Metallizing Engineering Co. v. Kenyon Bearing & Auto Parts Co.

HINCKS, District Judge. This is a suit for infringement of reissue patent No. 22,397, issued to the plaintiff on November 30, 1943. The original patent, No. 2,320,327, had issued on May 25, 1943 upon an application filed August 6, 1942 by John F. Meduna, the plaintiff’s assignor, and no claim is pressed of infirmity in the reissue not attaching to the original patent. It was conceded that the defense has been assumed by the defendants’ manufacturer, the Metallizing Gom-pany of America, a copartnership, which sold the apparatus by which the alleged infringement of the patented process was accomplished. The defenses are lack of patentable invention, noninfringement, public prior uses by Meduna, the inventor, and by others, abandonment, and ambiguity in the specifications. Findings 1. The claimed invention is a method for procuring a satisfactory bond between a metallic surface (such as a worn machine part which it is desired to build up) and metal which is sprayed thereon in molten form. Although all eleven claims of the patent are in issue, neither on trial nor in brief have the parties made any substantial distinction between them. The gist of the invention of each claim is as follows: (1) the preparation of the metallic base by stroking its surface with a metal electrode in a low-voltage circuit, thereby fusing to the base metallic elec-trodermaterial which forms a surface of irregular roughness having a multitude of projections with overhanging edges and minute craters with overhanging edges and (2) thereafter spraying molten metal upon the base part as thus conditioned. 2. Metallizing is an art of comparatively recent origin. To be sure, it had long been known that a metal spray if projected far enough would cause the metal droplets to chill in the air forming a metallic powder. But apparently it was not until about the close of the nineteenth century that it was discovered that metal droplets from a spray, if impinging upon an object while somewhat plastic, would coalesce into the form of a metal mass comparatively porous. By 1910 it was known that a metal spray could be used for coating surfaces with a protective or decorative metal coating. And probably the greatest contribution to the development of the art came in 1913 when Morf evolved an automatic spray-gun competent to melt wire of metals having a high melting point. Following this invention the practice of the art made great strides in the industrial field, both in this country and abroad. It long had been known that to obtain a satisfactory bond between the applied coat and the base, the base must be clean, and sand blasting had been used to accomplish this objective. By the middle of the 1920 decade it had been discovered that to ob — 44 — tain a satisfactory bond, not only must the base be clean, it must be sufficiently rough as well. 3. This requirement caused no particular difficulty as long as the process was confined to parts of comparatively soft metal. But by the end of the decade the art had become particularly concerned with the task of rebuilding worn machine parts. For those that were of metal comparatively soft it was feasible to roughen, as well as to clean, the surface by sand-blasting. But sand-blasting was a cumbersome and expensive procedure that required the removal of the work from the lathe and its preparation in a special chamber. Consequently, there came into vogue a practice of roughening the metal part by screw threading, — a mechanical process whereby threads were cut into the surface by a tool so adjusted as to tear into minute fissures the metal side walls of the thread. But screw threading was a process difficult, if not impossible, to apply to interior surfaces and recesses and more particularly it was inapplicable to the surfaces of metals of extreme hardness. Nor was it practicable to soften the metal by heat, then roughen the surface and harden the metal again, for such treatment resulted in warping strains. As a result, the art of metallizing was incapable of rebuilding machine parts of hardened metal for a period of many years after the rebuilding of softer parts had become common practice. Such was the state of the art at the time of Meduna’s discovery. 4. To be sure, at this time it was well known in the art of welding that a metallic electrode in a low-voltage circuit when brought into contact with a piece of metal introduced into the circuit would leave a deposit of electrode-material fused to the metal part. The McQuay-Norris machine was adapted to accomplish this very result. It is true that the deposit made by this machine was a porous, coral-like, metallic substance similar to that later applied by Meduna to roughen the metal base. But the porous character of the deposited metal was wholly adventitious and indeed undesired, for the McQuay-Norris machine was used principally to fill scores and cracks in cylinders, etc. To accomplish this result, it was necessary to lay on the porous metal deposit from the electrode layer by layer and subsequently by peening, grinding and chiseling to compact each layer, thus destroying rather than utilizing its porosity. 5. No prior patents have been shown which disclose a bonding method for sprayed metal obtained by a porous electrode-material electrically fused to a metal base by resistance heating. The Brewster patent, U.S. 1,327,267 (1920) covered a welding process “particularly adapted and intended for use in the filling of cavities which may occur in castings or other metal objects.” Its process was essentially that accomplished by the McQuay-Norris machine described in Par. 4 above. The Williams patent, No. 1,412,326 (1922) disclosed a method of treating a metal base, such as a tire rim, to improve its bond with rubber to be vulcanized thereon. The method consisted in fusing to the metal base a porous electrode-material by arc-welding, thus roughening the metallic surface. Nagin, No. 1,657,466 (1928) discloses a process for restoring the anti-slip properties of metallic floor surfaces by allowing drops of metal melted by an acetylene flame to fall and fuse upon a metal floor or stair, thus roughening the treaded surface. Schuman, No. 1,452,936 (1923) addressed himself to the problem of leaving a deposit of nickel electrode-material in the recess of iron machine parts without hardening the surrounding area of the base and thus destroying its grinding and machining properties. It contains no intimation that a porous electrode-material was obtained or even desired; on the contrary his objective was such as to imply a preference for a compact deposit. And both his objective and the method used for its attainment was wholly unrelated to the metal-lizing art. Andrus (1934) was concerned with the problem of a protective coating for underground pipe which should not “creep” or “strip” from the pipe as it expanded and contracted. His solution was to sprinkle the surface of the pipe with non-contiguous metallic pellets or spheres welded thereon. He indicates a preference for pellets 1/16 to 1/8 inch in diameter welded on to the base pipe at 1/2 inch intervals. He contemplates a coating of cement to be applied cold or a “bituminous coating” to be applied “hot”. Whether the bituminous coating was to be applied by casting, spray or otherwise is not indicated. Indeed, he says, “the manner of application of the protective coating is not a part of the present invention.” And certainly he suggests no solution, and least of all Meduna’s solu — 45 — tion, of the problem of a bond between sprayed steel and hardened steel (or between other metals) which shall withstand mechanical strains and pressures in machine parts. Robinson, No. 1,289,000 (1918) relates solely to the engraving art and discloses an electric “marking” pencil in an apparatus to prevent the roughening of the engraved surface by the pitting caused by arcing. For aught disclosed, Robinson was unconscious of any deposit of electrode-material from his pencil (electrode): his attention was confined to the “cutting” of the surface by his pencil. He does not disclose the deposit of any electrode-material, — certainly none of porous structure suitable and useful as bond for sprayed metal. Overlin, No. 1,350,734 (1920) is the converse of Robinson in that he seeks a method to cultivate and regulate pitting from arcing through the use of an “electric pen” useful to mark (for purposes of identification) metal tools and parts. But Overlin like Robinson is completely indifferent to the structural characteristics of the deposit (if any) left by his pen (electrode). The pitting which he seeks to accomplish is an end in itself; not a step or means in a process. Whether such pitting would be suitable for metallic bonding is not suggested or disclosed. The disclosure is outside the field of the metallizing art. Prior Publications 6. The defendant introduced an article from a 1933 issue of Commercial Car Journal, as Exh. 00. This refers to a method for sealing cracks in cylinder water jackets by use of an electric arc. It states: “In some instances, the weld is porous and these welds have been sealed by coating them with metal from the gun.” But the entire article, read in the light of the contemporary art, taught that the weld deposit should be sand-blasted before spraying, as defendants’ expert conceded. The defendant introduced excerpts from the “Automobile