Appellants brought suit for specific performance to require the France Manufacturing Company to carry out its obligations under a patent license agreement; to account for its operations thereunder; and to pay royalties required by its terms. They also asked for a declaratory judgment determining whether the above mentioned patent license agreement had been canceled by appellee; and, in case the court found that the contract had been canceled, then, they sought a decree enjoining appellee from infringing appellants' patents by use, sale, or offers of sale, and for an accounting of profits resulting from the infringement.

Appellee contended that the license agreement had been canceled; that the agreement was illegal as being contrary to public policy, in violation of the Sherman Anti-Trust Act, 15 U.S.C.A. §§ 1-7, 15 note; that appellee had been guilty of no infringement; and that the patents embraced in the license agreement were invalid for anticipation, and for want of invention. The district court, in an opinion, and in findings of fact and conclusions of law, held that the license agreement was void for ambiguity and failure to reflect a meeting of the minds of the parties; that if it were not ambiguous, it had been canceled as claimed by appellee; that it was violative of the Sherman Anti-Trust Act; that there had been no infringement; and that the patents in suit were invalid for anticipation and want of invention.

We shall discuss, in order, the issue of the validity of the five patents in suit; the question whether the license contract was void for ambiguity; the claim of appellee that the license agreement was canceled; the contention that the license agreement violated the Sherman Anti-Trust Act because of price-fixing provisions; the right of appellee, who was a licensee under the contract, to attack the validity of the patents; and the question of damages.

The patents in suit cover: a luminescent or fluorescent gaseous discharge tube lighting system, employing hot cathode fluorescent tubes, in which system provision is made for short-circuiting the flilamentary cathode of the tubes; such a system, in which the tubes are positioned in pairs of sockets and connected in circuit with the operating transformer primary and secondary windings, so that removal of a tube will disconnect the primary winding from the supply line, and will also disconnect the secondary winding; an improved fluorescent tube lighting system and an improved power unit; a form of fluorescent tube lighting system and apparatus; and the complete fixture including tubes, reflector, sockets, power unit, and necessary connections.1

The most important contribution, according to appellants, which their patents provide, consists of a fluorescent tube lighting system which will start instantly upon being switched on, in contrast with prior systems where the lighting of the tubes occurred only after several seconds' elapse from the time they were switched on. It is, moreover, submitted that appellants' patents have, in addition, accompanying advantages, including the elimination of starting switches, less and cheaper maintenance, operation at lower room, or outdoor, temperature, and the absence of stroboscopic effect and flicker, as well as safety in operation, through the use of low voltages, together with the advantage that when one of a pair of tubes is disconnected or removed from its socket, the circuit is broken, and danger of electric shock when changing tubes is greatly lessened.

The subject matter of the patents in issue, therefore, involves a fluorescent or luminescent tube lighting system and apparatus.

The fluorescent or luminescent tube now in general use is a gaseous discharge tube consisting of a glass tubular envelope, with its interior surface coated with fluorescent salts, filled with an inert gas under a certain pressure, and with a pair of electrodes, one in each end of the tube. The electric current from an outside source passes through the gas from one electrode to the other. The inert gas used is argon, and a small amount of mercury is also placed in the tube. The mercury becomes vaporized when the discharge starts in the argon gas. Since mercury has a lower ionization voltage than argon, the mercury takes over the entire discharge. An electrode, as is well known, is a terminal of the electric source, or, it may be said, of the terminals of the conductor by which the current leaves, and by which it enters the tube. Electrodes may be composed of wire filaments, plates, or other conducting objects.

Electrodes are designated either anodes or cathodes. An anode is the electrode through which the current passes to the cathode. In alternating current, which is used with fluorescent lighting, the direction of the flow of the electric current is periodically reversed, so that the electrodes at each end of the tube exchange the roles of anode and cathode with each alternation of the current with which they are supplied.

A description of the way in which the fluorescent tube lights or furnishes illumination requires some discussion of electrical phenomena. It is the theory, or working hypothesis, of science that all matter is composed of atoms; that all atoms contain a nucleus and one, or many electrons; that an electron revolves about the nucleus of the atom similar to the way of heavenly bodies in the planetary system. Different substances have either a positive or negative charge. Substances with like charges repel, while those with unlike charges attract each other. An electron is practically a pure negative charge of electricity. The anode is the positive electrode through which the electric current, on its way to the other pole, or cathode, enters the electrolyte, which, in the case of the fluorescent tube, is the inert gas within the tube. Since the anode is the positive pole or terminal, the electrons, being negative charges of electricity, are attracted to the anode and flow through the tube toward it from the cathode during the time when an electric current is passing between the electrodes; and particles, called ions, with a positive charge, are attracted to the negative cathode.

As far as this case is concerned, there may be said to be two types of operating fluorescent tube lighting, the hot cathode, and the cold cathode. Appellants' patents operate as cold cathode, and they refer to the hot cathode operation as the prior practice, in order to demonstrate how they differ from it and what contribution they have made to the art.

We shall, then, briefly describe the hot cathode fluorescent lamp, how it operates, and what happens in its operation; and compare it with cold cathode starting.

In a hot cathode fluorescent lamp, the two ends of each electrode are connected to the pins in the bases of the lamp. When the circuit switch is closed, current flows though the electrodes. The current warms the electrodes and lowers the required starting voltage. A bimetallicthermal automatic starter or switch is inserted in the wire connecting the electrodes. After a few seconds, the circuit is automatically broken at the starter. A surge of current at somewhat greater than the line voltage, which has meanwhile been built up by the ballast,2 is then sufficient to leap the gap and start a discharge through the argon. The discharge is quickly taken over by the mercury. After the stage of warming or preheating the electrodes, above mentioned, the starter automatically opens, and remains open as long as the lamp is operating satisfactorily. If the lamp does not light the first time, the starter circuit repeats its cycle until the lamp does light.

