Full opinion text
BARNES, District Judge. The patents in suit are Ruben No. 1,714,-191, issued May 21, 1929, on an application filed December 22, 1926, and Ruben No. 1.710.073, issued April 23, 1929, on an application filed March 21, 1927, which was a continuation in part of the earlier application. Both patents relate to electric condensers. Of patent No. 1,714,191, claims 4, 5, 6 and 7 are in suit. Of patent No. 1.710.073, claims 2, 5, 6, 7 and 8 are in suit. An electrical condenser takes advantage of the fact that when there exists between two conductors, insulated from one another, difference of potential or electric voltage, there is set up in the insulating medium between them a state of stress in which energy is resident. A condenser, therefore, consists of two conducting bodies separated by an insulating material. An electrolytic condenser is characterized by the fact that one of the conductors is a metal plate capable of having formed upon it a film of insulating material, while the other conductor is an electrolyte-containing material capable of forming the film on the plate. The word “electrolyte” means a chemical substance which, when dissolved in a liquid, will be dissociated and render the liquid conductive. The word is also used to mean the mixture of liquid and chemical substance. A wet electrolytic condenser is one in which aluminum electrodes are immersed in a mobile liquid aqueous electrolyte solution. The patents disclose dry electrolytic condensers of the aluminum film-forming type in which the aluminum electrodes separated by porous spacers of gauze or paper impregnated with a viscous pasty electrolyte composition are rolled up in spiral form. In patent No. 1,714,191 Ruben says: “For the electrode material I prefer thin aluminum sheets having an oxide film formed upon the surfaces before assembly, and as the electrolyte, glycerine having a relatively small amount of water and having mixed therewith suspended powdered sodium bicarbonate, a small percentage of boric acid being present to stabilize the paste which has a sodium borate content due to reaction of the boric acid with the sodium bicarbonate and to increase the conductivity of the mixture. In absorbing from the atmosphere sufficient moisture to maintain the desired conductivity, the ionic conductivity of the condensers is held constant. ., , * * * * * * * “The preferred method of assembling the element of this device is to apply the pastelike electrolyte mixture to a sheet of gauze spread over the oxide surface of the aluminum sheets, the electrolyte being then in the form of a layer about Jie inch in thickness. Over this should be placed the other electrode element, and the whole compressed for complete contact. If the elements are to be coiled another layer of the gauze and the paste electrolyte should be applied so as to keep the electrodes separated. The electrolyte permeates the gauze and fills the interstices establishing a layer over its entire area. ‘Forming’ the electrodes or coating the surfaces with aluminum oxide may be accomplished by connecting the sheets as the anodes in circuits in a bath of ammonium sulphate, the oxide serving to maintain the leakage at a low value.” In patent No. 1,710,073 Ruben says: “Glycerine has been found to be a particularly satisfactory material for use as a suspension medium for the film-forming electrolyte since it possesses in itself all the characteristics desired, being sufficiently viscous to form a suitable paste with a film-forming electrolyte, having a relatively high dielectric constant value to thereby increase the alternating current capacitance of the condenser, and being highly hygroscopic so that loss of life due to evaporation of electrolyte is effectively overcome. % % % j|C ^ “The particular film forming electrolyte employed may be any one of a large number of well known -electrolytes, but I have found a mixture of boric acid and sodium borate to be especially satisfactory. As a general rule, it may be said that film-forming electrolytes should preferably be either neutral or slightly acid in character and for this reason it is customary to employ for this purpose water solutions of the salts of weak acids. Similarly mixtures of weak acids with salts of the above mentioned character are used to replace the latter in order to reduce the alkalinity of the solution and thereby more nearly approach neutrality. I use a mixture of boric acid and sodium borate dissolved in water since such a mixture may be readily proportioned so as to be substantially stable chemically and possess the desired conductivity values. Also, a concentrated solution of these compounds with water mixes readily with glycerine to form a glacial like paste which is especially easy to spread. In forming the paste, the electrolyte is first prepared preferably by dissolving the film-forming electrolyte materials in hot water and this solution is then thoroughly intermixed with glycerine to form a product of uniform composition. s'fi >]; s>« ifc % "The preferred electrolyte composition is one in which the ratio of sodium borate to boric acid by weight is approximately 17 to 1 with the water content about 2 c.c. to each gram of boric acid. This solution is then added to and thoroughly mixed with glycerine whose content is preferably in the neighborhood of 2% times that of boric acid. These ratios and figures are of course only exemplary and the values used may depart materially from those given while remaining within the scope of the invention.” Radio receiving sets began to be manufactured and sold in substantial quantities about 1920. The electrical current to supply such sets was provided by dry batteries for the high voltage B current, and a storage battery for the low voltage A current. The B batteries had to be replaced frequently. The A battery had to be recharged every week or two, and it contained sulphuric acid which was ruinous to rugs and furniture if it spattered. The house lighting current (in most cases 110 volts alternating current) was not suitable to operate the radio set. There was a need for some means by which the house lighting current could be made suitable to operate the radio set. The fundamental electrical principles by which the house lighting current might be adapted to operate the radio sets were well understood. By means of transformers it could be raised or lowered to the appropriate voltage; by means of rectifiers the alternating current could be converted into pulsating direct current; and by means of filtering devices comprising condensers and chokes the pulsations could be smoothed out so as to provide the steady direct current necessary. Nevertheless a satisfactory A-battery eliminator did not appear on the market until the advent of Ruben’s dry electrolytic condenser, and the reason was that until that condenser was produced nobody knew how to make a condenser which had sufficiently large capacity in sufficiently small volume at a low enough price. Several other expedients were resorted to. First came the home charger which, by means of a transformer and rectifier made it possible for the set owner to recharge the storage battery in his own home without having to take it to a garage or battery service station. Then came the trickle charger, a lower powered unit consisting of a transformer and a small rectifier which were connected to the house-lighting current and the storage battery in such a way that the battery was