Opinion
No. 5407.
July 1, 1938.
Brockett, Hyde, Higley Meyer and John W. Meyer, all of Cleveland, Ohio, Drury W. Cooper, Sturges S. Dunham, and C. Blake Townsend, all of New York City, and W.D. Keith, of New Kensington, Pa., for plaintiffs.
John H. Bruninga, of St. Louis, Mo., and Kwis, Hudson Kent, of Cleveland, Ohio, for defendants.
Patent infringement suit by the Aluminum Company of America against Thompson Products, Inc.
Decree in accordance with opinion.
This case is before the court on exceptions to the master's comprehensive report. The first patent in suit is to Archer and Edwards for making castings of aluminum-silicon alloy, No. 1,572,459, dated February 9, 1926 on application filed November 20, 1920. The claims in suit are Nos. 1, 2, 3, 4, 6, 7 and 13, of which 1 to 4 are method, and 6, 7 and 13 product claims. It is sufficient to copy Nos. 1, 4, 7 and 13.
"1. The method of making castings of aluminum-silicon alloy, comprising filling the mold with the molten alloy and causing the silicon in the alloy to take the form of minute particles by solidifying the alloy rapidly."
"4. The method of making castings of aluminum-silicon copper alloy, comprising filling the mold with the molten alloy and by rapid cooling causing the alloy to solidify with a fine-grained structure and the precipitating silicon to take the form of minute particles."
"7. As a new product, a chill casting of aluminum-silicon alloy containing not less than about 3 nor more than about 15 per cent silicon, characterized by minute subdivision of the silicon, and high strength and ductility."
"13. As a new product, a chill casting of aluminum alloy containing substantial amounts of silicon and copper, characterized by minute subdivision of the silicon and increased strength and ductility."
The master found all these claims except 7 invalid for failure to state proportions of the constituents of the alloy, and refused to read such proportions into the claims from the specification. He held that only when the claim is ambiguous or where elements of an article claim are referred to in terms of means to perform a function can the specification be resorted to to interpret the claim; to this might have been added inoperative claims.
The specification recites: "The amount of silicon in the alloy can be varied considerably but should not, in general, be less than about 3 nor more than about 15 per cent."
As frequently held by the courts, the claims measure the invention. But they in turn are limited by the specification, and nothing not there disclosed can properly be claimed; consequently all claims in suit, either for the process, one step of which, or for the article, one element of which, is an alloy of aluminum-silicon in proportions substantially different from those stated in the specification, are void. Such claims, except No. 7, call for a combination of aluminum and silicon, the latter in any possible proportion; in that respect they are broader than the invention; and as this clearly appears on the face of the patent, and the claims are not ambiguous, the case is not one for interpretation, but for disclaimer under R.S. § 4917, 35 U.S.C. § 65, 35 U.S.C.A. § 65.
Our Circuit Court of Appeals, in Lamb Knit Goods Co. v. Lamb Glove Mitten Co., 6 Cir., 120 F. 267, 269, said: "If the patent, when construed upon what stands 'within its four corners,' claims more than the actual invention, the patentee must disclaim the excess in order to save that to which he is really entitled."
In Hildreth v. Lauer Suter Co., D.C., 204 F. 792, 797, Judge Rose remarked: "It is hornbook law that a claim which covers more than the patentee really invented is invalid."
As to the importance of the proportions of the alloy's constituents, see In re Lowry, 93 F.2d 909, 36 U.S.P.Q. 254, decided by the Court of Customs Patent Appeals. The court approved the views of the Examiner, saying: "He further stated that claims to an alloy should specifically state the percentages of the cobalt-nickel composition as well as the percentages of other ingredients because, he said, 'it is well known small changes in percentages of the ingredients often produce alloys of totally different characteristics." And it will be noted that although the application as amended stated a range of percentages of both nickel and cobalt, the court refused to read them into the claims "which were rejected as being indefinite and broader than appellant's invention."
In Ludlum Steel Co. v. Terry, D.C., 37 F.2d 153, there was a disclaimer of limitations in a claim of certain percentages of chromium and silicon going to make up an improved alloy.
