Swan Carburetor Co. v. Nash Motors Co.

WILLIAM C. COLEMAN, District Judge. This is a patent infringement case, involving intake manifolds used on automobile engines. .The sole question here presented is whether a single claim, No. 20, of Swan patent No. 1,536,044, has been infringed by the defendant, the Nash Motors — 26 — Company. The plaintiff company’s right to the patent, and the jurisdictional prerequisites are present. As a result of extended litigation in the Sixth Circuit, Reeke-Nash Motors Co. v. Swan Carburetor Co., 88 F.2d 876, the present defendant is estopped to deny the validity of the patent and of claim 20 thereof, since this claim, together with claim 13 of the patent, was held valid in a suit against another defendant, the Reeke-Nash Company, with which the present defendant is in privity. In that former suit such other defendant was held to have infringed, but not with the same devices that are in controversy in the present suit. For the purpose of convenience and clarity, those other devices are called defendant’s first group of manifolds, and the devices here in controversy, defendant’s second group of manifolds. As to the latter, this Court held, in an earlier proceeding, resulting from various motions filed by the plaintiff, that there had never been any adjudication in the prior litigation just referred to in the Sixth Circuit, and that, therefore, the present defendant, the Nash Motors Company, was entitled to its day in court on the question of whether these devices did infringe Swan patent No. 1,-536,044. See 25 F.Supp. 21. This decision was affirmed by the Circuit Court of Appeals for this Circuit. See 98 F.2d 115. In this Court’s prior opinion referred to, there is given a detailed chronology of the previous litigation involving this Swan patent and a related patent, No. 1,636,721, issued to Swan subsequently to the issuance of patent No. 1,536,044, but of earlier application, and therefore need not be repeated here. The patent in suit is titled “Method and Means to Facilitate the Distribution of Fuel in Internal Combustion Engines.” Claim 20, which relates to the means or apparatus, reads as follows: “In an inlet manifold, a distributing chamber having a single inlet conduit and three branch conduits, one of the walls of the chamber being opposite the inlet duct and symmetrically formed and situated with reference to the branch ducts so that entering fluid may be influenced by said wall uniformly in all directions transverse to the entering stream, and the branch conduits being of substantially uniform shape throughout and at any turn thereof presenting similar walls shaped and situated so that passing mixture may be influenced thereby in a manner to distribute equally to cylinders to which said turns may lead.” The intake manifold is the piping which leads from the carburetor and air intake to the several cylinders of the engine. It is the connection between the carburetor and the cylinders, conveying the fuel mixture from the former to the latter. The function of the carburetor is to mix the gasoline with air in proper proportion, and then the problem is how to insure that the resultant mixture will be most effectively and uniformly transmitted and supplied to all the cylinders where it is exploded, with resulting motive power. Generally speaking, an intake manifold has three main parts: (1) A pipe, called a riser, leading vertically from the carburetor to (2) a horizontal pipe, called a header, above the riser, leading out to (3) branches, three in number for a six-cylinder engine, to which the patent in suit is intended more specifically to apply — one branch at each end, and one in the middle, of the header, these branches in turn leading to the cylinder ports and intake valves of the engine. The primary object sought to be attained being to provide means which will enable the mixture of fuel and air to reach the cylinders in a uniform character throughout all of the cylinders, the problem is one of insuring not merely that the proportion of air to fuel in the mixture will be the same for all cylinders, but that the proportion of fuel in vapor and liquid form will be the same; and, also, that the quantity and quality of the mixture reaching each cylinder, will likewise be the same. Apparently, in the early days of the automotive industry, intake manifolds had, for the most part, been constructed as if the mixture leaving the carburetor were a gas and contained no liquid, more attention being attached to maintaining the same length of path or channel of flow to the various cylinders, or groups of cylinders, than to character of the path or channel. Shortly after the turn of the century, with the greatly increasing demand for gasoline, and the change in its quality resulting from the development of the so called “oil-cracking” process — that is, subjecting the crude oil to great heat and pressure, as contrasted with distillation of the crude oil — commercial gasoline became low in volatility. This meant that manifolds for internal combustion engines were called upon to distribute wet fuel mixture, which — 27 — included particles of unvaporized or liquid fuel from the carburetor to the several cylinders of the engine. This presented a new and serious problem, namely, how to distribute, equally, the wet or unvaporized constituents of the fuel mixture in its movement from the carburetor to the various engine cylinders. Since gases have so little inertia and are so mobile, as far as quantitative distribution is concerned it appears to have been long conceded that it makes little difference what form of piping is used, provided the passages to all of the cylinders are practically equal in length, and the bends in each are equal in number o and mean radius, and the effective areas