continuously recharged at a low rate (a trickle) at all times except when the radio set was in operation. Then came the Tigerman A and B current supply. Grigsby-Grunow-Hinds Company tooled up in 1925 to make these at the rate of one thousand units per day, but had to stop production soon after they started.because of troubles in practical operation. They sold 18,000 units but had to take back 16,000 of them for credit. On the B-power side of this device the difficulty was only in the inadequacy of the rectifying tubes. On the A side, the trouble was with a small storage battery having a j elly electrolyte which was used in the filter circuit in lieu of a condenser. The jelly electrolyte would dry out in operation and the filter circuit would no longer perform its intended function of smoothing out the pulsating rectified current. When this happened it would give rise to a loud hum in the speaker of the radio set. Shortly after the withdrawal of the Tigerman device from the market, the Grigsby-Grunow-Hinds Company was able to obtain from the Raytheon Manufacturing Company of Cambridge, Mass., a satisfactory rectifying device and with it was able to market a B-eliminator which was so satisfactory that 750,000 were sold in a period of twelve or thirteen months. There was still no satisfactory condenser available for an A-eliminator. If an attempt had been made to use the conventional paper condensers which were satisfactory for the B-eliminator, their bulk would have been “as large as a roll-top desk” and the cost would have been prohibitive. If a satisfactory condenser had been available,’ the market for the A-eliminator would have been as large as the market for the B-eliminator, and a profitable one. The Grigsby-Grunow-Hinds engineers tried to fix up the jelly storage battery in some way without success. Slepian, Bush and Edelman also tried to solve the problem. The Slepian and Bush devices never reached the market, and Edelman failed to produce anything of permanent value. Their work will be discussed more fully at a later point. And so from 1920 until after Ruben’s inventions, the storage battery remained an indispensable part of the radio set despite all its drawbacks. The plaintiffs contend, and the court thinks properly, that this chronology of effort and failure proves (a) an extensive period of recognized need and demand, and (b) the requirement for exercise of invention beyond the ordinary skill of the art to supply an improvement to satisfy such demand. When the Grigsby-Grunow-Hinds Company heard that Mr. Ruben had a new condenser suitable for an A-eliminator they sent their chief engineer to New York to find out about it. He witnessed a demonstration of.its effectiveness, learned what was in it and took specimens back to Chicago for test. The tests were satisfactory and Mr. Grunow and the chief engineer returned to New York to make a deal with Mr. Ruben. Mr. Grigsby followed and executed a license agreement which called for a minimum of $20,000 a year in royalties. With the Ruben condenser it was possible to make a satisfactory A-eliminator. The storage battery was no longer required. When this was introduced to the trade in June, 1927, 75,000 units were sold and gave satisfactory operation, less trouble so far as the condensers were concerned than the B-eliminator with the conventional paper condensers. The device was very successful and would have continued to sell in large numbers in that field except for the advent of a new type of radio tube which no longer required continuous direct current to heat the tube filament. It is clear that an insistent need for a satisfactory A-eliminator existed unsatisfied for at least six years; that the only reason for the failure of the skill of the art to provide an A-eliminator was the lack of a satisfactory condenser of large capacity and small volume; that others tried to solve the problem and failed; that the art had to get along as best it could with makeshift expedients until the advent of the Ruben condensers; and that when Ruben finally produced those condensers they were immediately recognized as filling a long felt want. Ruben began to devote his attention to the utilization of house-lighting current for operating a radio set in 1923. He first developed a magnesium copper sulphide rectifier without any electrolytes of any kind. This went into use in a trickle charger which eliminated the tantalum-sulphuric acid electrolytic charger which had been quite popular. This dry rectifier of -Ruben’s is still in use in large volume. Having made a satisfactory dry rectifier he undertook to produce a dry condenser. His first effort involved the use of aluminum plates, one of which had an oxide film on it, with copper sulphide powder between them, the same material he had used in his dry rectifier. He was able to get results but could not get satisfactory life tests which were necessary for a commercial device. He next tried to make a condenser for a B supply by using aluminum electrodes separated by an electrolyte of borax, boric acid and gelatin. This also had drawbacks to its commercial use and could not have been applied to the A-battery eliminator problem. He investigated the possibility of using wet electrolytic condensers for the A-filter circuit and made many of them with various kinds of salts. He found they all had the same inherent limitations, the low resistance (high conductivity) electrolyte tended to destroy the device in use. Finally, he made a condenser which obviated the inherent limitations of the wet condenser. To do this he used instead of a water solution a mixture of glycerin with sodium bicarbonate and boric acid. This mixture not having good conductivity enabled him to overcome the main limitation of the wet electrolytic condenser, namely, the tendency for the current to localize in one section; and thus he was able to wind up the condenser with a large area in a small space. Despite the high resistance of his electrolyte composition he was able to get an efficient condenser because of the close spacing of the electrodes giving a total low resistance structure. The use of the hygroscopic glycerin enabled him to overcome the evaporation difficulties which were inherent in the wet electrolytic condensers, because though some moisture might be driven off while the device was in operation, the hygroscopicity would quickly restore the proper amount and distribution of the moisture. He found that the sodium bicarbonate produced bubbles of carbon dioxide in the glycerin and that unless sufficient time were allowed for the gas to escape the bubbles would form pockets in the windings and reduce the effective area of contact. Being interested in the lowest cost of manufacture, he changed the salts to be added to the glycerin from sodium bicarbonate and boric acid to sodium borate and boric acid. The operation was identical in result but the manufacturing end was improved. Specifications for the manufacture of the condensers to go in the Majestic A-battery eliminator were prepared by Mr. Ruben on January 3, 1927 and forwarded to the Grigsby-GrunowHinds Company. These specifications disclose the subject-matter of the Ruben patent No. 1,710,073. On the question of validity of the patents in suit and particularly of anticipation, the defendant’s first reference is Greinacher Swiss patent No. 81,050 of May 1, 1919. The defendant refers to this patent as showing “in very early days the logical ways of making