In Crown Cork Seal Co. v. Sterling Cork Seal Co., 6 Cir., 217 F. 381, 386, 387, it is said: "Only when necessary to make the claims operative, or in case of ambiguity apparent on the face of the claims, or induced by their study in connection with the specification and prior art, is a court permitted to read in an element not expressly named therein, in order to narrow a claim, so as to make valid one otherwise invalid. McCarty v. Lehigh Val. R. Co., 160 U.S. 110, 16 S.Ct. 240, 40 L.Ed. 358. This court has applied this rule, and has refused thus to narrow a claim, and confine it to what turned out to be the real invention, in many cases (see National Cash Register Co. v. Gratigny (C.C.A.) 213 F. 463, 467); and we have, for that purpose, limited a claim beyond its apparent scope only in cases like Lamb Knit Goods Co. v. Lamb Glove Mitten Co. (C.C.A.) 120 F. 267, 269, when it contained some general term of reference which was capable of being so narrowed by interpretation without importing an element distinctly foreign."
The last expression quoted does not mean that our Circuit Court of Appeals holds that every claim capable of being given a narrow construction by reference to the specifications will be so construed in order to save it. What is decided on this point in Lamb Knit Goods Co. v. Lamb Glove Mitten Co., supra, is found on page 269. Judge Severens announces the general rule that claims must stand or fall as made, citing Keystone Bridge Co. v. Phoenix Iron Co., 95 U.S. 274, 24 L.Ed. 344; states the exceptions permitting construction by reference to the specifications, for certain limited purposes; and then says that when the claim specifically calls for matter not described at all, "this can not be done", such construction being allowable only where there is a description to which the claim, when correspondingly restricted, may apply. Having disposed of the matter of construction of the claim so far as called for by the case, he proceeds to deal with claims like those here, in the language respecting disclaimer quoted above.
In Union Switch Signal Co. v. Kodel Elec. Mfg. Co., 6 Cir., 55 F.2d 173, appears the following:
"Claims 1 and 2 call for 'copper having cuprous oxide formed thereon, said combination constituting in and of itself a uni-directional current carrying device.'"
"Claims 1 and 2, however, cannot be sustained. These claims would be met by a piece of copper of any size or shape having thereon a splotch of cuprous oxide of any size or shape. There is no suggestion in either of them of internal boundary rectification over a relatively wide area, which, as we have said, is the essence of the Grondahl invention." Page 176.
The last sentence suggests finding in the broad claims under discussion the essence of the invention, to wit, dispersion or precipitation of silicon in the form of minute particles, and therefrom an implied limitation of the proportion of silicon to such percentages as will cause or allow the dispersion or precipitation. But this would run counter to Wm. B. Scaife Sons Co. v. Falls City Woolen Mills, 6 Cir., 209 F. 210, 213. In that case, to avoid an anticipation, the patentee sought without success to have claim 10, which called for a filter, construed as intending a filter of specific construction shown and described. The court held that where the improvement is described in the claim in terms capable of either broad construction rendering it invalid in view of the prior art, or of a narrower construction preserving its validity, the courts will give it the latter, with this important qualification: "But where, from the specification or the history of the application or the language of the claim it is clear that the patentee intended to claim and the Patent Office intended to grant the broader monopoly which turns out to be invalid, the courts will not, for the arbitrary purpose of saving the claim, read into it a limitation which it does not have."
The Archer Edwards specification discloses the proportions of constituents and from it alone no purpose to seek the broader monopoly can be deduced. The history of the application contains some fairly strong indications of such purpose, for originally some of the process claims (now in unlimited form) were limited to proportions of silicon named in the specification. See Ex. 49, pp. 10-12. This is true of product claims, both as originally filed and as granted. Ex. 49, pp. 12-14. See the Scaife Case, supra, where the court regarded the fact that claims other than No. 10 called for the peculiar form of filter, as important. Symington v. National Mall. Cast. Co., 250 U.S. 383, 385, 39 S.Ct. 542, 63 L.Ed. 1045, and Smith v. Snow, 294 U.S. 1, 14, 55 S.Ct. 279, 79 L.Ed. 721.
Finally, the language of the claims under examination, so far as any evidence before the court goes, was deliberately employed for a definite purpose, and leaves no reasonable doubt that it was the intention of the patentees to claim and of the Patent Office to grant the broad monopoly expressed by such terms. A narrower construction will not preserve to the patent "the validity which it should have had". Scaife Case, page 213. (Italics mine).
These claims, except No. 7, are not entirely within the invention, never possessed validity, and the court cannot arbitrarily save them by construction. All call for chill casting or rapid solidification, both ambiguous terms which can properly be interpreted by the specification. Schick Dry Shaver, Inc., v. Dictograph Prod. Co., 2 Cir., 89 F.2d 643; Ex parte Brownell, 33 U.S.P.Q. 242. The proportions of silicon are mentioned in claim 7, and it is valid; but for want of a similar call, the others, while not lacking certainty, are void.
