RICH, Judge.
This appeal is from the Patent Office Board of Appeals decision affirming the rejection of all claims in application serial No. 833,023, filed August 11, 1959, for “Structural Elements for Metal Framing System.” This appeal is taken only on claims 48-54, the appeal of claims 37-47 having been withdrawn.
This application is a continuation-in-part of serial No. 806,969, filed April 16, 1959, which was a continuation-in-part of the parent application serial No. 317,-128, filed October 27, 1952. The parent case was carried through an appeal to the Board of Appeals which, on February 28, 1958, reversed the examiner’s rejection of three claims, directed to the same invention as the claims here on appeal but somewhat narrower in scope by reason of the inclusion of what appellant, after that board decision, came to regard as unnecessary limitations, not necessarily part of the novel subject matter sought to be protected. Before the intermediate application was filed a Notice of Allowance had been issued in the parent case but it was nevertheless abandoned.
The Invention
Broadly, the invention relates to metal framing systems which can be built up from stock structural members, nuts, bolts, and fittings to construct a variety of things such as storage racks for barrels, tires, bar-stock, pipe, and the like, billboards, catwalks, hand rails, tables, machine guards, ramps, scaffolding, shelving, tables and benches, partition walls, and even entire small buildings. Such systems appear to be adult versions of the childhood toys known as “Erector” or “Meccano” sets.
[366]*366The background of the invention is said to have begun with a system known as “Unistrut” described in U. S. Patent 2,345,650, issued April 4, 1944, to Charles W. Attwood, father of appellant. James W. Attwood, appellant’s brother, who has given affidavits in this case, is an officer and general manager of Unistrut Corporation. Appellant developed the invention at bar as a research engineer with Attwood Development Company which was formed by Charles W. Attwood “to conduct development work in general research in fields relating to Metal Framing.” There have been sales of “Unistrut” of over $80 million in the fifteen years it has been on the market and it now appears to be in competition with some 28 other manufacturers of metal framing systems.
Since the present invention was conceived as an improvement on “Unistrut,” we will indicate the nature of the latter. The basic element of the system is an elongated U-shaped channel member with inwardly turned edges, provided in various depths and metal gauges. These members can be cut to desired lengths with a hacksaw and joined to other pieces by means of nuts and bolts and various angle fittings, links, braces, etc. The basic fastening concept is to place a special nut in the channel with its ends underlying the inwardly turned edges thereof and to screw the bolt into it after passing it through a hole in the member to be joined. The nuts may have grooves to engage the edges of the channel. Thus the only tools required for erection of a structure from the parts of the system are a hacksaw and a wrench. It is said that some 2500 different fittings out of 10,000 that have been designed are regularly stocked, wherefore the system is highly adaptable to varying needs. It will be appreciated that since the nuts initially slide in the channels and can engage the inwardly turned lips at any point, the relative placement and connection of two members is infinitely variable. There is in the art, however, and apparently in competition with “Uni-strut,” another type of framing member which is provided with rows of holes or perforations of various shapes through which bolts may pass in order to join together structural members in the form of tubes, angles, channels, etc.
The present invention is also concerned with a metal framing system consisting of channels, angles, square tubes and the like, together with a variety of fittings, all to be held together by nuts and bolts. It differs from “Unistrut” in that the bolts are intended to pass through perforations in all the strut members. It differs from the known perforated type of strut members in that the struts do not initially have holes in them — that is to say, actual holes. Instead, the strut members are provided with rows of potential holes in the form of “knock-outs.” To provide á framing system which can be made up with facility in a large choice of dimensions each framing member is provided with a large number of knockouts by arranging them longitudinally in rows in regularly spaced relation, on 1-inch centers, for example. This is the first aspect of the invention. The second aspect is that the knock-outs are formed in a special way. We quote from the specification:
Briefly, my invention consists in punching out of the several structural members which comprises [sic] my metal framing system, a row or series of rows of so-called “knock-outs”, which are the slugs formed when the hole is partially punched out of the structural member, and the slug or “knock-out” is then replaced, filling its hole again, leaving the strength of the structural member practically unimpaired. [Emphasis ours.]