a dry electrolytic condenser” and quotes from this patent as follows: “It is possible, for instance, to introduce between the electrolytic valve cell a porous substance (such as pumice, sawdust, filter paper) or a powder which contains the electrolyte liquid in absorbed form, such as kieselguhr, for example.” Defendant states that one way, as outlined by Professor Greinacher, was to let the electrolyte be absorbed by filter paper located between the electrodes of the electrolytic cell, and that this is approved present day practice. The plaintiffs contend, and the court thinks that the merits of this particular controversy are with the plaintiffs, that, while this Swiss patent pointed out the logical way of approaching the problem, it inevitably results in failure and does not provide a satisfactory solution of the problem; that so far as the Greinacher patent goes the electrolyte to be absorbed in the filter paper was the same old low resistance wet electrolyte; that Greinacher apparently had no appreciation of the necessity of making any change in the electrolyte composition whatever; that in the Ruben patents and in defendant’s practice the electrolyte is of a permanently hygroscopic, viscous paste-like nature; that it is of high resistance to permit close spacing between the electrodes; that it can be rendered highly fluid by heating so as to be thoroughly absorbed into a porous spacer where it again becomes a viscous material of paste-like consistency; and that these characteristics are not to be found in the Greinacher electrolyte or any other prior art electrolyte. On the question of anticipation, defendant’s second reference is Roberts patent No. 1,650,740, the effective date of which is stipulated to be its filing date of June 15, 1925. The plaintiffs point out that the only pertinent portion of the Roberts specification is that which states that an electrolytic condenser can be made with a diaphragm of a film-forming metal in contact with an electrolytic filling which may be in the form of a liquid or a paste or jelly; that the patent will be searched in vain for any disclosure as to how to compound the liquid or paste or jelly which is to serve as the electrolytic filling; that a structure which would be compatible with a liquid electrolyte would be useless with a paste or jelly electrolyte and vice versa; and that jelly electrolytes produced with gelatin or starch or water glass are so poor as to be useless. In the court’s opinion Roberts does not anticipate. The defendant says that it was prior art to provide the anode with its dielectric film before assembling the condenser elements, and its third reference is Weinberg patent No. 979,906 of December 27, 1910. The defendant says that Weinberg first prescribed and patented the preliminary forming of the dielectric oxide film on the anode. The plaintiff says rightly that the Poliak British patent No. 1069 of 1896 disclosed the same idea in that year. The Weinberg patent is on a wet electrolytic cell. It does not disclose how to construct dry electrolytic condensers and does not anticipate the patent in suit. The defendant says it was prior art to use a wide variety of electrolytes and electrolyte solvents. Particularly, the defendant says that glycerin with ammonium octoborate in solution was the electrolyte of Hayden patent No. 996,583, dated June 27, 1911; that glycerin plus borate, such for example as ammonium borate, was the electrolyte of Peek patent No. 1,008,860, dated November 14, 1911; that glycerin plus phosphate of soda, phosphate of ammonium, carbonate of soda, and “like chemicals,” constituted the electrolyte of Hickley patent No. 900,278, dated October 6, 1908; that glycerin plus a borate and an oxyacid of boron was the electrolyte of Zimmerman patent No. 1,074,231, dated September 30, 1913; that borax and boric acid were in solution to form the electrolyte of Mershon patent No. 1,439,525, dated December 19, 1922; that boric acid and caustic ammonia (ammonium borate producers) were in solution to produce the electrolyte of Buttner patent No. 809,770, dated January 9, 1906; that sodium phosphate, or ammonium phosphate, were in solution to produce the electrolyte of Weinberg Patent No. 979,906, dated December 27, 1910; and that ammonium borate and boric acid were used in the electrolyte of Chubb patent No. 1,270,784, dated July 2, 1918. The plaintiffs rightly say that each one of these electrolytes is essentially an aqueous solution definitely intended for use in some sort of a wet electrolytic cell; that in none of these does the use of glycerin change the character from a highly mobile aqueous electrolyte; and that in any one of the patents mentioned the substitution of the viscous pasty electrolytes described in the Ruben patents would destroy the utility of the devices shown and described; and that, conversely, the use of any of these prior art electrolytes in a dry electrolytic condenser construction would not produce a dry electrolytic condenser but would ruin it. The defendant says it was prior art to provide “so-called dry or non-runny” electrolytes; that an electrolytic liquid absorbed by paper disposed between the electrodes was one form of “dry” electrolyte material disclosed by the Greinacher Swiss patent No. 81,050, dated May 1, 1919; that a gelatinized electrolyte liquid was another form of so-called “dry” electrolyte material disclosed in the same Swiss patent; that an electrolyte brought “to the ■consistency of a thick paste or jelly” characterized the battery of Cox patent No. 350,294, dated October 5, 1886; and that an electrolyte in the form of a paste characterized the battery, having spirally wound plate and separator elements, of the Crow■dus patent No. 1,173,651, dated February 29, 1916. The plaintiffs’ answer, and the court is •of the opinion that it is the correct one, is that in every instance referred to the ■electrolyte proposed by the prior art will be ruinous, if used in a dry electrolytic •condenser structure. The defendant contends it was prior art to provide compact condensers characterized by spiralled metallic leaves and liquid absorbing spacers of nonconductive material; that closely spiralled metallic condenser leaves separated by moist strips “of cotton or other textile or absorbent material” characterized the Wurts patent No. 503,186, dated August 15, 1893; and that anyone desiring to build any kind of compact condenser which must contain absorbed moisture held adjacent its plates has his requisite structure prescribed in this patent. The Wurts device is not an electrolytic condenser. It is on the contrary an electrostatic condenser in which the dielectric is an oil absorbed in the textile spacers. In the opinion of the court the Wurts patent does not anticipate. The defendant lists a number of men who by their patents on other works tried to solve the problem which Ruben claims to have solved. These men are Roberts, who applied for his patent No. 1,650,740 on June 15, 1925; Slepian, who applied for his patent No. 1,744,014, on September 12, 1925; Nodon, who published his French article on May 25, 1926; Bush who applied for his patent No. 2,134,273 on October 1, 1926; Edelman, who filed his application for Patent No. 1,753,920 on January 26, 1927; Miller, who on March 5, 1927, filed his application for patent No. 1,672,899, and on April 9, 1927, filed his application for patent No. 1,703,680; and again Edelman, who on May 11, 1927, filed his application