Rich, Judge.
This appeal is from the decision of the Patent Office Board of Appeals affirming the rejection of claims 1-4, 7, and 8 1 in appellant’s application serial No. 458,289, filed May 24, 1965, for “Edge De-Burring Roll.” We affirm.
Subject Matter Olaimed
Appellant claims apparatus for removing edge burrs from thin metal strips. Such burrs may be formed at the longitudinal edges of strips as a result of slitting or shearing operations used in forming the strips from wider sheet stock. These burrs are asserted to unfit the thin strips for many uses (e.g., for use as the magnetic cores of motors, transformers, and the like), and appellant’s specification states that
For sheet metal less than, substantially 0.0S0 inch thick, * * * de-burring has been done by time-consuming and expensive hand operation since there was no entirely satisfactory equipment for this purpose.
Appellant’s solution to the above problem may be undestood from his figures 1 and 2:
[868]*868
In Fig. 1,10 is the metal strip, 13 and 14 are power-driven feed rolls, and 11 and 12 are opposed de-burring rolls. Fig. 2 is a fragmentary vertical cross-section taken along tbe line 2-2 in Fig. 1. From it one may see that the edge rolls are fabricated from two parts and that the peripheral groove in which the work pieces are deburred is symmetrical and has a half-angle A. Figures 4 and 5 show alternative groove cross-sections. In Fig. 4 the bottom of the groove is flat, and in Fig. 5 the bottom is rounded, but in each of the sides of the groove are angled from the flat, abutting surfaces by the half-angle A, as in Fig. 2.
Claims 1-4 are directed to appellant’s de-burring rolls per se. Each of these claims recites an edging roll comprising a pair of individual roll members and means to secure them together with their respective inner ends contiguous to one another. They further require that the mating portions of the roll members define a strip-receiving peripheral groove with inwardly converging inclined surfaces at an angle, with respect to a plane perpendicular to the axis of the roll member, not exceeding 15° (i.e., the half-angle A in Fig. 2).2 Claims 1 and 2 are generic in scope; claim 3 (reciting that the peripheral groove is “V-shaped”) is drawn to the species shown in Fig. 2; and claim 4 (reciting that the sides of the peripheral groove have “annular bottom surfaces”) is subgeneric to the embodiments shown in Figs. 4 and 5. Claim 7 defines a combination comprising the edging roll of claim 1, means to force an edge of the strip into the groove of the roll, means to feed the strip past the roll, and means to support the roll for rotation about an axis perpendicular to the plane of the strip. Claim 8, [869]*869which depends from claim 7, adds a second, eoplanar edging roll which is used to force the edge of the strip into the peripheral groove of the first roll.
The References
The references relied on are:
Canda_ 1,428,142 Sept. 5, 1922
Wilson et al. (Wilson)- 2,326,715 Aug. 10, 1943
Wilson is the principal reference. It discloses apparatus for the manufacture of thin metal strips for use in Venetian blinds. Although the Wilson disclosure is primarily concerned with means for center-stretching such strips, prior to being advanced into the center-stretching means, the strips are passed between a pair of “edge rolls” which “are provided for removing the sharp edges or burrs from the side edges of the strip and for centering and guiding the strip into the stretching apparatus * * *.” These “edge rolls” are shown in Wilson’s Fig. 3, in which the rolls are 10 and the thin metal strip is 7.'
The half-angle of the V-shaped groove 10a measures about 6° on this drawing, but the specification says nothing about the angle.
Canda teaches that rolls for a ball rolling mill may be formed from a plurality of hardened steel disks secured together by means of bolts. In his Fig. 2, below, each operating surface 5 is partly on each of two adjoining disks.
[870]*870Furthermore, Canda expressly indicates that his “method of constructing rolls may be applied to other types, [of rolls], as for instance, such rolls having V-shapod grooves * *
The Rejection
The examiner rejected claims 1-4, 7, and 8 under 35 USC 103 as unpatentable over Wilson in view of Canda. It was his position that
Xlie only difference between Wilson’s deburring rolls and the recited rolls is that Wilson’s rolls are of a one piece construction while applicant’s recited rolls comprise two members releaseably secured together by a series of bolts.
This, difference he thought obvious in view of Canda’s similar expedient in the ball-rolling apparatus.
