On March 16, 1959, appellant filed application serial No. 799,483, entitled "Gasoline Composition." This appeal is from a decision of the board sustaining the examiner's rejection of all the remaining claims, 2-13. Claim 2 is typical and reads:

2. A gasoline composition consisting essentially of a branched chain di-lower-alkyl ether having from 4 through 8 carbon atoms in a mixture of naphthene, aromatic and paraffin gasoline boiling range hydrocarbons in which mixture the concentration of naphthenes is from about 2% to about 25% by volume, the concentration of aromatics is at least 15% and no more than 75% by volume and which mixture contains no more than about 10% by volume of olefins, the concentration of ether in the composition being no more than about 50% by volume.

It will be seen that what appellant is claiming is a gasoline composition. The position of the Patent Office here suggests the necessity for some basic thinking about the character of compositions of matter. From the standpoint of the patent law, a composition of matter is, as stated in Robinson on Patents, § 193:

* * * always a true combination. Each of its ingredients is itself a means whose operative forces manifest themselves through the chemical or mechanical properties by which it is distinguished. * * * It differs from all other combinations in that its ingredients, or elemental means, when once united in the combination, often become individually undiscernible by human sense, and can be recovered and distinguished only by the destruction of the combination as a whole. * *

Thus, at least in the patent law context, a gasoline composition as here claimed is necessarily a complete and separate entity whose existence as such entity is distinct from the substances of which it is composed. Even though these substances may be old, the composition may be new. The character of a new composition of individually old substances cannot, therefore, be determined from an examination of its component elements alone. It must be judged rather by the intrinsic attributes of the composition as a new combination. See Robinson, supra, § 194.

The gasoline composition as here claimed is but an embodiment of appellant's invention which resulted from an appreciation of an existing problem in the compounding of gasoline compositions and the knowledgeable selection of particular substances or ingredients named in the claims which, when blended within the limits clearly set forth in the claims, provided the new combination for which appellant seeks a patent.

These general observations take on particular significance in the context of the art to which the present invention relates. Gasoline compositions are readily evaluated for their so-called antiknock properties by standardized testing procedures from which an octane number is derived. Two accepted and standardized laboratory tests in general use to determine such octane numbers are: 1) The Research Method, which tests knock resistance under relatively mild, low-speed operating conditions (at 600 rpm with the air-fuel mixture entering at 125°F.); and 2) the Motor Method, which measures the octane number of a fuel under more extreme operating conditions (at 900 rpm with the air-fuel mixture at 300° F.).

While these measurements of the antiknock properties of gasoline compositions provide a satisfactory rating scale, they do not tell the whole story. A gasoline composition may show very satisfactory octane numbers by each method and yet may be unsatisfactory as an automobile motor fuel, since it may be utilized under widely varying operating conditions which may not be accurately reflected in either the Research Method or the Motor Method of octane number determination. The same gasoline composition usually has different octane numbers when tested by the foregoing test methods. It has been found that the difference between the two (Research octane number minus the Motor octane number) may be used as a quantitative measure of the "sensitivity" of the gasoline, i. e., its performance under conditions other than those simulated in these two methods. Thus, a gasoline having a Research Method number of 99 and a Motor Method number of 91 has a sensitivity of 8. In evaluating sensitivity, the rule is that the lower the sensitivity number, the better is the over-all performance of the gasoline composition.

As automotive fuel octane ratings have risen to conform to the modern trend toward high-compression engines, so has the sensitivity. Appellant's specification contains a table which purports to show the national average octane numbers and sensitivity of premium automotive gasolines for a number of years. This table shows the persistence of a problem in this art. It shows that as the octane numbers of the gasolines were increased, the sensitivity number also increased. Thus as the octane numbers increased from 85.2 (Research) in 1941-42 to 98.4 in 1957-58, and from 79.7 (Motor) in 1941-42 to 87.9 in 1957-58, this increase resulted in a progressive increase of "sensitivity" of from 5.5 in 1941-42 to 10.5 in 1957-58.

