Before MAGRUDER, Chief Judge, and WOODBURY and HARTIGAN, Circuit Judges.
WOODBURY, Circuit Judge.
This is an appeal from a judgment dismissing a complaint in a suit for patent infringement wherein it is ordered adjudged and decreed that claims 2 and 6 of U.S. Patent No. 2,252,210, and claims 1 and 4, of U.S. Patent No. 2,252,211, both issued to P. R. Seemiller on August 12, 1941, are invalid for want of invention, but if valid, have been infringed by the defendant. The first patent is for an improvement in the method of making heat exchange cores; the second is for the improved product resulting from the use of the new method.
Heat exchange cores are devices used primarily for the dissipation of heat, although they may be used for the reverse, the direction of heat exchange being immaterial. Hence they may be used either for heating or for cooling. We are here concerned with one specifically designed and adapted to cool the water used to dissipate part of the heat generated by the internal combustion engine in common use in motor vehicles. For convenience then we can refer to heat exchange cores of the specific type involved simply as automobile radiators. Admittedly the art is old, going back to about 1900, and it is also crowded.
For many years, perhaps from the beginning, automobile radiators have been classified in two categories, cellular and tubular. The difference between the two types may be more apparent than structural, as the defendant suggests, but for the purposes of our discussion we shall adhere to the classification established and accepted by the art. Both types, at least since the first crude devices of the early days, consist of two basic elements, known as heat exchange surfaces, which are denominated primary and secondary. A primary surface is any metal surface in contact with air on the outside and with the fluid used for cooling on the inside. These are called water passages.A secondary surface is a thin piece of metal in contact with and extending outward from the outside surface of a water passage, and hence in contact with air on both its surfaces. In accordance with usual terminology we shall call these secondary surfaces fins.
The water passages of the cellular type of core are not preformed. Cores of that type are made by clamping together a multiplicity of appropriately formed transversely corrugated strips or ribbons of the metal to be used, ordinarily copper, laid side by side on edge and of a length appropriate for the height of the finished radiator between the tanks soldered to the tops and bottoms of such structures, and then, after inserting the fins to be described presently, fastening the edges of the strips together by dipping the lateral surfaces of the assembly, which will be the front and back surfaces of the finished radiator, about one-eighth of an inch deep in a bath of molten solder. The fins in cores of this type are arranged vertically between the water passages. In a simple embodiment the fins consist of thin transversely corrugated strips as long as the water passages and as wide as the radiator is thick. These are positioned in the assembly in such a way that the convex portions of the corrugations of the fin are in contact with the convex portions of the corrugations of the water passages on either side of it. Thus cores of this type characteristically have a sort of honeycomb appearance when viewed from either front or back.
In radiators of the tubular type the water passages are preformed, and typically consist of a multiplicity of pre-tinned copper or brass tubes of a length appropriate for the height of the radiator between its top and bottom tanks. The fins in this type of radiator consist of a multiplicity of thin copper strips running across the radiator from side to side in parallel arrangement one above the other and spaced a minor fraction of an inch apart. The width of these strips corresponds to the thickness of the radiator, and the strips are perforated to receive the tubes. In assembling this type of core the strips are held parallel with one another on edge in a jig or holding device of some sort in such a position that the perforations therein are in alignment, and the preformed and pre-tinned tubes are pushed manually through the perforations. The fins are then soldered to the tubes by baking the whole assembly.
Seemiller combined the vertical corrugated fins of the cellular type of core with the preformed pre-tinned tubes of the tubular type. He used a tube oval in cross section formed from a strip of copper bent to tubular shape and sealed with an overlapping interlocking seam along one of its edges. In his radiator the tubes are arranged vertically in rows of three, one behind the other, and each row of tubes is separated by a vertical corrugated fin strip having rounded convex portions.Radiators of this type are assembled on their sides. In assembling, a corrugated fin strip is first laid on the bottom of a form or jig shaped somewhat like a flat open-topped box set on end, then a row of tubes is placed on the fin strip, then another fin strip is added, then another row of tubes, and so on until the form is full of fins and tubes in alternating layers. Then the assembly is compressed laterally, i.e., from the top of the form, in a clamp of appropriate design so that the corrugations of the fins are pressed into the bulging oval sides of the tubes, whereby both are so distorted that there is a relatively large area of contact between tubes and fins resulting in good heat conductivity between them. Uniform spacing of the alternating parts results from the relative stiffness of the interlocking edge seams of the tubes which prevents any row of tubes from being compressed too much while other rows are not compressed enough. Finally the whole assembly is back-soldered, like the conventional core of tubular type, to unitee its parts.
Full details of both the method of construction and the structure of all three types of core are to be found in the opinion of the District Court reported in 96 F.Supp. 438. The foregoing brief general description, however, is adequate for the purposes of this appeal.