In hot cathode operation, when there is a surge of voltage between the electrodes in the fluorescent tube, certain activity takes place: The electrodes — in this case, wire filaments — are coated with a special material consisting of oxides which emit electrons copiously when heated. The electric current causes the electrodes to become heated and, accordingly, to emit electrons which become accelerated and thrown from the cathode along the length of the tube toward the opposite electrode. With alternating current, the electrons travel backward and forward throughout the tube, and this continual reversal in direction is due to the fact that each electrode becomes anode and cathode in turn, as the current alternates. As billions of the electrons are so accelerated, they come into collision with the atoms of the inert gas in the tube. Such an atom, if struck hard enough, will have one of its electrons knocked out of its system, which may be compared with a planetary system. The atom then has one less negative charge because of the loss of an electron and, as a result, has a net positive charge equal to the amount of the negative charge which it lost by the collision. However, the collision with the atom may not be sufficiently powerful actually to knock an electron completely out of its connection with the nucleus, but, instead, may cause it to jump out of the orbit in which it naturally revolves and ascend into a higher orbit. Yet, if this be the case, such a condition is so unstable that within a fraction of a millionth of a second, the electron which had jumped into the higher orbit falls back to the original normal unexcited phase.3 Whatever the case, whether the electron has been knocked out of its planetary system or whether it has jumped into a higher orbit and back again, the event is accompanied by a radiation of light. This light may not be apparent in the fluorescent tube since the pressure, current density, and other considerations may affect the way in which the atoms are excited, and the light which is thereby produced may be in the ultraviolet range, that is, the so-called black light, invisible to the eye. However, the black light which is created by the innumerable collisions heretofore mentioned causes a radiation outward from the inside of the tube. In such radiation, the black light falls upon the fluorescent salts with which the inside of the tube is coated. When the black light falls on the fluorescent salts, they, in turn, become excited and radiate light, and in so doing, the fluorescent salts radiate light in the visible spectrum. In the operation of hot cathode, the discharge of electrical energy at the electrodes commences as a glow which develops when the current begins to heat the electrode and, as it continues, becomes a tiny "hot spot," a term generally used in the art for many years to describe the phenomenon.

The hot spot of the hot cathode operation, raising the temperature of the electrode to a high degree, causes the emission of electrons from the coating of the filament of the electrode. The electrons which are produced at the opposing filaments at each end of the tube permit the flow of current from electrode to electrode; and it is that flow of current that provides the illumination.

In cold cathode operation, the starting results, not from a hot electrode, but from the impressed voltage between the filamentary electrodes sufficiently high to strike an arc between them, causing ionic bombardment of the filamentary coating of the electrodes, and the knocking from that coating of sufficient electrons to permit the flow of the operating current; and such starting is achieved almost instantaneously as compared with the time requirement of the warming operation to preheat the electrodes of hot cathode starting which required several seconds. In cold cathode operation, as contrasted with hot cathode, no hot spot develops on the electrode; the discharge at the electrodes remains a glow. The glow in the cold cathode device results from the cloud of electrons and ions which come entirely from the gaseous medium itself, rather than being emitted from the electrodes at high temperature.

Appellants contend that, in their patents, an entirely different system of instant starting of fluorescent tubes is accomplished as compared with the prior art and practice. Although appellants use the so-called preheat hot cathode lamps, with the two ends of each electrode connected to the two pins in the bases of the lamp, the electrodes are not preheated. No current flows through the filament because each filament is short-circuited. Because of this fact, the conventional prior art of hot cathode starting can, of course, not operate with such a short-circuiting. Appellants' expert witness, Mr. Kelley, testified that the advantages of short-circuiting the terminal pins of the lamp or socket were that, in the instant start, with high impressed voltage, the "current is enabled to leave the lamp by means of both pins, so that more of the cathode is used for the production of electrons for the continued operation of the lamp after it is lighted, and it results in a more efficient use of the cathode, and is particularly adapted to the instant start cold cathode operation of these lamps." He further stated that since the tube filament of the electrodes is short-circuited, it does not participate as a filament, in the operation of the tube, except for the production of electrons; and the failure of the filament has no effect on the operation of the tube itself.

Appellants' starting operation, as above described in its patents, was cold cathode starting. For such cold cathode starting, they used a hot cathode tube which, however, was adapted for cold cathode by short-circuiting the electrodes in each end of the tube.

However, cold cathode starting was old in the art.

In 1856, Geissler, a German artist and glass blower, originated the electric discharge tube.4 He discovered that when a high voltage alternating current was passed through a sealed tube containing air at low pressure, the tube gave off light of a low luminosity until the vacuum deteriorated. His further experiments, together with those of Faraday, Crookes, and others, indicated that all individual gases would give off a fairly strong light, nitrogen and carbon dioxide giving the best results. Mercury, sodium, sulphur, chlorides, bromides, and other vapors also produced light of various colors. As early as 1862, patents were granted in Great Britain for using colored light in Geissler tubes for buoy lighting. The first electric discharge lamp containing fluorescent materials was made in Paris by Becquerel and was described by him in a paper published in 1859. Becquerel used a Geissler discharge tube containing various luminescent solids in fragments or in powder form. He suggested the preparation of tubes or bulbs in which a thin coating of such material was stuck to the inner surface of the glass; but the original Becquerel lamp had very low efficiency and a short life, and much progress had to be made before commercial success could be achieved. However, in 1893, an American electrical journal reported that Geissler tubes were receiving the earnest attention of electrical experimenters with the hope that, in the illumination of the tubes, lay the secret of phosphorescent light. In 1896, Edison patented a vacuum tube coated on the inside with calcium tungstate and containing two widely separated electrodes. Cathode rays excited the fluorescent tungstate to produce a light which was fluorescent when the discharge was passed. But Edison chose the incandescent lamp as the one to develop for purposes of illumination, with the success that holds the world in debt; and it was left for others to work upon the problem of gaseous tube illumination.

