This appeal is from the decision of the Patent Office Board of Appeals, adhered to on reconsideration, affirming the examiner’s rejection of article claims 1-6, all of the claims of application serial No. 535,835, filed March 21, 1966, entitled “Tantalum Carbide or Tantalum-Alloy Carbide Filament Mounting and Method.” We reverse.
The Invention
The claimed invention is a mounting structure for tantalum or tantalum-alloy carbide lamp filaments, described as follows in appellant’s brief before the board, with reference to Figs. 1 and 2 of appellant’s drawings below:
The filament comprises a coiled, elongated, incandescible filament (14) principally comprising tantalum carbide which has a coiled end extension (34). The coiled end extension (34) snugly overfits at least a part of a relatively thick, elongated tantalum carbide member (32). The inner surface of the overfitted turns of the coiled end extension (34) are welded to the relatively thick, elongated member (32). The relatively thick member (32) in turn is electrically connected to and mechanically supported by the main filament support and electrical connection means, such as the supporting frame (20). Because of its size, the member (32) can be readily clamped [between rigid tantalum supports 36] or otherwise supported without fracturing. The method for fabricating the filament and relatively thick member of tantalum carbide is * * * briefly, a tantalum metal, coiled end extension (34) is overfitted onto a relatively thick tantalum metal member (32). The two members are simultaneously carbided prior to incorporation into the lamp and during the carburization, the weld or bond is effected between the inner surface of the overfitted turns (34) and the relatively thick member (32).
*1250Claim 1, the sole independent claim, reads:
1. A combination connection for mechanically supporting and electrically connecting a coiled elongated in-eandeseible filament principally comprising tantalum carbide to a main filament support and electrical connection means, said connection comprising:
(a) a relatively thick elongated member principally comprising tantalum carbide;
(b) a coiled end extension of said elongated filament coil snugly overfit-ting at least a part of said relatively thick elongated member, the inner surface of the overfitted turns of said end extension coil welded to said relatively thick elongated member; and
(c) said relatively thick elongated member electrically connected to and mechanically supported by said main filament support and electrical connection means.
For the purposes of this appeal, the dependent claims need not be discussed.
The References and the Rejection The references relied upon are:
Allen 2,434,478 Jan. 13, 1948
Cooper 3,022,437 Feb. 20, 1962
Rlvely et al. (Rlvely) 3,183,395 May 11, 1965
Bird 3,237,284 Mar. 1, 1966
Sloan 3,287,591 Nov. 22, 1966
(Filed Dec. 13, 1961)
Allen discloses a method of attaching a projection lamp filament to its support members as shown in its Figs. 2 and 3:
Coiled end extension 39 of filament 38 of unspecified composition is slidably fitted on the support member 44 of molybdenum. The end extension is then bonded to the support member by conductive paste 62.
Cooper discloses attaching a coiled tantalum-alloy filament to tantalum-alloy leads by an unspecified method which appellant agrees might be spot welding. The filament and leads are assembled in a bulb, such as a sealed-beam headlamp, that contains a volatile hydrocarbon and hydrogen atmosphere. Cooper states:
The filament may then be converted to the carbide form by passing sufficient current through the filament to yield a filament temperature of about 3100 °C. In this way, the tantalum and tungsten, for example, comprising *1251the filament and lead ends associated with the filament will be substantially converted to carbides.
Bird is similar to Cooper except that it additionally includes a fine wire of smaller diameter (1 mil or less) helically wound around the larger diameter (5 to 7 mils) filament wire. The fine wire is cemented to the filament wire by an organic hydrocarbon thermoplastic which, on charring, produces carbon. A bond is formed between the wires when they are converted to carbides by self-heating as disclosed by Cooper.
Rively discloses a module assembly for projection lamp filaments of undisclosed material in which the filaments are spot welded to their leads.
Sloan discloses carburizing a tantalum filament coil in a furnace. Tantalum carbide support legs are cemented to the tantalum carbide coil in accurate orientation and molybdenum leads are cemented to the legs to complete the filament. Next, current is passed through the filament (enclosed in a hydrogen and hydrocarbon atmosphere) to “set the cement and sinter the tantalum carbide legs to the coil and leads.” The connection between the leads and legs is arc-welded with a carbon or tungsten electrode in an atmosphere of inert gas, such as argon, since electrode contact from spot-welding “could destroy the brittle filament.” However, it is to be noted that Sloan further states (reference numerals omitted for clarity):
The arc welding makes an effective permanent joint between the leg[s] and the support wire[s]. When the joint is simply heated by passage of current through it and [the] coil, the cement may not become sufficiently hot to siiiter the cement and set it properly, because the lead-in wires are thicker than the filament of [the] coil. For this reason arc welding is superior.
The cement used to affix the coil to the legs is nearer the hot coil and so will be more readily heated to the sin-tering point. However, even there, an improvement is effected by arc welding.
The examiner rejected the claims as unpatentable under 35 U.S.C. § 103 over Allen in view of Cooper and Rively. He considered it obvious to bond the extension 39 of Allen’s filament to support member 44 by spot welding, especially in view of Rively. He also considered it obvious to make the Allen filament and filament support of tantalum carbide in view of Cooper.