are the same throughout each of them. However, qualitative distribution is of even greater importance. Quantitative distribution is unaffected by the presence or absence of fuel in the mixture since it deals simply with volume of gas — or air as in the situation under discussion. But qualitative distribution has to do with both the fuel and the air, and with the maintenance of the proportions as determined by the carburetor. In other words, qualitative distribution is really a problem in carburetion, rather than in distribution, and the problem of manifold design would be completely solved, therefore, if the mixture were a gas or could be treated only as such, or if the mixture were a true one; that is to say, a mixture of fuel vapor and air. But time and temperatures involved prevent the attainment of a true mixture. Some of the fuel of necessity persists as a liquid in its passage through the manifold. Naturally, the farther the fuel is made to travel, the more favorable become the vaporizing conditions since the entrained globules continually become smaller through evaporation. But apparently, a sufficient length of straight passage, so that the real manifolding or branching would contain nothing but a mixture of air and fuel vapor to handle, was found to be virtually an impossibility. In practice, the entrained fuel and that which spread over the passage walls had also to be manifolded as well as that which was really vaporized. In short, it is that part of the fuel remaining as a liquid, when bends and branches in the piping had to be negotiated, that causes all the trouble. See “On Manifolds and Distribution” by P. S. Tice, in the April and May 1911 issues of “Motor.” Defendant’s denial of infringement is of a two-fold character: First, the usual defense of lack of identity in either (1) construction, (2) operation, -or (3) results between its device and that of the plaintiff under the patent as adjudicated; and, second, existence of identity between defendant’s device and the art prior to the Swan patent, as exemplified by the Matheson manifold (not • patented); or, in other words, defendant’s device is alleged to be nothing more or less than a replica or imitation of the prior art from which the Swan patented device in the prior litigation, already alluded to, was held to differ, and necessarily so, in order to constitute the Swan device a patentable one. Recognizing, of course, the well settled principle that a right, question, or fact distinctly put in issue, and directly determined in prior litigation, cannot be disputed in a subsequent suit between the same parties or their privies; and that even if such a second suit be for a different cause of action, any right, question, or fact once so determined previously must, as between the same parties or their privies, be taken as conclusively established, this Court, in the present proceeding, permitted the defendant to introduce testimony in regard to the prior art, not, however, for the purpose of disputing what had been established with respect thereto in the previous litigation by which, as already explained, the present defendant, by reason of privity, is estopped to deny the validity of the Swan patent in suit, claim 20, but solely for the purpose of permitting defendant to compare its alleged infringing device with the prior art, in an effort to demonstrate that there is identity between the two. In other words, in a case of the present type, a defendant is not confined to proof of dissimilarity between its alleged infringing device and plaintiff’s patented device, including all devices previously adjudicated to infringe the latter on the principle of equivalents, but has the additional right to submit proof of the identity of its device with the prior art, because, if successful in either effort, non-infringement is proved. Westinghouse Electric & Mfg. Co. v. Formica Insulation Co., 266 U.S. 342, 45 S.Ct. 117, 69 L.Ed. 316; Skelton v. Baldwin Tool Works, 4 Cir., 58 F.2d 221, 225; General Motors Corp. v. Swan Carburetor Co., 6 Cir., 88 F.2d 876, at page 884. — 28 — As explained in the previous opinion of this Court resulting from the preliminary-proceedings, the factual questions presented are, in their nature, most complicated, and capable of accurate determination only with the aid of scientists in the particular field of engineering, and with the aid of experiments conducted by such scientists; and therefore, it would be impossible for this Court, without such aid, in addition to what has already been disclosed as a result of the previous litigation involving the patent in suit and the other related patent, to say whether the alleged infringing devices of defendant are equivalents of plaintiff’s device within the meaning of the patent law. At the request of the Court, the parties to the present controversy have very happily agreed upon a neutral, interpartes expert witness, Professor Alexander Graham Christie head of the Department -of Mechanical Engineering at Johns Hopkins University, whose high standing in this field of engineering is unquestioned; he has observed extensive laboratory and road tests initiated and witnessed by each side, and his testimony has been elaborately developed in the present hearing, as well as the testimony of the expert witnesses for each side, who likewise are of high standing in their profession. The Court cannot over-emphasize the extent of the aid which such neutral expert has rendered to the Court by his clarification of the issues. First, we will consider whether, as defendant claims, its device is, in fact, a reversion to Matheson so as to bring it within the prior art as disclosed by Matheson. This entails naturally a determination of what Matheson was adjudicated to embrace that was different from the