delivered the opinion of the court:
The appellant, Smith, filed his application in the United States Patent Office and claimed invention in the construction of a heat ex*1008changer such as a radiator used for water cooling purposes in an automobile or the like.
The Primary Examiner rejected claims 10 to 14, inclusive,, which rejection was affirmed by the Board of Appeals. No claims were allowed. The rejection was based upon the ground that the claims were unpatentable over a number of references cited by the examiner. Erom the board’s decision, appellant has here appealed.
Appellant in this court conceded that all the elements of the claims which he has combined in his application, and which combination he •claims inventive over the prior art, are shown by combining different features of the references relied upon.
The examiner relied upon the following references:
Fedders, 992,763, May 23,1911.
Fliggins, 1,747,115, February 11,1930.
Freeman, 1,821,702, September 1, 1931.
Bundy, 1,892,607, December 27, 1932.
Saunders, 2,158,383, May 16,1939.
French patent, 575,142, April 16,1924.
French patent, 682,988, February 24,1930.
Technologic Papers of the Bureau of Standards, No. 211, May 26, 1922.
Claims 10 and 11 were properly regarded by the board as illustrative •and they follow:
10. A heat exchanger’ for an internal combustion engine comprising spaced cooling liquid tanks and a plurality of hollow members connecting the interiors thereof, each of said members being composed of steel which is about 28 gauge in thickness and which is coated with films of more electropositive metal which are a few thousandths of an inch thick; each member having a plurality of parallel lengthwise extending, straight, separated liquid passages which are about twenty times as long as wide in cross section, each member not exceeding about %i" in overall thickness, said members being arranged in parallel position and in the number of about five per inch.
11. A heat exchanger comprising liquid compartments and a plurality of straight hollow metal tube members connecting the interiors of said compart-’ ments, each side of each of said members having a plurality of longitudinally extending parallel smooth outer side surfaces arranged alternately in two .different planes and connected by smooth surfaces disposed at acute angles of said planes, the abutting surfaces of each member being secured together thereby forming in each member a plurality of straight lengthwise extending separate liquid passages, said members being arranged side-by-side and defining there-between air passages which are free from obstructions, which are venturi-like in shape in the direction of air flow therethrough and which are of such widths that- air passing therethrough is repeatedly agitated and mixed by engagement with the acutely disposed surfaces and its temperature is thereby repeatedly .equalized in lines transversely of its direction of flow.
Appellant’s heat exchanger or radiator is formed of a plurality of multiple tube members, of which the tubular portion is a flattened cross section. The multiple tube members are positioned between *1009headers in parallel edgewise relation crosswise of the radiator. It is stated that air flowing through such radiator is effectively agitated to cause good heat transfer between it and the walls of the tubing by- venturilike effect and also that the flattened cross section gives relatively extended surface.
Claim 10 differs from the other appealed claims, in that it provides for “each of said members being composed of steel which is about 28 gauge in thickness and which is coated with films of more electro-positive metal which are a few thousandths of an inch thick:”
Claim 11 adds the functional limitation that the members define therebetween air passages which are free from obstructions.
Claim 12 adds the limitation that the abutting surfaces on the ribbed plates are “secured to each other throughout their length” and that they thereby form “a plurality of straight separate liquid passages.”
Claim 13 is quite broad. It provides that the tube member is straight and recites that there is “such a space between adjacent members that air flowing through said spaces will be subjected to a venturilike effect and will be repeatedly agitated and mixed by said acutely inclined surfaces and the temperature of the air will be repeatedly equalized adjacent to said acutely defined surfaces.”
Claim 14 provides that the tube members shall be arranged in parallel position and in the number of about 5 per inch.
In affirming the decision of the examiner rejecting said claims, the board stated:
In regard to the shape and arrangement of tubular elements it is apparent that this detail is fully anticipated by each of Fedders, Higgins and French pateiit 682,988. The terms of the claims are regarded as fully satisfied by each of these references, also as a cumulative ground by the French patent 575,142.