The first commercial application of the Geissler tube was made about 1900 by Moore, a former Edison employee, who used tubes seven to nine feet long and two inches in diameter, with electrodes sealed at each end, and operated on high voltage to give what was later designated "instant start." The quality of the light was good but the tube life was short. This defect was remedied by an ingenious process in which, among other devices, an automatic valve permitted gas to flow into the tube when the pressure became too low, with resultant correct pressure and current flow. Moore later used tubes one hundred and eighty feet long with 16,000 volts on alternating current and step-up transformer. In a paper presented at a meeting of the American Institute of Electrical Engineers in 1907, Moore alluded to two prior papers he had presented on the subject of vacuum tube lighting in 1893 and 1898, and gave a detailed description of his process, stating, in the course of his remarks, that "A study of the history of the development of all light sources shows the vacuum-tube to be a logical conclusion of the progress of evolution. * * * Carefully tracing the evolution of all forms of light from remote ages down to the present time as regards the following seven most important factors: diffusiveness, color, steadiness, absence of heat, efficiency, safety and first cost, all — clearly prove this long tube system to be a natural sequence and that the times are now ripe for it." He mentioned further that installations had been made in department stores, theatre lobbies, and silk dyeing establishments; that artists were mixing their colors under the light of such tubes; and that it had been adopted by photographers, florists, and clothing merchants. He stated that "Aside from all theory, the facts show that electric illumination resulting from incandescent gases or vapors has vastly less limitations because their molecules or atoms are free to vibrate while those of solids are not, and therefore at the high temperatures necessary are subject to rupture. `The signs of the times' clearly indicate that we have reached the beginning of the end of lighting by solid conductors and the end of the beginning of lighting by gaseous-conductors." Moore's tubular lighting was so successful that it constituted a potential threat to the established business and heavy investment in incandescent lighting; but with Coolidge's development and introduction, in 1906, of tungsten filament lamps, the efficiencies of Moore's long tubes were no longer great enough to overcome their disadvantages of installation and auxiliary equipment over the improved incandescent lamps; and Moore's tubes, except for short carbon dioxide tubes for color matching, gradually disappeared from the market. However, Moore's tubes formed the basis for the later development of the lighting tubes brought out by George Claude, who demonstrated his first neon sign in 1910 at the Grand Palais in Paris. Appellee's expert witness, Dr. Hoover, pointed out that neon tubes can be either in the glow or arc state in operation; that, in starting, they can be either quick or slow; that the instant start was accomplished in the neon lamp by impressing voltage upon the tube sufficient to strike the arc, and that the original patent granted to Claude for his neon lamp was on the increase of the size of the electrode to such a point that the hot spot would not form; and the witness further stated that the fluorescent tubes that came upon the market prior to appellants' patents were merely modifications of the neon sign tubes. Dr. Hoover also pointed out that the Wiegand German patent, applied for in Germany in 1934, and which was subsequently granted an American patent No. 2,030,805 in 1936, disclosed a luminescent tube with a preheated hot cathode in one end of the tube and a cold electrode in the other; also, with two cold cathodes, one in each end, and a preheated hot cathode in the center of the tube; also, two cold sheet metal electrodes, one in each end, and a filamentary cathode which could not be preheated in the center of the tube; and, finally, two filamentary cathodes, one in each end of the tube, neither of which could be preheated, and a cold sheet metal electrode in the center of the tube. Fritze's German patent of 1937, granted a patent in the United States in the same year, patent No. 2,094,648, was a luminescent tube with a cold cathode in each end. The foregoing German patents thus disclosed numerous combinations of hot and cold cathodes in a tube, for different kinds of hot and cold cathode starting.

In 1938, G. E. Inman and R. N. Thayer published, in the Transactions of the American Institute of Electrical Engineering, a paper in which they discussed the lighting of fluorescent lamps and stated that at that time, the starting voltage of such a lamp "without preheating of the electrodes (cold starting) is about four times the operating voltage;" and they observed that "recent work has concentrated on preheated electrode (hot) starting because lamp life is greatly prolonged, and end blackening reduced, by eliminating the sputtering of electrode material that occurs with cold starting." From the foregoing, it is obvious that cold cathode starting was known long before the conception of the patents in suit; and Dr. Hoover testified that the problem that existed for years prior to appellants' patents was not concerned with the arrangement of circuits or design of transformers; that there was no problem of starting the fluorescent lamps; that "It was lamp life which was the big problem, which had always plagued designers of these fluorescent tubes, and it is solved only by improving and developing new types of cathodes, and not by any change in basic circuit design of the ballast or the other operating apparatus." It is clear that there was no invention involved in appellants' patents in applying the principle of cold cathode starting as such. Was the district court clearly erroneous in its findings that no invention was involved in applying the principle of cold cathode starting to a fluorescent tube that was constructed intentionally for hot cathode starting? Did the deliberate short-circuiting of the electrodes of a hot cathode tube in order to use it as a cold cathode tube give rise to invention? In arriving at our conclusion in this regard, we shall consider appellants' patent for short-circuited tubes and the manner in which such short-circuiting is accomplished, as well as a patented device which also embodies such short-circuiting, and which is cited as a reference by appellee in this case.

Boucher patent No. 2,351,914, one of the patents in suit, is concerned with a luminescent tube system and apparatus employing hot cathode fluorescent tubes with two spaced filamentary cathodes, each electrode being provided with a pair of terminals. In the system, the tubes are received in sockets, and each pair of terminals is short-circuited. With the short-circuiting, there is achieved, according to appellants' evidence, an excellent distribution of current, and a more efficient use of the filamentary cathode is had. Since the filaments of each electrode are short-circuited and the starting is had by an impressed voltage sufficient to strike an arc between the electrodes, the operation is by instant start cold cathode.

Russell patent No. 2,295,757 is the device which provides for the short-circuiting of the electrodes of appellants' hot cathode fluorescent tubes to permit instant starting. It is agreed by all parties that with respect to this patent providing for instant starting, the Boucher patent in suit, No. 2,351,914, reads directly on Russell. But appellants contend that the Boucher patent was prior in conception to the Russell patent.

It is contended by appellants that the Russell patent was not pleaded and cited in connection with the above mentioned Boucher patent, and, therefore, cannot be considered in determining whether the Russell patent anticipated Boucher. With regard to the question of citing and pleading prior patents as anticipating patents in suit, this, like many other questions of pleading, goes to the admissibility of evidence when a defense, which is required to be pleaded, is not pleaded; and such evidence, on objection by counsel, may be excluded or admitted on terms, in the discretion of the trial court. In any event, it is to be said, aside from the question of pleading patents as anticipatory, that where the defense is one of invalidity of the patent because of lack of invention, other patents can be introduced in evidence to show the prior state of the art without being specially pleaded for that purpose. Oswell v. Bloomfield, 7 Cir., 113 F.2d 377; Minnesota Min. & Mfg. Co. v. Industrial Tape Corp., 7 Cir., 168 F.2d 7.

There is much confusing testimony with respect to the priority of the Russell patent, it being contended by appellants that the Boucher patent was entitled to an effective date prior to the filing of the Russell application for patent, and prior to the date of the Russell conception.