In more detail:
Before the forming operation [which shapes the structural element], * * the strip of sheet metal is subjected to a punch press or roller die operation, whereby a multiplicity of so-called knock-outs 40 or slugs are formed when a hole is partially punched out of the structural member, and the slug or knock-out is [367]*367then replaced, filling its hole again, and leaving the strength of the structural member practically unimpaired. * * * With a hard steel the punch normally will penetrate the structural member approximately one-third of its thickness, and with a softer steel will penetrate the structural member to approximately two-thirds of its thickness. The slug or knock-out 40 is then pressed back into place in a suitable manner, as by pressing rolls, to present a substantially smooth surface on the flat sides of the structural member 30. To restore or retain the original strength of the structural member after the knock-outs are severed, and are then pressed back into place, it is important that the knock-outs closely contact the sides of the hole which is formed in the structural member by punching out a knock-out. The closer the fit of the knock-out in its hole, and hence its contact with the sides of the hole, the more the original compressive strength of the structural member is retained or restored. [Emphasis ours.]
As compared to “Unistrut,” which requires for assembly only a saw and a wrench, the system of the invention, which goes by the name of “Perf-OStrut,” requires in addition only a hammer and a drift pin with which to remove the knock-outs. It is stated to be the great advantage of “Perf-O-Strut” over “Unistrut” that “at least ninety percent” of the fittings used with the latter can be eliminated, which effects an enormous saving in stocking requirements. Another major advantage is that panels can be attached directly to any side of a “Perf-O-Strut” member, whereas with “Unistrut” this cannot be done unless special members or fittings are used.
The invention has been defined in a single main claim, 48, and six dependent claims. Claim 48, which we have broken into numbered elements, reads:
48. In an elongated unitary load-supporting metal frame member for an adjustable metal framing construction,
said frame member having one or more flat sides,
a plurality of spaced removable knock-outs provided in at least one of said sides and forming a hole therein when removed,
each of said knock-outs being completely severed from said member for at least a major portion of the common periphery of the knockout and the associated hole,
said knock-out being substantially coplanar with said side and the edge of said knock-out closely contacting the side of said hole whereby said frame member presents a substantially smooth surface and each said knock-out when retained contributes to the strength and rigidity of the frame member,
said knock-outs when removed providing optional holes for the attachment of additional frame members to said frame member.
Additional limitations added by the dependent claims are: in 49, that the knock-outs are arranged in one or more longitudinal rows; in 50, that they are attached to the frame members by a connecting web; in 51, that they are completely severed and not thus connected; in 52, that there are two or more angularly bent side elements; in 53, modifying 52, the same limitation as in 49; in 54, modifying 52, that there are at least three side elements, two being parallel, at least one pair of parallel elements having their knock-outs directly opposite each other.
Refection and References
Since the patentability of this invention has been before the board on two different occasions, on different sets of claims and different references, in part, we prefer to approach the problem historically, beginning at the beginning. The solicitor would prefer that we do [368]*368not, but the patentability issue in this case is a close question, difficult to decide, and merits a careful look at the entire picture.
The first appeal to the board was taken in 1956 on three claims in parent application No. 317,128. Three references were relied on:
Clayton Jan. 23, 1934 1,944,707
Andrew et al. Sept. 4, 1951 2,567,141
Raucati (Italian) Feb. 15, 1947 418,418
Raucati then was and still is the “primary” reference. There has never been any dispute about what it shows. It is in the same broad field as appellant’s invention and is concerned with “a metal structure with demountable joints for the framing of bridges, scaffolds and similar structures. * * * portable metal structures which are easily dismantled and transported.” Illustrated are structural framing members in the form of square-section metal tubes, opposite sides of which are provided with “one or more series of holes for the passage of bolt- and-nut fasteners which join together two or more of the tubular elements constituting the structure.” The holes are spaced at regular intervals, the holes on opposing sides have their axes lined up, and the axes of holes on adjacent parallel sides may be staggered half a space, arrangements appellant employs in his system. Perforated plates may be used as fittings to make joints where the holes in the structural members do not line up.