The significance of this progressive increase in gasoline sensitivity is apparent in view of modern design trends in automotive internal combustion engines. This is referred to in appellant's specification which points out that the "result of all these factors is that modern automotive engines are not only coming to require fuels of increased knock-resistance but are coming to require fuels of reduced sensitivity also." Appellant then goes on to state:

The principal object of the present invention is to provide a commercially practical high-octane-number gasoline of low sensitivity. * * *

Appellant asserts that he is able to attain the objective of decreased sensitivity by adding a branched chain di-lower-alkyl ether to a particular gasoline mixture, to form a gasoline composition as claimed. In an experiment described in the specification, 20 parts by volume of methyl tertiary-butyl ether were blended with 80 parts of a base gasoline containing 20% by volume of naphthenes, 47% by volume of aromatics, 1% by volume of olefins and 32% by volume of paraffins. Pure methyl tertiary-butyl ether alone has a Research Method octane number of 110.1, a Motor Method number of 100.6, and therefore has sensitivity factor of 9.5. The corresponding numbers for the base gasoline mixture alone were found to be 97.1, 88.0 and 9.1 respectively. However, when the combination of these materials was tested, the Research Method number was found to be 101.4, the Motor Method number was 93.6, and the sensitivity factor was 7.8. Appellant comments upon these results as follows:

It is, of course, not suprising [sic] that the higher octane number ether raised both the Research and Motor Method octane numbers. However, it was unexpected to observe that the ether, having a higher sensitivity than the base gasoline, lowered the sensitivity of the base gasoline instead of raising it further, as is usually the case with conventional hydrocarbon blending gasoline components. But what is most surprising is the fact that the sensitivity of the bland [sic] was lower than either that of the ether or the base gasoline separately.

The examiner based his rejection of the appealed claims upon the following references:

  Buc             2,046,243       June 30, 1936
  Ross et al.     2,360,585       Oct. 17, 1944
  Evans et al.    2,409,746       Oct. 22, 1946

Sachanen, The Chemical Constituents of Petroleum 260 (1945).

Wagner et al., "Improved Motor Fuels Through Selective Blending," pp. 8-13 (paper presented before American Petroleum Institute, Nov. 7, 1941).

Morris, "Effect of Gasoline Sensitivity on Road Octane Number," The Oil and Gas Journal, Nov. 26, 1956, pp. 33-34.

All the claims were rejected, in the words of the examiner:

* * * as being unpatentable over Buc and Evans et al. in view of Ross et al., Wagner et al., Sachanen, and Morris. Buc and Evans et al. teach the use of the claimed alkyl ethers as octane improvers and anti-knock agents in lead [-ed] and unleaded gasolines. Wagner et al., Ross et al., and Sachanen show the claimed gasoline blends to be old. Morris and Wagner et al. show fuel sensitivity and problems related thereto to be old and well known. Hence, in view of these comprehensive teachings, it would not involve invention1 to improve the octane and lower the sensitivity of the gasolines shown by Wagner et al., Ross et al., and Sachanen by adding the alkyl ethers of Buc and Evans et al. to them.

The Board of Appeals affirmed the examiner's rejection and stated the essence of its position as follows:

* * * We are of the opinion that the Buc and Evans et al. ether additives for gasolines would inherently function in the gasolines of Ross et al. of Wagner et al. to give the additional benefit of reduced sensitivity set forth in appellant's specification. The mere determination or discovery of the latter benefit which would flow naturally from following the teachings of the applied references does not render the claimed gasoline composition unobvious and patentable. [Emphasis added.]

We agree with the Patent Office that the prior art shows that the octane number of gasoline motor fuels can be improved by the addition of ethers of the type defined in appellant's claims. We also agree that the references show the base gasolines of appellant's claims to be old.

We note, however, that none of the references relied on by the Patent Office provide the slightest teaching to one of ordinary skill in this art as to what might happen to the sensitivity factor in a gasoline composition consisting of such ethers with such base gasolines. Thus, while the references would teach one of ordinary skill in this art that the blending of appellant's ethers and base gasolines would result in octane numbers for the blend falling somewhere in between the individual octane numbers for the individual components, they contain absolutely no information which would suggest that the sensitivity factor of the resulting blend would be lower than the sensitivity factor of either of the individual components. In brief, the art does not show that it would have been obvious to one of ordinary skill in this art, under the test of section 103, to combine known ingredients to produce a new gasoline composition which is characterized by the unique and unobvious combination of high anti-knock properties with decreased sensitivity to engine operating conditions.

