delivered the opinion of the court:
This appeal is from the decision of the Patent Office Board of Appeals .affirming the examiner’s rejection of claims 1, 2, and 4-18 in appellant’s application serial No. 321,353, filed November 4,1963, for ultrafine particles of polytetrafluoroethylene and a method for making them. We reverse.
THE INVENTION
Polytetrafluoroethylene (hereinafter “PTFE”) is the plastic well known under the trademark “Teflon.” We are told that it is sold commercially in both powdered and powdered ultrafine form and that the powdered ultrafine PTFE as is and the powdered PTFE after further comminution are used to fabricate shaped articles. PTFE powder does not melt and flow even when heated under pressure, but ultrafine PTFE particles can be shaped under high pressure in a mold to a preform sufficiently strong and coherent to be removed from the mold. The preform can then be heated in a furnace to a temperature at which sintering occurs, resulting in coalescence of the particles into usable articles of the same shape as the preform.
*1184Appellant claims PTFE in finely powdered, non-fibrous form and a method of preparing the powdered product consisting of grinding the PTFE powder commercially available in a particular type of air mill which is concededly old for other purposes. Commercially available PTFE powder has been ground in other types of air mills into ultrafine, fibrous particles before premolding, but appellant maintains that grinding such powder in this particular type of air mill, which mills and classifies sequentially, rather than simultaneously as in the previously used air mills, results in non-fibrous, ultrafine PTFE powder with significantly improved preform strength (unsintered flex strength). This is said to be advantageous because it reduces scrap reworking due to cracking on removal from the mold. Claims 1 and 2 are reproduced as illustrative, with subparagraphing supplied and the recitations principally in controversy emphasized:
1. A method for preparing polytetrafluoroethylene adapted for molding precision parts and thin sheeting, comprising the sequential steps of
(a) subjecting particles of polytetrafluoroethylene, which particles are at least 100 microns in their smallest dimension, to milling effected substantially solely by interparticulate collision, at a temperature of less than about 200° F;
(b) classifying the milled particles to separate therefrom particles having a maximum dimension of up to and including about 50 microns, wet-sieve size; and
■(e) subjecting the unseparated particles of larger than about 50 microns, wet-sieve size, repeatedly to the milling effected substantially solely by in-terparticulate collision, at a temperature of less than about 200°F.
2. Ultrafine polytetrafluoroethylene adapted for molding precision parts and thin sheeting comprising
porous non-fibrous particles with rounded countours having
a particle size, wet sieve, no greater than 50 microns, said polytetrafluoroethylene material having
■a distribution function of no greater than about .40,
a sub-sieve size of no greater than about 5.0 microns,
a ratio of wet-sieve size to sub-sieve size in the range of about 2 to about 10;
having, upon molding to a zero void content at 6,000 psi,
an unsintered, fleco strength of at least 860 psi; and
having, after sintering,
a tensile strength of no less than about 4000 psi,
'a percent elongation of no less than about 300,
an anisotropic expansion factor of no greater than about 1.13,
a dielectric strength of no less than about 1200 volts per mill,
a surface roughness of no greater than about 82 microinches at 500 psi and no greater than 32 microinches 'at 2000 psi and
a percent void content of no greater than 0.1% at pressures greater than 2000 psi.