As to the other two references, Andrew et al. is entitled “Closure Cap Lining Machine” and relates to machines for lining receptacle closure caps with faced cardboard discs to make them fluid-tight. Clayton is entitled “Removable Knockout” and relates to the common metal boxes used for electrical wiring elements such as switches and outlet receptacles. In 1929, when Clayton filed his application, the prior art on knock-outs was thus described in the specification:
It has become the standard practice among manufacturers of such boxes to punch a portion of the metal in the walls of such boxes entirely through, with the exception of a small section [a web] which is left unpunched in order to retain the punched portion in the box. Such punched portions are known as knock-outs and are made in a variety of shapes and sizes to suit large varieties of wiring materials. These knock-outs when removed, provide an opening in the box wall by which wiring materials of various sorts may enter the box. Due to the disposition of the knock-outs in the box wall in the plane of the wall or parallel to it, difficulty has been experienced in removing such knockouts. [Emphasis ours.]
Clayton’s invention was to put a slot in the knock-out into which a screwdriver could be inserted to twist out the knockout by rocking it back and forth to break the webs.
In 1958, in its earlier opinion, the board disposed of Clayton and Andrew et al., the “secondary” references, in the following passages which we quote because these references are again involved in the present appeal:
Clayton discloses a sheet metal switch box having removable knockouts located in the plane of the side of the box in which they are formed.
Andrew et al, as it is pertinent to the instant rejection, discloses punching lining disks for closure caps from a strip of yieldable material, returning them to their original position in the plane of the strip and then feeding them forward while in that position by the feeding of the strip.
*#***•«
[369]*369We agree with the appellant that neither of the secondary references is concerned with the strength characteristics of structural members. The Andrew et al patent is wholly foreign in this regard since it deals with a flexible strip and with disks completely severed from the strip. Clayton, while showing a sheet metal box having knock-outs in the plane of the side in which they are formed, it is deemed apparent that the knockouts therein are not held in that plane for the purpose of strengthening the box against load stress while in use, since electric outlet boxes in their normal use do not carry any loads. Thus, in the absence of appellant’s own disclosure we do not believe that one skilled in the art of structural members would be led by the Clayton disclosure to modify the structure of the Italian patent, by substituting knock-outs for the apertured side portions thereof and positioning the knock-out in the plane of the sides, for strengthening the structural members of the Italian patent. Since appellant’s concept and the result obtained in effectuating it are lacking in the references before us, the rejection of the appealed claims will not be sustained. [Emphasis ours.]
At this point we express full agreement with the board’s reasons for then reversing the rejection, noting particularly that it could not find either the inventive concept or the result obtained in the references before it, notwithstanding Raucati had the framing members with holes and Clayton had the knock-outs in the plane of a metal box wall. Knockouts had then been common practice for a long time yet it was not felt one of ordinary skill in the art would have found it obvious to use them in structural members in the manner and for the purpose found in appellant’s invention. These reasons for reversing the rejection are, to our minds, the important ones.
In the decision just discussed, the three claims involved contained limitations to structural members which are “tubular,” or to staggered rows of holes, or to parallel walls with aligned holes, or to holes along the center line of a member, or a plurality of these limitations. We note, however, that none of these restrictions underlay the reasons for the board’s reversal of the rejection. Clearly, the claims with such limitations were inadequate to protect the invention and so the application was refiled instead of being issued, though allowed. We come, then, to the present appeal wherein adequate protection is sought on claims omitting the aforesaid limitations.
The record now contains a considerable volume of data on the commercially developed Attwood invention and on the relevant commercial art. The status of “Perf-O-Strut,” i. e. the invention in commercially developed form, as of the time this data was made of record in 1960-1962, was that extensive catalogs with complete engineering data had been prepared and printed but not publicly distributed (they are of record), and, as stated in the latest affidavit filed, “Experimental • machinery has been developed and built for manufacturing Perf-O-Strut Metal Framing and fittings therefor. * * 1 This potential business had not yet reached the point where its market impact could be known, notwithstanding the passage of eight or [370]*370nine years from the filing of the parent application.
The Examiner’s Answer in the present appeal relied on the following references:
Young 1,764,134 June 17, 1930
Holmstrom 2,508,066 May 16, 1950
* Andrew et al. 2,567,141 Sept. 4, 1951
Cripe 2,605,868 Aug. 5, 1952
* Raucati (Italian) 418,418 Feb. 15, 1947
* involved in former appeal
In presenting these to the board the examiner said, after reference to the prior appeal in which he was reversed,
* * * in the instant appeal a different fact situation is before the Board due to the introduction in the prosecution of the patents to Holmstrom, Cripe or Young, all of whom show structural load-bearing members having knock-outs.