The recent decision of the Supreme Court in General Elec. Co. v. Wabash Appliance Corp. et al., 304 U.S. 364, 58 S.Ct. 899, 82 L.Ed. 1402, cited by both parties, emphasizes that a court trying a patent case should do nothing that "would constitute an improper importation into the claim[s] of a factor nowhere described there" [page 904]. I do not see that it has any further direct application.
Assuming the foregoing views incorrect, the question of validity of the claims in view of the prior art arises.
The patent states that the invention relates to aluminum alloys containing silicon in substantial amount; and that its chief object is to provide a simple method of making aluminum-silicon castings of improved physical properties, such as tensile strength and ductility, and that the invention is based on the discovery that such properties "are markedly affected by the rate of solidification of the molten alloy, rapid cooling to complete solidification producing a very notable and advantageous increase in tensile strength and elongation".
On its face the patent discloses much if not all pertinent knowledge on the subject then possessed by the art, for example. that casting an aluminum alloy (not limited to silicon only) in a chill mold produces a finer grain and somewhat better physical properties.
The patentees allege the discovery that in the case of the aluminum-silicon alloys, rapid cooling in a chill mold instead of producing merely a general reduction in grain size with advantages to be expected from such moderate alteration in structure, "actually produces a fundamental change in the character of the structure of the alloy, and not a simple reduction in the size of all the grains, as would be expected from the prior art".
It is also stated that rapid chilling produces smaller particles and changes the relative sizes, arrangements, and quantities of the silicon and aluminum constituents. It recognizes "permanent" iron or steel molds in ordinary use for casting as proper for the practice of the invention. The different effects produced by slow and rapid cooling upon the granular structure and properties of the alloy are stated, with assertion that with rapid freezing "the granular structure of the alloy is greatly refined, with concomitant increase in the desirable properties of tensile strength and ductility." This in contrast with the result of slow cooling wherein "under the microscope silicon appears in the form of relatively large plates or needles which weaken the metal and induce brittleness".
The specification, p. 2, line 5, contains this important provision: "The amount of silicon in the alloy can be varied considerably but should not, in general, be less than about 3 nor more than about 15 per cent."
The alleged difference between the results of the two methods of chilling are further elaborated on p. 2, where it is said, line 101: "This difference in structure gives us in the case of a chill casting a material composed of ductile dendrites and a very fine grained matrix with none of the particles of hard and brittle silicon which characterize high silicon alloys that have solidified slowly."
It is stated that the invention is not limited to alloys composed of aluminum and silicon alone, copper and zinc being as examples of other alloying metals.
Claims 1, 2, 6 and 7 are directed to an alloy of aluminum and silicon only, while claims 3, 4 and 13 call for an alloy containing other metals as well.
Bamberg patent 1,296,589, item 63, Ex. 47, filed March 7, 1916, for casting and process of making same. This is a very important item of the prior art. The master held, and I think properly, that it does not constitute an anticipation of the patent in suit, for the sole reason that the proportions of the constituents of Bamberg's aluminum-silicon alloy are not shown. The fact that rapid cooling produces a relatively fine grained structure, as stated by Archer and Edwards, was understood and is disclosed by Bamberg. In my opinion Bamberg does disclose a method of chill casting aluminum-silicon alloy.
On p. 1, line 69, he states: "Another object of this invention is to provide a process of making piston castings for internal combustion motor pistons from an alloy of aluminum and magnesium, or either of them, containing copper, tin, zinc, iron, manganese, titanium, nickel and silicon, or either or any of them, and in which the resulting structure will be relatively fine-grained", etc.
And on p. 3, line 65 et seq: "Satisfactory alloys for the production of such piston castings may contain about eighty per cent. or upward of aluminum, the remaining constituents being iron, titanium, silicon, magnesium, zinc, tin, copper, nickel and manganese, or any of them, as may be found advantageous for the particular piston job having the desired internal structural characteristics. Aluminum alloys which have been found especially suitable for piston castings contain, in addition to the aluminum, copper up to 11.0%, magnesium up to 1.0% and iron up to 1.5%. These alloys may contain a small amount of silicon which appears as an unavoidable impurity in commercial ingot aluminum."
I agree with counsel for defendant that "These alloys" in the last sentence mean the alloys mentioned in the preceding sentence only, which are not aluminum-silicon alloys. The other statements, to my mind, clearly teach the use of aluminum-silicon alloy in the process of making piston castings.