for Patent No. 1,658,976. Concerning the Roberts patent No. 1,650,-740, it has been noted above that the only pertinent portion of the Roberts specification is that which states that an electrolytic condenser can be made with a diaphragm of a film-forming metal in contact with an electrolytic filling which may be in the form of a liquid or a paste or jelly; that the patent will be searched in vain for any disclosure as to how to compound the liquid or paste or jelly which is to serve as the electrolytic filling; that a structure which would be compatible with a liquid electrolyte would be useless with a paste or jelly electrolyte and vice versa; and that jelly electrolytes produced with gelatin or starch or water glass are so poor as to be useless. Dr. Slepian, the patentee in patent No. 1,744,014, testified in the case at bar. He received his bachelor’s, master’s and doctor’s degrees at Harvard, where he majored in mathematics and physics. He studied abroad in Germany and France for two years and after teaching mathematics at Cornell University for two years began work in the Research Engineering Department at Westinghouse Electric & Manufacturing Company in the summer of 1916. In 1920 he was placed in charge of the development of electrolytic condensers and worked continuously on that problem until about 1926. They were of the wet electrolytic' type utilizing electrolytes of various kinds, among which he preferred a concentrated aqueous solution of borax and boric acid similar to that which was used in Mershon wet electrolytic condensers. He made and tested hundreds of small wet electrolytic condensers to determine the effects of changes in the electrolyte and to study their performance. Dr. Slepian stated that the reason why a current blocking film could be formed electrolytically on an aluminum surface in a film-forming electrolyte was rather mysterious and there was no generally accepted theory to explain the existence of such a film and its current obstructing effect. Granting the existence and obstructing effect of such a film, however, the capacitance effect would follow at once from usual electromagnetic theory; and the capacitance measurements indicate that the film must be of the order of a millionth of a centimeter in thickness. Dr. Slepian was familiar with the Zimmerman patent 1,074,231, the Peek patent 1,008,860, and the Chubb patent 1,270,784, all of which relate to wet electrolytic condensers with various electrolytes. His work with wet electrolytic condensers taught him that extreme cleanliness and purity of materials were very important, that there was a gradual deterioration of the cell with time; that on standing idle there was a deterioration which was not wholly recoverable when the cell was put back on circuit; that the liquid electrolyte evaporated and the loss had to be made up with distilled water despite special precautions to prevent evaporation, and that contact of the film on the aluminum electrode with any solid foreign body would produce sparks and usually rapid corrosion at the point of contact. There had been an insistent demand for apparatus to enable the A-power and the B-power for radio sets to be obtained from the ordinary house lighting circuit since about 1920. When he tackled the problem B-eliminators were on the market and were reasonably satisfactory, utilizing rectifiers and suitable filters containing paper condensers. The development of the device of his patent No. 1,744,014 was the result of the need by the Radio Engineering Department at the Westinghouse Company for a condenser of very large capacity for smoothing out the ripples in the filament supply for the radio tubes in order to prevent the hum which would otherwise be produced. Dr. Slepian testified: “I think it was generally recognized that the success of radio broadcasting or the wide use of radio receiving sets depended upon some satisfactory solution of this A-power supply problem. After studying the problem, I proposed the use of a polarization cell to function as the condenser in this application. I considered the wet aluminum cell but knowing the extreme sensitivity of’ the aluminum, the necessity for extreme purity and cleanliness and particularly the necessity of avoiding contact with foreign materials, I thought it rather an uncertain question as to whether an aluminum electrolytic condenser could be built that would be sufficiently trouble free in the compact form that would be required for radio sets. On the other hand, the polarization cell, which I described in my Patent No. 1,744,014, is quite insensitive to impurities and in such a cell there is no great difficulty about using porous substances between the polarization electrodes to hold the necessary moisture to keep up electrolytic conductivity.” There was no difficulty in understanding the electrical requirements for an A-power supply and the necessary sizes of the condensers could be calculated. The other component parts were readily available. The only thing lacking was a condenser of sufficiently large capacity and sufficiently small size to be practical, although the rectifier perhaps needed some further development. Dr. Slepian described the polarization cell condenser much as it is described in his patent, namely, iron plates separated by paper and soaked in an alkaline solution, then sealed up in wax. The glycerin was added to the aqueous solution of sodium hydroxide only for the purpose of reducing the evaporation and he did not detect any other effect of the glycerin. Dr. Slepian outlined the generally accepted theory as to why the polarization cells exhibit a condenser effect and pointed out that his patent differentiated between electrolytic condensers of the film-forming type (wet electrolytic condensers) and polarization cell condensers, and said it was proper to say that the two are “different breeds of cats”. Dr. Slepian further testified: “D.Q.113. Dr. Slepian, it has been testified in a previous litigation that dry electrolytic condensers, made in accordance with the disclosure of the Ruben Patent No. 1,710,073 are capable of tolerating amounts of impurities which would be ruinous to wet electrolytic condensers and that the close spacing of the aluminum foil electrodes and the direct contact of the spacer material with the dielectric film formed on the aluminum is not injurious and I believe that these perhaps are in some way referable to the use of a viscous, pasty electrolyte made with glycerin. Did your experience with wet electrolytic condensers give you any inkling that such results were possible? A. No. “D.Q.114. Is the achievement of such results any cause of surprise to you? A. Yes, I am astonished at the statement. “D.Q.115. What leads you to feel astonishment at that statement ? A. All my experience with aluminum electrolytic condensers indicated to me that the aluminum cell was extremely sensitive to impurities and contact with foreign bodies. “D.Q.116. Did any of. your experience lead you to think that there was any way of making aluminum film forming electrolytic condensers that would minimize those difficulties? A. No. I did not know of any obvious way or any likely way. “D.Q.117. Is that the reason you turned to the polarization cell condenser when you were asked to develop the device shown in your Patent, Plaintiff’s Exhibit No. 4? A. Yes, precisely.” Dr. Slepian identified himself as the patentee of patent No. 1,543,729 which is in evidence in the