We do not need to go deeply into mechanical details for the court below found and the appellant conceded that no single element of either Seemiller's radiator or his method of making it is in itself new.Furthermore it is conceded that the various elements of the Seemiller patents are to be found in one old patent or another of the crowded automobile radiator art. But the appellant says that although many men highly skilled in that art had for years tried and failed, no one before Seemiller had the insight to select and fashion the right elements of the known art, and to put those elements together in the right way, to achieve the outstanding advantages resulting from Seemiller's teaching.
And the advantage claimed for Seemiller is greatly increased efficiency. It is not contended that Seemiller's radiator performs any function that automobile radiators had not performed before, if indeed there can be any other function for an automobile radiator to perform except to cool the fluid used to dissipate part of the heat generated by the engine. Nor is it claimed that Seemiller's radiator functions on some novel principle. The claim for Seemiller is that his radiator can be manufactured with less solder out of lighter stock by less skilled workmen, and hence is less expensive to produce, and furthermore that his radiator is not only lighter in weight, but also structurally sturdier, and moreover that it is more efficient as a colling device per square inch of surface so that smaller radiators can be used thereby effecting a further saving in weight as well as a saving of space. The existence of these advantages, which are characterized as outstanding, are said to be convincingly established by the evidence, and also amply attested, not only by the remarkable commercial success of Seemiller's radiator due not to advertising or trade practices but solely to the inherent superiority of the product, but also by the defendant's tribute of copying.
The District Court is charged with failing to appreciate the significance of the foregoing facts, and also with misapplying the law established by the cases, particularly Great Atlantic & Pacific Tea Co. v. Supermarket Equipment Corp., 1950, 340 U.S. 147, 71 S. Ct. 127, 95 L. Ed. 162, upon which it principally relied. The argument is that in regarding Seemiller's achievement as within the expectable skill of the calling the court relied solely upon ex post facto reasoning (hindsight) disregarding the facts of record, such as the evidence of repeated prior attempts and failures by those well versed in the art, Seemiller's outstanding commercial success based on the inherent merit of his inventions, and the spontaneous, sincere tribute paid him by the defendant's open infringement, all of which it is contended are convincingly persuasive that what Seemiller did was in actual fact far from obvious. Furthermore it is said that the court below erred in interpreting the Atlantic & Pacific decision, supra, as establishing the rule that there can be no invention in making a new combination of old elements or methods unless the new combination performs some new function, not just the aggregate of the separte functions of the old methods or elements, whereas that case did not change the old rule that a novel and ingenious selection of old parts or techniques, plus combining them in a particular way to produce superior performance, long sought for but not found, is enough to establish patentability.
As already clearly appears Seemiller's sole contribution consisted of a new combination of old methods and elements. He did not substitute any new material of known characteristics for an old one in an old combination. Hence we are not concerned with the closely allied question which divided this court in Lincoln Stores v. Nashua Mfg. Co., 1 Cir., 1946, 157 F.2d 154. Seemiller used only the materials which had been used before. Furthermore he did not draw from remote arts in making his combination. See Scott & Williams v. Lasticnit Co., 1 Cir., 1951, 186 F.2d 416. Every feature of his method and every element of his product appear in the prior automobile radiator art in which he worked. Neither did his radiator do anything that radiators had not done before, nor did it operate on any novel principle. His radiator cooled the water used to dissipate a portion of the heat generated by the internal combustion engine in common use in automobiles in the same way and by the application of the same principles that automobile radiators had always operated, and it did nothing else. At the most, he combined the best features of both the cellular and the tubular types of radiators as known to the prior art, and by so doing produced a radiator having the best features of both types without the disadvantages of either, and thus he produced a superior product. The question therefore clearly emerges as to whether today a patent covering a combination of old elements can be sustained when the new and useful result of the combination is nothing more than an improvement, perhaps a substantial improvement, in performance over anything which had gone before.
Several earlier decisions of the Supreme Court certainly stand for the proposition that marked improvement in the performance of a machine or device constitutes a new and useful result sufficient to sustain a patent for a combination of old elements. Loom Co. v. Higgins, 1881, 105 U.S. 580, 26 L. Ed. 1177; Diamond Rubber Co. of New York v. Consolidated Rubber Tire Co., 1911, 220 U.S. 428, 31 S. Ct. 444, 55 L. Ed. 527; and Eibel Process Co. v. Minnesota & Ontario Paper Co., 1923, 261 U.S. 45, 43 S. Ct. 322, 67 L. Ed. 523 may be cited as typical. In recent years, however, the Court has indicated that at least extreme caution is to be observed in sustaining combination patents when the new and useful result is merely improved performance.
