This appeal is from the decision of the Patent Office Board of Appeals affirming the examiner’s rejection of claims 4 and 5 in application serial No. 821,358, filed June 19, 1959, for “Halo-Substi-' tuted Phenoxymethyl Oxazolidones.” No claim is allowed.
The sole issue is obviousness under 35 U.S.C. § 103 in view of the following references :
Beasley et al., Chem. Abstracts, Vol. 51, col. 8723 (1957)1
Karrer, “Organic Chemistry” (Else-vier), pp. 434, 445 (1950)
Noller, “Chem. of Organic Compounds,” 2nd Ed., (Saunders), pp. 440-1 (1957)
Fieser et al., “Organic Chemistry,” (Reinhold), pp. 624-626 (1956)
The subject matter on appeal and its relation to this art can best be understood by referring to the following formula:
Claim 4 defines the compound shown, namely, 5- (p-chlorophenoxymethyl) -2-ox-azolidone.
Claim 5 defines the compound having the same structure except that the chloro atom is in the meta or 3 position rather than the 4 position as shown.
Beasley, the principal reference, discloses the compound having the same structure except that the chloro atom is in the ortho or 2 position.
As to the remaining references, we agree with the board that they “pertain to a fact so well known in organic chemistry, namely the existence of ortho, meta and para isomers of disubstituted benzene compounds, that it is hardly necessary to refer to them further.”
The board also said:
* * * we have come to the conclusion that, under the facts of this particular case, the appealed claims are not patentable over the cited Beasley et al. publication.
This issue therefore is: are appellant’s m- and p-chloro compounds patentable ' over the o-chloro isomer, and other disclosure, of Beasley?
Appellant admits “that, as pure chemical compounds,” his compounds “have a relationship sufficiently close to the compound of the prior art so that they would be held obvious under 35 U.S.C. § 103 absent a showing of a significant difference in properties.” (Our emphasis.) He urges patentability however, on the grounds: that his compounds are anti-convulsants, particularly effective against electroshock-induced convulsions, which fact is said to be indicative of ultimate anti-convulsant use against epilepsy; that there is no teaching in the art that the reference compound is at all active as an anti-convulsant of any type; that, while he discovered the Beasley compound to have anti-convulsant properties, he also found his compounds to be 4.4 and 7.0 times as potent as that compound against electroshock-induced convulsions ; and that the claimed compounds are therefore patentable because (1) based upon what the reference teaches, appellant’s compounds very clearly possess an unexpected property, and (2) even if one assigns to the “prior” art that information about the Beasley compound which appellant himself discovered, his compounds have been shown to possess unexpected superiority in potency which itself is conclusive of non-obviousness.
The examiner’s basic position was that Beasley discloses not only the o-chloro isomer of appellant’s compounds but other “related” compounds as well. One of these shows “a para-chloro positioning” and “mephenesin is mentioned”, this latter compound being 3-o-toloxy-l,2-propanediol. The examiner alsó stated that 5-(o-mei/losyphenoxymethyl) -2- ox-*382azolidone “is a well-known pharmaceutical.”
From this the examiner argued that the “prior art has found closely related compounds to have potent pharmacological properties” and contended that “When an important medicine is discovered it is customary practice to produce various molecular variants of it seeking the optimum as well as seeking other medicines which may derive from the fortunate association of atoms.” In a compound, according to the examiner, “there is a fundamental association which gives the ‘personality’ of the compound and, usually, minor substituents then merely shade the properties in matter of degree.” That being so, the examiner found the so-called “Second Ward Affidavit,” showing a number of quite startling changes in pharmacological properties resulting from seemingly minor structural changes in compounds appellant believed to be “related,” unpersuasive of patentability saying:
* * * the exceptions sometimes prove the rule and that there are exceptions does not negative the fact that generally changes as in isomers or homologues only affect properties in degree only. This does not make the compounds any the less obvious2 as such.
The board, in an opinion dated February 28, 1964, took somewhat the same approach. While stating that “properties of the compounds must be considered together with the chemical structures and, in a proper case, a patent may be granted by virtue of the unobvious properties,” it found affidavits submitted by appellant showing the compound of claim 4 to be 4.4 times and the compound of claim 5 to be 7.0 times as potent as the Beasley compound not convincing that the claimed compounds are “possessed of properties not possessed by the ortho compound of the prior art,” and affirmed the examiner on the ground that “A mere improvement in the same property does not establish unobviousness * *
Our principal difficulty, at this point, in accepting the examiner’s and board’s positions is that their reasoning is predicated on the premise, to use the board’s own words, that “the prior art compound has significant potency as an anticon-vulsant against electroshock-induced convulsions * * * ” While that is the fact, as shown by some of the above-mentioned affidavits, we are unable to find a dislosure of it in the art of record. Clearly, appellant’s discoveries in this regard cannot be relied upon to show the obviousness of his invention. Those discoveries are not prior art within the purview of section 103, and are made known to us in the record only through affidavits, not references.