Engineer” for May, 1941. Exh. PP. This included text under the heading “Heavy Duty Bearings” and other text under the heading “Cladding by Spray”. Both texts related to the rebuilding of bearings. The first described a removable bearing comprised of a steel casing externally dimensioned for a pressed fit into its seat and interiorly^ surfaced with a rough weld deposit of bronze material produced by the application of flame (blow-pipe). The roughness of the bronze weld-material, it was suggested, “helps to anchor the babbitt” which was later to be cast (not sprayed) into the bearing. The article gives no support to defendants’ contention that in this process sprayed pabbitt referred to in the second text (“Cladding by Spray”) might be substituted for the cast babbitt expressly specified in the first text. Indeed, the structure described was obviously such as to make the application of sprayed babbitt impractical if not impossible. Moreover, there is utterly nothing in the text to suggest that this disclosure was addressed to the problem of bond between the babbitt and the body of the bearing. Instead, it specifies that the advantage of the bronze layer of weld material interposed between the steel casing of the bearing and the cast babbitt was that it “provides a soft cushion * * * in case the babbitt wears entirely through as it often does.” And that, obviously, explains the specification that the bronze weld layer should be as thick as 3/16 inch; surely a deposit of such thickness would not be required if the bronze layer were interposed only as a means for bonding with the babbitt. 7. Phelps-Dodge Shop Use. The testimony is that this company in Washington, D. C., long prior to 1940 was engaged in the business of rebuilding motors. In the course of that business it developed a practice for rebabbitting burned-out connecting rods. To make the babbitt stick to the steel rod it was necessary, as was well known in the art, first to coat the steel with tin by immersion in molten tin, and this was its general practice. But for rods of especially hard steel the difficulty of making the tin stick to the steel it overcame by first depositing bronze electrode-material by an arcing process on the steel. The rod was then washed with acid and dipped in liquid tin, and thus coated was wire-brushed and placed in a mold into which molten babbitt was cast. No documentary evidence or contemporaneous specimen of the practice was produced. Not until June, 1942 did this company do any metal spraying. In its metallizing, it prepared the parts by sand-blasting,— not by the arc-welding process which the witnesses say was used to prepare parts for tin-plating. And there was no evidence to show that prior to August, 1941 the Phelps-Dodge employees ascertained or — 46 — knew that an electrical means for roughening rods, even if useful in improving a bond with liquid tin, was also an effective means to obtain a satisfactory bond with metals, and especially metals having a high melting temperature, sprayed thereon in minute droplets, each rapidly cooling as it travelled to its target. At most, the testimony shows that an indeterminate number of connecting rods, none of which was offered in evidence, after treatment as above described, left the shop into the great unknown and were never heard of again. The practice, if it once existed, left no impression on any art relevant to the Meduna process either with the general mechanical public or even with the very employees who used the practice. Thus I find this alleged shop practice not sufficiently proved to constitute a part of the relevant prior art. Likewise, as to the alleged shop practice of Heskett Machine and Engine Company. Alleged Prior Use by the Inventor 8. Early in the spring of 1940, the inventor purchased a small machine shop with its business which included some custom in the metallizing of machine parts. At the time, the inventor was a mechanic skilled and experienced in the art of welding and machine-work but without experience in the more recent art of metal-lizing. In the prosecution of his newly acquired business he took a small order to resurface some tractor pins of hardened steel. He then learned for the first time from a mechanic whom he had acquired with the business the difficulties inherent in procuring a satisfactory bond between sprayed metal and a hardened base. However, in this quandary he recalled the characteristic ■ roughened surface of deposited electrode-material produced by the McQuay-Norris transformer which he had previously used and conceived the notion that a roughened 'surface thus fused upon the base might constitute a suitable bond for the sprayed metal. After a few experiments, personally conducted, along this line, he obtained a result which for the moment seemed satisfactory. This was in March 1940 (with reference to this date it will be borne in mind that the application for his original patent was filed August 6, 1942). Elated by the apparent results of his experiment, he sought .the advice of a friendly engineer who told him that his process would indeed be valuable if it served to produce a satisfactory bond but advised him to test it out thoroughly in actual service before attempting to patent it. 9. Agreeable to this advice, the invent- or, personally and through a “free lance” salesman, solicited metallizing jobs for hardened metal as well as soft. In the period from March, 1940, to August, 1941, in the commercial conduct of his business sixty-six orders involving hardened metal were received all at prices which aggregated $1,100, — a figure which was about 10% of his gross business for that period. Some of these jobs were done for firms which knew that the inventor was using some new process; of these some were done under the inventor’s agreement to make no charge if the job was not satisfactory. But most of these jobs were done by sub-contract through repair stations for owners who were without knowledge as to the process to be used, and whose identity was never known to the inventor and whose identity he never sought to ascertain. As late as December, 1941, a job was done for a customer who understood that the job was experimental. This particular job proved unsatisfactory and the inventor did it over at his own expense leaving the incidental expense of installation, etc., to be absorbed by the customer. Whether the second performance on this job proved satisfactory does not appear; for aught that appears no one ever took steps to ascertain. For at least one of these jobs, partly to get the work and partly because of uncertainty as to whether a job done by the new process would prove satisfactory, the inventor made a charge less than that which otherwise would have been appropriate. As early as April, 1940, the inventor caused a patent search to be made with a view to determining whether his invention was patentable. In the fall of 1940, one Jensen, an executive of the defendants’ manufaoturer, dickered with the inventor and offered him a contract whereby the inventor was to apply for a patent and give an exclusive license thereunder. This contract was declined by the inventor, not because he was averse to patenting his process, but because he was not satisfied with the terms and because of unwillingness to deal with Jensen personally, rather than his corporate principal. On April 23, 1942, the inventor agreed to sell to the plaintiff — 47 — the patent rights in his invention which was described as a “secret process”, and the plaintiff agreed to investigate the pat-entability of the process, which theretofore had not been disclosed to it, and if it should find the invention patentable to make application for the patent thereon and to prosecute the same diligently. On July 22, 1942, the inventor gave the plaintiff a formal assignment of his rights as the inventor which was filed with the application on August 6, 1942. 10. In October, 1940, and again in October, 1941, the inventor was induced to place an advertisement in a trade publication called “Washington Purchasing Agent”, announcing to the trade that he had a new process for metallizing hardened metals. These advertisements contained no intimation whatever that his process was in the experimental stage. They produced no orders for the inventor. And there was no evidence other than the inventor’s testimony that the business thus solicited was sought for experimental purposes. 11. Throughout the entire period prior to August, 1941, there was no evidence of systematic practice to follow up the commercial jobs on which the new process had , been used. Only occasionally, when the/ inventor happened to be in personal contact with a customer, did he inquire if a job previously done was holding up satisfactorily. And as to jobs coming to the inventor through middlemen from owners whose identity was never known to the inventor, there was no follow-up whatever. 