*1185THE REJECTIONS
The references relied on are:
Thomas et al_ 2, 936,301 May 10,1960
Wallace_3,178,121 Apr. 13,1965
Filed Apr. 24,1962
The examiner rejected all claims now appealed “as indefinite under 35 U.S.C. 112,” reasoning that, “If a product is to be claimed, it must be defined in terms of its own properties” and that “Applicants’ [sic] recitation of an unsintered flex strength is not * * * indicative of the properties of the PTFE powder, rather of the unsintered premolded article the powder will be used for.” He additionally rejected claims 1, 2, 4-8, and 15-18 as unpatentable over Thomas under 35 USC 102 on the ground that “These claims are not considered to define over powder G of Thomas et al in that unsintered flex strength does not define the PTFE claimed” and all claims as unpatentable over Thomas under 35 USC 103 on the ground that:
It would be obvious to one skilled in the art to employ any mill, including a “Jet-O-Mizer” [1] to prepare the powdered PTFE products of Thomas et al. Applicants’ [sic] affidavit does [not] present any significant results that would refute this contention. Routine experimentation by one skilled in the art would lead to the use of any air mill to grind PTFE in view of the fact that Thomas et al disclose air mill grinding of PTFE and more particularly to a “Jet-O-Mizer” admitted by applicant to be an old milling device.
The board affirmed all three rejections. However, it advanced what we consider to be three separate rationales for the section 112 rejection. The first, which it denominated “The more important point,” was that the unsintered flex strength recitation is “directed at a result in the preform that is obtainable from the claimed powder upon forming it” and is therefore “Obviously * * * not helpful in defining the particulate composition or the method for producing it.” Second, even if such recitation might be permisible as a last resort when the product could not otherwise be identified,
* * * the characteristics of the composition which contribute to the resulting flex strength and the steps of the process used in producing it are all physical or mechanical, presenting no apparent difficulties in direct identification * * * [and] thus require no resort to indirect modes of identification.
Finally, the flex strength limitation is indefinite because it “fails to recite the temperature of molding and the compositional characteristics of the material molded, which obviously will profoundly affect the preform strength.”
*1186As to the prior art rejections, the board affirmed both the section 102 and section 103 rejections without clearly distinguishing between them. However, it would seem that the board, like the examiner, first determined the flex-strength recitation to be impermissible, thereafter ignored it, and found everything else to have been anticipated by Thomas. As far as it considered the possibility that appellant’s product and process were “not identically disclosed or described” in Thomas, it focused on the degree of interparticulate collision in appellant’s mill and the mill disclosed in Thomas, which is a characteristic on which appellant expressly did not predicate patentability, and ignored sequential versus simultaneous classification, the characteristic on which appellant did rely for patentability. Additionally, the board questioned the sufficiency of appellant’s comparisons, which indicated that a flex strength of 828 psi was the maximum obtainable by the method disclosed in Thomas in contrast to the minimum of 860 psi recited in appellant’s claims, but held that:
* * * even assuming a valid comparison in all respects, a difference between 828 and 860 in the “unsintered flex strength” would hardly have been an unobvious one in the use of an apparatus more effective in the degree of interpartieulate reduction achieved.
OPINION
The Section 11%, Rejection
The principal rationale for the section 112 rejection was simply that unsintered flex strength is a property of compacted preforms made from PTFE powder rather than a property of the powder itself, and that only properties of the powder, as a powder, could be used in defining the powder claimed. Appellant argues that the powder’s capability of being compacted into a preform having at least the specified “unsintered flex strength” of the preforms is a property of the powder which may be used to define the powder.
The first sentence of the second paragraph of 35 USC 112 requires only that claims “set out and circumscribe a particular area with a reasonable degree of precision and particularity.” In re Moore, 58 CCPA 1042, 439 F. 2d 1232, 169 USPQ 236, 238 (Mar. 31,1971). In the absence of evidence to the contrary, we will assume, as we said in Moore, that what the claims define is what the applicant regards as his invention. If those skilled in the art can tell whether any particular PTFE powder is or is not within the scope of a claim, the claim fulfills its purpose as 'a definition. As we remarked recently in reversing a section 112 rejection made on the ground that particular language was “functional” and thus “indefinite,” “we are unable to see merit in any proposition which would require the denial of a claim solely because of the type of language used to define the *1187subject matter for which, patent protection is sought.” In re Swinehart, 58 CCPA 1027, n. 4, 439 F. 2d 210, 169 USPQ 226, 228 (Mar. 31,1971). See also In re Wakefield, 57 CCPA 959, 967, 422 F. 2d 897, 904, 164 USPQ 636, 641 (1970) (holding the scope of a claim to be definite “because each recited limitation is definite”).