The solicitor’s brief gives signs of his having struggled with the same problem. There, and at oral argument in an effort to show that the examiner and board “must have had some basis in mind” for their statement, the solicitor contended that appellant “should not make such an argument [that anti-convulsant activity in the prior art compound is taught only by him] because he should know that the reference compound would be expected to be active as an anticonvulsant.” (Our emphasis.) This follows, according to the solicitor, because (1) appellant has suggested no other activity for the reference compound and his specification states that compounds of similar structure to his were “known, and previously most active,” (2) the original Beasley article appeared in the Journal of Pharmacy and Pharmacology, which fact is *383noted in the abstract and therefore suggests pharmaceutical activity of the compounds disclosed therein, (3) the abstract describes experimentation with mephenesin which is a known anti-convulsant according to The Merck Index, 6th ed., 1952, and thereby also suggests anticonvulsant activity of other compounds disclosed in the reference, especially those prepared from mephenesin, and (4) appellant himself took out a prior art patent3 showing that compounds of similar structure are known to possess anticonvulsant activity. This, says the solicitor, is “some evidence” of what appellant knew and should be considered, notwithstanding the fact that that patent, technically speaking, was not used as a reference after the first action. Finally, the solicitor argues, since appellant has not submitted evidence to show the reference compound would not be expected to have anti-convulsant activity, the “finding” of the board that it has known activity should prevail, citing In re Knapp-Monarch Co., 296 F.2d 230, 49 CCPA 779.
Without implying our acceptance of the arguments above summarized, we defer for the moment further comment on them except to point out that Knapp-Monarch involved the issue of likelihood of confusion under section 2(d) of The Trademark Act of 1946 (15 U.S.C. § 1052 (d)), in which we said that “Factual matters of which judicial notice is taken” can be challenged by production of contrary evidence, but, in the absence of such evidence, a finding by the board based on its judicial notice is conclusive. We are unable to see, in the case at bar, any finding by the board based on such judicially noted facts. Rather, we tend to construe its statement that the prior art compound has anti-convulsant activity as an allegation, unsupported by any facts, unless of course the solicitor is suggesting that the examiner and board were somehow impliedly taking judicial notice of such publications as The Merck Index, supra, or appellant’s issued patent, and which he seems to be asking us to do now.
The solicitor’s brief, like the examiner and the board, makes a point about enhanced potency in the same kind of property, saying:
Appellant also urges throughout his brief that the increased potency of 440% and 700% of his compounds is necessarily an unexpected result which makes his compounds patentable, and he finally poses the question as to what greater degree of activity would have to be shown to make the compounds patentable * * *. It is clear from the decisions of this Court that a showing of up to nine times greater activity does not necessarily make a compound patentable, In re Carabateas [345 F.2d 1013], 52 CCPA [1386], 145 USPQ 549; In re Rosselet et al., [347 F.2d 847] 52 CCPA -, 146 USPQ 183.
While the statement in the last sentence above quoted is true, it becomes so merely by reason of the word “necessarily.” To dispel any misunderstanding, we wish to point out that what Carabateas held was that the two substituted-piperidines there claimed, differing from the primary reference by a so-called “reverse ester” configuration in the 4-position, and which were shown to be six and nine times more potent as analgesics than the primary reference compound, constituted “obvious” subject matter within the meaning of section 103 because the secondary references were believed to teach that esters having the structure of the claimed compounds would possess greater analgesic activity than their “reverse esters” by an amount approximating that found by the appellant. In brief, the enhanced potency of the invention was considered to have been reasonably suggested to one of ordinary skill in the art. Stated differently, the very property there relied upon for pat-entability was suggested in both its *384qualitative and quantitative sense. That is all Carabateas says.
In Rosselet, we affirmed because of a lack of proof, saying “that the Spero compounds [against which no comparisons had been made] are * * * sufficiently close to the claimed compounds to make the claimed invention prima facie obvious and * * * a showing of some unexpected effectiveness or property relative to those compounds is required to support a finding of patentability.” We expressly said that we were not holding “that a showing of activity 4.7 and 1.6 times that of the prior art is insufficient as a matter of law to establish patentability * * *”
We have neither of those situations here. Unlike Rosselet, appellant has presented evidence of enhanced potency compared with the only prior art compound. That evidence is “clear and convincing evidence of substantially greater” effectiveness, as required by this court in In re Lohr, 317 F.2d 388, 50 CCPA 1274. Unlike Carabateas, we find nothing in the art of record to suggest to one of ordinary skill that appellant’s m-chloro and p-chloro isomers would have anti-convulsant activity substantially greater than the prior art o-chloro compound in an amount approximating that discovered by appellant. We therefore find Gara-bateas and Rosselet are not controlling or persuasive precedents in this appeal.
