henze, the application was for claimed compounds one of which was the next higher homolog of a compound disclosed in a prior publication.

The examiner had held, * * * that the claims in issue were unpatentable, lacking invention over the disclosure of the publication, because the appellant had failed to show that the compounds claimed had an unexpected beneficial property not possessed by the lower adjacent homologue disclosed in the publication. The examiner held that the affidavit did not aid in conferring patentability upon the rejected claims because there was no showing that the lower homologue would not yield results similar to the claimed compounds under certain dosages. The burden is on the applicant to rebut that presumption (of unpatentability) by a showing that the claimed compound possesses unobvious or unexpected beneficial properties not actually possessed by the prior art homologue. It is immaterial that the prior art homologue may not be recognized or known to be useful for the same purpose or to possess the same properties as the claimed compound.

Id. at 200-201.

The rationale articulated for the rule was that the characteristics normally possessed by members of a homologous series are generally the same, varying gradually from one member to the next. Thus, a chemist knowing the properties of one member of the series would have little difficulty predicting the properties of the adjacent members of the homologous series. Ibid.

The Henze case and its progeny have brought little clarification to the field of patent law. See note 20, supra. The case itself applied its doctrine to the question of adjacent homologs. Isomerism, as involved in the present case, is somewhat different; it is basically a structural classification not dependent upon similarities of properties as is homology. See MORRISON & BOYD, ORGANIC CHEMISTRY 32 (2d ed. 1966); Collins, supra note 20, at 286. The Commissioner asserts that Henze is applicable to isomerism as well as homology and points to specific language in that opinion, 181 F.2d at 200-201. Such language is present, but on the basis of subsequent decisions of the Court of Customs and Patent Appeals, we think that it referred to the general problem of chemical obviousness, not to the specific 'presumption' or 'inference' of obviousness derived from the structural similarity present in the case of adjacent homologs.

In one line of cases, that court has made it quite clear that Henze should be limited to its factual situation, even to the extent of excluding the question of non-adjacent homologs from its application.37 In another line of cases the court has indicated that the definitions attached to the relationship are unimportant, Application of Druey, 319 F.2d 237, 240, 50 CCPA 1538 (1963); rather it is the proximity of the relationship in terms of the non-obviousness of the subject matter as a whole. Application of Lohr,317 F.2d 388, 389 n. 2, 50 CCPA 1274 (1963); Application of Mehta, 347 F.2d 859, 864, 52 CCPA 1615 (1965). We think these cases demonstrate that the genesis for the Henze rule was found in the homology of hydrocarbon chains alone, and that no mechanistic formulae can be indiscriminately applied to questions of chemical obviousness.38

Although not cited by the parties, we think Application of Riden, 318 F.2d 761, 50 CCPA 1411 (1963), has some bearing on the issue. In Riden the application was for certain compounds useful as pesticides. The primary reference disclosed no use, either as an insecticide or for any other purpose, id. at 762. Other references, however, did disclose utility as fungicides. The court found

There is no showing in this case that the compounds of claims 3 and 8 differ by more than a matter of degree from the compounds of Boehme et al. (the prior reference) in any properties. It is true that Boehme et al. does not teach that the compounds have insecticidal properties, but there is no showing that they do not. Had there been evidence of a new property of the closely related compounds not possessed by the analog, a different case might arise. * * * Such evidence was not forthcoming during the prosecution of the case. Indeed, the specification of appellants' application indicates that the Boehme et al. compounds have the same properties but differ only in degree, * * *.

Id. at 764. Thus, Riden would seem to indicate that even when the prior art was not alleged to possess the same properties, it is incumbent upon the applicant to make a comparison to see if it possessed the same properties. However, this does not appear to be the rule of the case since a patent was allowed as to the pesticidal use because such use was not suggested in the primary reference; nor did the court find it would have been obvious from other references which did disclose related uses. Id. at 768. Because of this difference we do not believe that Riden compels comparative testing to the extent here sought.

We recognize, of course, that the best method of demonstrating patentability under Papesch and its progeny by the existence of a novel and unobvious property is through the use of comprehensive comparative testing. Nevertheless, on the whole record we are satisfied that the District Judge's conclusion of law concerning comparative testing was correct as applied to the facts of this case. We reach this conclusion on the basis of the limited comparative testing here, which, although it did not disprove the existence of the same properties in the prior art, strongly suggests, by reason of the vastly superior results obtained, that those new results derived from the existence of a novel property in the claimed compound; by limiting Henze to its facts, as the Court of Customs and Patent Appeals has done; and by what we read as to the meaning of the additional evidence adduced in this case. It should be recalled that Appellee attempted to overcome the burden of structural obviousness through two evidentiary showings. In the first place, as we have indicated, evidence was offered to show new and unexpected properties and novel utilization of the claimed compounds. Had that been the only evidence offered we, and the District Judge as well, might have required further comparative testing. But Appellee went on to offer further evidence of a secondary nature-- the indirect utility of the claimed compound as an intermediate for the production of known drugs with greater effectiveness than when prepared under then-existing methodology. In light of this dual approach we find that the District Judge's conclusion of law not requiring comparative tests to be correct.