In this case, it seems to us that those skilled in the art will be in no uncertainty concerning what subject matter falls within the scope of the claims.. The claims in question recite essentially that the PTFE particles have, upon pressure molding to a zero void content at 6,000 psi, an unsintered flex strength of at least 860 psi. It is clear that PTFE particles which exhibit an unsintered flex strength of less than 860 psi, when subjected to such compression, do not fall within the scope of the claims even if they meet the other claim requirements, while those that exhibit an unsintered flex strength of 860 psi or more When so compressed, and meet other requirements, do. The mechanics of determining the unsintered flex strength of any given batch of PTFE particles are set forth in the specification, and they do not appear to require unreasonable effort. Accordingly, we hold that the challenged language is not fatally defective solely because it recites a property which can be measured only after the PTFE powder has been compacted into a preform.
However, the board also reasoned that the flex strength limitation is indefinite because it does not contain “the temperature of molding and the compositional characteristics of the material molded, which obviously will profoundly affect the preform strength.” Appellant has argued that “in the art, unless a temperature is specified it is generally understood that room temperature is intended.” If this is so, of course, the claims should be read as though the room-temperature limitation were expressly contained therein. However, appellant did not ask for an opportunity to submit evidence supporting the above assertion, and we will not take judicial notice of its accuracy. Nevertheless, we cannot agree with the board that the omission of the temperature of molding 2 renders the claims indefinite. Even if it is not true, as appellant asserts, that it is generally understood in the art that omission of temperature from such a recitation indicates that room temperature is intended and the claims are therefore broader than they otherwise would be, breadth is not to be equated with indefiniteness, as we have said many times.
The Prior Art
Thomas discloses PTFE in ultrafine, fibrous form and a process for preparing this product by comminuting commercially available *1188PTFE powder in a rotary blade type grinding and classifying apparatus. To demonstrate the superiority of their particulate PTFE over the prior art, Thomas et al. compare their product to, among others, particulate PTFE “obtained by subjecting * * * [a commercial grade of PTFE granular powder] to the disrupting action of high velocity gas jets in a ‘Micronizer.” According to Thomas, the result is a powder which “can be readily leveled and compressed, but [which] yields fragile preforms.”
The examiner’s “use of the Wallace patent,” according to the board, was “merely explanatory of the [reference to a] ‘Micronizer’ ” in the above quotation. A “Micronizer,” as described in Wallace, is a fluid energy mill which “has a circular disc-shaped working chamber in which both grinding and size-classifying steps are carried out simultaneously.”
The /Section 102 Rejection
As above mentioned, the basis of both the examiner’s rejection under 35 TJSC 102 and the board’s affirmance thereof seems to have been that since the unsintered flex strength recitation in appellant’s product claim is improper it can be ignored and that, with the un-sintered flex strength recitation out of the way, what was left did not “define over” the Micronizer-prepared powder disclosed in Thomas. In this respect, the section 102 rejection in this case is not unlike the section 103 rejection which we reversed in In re Wilson, 57 CCPA 1029, 424 F.2d 1382, 165 USPQ 494 (1970), and is unsupportable for the same reason: “All words in a claim must be considered in judging the patentability of that claim against the prior art” (emphasis added). Id. at 1032, 424 F.2d at 1385, 165 USPQ at 496. Moreover, for reasons we have already set forth, we see nothing improper about appellant’s recitation of unsintered flex strength as one of the properties of the powder he claims.
