Armstrong v. De Forest Radio Telephone & Telegraph Co.

MANTON, Circuit Judge.

The patent in suit was granted for a wireless receiving system on October 6, 1914, and on. an application •filed October 29, 1913. There are 12 claims in suit. All are held to be infringed by the decree below. The invention relates to improvements in the arrangement and connections of the electrical apparatus and receiving station of a wireless system and particularly a system in which the so-called audion is used as the Hertzian wave detector; the object being to amplify the effect of the received waves upon the *586current in the telephone or receiving circuit, to increase the loudness and definition df the sounds in the telephone or other receiver, whereby more reliable communication may be established or a greater distance oí the transmission becomes possible.

There are several defenses interposed, including the one of nonin-fringement. The primary question is whether or not the patentee invented the feedback circuit, with an explanation of what it is. The patentee states that he has modified and improved upon the arrangement of the receiving circuits; that the improved arrangement corresponds with the ordinary arrangement of circuits in connection with an audion conductor, in that it consists of two interlinked circuits, a tuned receiving circuit in which the audion grid is included, and which is referred to as á tuned grid circuit, and a circuit including the battery or other sources of direct current and the wing of the audion, which bs referred to as the wing circuit. The grid circuit passes through a stopping condenser, then through the secondary of the receiver, in parallel with which is the tuning condenser. Thus the circuit passes to the filament, and from the filament to the grid, through the space inside the audion. The plate circuit, starting from the plate, passes through the battery and the telephone receivers in series. Thus the plate circuit goes to the filament, and from the filament through the space inside the audion bulb to the plate. The path of the battery current is from the upper positioned terminal of the battery, through the plate of the audion, through the intervening space of the audion to the filament, from the filament to the telephone receivers, and then to the negative side of the battery. The grid is positioned in the middle of that path of the current from the plate to the filament and inside the audion. In other words, in the space between the filament and the plate of the audion, the incoming high frequency is impressed upon the grid, which is in the path of the flow of the battery curfent. The high-frequency current or oscillation is received by the antenna or aerial. An oscillation current flows through this aerial, the primary of the transformer and ground. This primary oscillating current induces a secondary oscillating current in the circuit formed by the secondary of the transformer and the tuning condenser. This oscillating current builds up by resonance a potential difference across the coil and condenser, which is transmitted through the stopping condenser, so that it exists between the grid and the filament of the audion. It is the potential on the grid which varies the steady flow of the battery current through the plate circuit. This is the fundamental operation of the audion. A tuned circuit has the same significance as a resonant circuit. It is a circuit whose inductance and capacity are so adequately related as to give it a natural frequency in accordance with the frequency of an impressed oscillating current. The usual way of a single oscillation is through the medium of an antenna and oscillation transformer, a tuned or resonant grid circuit, and a stopping condenser to the grid and filament.

The Armstrong invention consists in applying the variations, in the current, which are of radio frequency, in such mariner as to transfer energy bade into the grid circuit.. This was a wholly novel idea. This *587result was accomplished by two specific means: First, the provision of coupling between the grid circuit and the plate circuit. The specific form of this coupling consisted of the inclusion of the telephone receivers in the comiñon portion of the grid circuit and the plate circuit. This had the far-reaching consequence of automatically transferring radio — that is, high-frequency — energy from the plate circuit to the grid circuit. The second means disclosed by Armstrong for achieving regeneration consisted in making use of the inherent coupling present in the audion bulb itself; this coupling being due to the electrostatic capacity between the plate and the grid. This capacity is inherent, and Armstrong’s means made it useful and accomplished the result by the insertion of a radio frequency inductance in the plate circuit. From the sending station, radio signals are sent out in the form of radio frequency waves — waves of a frequency above audibility. These waves are receiyed by the antenna, are impressed upon the grid, and alter the potential on the grid. The grid is interposed in a gap between the filament and plate. The plate current flows across this gap, and the variations in the grid potential cause variations in the plate current, which is a current through the telephone. It has been found that, if radio frequency waves are emitted and received continuously in unbroken trains, they would produce a continued unvarying alteration of the grid potential, and a single continued unvarying alteration of the plate current, so that the only response of the telephone would be a “click” when reception began and another “click” when it ceased.