Swan patent. If we take literally the language employed in the only written opinion rendered by a District, that is, a trial court in the Ohio litigation, it would appear that the distinction was understood to exist primarily in the fact that the Swan patent required, in order to insure uniform distribution of the gasoline mixture to all cylinders — -the primary object sought to be attained and found by the courts of the Sixth Circuit to have been attained by Swan but not previously attained by anyone else — rectangular corners on the inside or “near” sides of the inlets from the header to the cylinders, whereas such corners are rounded in Matheson. For example, the trial Court (Judge Westenhaver) said (Swan Carburetor Co. v. General Motors Corp., D.C., 42 F.2d 452, at page 455): “The gist of his [Swan’s] invention consists in bringing the gaseous mixture from the carburetor to the header in perpendicular or straight lines, then abruptly changing its course at right angles in the header, and then again changing its course at right angles from the header into the branches. All other features of Swan are subsidiary. The dome or flat wall at the intersection of the riser and header, the recesses in the outer bend at the intersection of header and branches, the flat or level floor, and even a square of rectangular cross-section, while stressed, are not made by Swan the substance of his invention.” And continuing, the Court said (page 455): “His [Swan’s] fundamental conception is and was that by changing abruptly at right angles the line of travel of the mixture, and by eliminating all obstructions, curves, or recesses in that line of travel, the tendency of the wet particles to collect at one place more than another would be avoided, and that, whenever the line of travel was thus abruptly changed at right angles, the wet particles would be thrown toward or beyond the main current and be mixed with the air and fully vaporized. In this way he asserts that an area of agitation or turbulence or mixing zone is created at each intersection where a sharp right angle turn is made, and thereby the mixture is completely vaporized before it reaches the engine ports, and that no branch or port is more favored in the distribution than another.” And still further (page 456) : “He [Swan] further notes that, while the outer angle of the branches leading to the cylinder may be curved or rounded, the inside angle of the turn must be sharp. He says: ‘Good results may be obtained in a modification of this character, in view of the fact that the wet constituents of the fuel will be caused to be projected beyond the sharp inside angle into the air at the turn, effecting a remixing of the constituents incident to the changed direction thus caused at the turn.’ ” However, a closer analysis of that opinion clearly shows that the Court was not called upon to, and did not, give to claim 20 the consideration now required. In fact, that claim is not only not analyzed in the opinion, but not even separately referred to, and this for the reason that the Court — 29 — was concerned, as hereinafter explained, with a different question from that here presented, namely, whether a manifold with both header and branches or inlets to the cylinders, rounded in form, infringes the Swan patent. When we analyze the opinion of the appellate court in that litigation, we find the patent to have been interpreted in its broader aspects, embracing specific reference to the different claims. The Court said (General Motors Corp. v. Swan Carburetor Co., 6 Cir., 88 F.2d 876, at page 886), that Swan’s claim was successfully proved that he was the first to secure “substantially complete vaporization. He did this mainly by the creation of a maximum turbulence at all points where the direction of the flow was changed. The result was that the vaporized fuel which tended in the manifold to revert to liquid was revaporized. Swan also secured equal distribution by directing the properly mixed fuel through pipes in which there was a complete absence of curves or pockets in the longitudinal line of travel in either header or branches.” The Court further said (page 887): “Swan’s inventive concept consisted in bringing the fuel mixture from the carburetor to header in substantially straight lines, and so abruptly changing its course in the header and at the branches as to create a maximum turbulence at the points where direction of flow changed with resulting re-mixture of the heavy particles of gasoline and air. This concept was new. With it Swan combined forms which, while old, produced a straight line in the longitudinal flow of fuel mixture. This combination was new. The turbulence and its creation within a structure which avoided the dangers of excessive puddling were his main contribution to the art. * * * A materially better result was obtained, and this result is accomplished alike by the round and the square cross-sectional form. * * * * * * * * “It is true that claims 11, 12 and 13 of this patent describe ‘walls the intersections of which form straight lines,’ but the remaining claims are not thus restricted. “Whether or not patent No. 1,536,044 described a pioneer invention, as in effect held in the General Motors Case, [42 F.2d 452], under the specifications, which are to be read with the claims, though they may not expand or limit them, Swan is entitled to a range of equivalents broad enough to cover appellant’s manifold. “This is true even as to claims 11, 12 and 13, in which, as in the other claims, the gist of the inventive concept was' the creation of the maximum turbulence as above described, not restricted to any particular form" (Italics inserted.) When we examine these same opinions in the Ohio litigation for the purpose of ascertaining what was concluded