Of course, Thomas does not quantify the unsintered flex strength of preforms made from the PTFE powder he milled with a Micronizer air mill, but appellant has submitted evidence, which we will consider more fully in the next section, proving that the unsintered flex strength of Thomas’s air-milled powder could not have been as high as the unsintered flex strength recited in his claims. Accordingly, we will not affirm a section 102 rejection on the evidence in this case.3
*1189 The Section 108 Rejection
The logic of the rejection under 35 USC 103 was (1) that, since Thomas disclosed that commercially available PTFE powder could be further comminuted prior to molding and sintering either in some form of hammer mill or in a Micronizer air mill, it was prima facie obvious to use any prior art air mill for the same purpose and (2) that appellant’s evidence of the superiority of the particular type of prior-art air mill he used did not rebut the prima facie case. We do not understand appellant to controvert the first step in this reasoning, but he does vigorously assert that his affidavit evidence together with the showing in his application is sufficient to overcome the inference of obviousness to be drawn from the similarity of his process and product disclosed in Thomas.
Appellant’s application and affidavit compare ultrafine PTFE powder prepared in a commercially available, prior-art air mill (a Jet-O-Mizer) of the type recited in his process claims with ultrafine PTFE powder prepared in a commercially available, prior-art air mill (a Micronizer) of the type disclosed in Thomas and which his application indicates was conventionally used in the art ‘prior to his invention.4 The difference between these two devices which is of importance here is that Jet-O-Mizer type air mills sequentially grind and classify the particles on which they operate, whereas Micronizer-type air mills siirmltcmeously grind and classify. To effectuate this difference, Jet-O-Mizer air mills are shaped like elongated, hollow toroids, in one end of which grinding takes place and in the other, classifying, whereas Micronizer air mills, as previously mentioned, have a circular working chamber in which both grinding and classifying are carried out simultaneously.
According to appellant’s application, “The highest flex strength achieved for the circular air mill material [in 6 runs] was 737 psi whereas the lowest flex strength for the elongated toroid air milled material [number of runs unspecified] was 860 psi.” The examiner, however, was not convinced by the figures in the application, so appellant filed his own affidavit, in which he set forth his own qualifications as an expert in PTFE technology, including the assertion that he had “worked extensively with particle grinding and classifying apparatus,” described further testing done under his supervision on Mi-cronizer apparatus “with the objective of preparing a PTFE powder *1190falling within the characterization given by Thomas et al. * * * [for the powder they had prepared in a Micronizer] ” and with the further objective of attempting
* * * to maximize the unsintered flex strength of the resulting “Micronized” PTFE product, by controlling', to the best of his ability, those conditions which he would have expected to favorably influence unsintered flex strength; * * *
and asserted that he was “unable * * * to obtain values for unsintered flex strength in the ‘Micronized’ product, of 860 psi. or greater * * According to the table setting forth his results, the highest value he was able to obtain was 828 psi.
In his final action, the examiner accepted appellant’s affidavit “as establishing the limit of a ‘Micronizer’ in providing unsintered flex strength data.” In his Answer, however, the examiner questioned the sufficiency of the affidavits on two grounds: first, that the number of runs made with the Micronizer “may not have established enough data to prove anything” and, second, because “Wallace discloses that the Micronizer’ is capable of very effective interparticulate collision allegedly responsible for the unsintered flex strength of the premolded article.”
Despite appellant’s protestations in his reply brief before the board that he did not rely on the recitation that, in Ms process, “milling [is] effected substantially solely by interparticulate collision” to distinguish his process from prior art Micronizing operations, the board stated that “For the reasons given by the Examiner, * * * [it was] not impressed that the results as to ‘unsintered flex strength’ shown in the Rule 132 affidavit represent the best obtainable in the use of the Micronizer’ ” (emphasis supplied). Moreover, the board criticized the affidavit for not indicating that the conditions of the runs (air pressure, input particle size, etc.) compared were identical.