case at bar. He summarized the disclosure of that patent as the addition of a small amount of protective colloid to the electrolyte of a wet electrolytic condenser, of a type similar to the condensers of the Peek and Zimmerman patents. The protective colloids mentioned in the patent, p. 1, lines 53 to 64, are gelatin, agar-agar, albumen, gum arabic or the like. Dr. Slepian stated that nothing in this patent ever led him to the development of a dry electrolytic condenser of the aluminum film-forming type. Dr. Slepian’s polarization cell condenser never went into commercial use in an A-battery eliminator because of the advent of the A.C. tube, nor has it ever found any other application of any kind. The critical voltage for a polarization cell with an iron electrode and sodium hydroxide is in the neighborhood of one volt. If this decomposition voltage is exceeded, any condenser effect is practically absent. The theory of such polarization cells indicates that high voltages cannot be withstood by a single cell whereas the aluminum electrolytic condenser can withstand several hundred volts. A considerable portion of the time devoted to the trial of this case and all of the time devoted to the supplemental hearing which was had was given over to the receiving of evidence bearing on the Nodon French article. The question with which the parties concern themselves is, Does the Nodon article disclose a condenser? Prior to the trial certain tests were conducted, under the direction of counsel for the plaintiffs, which tests counsel for the defendant was invited to attend. Further tests were made in open court during the trial. On the trial and in its brief the defendant contended that, because an oxide film had not been preformed on the anodes of the devices tested, the devices did not conform to the disclosures of the Nodon article and that consequently the tests had no evidentiary value. In its brief filed after the trial but before oral argument the defendant said: “Your Honor may rest assured that the plaintiffs would have tested a Nodon condenser, after providing it with the customary oxide film on the anode, had they not been certain that the presence of such film would have insured its creditable performance.” After the oral argument the plaintiffs moved to open up the case so that the parties might have opportunity to make further tests and report their results to the court. The case was opened; further tests were made of devices in which oxide films had been preformed on the anodes; these tests, which were made both out of court and in court, were attended by counsel and expert of the defendant, as well as by counsel and experts of the plaintiffs; and supplemental briefs have been filed since the supplemental hearing. The court has arrived at some conclusions in respect of the disclosures of the Nodon article. In the court’s opinion, the Nodon article neither requires nor permits the preforming of an oxide film on the ano.de. But whether it does or not makes no difference. The tests of the devices, without the preformed oxide film on the anode and with the preformed oxide film on the anode, all demonstrate to the satisfaction of the court that the Nodon article does not disclose a condenser. Evidence of actual tests on devices constructed in accordance with the Nodon disclosure (that is, without a preformed oxide film on the anode) was presented on the trial. The court was shown the comparative effect of a condenser made in literal accordance with the precise disclosure of the Ruben patent No. 1,710,073, a Nodon device, and a dry electrolytic condenser of current production of P. R. Mallory & Co., Inc., one of the plaintiffs, on the hum in an old radio set powered with an old Majestic A-eliminator. The device made in accordance with the disclosure of the Ruben patent No. 1,710,073 eliminated the hum in the same way as did the up-to-date dry electrolytic condenser manufactured by the Mallory Company. The Nodon device produced no sensible effect so far as hum elimination was concerned. For use in an A-eliminator, the Ruben device of April, 1929, was eminently satisfactory, and as effective as the dry electrolytic condenser of 1940. The Nodon device, by contrast, was useless. The next tests by means of a cathode ray oscillograph furnished the scientific explanation as to why the Ruben device was effective in eliminating hum and the Nodon device was ineffective. The output of the rectifier in the A-eliminator was shown to produce pulsating direct current which varied in voltage from minimum to maximum about 9 volts. With just one Ruben device across the rectifier output the wave form was changed in such a way that the variation from minimum to maximum was reduced to only 2 volts. With the Nodon device across the rectifier output however, the variation from minimum to maximum was still between 7 and 7% volts. And further tests showed that the same effect on the wave form of the rectifier output was produced by a resistance in series with a small battery of 4 volts. These tests demonstrated why the Ruben device was effective in solving the problem of eliminating the A-battery from the radio set and why the Nodon device was incapable of accomplishing such result. Further tests showed the characteristic effect of a standard electrostatic condenser in displacing the current wave of an alternating current of 60 cycle frequency with respect to the voltage wave, a phase displacement of approximately 90°, and the characteristic effect of a pure resistance, without any capacity effect, no phase displacement. The tests then showed that a condenser made in accordance with the disclosure of the Ruben patent No. 1,710,-073 produced the same sort of phase displacement as is produced by the true electrostatic condenser, approximately 90°: The Nodon device on the other hand produced the result, no phase displacement, of a pure resistance. The same results were obtained when the test was made with alternating current of much higher frequency, 500 cycles. Further tests showed the effect of varying the frequency of alternating current on the impedance of the Ruben condenser and the Nodon device. With a true electrostatic condenser the impedance varies with the frequency of the alternating current applied and the values obtained when plotted on log-log paper give a sloping line with a downward inclination of 45° from left to right. The impedance of a pure resistance however does not vary with the frequency, and the values observed when plotted on the same log-log paper give a horizontal line. The Ruben condenser gave values that produce the sloping line characteristic of a true electrostatic condenser. The Nodon device gave values which produce the substantially horizontal line characteristic of a pure resistance. Finally, the tests disclosed the power factor of the Ruben device and the Nodon device. The power factor is a measure of the amount of energy consumed in the device. Ruben’s device had a power factor around five to six per cent, whereas the Nodon device had a power factor of ninety-seven to ninety-nine per cent. The Nodon device consumed practically the entire energy imparted to it and did not act like a condenser that stores and restores energy from and to the circuit. Similar tests on devices made in accordance with the disclosures of patent No. 1,714,191, i. e., with an electrolyte composition consisting of glycerin, sodium bicarbonate and boric acid, gave the same characteristics