delivered tbe opinion of tbe court:
This is an appeal from the decision of the Patent Office Board of Appeals affirming the rejection of the only remaining claims in appellant’s application which are as follows:
22. A hollow sectional plaster board partition each section composed of two spaced paper sheathed face plaster boards of equal width and laterally offset, and a spacer at each lateral edge comprising a relatively narrow- filler strip of paper sheathed plaster board between and adhesively secured to each face board and having its outer edge disposed midway between the offset edges of the face boards, the thickness of the filler strip and of the face boards being substantially the same and the exposed face of the filler strip being at least three times its thickness, whereby the edges of the section have two steps of equal length and *815height, two adjacent sections being joined in the same plane with their two-step complementary edges fitting together in face to. face contact across the full width of the steps with the overlapping faces adhesively joined and the edges of the face boards and spacers of the -two sections abutting and making a solid joint, and a vertical series of screws entered through the overlapping step portion of each face board and passing freely through said overlapping portion and extending through the spacer and into the inner layer of the face board of- the opposing section, whereby the screw threads obtain a purchase on the paper sheathings and adhesive between the overlapping faces of the joints and serve thereby as clamping means to hold the faces of said joints firmly together in contact while the adhesive on said faces is hardening in the erection of the partition.
23. A hollow sectional plaster board partition as described in claim 22, wherein each of the face plaster boards and of the filler strips is of laminated construction, each lamination comprising a plaster sheet entirely sheathed on its opposite faces with paper.
The following references are relied on:
Muller, 2,133,108, Oct. 11, 1938.
Wittner, 2,295,248, Sept. 8, 1942.
Bartholomew, 2,361,733, Oct. 31, 1944.
The invention, the nature of which will appear from a reading of claim 22, relates to a non-bearing building partition or wall which is made of prefabricated panel units of gypsum wallboard which, the record shows, is well known to be a type of plaster board sheathed in paper. Appellant’s application says, “A well known form of plaster board is gypsum board, which is gypsum in sheet form with a sheathing of paper on both sides.” When we use the term “wallboard” herein we refer to this product. Such wallboard is usually nailed to wood, studding a place of the traditional lath and plaster construction. The partition described by the applicant, on the other hand, is intended to be used without a separate support such as studding and is merely anchored at top and bottom to the ceiling and floor and otherwise supports itself. Appellant obtains the rigidity required of such a partition wall by laminating a plurality of sheets of wallboard together and in doing this the outer or face sheets are laterally offset, thus producing complementary lap joints at the edges of the panels which make up the partitions. The outer sheets of each panel are not secured directly to each other, however, but are spaced apart by filler strips, one at each edge and positioned between the face sheets to extend half way across the distance between the edges of the outer panels so as to form a two-step lap joint. A center filler strip is also disclosed as preferable but this is not a limitation of the claims. Thus, the panels of the partition contain dead air spaces between the filler strips. Each of the outer or face layers may be made of two or more sheets of wallboard adhered together, in which case the filler strips, as *816claimed, are similarly laminated to have the same thickness by being made of the same number of' plies as the outer panels. This is the Subject matter of claim 23. The preferred embodiment shown in the drawings has panels and fillers each of two layers of wallboard so the total'thickness of the partition is six times the thickness of one sheet of gypsum board. All six layers being adhesively secured together, the panels and a partition made from them will have considerable rigidity.
Emphasis is placed particularly on the manner of joining the unit panels together in making a partition therefrom. This is done by coating the two-step lap joint surfaces with glue and inserting screws, shown in the drawing as ordinary wood screws, freely through holes in the outer lap of each joint and screwing them through the underlying portion of the filler strip and part way through the opposing panel, that is through all of the paper layers except the outer surface layer, the screws holding because they get a grip in the several layers of adhesive and paper through which they pass. Since the screws, which may be placed at eight-inch intervals, pull the faces of the joint together, the proper conditions exist for making a strong glued joint when the glue sets.
Affidavits were filed, made by the applicant and another, which indicate that the claimed construction is commercially successful, having been installed in over 275 houses, since 1947 and prior to March 1952. The construction is said to have Federal Housing Administration approval. It is stated that cost “compares favorably with” stud and plaster construction. It it not said specifically to be cheaper. The affidavit of one Peterson, president of a construction company, says that he has installed the partition in 250 houses and has found it to stand up better than ordinary lath and plaster partitions on studding. He gave it as his opinion that it is one of the most practical and valuable contributions to the building industry that he has encountered. With the affidavits, records of strength tests were also submitted.
The building art is one of the oldest known to man and it can aptly be described as “crowded.” It will be observed from the limitations of the claims that the applicant seeks a patent on only a narrow improvement. Progress is as important, however, in crowded arts as well as in those which are in the pioneer stage, In re Tamarin, 38 C. C. P. A. (Patents) 872, 187 F. 2d 160, 88 USPQ 490, and such progress is usually made in small increments. The question before us is whether applicant’s limited advance is such as is entitled to patent protection under the law.
The principal reference relied on by the Patent Office, and the only one dealing with the art of wallboard construction, is Bartholomew. *817It discloses a self-supporting, laminated wallboard unit made from the same kind of wallboard used by applicant, having spaced outer face panels and relatively narrow spacers made of the same material, the parts all being adhered together and intended for use without studding or other support. The edge joints disclosed are simple square edges to be abutted, tongue and groove of the conventional type and a modification thereof which is a “V” joint. None, of these joints as we have termed them, function to secure the panel units together but are at most a form of edge interlock. On the subject of joining, all the patent discloses is the following:
It is within the purview of the invention to employ any suitable type of dowel means between the units for purposes of joining the same at adjacent edges and reinforcing the edge connections at the joints.