Inventions of this type cannot, we think, be passed off as “mere improvement in the same property” or as mere shades “in matter of degree” which compel, ipso facto, a finding of unpatenta-bility.
In the first place, the discovered activities or properties are part of the things claimed, the compounds. Whereas the inventor may or may not know what part of the compound is responsible for the enhanced activity, the important point, under the law, is that there be proof that the compound in fact possesses that activity. Once that has been done, we, like the Patent Office, are required by the clear wording of section 103 to regard it as a relevant portion of the invention “as a whole.”
Second, we find no authority in section 103 for treating “improvement” inventions, or inventions differing from the prior art only “in matter of degree,” any differently from other types of inventions.
Third, we are inclined to agree with appellant that inventions of the type here involved are frequently made only after the expenditure of vast amounts of research time and effort; in short, they represent the very kind of invention some industries, most notably the drug industry, seek in order to obtain, or maintain, the kind of competitive advantage which promotes progress in the “useful arts.”
Fourth, we agree “that new compounds are nothing more than a novel combination of elements, all of which elements are per se old,” to quote appellant, “a fundamental association [of atoms],” to quote the examiner, and we see no basis in logic or in law for distinguishing mechanical and chemical inventions on an issue such as that before us. Certainly a novel combination of old mechanical elements which so cooperate as to produce a new and useful result, or merely a substantial increase in efficiency, can be patentable. In re Wright, 268 F.2d 757, 46 CCPA 955, and cases there cited. Furthermore, like the Patent Office, we have frequently found novel chemical processes producing the same product, but in unexpectedly higher yields, to be patentable by reason of that yield, a “matter of degree.” Should not chemical products, also displaying an unexpectedly higher degree of effectiveness, be treated in like manner?
In the final analysis, as in all section 103 cases, we must look first to the differences between the prior art and the subject matter sought to be patented and then determine if what appellant did, or made, as a whole, would have been obvious.
In the pre-1952 Act case of Schering Corp. v. Gilbert, 153 F.2d 428, 430 (2d *385Cir.) where, on the question of patenta-bility of a compound proved to possess unobvious and highly useful properties as an x-ray contrast agent in chole-cystography, the majority said:
* * * it is necessary to understand what the inventors did as well as what they sought to accomplish and give recognition to their end, result as a novel and useful improvement, not in the art of organic chemistry but in that of cholecystography. [Our emphasis.]
We followed that rationale, applying the present statute, in In re Papesch, supra footnote 2, and have continued to follow it since.
In Papesch we reversed the board’s af-firmance of a rejection of claims to a triethyl compound on a reference disclosing the corresponding trimethyl compound because the invention possessed unexpectedly potent anti-inflammatory activity. While Papesch did not involve the “so many times greater” situation we have here, that case and the ensuing cases cannot be distinguished from the instant case on that ground. Appellant has here proved his compounds to possess anti-convulsant activity substantially greater than the prior art compound. We find this to have been unpredictable from the prior art, even upon considering the solicitor’s arguments about suggestions of some activity allegedly derivable from knowledge about “related” compounds mentioned in the reference. Therefore, as in Papesch, there exists a significant, advantageous, unexpected difference between appellant’s invention and the prior art, which renders the invention patentable.
The decision of the board is reversed.
Reversed.
WORLEY, C. J., concurs in the result.
. It is conceded that the abstract itself, not the article abstracted, is the reference relied on.
. Just how one finds the compounds “obvious” in the first instance, the examiner does not say, but apparently he envisions a comparison of structures only. That such an aproaeh is not sanctioned by this court, although concededly the law was less well-defined in June, 1961, the date of the Examiner’s Answer, can be seen, e. g., in In re Krazinski, 347 F.2d 656, 52 CCPA 1447; In re Ruschig, 343 F.2d 965, 52 CCPA 1238; In re Ward, 329 F.2d 1021, 51 CCPA 1132; In re Lunsford, 327 F.2d 526, 51 CCPA 1000; In re Riden, Jr., 318 F.2d 761, 50 CCPA 1411; In re Papesch, 315 F.2d 381, 50 CCPA 1084; In re Petering, 301 F.2d 676, 49 CCPA 993; In re Lambooy, 300 F.2d 950, 49 CCPA 985.
. U. S. patent 2,895,960, issued July 21, 1959, disclosing and claiming a method of preparing 5-o-mei/io®i/phenoxymethyl-2-oxazolidone.