12. From the foregoing facts, I find that the inventor’s main purpose in his use of the process prior to August 6, 1941, and especially in respect to all jobs for owners not known to him, was commercial, and that an experimental purpose in .connection with such use was subordinate only. 13. At all times prior to August 6, 1941, the practice of the process was so guarded as not ito come to public knowledge; its nature was disclosed only to a few employees and advisers of the inventor, less than half a dozen in number, in all cases under a promise of confidence which was not abused. Although there was some conflict in the evidence on the point, I find that prior to August, 1941, the nature of the process could not have been deduced from inspection or physical tests upon specimens of the processed product in the hands of the public, although later, to be sure, after the process had come into public knowledge through the publication of the patent application, parts treated by the process were then easily identifiable by those familiar with the art. 14. Since the date of Meduna’s original application, the process disclosed therein has had a wide commercial application. Both the plaintiff and the defendants’ manufacturer have developed electrical bonding machines adapted to facilitate the practice of the process which have been widely distributed through commercial channels. As a result, a great volume of worn machine parts of hardened metal which under the earlier art were junked as being not susceptible of metallizing it is now economically advantageous and mechanically possible to rebuild by metalliz-ing. The process also has a wide application to many parts of metal soft enough for machining and rough-threading, and especially to those of such shape and assembly that machining and sand-blasting is either unfeasible or inconvenient. 15. Notional Welding and Grinding Company (Texas) Prior Use. All of the evidence relating to this prior use comes from three employees of the Texas Company whose task it was to grind, weld, metallize and otherwise work on repair jobs undertaken by the Texas Company, a customer of the defendants’ manufacturer. They testified as follows. That in the early part of 1939 an automotive crank shaft came into the shop for repair. That in order to straighten the shaft heat was applied which created a hard spot on the shaft about as big as a half dollar piece. That thereafter the shaft was ground and rough-threaded preparatory to metallizing but that in the rough-threading process the tool rolled over the hardened spot leaving it less rough than the remainder of the ground surface. That the part was then metallized and ground down to specified size and in the grinding process it was noticed that the sprayed metal had not bonded to the surface of the hardened spot. That the shaft was again reground preparatory to metallizing and again rough-threaded. That thereupon at the suggestion of one of the trio, whose identity no one remembered, the McQuay-Norris machine in the shop with which they were all familiar was used to fuse a deposit of metal electrode-material on the hardened spot. That thereupon the shaft was again metallized and again ground down to spe48 ci'fied size and this time no lack of bond appeared in the grinding process, — a fact demonstrated, however, only by tapping with a hammer. And that the shaft then left the shop, presumably being returned to its owner, whose identity had never been known by any of the witnesses, and who was never heard of again. There was also some testimony that another crank shaft was similarly treated on some subsequent occasion as to the date of which there was no testimony whatever. Of these uses there was no documentary evidence whatever, all records of the transactions, according to the testimony, having been destroyed years before in the usual routine of the 'business. Merely because the job looked satisfactory when it left the shop, it does not follow that the process had been practiced successfully. There was no evidence of any real test in the shop. And for aught that appears when the shaft left the shop it may never have been put into a vehicle. If put into a vehicle, there is no evidence that it was given a fair working test; the vehicle may have been junked as the result of a collision. If the shaft had failed in use, in the absence of evidence that the repair job had been guaranteed, it is reasonable to believe that the owner either junked the shaft or took i,t elsewhere for repair. If on' failure it had been returned to the Texas Company, it does not necessarily follow that it would have been brought to the attention of the witnesses or would surely have been recognized by them. The use of the process was plainly by way of experiment: it was so characterized by one of the witnesses. Apparently no one in the shop had confidence in the success of the experiment for thereafter the general practice was to use rough-threading wherever possible and no effort was made to attract and handle 'business in metallizing hardened parts. Indeed, on January 3, 1941, the National Company wrote to the plaintiff to inquire if its knurling tool was a suitable means for preparing crank-shaft journals for metalliz-ing in lieu of rough-threading. The date of this use as fixed by defendants’ witnesses was some four years before their first .testimony was given. In a letter from one of these mechanic-witnesses to defendants’ manufacturer, dated September 17, 1943, the witness seemed to stress the fact that ,the 1939 date was “only to the best of my recollection”. And there is nothing in the record to suggest that their memory had been aided by any reliable and truly significant landmark. There was, ,to be sure, some testimony that they remembered the date as being prior to a change in the location of the shop which took place in December, 1940. But as to this there was nothing in the situation to explain a mental association between an hour’s work on an unidentified crank shaft and a change in the location of the business. One witness said he was able to fix the date because he remembered it was a few months after his purchase of a certain automobile. But here again there was nothing in .the situation or the background to lend plausibility to this claimed mental association. We have nothing but the word of the witness speaking apparently from unaided memory as to the' date on which he did purchase his automobile. Altogether, while I find nothing in the record to make me doubt that the witnesses have testified as nearly as their actual memory permits, I feel that their testimony, especially as to .the date, perhaps unconsciously was affected by suggestion and does not satisfy the high degree of the burden of proof which rests upon the defendant. And the only evidence of any subsequent practice within the knowledge of these witnesses was wholly undated. There was no evidence whatever .that the alleged use had been practiced successfully. I am unable to find that the defendants by the required burden of proof have established this alleged prior use. 16. Reinhard Brothers Prior Use. Reinhard Brothers is a large machine repair shop in Minneapolis. The defendants’ evidence consisted principally of the oral testimony of two employees, — Latourell before trial and at trial, and Kingsberg before trial. The substance of their testimony was as follows. That on June 24th they had a rush job to do for the Minnesota State Highway Department consisting in the rebuilding by metallizing of a hardened steel shaft from some automotive vehicle. That they first ground the shaft and then prepared it for metallizing by rough-threading. Metallizing was then accomplished but was unsatisfactory since full of minute cracks. That they then removed .the sprayed metal by grinding and Latourell suggested that its surface might be prepared for metallizing by arcing with a welding machine which was at hand. That Latourell and Kingsberg in conjunc — 49 — tion experimented on the shaft for several hours using various types of electrodes until a satisfactory result was indicated by a nickel steel electrode. That Kingsberg then completed the preparation of the shaft with the nickel steel electrode and metallized it. That the shaft then appeared to be satisfactory and shortly left the shop in the usual course of business and was never heard of again. Reinhard Brothers had purchased a metallizing gun from defendants’ manufacturer in December, 1940. They acquired from the same source a bonding machine, the use of which is claimed to infringe Meduna, in the fall of 1942. By letter dated May 4, 1943, Parkinson, a representative of defendants’ manufacturer, wrote Latourell saying: “In order for us to complete our files on our Electric Bond machine, we would appreciate your advising us when you first used an electric arc machine for preparing a shaft to receive sprayed metals or sprayed coatings.” It may be noted that this was a scant month before the complaint in this case was served. On May 27, 1943, Latourell replied by letter, saying: “Our first attempt to employ the arc method of roughing a shaft ,to receive metallizing was in the latter part of 1941. At that time while working on a number of hard surfaced shafts and not having any method for grit blasting, we were forced to try some way to get a satisfactory bond and resorted to the use of our A. C. arc welder.” By letter dated September 22, 1943, Kings-berg wrote Parkinson “With regard to the method of electric or arc bonding surfaces before metallizing, my first experience with .this was in the early part of August, 1941, as near as I can remember.” In the spring of 1944, after defendants’ manufacturer had been exceedingly active in preparing the defense, its representative Parkinson was in conference with Latour-ell and Kingsberg in the Reinhard shop. They then caused the shop invoices covering metallizing work in or about the month of June, 