In affirming, the board first found appellant to have admitted in his specification that tantalum carbide and'tantalum-alloy carbide filaments were known in the art. It also regarded statements in the specification that “a very thin tantalum carbide filament is quite difficult to weld to a supporting member” and that welding of such filaments to supporting members is “deleterious to the life of the filament” to constitute a concession that welding of such carbu-rized filaments is possible. The board relied on the “admissions” and the Allen and Cooper patents. In its reasoning, as will appear from the statements quoted below, the board suggested two ways of making appellant’s structure:
In our opinion, to one ordinarily skilled in the art appellant’s admissions and the Allen and Cooper patents would suggest making Allen’s members 39, * * * [and] 44 * * * of tantalum or tantalum alloy as in appellant’s admissions or as in Cooper; welding them together as in appellant’s admissions instead of cementing them together as in Allen, the welding necessarily being between the interior of the coils and the exterior of the elongated member; and then carbiding the tantalum or tantalum alloy members as in appellant’s admissions or as in Cooper, thus to provide the substance of appellant’s claim 1.
•X- -X- -X- -X- -X- -X-
Furthermore, in our opinion, to one ordinarily skilled in the art appellant’s admissions and the Allen and Cooper *1252patents would suggest making Allen's members 39 * * * [and] 44 -» * * 0f tantalum carbide or tantalum alloy carbide as in appellant’s admissions or as in Cooper; and welding them together as in appellant’s admissions instead of cementing them together as in Allen, the welding necessarily being between the interiors of the coil turns and the exteriors of the elongated members, thus to provide the substance of appellant’s claim 1.
Appellant filed a request for reconsideration accompanied by an affidavit asserting that the elements of Allen, if made of tantalum carbide, could not be welded as discussed by the board. In its response, the board declined to change its decision and additionally suggested yet another way to provide appellant’s structure, stating:
It is further our opinion that, in view of appellant’s admissions and the Cooper patent, to one ordinarily skilled in the art it would be suggested to make Allen’s member 44 of tantalum carbide and his filament 38, 39 of tantalum, welding these elements together, and then carbiding the filament 38.
* -X- -x- # * *
In our opinion, appellant has not proved that at the time of his invention the obvious claimed structure could not have been made by one ordinarily skilled in the art by some method which was known or was obvious in view of the cited and admitted prior art.
In his brief, the solicitor argues that the claims are obvious in view of Allen, Cooper, Bird, and Sloan.
OPINION
It is apparent that the board based its conclusion that appellant’s claimed structure was obvious largely upon reasoning that it could be made by methods which it found suggested by Cooper and Allen and the “admissions” it found in appellant’s specification. We do not think these references and “admissions” show the structure to have been obvious, and, in particular, we do not find that the methods disclosed in Cooper and Allen would have made that structure obvious.
It is true that appellant did concede that incandescent lamp filaments of tantalum and tantalum-alloy carbides are old. But that is about as far as any admissions made by him are of any significance. Certainly he did not concede that it would be obvious to weld such carbide filaments to supports of the same composition; he has instead steadfastly urged that spot welding a tantalum carbide filament to a support is not practical because the filament tends to shatter during the process. That appellant did not state such a weld was impossible, while maintaining it was difficult, plainly does not amount to the admission which the board apparently found.
The desirability of making the Allen filament 38, 39 of tantalum or tantalum-alloy carbide might very will have been apparent to a person skilled in the art. However, we do not find Cooper to suggest either that both the particular filament structure, including both its coiled end 39 and support member 44, be made of such material or that they could be welded together as required by the claims.
Appellant submitted an affidavit of Johansen, a highly trained scientist experienced in the development of incandescent lamps which utilized tantalum carbide filaments. The affiant states that it would be “technically impossible” to provide the claimed structure by first fabricating the filament and supports of tantalum or tantalum-alloy and then heating them by passing current there-through in series to carburize them, the first method suggested by the board, because the support members, being relatively large cross-section and relatively low resistance compared to the filament, would not be heated to a high enough temperature. Affiant further states that making the filament and supports of tantalum carbide and then welding them together, the board’s second sug*1253gested method, would be impossible “due to the extreme brittleness of tantalum carbide and the difficulties encountered in welding same.”
We find nothing in the prior art relied on by the board to indicate that a person of ordinary skill in the art would have any opinion contrary to the affiant’s on these points. While Cooper converts his tantalum filament to carbide form by self-heating in a suitable atmosphere and refers to the “lead ends associated with the filament” as also so converted, it is apparent that the latter conversion would not result in a lead “principally comprising tantalum carbide” as required by the claims. We therefore do not agree that a person skilled in the art would have found it obvious to modify Allen by either of the first two methods suggested by the board.
The board’s third method, suggested after the submission of the affidavit directed against the first two, was to make Allen’s support member 44 of tantalum carbide, assemble a tantalum filament thereon, and then convert the filament to carbide form and weld the members together. We cannot determine from the record whether this method would be operative or not. But, as we have already noted, Cooper does not suggest the particular support and coiled filament end be formed of two carburized materials with the inner surface of the turns of the filament coil welded to the exterior of the support member. The ability of the board to originate a method, not suggested by the prior art, to make the structure derived from appellant's own disclosure, does not demonstrate that the structure would have been obvious.
Only the Allen, Cooper, and Rively references were discussed in the examiner’s Answer and the board’s opinion. The solicitor, in his brief, undertakes to apply a new rationale to support the rejection based on the Bird and Sloan patents, applied neither in the examiner’s Answer nor in the board’s opinion. While appellant has expressed no objection to our considering these references and their application by the solicitor, it is too well settled to require citation of authority that we review only rejections made by the Patent Office in its examining and internal appellate functions and do not consider grounds of rejection presented for the first time in the brief of its solicitor. See In re Nygard, 341 F.2d 924, 52 CCPA 1032, 144 USPQ 586 (1965).
The decision of the board is, accordingly, reversed.
Reversed.