(b) Indirect Utility as an Intermediate

The second phase of Appellee's evidence consisted of comparative tests between the claimed compound and its three prior art isomers with respect to their value as an intermediate in the preparation of two end products, a patented sedative and a patented antihistamine. This evidence disclosed that when the patented compounds were prepared from the claimed compounds they had a greater sedative and antihistaminic effect and greater therapeutic ratio than when the patented compounds were prepared from the three prior art isomers. Appellee claims that these tests demonstrate some novel and unique property in the claimed compound and that one skilled in the art would ascribe to the claimed compounds the contributing cause for the differences in end-product results.

The Commissioner's argument is tied in with his earlier evidentiary contentions just discussed. There he claimed that the direct evidence comparing the prior art with the claimed compound did not sufficiently disclose a novel property in the prior art. Since that evidence was insufficient, the Commissioner continues, the only basis for granting the patent was the District Judge's reliance on the evidence in question in this aspect of the appeal. The error here, the Commissioner claims, is a violation of the 'best evidence' rule. Relying on a series of cases, the Commissioner argues that the District Court opinion violated the well-known rule that evidence of new properties or unexpected results possessed by final products made directly from intermediates or derivatives (i.e., the claimed compounds) is inadmissible to prove patentability because of the remoteness of the relation back to the intermediate.

The Solicitor in his brief relies upon the rule which had its genesis in Application of Finley, 174 F.2d 130, 36 CCPA 998 (1949). In Finley the claimed compound was an ester which, when converted to its calcium salt, produced an additant for lubricating oils with higher thermal stability than the prior art references. The examiner, the Board of Appeals, and the court agreed that the compound itself was too far removed from its calcium salt to justify attributing the utility of the end product to the claimed compound.39 The Patent Office subsequently adopted this rule in several cases,40 and the Court of Customs and Patent Appeals has reaffirmed the rule in several recent cases.41

We agree with the Solicitor that the best evidence rule has a sound basis for application in the patent field. But, the rule is not 'hard and fast,' Application of Widmer, supra note 41, 353 F.2d at 759. Secondary evidence runs a greater risk of failing the relevancy test of admissibility in proportion to its remoteness from the claimed subject matter, but the question of remoteness is a question of law to be determined on the basis of the subject matter as a whole.

We hold the District Court was correct in its application of the rule. Several reasons compel this conclusion. In all of the cases relied upon by the Solicitor, the secondary evidence was the only evidence presented, or the secondary use was the only use specified. Here, however, both direct evidence and secondary evidence was presented. In such a situation we think that the trier of facts is permitted to consider the secondary evidence as admissible evidence. Since this case was decided, the Court of Customs and Patent Appeals has expressly relied upon the opinion now under review for the proposition that the secondary evidence is admissible. Application of Diamond, supra note 41,360 F.2d at 219-220. Certainly, if the District Judge's application of the best evidence rule had been in conflict with the views of the Court of Customs and Patent Appeals, they would hardly have cited his decision in this case. Having amply predicated admissibility on the basis of these two facts, we cannot say that the District Judge was clearly erroneous in finding that the secondary evidence also demonstrated the unique and novel properties claimed for the compound, and that a person of ordinary skill in the art would reasonably ascribe to the claimed compounds the contributing cause for the differences between the end product when made from the prior art and when made from the claimed compound. We hold the District Court was correct both in his findings of fact and conclusions of law; we should make clear that our opinion with respect to this issue is predicated on the availability of both primary and secondary evidence, and on the sufficiency of this evidence to show the existence of a novel and unobvious property.

Affirmed.

In Graham v. John Deere Co., the Supreme Court stated that the 'inquiry which the Patent Office and the courts must make as to patentability must be beamed with greater intensity on the (non-obviousness) requirements of 103 * * *.' 383 U.S. 1, 19, 86 S. Ct. 684, 694 (1966). I respectfully submit that the majority has ignored the Supreme Court's admonition.

There is no question that the claimed compound is structurally obvious. The majority concludes that such a compound can meet the requirements of 103 if it has novel properties. I need not decide whether this conclusion is correct because I believe there was no adequate showing that the claimed compound in fact possessed novel properties.

The only method of proving that a compound has properties not possessed by the prior art is comparative testing. Here the majority concedes that the comparative testing was not comprehensive. Furthermore, the trial judge's ability to evaluate the limited comparative evidence was undoubtedly undermined by his belief that:

When the prior art does not contain the slightest suggestion that the properties of a claimed compound would make it useful for a certain advantageous purpose, then the utility itself is unobvious, and it is unnecessary for an applicant to prove by comparative tests that his compound is more useful for the certain purpose than closely related prior art compounds.

Finally, the district judge only concluded that the claimed compound was 'equal or superior as a vermicide' to the prior art compounds. This is not a clear cut finding of novel properties.

The majority refuses to require further comparative testing in this case because the appellee 'went on to offer further evidence of a secondary nature-- the indirect utility of the claimed compound as an intermediate for the production of known drugs with greater effectiveness than when prepared under then-existing methodology.' There is some question as to whether such secondary evidence is admissible.1 But even if it is, I do not believe it makes up for the lack of comprehensive testing on the primary issue-- the claimed compound's novel properties when utilized as a vermicide. Thus I think the Court's opinion today fails to heed the admonition of the Supreme Court in Graham v. John Deere Co.