The application for the Russell patent was filed March 1, 1941. The application for the Boucher patent was filed July 14, 1941. Appellants contend, however, that the Boucher patent is entitled to an effective date prior to the date of its application, and prior to the date of the Russell application, inasmuch as the Boucher invention was conceived prior to that time, as shown by correspondence between the inventor, his counsel, and Russell. The correspondence commences October 8, 1940, and is primarily concerned with sockets and circuits, and mentions short-circuiting a hot cathode lamp, resulting in cold cathode operation. In reply to the above letter, appellants' counsel, on October 10, 1940, acknowledged the communication, together with certain memoranda enclosed, and stated that the transformer mentioned therein seemed to have some novelty and that the circuit was novel; and he requested further information as to the desirable features of the transformer and circuit and tubes. In replying, on October 16 and 18, 1940, appellants told of the short-circuiting arrangement. They pointed out that it avoided the inconvenience of the poor appearance of certain fixtures arising out of ballast considerations and that the short-circuiting had an additional advantage in that the breaking of the filament did not impair the operation of the lamp, and that "the operation of the lamp becomes the same as an ordinary luminous tube connected to a high tension transformer." To this letter, appellants' counsel replied, after having received three letters from the patentee with regard to the short-circuiting feature, and stated: "It seems to me that what you have done is to take a hot cathode tube and do away with its hot cathode characteristics through short-circuiting the filament. In other words, it looks to me like you are using a hot cathode tube for cold cathode operation. This would seem to be inefficient and costly. I should very much appreciate your writing me fully as to the advantages of your new system, paying particular attention to the advantages of using a hot cathode tube in this system." While, in subsequent correspondence, the patentee pointed out the advantage of a cold cathode operation generally, with a short-circuit of a hot cathode lamp, there seems no explanation why a short-circuited hot cathode lamp should have any advantage over the known cold cathode lamp, for cold cathode operation, or why, in the words of the patentee's counsel, it would not seem to be inefficient and costly.

As to Russell, it appears that he wrote appellants on December 16, 1940. In his letter, Russell speaks of receiving a transformer from appellants for some tests, and asks them to send on a wiring diagram. He also states that he is "submitting our socket models to the underwriters." This is the first mention of Russell's sockets in any of the correspondence. Had Russell conceived the short-circuiting socket prior to Boucher?5

From the evidence quoted in the margin, the district court was certainly justified in concluding that the Russell patent for the short-circuiting socket was prior in the art to the Boucher patent for the same purpose, and, in fact, it is difficult to escape the conclusion that appellants, on the record, admit that Russell was the inventor of such socket. The findings of the district court that Russell's short-circuiting socket was an invention that antedated Boucher's patent for such a device, and constituted prior art, are amply sustained by the evidence, and, certainly, are not clearly erroneous.

Under all of the evidence in this case, we are unable to see how there was any invention in applying the principle of cold cathode starting to a fluorescent tube that was constructed for hot cathode starting, merely by short-circuiting the filaments of the electrodes. There was no essential difference, or advantage, or contribution to, or advance in the art over what was already known and being practiced. Instant starting cold cathode had long been accomplished by impressing a high voltage on the tube sufficient to cause an arc. But this high voltage had imposed such a sudden and heavy impact upon the electrodes that the filaments of the earlier model lamps would not stand up under such treatment and the result was that such cold cathode starting reduced the life of the tube by one-half, as compared with the slow start by preheating; and the only reason preheating had been resorted to was because the filaments of the electrodes could not, for a long period, withstand the effects of the high voltage starting. Lamp life, as Dr. Hoover stated, has always been the big problem in fluorescent lighting; and this problem has been solved by electrodes of sturdier and more rigid construction, with thicker coils having a loose overwind of fine tungsten wire, heavily coated with an emission mix of alkaline-earth oxides, which constitutes the emission material when the arc strikes. To avoid the blackening of the lamp ends which, without preheat, reduces lamp life, the cathode is often shifted to an axial mounting, instead of transversely, as in the heretofore conventional mounting, and such axial mounting is surrounded by a close-fitting cylindrical shield of nickel or iron, so that all material leaving the cathode is trapped by the shield, and end discoloration is consistently eliminated until the last few hours of lamp life. Cold cathode tubes so equipped are now operated with as long a life as the preheated hot cathode tube. It is the more sturdy construction of the electrodes that has solved the prior difficulties and made cold cathode starting a success, rather than the arrangement of circuits or anything resembling the short-circuited hot cathode lamps of appellants.

Dr. Paul L. Hoover, appellee's principal expert witness, testified that all of the circuits involved in the patents in suit are the same as those that had been used for years in the lighting of gaseous discharge tubes and devices; that the circuits of the patent in suit consist of some form of a transformer with a primary winding and secondary windings, all operating two gaseous tubes, with a condenser in one of the circuits to correct for power factor, and also to produce non-stroboscopic light.

With respect to appellants' patent in suit, Boucher No. 2,317,844, Dr. Hoover stated that Claim 11 was representative of, and, except for minor variations in phraseology, identical to the other claims in suit, 6, 7, 9, and 12, of this patent.

"11. In combination, a transformer having a core and a primary winding and two secondary coil sections mounted on said core, two luminescent tubes individually connected across the terminals of individual secondary coil sections, and a condenser connected in series with one of said tubes and its associated secondary coil section, said condenser being of sufficient capacity to substantially limit the current in said one tube and coil section, said transformer core having magnetic core shunt means around the other of said secondary coil sections for limiting the current in that coil section and its associated tube, said shunt means providing a shunt path of substantially lower reluctance than any shunt path around the secondary coil section connected to said condenser."

With regard to this circuit arrangement, Dr. Hoover specifically pointed out wherein the Osborne patent No. 2,025,471, Freeman patent No. 2,298,935, and Holladay patent No. 755,775, embraced the same claims and teachings as Claim 11 above mentioned, and constituted the prior art. He testified that Claim 11 of the Boucher patent read identically on the Osborne structure, and that the other claims in issue of the Boucher patent also read on Osborne, except for what he called insignificant details such as an auto transformer which was really the same thing as the Osborne transformer, and a shell type transformer, which Osborne did not mention, but which was a matter of choice. He also stated that the circuit of Osborne was practically identical to the circuit and transformer used by appellee except for features resulting in no difference in operation. Osborne showed no sizes or dimensions.