The radiator tubing of the above patents appears to be formed of any suitable metal merely as a matter of choice. In order to show the use of steel coated with copper in connection with tubing the examiner relies upon patent to Bundy. That patent clearly discloses a tube of ferrous metal or alloy coated with copper. This satisfies the requirement of claim 10 in respect'to steel coated with an electropositive metal Saunders discloses a radiator formed of sheet steel sections united into tubular form by use of copper as a solder. The copper to some extent may be regarded as forming a surface coating.
Claim 10 has been rejected in one instance as unpatentable over Fedders, Higgins or French patent in view of Bundy or Saunders. This ground of rejection is affirmed. It is considered to be no more than an optional matter of choice to form a heat exchanging unit of any desired metal, that is, to form a radiator of any of the first three patents out of tubing of steel or ferrous metal.
In connection with this structure applicant has submitted an affidavit setting forth comparative tests as between a conventional radiator for tractors and a radiator in accordance with Claim 10. Temperatures at the top and bottom of the two kinds of radiators are given. It appears that the temperature difference in the several tests are of the order of 2° to 5° between the top and *1010bottom, where the temperature of the water in general in the radiator was up close to boiling temperature. This difference in temperature is¿ considered to be relatively small and not of much practical significance. We are also in agreement with examiner’s conclusions that within the broad descriptive terms of the experiment many unknown factors might have intervened. One important factor would seem to be that no comparison was made as to the condition of the radiator in respect to being entirely clean or possibly coated with a film of sediment in case the so-called Ford radiator was old or not in good condition. We are not convinced that the affidavit is sufficient basis upon which a conclusion as to invention may be predicated.
Claims 11, 12 and IS are also rejected upon the patent above noted relating to the mere form of radiator without composition of tubing. This rejection upon Fedders, Higgins, Saunders and French patent 682,988 is affirmed.
Claim 14 includes some limitation as to the number of vertical tubular elements in each tubular unit. This feature appears fully satisfied by Higgins, disclosing seven tubular units and French patent 682,988, disclosing a choice between three, four or five tubular units. This feature is regarded as being purely a matter of choice without producing any new or unobvious result. The reference to venturi-like effect also referred to in Claim 14 is necessarily true of the structures shown by Fedders, Higgins and French patent 575,142. The discussion set forth in the brief has been carefully considered without finding reason for reversing the examiner’s action.
After the board’s decision appellant requested reconsideration and - filed with the board a supplemental affidavit, asking that it be considered in conjunction with the former affidavit referred to in the above quotation. The board, in its decision on reconsideration, stated:
It is found, as pointed out in the communication of July 2, 1945 that we misconstrued the reference to “tube members” occurring in Claims 10 and 14, as meaning tubes 10 rather than tube units 3. However on consideration of Claims 10 and 14 it appears that the selection of about five units per inch is no more than mere optional choice of á variable feature. The reference Tech-nologic Papers shows that it has been recognized that various numbers of tubes per inch in radiators have been proposed. One of the examples is four tubes per inch, another about six per inch. Others of the ten examples run higher and lower respectively. In Claim 14 this feature has no particular dependence or cooperation with any other feature. It is merely arbitrary.
In Claim 10 it might be regarded as related to the limitation of making the radiator of galvanized iron which because of its poorer conductivity of heat as compared to copper would require more surface area in a radiator of certain overall dimensions. But we hold that even this is a feature obvious to one skilled in the radiator art and to be a mere matter of test and without inventive merit.
We find no reason for changing our conclusions in regard to the venturi effect in the references. It is regarded as true of each of the structures of the references, Higgins, Fedders and the two French patents.
The above quoted first decision of the board substantially follows the examiner’s application of the references to the claims.
In this court some of the claims containing limitations which distinguish those claims from others are not pressed with much enthusi*1011asm as being patentable over the other claims by reason of such limitations ; in other words, the claims stand or fall together.