with respect to Matheson, we find complete accord between the trial and the appellate court on the point that, as respects equal distribution, Matheson, as demonstrated, had failed. For example, the trial court said (Swan Carburetor Co. v. General Motors Corp., D.C., 42 F.2d 452, at pages 456, 457): “On this hearing, much testimony was introduced pertaining to the Matheson manifold, including an original specimen. It is illustrated also in the Tice article, and its priority and use are beyond dispute. * * * The riser, header, and branches are circular in form, but they do not embody the substance of Swan’s invention. The riser is widely flared at its intersection with the header, and the branches turn towards the engine on a gradual curve. The flaring of the riser top would slow down the travel of the mixture, and would permit the wet particles to adhere to and follow the flaring of the curve rather than be projected perpendicularly from the riser into the zone of turbulence formed by the intersection of the riser and header. The same tendency is present at the curved intersections of the header and branches. These differences distinguish fundamentally Matheson and Peerless from Swan. * * * As to Matheson and Peerless, it may be said that both disappeared from the market years ago and were supplanted by other manifolds, while Swan’s, since its appearance, has not only been adopted by defendant, but by numerous other makers of automobiles, and is enjoying a wide and extended commercial success.” Likewise, the appellate court said (General Motors Corp. v. Swan Carburetor Co., 6 Cir., 88 F.2d 876, at pages 886, 887) : “A great number of prior art manifolds was introduced in evidence, only the most pertinent of which we will consider. The Murray and Tregurtha manifold has heretofore been discussed in the royalty case. It did not perform according to the Swan principle, and its distribution — 30 — was faulty. This is equally true of Matheson, Peerless and Fay & Bowen, which must rank as prior efforts and failures.” Thus it is clear that the Ohio courts concluded and stated without equivocation that Matheson was a failure in so far as the result sought to be attained and actually attained by Swan as early as 1910 or 1911 was concerned, as of which time we must consider the status of Matheson. If a prior device, although of like construction, is not proved to be capable of performing the same function, it does not anticipate. Carnegie Steel Co. v. Cambria Iron Co., 185 U.S. 403, 424, 22 S.Ct. 698, 46 L.Ed. 968. There does not appear to have been in connection with the Ohio litigation any tests of the actual Matheson device under operating conditions that were comparable with the tests that have been made in connection with the present litigation with respect to the Swan patent, and the Nash accused device. That is to say, whilq there was testimony in the prior litigation from persons who drove automobiles equipped with Matheson manifolds, that they attained a certain degree of satisfactory operation from such, equipment, and also proof of certain tests showing the relative flow of the mixture to the different cylinders, there was no actual proof that satisfaction in the sense of equal distribution to all cylinders could be attained under the different extremes of temperature or of load and speed. In short, while the Matheson manifold was, it is true, used commercially for a number of years, actual proof is lacking as to the exact character of distribution of the gasoline mixture to all cylinders throughout the range of operating conditions under which a motor car was then normally used. Nevertheless, the proof is most presumptive of the fact that equal distribution could not be obtained. Holes were even bored experimentally into the header in an effort to get equal distribution. When we subject the new evidence produced in the present case as to what the Matheson device accomplishes to the same wort of analytical review, we likewise find no proof of precisely what that device, as constructed and used as early as 1910 or 1911, could accomplish, because the tests purporting to demonstrate these results were made upon a device that deviated in many respects from the actual Matheson manifold of that day, and, furthermore, they were not made under different extremes of temperature. For example, it appears that the bore of the riser was changed and a heater was added to the riser, and, also, that the device was not tested under cold weather conditions. It is true that, when we turn to the matter of results, Professor Christie, whose testimony, especially on this phase of the case, is believed to be of great probative weight because he testified without the slightest bias that might attend an advocate, stated that he thought substantially the same results would have been obtained as were obtained with the Nash accused device, had the Matheson device, constructed precisely as known to the courts in the previous litigation, been capable of operation with substantially the same velocities and the same rates of flow, and had it been so operated. However, this is a speculation, not even an asserted, much less an established, fact. In short, Professor Christie said that little emphasis should be placed upon the fact that a splashing motion given to the mixture is more extensive or dominant when rectangular inside or “near” corners are employed where cylinder inlets and the header join, than when round corners are there employed. His drawings of what he observed throughout all of the comprehensive tests show the same general characteristics in the turbulence, the amount of splashing of the mixture as contrasted with the rolling motion imparted to it under the same velocities, decreasing in extent as the corners of the inlets