As to the Freeman patent, appellants say that Freeman's disclosure which is relied upon by appellee, is Fig. 1 of the drawing, but that no such disclosure is found in the disclosure sheets of 1937 or in any of the sheets at any time prior to the filing of the application for patent. But the last page of Freeman's disclosure to the Patent Department, dated March 11, 1937, shows a drawing identical to Fig. 1 of the Freeman patent. In addition, the report of tests on the unit is dated August 24, 1939, and refers to the circuit, which it states was, at that time, recently devised for operating two fluorescent lamps from a 118 volt line at a high power factor and with a minimum of flicker. The Freeman patent, in its opening lines, states: "My invention relates to the circuit arrangement for discharge devices, and especially to the circuit for fluorescent lamps." The cathode in the Freeman patent is of the filamentary type. No starting switches are shown in Fig. 1, so that the tubes are started cold cathode, which is the same arrangement as that charged to be infringed. Dr. Hoover testified that the typical claims of the above mentioned patent in suit read literally on Fig. 1 of the Freeman patent, and that there is nothing in any of the patents in suit which is not shown in Fig. 1 of Freeman except that Freeman does not show his arrangement mounted in a housing and equipped with a reflector. Moreover, since there are no starting switches shown in Fig. 1, the starting cannot be by preheated cathode, but must be cold cathode starting.

Holladay patent 755,775, issued as early as 1904, shows a system of electrical distribution for supplying two series of arc lights with carbon electrodes. This was, of course, very much different in size and operation than the patents in suit. The transformer itself would weigh a couple of tons instead of appellants' transformer which weighed only a few pounds. As Dr. Hoover testified, this did not envisage fluorescent lighting; but the characteristics of the arc lamp were the same as far as the need for controlling the current was concerned after the arc was struck. He further stated that if fluorescent lights were substituted for arc lights in Holladay, it would be merely necessary to increase the voltage by the transformer and reduce considerably the current. The former is achieved by selecting the proper number of turns used on the secondary coils, and the latter is obtained by selecting the proper size of the shunt and condenser. He further testified that the exact circuit of Holladay could be used for starting fluorescent tubes; that there is no difference in basic theory and no difference in operation between the Holladay transformer or Osborne, and appellants' transformer. "If you, in 1904, * * * had told Mr. Holladay: `Instead of operating your arc lights, * * * I would like you to operate this light.' You wouldn't have to tell him what kind of light it was, you would just give him the voltage and current rating of that light, and he would have designed you the same circuit and worked it in exactly the same way; it was merely a matter of redesigning for the correct voltage and current to suit the particular light, a light available;" and the witness went on to declare that such redesigning would not have involved any problem.

He also testified as to the Boucher patent in suit, No. 2,317,845, which, he stated, differs from the above mentioned Boucher patent No. 2,317,844 only by reason of the fact that the former has the lamps mounted in a fixture having a reflector, with the transformer positioned in a housing mounted on the back of the reflector, which was common practice in the fluorescent art. McCarthy patent No. 2,195,114 was for a gaseous discharge lamp unit with two fluorescent hot cathode lamps, with filamentary electrodes, mounted within a common fixture, which was a reflector, and carried the auxiliary equipment in a housing on top of the reflector. McCarthy showed a condenser, but not a transformer, and Dr. Hoover testified that if the voltage was sufficiently high, the unit was operable without a transformer, although normally a transformer was used; and that if the voltage were sufficiently high to strike an arc, it would not be necessary to use the starting switches.

Bridges patent No. 2,370,635, the latest in suit, is concerned with a lighting system as well as with a power unit employed in that system. As to the system, the parties met on the issue whether the circuits set forth in the Bridges claim constituted invention. Dr. Hoover testified with regard to the circuits of the Bridges transformer and those in Osborne patent No. 2,025,471, owned by General Electric Company, under which company, appellee was licensed, and stated that, in comparing Osborne's circuit to the circuits and construction of Bridges' transformer unit, he found them identical except that in Osborne, the secondaries were separated; that, in appellee's structure, there was a common connection between the two secondaries and the primaries; but that this was merely a connection difference and was of no importance in the operation. He further testified that Osborne would operate a pair of fluorescent lamps in the same way as the Bridges type of transformer. Dr. Hoover also testified that, in principle, the circuit arrangement of a structure in prior public use by the Sola Corporation was exactly the same as that on which appellants were charging infringement, and that, except for starting switches, it operated and functioned in the same manner. Bridges himself, the patentee of the patent in suit, testified that he had not known that the Sola Corporation had such a unit on the market prior to the date of his invention, but only learned of that fact after his patent had been granted. Moreover, Dr. Hoover, in his testimony, also specifically applied a General Electric structure in prior public use to the Bridges patent and stated that it read directly on the Bridges patent; and several other expert witnesses testified on behalf of appellee to the same effect.

With regard to the power unit embraced in the Bridges patent, appellants claim, first, in the language of their chief expert witness, that the difference between the Bridges patent in issue and their earlier Boucher patent No. 2,317,844 is "Fundamentally * * * in the nature of an improvement [on its prior patent], in that the very compact and inexpensive unit is the result following the teachings of this patent," and the witness then goes on to give instances of the elimination of shunts and the central core leg which "all contribute to this more compact and rugged unit." That the improvement of providing the compact, inexpensive, and rugged unit, as above related, amounts to invention for which the reward is a monopoly, is not tenable; and the district court immediately informed counsel that such was its view when the question first arose during the course of the trial.

With respect to the Sola prior use, appellants argue that, while the specification and a wiring diagram of the Sola structure in prior public use were before the court, the experimental unit itself was not submitted, nor a test report on the unit. It was testified by those having knowledge of the Sola structure that the unit and test report could not be found. Appellants, accordingly, claim that the Sola public use was nothing but an abandoned experiment. Yet Bridges himself testified that he learned after his patent was granted that the Sola Corporation had such a unit on the market, before that time. As to the public use of the General Electric structure in question, appellants claim that the evidence in this regard is no better than that of the Sola public use; that although one branch of the General Electric Company made up ten such units for operating mercury arc tubes and sold them to another branch, and the unit was thereafter placed in large-scale production starting in 1938, the units were not power units and the system was not the system of the patent in suit. Appellants seek to construe the testimony of one witness, as referring to a sample unit which was unsatisfactory, rather than to a public use of the structure, which appellee contends to have been the case; but, at most, this interpretation can only be said not to be free from doubt on the subject. There certainly was evidence upon which the trial court could conclude that the Osborne patent and the public uses of Sola and General Electric anticipated Bridges, in some instances, as to the apparatus, and in others, as to the circuits and system.