He thus stated the basis of the rejection :
As stated by applicant in his appeal brief on page 13, “the sole question raised in this appeal is whether in the absence of appellant’s own disclosure would one skilled in the art of structural members be led by the disclosure of any or all of the secondary references to modify the structure of the Italian patent, by substituting knock-outs for the apertured side portions thereof, and positioning the knock-out in the plane of the sides, for strengthening the structural members of the Italian patent.” It is the Examiner’s position that in view of the secondary references such a modification would be obvious to one having ordinary skill in the art. [Emphasis ours.]
The board2 disposed of Andrew et al. and Young summarily, saying they “are not deemed pertinent to the subject matter claimed, nor to the rejection.” Therefore we need not discuss them. It affirmed on the basis of Raucati taken with Holmstrom or Cripe but it also relied on Clayton, which it had discarded in its first opinion. The board agreed with the examiner’s position but it also went further (without treating it as a new rejection) and held that all claims except claim 51 “are fully structurally met by Holmstrom per se * A complete disposition of the case requires us to deal with both positions. We will first consider the obviousness issue, assumed by us to be predicated on 35 U.S.C. § 103, and will discuss the new reference disclosures in the course of doing so. We will then consider the “fully structurally met” argument, which would presumably be based on 35 U.S.C. § 102.
OPINION
The Obviousness Issue
The solicitor’s brief dismisses the board’s first decision, which was clear, brief, and directed solely to the obviousness issue, with a phrase, saying simply that it lacks significance to the case at bar for reasons which should be evident from the examiner’s answer and the board opinion herein. But it does not brush off that easily.
The examiner, having been reversed because none of the references originally relied on suggested the use of knock-outs pressed back in place in the kind of structural members involved in adjustable metal framing systems, sought to sustain his position on patentability and at the same time answer the board by citing new references showing “structural members” in which there were knoek[371]*371outs. Young, the new reference summarily dismissed, like the old Andrew et al. reference, showed a concrete I-beam with the usual metal reinforcing rods. When the beam was cast it was provided with circular thin spots in the web, optionally defined by metal rings, which could be removed to produce holes to receive brackets for the support of cross beams. The board found knockouts in this sort of a “structural” member entirely irrelevant. The main question is whether the other two new references are, realistically, any more relevant. We think not.
The Holmstrom patent is entitled “Building Construction” and describes a single form of four-part, pressed-steel, box girder. It will best be understood from Fig. 1, here reproduced, which is a vertical sectional view through an assembled girder.
There are two box-like side members A and B with sloping upper and lower faces 8 terminating in notches 10. The members are reinforced by internal struts 14. Top and bottom cap members C and D have overhanging edges 22 and projecting plates 26 with hooked edges 28 which form latch means. When the sides are placed face-to-face and the caps put on they latch together irremovably. Not shown here are wedges which can be driven through slots in the caps to spread plates 26. To be sure, this is a “structural member”, but nothing like appellant’s. The Patent Office relies on the presence of knock-outs, the total description of which is:
The walls 2 and the abutting faces 6 are provided with aligned knockout disks 12 which, when removed, define aligned openings that are adapted to receive piping, electric conduits and the like * * *.
A conduit is shown in broken lines passing through the girder. The drawings that show the knock-outs before removal also show their surfaces in the plane of the walls of the girder members, a structure also disclosed to be old by Clayton, over which the board found patentability on the first appeal. Fig. 5 shows that the knock-outs are arranged in rows along the median lines of the box-like girder side members A and B.
[372]*372Cripe’s “structural member” is a prefabricated steel unit from a plurality of which a flight of stairs can be assembled by bolting a number of units together and fastening them to stringers. Figs. 1 and 2 show, respectively, a portion of a stairs and one of the units, a portion of the tread 10, riser 11, and front 12 being broken away.