The main contention of appellant is that he has taken from the prior art certain features and applied them in a different relationship in a new structure and has brought about a new improvement of a patentable character. Appellant argues that not in fifty years of the automobile radiator industry has anyone made a similar combination and he particularly stresses the importance of the use of steel as provided for in claim 10, premising this argument upon the alleged prior understanding that by reason of the corrosive character of steel it was not thought feasible to use this metal in forming the tubes of a heat exchanger. Appellant argues that he has proceeded against the teachings of the art to produce an unexpected result and a substantially new improvement which should be regarded as rising to the dignity of invention.
Appellant teaches in his application the use of tubes in parallel rows to the direction of air flow, which tubes are to be flattened and placed in such manner that the corrugations form opposite halves of parallel flattened tubes. Each assembly of two ribbed plates forming several tubes is referred to as a tube member.
The patents to Fedders, Higgins, and Saunders and the two French patents all show variations of the same general type of radiator. Freeman shows lead-coated steel tubes suitable .for use in heat exchangers, and Bundy shows copper-coated steel tubes suitable for such use. The Technologic Papers of the Bureau of Standards, No. 211, discusses at length several factors involved in heat exchanger design and mentions various tests with reference to spacing of plate type tube members particularly for use in aircraft engines, where tube members of various cross sections are employed.
It would serve no useful purpose here to discuss in further detail the disclosures of the art cited since it is admitted that every element of every claim is disclosed somewhere in the cited art.
The sole question presented is whether or not what appellant has done by combining what he regards as the desirable features of the prior art, and, in so doing, producing alleged improved results, amounts to invention.
Appellant challenges the holding of the Patent Office tribunals to the effect that what he has clone is the obvious thing for one skilled in the art to do. He calls attention to a citation in a brief in another apqieal in this court (In re Horney, Patent Appeal No. 5240) of In re Buser, 29 C. C. P. A. (Patents) 1196, 129 F. (2d) 710, 54 USPQ 319, wherein this court held that a finding of obviousness by the Patent Office tribunals is not conclusive.
*1012This, of course, is the holding, and a proper one, in that and in many 'other decisions, and, as appellant frankly concedes, it becomes a question of opinion for the one who is called upon to decide whether or not such a combination of old elements into a new structure involves invention.
While-appellant has not complained directly about the combining of references, as so frequently occurs in this court, the subject matter is one that is implicit in the argument presented. It has been held so often as to require little discussion here that in the rejection of claims in the Patent Office it is proper to combine references to meet the limitations of combination claims, and then the question occurs, as in this case, whether or not the applicant’s accomplishment involved invention.
It seems to us that after eliminating the immaterial limitations which are not relied upon as carrying with them invention over the other claims there is nothing in the combination of the elements of appellant’s construction which might not be, by the skilled mechanic, employed in a combined structure such as we have at bar.
Surely it cannot be urged that because appellant has picked out of the art certain well-known features with well-known functions and combined them into a new structure, with possible improved results, this necessarily connotes invention. Assuming that appellant has achieved improved results and that his structure has advantages over anything produced in the prior art, this does not necessarily signify that his production required the exercise of the inventive faculty. .The old elements all perform the same functions in the new combination that they performed in the prior art; there is no new or unexpected coaction between them, and the results obtained can hardly be said to be those which are unobvious or unexpected.
Of course, it is well-settled law that if one selects from the prior art well-known features in such a manner as to produce new, unobvious, unexpected and useful results which are not merely the accumulations of the good results taught by the art he has made an invention. Frequently combinations of old elements in new structures have resulted in inventions of a high order.
However, it must be understood that the radiator art is a crowded one and that one who conceives of a new combination of old elements is rarely justified in claiming patentability merely because he has produced some improvement and no one else has made the same combination.