To produce intelligent signals or speech sounds in the telephone, the transmitting waves, either when they are transmitted or after they are received, must be broken up and modulated, to produce interrupted or varying wave trains, and in consequence interrupted or varying alterations of the grid potential. In the use of the detector, wave trains must thus be interrupted or modulated. The effect of the audion as a detector was that the integrated effect of a train of received radio frequency waves caused an alteration of the grid potential which produced a single pulse of current in the plate or telephone circuit. By audio frequency interruption or modulation of the radio frequency waves, either at the transmitting station or at the receiving station, audible sounds are produced in the telephones. By interrupting or modulating the transmitted waves according to a prearranged system, intelligible signals are transmitted. Heretofore the audion, used only as a detector, was thought to be a device in which radio frequency oscillations existed in the grid circuit and audio frequency impulses — the result of integration of the radio frequency wave trains — existed in plate or telephone circuit. The. idea of radio frequency oscillation in the plate circuit did not exist.

The patentee, while a student of Columbia University, living in Yonkers, was an amateur wireless operator and had a station at his home. There he made observations which led him to suspect that the radio frequency oscillations might be carried over into the plate circuit with some improvements in the detecting action of the audion. He tuned the plate circuit to radio frequency by inserting in the plate circuit such inductance and capacity as to make it responsive to the radio frequency *588waves. Then he found, not only that the radio frequency waves could be carried over into the plate circuit, but that they could be there amplified by the energy derived from the local battery in the plate circuit without change of frequency or wave form, ana that they could be fed into the grid circuit, where they increased the potential variations on the grid and the operation continuously repeated itself, producing the feedback regeneration which increased normally the sensitiveness of the device and the loudness of the receiving signals. It was in this way that he thought out his invention, which has been a great advance in the wireless art. In this feedback audion circuit, as disclosed by the patent, the incoming signal is impressed upon the grid in the same manner with the simple audion circuit, but the flow of current in the plate circuit is varied in a very different manner. Now, because of the feedback connection, regenerative amplification begins. The first half cycle of plate current variation will feed energy to the grid circuit, and will cause a larger subsequent variation in the grid potential. In turn, this increased variation in the grid potential will increase the variation in the plate circuit, and this again feeds back energy to the grid and increases its potential generation in the third and succeeding half cycles. This cyclic process continues to build up the variations in the plate current and grid potential as shown. The building up of the oscillations continues indefinitely, and the limitations are due only to the vacuum tube or audion, which weaken the feedback or regenerative action, as strong oscillations áre built up. If the regenerative or feedback coupling is sufficient to transfer enough energy from the plate circuit to the grid circuit, the energy released from the battery in the plate circuit will, be greater than the entire loss of energy in the system of circuits, whereupon continuous independent electrical oscillations will persist in the system, irrespective of the oscillations which may be impressed thereon from the antenna, and the audion in the regenerative feedback circuit becomes an independent generator of continuous oscillations.

[1] The invention, though simple in its instrumentalities, is a radical modification of an instrument which was little understood. The action of the audion was very obscure, and was not understood in the art, until Armstrong wrote his article in the Electrical World on December 12, 1914, and later an article in the Radio Institute paper in September, 1915. These articles demonstrated the action of the simple audio detector, and of the feedback audio as a detector, amplifier, and oscillator. That Armstrong has given a correct statement of the operation of the audion, with or without the feedback circuit, in these papers,- has been established to judicial satisfaction. Marconi v. De Forest (D. C.) 236 Fed. 942. What this patent discloses, and what Armstrong did, clearly amounts to patentable invention.