We are not impressed by the board's view, here urged by the solicitor, that because the ether additives would "inherently function" to produce the decreased sensitivity, the claimed blends are therefore obvious. Indeed, can it not be said that all functions of a substance are "inherent"? The fact remains that appellant has discovered that when certain ethers are blended with certain gasolines, the sensitivity of the resulting gasoline composition is less than the sensitivity of either the ether or the gasoline alone. The mere notion that such discovery would "flow naturally" from what appellant did does not mean that the discovery could be predicted from what the art had done. As we said in In re Papesch, 315 F.2d 381, 50 CCPA 1084: "From the standpoint of patent law, a compound and all of its properties are inseparable; they are one and the same thing." Or, as we stated recently in In re Huellmantel, 324 F.2d 998, 51 CCPA 845:

* * * This court has frequently stated that properties of chemical compounds must be considered in determining obviousness under 35 U.S.C. § 103. * * * We think this also applies to such compositions as we have here, i. e., admixtures of two or more compounds. Therefore, the issue of obviousness in this case can be resolved only after considering the differences between the prior art and the claimed compositions chemically and in view of their pharmaceutical properties. * * * And, in a footnote to the foregoing statement:

* * * The examiner concedes that appellant's composition "shows improved results." The board accepts this as "fact," but concludes it is immaterial since "it is obvious to at least try" (our emphasis) to substitute one steroid for another in the prior art composition.

We believe this reasoning, insofar as it negates consideration of properties in determining obviousness under section 103, flys in the face of the plain language of the statute as interpreted by this court * * *. Section 103 says, inter alia, "the subject matter as a whole would have been obvious * * *." Nothing is said about "obvious to try." Consideration of the subject matter "as a whole" in chemical cases requires comparison of properties, pharmaceutical or otherwise, as well as comparison of chemical structures.

We do not mean to imply that every variance in property of a new compound or composition will tip the balance for patentability where otherwise closely related compounds or compositions are involved. However, all relevant property differences must be considered in the light of the facts of each case in the determination of statutory obviousness.

To the foregoing pronouncements of this court, we would add the pertinent statement of the court in Schering Corp. v. Gilbert, 153 F.2d 428, (2d Cir. 1946) that:

* * * The utter novelty of this product makes the seemingly slight departure from the old deceptive unless it is kept in mind that such things must be evaluated not alone by the degree of the change but also by reference to the purpose sought to be accomplished.

Steering our determination of this appeal by the guidelines laid down in Huellmantel, we find that the reduced sensitivity of appellant's claimed gasoline composition, along with its high octane numbers, is cogent evidence that the claimed composition is an entity distinct from its individual components and we thus find that such a combination would not have been obvious from the teachings of the prior art. This combination of properties is so totally unexpected from the art and is of such significance that we think that the board erred in sustaining the rejection of the appealed claims. The decision of the board is accordingly reversed.

Reversed.

MARTIN, Judge (dissenting).

Assuming arguendo that the decrease in sensitivity factor is unexpected, I cannot agree with the legal conclusion reached by the majority opinion since it fails to adequately weigh all the evidence in the case. Finding a single unexpected result to be conclusive of non-obviousness without weighing contrary evidence is an approach opposite to that so recently taken by this court in In re Cline, 345 F.2d 847, 52 CCPA ___, In re De Montmollin, 344 F.2d 976, 52 CCPA ___, and In re Graf, 343 F.2d 774, 52 CCPA 1206.

Both the appellant and the majority agree that 1) the very ethers used in the composition are old, 2) that the base gasolines used are also old, 3) that the concept of the combination is old, i. e., the concept of combining those ethers with other gasolines is old, and 4) that it is entirely expected that the ethers raise both the Research and Motor Method octane numbers.

What the majority finds conclusive of non-obviousness is the relative unexpectedness of the drop in sensitivity factor, that the ether has raised the Motor Method octane number more than the Research Method number, or, the latter less than the former, over the base gasoline, although it is acknowledged to be expected to raise both. While the amount of the reduction in sensitivity factor, about 1.5 points,1 goes to whether it is unexpected or not, I do not think that that single factor set against the others above listed is of such significance as to be conclusive of non-obviousness.