as the condenser of the Ruben patent No. 1,710,073. Even when boric acid was omitted, and the electrolyte composition contained only glycerin and sodium bicarbonate, the characteristic condenser action was produced. The only objection that the defendant made to these tests was that first voiced on the trial that the devices tested did not have preformed oxide films on the anodes. As has been indicated, the court is of the opinion that the objection was not well founded. Furthermore, the court is of the opinion that these tests demonstrated that Nodon had nothing more than a resistance with a slight battery effect and that the Nodon device was incapable of solving the problem which Ruben solved with his dry electrolytic condenser. After the order reopening the case, a number of tests were conducted. The D. C. Leakage Test. One of the recognized properties of any electrolytic condenser is its ability to produce a dielectric film on the anode when it is subjected to direct current, even though tire aluminum electrode is not provided with any preformed film. In the original tests the Ruben samples and the Nodon samples for direct current tests had both been prepared without preformed films, yet the Ruben samples exhibited the properties of a true condenser while the Nodon samples showed no discernable condenser effects. In other words, during operation of the devices the Ruben electrolyte formed and maintained a film on the anode, while the Nodon glycerin and ferric oxide mixture did not. In the opinion of the court this was proof that Nodon did not provide a film-forming electrolyte. For the reopened hearing the tests went much further. There were prepared Ruben samples and modified Nodon samples with 12-volt preformed films on the anodes and there was observed the change of leakage with time in both of these at 8-volts D. C., a voltage well within the capabilities of the 12-volt preformed films. The results are in evidence. The initial leakage of the modified Nodon device was very much higher than the initial leakage of the Ruben device, a significant factor since both devices were provided with the same 12-volt preformed film. With the passage of time the leakage of the Ruben device decreased rapidly to a very low minimum, showing that any defects in the preformed film were rapidly repaired by the film forming electrolyte in the Ruben condenser. The modified Nodon device however behaved in just the opposite way. Its leakage increased with time to a considerably higher value, indicating that the preformed film was losing any effectiveness it originally had. This test shows that the modified Nodon device is not an electrolytic condenser, since the Nodon paste instead of forming or maintaining the preformed oxide film (as would be the case with any electrolytic condenser) actually brings about the destruction of any preformed film. A-Eliminator Tests. Input, output and ripple voltages. Modified Nodon devices were prepared and tested as to their ability to “filter” the output of a rectifier such as is used in an A-eliminator. To get an output of six volts using the modified Nodon device, an input of 12.6 volts was required as against 12.9 volts for the Nodon device without preformed films and 9'.9 to 10 volts for the Ruben condensers. With the true Nodon devices a ripple voltage of 1.3 volts remained; with the modified No-don devices the ripple voltage was 1.15 volts, whereas with the Ruben condensers the ripple voltage was only .25 to .3 volts. With the modified Nodons, as with the pievious ones, there was no tendency to produce the efficient filter action of the Ruben condensers. Effect on rectified wave form. The effect of the modified Nodon devices on the wave form of a rectifier output was determined. A photograph in evidence shows the wave form of the rectifier output without any filter device. A second photograph in evidence shows the effect of the original Nodon device on this wave form. A third photograph in evidence shows the effect of the modified Nodon device on this wave form when initially connected; and a fourth photograph in evidence shows the effect of the same modified Nodon device after 15 minutes of operation. With no filter device the rectifier output has a peak to peak ripple of 9 volts. The true Nodon device gave a ripple of 7.4 volts. The modified Nodon device gave a ripple of 6 volts when initially connected, which increased to 6.8 volts after 15 minutes operation, thus indicating that the modified Nodon device tends to approach the original Nodon device over a period of time of operation. This confirms what might be expected from the leakage test, since the Nodon paste apparently tends to destroy the effect of any preformed film. By way of contrast, the Ruben device gave a ripple of only 2.1 volts. Moreover, as shown by a photograph in evidence 17, the Ruben condenser charges to the peak voltage with each wave and gives a sloping trace from the peak of each wave to the intersection with the next wave. Other photographs in evidence show, on the other hand, that the Nodon devices, whether with or without preformed oxide films, do not charge a peak voltage and do not exhibit any condenser action. Hum test in open court. On the basis of the preceding tests, the electrical engineer who conducted the tests concluded that the modified Nodon device would not be useful for hum reduction in an A-eliminator, and demonstrated the correctness of his conclusion by actual test with a radio set. The hum was very loud with the modified Nodon devices in circuit, practically identical with the hum when the true Nodon devices were in circuit. When the Ruben devices were cut in, the hum was reduced to a very low level which did not interfere with radio reception. Alternating Current Tests. Phase displacement. The electrical engineer who conducted the tests also paralleled his alternating current tests of the true Nodon devices with similar tests of the modified Nodon devices provided with preformed oxide films on both plates. At 60 cycles the modified Nodon device produced some phase displacement, approximately 30°, with a power factor of 86%. At 500 cycles the same device exhibited little if any phase displacement and a power factor of at least 98%. Phase displacement with sesquioxide of iron omitted. To find out what the phase displacement at 60 cycles was due to, the electrical engineer who conducted the tests prepared another device, which was like the modified Nodon device tested, except that the sesquioxide of iron was omitted and the gauze spacer impregnated only with chemically pure glycerin. The phase displacement produced by the device without the sesquioxide of iron at 60 cycles was approximately 37y2°, with a power factor of 79%. At 500 cycles it produced the same results as the modified Nodon device. Obviously the phase displacement produced by the modified Nodon device at 60 cycles was due to the preformed film and not to any effect of the colloidal sesquioxide of iron in the Nodon paste. The tests showed that a device with preformed films made without Nodon’s sesquioxide of iron is better than the device described by Nodon; but neither device approaches the operation of a condenser. Impedance-frequency relation. The electrical engineer who conducted the tests also determined the impedance-frequency relation of the modified Nodon device and the similar device without sesquioxide of iron at frequencies of 60, 330, 500 and 690 cycles. The