Haynes Stellite Co. v. Chesterfield, 6 Cir., 22 F.2d 635, is cited by plaintiff because of its statement on page 637: "Such hit or miss reference does not disclose the combination of cobalt and two of the chromium group." The case is distinguishable. If it is hit or miss to say that an alloy may contain aluminum, the remaining constituents being nine other ingredients or any of them, how much less hit or miss is it to say aluminum with silicon, zinc, or copper, or any of them, as may be found advantageous, etc.; or aluminum with silicon or copper, or either of them, as may be found advantageous? The invention in the Haynes Case, supra, "consisted in uniting cobalt with any two of the metals of the chromium group." Page 636. In the British patent for hardening steel, other metals could be employed without the use of chromium.
Judge Denison says of the latter patent: "He used pig iron as a base, into which he says he introduced two or more of any of a long list of metals — nickel, cobalt, chromium, tungsten, molybdenum, uranium, etc. This seems to be one of those general descriptions, which are intended to be all-embracing and which produce nothing concrete. Haynes' idea is not present."
Two or more of an indefinite long list of ingredients from this British patent is a very different thing from Bamberg's list. And not only in the portions quoted but elsewhere in Bamberg's patent many of the Archer and Edwards ideas are found.
Consequently I do not agree with the master that Bamberg teaches nothing of the effect of chilling aluminum-silicon alloys. True, his patent deals almost exclusively with aluminum-copper alloy, but also recommends aluminum-silicon for piston castings. It also teaches much of the method of making them in metal molds and, if the Archer and Edwards proportions were shown, would constitute a full disclosure of their process. However, the proportions being absent, there is no anticipation.
Richards on Aluminum — Items 43-44-45 — Exhibit 48. This work was cited against the application, but after some explanation by counsel, the patent was allowed over it. It discloses an aluminum-silicon alloy containing over 3% silicon by analysis of Pittsburgh Reduction Company, Item 43, p. 56; that 5% silicon does not prevent casting, p. 518, and the effect of chill casting apparently contrary to the theory of the patentees, p. 519, to which counsel's explanation to the Patent Office was directed. Aside from the disclosure that some experiments had been performed with an alloy in the proportions of the patent, it contains nothing detrimental to the plaintiff's position. I agree with the master that the general inference may be drawn from this publication that the specific discovery of Archer and Edwards was new with them.
Schirmeister article — Item 49, Exhibit 47. This is a lengthy thesis in German submitted for a degree as "A study of binary aluminum alloys in 1915" and states the author's conclusions from a series of laboratory experiments a few years earlier with twenty aluminum alloys; as to their ability to be rolled, strength when rolled, alloying characteristics, shrinkage, change in structure, and the like.
The master treats the article fully and much of the expert testimony is directed to it.
This foreign publication does not anticipate, for, as the master says, it contains no clear teaching as to time limit or that cooling must be rapid. Further, Schirmeister determined the properties of the alloy after the cast ingots had been rolled; and rolling, according to the testimony, changes and breaks down the structure of an aluminum casting. Rec. 230, 473; Wisconsin Alumni Research Foundation v. George A. Breon Co., 8 Cir., 85 F.2d 166; Loew Filter Co. et al. v. German-American Filter Co., 6 Cir., 164 F. 855, 860.
There is a dispute between the experts as to whether Schirmeister's experimental slabs were cast in open or closed top iron molds, consequently whether the molds were of the type required to produce the "rapid" cooling or solidification of the claims in suit. The article does not definitely disclose this feature of the mold. The conflict so impairs the evidence on the point of actual cooling time employed by Schirmeister as to leave it practically worthless on that point. Permutit Co. v. Paige Jones Chemical Co., D.C., 292 F. 239, 241; Cimiotti Unhairing Co. v. Comstock Unhairing Co., C.C., 115 F. 524.
Schirmeister's efforts were serious experiments which measurably added to the knowledge of casting aluminum-silicon alloy and must be taken into account on the question of invention over the prior art. Even though primarily intended to develop characteristics of the alloy when rolled after being cast in metal molds of uncertain construction, yet his statements — "* * * aluminum is uniformly and favorably affected by the addition of silicon * * * The alloys with 5 to 7 per cent silicon will be best suited for rolling stock and those with 10 to 12 silicon for casting purposes" could not be ignored by the patentees.