Thus in Lincoln Engineering Co. of Illinois v. Stewart-Warner Corp., 1938, 303 U.S. 545, 549, 58 S. Ct. 662, 664, 82 L. Ed. 1008 the Court said citing cases: "The mere aggregation of a number of old parts or elements which, in the aggregation, perform or produce no new or different function or operation than that theretofore performed or produced by them, is not patentable invention." And in its most recent pronouncement on the subject in Great Atlantic & Pacific Tea Co. v. Supermarket Equipment Corp., supra, 340 U.S. at page 151 et seq., 71 S. Ct. at page 129, the Court after quoting the above sentence from the Lincoln Engineering Co. case with the prefatory remark that it condensed the "negative rule accrued from many litigations * * * about as precisely as the subject permits" went on with citation of intervening cases to say:
"The conjunction or concert of known elements must contribute something; only when the whole in some way exceeds the sum of its parts is the accumulation of old devices patentable. Elements may, of course, especially in chemistry or electronics, take on some new quality or function from being brought into concert, but this is not a usual result of uniting elements old in mechanics. This case is wanting in any unusual or surprising consequences from the unification of the elements here concerned, and there is nothing to indicate that the lower courts scrutinized the claims in the light of this rather severe test."
Then, after pointing out that neither court below had made any finding that the old elements which made up the device under consideration performed any additional or different function in the combination than they performed out of it, and that no such finding was possible on the evidence, which the Court summarized by saying "Two and two have been added together, and still they make only four", the Court went on to say:
"Court should scrutinize combination patent claims with a care proportioned to the difficulty and improbability of finding invention in an assembly of old elements. The function of a patent is to add to the sum of useful knowledge. Patents cannot be sustained when, on the contrary, their effect is to subtract from former resources freely available to skilled artisans. A patent for a combination which only unites old elements with no change in their respective functions, such as is presented here, obviously withdraws what already is known into the field of its monopoly and diminishes the resources available to skillful men. This patentee has added nothing to the total stock of knowledge, but has merely brought together segments of prior art and claims them in congregation as a monopoly."
We do not think the court by its language in the above quotations from the A. & P. case meant to lay down a hard and fast positive rule for resolving the question of invention in all cases involving combination patents. Indeed at the outset of its opinion it said that while the Court had sustained combination patents in the past it had never "ventured to give a precise and comprehensive definition of the test to be applied in such cases", and then it indicated the futility of any such attempt. Nor do we think the Court in its reiteration of what it called the negative rule as stated in the Lincoln Engineering Co. case intended to summarize all the law and the prophets on the subject. Therefore it may be that a patent covering a far from obvious combination of old mechanical elements, or one covering a combination which called for more than the ordinary competence of those skilled in the art to put together, can, if the combination be new and useful, still be valid even though the old elements do not unite to produce a device which performs some novel function in addition to the separate functions of its component parts. See our discussion and the cases cited in Abbott Machine Co. v. Universal Winding Co., 1 Cir., 1943, 137 F.2d 166, 167-170. Thus perhaps increased efficiency without performance of a novel function may still in rare instances be a sufficiently new and useful result to sustain a patent covering a combination of old mechanical elements. But however this may be, certain it is that the increase in efficiency of the new combination must be an "unusual or surprising" consequence of the unification, and hardly that much can be said for Seemiller's core.
The automobile radiator art has been far from static over the past fifty odd years. The history of the art is studded with improvements in the direction of lighter, stronger, and more efficient radiators per pound of weight and square inch of surface, which are cheaper to produce. Seemiller took another step in the same direction. But his step was not a revolutionary one. It was not a leap or a bound or even a stride ahead, but was only another step in the regular march of progress in the art. This is not enough to warrant a finding of invention under the "severe test" applicable in the present day of rapid advance in the arts.
Nor does the evidence that Seemiller's core filled a long-felt want, enjoyed commercial success, and was copied by the defendant, upon which the appellant heavily relies, indicate to the contrary. In the first place, these considerations are at the most weak reeds for a patentee to lean upon. They do not, either separately or in conjunction, establish invention, but only serve to tip the scales in favor of a finding of invention when that question on the other evidence hangs in the balance, as the Supreme Court has repeatedly said in the past. See Great Atlantic & Pacific Tea Co. v. Supermarket Equipment Corp., supra, 340 U.S. at page 153, 71 S. Ct. 127, and cases cited; Altoona Publix Theatres, Inc., v. American Tri-Ergon Corp., 1935, 294 U.S. 477, 487, 488, 55 S. Ct. 455, 79 L. Ed. 1005, and cases cited. In the second place the evidence of filling a long-felt want and of commercial success is not too impressive. Seemiller's core fell far short of dominating the market. Ford and General Motors, two of the largest automobile manufacturers, continued after Seemiller to make and use cores of their own of a type not covered by the patents in suit. And it does not appear that Seemiller came to dominate the field left vacant by these two manufacturers. While no doubt he filled a known want and enjoyed success, his achievements in his field would seem to be less spectacular than Turnham's achievement in the "cash and carry" grocery store field which the Court in the A. & P. case said was insufficient to tip the scales in favor of invention.
Thus we conclude that although the court below may perhaps have been in error in construing the A. & P. case as requiring as a sine qua non of invention a finding that the combination performs some new function in addition to the sum of the separate functions of its parts, nevertheless the court did not err in its ultimate conclusion that the claims in suit are invalid for want of invention, for, "[scrutinizing] the claims in the light" of the rigorous standard required by the law as it stands today, no other finding is possible on the evidence of record.
The judgment of the District Court is affirmed.