from the header to’ the cylinders were less acute. As to ultimate results, Professor Christie is even more emphatic in his conclusion that they were the same, except for fractional variations which he finds immaterial. He admits there must be some limit of efficient radius, depending upon how far the curves are “flattened,” but he did not determine precisely where the limit was — merely that it had not been reached with any of the devices, including Matheson, exhibited to him, and upon which tests were witnessed, but had been exceeded with a model of his own design, constructed with the “flattening” of the curves very much more exaggerated, and similarly tested. While it is axiomatic that the sustaining of a patent upon a differentiation from the prior art, does not authorize the successful party to gather to himself a monopoly of what was old when he entered the field, it is equally axiomatic that when it is sought to avoid the charge of infringe — 31 — ment through reliance upon the prior art, there must be clear proof that the alleged infringing device falls within that prior art, rather than within the scope of the patent, especially when the patent is, as we believe to be true in the present case, if not actually a basic patent in the strict sense, entitled, nevertheless, to a broad range of equivalents. We conclude that such clear proof is lacking from the point of view of functioning and results. Turning from a consideration of functioning and results, to a consideration of the actual means employed, namely, form of construction, it must be conceded that in this respect the accused Nash device is more like Matheson as exhibited both to the Ohio courts and here, than it is like the Nash adjudicated device. First, as we have seen, it embodies, like Matheson, curved corners such as exist in none of the adjudicated devices which are res adjudicata in the present case. However, in the second suit brought by the present plaintiff against General Motors Corporation in Ohio, devices with corresponding inside or “near” corners similarly but not as much rounded, were held to infringe (see General Motors Corp. v. Swan Carburetor Co., supra), this fact being definitely established by exhibits taken from that case and introduced into the present case. By the doctrine of comity, much respect, at least, should be accorded this decision. Furthermore, in the proceeding against the Reeke-Nash Co. (the rulings in which bind the present defendant), the Master found the following (finding No. 21) : “Nothing in the prior art restricts the claims of the patents in suit heretofore enumerated, which define the apparatus or combination of elements patented therein to the preferred form of Swan’s patented manifold with square or rectangular cross-section, or with sharp right angled inside corners or with flat walls opposite the riser and at the ends of the header, and nothing in defendants adoption of the manifolds here charged to infringe of round cross-section with slightly rounded inside corner and wholly or partly rounded walls opposite the riser and at the ends of the header, is in such accordance with any prior art mmiifold or the teaching of any prior art patent or publication or is in such accordance with any prior art construction with respect to function, mode of operation, or results that any of defendanfs manifolds can be said to fairly differentiate from the patented construction or combination, or can be said to be made in accordance with or in substantial accordance with any manifold or complete combination shown to have existed in the prior a/rt.” (Italics inserted.) This finding, along .with the Master’s other findings, was adopted without opinion by the District Court, and presumably was affirmed by the appellate court, for the latter said, as we have already pointed out, that this Swan patent, whether or not it be called a pioneer invention, is, under the specifications, when read in connection with the claims, “not restricted to any particular form.” We see no reason to believe that this statement is not to be taken literally. That is, the words, “any particular form”, must be interpreted as intended to embrace the structure forming the connection between the header and the cylinder inlets which is the subject of the present controversy, and not confined to formation of the header itself, or other parts of the manifold, as are clearly the words, used earlier in the opinion, “the rounded and the square cross-sectional form.” We do not attempt to give to the patent, by virtue of the Master’s finding, above quoted, scope which the appellate court did not give to it. But, as explained, this finding was affirmed without opinion by the trial court, and the decree of the latter which was issued on these findings, was affirmed by the appellate court. So, we are forced to the logical conclusion that the scope which the Master gave to the range of equivalents of the Swan patent, was intended to be adopted by both the trial and the appellate court, although we recognize, and so held in our earlier opinion in the present case, which was affirmed by the Circuit Court of Appeals, 98 F.2d 115, that, since the Nash accused device as presented in the present case was never, in fact, introduced in the prior litigation except for the purpose of illustrating the operation of the Matheson patent in an attempt to show that the latter was not a failure, defendant is not estopped by anything that was said or done in the prior litigation to disprove, if it carr, infringement by the Nash device here accused. Second, the top of the chamber over the riser in all of the varieties of the Nash accused devices is also like Matheson, namely rounded; and, also like Matheson, those varieties have wide of rounded — 32 — curves at the bottom