1941, to be assembled and from them selected an invoice to the Minnesota State Highway Commission as being the one covering the job on which their alleged use of an arc-bonding process had first been made. This invoice is in evidence as Exhibit C in Latourell’s deposition of evidence before trial. It shows that the job was one received “to metallize and grind”; that the shaft was worked by two employees, one of whom was Kingsberg; that the shaft was “turned down” on June 23, machined and key way cut by Kingsberg on June 24th involving 2% hours; sprayed on June 24th' (J4 hour) ; keyway and end of shaft welded up by another workman, June 24th, *4 hour. But there is nothing whatever on the invoice to show that arc-welding was used to roughen preparatory to metal-lizing. In this respect the invoice was in no essential respect different from a number of invoices on which concededly arc-welding had not been used. Thus neither the invoices nor any contemporary writings nor even the letters of the witnesses quoted above, gave any support whatever to a use of arc-welding preparatory to metallizing prior to the critical date of August 6, 1941. Latourell testified with respect to this job .that a substantial amount of time was spent in experimenting with various electrodes on the morning of June 24th; Kingsberg testified that the experimenting was in the afternoon. Latourell testified that the job was finished before five o’clock in .the afternoon and that he saw it after it was finished. Kingsberg testified that he himself completed the arc-welding and subsequently sprayed it and ground it on June 24th, completing the job at 7:30 or 8:00 in the evening. Kingsberg’s time sheet for that day is in evidence showing that he checked out at 5 :00 P.M. Latourell testified that he placed the date of the job in June because Kingsberg told him it was about the time of a strike in the shop which .took place in June. Kingsberg testified that Latourell watched him do the spraying. Latourell' testified that he did not see the spraying applied but saw the job on the same day. On February 17, 1944, Latourell wrote the defendants’ manufacturer: “One of the particular reasons for our definitely remembering this job is that it was done during a period when the employees of this company were on strike which was settled on June 19th, 1941, and only a few of the employees were allowed to enter the building or property at that time, to finish or complete various of the jobs which were in process and which were required in a rush to be put back in service.” According to the latest testimony of defendants’ witnesses the job was not begun until June 23d and there was no evidence that at that time only a few of the employees were in — 50 — the building. And Kingsberg’s time sheet shows that he was one of the employees who was out on strike and hence not one of the few who were allowed to enter to finish pressing work on hand. No evidence whatever was offered to show the subsequent history of the shaft nor was there any explanation for this failure. Although there was evidence that other hardened shafts had come in for metalliz-ing before June, 1941, there was no explanation as to why neither Latourell nor Kingsberg suggested preparing them by arc-welding. After June 19, 1941, the shop continued its practice of preparing hardened shafts by annealing and rough-threading, sending the larger pieces to outside firms for annealing although fusion bonding, if known and available within the shop, would have been quicker, more economical and generally desirable for its capacity to prepare the shaft without the danger of warping and softening which was incidental to annealing. Upon the foregoing facts I find that the defendant has failed to sustain the burden of proof as to this alleged prior use. 17. The Capital City Welding Works Prior Use. Most of the testimony as to this use comes from the defendants’ witness Kelling who was examined by the plaintiff in the presence of defendants’ counsel before trial. The alleged use upon which the defendant relies related to a job for the Missouri Power and Light Company done on July 15, 1938. The witness testified that the job was recorded in an invoice of Capital City to the customer which under date of July 15th included the item “Build up two shafts by metalliz-ing $50.70”. As to this job the witness testified that after rough-threading the first shaft and metallizing he noticed that ,the metallized coating had cracked. That having had a metallizing gun for only a few months, he called his friend Schroeder in St. Louis and told him he was having difficulty with a hardened shaft and asked his advice. That Schroeder told him to roughen it with an electrode preferably of nickel steel. That having no nickel steel rod on hand, he thereupon experimented on a piece of scrap metal with various rods on hand and after ten or fifteen minutes got what appeared to be a satisfactory result from a copper rod. At .this 'stage in his testimony it developed that the two shafts about which he was talking were not shafts at all but were two sleeves in which the shaft reciprocated. This, although his only means for dating and otherwise identifying this particular job was the invoice referred to which specified the building up of two shafts by metallizing. He then proceeded to testify further as follows: that using a copper rod in accordance with the method suggested by Schroeder over the telephone,he roughened one sleeve and metallized it; the result being apparently satisfactory he then proceeded to metallize the other sleeve by the same method; and that both sleeves were delivered to the customer at the close of the same day. At this stage the plaintiff examined a superintendent of the Missouri Power and Light Company who had given the foregoing job to Kelling. His testimony was that the job consisted of the shaft and the two sleeves from a pump. That these three parts were given Kelling for building up at substantially the same time and were all returned at substantially the same time. However, it was found that the metallizing on the shaft was loose and consequently it was junked and a new shaft procured; the sleeves, however, were installed and served for four years. That he had not been informed that any electrical process had been used on one or more of these pieces. And that he was unable to say whether any of the part9 had been built up by metallizing. Kelling at the trial admitted that he had worked on the shaft as well as the sleeves. He testified further that the shaft came into the shop some little time before the sleeves and that he had learned of the failure of the shaft before he finished the first sleeve. To explain the absence of any job sheet of his own which showed the job to include three pieces, and not two only, he said that since the shaft had failed no charge therefor was made or recorded. But he also testified .that it had been his practice to make out a job slip when the work was done or at least when it was delivered to the customer. That prior to his acquisition of a spray gun by purchase from .the defendants’ manufacturer in the spring of 1938 he had used a method of building up metal parts by a welding process without metallizing. That the sleeve on which he claimed to have used the process learned from Schroeder as preparatory to metallizing a hardened shaft had been so badly worn that it had completely worn through, so that even before — 51 — his first attempt to metallize it he had built it up to a certain extent by welding. That when informed of the failure of the shaft he besought the customer for an opportunity to do it over without avail. In that connection, however, he did not testify that he informed the customer that one of the sleeves which the customer had accepted as satisfactory had been done by a different process which might have been expected to be as successful on the shaft as on the sleeve. That his work on the sleeve had been witnessed by a specified assistant, but when confronted with his own social security records he admitted that the assistant specified was not in his employ until a later time; that the true witness was another assistant now in military service in the Pacific. That after plaintiff’s attorneys had made inquiries of him as to his shop job sheets relating to jobs which might show the use of the fusion bonding practice, he had looked through a mass of his old job sheets and because he found none containing a record showing the use of the process, he destroyed the lot. He admitted that when questioned by plaintiff’s attorneys in his shop before his first testimony he had undertaken to illustrate the means which he had used to roughen .the first sleeve which he had rebuilt for the Power and Light Company and that in this demonstration his electrode had frozen to the work surface. Kelling’s testimony throughout, both out of court and in court, was exceedingly confused, vague and contradictory. It is supported by no other witness nor by any contemporary writing — -unless contrary to my view an invoice relating to “2 shafts” may be deemed to support the use of a fusion-bonding step not at all indicated on the invoice and now claimed to have been practiced only on the sleeves. On the whole, I find that this prior use has not been sufficiently proved. 