Appellants' reply to the contentions of appellee is that Bridges, unlike Osborne, was concerned with fluorescent lighting, and particularly, instant start fluorescent lighting. Appellants attack the Osborne disclosures as being entirely diagrammatic and say: "A transformer having a primary winding and two secondary windings mounted on a common core is neither shown nor suggested. Apparently it was the contention of the defendant that such a form of construction must be implied." However, Dr. Hoover testified, from the Osborne patent and from exhibits which he explained, that Osborne disclosed a transformer having a primary winding and two secondary windings mounted on a common core.

In other particulars, appellants attack the Osborne patent as being "completely silent on the employment of a `transformer of the shell type.' * * * And it fails to show `magnetic core shunt means' around one secondary coil section. * * *" As to the difference of a shell transformer, Dr. Hoover said that was a mere minor consideration, and as to the failure to show the magnetic core shunt means, he said it was indicated in the Osborne drawings by vertical lines, disposed between certain numbered coils, and that such lines represented magnetic core shunt means. Were there magnetic core shunt means in the Osborne patent? Did the vertical lines above mentioned indicate, in the Osborne drawing, such means? Was the Osborne patent anticipatory of, and prior art to the Bridges patent? The district court's findings could not be said to be clearly erroneous for having accepted the undisputed testimony of an expert witness on this subject and we find no error in its conclusions that the apparatus in question lacked invention.

The necessity of reliance by the court upon witnesses which it finds to be the most credible upon many important aspects of electrical patent cases, is illustrated in this case by testimony that while, in schematic showings of wiring diagrams, such features as, for instance, transformers are sometimes shown by parallel lines, and coils are shown by loops, that is only one way of showing them. There are various ways. The Patent Office has rules for shortening the making of drawings by the use of certain symbols other than the symbols ordinarily used by engineers; and one of the expert witnesses admitted that he could not be sure whether he could tell, by an inspection of a patent drawing, whether, for instance, there was more reluctance in one circuit than in another, although such reluctance could be shown in a patent drawing.

Boucher patent No. 2,351,915, in addition to the system or circuit, is concerned with a safety feature in which it is provided that when one lamp is removed from its socket, no voltage is left on in any of the sockets. This, it is stated, avoids the danger of receiving an electric shock to anyone removing lamps who might touch one of the exposed sockets. Dr. Hoover testified that the Hendry patent No. 1,689,485 was directed toward a system for electrical discharge illuminating lamps; that, "since they had relatively high voltage in the discharge tubes, Hendry considered it desirable to have some device so that if you had to remove the tube from the socket you can interrupt the power source and remove that high voltage, * * * when the tube is out of the socket those contacts are open and therefore there is no power supply to the transformer." He further stated that if the tube were taken out of the socket in the Hendry device, the contact was broken, and that it was purely for safety purposes "so when one removes the tube from the socket the high voltage is automatically cut off so that one can't get injured when replacing the tube or for whatever purpose one may want to remove the tube." The witness summed it up by saying that if you would apply or use "the socket arrangement of Hendry, to a fluorescent tube, then you would have precisely the same operation as described in Boucher Patent 2,351,915." In conclusion, Dr. Hoover stated that this patent device was quite old in 1927; that it was what is called, in the electrical art, an interlock which had been used in numerous ways in the nonilluminating art; and that the Hendry patent was merely an old idea of interlocking applied to the illumination field.

On cross-examination, appellee's expert witness admitted that in the Boucher patents, the electrical contacts were connected with the sockets — that the sockets were "connected to the secondary winding," while, in Hendry, the secondary winding was connected with the terminal of the lighting tube and not to the socket. He stated that the contacts, however, were "associated with the socket" and that he would call the whole thing a socket assembly. With respect to the language of the claims of the Boucher patent calling for "a pair of sockets into which said tube is positioned connected to said secondary winding," Dr. Hoover observed that the claim said the "`tube is positioned' meaning in the socket, — `connected to said secondary winding.'" He further stated that "in the structure of the patent I found a pair of sockets into which said tube is positioned * * * [it] certainly is connected to the secondary winding. It is a question whether or not it is necessary to have an intervening member on the socket or not, I mean an intervening connection, something fastened to the socket and then to the tube. Certainly in the socket which you had here * * * there is an intervening inset, intervening connecting link, but I don't know if that is necessary to meet the words of this claim. * * *" He pointed out that in Hendry, there is a pair of sockets into which the tube is positioned and it is connected to the secondary winding. "The tube is connected to the secondary winding. * * * The socket is not connected to the secondary winding."

When asked whether it wasn't impossible to remove the tube from the socket and still maintain the primary circuit in Boucher, Dr. Hoover stated that the removal of the tube broke the whole circuit, but that such was also the case with Hendry; that, while the terminals in Hendry might sometimes weld together so that, when the tube was removed, the full secondary voltage would continue with the closed contacts and the primary circuit would not be broken, such a condition would be a "mis-operating" of the Hendry patent, and that the situation suggested by the question would be brought about only when Hendry was inoperative; that the secondary circuit in Boucher could be "locked out," and that "in Boucher you have another danger which you don't have in Hendry."

In considering whether the findings and conclusions of the trial court were clearly erroneous, there remains, out of the foregoing, the significant fact that Dr. Hoover's testimony that the specifications of Boucher's patent of the safety feature read on the specifications of the prior patents; and that, according to such specifications, the safety feature in question was old in the art; and no testimony appears in the record to contradict this evidence.