The boar d affirmed the examiner’s rej ection, stating that
, 'It appears to us that the Wilson-et al. teaching of forming rolls 10 is adequate to. direct persons .of ordinary skill in the art to the use of rotatable edge contacting members having the forming groove angularity claimed. Such teachings have not been challenged' by appellant. We agree with the Examiner that it would be obvious to form such rolls of two parts with the separation at the apex of the forming groove. In the formation of such a grooved forming roll the problems of fabrication appear to us to reasonably lead to the use of two abutted parts as the obvious way of providing the required inner sharp angled recess. Machining of the roll 10 of Wilson et al. from an integral blank would present problems which are commonly avoided in the preparation of such shapes by the use of separable parts.
Appellants Arguments
Appellant makes four arguments. First, lie argues that the board erred in that it'
* » * treated diagrammatic drawings of the principal reference, Wilson et al., as working drawings, scaled the drawings to find a disclosure meeting of critical maximum angular limitation in all of appellant’s claims but otherwise not disclosed in the specification of Wilson et al. either as to that critical maximum angle or its function and effect.
In support of this proposition, appellant cites In re Bager, 18 CCPA 1094, 1097-98, 47 F. 2d 951, 952-53, 8 USPQ 484, 486 (1931), In re Wilson, 50 CCPA 827, 834, 312 F. 2d 449 454, 136 USPQ 188, 192 (1963), and In re Chitayat, 56 CCPA 1343, 1346, 408 F. 2d 475, 478, 161 USPQ 224, 226 (1969). Second, he argues that
The reliance on Canda * * * assumes the untenable premise that the art knew what appellant alone teaches, namely, the maximum total groove angle permissible for de-burring by coining thin metal strip was 30°, and the problem was merely how to obtain it.
Third, he argues that
* * * the Board wholly overlooked that none of the references, singly or in combination, suggest the combination called for in Claims 7 and 8 * * *.
[871]*871And fourth, lie similarly argues that
* * * the Board wholly overlooked that none of the references, singly,or in combination, suggest * * * the specific groove root configuration called for by Claim 4, sub-generic to appellant’s embodiment [s] disclosed in Bigs. 4 and 5 (R. 15) of appellant’s drawings. • .
Ofinion
Answering the first argument, as we said in Wilson, supra, “Patent drawings are not working drawings * * However, we did not mean that things patent drawings show clearly are to be disregarded. In re Eager, also cited by appellant, is an example of a case in which the teachings of patent drawings, even as to features unexplained by the specification, proved dispositive. As this court there said, “Description for the purposes of anticipation can be by drawings alone as well as by words.” In Wilson, as our opinion emplia-' sizes, the attempted reliance was not only on a patent día wing per se, it was on “a greatly enlarged section of a small dr awing, obviously-never intended to show the dimensions .of anything.” Here, Fig. 3 in the Wilson reference focuses on the edge rolls, showing them with great particularity and showing the grooves thereon to. have an angularity well within the range recited in appellant’s claims, Wilson, therefore, shows this aspect of subject matter lying within appellant’s claims tó be old.
Appellant’s second argument stresses the criticality, of the groove-angle maximum recited in his claims. However, claims are unpatentable when they are so broad as to read on obvious subject matter even though they likewise read on non-obvious subject’ matter. Here, appellant’s claims read on edge rolls having a 6°' groove angle and constructed from two mating parts. Wé have found that Wilson shows an edge roll having a 6° groove angle, and appellant no longer seriously relies on the two-part construction limitation, Ap-’ pellant’s claims are therefore unpatentable under § 103. Appellant’s alleged discovery that the maximum groove angle for effective deburring is 30° cannot affect this result.
As for appellant’s third argument concerning claims 7 and 8 in particular, we agree with the solicitor that Wilson clearly discloses (in a drawing we have not reproduced) a coplanar pair of edge rolls, means to feed the thin metal strip past the rolls, and, at least implicitly, means forrotationally supporting the rolls. Wilson thus shows the combination in which there is no novelty except the details' of the roll of claim 1, which we have already held to be obvious. ’ The added elements of these claims add nothing, so far as patentability is concerned.
As for appellant’s fourth argument which is predicated on-the annular groove root configuration parallel to the periphery recited in [872]*872claim 4, the abbreviated record we have on appeal does not indicate that appellant argued that there was any patentable significance to that feature before either the board or the examiner. As we said in In re Touvay, 58 CCPA 809, 811-12, 435 F. 2d 1342, 1344, 168 USPQ 357, 359 (1971), “Ordinarily we do not consider arguments directed to the significance of a particular limitation at this late stage unless they were raised below.” Here, as there, we find that “The circumstances of this appeal d'o not warrant deviation from this rule.”
The decision of the board is affirmed.