18. The Schroeder Prior Use. The Schroeder Welding Company in St. Louis was a small machine repair shop of which Armin Schroeder has long been the proprietor. Since 1932 he has been assisted in the work of the shop by hi's two sons, Ira and Harold, and at irregular intervals by an employee named Burns. The testimony as to this use comes from Armin and Ira Schroeder, Selhorst, Moore and Kunkler. The witness, Kunkler, testified that in 1933 he saw Armin Schroeder repair cracked, automotive cylinder-blocks and valve-seats by notching out the crack, filling it in with electrode-material from a McQuay-Norris machine and then applying sprayed metal with a metallizing gun. As to .this testimony, it will be noted that Kunkler scarcely qualifies as a disinterested witness; he is the chief executive officer of the defendants’ manufacturer and Schroeder was his customer. Moreover, if a sight of such technique had indeed served to teach Kunkler the process later developed by Meduna, it is hard to understand why a teaching of the practice was not included in the instructions for the use of the metallizing equipment which accompanied the equipment sold by the defendants’ manufacturer in the intervening period. However that may be, it clearly appears from the Schroeders’ testimony, and is not disputed by Kunkler’s testimony, that the method which Kunkler claims to have witnessed in 1933, if practiced at all, proved so unsatisfactory (ninety per cent failures according to Armin Schroeder) that it was abandoned in about six months as unsuccessful. There is testimony from the Schroeders, with some support from the witness Sel-horst of a process practiced in the Schroeder Company beginning in about 1938 (Sel-horst placed the date as 1935) for the repair of cylinder-head valve-seats. According to this testimony, after the crack was notched out, tapered pins were drilled into the side-wall so as to interlock. Thereafter, according to the testimony, electrode-material was deposited in the crack by a McQuay-Norris machine and thereafter, but only if the crack still leaked, a coating of sprayed metal was applied. It is apparent that the essence of this method was the interlocking arrangement of the tapered pins. Indeed, there was testimony that an earlier variation of the practice, in which studs without any interlocking arrangement of the exposed ends had been used, had proved unsuccessful. Thus, if used at all the gist of the practice was confined to the interlocking tapered pin feature; the electrode-material was used as a seal to stop leaks and the sprayed metal, if used at all — which was seldom — was to seal a fine crack not filled by the electrode-material. I doubt, therefore, whether this claimed practice of using electrode-material to fill cracks with, an occasional and incidental application of — 52 — sprayed metal to fill the finer cracks not stopped with electrode-material, even if it existed prior to August, 1941, was the Me-duna process. For the interlocking pins of the Schroeder practice were its indispensable bonding means. However that may be, I find that such a use, if it existed, was not public but secret, —-at least until 1942. The testimony of the Schroeders is to that effect. Kunkler himself testified that he was cautioned against interrupting Schroeder while engaged on his secret process. And there is evidence that Selhorst, the mechanic from a neighboring shop who came in to reseat the valve after the repair job had been done, had been requested not to disclose the process of repair and had respected the confidence reposed in him. The next evidence of a prior use by any of the Schroeders relates to the work done on cam shafts for a bus company. Armin Schroeder testified that beginning in 1938, the bus company began furnishing him with hardened shafts for metallizing. Some of these which had previously been rebuilt by welding were sufficiently soft to permit of rough-threading which was his customary method of preparation for met-allizing. But some of these shafts, apparently those which had not been previously welded, according to his testimony, proved to be too hard for rough-threading. With these he admitted fruitless attempts to prepare the shaft by machining and by use of a special knurling tool, and testified that he had eventually developed a practice of grinding the shaft under-size, then roughening it by fuse-bonding with the nickel electrode of the McQuay-Norris machine, and thereafter applying sprayed metal. He testified further that all the cam shafts received from the bus company for metallizing after his development of the fuse-bonding process, were treated by that process. The reliability of this testimony, however, is brought into doubt by other evidence. Thus after this controversy arose, when the defendants’ manufacturer was making active efforts to obtain evidence of a prior public use of the Meduna process, inquiry was made of Schroeder, its friend and customer, and he replied by a letter written on June 25, 1943, in which after having “given this a lot of -thought” he said that he had used fuse-bonding to repair cracked valve-seats and water jackets iri 1933 but “did not use it on hard shafts untill (sic) about two years ago and then only when there was no other way to get a good surface.” If, as he testifies, he had used fuse-bonding on cam shafts for the bus company in 1938 and for years thereafter, it is indeed strange that in this letter after much thought and without any mention of its continued use on cam shafts, he should have harked back to his unsuccessful and abandoned process of 1933. Moreover, the plaintiff confronted Schroeder with four cam shafts from the bus company which, according to the testimony of an officer of that company, had been installed after metallizing by Schroeder not later than May 9, 1941. According to the usual practice of the company, such shafts were not held in stock after rebuilding for longer than six months. It is thus reasonable to believe that Schroeder’s work on these shafts had been completed not earlier than December, 1940, — - some two years after he claims to have developed the fuse-bonding process for such shafts. Yet, when confronted with these shafts he was unable to identify any one of them as having been prepared for metallizing by the fuse-bonding process and one at least he identified as having been rough-threaded, — a fact inconsistent with his earlier testimony that after 1938 all such shafts he had prepared by fuse-bonding. Moreover, there was testimony from the Schroeders that they resorted to the fuse-bonding process on these cam shafts of the bus company only after they had unsuccessfully attempted to secure the desired result by a knurling tool. Ira Schroeder in his deposition was explicit on that point. He was then confronted with a letter which he himself had signed in behalf of his father’s company showing that the knurling tool thus claimed to have been used had not been purchased until February 20, 1941. Later, in court, he testified that he had been mistaken in his earlier testimony that the fuse-bonding process was used on the bus company’s cam shafts before — instead of only after — the acquisition of the knurling tool. But if the Schroeders had already developed a satisfactory process for hardened shafts with the McQuay-Norris machine which they already had on hand, why then should they have purchased the knurling tool in 1941? May it not be that Ira’s first recollection is right and that the fuse-bonding was not developed until after 1941? One would expect his recollection of the sequence of the development to be — 53 — more accurate than his recollection of the date. Ira Schroeder in court, in what appears to me to be an apparent effort to counteract features of his father’s earlier testimony on deposition damaging to the defendants’ case, testified that instead of using the fuse-bonding process on all of the cam shafts received from the bus company, they used it only on four such shafts which alone were too hard to rough-thread, all the others, apparently, — like one of those which had been exhibited to Armin Schroeder on deposition — having been softened by a welding process. Such testimony, to be sure, is not necessarily inconsistent with the testimony of the bus company officer which the plaintiff took on deposition. It is, however, inconsistent with the testimony of the same witness given earlier on deposition and with the earlier testimony of his father. The witness, Moore, was a lawyer without mechanical experience, who was engaged in a business enterprise next door and apparently spent much of his working day visiting with his friend Armin Schroeder. He testified that he saw Schroeder prepare the cam shafts, apparently those for the bus company, with a McQuay-Nor-ris machine and then metallize it. But when pressed to give the detail of the whole process he made it appear that the electrode-material was applied directly to the worn surface, instead of first grinding off th'e worn surface as was Schroeder’s testimony. Nor am I impressed by the credibility of this lay-witness as to matters of mechanical detail. If he spent as much time in the shop as he says, he must have seen Schroeder do hundreds of jobs and it is highly unlikely that he has an accurate memory as to the date and detail of the work done on the four particular shafts to which the defendants’ claim of prior use is now reduced by Ira Schroeder’s latest testimony. To summarize the numerous and voluminous depositions relating to this prior