The majority opinion in effect stands for the proposition that a single unexpected result is conclusive of non-obviousness, even where the concept of combining the components, i. e., the ether with an equivalent gasoline, is old, and one, if not two, of the resultant properties are acknowledged to be entirely expected. That approach of the majority opinion does not give due weight to each and every bit of evidence, contrary to what I understand to be the law. In re Cline, supra, In re Montmollin, supra, In re Graf, supra. Thus, upon weighing the total circumstances of this case, I am compelled to say that the rejections made were proper and should be affirmed.

But further, I do not agree that the unexpectedness of the reduction in sensitivity is as clear as the majority assumes it to be.

The sensitivity factor would be of some significance if the "rule," "that the lower the sensitivity number, the better is the over-all performance of the gasoline composition," were not extremely general. The statement would be more accurate if the two gasolines compared had the same research octane number. Since it is common knowledge that the cars of today rarely operate at 600 RPM, most idling around or above that figure, the Motor Method octane number is the better index of all-around performance. Clearly if two gasolines have the same research octane number but one is less sensitive, that one will give superior over-all road performance since the Motor Method octane is higher.2 Yet the ethers used herein are acknowledged to raise the Motor Method octane number, as is entirely expected. Appellant states:

It is, of course, not suprising [sic] that the higher octane number ether raised both the Research and Motor Method octane numbers. * * *

The sensitivity factor, to me becomes more and more unreal as a truly significant factor, on which non-obviousness here is pegged.

The Evans et al. reference (not described in the majority opinion) is admitted by appellant to show the use of the claimed di-lower-alkyl ethers in gasoline. Evans et al. blend with a motor fuel:

* * * an agent which tends to reduce the knocking of the fuel in supercharged gasoline engines. [Emphasis mine.]

"Methyl tertiary butyl ether" is disclosed therein as such an agent; compare appellant's claim 5. The ether amount in the Evans et al. blend varies from "5% to 50% * * *." Evans et al. state:

We have now found that certain blends of ethers with hydrocarbons, which blends possess octane numbers not substantially greater than those of the hydrocarbons alone prove vastly superior to the straight hydrocarbons in supercharged engines. * * *

* * * * * *

As is known, supercharging is one method of markedly raising the power output of gasoline engines. This method comprises boosting the intake pressures to above atmospheric pressure, thus increasing the compression pressures. The degree of supercharging is limited by the knocking property of the fuel, for all fuels have a critical compression pressure above which they detonate spontaneously. Obviously the fuel which permits the highest degree of supercharging is the most desirable.

* * * * * *

It will be seen that the addition of various quantities of our ethers, while failing to improve the octane rating, markedly increases the allowable boosting of the intake pressure without causing knocking. [Emphasis mine.]

Thus it is clear to me that Evans et al. teach that the addition of appellant's ethers to a gasoline would be expected to reduce knocking under severe engine operating conditions (albeit conditions different than those of appellant's). The Evans et al. base gasolines, ether, and blend all have about the same octane number (about 100) based on an "ASTM" method. Regardless of which ASTM method is used in determining the same octane numbers for base gasoline and blend, the fact that the Evans et al. blend (gasoline plus ether) then endures an increase in severity of engine conditions which causes the reference iso-octane fuel to knock, clearly indicates the fundamental result Evans et al. teach is the same as that here on which non-obviousness solely depends. That is, Evans et al. teach that under engine conditions which more nearly approximate the rigors of true performance the blend of gasoline plus ether will perform better in terms of its resistance to knocking. Compare that to the fundamental result appellant seeks:

* * * when testing two gasolines of different compositions in the same engine one may find that both gasolines have just adequate knock-resistance in the engine at low speeds and that as the speed of the engine is increased, resulting in higher combustion chamber temperatures, one of the fuels is still adequately knock-resistant while the other detonates or knocks. The term "sensitivity" is used to define this difference among gasoline fuels; in this example the second gasoline has a higher "sensitivity" than the first gasoline, i. e., it is more "sensitive" to a change in the engine operation toward conditions of greater stress on the fuel.

Both Evans et al. and appellant show the same fundamental result attained by the addition of the ether, and I see no magic in defining it in terms of sensitivity factor. Cf. In re Summerson, 345 F.2d 222, 52 CCPA ___. Since I have not been shown that the board clearly erred in its weighing of all the facts, I would affirm.