of the two right and left exits from this chamber, but have a sharper curve into the center branch or outlet, as opposed to a sharp corner at the bottom of all three exits of the chamber in the Nash adjudicated device; and in at least some of the adjudicated devices the tops of this chamber are of different construction. We adopt Professor Christie’s view that such features are rather inconsequential variations and such as are permissible under the Swan patent. This brings us appropriately to a consideration of precisely what the specifications and the claim in suit in the Swan patent prescribe. It is true that Swan specifies a preference for the rectangular inside corner, but the most analytical search of the specifications fails to remove all doubt that he intended to make it a prerequisite in all types of construction under the patent, and, indeed, such an intent is squarely contradicted by the very language of claim 20, as will hereinafter be more fully alluded to. Swan illustrates in one of the drawings accompanying the patent, a detailed section of one of the branches, or inlets, to an end cylinder or cylinders having the inside angle of the turn sharp and the outside rounded. Explaining this in the specifications, Swan says: “Good results may be obtained in a modification of this character in view of the fact that the wet constituents of the fuel will be caused to be projected beyond the sharp inside angle into the air at the turn, effecting a remixing of the constituents incident to the change of direction thus created at the turn.” (Italics inserted.) There is no other passage in the specifications which lays as great emphasis upon the employment of a “sharp inside angle.” The nearest is the following succinct statement of the functional theory of the patent: “The mixture from the carburetor will move in rectilinear lines to the distributing zone where it takes a right angle turn and after it leaves said zone and in passing through the manifold branches the mixture will not pass by any cylinder port but will flow in similar rectilinear lines through the branches and where required, take right angle turns into the outlets and cylinders as demanded by the induction cycles of the engine.” (Italics inserted.) When we turn to the language employed in the claim (No. 20), we find no mention whatsoever of right angle turns into the outlets, but merely that “the branch conduits” shall be “of substantially uniform shape throughout and at any turn thereof presenting similar walls shaped and situated so that passing mixture may be influenced thereby in a manner to distribute equally to cylinders to which said turns may lead.” With respect to the characteristics of that part of the header opposite, that is, just above the riser, all that is required is that it shall be “symmetrically formed and situated with reference to the branch ducts so that entering fluid may be influenced by said wall uniformly in all directions transverse to the entering stream * * * This is affirmed by the specifications, wherein it is expressly provided that the bore of the riser, although preferably square, may be of other shades and may have its top recessed to provide a dome, provided only that “the top surface is of a shape not to favor any particular outlet.” We concur in Professor Christie’s construction of the language just quoted from the claims and specifications as obviously permitting a considerable latitude of equivalents, the word “symmetrically” as used in the claim being entitled to a practical, engineering interpretation, rather than a literal or dictionary interpretation. Defendant’s expert, Professor Brown, asserted that the Nash adjudicated is a “hybrid,” because, having rounded outer edges or corners and sharp inside corners for the cylinder inlets, it possesses to a certain degree the characteristics of the Swan device as preferably constructed, this so-called hybrid construction forcing the mixture to turn the corners sharply, but giving little evidence of what was described as a double cork-screw, spiral movement imparted to the mixture where rectangular corners, both inside and outside are employed. On the other hand, he testified that, in his opinion, the Nash accused may not properly be classified as a “hybrid” for the reason that it has rounded inside corners. That is to say, he claimed that the Nash adjudicated is in a twilight zone, and that we definitely pass out of that zone with the Nash accused. However, this distinction seems too artificial and tenuous, and we are not persuaded by it. Professor Christie’s testimony, supported by his drawings, is to the contrary, namely, that in operation and flow of the mixture there is no material difference whether the inside corner be square or round. This we accept, because we do — 33 — not believe it has been successfully refuted. It is not necessary that an inventor, in order to secure a patent, shall disclose, or even know all the scientific principles involved in his invention. This means that his invention may, after issuance of a patent upon it, and its commercial use, be found to embrace scientific principles and to produce practical results which the inventor himself never contemplated it would embrace or produce. Diamond Rubber Company v. Consolidated Rubber Tire Company, 220 U.S. 428, 435, 436, 31 S.Ct. 444, 55 L.Ed. 527. It must be conceded that Swan, when he conceived his idea and disclosed it in patent application form, thought the best way to attain the desired end, namely, equal distribution of the full mixture to the various cylinders, was to direct the mixture through pipes in which there was a complete absence of curves or pockets. The Circuit Court of Appeals for the Sixth Circuit concluded, as we have seen, that this was at least one of the