use there is in the record no writing or documentary evidence that it ever existed prior to August, 1941. Nor did the defendants produce a single contemporary specimen of the highly durable material to which the process had been applied. This failure, as also the failure to produce Harold Schroeder or Burns, the Schroeder employee, as witnesses gives some ground at least for inference unfavorable to the defendants’ claims. The testimony of the several witnesses is self-contradictory and contradictory of each other in a number of details. These contradictions I do not attempt, to resolve; I find only that the defendants’ proofs are not sufficiently clear and convincing to establish this claimed prior public use. 19. Infringement. Prior to action brought, the defendants obtained the use of a bonding machine from the Metallizing Company of America, the manufacturer of the machine, which has assumed the defense in this case. This machine the defendants used to fuse to machine parts electrode-material having the characteristics described by the Meduna patent as a preliminary step in the process of spraying metal thereon. In so doing they infringed claims 1, 2, 3, 4, 5, 6, 7 and 11. Claims 8, 9 and 10, however, are limited to a fusing step accomplished electrically by a voltage specifically limited (1/2 volt to 20 volts) and in this voluminous record I am unable to find evidence to show definitely that the defendants operated within this voltage range. Consequently, I find that the infringement of claims 8, 9 and 10 has not been proved. Conclusions of Law 1. None of the claims in suit are invalid for lack of patentable invention or for anticipation. For its opposition to this conclusion, the defendants rely largely on Emmett v. Metals Processing Corporation, 9 Cir., 118 F. 2d 796. That was a case involving the validity of a patent for a process in the metallizing art. But by the date of that invention, the art had already learned the necessity of roughening the surface of the metal part to which sprayed metal was to be applied: the use of sand-blasting for this purpose was well known. And the prior patented art also disclosed a method of screw threading by a cutting tool as a preparation for bonding with molten aluminum. To this prior art, all that the patent in suit contributed was the suggestion that the cutting tool should be so set as to tear the side walls of the V-shaped thread. But this was at most a meagre contribution: like the prior art it was restricted to a mechanically produced variation or roughening of the surface to be sprayed. As such, it was rightly held to be within the range of mechanical skill rather than a demonstration of patentable invention. — 54 — But Meduna’s contribution was more substantial and less conventional. Instead of roughening the base metal by mechanical means he used electrical means; and even more important, instead of merely roughening the metal of the base and applying sprayed metal directly thereto he conceived the idea of interposing a layer of electrode-material fused to the base having a metallic structure peculiarly suitable to bond with the sprayed metal. Although the deposit of porous electrode-material was concededly old in the field of welding and the spraying of molten metal was old in the field of metallizing, the combination of these steps as taught by Meduna was indeed a new and highly useful addition to the metalliz-ing art. Its patentability, of course, is not to be denied simply because Meduna was fortunate enough fortuitously to stumble upon a clue to the solution, without imposing laboratory and laborious research. It should be affirmed for the new result obtained from the coordinated combination of old practices never theretofore associated. For, as my findings show, the use of the McQuay-Norris transformer (Par. 4) and the disclosures of Brewster and of Williams (Par. 5) contain no teaching that the deposit of porous electrode-material which they envisaged was suitable as a means for bonding solid metal to' sprayed metal. For them the porosity of the deposit was a liability to be overcome by peening rather than an asset to be directly utilized. Indeed, I understood the defendants’ expert to concede that the Brewster patent did not anticipate. Williams, I think, comes closest to the Meduna invention. But although he taught the utility of a porous deposit as a bonding means he taught only arc-welding and failed to disclose Meduna’s superior means for producing a deposit by heat-resistance fusion. However that may be, there is nothing in the Williams specifications or indeed in the entire record of fact to show that the problems inherent in the bonding of metal to metal are identical with those involved in bonding rubber to metal. We do not know whether Williams contemplated that the rubber should be applied in liquid, plastic or sprayed form, nor its capacity for flowing into the craters and interstices of the porous electrode-material, nor whether the rubber if capable of penetrating the interstices may there be effectively hardened by vulcanization and so anchored. I cannot hold that the Williams disclosure is within the field of the metallizing art without convincing evidence that the bonding of metals and the bonding of rubber involve common problems and hence lie in a common field. I had not supposed that that was so: certainly I find no evidence to that effect in the record. On the whole, I view the Williams disclosure neither as an anticipation nor as derogative of the inventive quality of Meduna’s claims. The other prior patents, (Par. 5) as indeed the prior publications (Par. 6) upon which the defendants rely, seem to me too remote to justify discussion. The article in Commercial Car Journal, like Brewster, evidently viewed the porous character of crack-filling weld material as a defect to be “sealed” by sprayed metal — not as a recommended bonding medium. And the article in “Automobile Engineer” seems to me far afield. In short, the patent here seems to have more intrinsic claim to classification as a patentable invention than that sustained in Goodyear Co. v. Ray-O-Vac Co., 321 U.S. 275, 64 S.Ct. 593, 88 L.Ed. 721, and substantially more claim to such classification than that invalidated in Cuno Engineering Co. v. Automatic Devices Corporation, 314 U.S. 84, 62 S.Ct. 37, 86 L.Ed. 58, or Picard v. United Aircraft Corporation, 2 Cir., 128 F.2d 632. This conclusion is corroborated, but by no means dependent upon, the extent to which the process .has been commercially applied since the application for the original patent. There is no evidence that since that date any rival process for metal-lizing hardened metal has come into vogue. The inherent economy and celerity of rebuilding by metallizing gives good ground for inference that the commercial success attained is not temporary only, due to the scarcity of new parts for replacement in time of war. Cf. Justice Jackson’s dissent in Goodyear Co. v. Ray-O-Vac Co., supra. And the sustained and successful activity of defendants’ manufacturer, since the publication of the invention, in spreading (for its own commercial purposes) the practice of the process, although not a concession of its inventive quality, at least ¡tends to demonstrate that a real need for the process existed which without aid from Meduna it had never theretofore been able to fill. — 55 — 2. None of the claims in suit are invalid because too ambiguous and indefinite, or because functional. In raising this defense, the defendants are hypercritical. The patent is sufficiently definite to teach any skillful mechanic how the claimed process may be practiced. That the fusion step is stated to surface the base with a deposit of material having “minute” craters, etc., is merely descriptive of the result which will be obtained by following the specifications. I incline to believe that such descriptive matter might properly be treated as surplusage not affecting the validity of the claims. In any event, as against the background of the prior art and in the light of the specifications, the word “minute" as appearing in the claims, even if viewed as a limitation, sufficiently defines the claimed invention. And the defendants’ objection that the claims state nothing but the function of an existing machine (such as McQuay-Norris) is wholly without merit. Without drawing upon the introductory phrase of each claim, the invention is plainly stated to include as one step of the process the application of spraying metal to a metal surface previously conditioned as specified. Thus the claimed invention is not the mere function of an existing machine (such as the McQuay-Norris machine) ; nor is it the single, preliminary step of pre-conditioning the surface to be sprayed. The claimed invention is narrower. As plainly stated, it is the combination of at least two steps. Granted that one step is a result produced by McQuay-Norris and that the other step is the use of a spray-gun, neither step alone produces Meduna’s results. No possible use of any prior machine would in itself without the other step constitute an infringement. This is so even if, as the defendants urge, the introductory phrase of each claim is treated as descriptive only. For the body of each claim states a patentable process. 