It is contended that although Boucher patent No. 1,950,395 for a transformer was one of the many patents cited by the defendant-appellee in its answer, there was no testimony respecting the patent, and that, in view of the absence of any showing of the pertinence of the patent to the licensed patents and the claims sued upon, there is no basis for the finding of the court that that patent discloses and anticipates the later patents of Boucher and Bridges. It is, however, for the trial judge to decide whether he needs the assistance of an expert to understand the patent ruled upon as anticipatory or as prior art. Kohn v. Eimer, 2 Cir., 265 F. 900. Appellee points out in its brief that the early Boucher patent shows a core type and shell type transformer, used interchangeably, in different figures of the patent drawings, wherein the primary coil is in the center, the two secondary coils are on the outside, and wherein two lamps are connected to the respective secondary coils; that the circuit includes condensers as well as magnetic shunts; that the condensers operate to limit the current and to improve the power factor of the circuit; and that nothing inventive was left, as far as the patents in suit are concerned, after the full and complete disclosures of the earlier patent which was issued in 1934. Appellee also declares that the circuits in the above patent were arranged for instant starting of the tubes and that, while mention is made of neon tubes as an example, nevertheless, it is also stated that the invention relates to luminescent gas-filled tubes, and that the use of neon gas is specifically stated as used merely for the purpose of illustration. The only permissible use of expert testimony is to assist the trial judge in understanding what the specifications say. "When the judge has understood the specifications, he cannot avoid the responsibility of deciding himself all questions of infringement and anticipation, and the testimony of experts upon these issues is inevitably a burdensome impertinence." Kohn v. Eimer, supra, 265 F. at page 902. In this case, the trial judge heard an abundance of testimony upon all patents in issue and a multitude of patents referred to as prior art and anticipatory, with respect to which he concluded, in his opinion: "Transformers, of course, are old. Perhaps new uses or different constructions and applications have been apparent and designed as the art advanced. The control of electric current circuits in lighting, with lags and leads, shunts, and disposition of primary and secondary coils of the transformer, and connections, spacing, or superimposed primary and secondary coils, hot and cold cathode lighting means, all seem revealed in earlier patents and these and other circumstances have led to the opinion and conclusion that there is no precise novelty of an inventive character that can be identified beyond an aggregation of known elements and electrical current control and use, accomplished by rearrangement of essential means for producing the desired light;" and he concluded: "The extensive descriptions and claims of those patents cannot be read without being impressed by the fulsome coverage of the various phases." We have been referred to no part of the evidence with regard to the earlier Boucher transformer that disclosed the findings of the district court to be clearly erroneous; and in default of such a showing, they will not be set aside. Rule 52 of the Federal Rules of Civil Procedure, 28 U.S.C.A.

As to the contention that the system of the Boucher patent No. 2,351,914, consisting of a combination, accomplishes a result that is not achieved in the devices of the prior art, we concur with the views of the district court that the combination of the old elements in the patent was a mere aggregation, and that it does not partake of the quality of invention. Great Atlantic & Pacific Tea Co. v. Supermarket Equipment Corp., 340 U.S. 147, 71 S. Ct. 127, 95 L. Ed. 162; General Motors Corp. v. Estate Stove Co., 6 Cir., 201 F.2d 645.

On the issues of the prior art and anticipation, as to all of the patents in suit, we find that appellee's expert witness, Dr. Hoover, is the only witness who can be said to have testified clearly, emphatically and without qualification regarding the patents referred to as embodying the prior art and anticipating the patents in suit; and his testimony was undisputed that the specifications of the patents in suit read upon the prior patents, and that the prior public uses of other circuits and transformers were the same as those embodied in the specifications and claims of the patents in suit. Appellants, indeed, in their briefs, criticize Dr. Hoover and point out, in some instances, differences in the nomenclature used by him in his testimony and in the language of the specifications and claims of the reference patents, as compared to the specifications and claims of the patents in suit; and they show, for instance, that some of their lamp patents include reflectors and housings while some of the reference patents do not explicitly provide for these features. But these latter combined features do not amount to invention. Appellants fail to show what problems were solved and wherein invention resided, through the short-circuiting of hot cathode lamps for cold cathode operation. Throughout the entire case, Dr. Hoover's evidence, as to the reference patents, in all essentials — if not entirely — was undisputed. No cross-examination broke his testimony down. The proofs introduced by appellee remained unchallenged by contrary proofs or by destructive analysis; and it cannot be said that the trial court's findings, resting on his testimony, were clearly erroneous; and such findings that appellants' patents were invalid are, accordingly, affirmed. Where patents are found to be invalid, it is not necessary to pass upon the issues of infringement. General Motors Corp. v. Estate Stove Co., 6 Cir., 201 F.2d 645, 659.

We come, then, to the licensing and royalty agreement between the parties, concerning which the district court was of the opinion that it was void because of ambiguity and failure to reflect a meeting of the minds. While this opinion was not of controlling importance in the determination of the trial court inasmuch as he decided in favor of the defendant-appellee on other grounds, nevertheless, if such contract were void, there would be little to discuss with respect to the other contentions. We are of the view that the contract was not ambiguous but was, as far as the stipulations agreed upon are concerned, a binding contract. It is, indeed, to be said that it was a peculiar contract. Appellee wished a non-exclusive right to make, use or sell transformers, systems, or apparatus that might embody the invention of patents owned by appellants. It appears that both parties to the contract, as of the date of its execution, had patent applications pending in the Patent Office. Appellee was concerned with the possibility of its infringement of a patent that might be granted to appellants arising from one of the applications on file. A proposed license agreement had been submitted by appellants but appellee, being dissatisfied with its terms, prepared and submitted a new agreement that was finally accepted. In this agreement, specified royalties were payable by appellee to appellants, commencing with the determination of an infringement by appellee; and it was provided that if any disputes arose as to whether such infringement existed, the matter would be submitted to arbitration. It was further provided that in the event that defendant-appellee made a transformer for use in any structure covered by any of appellants' patents, but did not complete the construction of the power unit — leaving such construction up to a purchaser or other party, then appellee would pay a royalty on such transformer in accordance with the selling price of the complete power unit, as provided in the contract. In brief, it can be said that appellee agreed to pay stated royalties on the manufacture, use, or sale of certain transformers, systems, or apparatus contingent upon it being subsequently found that such transformers, systems, or apparatus embodied patents owned by appellants. Both parties understood the contract in the same way and bound themselves to carry out its provisions. It was a valid contract.

Appellants contend that appellee never canceled the license and royalty agreement; but on May 1, 1945, appellee wrote appellants complaining that licenses had been granted to third persons without certain safeguarding provisions, contrary to the agreement between the parties, and that, in view of such action, "we hereby give notice to you of our cancellation of the Agreement," in accordance with the provisions of the contract which gave either party the right to cancel for breach of any of its covenants on sixty days' notice. The evidence discloses that the contract was effectively canceled.