I further note that the portion of the article containing Table 10, the analysis of aluminum-silicon alloy, within the range of the patent in suit, was cited against certain amendments proposed by Pacz to his patent 1,572,502. See File Wrapper Pacz patent, Ex. 545, p. 19. These amendments related to aluminum-silicon alloy within some of these claims, and the rejection occurred July 15, 1922. Why it was not also cited against the patent in suit is not understood; perhaps due to oversight. However this is academic, as it was not sufficient to invalidate either set of claims.
Pacz patents 1,572,502, Ex. 529, and 1,387,900 — Item 66, Ex. 47. Applications for both these patents were filed before that of the patent in suit. Plaintiff contends neither can be considered as in the prior art unless found to fully describe the invention of the Archer and Edwards patent, as required by R.S. § 4888, 35 U.S.C.A. § 33, citing Alexander Milburn Co. v. Davis-Bournonville Co., 270 U.S. 390, 46 S.Ct. 324, 70 L.Ed. 651. The master thought the disclosures, though incomplete, were in the prior art, and relied on Denaro v. Maryland Baking Co., D.C., 40 F.2d 513. Hazeltine Corp. v. Electric Serv. Eng. Corp., D.C., 18 F.2d 662, 664, seems to the contrary. With some doubt, I concur in this view of the master.
Plaintiff makes a great deal of the fact that after application for the patent in suit was filed, Pacz attempted to amend his applications, and properly insists the patents must be held to read as the applications stood when Archer and Edwards filed. But, as all amendments in 1,572,502 were rejected (see Ex. 545) and do not appear in the patent, it is not affected.
When No. 1,387,900 was filed, it contained this language, p. 2: "My alloy is also excellent for die castings, it can be forged, and drawn into wire, and possesses considerable ductility when cold." This patent related to Pacz' chemically modified alloy of aluminum and silicon. On May 26, 1921, after the application for the patent in suit was filed, Pacz amended, and the amendment being carried into the patent, appears, p. 2, line 57: "My alloy is also excellent for die castings, and after the alkali-fluorin treatment herein described, it can be forged and drawn into wire and possesses considerable ductility when cold". This must be held to refer as originally, to Pacz' chemically modified alloy; not to his alloy before modification. He shows, p. 2, line 13, that he modifies before casting.
No. 1,387,900 shows and claims both process and product. As it requires chemical treatment, it does not anticipate the process claims in suit. If the product was the same in all essentials as that of claims 6 and 7, those claims would be anticipated. See Lamb Knit Goods Co. v. Lamb Glove Mitten Co., supra, at page 269. Though having many of the characteristics of the product of chill casting, it is not proven to be the same.
Pacz patent 1,572,502 is for a product only, and covers alloys containing copper as well as silicon. When it was attempted to add claims for aluminum-silicon, they were rejected as involving new matter in that they did not contain copper. Ex. 545, p. 19. Being for a product only, it cannot anticipate the process claims of the patent in suit, and as Pacz' alloy includes copper, product claims 6 and 7 drawn to aluminum-silicon alloy, are not anticipated. Claim 13 seems to be anticipated, for the patent, p. 1, lines 73-77 says: "As my alloy has a low melting point and is very fluid when melted, it is a good material for use in die casting and in general in the manufacture of articles in permanent molds."
And, line 106: "I have obtained exceptionally good results from an alloy of 67-½ parts aluminum, 25 parts copper and 7-½ parts silicon".
Claim 13 is drawn to a chill casting of aluminum alloy containing silicon and copper characterized by minute subdivision of silicon. Plaintiff's expert testified, R. 832, that chilling aluminum-silicon alloy 3 to 15% silicon as practiced with permanent molds will produce the micro-structure of the patent. And such effect is asserted in claim 13 as to aluminum-silicon-copper alloy. Consequently, as the master found, Pacz 1,572,502 which did not relate to chemical modification, but to casting of such alloy in permanent molds, anticipates claim 13 in suit. I cannot agree, however, with the master's view that this patent anticipates claims 3 and 4.
Clerget piston — Item I — Exhibit 48. The Aeronautical Journal of November 1919 contained an article which is in evidence, on aluminum alloys for airplane engines; and set out the analysis of a binary piston alloy made at Birmingham, and of the so-called Clerget piston, which appears to be a French product. Both include an inconsequential, .31% silicon, and the latter also sets out a separate, additional item, "silicon 5.39%". There is a dispute between the experts, one saying that the Clerget piston evidently contained two kinds of silicon, such as is shown in many of the exhibits to have been at one time taken into account, while the other expert says the double mention of silicon is a blunder. On authority of Badische Anilin Soda Fabrik v. Kalle Co., 2 Cir., 104 F. 802, I think the master was correct in accepting the article as written, and as showing an aluminum piston containing, along with 11.63% copper, 5.7% silicon. No weight is given to the opinion of plaintiff's witness that in France, according to his best information, these pistons were supposed to contain no silicon, this being hearsay.