The usual wood stringers 13 are employed and the metal units used to form steps. Flanged, triangular end-plates are screwed to the stringers and welded to the bent sheet which forms the tread, riser, and front. A channel rail 18 on the back of each unit is bolted to the front 12 of the unit above it by bolts 20 which pass through selected holes 19, the selection depending on the slope of the stair. They are thus described:
Preferably, the openings 19 are of the knock-out type, that is, each opening is initially closed by a weakened section which can be removed by a blow thereon. The openings or holes 19 are arranged in horizontal rows and vertical tiers.
We believe there is an inventive concept involved in appellant’s particular application of a particular construction of knock-out to elongated load-supporting metal frame members of an adjustable framing system which is not suggested or rendered obvious by the three references relied on. It must be confessed that having been told about the invention and when at first introduced to the metal framing system of Raucati and the various uses of knock-outs long known in various somewhat related metal goods arts — that is to say, to the elements which are combined in the invention — it has an aura of prima facie obviousness about it. But upon further contemplation there appears to be more to it than merely bringing these elements together. In the first place, we find no suggestion to bring them together.
Admittedly there is a whole class of adjustable framing system frame members which are provided with rows of holes for the same purpose as the holes ultimately made, but only as needed, in appellant’s system. Knock-outs have been used in electric wiring boxes for a very long time and they also appear iri Holmstrom’s girder and Cripe’s stair [373]*373units in 1946 and 1947. There is more to appellant’s invention, as we view it, than merely employing the old knockouts. As pointed out in two passages in the portions of the specification quoted above, appellant’s main idea is that by forcing the knock-out slug tightly back into its hole, having so formed it in the first place as to have the closest possible fit, the compressive strength of the member, though containing many such potential holes, is left practically unimpaired. Tests made at appellant’s instance demonstrated this fact. With this concept appellant has created what appears to be an entirely new and advantageous type of adjustable framing system. We have searched in vain in the references for any suggestion of forcing knock-outs back into their holes to restore strength to a member. It is noted, moreover, that this feature is not utilized in just any “structural” member but, as claimed, in the elongated load-supporting member of an adjustable metal framing system, element [1] in claim 48. Such members do not resemble in size, shape, or proportions the type of girder shown in Holmstrom nor the stair unit in Cripe. In these two references we, in view of the relative proportions of the units as a whole and the knock-outs therein, do not see that the knock-outs, even when removed, would have an appreciable effect on their structural strength. For this reason we do not regard it as significant that the knock-outs in place in Holmstrom and Cripe inherently provide some added strength, as found by the board. They do so where it doesn’t matter to anyone. The inherent fact seems to have gone unrecognized. There is nothing in either reference to indicate that such is the fact and therefore no teaching of appellant’s concept that an adjustable framing member having all the convenience of the perforated type of member but with greatly added strength could be produced by using replaced knock-outs in place of actual holes. We feel the same about the new so-called structural-member secondary references as the board did in the first appeal about Clayton, whose knock-out disclosure the board held did not suggest either the inventive concept or the result obtained. We feel this way because Holmstrom and Cripe equally fail to make such suggestion even though they do disclose “structural” members.
In our view Holmstrom and Cripe really add nothing to Clayton and these three patents, taken together, merely emphasize that knock-outs have been well-known in the metal-working arts for a very long time without anyone recognizing, or at least making use of, the characteristic of a fully replaced and tightly fitting knock-out slug to add strength to a member as compared to the strength of the same member having perforations. This is a different concept from that normally involved in the use of knock-outs, which is merely to maintain in a closed condition a closed member such as a box or girder or stair unit until such time as an opening in it is needed.
The board’s use of the Clayton reference, not relied on by the examiner in this appeal, was with reference to the limitation in claims 48-54 that the knockouts are “severed” — clause [4] supra. The board said Clayton shows the forming of knock-outs by punching. While this is true, we do not see that it adds anything to the force of the argument in support of the rejection. Knock-outs have apparently always been made that way.