3. The plaintiff has failed to prove that the inventor’s use of the invention prior to the original application was mainly for purposes of experiment. The statute, 35 U.S.C.A. § 31, sanctions the issue of a patent only upon inventions which have not been “in public use or on sale in this country for more than one year prior to” the application. The courts, however, have excepted from this limitation of the statute a use or sale by the inventor “which may be properly characterized as substantially for the purposes of experiment.” Smith & Griggs Mfg. Co. v. Sprague, 123 U.S. 249, 256, 8 S.Ct. 122, 126, 31 L.Ed. 141. In determining whether a. given case comes within this exception the fact that the inventor has prior to the privileged period used his invention in the course of commerce, is not necessarily fatal. For plainly a given use may at the same time be both for purposes of experiment and for purposes of commerce. And where the use is motivated by both of these considerations, the case will fall within the implied exception only when the “main purpose” of the inventor was one of experiment, the commercial feature of the use “being merely incidental and subsidiary.” Smith & Griggs Mfg. Co. v. Sprague, supra, 123 U.S. 249 at page 266, 8 S.Ct. 122, at page 130, 31 L.Ed. 141. As Judge Learned Hand observed in Aerovox Corporation v. Polymet Mfg. Corporation, 2 Cir., 67 F.2d 860, this test is indeed hard to apply. Here there are some facts which tend to support Meduna’s direct testimony that his use of his process prior to August 6, 1941, (his' application date was August 6, 1942) was influenced by a desire to ascertain whether the processed product would stand up under actual service conditions. But the facts also compel the finding (Pars. 9-11) that throughout this period his primary purpose was commercial; that his experimental purpose was only “incidental and subsidiary”. The conclusion seems necessarily to follow the inventor’s lack of any system for following up and keeping trace of jobs done by the new process. Some of these were done for owners known to the inventor. But even as to these, he made inquiry only at casual intervals at meetings not especially sought for that purpose. And he seems to have made no inquiry— certainly no systematic inquiry — as to the extent of work to which each of these rebuilt parts was subjected, — an item which, as even the plaintiff’s evidence shows, was an indispensable factor in gauging the success of these “experiments”. Nor did he keep, for aught that appears, written record of the subsequent history of these jobs showing the work to which each was subjected and the state of its wear at stated intervals. Indeed, his only record of these jobs was just as meagre as his other commercial records. — 56 — But even if we lay out of account the jobs done for known owners, we are faced with a number of jobs for owners unknown to the inventor which came to him through repair stations and middlemen. As to these jobs, at no stage did he make any effort to ascertain the names of the owners or to follow the subsequent history of the jobs. The plaintiff suggests that the inventor might properly assume that, if these jobs subsequently proved not satisfactory, they would be brought back to the same repair station and hence come to his attention. But as to this it was equally plausible to attribute the absence of returned work to other causes. For instance, on the failure of a rebuilt part, the owner might come to the conclusion that it was more satisfactory to replace the worn-out part by a new one rather than to have it again rebuilt, and so register no complaint to the particular repair shop through which the metallizing job had been done. Perhaps if the inventor had guaranteed a definite life to the parts rebuilt by the process, the absence of failures brought to him for reprocessing would have had more significance. But there was no evidence that all the jobs were done by the inventor under such a guarantee. Moreover, if a rebuilt part went into a motor vehicle — as many did— which happened to come into collision and hence was junked, the inventor’s absence of such knowledge would not warrant an inference that the part was still in use and that his. “experiment” was proving successful. Thus considered, the inventor’s practice of sending out the product of his process to unknown owners on a commercial basis is a fact tending to show that at least as to these jobs his primary purpose was commercial. In any event, the fact of the inventor’s use having been fully established, the burden is on the plaintiff to prove the predominance of the experimental motive. This burden the plaintiff must discharge by evidence that is “full, unequivocal, and convincing.” Smith & Griggs Mfg. Co. v. Sprague, supra. Surely I cannot hold that this high degree of proof has been maintained. For even the facts tending to sustain the presence of an experimental motive fall short of demonstrating the predominance of that motive. On that issue, at best, the proofs are highly equivocal. 4. The defense of a prior public use by the inventor has not been sustained by the defendant: From this defense, the plaintiff seeks another escape by contending that prior to August, 1941, the inventor’s use of the process was not public in that it was practiced by him only in secret. That the only practice of the process was indeed secret, I have found. I found further, though to be sure on conflicting evidence, that prior to the critical date those skilled in the art could not have detected the process from such processed parts as the inventor put into commercial circulation. On this point I accept the testimony of plaintiff’s expert. To me the testimony of defendants’ expert on the point was wholly unconvincing. And it is not without significance that defendants’ manufacturer, who knew of the existence of the process even before its acquisition by the plaintiff and whose eagerness to use the process is fully apparent, offered no evidence to show that it had fathomed the process and begun its practice prior to its publication. On the facts thus found, did the sales of the product made from the invention practiced in secret constitute a “public use” within the meaning of the statute? As to this, on the one hand it is definitely established that to put an invented article subsequently patented into the hands of the public by sale or gift will constitute a public use of the invention even though the essence of the invention is not thereby disclosed to the public. Hall v. MacNeale, 107 U.S. 90, 2 S.Ct. 73, 27 L.Ed. 367, and Egbert v. Lippmann, 104 U.S. 333, 26 L.Ed. 755. It is also established that the sale of the product of an invented machine subsequently patented does not constitute a public use if the machine in producing the product was secretly operated. Gill-man v. Stern, 2 Cir., 114 F.2d 28, Peerless Roll Leaf Co. v. Griffin & Sons, 2 Cir., 29 F.2d 646. So much is clear. But a question relating to sales of the product of a process secretly practiced, is one of considerable difficulty. The defendant relies heavily upon a passage of the opinion in Grasselli Chemical Co. v. National Aniline & Chemical Co., 2 Cir., 26 F.2d 305, on page 309, which states: “Once the invention has been embodied in goods which are put in public use it becomes impossible for a later inventor to secure a patent.” And the text — 57 — further indicates that this principle applies even if the product was such as not to teach the embodied invention even though the product were to be destroyed to facilitate the search. It should perhaps be noted that this passage is avowed dictum. The opinion states expressly that the court prefers “not to rest our decision on this feature of the evidence (i. e., on the prior use).” Instead the claims were invalidated on the ground of a prior patent. However that may be, on analysis it appears that the reasoning of the dictum is not applicable to the case at bar. In the Grasselli case, there was in issue an article claim (4) for vulcanized rubber, as an article of manufacture, made by the use of a specified accelerator, and a process claim (1) for producing vulcanized rubber by the use of the same specified accelerator. The specified accelerator was the gist of the invention of each claim and the novelty of the accelerator was the crucial issue. Thus the facts there were such that if the article claim had been anticipated by a prior use, the process claim must necessarily be held anticipated also: if rubber vulcanized by that particular accelerator was old, of course it was old practice to apply that accelerator to latex to produce vulcanized rubber. Substantially the same situation seems to have been involved in Macbeth Evans Glass Co. v. General Electric Co., D.C., 231 F. 183, cited in the Grasselli dictum. For the Macbeth case also involved a “formula” as well as a process which included the formula. As to the article claim, the Grasselli dictum naturally followed Egbert v. Lippmann and Hall v. MacNeale referred to above, which declared the law defining a public use of article inventions. But the dictum did not go so far as to say that the same rule applied to all process inventions. It failed to distinguish between the article inventions and process inventions because on the facts then before it there was no room for distinction; as I have just pointed out, on those facts a prior public use of the article necessarily implied public use of the process. But later, in Aerovox Corporation v. Polymet Corporation, 67 F.2d 860, 863, the same court was