Transactions of American Institute of Mining Engineers, 1889-1890 — Exhibit 541. What is found here is to some extent taken from Richards' book. While it shows employment of silicon up to 6 to 8 per cent, in connection with aluminum as being possible in some special cases of particular surface treatment for protection from oxidation, and also contains a table at p. 535 showing silicon above 3% combined with aluminum, yet the statement appears at p. 551: "In all ordinary work to which aluminum is put, the more nearly free the metal from silicon the better". And scattered through the exhibit will be found various references to silicon as an impurity.
In all this prior art the only anticipation found is of claim 13, and, as I see it, nothing else of consequence which adds to the disclosures and acknowledgments made by Archer and Edwards on the face of their application.
I agree with the master that in this limited field, where investigation and experiment by most competent technicians was intense, the patentees discovered, what others had overlooked, that rapid chill casting of an aluminum-silicon alloy produced by a simpler method than was then understood, the desirable features of physical structure and properties set forth in the patent, and that having been granted a patent clothed with the presumption of validity, on which has been built what is well termed an aluminum-silicon industry of substantial magnitude, it should be held that their advance in the art was inventive and not the result of the skill of the art. The court is not impressed with defendant's argument often repeated, that all the patentees did was to pour an old alloy into a well known mold and by a perfectly well known process of casting secure a result that was inevitable. Certainly many extremely valuable patents involving obscure chemical reactions and what to the courts have appeared genuine discoveries, would be stricken down on this theory.
There was an invention, but the claims for its protection, except No. 7, are not in proper form to protect it.
Infringement of patent 1,572,459. Product claim No. 7, the only one found valid, is limited to a chill casting of aluminum-silicon alloy, especially as other product claims call for additional ingredients. Covering a slight advance in the art, the patent should be strictly construed. Defendant's alleged infringing pistons contain copper, magnesium, manganese, iron and zinc, as well as aluminum and silicon. These additions substantially reduce the desirable qualities of an aluminum-silicon alloy. While defendant's piston alloys, if rapidly chill cast, will realize the benefits of the invention in a limited and possibly substantial way, I do not think they infringe claim 7.
As found above, claims 1, 2 and 6 are believed invalid for undue breadth, being drawn to a process or product in which aluminum-silicon alloy with a limited silicon constituent is essential and to which composition they must be restricted, if valid. The same may be said of claims 3, 4, and 13. If valid, they will be restricted to their terms, and are not shown to have been infringed by any of the defendant's pistons described in Section 1, Exhibit 4, or in the making thereof. All such pistons contain not only aluminum and silicon together with a soluble metal or copper, called for in the latter three claims, but also the other alloying ingredients mentioned above.
Plaintiff virtually concedes that at most the proof shows the extreme lower part of the skirts of the defendant's pistons and parts of some of the heads possess a micro-structure more or less similar to that of plaintiff's test bar, R-5065, Exhibit 4, when the latter was chill cast at some uncertain rate of cooling time. It is noted that none of plaintiff's many pistons or other products actually cast for sale and use was tested or analyzed, and the actual cooling time employed by plaintiff in practice is uncertain.
Plaintiff's argument based on the photo-micrographs is generally understandable. But it is not entirely clear to the court that Exhibit 40, chill casting of R-5067, said to be a synthetic composite of defendant's pistons, "bears Substantial resemblance to" Exhibit 16, chill casting of plaintiff's test bar, or that Exhibit 30, skirt of defendant's piston R-2, claimed by plaintiff to be chill cast, really resembles Exhibit 16 at all, though, if it does, it undoubtedly bears more resemblance to it than to Exhibit 15, plaintiff's test bar when slowly cooled in sand mold.
Plaintiff's expert, Dr. Jeffries, said 75 seconds used in casting an aluminum-silicon piston would produce an unsatisfactory result, but that with defendant's alloy in which copper, manganese, etc., as well as silicon, is found, slow cooling produces quite good physical properties, though it would not be within the teaching of the patent in suit. Rec. 361-2.