results are plotted on an exhibit in evidence in comparison with the impedance-frequency relation of a true condenser and a true resistor. These graphs show indeterminate hybrid characteristics of neither a true resistor nor a condenser, but much more closely resembling a resistor than a condenser. Some slight capacity effects are noticeable at lower frequencies but these are greater when Nodon’s sesquioxide of iron is omitted. The defendant made no objection to these tests but on the supplemental hearing for the first. time complained “that plaintiffs did not subject the ‘Nodon Devices’ to the fundamental ballistic galvanometer test.” The answer to this objection is that the question under investigation was whether the Nodon device was a condenser at all. The broad property of storing electricity and restoring some of it to a circuit does not differentiate between a condenser and a storage battery or polarization cell. In the opinion of the court these tests which have been detailed show that the Nodon device, with or without an oxide film on the aluminum electrode, is not an electrolytic condenser and will not function as such. The Bush application discloses a device intended for use in the filter circuit of an A-eliminator. It comprises two sheets of a metal such as aluminum or nickel, but preferably nickel silicon, interleaved with a porous substance such as asbestos or blotting-paper which is soaked with a hygroscopic or deliquescent liquid electrolyte such as potassium acetate. It is designed to operate at about 1.5 volts per unit and the application shows four units connected in series so as to be capable of withstanding the low voltage of the filter circuit. One Marshall testified in the case at bar that Dr. Vannevar Bush is an outstanding scientist in the electrical field; that in 1922, when the Raytheon Manufacturing Company was organized, Dr. Bush was a professor of electrical engineering at Massachusetts Institute of Technology; that since then he has held various responsible positions in technology, including Dean of Engineering and vice-president of the institution; that he is now president of the Carnegie Institute at Washington, one of the outstanding scientific institutions in this country; that he is also Chairman of the Research Committee of the National Defense Council; that he was consulting engineer for the Raytheon Manufacturing Company from the time of its organization until he became vice-president of the Massachusetts Institute of Technology; that he was such engineer for the Raytheon Company in 1926 when he filed the patent application in question; and that none of the devices described in the Bush patent were ever sold on a commercial scale. The evidence further shows that any liquid electrolyte is incompatible with closely spaced electrodes and the Bush device is subject to this drawback; and that although the Raytheon Manufacturing Company was able to supply the Grigsby-Grunow-Hinds Company with satisfactory rectifying tubes which went into 750,000 B-eliminators, and that company was looking anxiously for an A-condenser, it did not turn to the Bush device. The Edelman application filed January 26, 1927 was too late to antedate Ruben. The plaintiffs contend and the defendant does not deny that for the purposes of this case the effective date for the Ruben patent No. 1,714,191 may be taken as December 20, 1926, the date on which he executed the application for that patent; that the application for the Ruben patent No. 1,710,073 was executed on March 18, 1927, but the memorandum which was prepared by Mr. Ruben on January 3, 1927, shows that he was in possession of the information contained in his second application at least as early as the date of the memorandum; and that, accordingly, for the purposes of this case, January 3, 1927, may be taken as the effective date for Patent No. 1,710,073. Furthermore, the evidence shows that the Edelman device resembles “a rock in harness,” and contains no glycerin or any equivalent of it; and that while it went into commercial production it is now obsolescent if not obsolete. The Miller application filed March 5, 1927 was too late to antedate Ruben whose effective dates, as has been observed, were December 20, 1926, and January 3, 1927. The Miller patent teaches that three parts of gelatin should be dissolved in one part of glycerin and that the resulting mixture should then be saturated with sodium bicarbonate or other salts such as ammonium borate. The undisputed testimony shows that when the electrolyte liquid of a standard wet electrolytic condenser is gelatinized with gelatin, the units so made up are “failures from the start” in that it is hard to impress voltage above 5 or 10 volts and the leakage may be 30 or 40 milliamps. No commercial use of the Miller device has been shown. The second Miller application filed April 9, 1927, later than the Ruben effective dates, retains the gelatin-glycerin-sodium bicarbonate mixture but reduces the relative quantity of gelatin and adds a small amount of lithium hydroxide. There is no evidence of commercial use of this device. The second Edelman application filed May 11, 1927, later than the Ruben effective dates, discloses a device substantially the same as the device of the earlier Edelman patent, rock-hard and containing no glycerin. The evidence shows that this device is obsolescent if not obsolete. The court has now considered all the references that were referred to by the defendant in its brief. Additional allegedly prior art was introduced on the trial but has not since been referred to. The court is of the opinion that the Ruben invention is a new combination of aluminum plates closely spaced and separated by porous spacers impregnated with a viscous pasty composition containing glycerin and a film-forming electrolyte, which may be rolled up in a compact spiral form and which is free from the drawbacks of spillage and evaporation. This combination is not found in Nodon or any other prior art reference. If commercial success were needed to resolve a doubt (which does not exist in the court’s mind) in favor of validity, the record discloses that commercial success. The evidence shows that through 1940 over $21,000,000 of dry electrolytic condensers have been manufactured under sub-license under the patents in suit, and that royalties paid up to June 30, 1936, amounted to $460,000, plus $125,000 paid as consideration for granting the sub-licenses, plus $35,000 in royalties paid by the Grigsby-Grunow-Hinds Company. Royalties paid since June 30, 1936, have amounted to $180,000. These total $800,000 in all. Evidence has been presented showing the materials used, the type of construction involved, and the nature of the electrolyte composition contained in condensers manufactured by the various licensees under the patents in suit. Generalizing for the moment, this evidence shows that all such condensers have been made with the use of two electrodes of aluminum, at least one of which has been electrolytically formed with a dielectric film, the electrodes being interleaved with porous spacers either of gauze or paper which are impregnated with a viscous pasty electrolyte composition made with the ingredients glycerin or ethylene glycol, and ammonium borate plus boric acid, or ammonium borate alone, the whole unit being in rolled form and enclosed in a suitable container either of cardboard or metal. It appears in the instant case that glycerin and ethylene