There sometimes is a very narrow division line between what amounts to invention and what is within the realm of performance of the skilled mechanic, and it is difficult to lay down -precise rules *1013ior determination of the issue. As appellant suggests, it is largely a matter of opinion. Appellant’s counsel lias called attention to Ms-argument and citation of authorities in a brief in another appeal pending in this court to the effect that the finding of obviousness by the Patent Office is not conclusive upon this court. It is usually persuasive, however, since the Patent Office tribunals presumably are-familiar with the advancement of the art.
Appellant, as before stated, emphasizes that' the metal used in constructing his tubes — steel, coated with noncorrosive metal such as-copper — has produced a new and useful result. Tubes of steel in this relationship, or tubes coated with noncorrosive metal in the same relationship, are not new, as is shown in the art. Certainly their use is suggested by the art and the mere fact that these tubes are employed in a slightly different arrangement (although along the same general lines employed in the art) does not indicate invention.
It would be unfortunate, in cases like that at bar, to handicap the builders of heat exchangers by granting patents for mere combinations of old elements and thus restrain the freedom of those skilled in the art, unless such combinations produced a new, useful and unob-vious result. The instant record affords no convincing showing that appellant’s activities have gone beyond the realm which should be left open to those skilled in the art.
Appellant submitted certain proof with respect to the utility of his new combination by way of affidavits, one of which was considered by the examiner and the other of which accompanied the request for reconsideration by the board. These were concerned with. comparative tests between a standard Ford tractor radiator and a radiator allegedly constructed according to appellant’s specifications. The board, in commenting upon the first affidavit, stated:
Temperatures at the top and bottom of the two kinds pf radiators are given-It appears that the temperature difference in the several tests are of the order of 2° to 5° between the top and bottom, where the temperature of the water in. general in the radiator was up close to boiling temperature. This difference-in temperature is considered to be relatively small and not of much practical significance. We are also in agreement with the examiner’s conclusions that within the broad descriptive terms of the experiment. many unknown factors-might have intervened. One important factor would seem to be that no comparison was made as to the condition of the radiator in respect to being; entirely clean or possibly coated with a film of sediment in case the so-called Foret radiator was old or not in good condition. We are not convinced that the-affidavit is sufficient basis upon which a conclusion as to invention may be-predicated.
We are in agreement with the holding of the Patent Office-tribunals as to the effect to be given that affidavit. In doubtful, cases it is very helpful to the Patent Office and to the courts, even on *1014a question of patentability, to have sworn proof to sustain claims of improvement over the prior art. However, as we have said before, all improvements do not necessarily connote invention and assuming that there was the degree of improvement claimed by appellant, which he urges is shown by the affidavits, we are nevertheless of the opinion that what appellant has done is nothing more than that which is within the realm of the skilled mechanic and does not involve invention. In view of our conclusion in this respect, it is unnecessary for us to point out, as did the tribunals below, the deficiencies apparent on the face of said affidavits relating to the tests with the Ford tractor.
Appellant has stressed the importance of having about five tube members per inch in his radiator structure. On this phase of the case the board said:
* * * the selection of about five units per inch is no more than mere optional choice of a variable feature. The reference Technologic Papers shows that it has been recognized that various numbers of tubes per inch in radiators have been proposed. One of the examples is four tubes i>er inch, another about six per inch. Others of the ten examples run higher and lower respectively. In Claim 14 this feature has no particular dependence or cooperation with any other feature. It is merely arbitrary.
In Claim 10 it might be regarded as related to the limitation of making the radiator of-galvanized iron which because of its poorer conductivity of heat as compared to copper would require more surface area in a radiator of certain overall dimensions. But we hold that even this is a feature obvious to one skilled in the radiator art and to be a mere matter of test and without inventive merit.
With this holding of the board we agree.
Further discussion of the details of some of the other appealed claims is not deemed, to be of importance here.
The decision of the- Board of Appeals affirming the examiner’s re- , jection of the appealed claims is affirmed.