factors in Swan’s success in attaining the desired end. See General Motors Corp. v. Swan Carburetor Co., supra, page 886. But Swan did not claim this was the only way he could do it, because, as we have also seen, claim 20 of his patent, here in suit, describes a construction that permits of curves, including curves at any turn where the branches or inlets to the cylinders leave the header, provided only these branches are “of substantially uniform shape throughout, and at any turn thereof” present “similar walls shaped and situated so that passing mixture may be influenced thereby in a manner to distribute equally to cylinders to which said turns may lead.” It seems reasonable to say that these words should be given such meaning, and not merely interpreted as a re-statement of all the other apparatus claims, notably five of them (Nos. 11, 12, 13, 15 and 21), which expressly require straight lines at these intersections, the validity of one of them (No. 13) the present defendant being here estopped to deny. Claim 13 reads: “In an inlet manifold, a distributing chamber having a single inlet conduit and a plurality of outlet conduits, said chamber being formed of walls the intersections of which form straight lines, one of said walls being opposite the inlet duct and symmetrically shaped and situated relative thereto, so that entering mixture may be influenced by said wall uniformly in all directions transversely to the entering stream.” Clearly, this claim permits of acute, or obtuse, as well as right angles. The restriction is merely that the intersections shall form straight lines. There is nothing in claim 20 requiring that the intersections shall do so. We are not unmindful of the fact that the language of every claim in a patent must be viewed in the light of the specifications and drawings, and that a claim broader than the invention as described in the specifications is void. Carlton v. Bokee, 17 Wall. 463, 21 L.Ed. 517; Victor Cooler Door Co. v. Jamison Cold Storage Door Co., 4 Cir., 44 F.2d 288. That is to say, the claims must be responsive to the specifications. They must have a reasonably sensible coordination with the specifications. They are, however, entitled to reasonable flexibility of interpretation, depending upon the state of the art in question; and similarly and conversely, the specifications are entitled to a certain flexibility of interpretation. Applying to the present case the rule as we thus understand it, we believe that the specifications and claim 20 remain sufficiently responsive one to the other, when the interpretation herein given is placed upon them. The validity of claim 20 is no longer subject to question or review, so the problem before this court is solely one of interpretation of the specifications and the claim as applied to a new set of facts. The correctness of the scope which we give to claim 20 is further strengthened by the second, that is later in issuance, Swan patent, No. 1636721. It issued on an application filed more than three years prior to the application in No. 1536044, and expressly provides both in its specifications and in all of its claims that the mixture, in turning into any cylinder, must always make an abrupt bend, preferably at a right angle. General Motors Corporation took a license in 1922 under this application, and at first made only square shaped and square cornered manifolds, but later adopted a manifold rounded in shape as to cross-section, and with the other modified characteristics which we have heretofore considered under the name of the Nash adjudicated device, because it was this device, upon whch General Motors Corporation was sued, and which was found to in — 34 — fringe, in the Reeke-Nash Case, 88 F.2d 876. That is to say, after General Motors Corporation adopted this different type of manifold, Swan applied for, and obtained his patent No. 1,536,044, and thereupon sued General Motors Corporation. In short, Swan wanted patent protection for improvements upon his original application in No. 1,636,721, which, it is to be noted, had not ripened into a patent at the time of this previous litigation just referred to. Now that we have, in the present litigation, the complete history of both patents clearly before us, the problem is somewhat clarified, and, as we have said, we are strengthened in our conclusion. Were we to sustain defendant’s contention, it would be tantamount to saying that the prior art prevents Swan from having the broad protection for his invention which we clearly believe “the improvement” patent in suit gives him; or, in other words, that this court should, in effect,- invalidate a claim of this patent, although the parties are bound by the previous adjudication of its validity. It is fundamental that identity of means and of operation must be -combined with identity of results in order to constitute infringement. Westinghouse v. Boyden Power Brake Co., 170 U.S. 537, 569, 18 S.Ct. 707, 42 L.Ed. 1136. It is true, identity of means, for the reasons just stated, is closer between the Nash accused and Matheson than between the Nash accused and Nash adjudicated, but close enough between the latter to bring the Nash accused within the scope of the patent and its reasonable equivalents; and certainly, identity of operation and results, as has just been shown, is much closer between Nash adjudicated and Nash accused than between Matheson and Nash accused. A close copy which uses the substance of an. invention, and substantially the same device performing precisely the same offices, with no change in principle, constitutes infringement. Sanitary Refrigerator Co. v. Winters, 280 U.S. 30, 50 S.Ct. 9, 74 L.Ed. 147. Such we find to be the case here. Furthermore, the commercial success that has attended the Swan patent is another