glycol are full equivalents for the purposes in question. Each is a hygroscopic and viscous liquid. Each has a high dielectric constant of about the same order of magnitude. Each reacts with sodium borate or ammonium borate in the same way to form sodium or ammonium glycerol or glycol borate. The finished electrolyte compositions made with glycerin and ethylene glycol are practically indistinguishable except for certain very fine electrical measurements or chemical analyses. The evidence shows that the combinations claimed by the patents in suit have been utilized in all dry electrolytic condensers made under the licenses granted by the Mallory Company. The evidence with regard to other patents which are included in the licenses shows that the patents in suit and the glycol patent No. 1,891,207 are the only ones which have been followed by ' all manufacturers of dry electrolytic condensers. Some few manufacturers have taken advantage of a few of the additional patents, but none of these are essential to manufacture of satisfactory commercial devices. No extra royalty has been charged for these additional patents; and there was no diminution of the royalty when the glycol patent No. 1,891,207 was held invalid. Sprague Specialties Company was a licensee. Specimens of its condensers are in evidence. A specimen of the Sprague electrolyte paste is in evidence. Comparison with a specimen of electrolyte paste in exact accordance with the preferred formula of the Ruben patent No. 1,710,073, shows that they are almost identical in appearance and physical condition. The Sprague paste contains ethylene glycol and ammonium penta borate. Over $6,300,000 of dry electrolytic condensers have been sold by Sprague since it acquired the license under the Ruben patents, all of them have been marked under the patents in suit, and royalties have been paid on all of them. The Magnavox Company Incorporated is' a licensee. It produces dry electrolytic condensers of which specimens are in evidence. They are made up with a preformed anode, an unformed cathode, both of aluminum, a spacer of porous material impregnated with a viscous pasty film-forming electrolyte, all of which is wound into a form of a roll and packaged in an aluminum can, card-board box or cardboard cylinder. The electrolyte ingredients are boric acid, ethylene glycol and ammonium hydroxide. The ammonium hydroxide combines with the boric acid to make ammonium borate. A specimen of the electrolyte paste is in evidence and on comparison is seen to be closely similar in appearance and consistency to a paste made in accordance with the preferred formula of Ruben patent No. 1,710,073. The Magnavox Company has sold approximately $3,500,000 worth of dry electrolytic condensers since 1932, all of which have been marked with the numbers of the patents in suit and on all of which royalties have been paid to the Mallory Company. Mallory dry electrolytic condensers are in evidence. A specimen of a Mallory electrolyte composition is in evidence, and, on comparison, is seen to be closely similar to a paste made in accordance with the preferred formula of Ruben patent No. 1,-710,073 in appearance and consistency. The ingredients are boric acid, ethylene glycol and ammonium hydroxide. The Mallory condensers are made up with two aluminum electrodes, one of which has an oxide film formed on it, interleaved with porous spacers, either gauze or paper, the spacers being impregnated with the electrolyte composition by pasting, wet winding or impregnating by dipping in the heated electrolyte composition. Condensers made by other licensees have contained a hygroscopic material such as glycerin or ethylene glycol with ammonium borate, or borax with boric acid, or ammonium borate with boric acid. Electrically, they have the same value of capacitance per unit area at the given voltage. They are all of low power factor due to the closely spaced electrodes. They will main • tain their capacitance whether in use or' not in use, which was an inherent property) of Ruben’s dry electrolytic condensers, due. to the use of a pasty electrolyte. This applies to condensers made by Mallory, Sprague, Magnavox, Aerovox, CornellDublilier, Micamold and Solar. Ruben said in his patent: “The particular film forming electrolyte employed may be any one of a large number of well-known electrolytes, but I have found a mixture of boric acid and sodium borate to be especially satisfactory.” Ammonium borate is such an electrolyte and the equivalent of the sodium borate and boric acid specifically mentioned in the Ruben patent. Ethylene glycol has been shown to be the equivalent of glycerin in respect of its hygroscopicity, viscosity, high dielectric constant, and its ability to react in the same way with sodium borate and with ammonium borate. The evidence shows that all commercially practical dry electrolytic condensers on the market today stem directly from and embody the Ruben inventions of the patents in suit. So far as the art knows today, there is no way of making a commercially practical dry electrolytic condenser except by following the fundamental teachings of the Ruben patents. The defendant manufactured dry electrolytic condensers comprising two pieces of aluminum foil, one having an oxide coating applied in an electrolytic bath, interleaved with gauze spacers, rolled up and impregnated with a mixture of boric acid and ammonium hydroxide in glycerin. After the impregnation with the electrolyte composition the units were provided with terminals and enclosed in suitable containers. The boric acid and ammonium hydroxide used as starting materials combined to form ammonium borate, leaving an excess of boric acid. The ammonium borate and the excess of boric acid combined with the glycerin to form ammonium glycerol borate and glycerol borate. There was some free (uncombined) glycerin in the defendant’s electrolyte composition. Borax (sodium borate) and boric acid when mixed with glycerin immediately and inevitably combine with the glycerin to form sodium glycerol borate and glycerol borate. The defendant used to some extent another formula for its electrolyte composition in which a small part of the ammonium •borate was replaced with ammonium phos“phate and sodium phosphate, both of which rare in themselves film-forming materials. >-The electrolyte composition containing phosphates was found to be less satisfactory than the one containing borate alone and was eventually abandoned. The court, for purposes of study of questions of infringement, having set down on paper each of the claims in suit, together with the contentions of the parties in respect of each claim and finally the conclusions of the court on the question of infringement of each claim, realizes that all of the claims in suit might, if one had the time, very well be considered en masse. But the court does not have time to be brief. Accordingly, the claims, and the contentions of the parties in respect of infringement thereof, will be set forth, as they have been studied, separately, together with the conclusion of the court as to the infringement of each claim. Claim 4 of patent No. 1,714,191 is as follows: “4. An electrical condenser comprising film forming electrodes separated by a paste electrolyte comprising a mixture of an hygroscopic and viscous material and a neutral salt, said material to afford a paste like mixture with said neutral salt and to maintain the water content of the electrolyte subs