Prior to the execution of the contract, appellee company, which had manufactured transformers for approximately forty years, was concerned over the possibility that appellant, a patent holding company, might issue similar licenses to others who might be price-cutters, and, accordingly, in order to protect its interests as a manufacturer, appellee insisted upon the insertion of a price-fixing clause, which was required to be embodied in any future licenses that might be granted to others by appellants. But such a price-fixing provision was in violation of the Sherman Anti-Trust Act, according to the ruling of the Supreme Court in Sola Electric Co. v. Jefferson Electric Co., 317 U.S. 173, 63 S. Ct. 172, 87 L. Ed. 165, a case which was decided after the execution of the contract between the parties in this case. In order to avoid the effect of this holding by the Supreme Court, appellants informed appellee that they were canceling the price-fixing provision; and they now contend that since the provision was inserted in the contract for their own benefit to protect them against price-cutting by their licensees, they could cancel the provision, under the familiar rule that a licensor may waive conditions or stipulations which are made in its own favor. However, the trial court found, and the evidence clearly shows, that it was the appellee who insisted upon the price-fixing provision, and that it was inserted at its behest and for its protection against possible competitors who might thereafter be granted licenses by appellants. The cancellation of this provision by appellants, therefore, constituted a breach of contract, for such provision could not be unilaterally removed from the agreement.

It is mildly intimated by appellants that appellee, upon the decision of Sola Electric Co. v. Jefferson Electric Co., supra, thought it saw a way to keep its license, and yet squirm out of paying royalties; and that there is no reason, in good conscience, why it should be permitted to do so. Appellee, however, did not keep its license, but, rather, canceled it. As to whether appellee is justified, in good conscience, in challenging the validity of the patents, in view of its having accepted a license thereunder, this suggests the further inquiry: whether appellee, in contesting the validity of the patents, by reason of an illegal provision in the license agreement which it insisted upon for its own protection, is thus to be enabled to profit by its own wrong. In other words, the question implied is whether one seeking the aid of equity must do equity — or in this case, whether appellee must accord to appellants all the equitable rights to which they are entitled.

It may be here emphasized that a contract which cannot be performed without violation of a statute is illegal and void. Where parties are in pari delicto, the law will leave them where it finds them, and all relief is refused because of the public interest. Twin City Pipe Line Co. v. Harding Glass Co., 283 U.S. 353, 51 S. Ct. 476, 75 L. Ed. 1112. The defendant may assert the invalidity of an agreement, even though he is a participator in the wrong. Standard Lumber Co. v. Butler Ice Co., 3 Cir., 146 F. 359. The defense of illegality is allowed, not as a protection to the defendant, but as a disability to the plaintiff. Roberts v. Criss, 2 Cir., 266 F. 296, 11 A.L.R. 698.

We come, then, to the application of the maxim that he who seeks equity must do equity. But the maxim is not applicable as against a party seeking purely defensive relief. Connecticut General Life Ins. Co. v. Benedict, 2 Cir., 88 F.2d 436. It does not bind one who does not seek affirmative equitable relief. First Trust & Savings Bank v. Iowa-Wisconsin Bridge Co., 8 Cir., 98 F.2d 416. By relying upon the invalidity of the license agreement and the right to challenge the validity of the patents licensed thereunder appellee seeks no affirmative relief; and the maxim, applicable only to those who seek such relief, does not here bind appellee. Moreover, whatever may have been the appellee's self interest in denying validity of the patent in suit, there is also the public interest, for the public should be relieved from the harassment of an improvidently issued patent monopoly in the absence of a clearly recognizable estoppel. It is to be said that there appears no turpitude on the part of appellee in requiring the stipulation in the license agreement for its protection against the possibility of subsequent licenses by appellants to price-cutting competitors. In fact, appellants insist, contrary to the evidence, that the stipulation was inserted in the license agreement at their behest. It seems clear that it did not occur to appellee, at the time of the execution of the license agreement, that the price-fixing stipulation might enable it to challenge the validity of the patents licensed. That possibility did not arise in the minds of either of the parties until the decision in Sola Electric Co. v. Jefferson Electric Co., supra. But it is that decision that controls the determination of the issue here; and appellee may properly, under that case and the authorities above cited, rely upon invalidity as a defense to appellants' action.

This brings us to the question whether defendant-appellee, a licensee under the contract with appellants, had a right to attack the validity of appellants' patents. The general rule, of course, is that a licensee is estopped to attack the validity of patents licensed to him. However, when the licensing agreement is contrary to public policy for being in violation of the Sherman Anti-Trust Act, the licensee is not estopped to set up the invalidity of the patents against the licensor, in a suit by the latter to recover royalties and to enjoin sales not in conformity with the license agreement. Sola Electric Co. v. Jefferson Electric Co., supra. Appellants attempt to bring the instant case under the rule announced by the courts that when a price-fixing stipulation in a license agreement is abandoned, the licensor may maintain his action on the contract; but these are cases where such a stipulation was inserted for the benefit of the licensor, or where the parties, subsequent to the commencement of the suit, abrogated the price-fixing provisions by a new agreement, and there was no evidence that any of the parties had ever relied upon the price-fixing stipulation. Appellants also submit that inasmuch as the right of price control never came into being, appellee never had the right to challenge the validity of the licensed patents. The fact is, however, that appellee always had the right, under the contract, to invoke the price-fixing provision whenever its interests required it, and we are of the view that, under Sola Electric Co. v. Jefferson Electric Co., supra, the price-fixing stipulation was unlawful and gave appellee the right to challenge the validity of the patents.

In its decision, the district court stated that each party should pay its own costs, "because I think that this litigation is the result of the fault of each party in permitting their relationship to deteriorate into prolonged, burdensome and costly controversy." Rule 54(d) of the Rules of Civil Procedure provides that "* * * costs shall be allowed as of course to the prevailing party unless the court otherwise directs * * *." The matter of allowance of costs on the trial is within the discretion of the district court; and we are not convinced of any abuses of discretion on its part. "Moreover, this conclusion was embodied in the judgment to which the defendant has prosecuted no appeal. It cannot therefore, challenge the decision." White Cap Co. v. Owens-Illinois Glass Co., 6 Cir., 203 F.2d 694, 699. Appellee may tax its costs upon appeal.

In consideration of the foregoing, the judgment of the district court is affirmed.