[2] But it is sought to defeat the patentee by the claim of prior date of invention by De Forest, and some patents in the prior art also submitted as- defenses. Since it is important to fix a date of the conception of the patentable idea, we shall review briefly his early study of the subject-matter. While a student in high school in 1906, Armstrong erected a radio station at his home, operating it as an amateur during *589tliat year. His first vacuum tube detector, a two-element vacuum tube, lie used in 1908, and continued using it until 1910 or 1911, when he obtained a three-element vacuum tube detector, a De Forest audion. lie used this last secured audion in experiments, trying out various circuit connections in order to improve its sensitiveness. It was in the spring oí 1912 that he started an intensive study and technical analysis of the action of the audion, and during this time he connected a condenser across the telephones of the simple audion receiving station, and noticed that on some bulbs an increase in signal strength would result. In the summer of 1912 the idea occurred to him to tune the plate circuit of the audion system by means of an inductance, and in September, 1912, he set up an apparatus at his home to try the experiment on a small antenna and on the press signals sent out by a radio station located on the Atlantic Coast, at Cape Cod, Mass. The signals became very much louder than any he had ever been able to receive before. In the fall of 1912, in his experiments, he found that an inductance was gradually introduced in the plate circuit, and, as the signals became stronger and stronger, a certain point would be reached where the character of the tone of the radio station signal, which was normally clear and musical, changed into a hissing or mushy sound. At the point of adjustment at which this phenomenon occurred, the tuning of the system became most exceedingly critical. This effect was due to heterodyne beat action on spark signals, and resulted from the fact that the* audion feedback was so adjusted as to release from the wing or plate circuit, or battery circuit, sufficient energy to maintain within the system continuous high-frequency oscillation, independent of the received oscillations. It was in February, 1913, when he learned to understand the cause of this phenomenon.

After obtaining these characteristics and results, he continued to use this arrangement of apparatus, particularly at night, in testing out the range of his apparatus and trying experiments with it. He made many modifications in the audion receiving circuit. One which proved to be very effective was the placing of the telephone receivers in the path common to both the plate and grid circuits of .the audion, and, because of its effectiveness, he used it permanently thereafter. His apparatus was set up in his bedroom, located in his home in Yonkers, N. Y., and the antenna was very small and of a primitive type, being swung between the roof of his house and that of the house next door. In the fall of 1912 he replaced this small antenna by a 110-foot mast. His first experiments were conducted on wave lengths of from 2,500 meters down to the ship radio service wave lengths of about 600 meters. He was able to deceive all of the Atlantic coastal radio stations within a very short time thereafter. Then he constructed a suitable apparatus to be used in the reception of longer waves; that is, wave lengths of the order used by the then-existing transatlantic radio stations.' This apparatus, which he constructed, was offered in evidence. It was identified by amateurs, who saw it in his home. He told his father of the discovery, and showed him his apparatus, and had him listen to the signals about the 1st of October, 1912, and asked financial aid to secure a patent. This was not granted him. On December 7, 1912, he told *590a college mate of his success in improving the sensitiveness of the audion receiver by means of a new connection and an entry was made by that college mate of that fact in his diary. Two days later, December 9, 1912, he told the same college mate of having heard Clifden, Ireland, and a notation of this was made in the diary. There is corroboration of his testimony as to experiments in 1912. He showed his receiver to his uncle, and tried to get him to advance money for a patent application. He did not succeed, but was advised by his uncle to go before a notary public and make a drawing of the connections and have it witnessed by such notary. This was done on the 31st of January, 1913. A drawing on a tracing cloth of his circuit connections was witnessed by a notary and the date was fixed thereon. The sketch shows his mental conception showing the essential features of an invention which was already a completed, successfully operative instrumentality. It fixes, beyond dispute, a date prior to which the invention was completed. He was without means to file his patent application, because he was unsuccessful in getting the money from his father or his uncle. He realized the importance of his invention, made the drawing, and had it witnessed before a notary, and there made a record of his invention, the best he could under the circumstances. Therefore we have a reduction to practice at least as early as January 31, 1913, the date of the drawing.