After such consideration as the court is able to give this question, I cannot say that the master's views as to infringement are incorrect, though I would not attach the importance he does to Dr. Jeffries' inability to distinguish those portions of the defendant's piston R-2, wherein evidence of the process of the patent appears, from those parts in which it does not appear. I think the master was right in his view that infringement of the claims in suit was not satisfactorily established.
Patent 1,799,837 to Archer and Kempf, issued April 7, 1931 for an aluminum base alloy and piston made therefrom. Claim 1 only is involved. It reads: "An aluminum base alloy containing from about 7 to 15% silicon, from about 0.2 to 3% of magnesium, from about 0.5 to 7% of nickel and from about 0.3 to 4% of copper."
It is admitted that this patent is not anticipated, but the master sustained defendant's contention that it involved no inventive advance. In his report the master cites two patents and two publications. None of these contain disclosures at all sufficient to maintain the claim of non-invention, for example, Bamberg's patent omitted any mention of percentages of ingredients, and also said nothing about low thermal expansion, which appears to be the main object of the patent in suit; Wilm's patent mentions magnesium as desirable, but again low thermal expansion is not an object of that patent, which is for a process for treating aluminum-magnesium alloy by high temperature followed by cooling and storage prior to use. And the publications are even further away than these patents. Pacz 1,387,900 sought low heat expansion, but by chemical treatment. Defendant's counsel files an argument to bring in points not mentioned by the master, but it is not convincing. Principally he presents a table of percentages of Archer and Kempf's ingredients and those of certain prior art uses and references and of defendant's own pistons. By substitutions and transpositions made in the name of equivalency it is sought to prove that this prior art so resembles the patent that invention is absent. The evidence discloses that in the case of complex alloys it is impossible to say in advance just what the effect of changing ingredients or their proportions may be. For example, to replace 8% of the copper content of the Clerget piston by 1% of magnesium because, generally speaking, magnesium may be said to possess eight times the hardening quality of copper, will evidently produce a result that cannot be foretold with certainty. So of other transpositions suggested. The tabulation contains many errors which necessarily affect its result; to what extent the court cannot say nor ascertain. Counsel supports this method of arranging the prior art to his satisfaction by the testimony of defendant's expert, who was able to produce with blackboard and chalk in a few moments with little if any calculation two tables, Exhibit 533, which, to my mind, merely shows the ease with which "America could have been discovered after 1492". Contentions based on such hindsight have often been answered and do not require discussion.
If defendant is allowed to treat nickel 40% as being within the patent in order to justify re-arranging silicon so that Clerget's piston or other of the prior art will more nearly resemble the patent, it will be compelled to admit that some of its pistons contain an infringing amount of nickel. This proposed re-arrangement of silicon based on plaintiff's contention that 40% nickel is within the patent, appears entirely arbitrary. If because of plaintiff's contention respecting nickel, silicon is to be reduced 20%, why not all other ingredients as well?
I think a finding in favor of validity of this patent is required and that plaintiff's authorities, which are along the usual lines, amply support its position.
The master held that even if claim I is valid, it is not infringed. That defendant's pistons contain manganese and iron, elements not called for by the claim, the former having alloying properties, and the latter present as an impurity; and that nickel is found in defendant's pistons in such small quantities as to indicate that its presence was unavoidable. Claim 1 calls for nickel from about 0.5 to 7%, and the evidence, Ex. 4, shows nickel in defendant's pistons from .03% to .49%. The master thought infringement should only be found here if defendant's alloys appear to be precisely those of the claim, without stating any reason for that view. It may be because he believed, as in the case of Archer and Edwards, that if any advance be shown, it is very slight. It was also his view that effects of the constituents of claim 1 were not certainly known and can only be ascertained by experiment, even though the materials employed by patentees and, generally speaking, their properties and effects in alloys of this kind, were well known; that experiments to secure the objects desired in this case were indicated as necessary, but that their results, if successful, should not under the circumstances here present be entitled to protection.
The patent law has always recognized the propriety and necessity of investigation and experiment to reach a definite end; and a patent is not to be condemned merely because it is the result of that sort of research. A patent is not granted only for results obtained by chance. The fact that the experimenters are highly skilled and probably well paid technicians, operating with expensive laboratory equipment, should not affect the result. In this case it appears that notwithstanding these advantages, their investigation covered a period of several years during which the art appreciated the desirability of an aluminum-silicon alloy with low thermal expansion, etc., but did not fill the need.