confronted with a process claim which was not inextricably interwoven with an article claim or a formula. There a prior use of the invention had been proved which the plaintiff sought to avoid by proof that the use was not public but secret only, notwithstanding that the product of the process had been put upon the market. To accomplish this avoidance, the plaintiff had tried to prove that at the time (1926) when the product was marketed the invented process could not have been learned from its product. But as to this, the court concluded its opinion by saying: “The plaintiff argues that this was not true in 1926, but the record does not affirmatively bear this out; once more it has failed to carry the burden of proof.” From this I conclude that a process invention which, as here, does not involve the use of an article invention may be practiced in secret, and the sale of its product will not be held to constitute a public use of the invented process if the plaintiff sustains the burden of proving that at the time the product is sold the process could not have been learned from the product. And that burden, as I have found (Par. 13), has been amply sustained by the plaintiff here. This conclusion, I think, not only fully recognizes the inherent consistency between the Grasselli and Aerovox opinions but also is in harmony with other aspects of the law defining public use. If, as' is well established (Gillman v. Stern, supra), an invented machine may be secretly operated and its product freely sold without involving a public use or sale of the invention inherent in the machine, I can see no reason whatever for withholding the same immunity from an invented process, provided it is proved that the inherent invention could not be learned from the product sold. Certainly there is nothing in the statute to require a distinction. Nor does my conclusion offend a dictum in the opinion in Peerless Roll Leaf Co. v. Griffin & Sons, 2 Cir., 29 F.2d 646, at page 649, upon which the defendant has placed a reliance not warranted, I think. There in dealing with a machine patent the court said: “The sale of the product (of the machine) was irrelevant, since no knowledge could possibly be acquired of the machine in that way. In this respect a machine differs from a process”, citing the Grasselli case. To be sure, as is implicit in my conclusions, the sale of the product of the process is relevant, not irrelevant. Evidence of such sales will indeed be held to demonstrate a public use of the process unless, as the Aerovox opinion recognizes, — 58 — the plaintiff proves that the process could not be learned from the product sold. Agreeably to the dictum, I have freely admitted all proffered evidence of the inventor’s sales, as being relevant to the defense. But I have found that the plaintiff’s proofs constitute a sufficient avoidance of the defense. 5. The defense of abandonment has not been sustained. On this issue the ultimate test is the intention of the inventor. In Macbeth Evans Glass Co. v. General Electric Co., 6 Cir., 246 F. 695, it was held that conduct whereby a formula was practiced in secret for ten years while its product was widely sold demonstrated an election to forego a monopoly under the patent laws. It does not follow, however, that in every case the secret practice of a process prior to application for a patent thereon, or even for more than a year prior to the application, is conclusive evidence of an election to forego patent protection and hence of abandonment. Indeed, in Peerless Roll Leaf Co. v. H. Griffin & Sons Co., 2 Cir., 29 F.2d 646, 649, it was said: “It is possible, even after the time of experiment has passed, to practice the invention without abandonment.” This, I think, is such a case. For here it is apparent that from the time he stumbled upon his discovery the plaintiff sought jealously to preserve his patent rights. (Pars. 9 and 13) Although I have found that as to a part of his secret practice, at least, his experimental purpose was subordinate to a commercial purpose, the experimental purpose was nevertheless present and genuine. The presence of such a purpose, even if not inconsistent with an intent to abandon, at least serves to show that the delay in making application for a patent was not necessarily attributable to an intent to forego a patent. Here the delay was at most of moderate dimensions. After the inventor stumbled upon his discovery in March, 1940, he was entitled to a reasonable time within which to reduce his invention to practice. Agawam Co. v. Jordan, 7 Wall. 583, 607, 19 L.Ed. 177. In determining how long a period might reasonably be allowed the inventor for this task we may take into account his circumstances. While developing his invention it was necessary for him to continue to earn a living in the little two-man machine shop which he had recently acquired. Nothing in the law required him to give up all his other work and devote his whole time to the task. He was without means to accelerate the task by the purchase and use of laboratory equipment. And while on the whole it seems clear that the invention had been reduced to practice by some time before August, 1941, his secret practice of the invention for the time intervening was of slight weight in itself, and of wholly inadequate weight in the light of the entire situation, to support a finding of an intent to abandon. National Aluminate Corporation v. Permutit, 6 Cir., 145 F.2d 175, 180. And the burden of proof as to this defense is, of course, on the defendant. Electric Storage Battery Co. v. Shimadzu, 307 U.S. 5, 6, 16, 59 S.Ct. 675. 6. The defense of a prior use has not been sustained. As my findings show (Pars. 15-18), each one of the four prior uses pleaded is supported only by the oral testimony of customers and friends, and their employers, of the defendants’ manufacturer. I do not mean to imply that all this testimony is deliberately perjured. But common experience demonstrates that such witnesses are susceptible to sugestión. Some of them are susceptible to suggestion. Some of them outcome of the case: if the validity of the patent is sustained they may become liable as infringers. The contradictions in their testimony are numerous and even those relating to items not of controlling importance yet serve to show that the substance of their testimony, and particularly their testimony as to the claimed dates of the alleged uses, was inadequate to sustain the burden of proof under the doctrine of the Barbed Wire Patent Case (Washburn & M. Mfg. Co. v. Beat ’Em All Barbed Wire Co.), 143 U.S. 275, 12 S.Ct. 443, 36 L.Ed. 161, and General Motors v. Leer, 2 Cir., 60 F.2d 902. Moreover, the universal absence of any contemporaneous writing or document sufficient in itself without oral supplement to attest a use of the process, or of any commercial specimen of some product of the claimed use, is of especial significance in a case such as this where the relevant period terminated as recently as August, 1941 (the action having been brought in June, 1943) and the product of the process being of durable material such as hardened steel. In this connection it may be observed that in connection with the Schroeder use the — 59 — plaintiff was able to bring into court commercial specimens of Schroeder’s handiwork made in the relevant period though not shown to embody the Meduna process. But the defendants’ manufacturer who combed the country from coast to coast has failed to produce a single commercial specimen. Moreover, except for Schroeder’s, none of the other claimed prior uses were shown to have been successfully operative, or to have constituted more than an abandoned experiment, or a casual practice soon discontinued. In this respect also the defense of prior use was inadequate. Electrical Engineers’ Equipment Co. v. Champion Switch Co., 2 Cir., 23 F.2d 600; Concrete Mixing, etc., Co. v. Powers-Kennedy Contracting Corporation, 2 Cir., 27 F.2d 668; Wenborne-Karpen Dryer Co. v. Cutler Dry Kiln Co., D.C., 285 F. 73, reversed on other grounds, 2 Cir., 290 F. 625. As to Schroeder’s prior uses, that of 1933 was, according to his own testimony, unsuccessful. As my finding indicates, his use of sprayed metals as an occasional supplement to the use of interlocked tapered pins for the repair of cracked valve-seats was not the use of the Meduna process at all; according to his testimony the tapered pins were an indispensable element in his process. But, however that may be, this use was secret and cannot serve as an anticipation. Gillman v. Stern, 2 Cir., 114 F.2d 28. This leaves only for consideration the four cam shafts to which Schroeder claims to have applied the Meduna process. There was, to be sure, credible evidence from Schroeder’s customer that whatever method Schroeder used to rebuild these shafts was satisfactory. But the evidence that the method used was the substance of the Meduna process is subject not only to all the frailties of the defendants’ proofs to which reference has been made, but also to the devastating contradiction that the method was developed in 1938, after experiment with a knurling tool which concededly was not purchased until some time after June 1941. 7. All the claims in issue are valid. 8. The defendants have infringed claims 1, 2, 3, 4, 5, 6, 7 and 11. The infringement of claims 8, 9 and 10 has not been proven. A decree accordingly may be submitted, on notice, unless notice is waived.