No such joining means are shown. The reference, therefore, clearly falls short of anticipating the claims, which call for a specific joint structure which has the following principal features:
(1) a two-step lap joint wherein the width of each step is at least three times the thickness of the filler strip (the face boards having the same thickness as the filler strips),
(2) the overlapping surfaces of the joint, which are of paper-sheathed plaster board, being adhesively joined,
(3) a series of screws passing freely through the overlapping portion of each face board, through the spacer and into the other face board to clamp the faces of the joints together while the adhesive hardens.
As the Board correctly said, “The differences over Bartholomew pertain to the joint.” Obviously appellant is relying on the joint structure for patentability and clearly the Board was of the opinion Bartholomew does not disclose it.
The Board then turned to a consideration of the Wittner and Muller patents on which the examiner relied for an alleged disclosure of the claimed joint features, saying, in effect, that with respect to joints for panels, joints are joints regardless of materials and joints for plywood and gypsum board are deemed to be from the same art. Considering the vast differences in the physical characteristics of wood or plywood and gypsum or plaster board, we are not convinced that this is a sound approach.
The Wittner patent relates to a prefabricated plywood panel unit put together entirely with adhesive and consisting of three thin sheets of plywood united with two wooden frame structures in the form of rectangular grids, a five layer structure. There is a center layer of plywood on either side of which is a rectangular grid or box frame, consisting, for example, of seven horizontal members and thirteen *818uprights on sixteen inch centers. Adhered to the outer face of each grid are four 4 ft. by 8 ft. plywood panels, the two outer layers of plywood being so offset with respect to each other and to the center layer between the grids as to form stepped ends making a kind of lap joint. When panel units of this construction are joined end to end the plywood surfaces, however are not in contact but are spaced apart by the thickness of the grids. The edge of each layer of plywood falls on the center of a parallel grid element and when a joint is made the plywood of the adjoining panel is adhered to the exposed surface of the grid member. No fastening of any kind is disclosed, other than adhesive. As the Board said, “Wittner does not disclose how pressure is applied during setting of the adhesive.” The Board took the view that Wittner contains a clear teaching of applicant’s lap joint but to find this teaching, which is far from apparent to us, the Board suggested first that one of Wittner’s'panels and one of his spacing grids could be omitted, which is but to omit parts and their function, and then the spacers could be made to abut at the joint, which would be an obvious variation. This seems to us to be a double departure from the actual teaching of Wittner aixd a reconstruction of his disclosure in the light of applicant’s specification. Wittner does not suggest a structure modified in accordance with the Board’s reasoning. We agree with the analysis in appellant’s brief that the Board “proposes to modify the secondary reference before it is employed to modify the primary reference.” Wittner is a reference only for what it clearly discloses or suggests and such modification as has been attempted is not a proper use of this patent as a reference. In re Demarche, 42 C. C. P. A. (Patents) 793, 219 F. 2d 952, 105 USPQ 65.
As indicated above, the claims also- call for a specific use of screws in the joint which is not suggested in either Bartholomew or Wittner, for which reason Muller was cited. Muller discloses making large sheets of plywood out of smaller ones by rabbeting adjoining edges over a considerable area, applying adhesive thereto and driving rows of clinched metal staples along the edges of the joint and through the middle to hold it while the adhesive hardens, thus obviating the need for a press. The staples along one edge are inserted from the side opposite those along the other edge. This is a disclosure of neither screws nor a wallboard panel joint. In applying Muller, the Board used it for its teaching of the desirability of drawing a glued joint together with fasteners, saying, “This being the joint art, it is the same art and can properly be considered for its teaching. The reference discloses staples, not screws, in plywood, not in wallboard. So the Board took the further step of holding screws to be a’ recognized *819equivalent fo.r staples and concluded that their use in the wallboard joint as claimed would be within the expected skill of the art.
If we were dealing with the woodworking art, we might feel that what the applicant did was within the skill of that art. But we are not. The only reference in the field of wallboard utilization and handling is Bartholomew and that patent, though acknowledging, as of its 1943 filing date, that the material was then in “universal” use, failed utterly to meet the problem of joining self-supporting walk board panels together otherwise than with some undisclosed form of “dowel means.” While glue might be an obvious expedient for joining paper-covered wallboard elements, we are inclined to believe that it would not normal],y occur to one having an elementary acquaintance with the characteristic behavior of screws in a plaster composition, which is what gypsum board is, that screws were the solution to the problem of holding wallboard panels together. What we have here, moreover, is a joint involving a combination of several elements, each perhaps seemingly obvious in itself, in the light of hindsight wisdom, which no one prior to applicant has suggested putting together for the purpose of joining wallboard panel units together to make a strong partition, the joints of which the evidence in the record shows, are its strongest parts. The claimed combination includes the two-step lap joint, plus the adhesive, plus the screws, inserted in a particular way with relation to the panel elements. We do not see that the references, without the benefit of the applicant’s disclosure, fairly suggest this combination in the wallboard art.
What we said in In re McKenna et al., 40 C. C. P. A. (Patents) 937, 943, 203 F. 2d 717, 97 USPQ 348, is appropriate here,
The mere fact that elements of applicants’ process may be found in various patent does not necessarily negative invention. [4] * * Moreover, the concept of doing a thing, the result of which is new and useful and unexpected, must be considered along with the actual steps of doing it in considering the presence or absence of invention and patentability.
The same principles are applicable to a mechanical combination. We think any doubt on the question of patentability should be resolved in favor of the applicant. In re Pappas et al., 38 C. C. P. A. (Patents) 746, 185 F. 2d 695, 88 USPQ 108.
The decision of the Board of Appeals is reversed.
Jackson, J., retired, recalled to participate in place of Cole, J., who, because of illness, did not participate in the hearing or decision of this case.
Worley, J., dissents.