On the question of infringement the plaintiff has decidedly the better of the argument. The patent states the preferential amounts of ingredients, to wit, of silicon between about 10 and 15%, magnesium 0.2 to 1%, nickel 2 to 5%, copper 0.5 to 2%. As to nickel the patent further states: "For some purposes nickel may be advantageously added in amounts as high as 7% and as low as 0.5%". And as to copper — "Where machining is not the deciding factor * * * copper may be added in amounts as high as about 7%". It then states the patentees' preference for an aluminum base alloy specially adaptable for pistons containing 14% silicon, 2% nickel, 1% magnesium, and 0.75% copper; and that other constituents such as iron, usually found as an impurity in silicon alloys, may be present without appreciably affecting the essential properties of the alloys. Plaintiff's commercial 132 alloy contains 12 to 14% silicon, about 1% magnesium, about 2% nickel, and 0.75% copper. As to all ingredients save nickel, defendant's pistons are near the middle of the preferred range of compositions stated. In 23 of the first 24 of its pistons on Exhibit 4, nickel runs from 0.41 to 0.49. The claim in suit calls for nickel "from about 0.5 to 7%". This should include any percentage over 0.40, especially in view of the admission of defendant's expert Tour that adding or deducting 1% nickel would make "no particular difference". Rec. 718.
Defendant's alloy appears to secure all the advantages of the patent. There is ample testimony, supported by Plaintiff's Exhibit 11, that the coefficient of thermal expansion of defendant's R-2 composition is the same as that which the patent gives for the 132 alloy. Rec. 506-507. Under the foregoing facts defendant does not escape infringement because its nickel is slightly below 0.5%. Crozier-Straub v. Graham, 3 Cir., 28 F.2d 321, 324-326, certiorari denied 279 U.S. 840, 49 S.Ct. 253, 73 L.Ed. 987; Lampus v. Crozier-Straub, 3 Cir., 41 F.2d 746.
The more serious question arises from the presence in defendant's pistons of small quantities of manganese, iron, and zinc. These are not shown to have impaired its pistons, which retain all the advantages secured by what is otherwise an infringement, and what I believe should be regarded as still an infringement notwithstanding the presence of these other elements. Haynes Stellite Co. v. Chesterfield, 6 Cir., 22 F.2d 635; General Elec. Co. v. Hoskins Mfg. Co., 7 Cir., 224 F. 464, 472, 473. If the presence of nickel was accidental, that fact constitutes no defense. Thompson et al. v. Bushnell Co., 2 Cir., 96 F. 238, 243; Kawneer Mfg. Co. v. Toledo Plate Window Glass Co., D.C., 232 F. 362, 367. Rowley Co. v. Rowley, 39 F.2d 865; Fowler v. Detroit Bedding Co., 47 F.2d 752; and Firestone Tire Rubber Co. v. United States Rubber Co., 79 F.2d 948, all by the Sixth Circuit Court of Appeals, are cited to the contrary view. The facts in the first two cases distinguish them from this, and I am not able to find the point discussed in the last case. If it is somewhere in the ten pages of the opinion, counsel should have pointed out its location. Nor does the evidence satisfy me that nickel was present by accident only.
Plaintiff's exceptions 1 to 11, 15, 17, 18, 20 to 25, 26 (as to 1st par.), 27, 28 (as to 1st clause), 30 to 34, 36, 37, 39, 40 to 42, 43 (as to 1st ground alleged invalidity), 44, 45, 51 (as to Bamberg), 56 and 57 (as to pat. 1,572,459) are overruled.
Its exceptions 12 to 14, 16, 19, 26 (as to pars. 2 and 3), 28 (as to 2nd clause), 29, 35, 38, 43 (as to 2nd and 3rd ground alleged invalidity), 46 to 50, 51 (as to 1st and 3rd sentences), 52 to 55 and 56 and 57 (as to pat. 1,799,837) are sustained.
Defendant's exceptions are overruled.
The court holds claims 1 of Archer and Kempf patent 1,799,837 valid and infringed, and in respect to it, the usual decree against the defendant for an injunction and an accounting will be entered and the matter again referred to special master Friebolin to take the account.
As to Archer and Edwards patent 1,572,459 the bill will be dismissed in conformity with the master's recommendation.
Pacz patent 1,595,218, claims 1 and 2, is involved in this litigation. In respect of this patent the master reported in favor of the defendant, and plaintiff has filed no exceptions. Consequently the report is confirmed and the bill dismissed as to such patent.
Plaintiff will be ordered to pay two-thirds and defendant one-third of the costs up to this time.