Taking these objections in reverse order, we note that appellant was not trying to prove that, for the same process limitations, the use of a Jet-O-Mizer type air mill would always produce ultrafine powder having a higher unsintered flex strength than ultrafine powder produced by a Micronizer-type mill. Rather, he was trying to prove that by using a Jet-O-Mizer type air mill with appropriate process limitations a person skilled in the art could obtain ultrafine PTFE powder wMch exHbited better unsintered flex strength than could be obtained by a person skilled in the art using a Micronizer-type air mill and the optimum process limitations therefor. Thus, we think it of no moment that the process limitations in the Jet-O-Mizer runs and the Micronizer runs were not identical. The Patent Office has not shown that the optimum results with these two air mills are obtained when using the same process limitations. Appellant has sworn that the limitations used *1191with a Micronizer were chosen with a view to maximizing the unsin-tered flex strength of the resulting powder, and we have no reason to doubt that appellant, a person of at least ordinary skill in the art, was successful in so doing. Certainly the statement in Thomas that powder comminuted in a Micronizer “yields fragile preforms” supports this view.
As to the number of runs, we note that Example III in appellant’s specification sets for the results of six runs which were made under conditions which appellant swore in his affidavit were, in his opinion, “suitable for preparing a milled PTEE product with good properties, particularly with a property of high unsintered flex strength” and that Table II in the affidavit itself sets forth the results of three runs in which appellant swore that he “attempted to maximize the unsintered flex strength of the resulting ‘Micronized’ PTFE product * * In this ex parte action, too much should not be required of an applicant to rebut the Patent Office’s mere suspicions, based on no suggestion in the prior art of record, that some other combination of process limitations might result in higher unsintered flex strength. We accept appellant’s evidence as establishing that his invention enables those skilled in the art to produce PTFE preforms of unsintered flex strength greater than the strongest preforms obtainable with the prior art method.
The final question is whether the difference shown is sufficient to establish non-obviousness. Appellant’s evidence indicates that, by using an air mill which sequentially grinds and then classifies, he has been able to produce PTFE preforms having an unsintered flex strength of as much as 1100 psi, but his product claims encompass all ultrafine PTE powders which, in addition to meeting the numerous other claims criteria, have an unsintered flex strength of over 860 psi. The difference between the claim lower limit of 860 psi and the maximum appellant was able to obtain using prior art techniques is only a little less than 4%. However, the evidence clearly indicates that an increase in unsin-tered flex strength is of great practical utility in the PTFE molding art, since weakness in prior art preforms led to cracking on transportation from the mold to the sintering furnace, causing extensive scrap reworking. Under the circumstances, appellant having demonstrated substantial improvement in unsintered flex strength due to his invention and the prior art failing to show how to attain even the minimum encompassed by the claims, we think he has established unobviousness. Compare In re Fisher, 51 CCPA 1099, 427 F.2d 833, 166 USPQ 18 (1970).
Accordingly, we reverse.
1 Appellant’s specification discloses that a mill of the type employed in his process is commercially available under the trade name “Jet-O-Mizer.”
We do not understand the implication in the board’s opinion that the claims omit “the compositional characteristics of the material molded.” On the contrary, the claims seem to set forth “the compositional characteristics of the material molded” in great detail.
Appellant has also pointed out that the rejection under 35 USC 102 of claims 1, 6-8, and 15-18 is not sound because these process claims recite that the milling is carried out in sequential steps, whereas in a Micronizer air mill, grinding and classification take place simultaneously in the same zone. Whether this difference is sufficient to make appellant’s invention as recited in these claims non-obvious in view of the prior art will be discussed in the next section, but appellant's sequential process is clearly not “identically disclosed or described” in Thomas becasue of that difference.
Appellant’s application also compares his powders with commercially available ultra-fine powders and with ultrafine powders obtained by the use of various types of hammer mills. While his tables indicate that his powders were superior in unsintered flex strength to either of the above, this evidence need not be considered further because the section 103 rejection of both product and process have been made solely on the basis of the prior-art air milling and air-mill-produced ultrafine powder.