The Board’s “Fully Structurally Met’’ Idea
We refer to this as an idea rather than a rejection as it is not clear whether the board was making a new rejection or merely supporting the examiner’s obviousness rejection by a sort of a fortiori argument, regarding a reference that fully meets a claim as the ultimate in showing obviousness. This idea was not applied to claim 51. As to the other claims, the board could reach its conclu[374]*374sion only by ignoring parts of the claim which do not, in fact, read on Holmstrom. These parts are in clauses [1] and [6] of our claim analysis, supra, namely, the parts beginning with the word “for.” The board deemed these to be “use limitations” which merely denoted intended use, of no patentable significance, which cannot be relied on to sustain patentability. We do not so regard them.
The board supported its view as to the “use limitations” by reference to three cases: In re Dalton, 188 F.2d 170, 38 CCPA 953; In re Bisley, 197 F.2d 355, 39 CCPA 982; and In re Moreton, 288 F.2d 708, 48 CCPA 875.
We will first dispose of the Moreton case as not at all in point. The only thing of possible apparent relevancy in the opinion is the statement that under the statute one cannot claim a use per se. As the opinion states, this is because a use is not among the categories of inventions which can be patented. 35 U.S.C. § 101. It is not a holding that use limitations in claims to structures are of no significance in determining patentability.
The Dalton case was relied upon apparently for some rather sweeping statements it contains to the effect that “Properties, functions, uses, and results” may not be solely relied upon for patentability and the court in that case was relying heavily on the binding effect of Halliburton Oil Well Cementing Co. v. Walker, 329 U.S. 1, 67 S.Ct. 6, 91 L.Ed. 3. All this was prior to the Patent Act of 1952 which introduced a new statute with respect to “functional” claims in the last paragraph of 35 U.S.C. § 112. One of the purposes of this statute was modification of the Halliburton rule. As stated in Federico’s Commentary on the New Patent Act, 35 U.S.C.A., at p. 25:
It is unquestionable that some measure of greater liberality in the use of functional expressions in combination claims is authorized than had been permitted by some court decisions, and that decisions such as that in Halliburton Oil Well Cementing Co. v. Walker * * * are modified or rendered obsolete, but the exact limits of the enlargement remain to be determined.
We have here a combination claim and the limitations ignored by the board as use limitations we think are functional expressions which must be given weight. We consider that 35 U.S.C. § 112 has rendered much if not most of what was said in Dalton on this point obsolete. When these clauses are considered, there is no foundation for the board’s view that the claims read on Holmstrom.
Furthermore, we believe that the claim limitations to frame members for use in adjustable framing construction etc. should be given weight under our decision in Kropa v. Robie, 187 F.2d 150, 38 CCPA 858, where, contemporaneously with Dalton, the court said:
In the case before us, the words “An abrasive article” are essential to point out the invention defined by the counts. In our judgment those introductory words give life and meaning to the counts, for it is only by that phrase that it can be known that the subject matter defined by the claims is comprised as an abrasive article.
Here we think it is of patentable significance that what is claimed is “an elongated unitary load-supporting metal frame member for an adjustable metal framing construction,” rather than some other kind of structural member. This clause the board disregarded.
The Bisley case does not support the board. While it contains the sentence, “A limitation reciting only manner of operation or use will not sustain patent-ability of a claim,” that does not appear to us to be the case here. The opinion also says, “Definite limitations in a claim should not be ignored or construed out of the claim.” Another pertinent passage applicable here is:
Counsel for appellant in his brief contends that in a case like this in[375]*375vention is to be gauged not alone by the extent or simplicity of the physical changes made, but also by the perception of the necessity or desirability of making such changes to produce a new result. We think that this contention has merit.
And finally, having reference to the board’s previous allowance of other claims, for reasons apparently unrelated to the limitations they contained which are not in the appealed claims, we quote the following passage from Bisley:
Allowability of an appealed claim is not controlled by the fact that similar claims have been allowed by the Patent Office, since an appealed claim must be patentable in its own right in the opinion of this court. However, similar claims allowed by the Patent Office tribunals furnish evidence of what features those tribunals regarded as patentable, and we think it proper, and sometimes necessary, to consider allowed claims in order to fully determine the views of the board and the examiner. [Cases cited.]
Here the board’s two decisions arrived at essentially opposite results on the premise that the new secondary references made a difference because they related to “structural” members. We have concluded that this fact is not significant in view of other factors. We therefore express agreement with the first decision of the board on patent-ability and disagreement with the second, which we reverse.
Reversed.