important factor that bespeaks for it, in the patent law, liberality of interpretation. “In administering the patent law, the court first looks into the art, to find what the real merit of the alleged discovery or invention is, and whether it has advanced the art substantially. If it has done so, then the court is liberal in its construction of the patent, to secure to the inventor the reward he deserves. * * * Indeed, when one notes the crude working of machines of famous pioneer inventions and discoveries, and compares them with the modern machines, and processes exemplifying the principle of the pioneer discovery, one hesitates in the division of credit between the original inventor and the improvers, and certainly finds no reason to withhold from the really meritorious improver, the application of the rule ‘ut res magis valeat quam pereat,’ which has been sustained in so many cases in this court. * * * ” Eibel Process Co. v. Minnesota & Ontario Paper Co., 261 U.S. 45, 63, 43 S.Ct. 322, 328, 67 L.Ed. 523. If it be said that by the conclusion here reached no recognition is given to the existence of a prior art, as exemplified by Matheson and accepted by the rulings in the Sixth Circuit by which the parties to the present litigation are bound to the extent heretofore stated, suffice it to say that we do recognize the existence of such prior art to the following extent, and that only to such extent is it clear from those prior rulings that we are bound, namely, that Matheson had never demonstrated that he had secured that degree of turbulence of the mixture necessary to obtain equal or substantially equal distribution to all cylinders, although he had attempted to do so by a device whose construction was closely akin to that of the present Nash accused device. This coincides with the proof as to Matheson that has been adduced in the present case, because, just as in the previous litigation, there is no proof that Matheson’s device as constructed and used at the time Swan conceived and patented his device, was in fact other than an unsuccessful effort at equal distribution as then and still understood, notwithstanding speculative evidence to the contrary as a result of tests made upon so called Matheson devices of definitely modified design. Clearly, such unsuccessful effort is not anticipation, as understood in the patent law. If it were, an inventor’s success would no longer be “his title to consideration.” See Diamond Rubber Company v. Consolidated Rubber Tire Co., supra, page 437, 31 S.Ct. 444. — 35 — Nor is this court here called upon to suggest what it might have done, had it been called upon to consider the validity of the Swan patent, in the light of the prior art as disclosed by the additional proof in the present case. Suffice it to say on this'point that we do believe the method of presentation of the testimony to this court, controlled and restricted as it was, to a large extent, by the help of the neutral inter-partes expert witness, has greatly clarified the precise issue, and apparently has removed from the case a welter of testimony which could serve no purpose other than to confuse that issue. Shorn of all confusing technicalities the present case comes down to this: (1) The patent claim is valid, — that can no longer be disputed; (2) tests conducted in connection with the present litigation appear not only to have been of a more incisive character, due largely to the guidance of the neutral inter-partes expert — lacking in the previous litigation — than were the related tests in the prior litigation, but to have proved quite conclusively that overemphasis has heretofore been given to the effect of right angle, as opposed to curved inside corners where the header and the cylinder outlets meet, because the desired turbulence is now known to be capable of creation substantially as well by the one form as by the other, with substantially the same desired result in either case, namely, substantial equality of distribution of the mixture, both quantitatively and qualitatively, to all cylinders; (3) this turbulence is not itself defined in the claim nor in the specifications in any such manner as to restrict the patent to a device that will only create it in given amounts; therefore, (4) even if it be assumed that the claim, when read in the light of the specifications, must be said to contemplate the use of the right-angle inside corners exclusively, the use of curved ones is but a practical, engineering equivalent, and, therefore, an infringing equivalent within the meaning of the patent law. “It is only ingenious diversities of form and proportion, presenting the appearance of something unlike the thing patented, which give rise to questions; and the property of inventors would be valueless, if it were enough for the defendant to say, your improvement consisted in a change of form; you describe and claim but one form; I have not taken that, and so have not infringed. ****** “Where form and substance are inseparable, it is enough to look at the form only. Where they are separable; where the whole substance of the invention may be copied in a different form, it is the duty of courts and juries to look through the form for the substance of the invention; for that which entitled the inventor to his patent, and which the patent was designed to secure; where that is found, there is an infringement; and it is not a defence, that it is embodied in a form not described, and in terms claimed by the patentee.” Winans v. Denmead, 15 How. 330, 342, 14 L.Ed. 717. For the reasons above set forth, plaintiff has sustained the burden cast upon it of proving that its patent has been infringed. A decree will be signed in accordance with this opinion, including the appropriate reference to a Master for an accounting.