[3] There is no reason to doubt the testimony of the patentee as to-the date of his inventive thought. While it is true that uncorroborated testimony of an inventor is to be accepted with caution (Clark Thread Co. v. Williamantic Linen Co., 140 U. S. 481, 11 Sup. Ct. 846, 35 L. Ed. 521), yet there is no rule of law that requires the rejection of the-uncorroborated testimony of an inventor as to the date of his "conception. The court must be satisfied as to the fact of demonstration and that the conception was complete. A mere assertion of the inventor at a given time, unaccompanied by contemporaneous acts or words indicating its existence, is not sufficient evidence; but if a court of equity is satisfied of the good faith, and the truth of the apprehension .of the invention is definite enough to have enabled him to reduce it to practice-without a further exercise of inventive skill, that is sufficient. In other words, an earlier conception, when alleged, is a matter of fact, and must be established by sufficient proof. Where there has been disclosure to others, or embodiment of the invention in some clearly perceptible form, such as a drawing, with sufficient proof of identity in point of law, that is sufficient. While the mere unsupported evidence of the-alleged inventor on an issue of priority, as to the fa?ct of conception and time, cannot be received as sufficient proof of prior conception, any full and accurate description of the invention, either in words or drawings or by 'model, will suffice, although the attempt to represent it may have failed. All that the law requires is that the drawings exhibited and relied on as evidence of the conception of the invention-must show a complete conception, free from ambiguity or doubt, and such as would enable the inventor or others skilled in the art to reduce the conception to practice without any further exercise of inventive-*591skill. The rule as to corroboration is one of caution and discretion, and is not a statutory provision.

Here the patentee had a complete disclosure of the invention in his sketch. He had an actual apparatus, which was in existence at the time of the sketch, and which is received in evidence. Further than this, on February 17, 1913, the patentee wrote a letter to one Underwood, telling of his experiments and results. It sets forth sufficient to justify the claim that he was referring to the regenerative audion, and stamps the patentee as the original inventor of the feedback circuit. He expressed his intention to make quantitative measurements to determine the cause of the choke or hissing state. He suspected that the audion was independently generating high-frequency currents, and after discovering that the audion was generating high-frequency oscillations, the explanation of the whistling note on the telephone occurred to him. He said he became convinced that he was observing a beat effect between the incoming signals and the local oscillation. To test this out, he set up a second audion system similar to the first, and by adjusting both audion systems in the hissing state, he obtained beats between them, which could be carried to any desired pitch. In March, 1913, he constructed apparatus with constants suitable for receiving messages from Clifden, Ireland, and Glace Bay, in conjunction with the small antenna he was using, and succeeded in receiving the Clifden signals with regularity. Prof. Mason confirms this as of March, 1933.

We think this was sufficient to fix the date of the invention as to the date of the sketch, to wit, January 31, 1913. He demonstrated diligence in making known his idea and in protecting it. He was handicapped financially, and we do not think he was guilty of any laches which worked to the disadvantage of any competitor in inventive thought in the art or himself. He used the apparatus upon which he made his invention, subsequently in commercial use at Sayville, N. Y., with the single exception that some of the original induction coils used at his home were bulky and wound upon holsters or cardboard cylinders, and in the apparatus used at Sayville they were replaced by coils which were physically smaller, but possessed the same electrical constants. He was called into this practical service at Sayville to overcome difficulties in the receipt of signals from Nauen, Germany, in 1914. His apparatus continued in use there for a considerable length of time, until the engineers were able to construct a duplicate system, when the apparatus was returned to Armstrong. The engineers there preferred to use this apparatus in the oscillatory state, even in the reception of spark signals. Since then apparatus built under his invention have reached a high degree of commercial success.

The testimony of De Forest has been offered in evidence, by which it is attempted to show that he had conceived the invention in 1912 and 1913, and that he is, in-point of fact, the prior inventor. His testimony shows that he experimented, having in mind two uses of the audion: First, the use of the audion as a telephone relay or repeater; and, second, the development and use of the audion as an amplifier of telephone and detector of radio signals, and the use of the amplifier to record detected radio signals on a telegraphone wire. The invention *592involved here has the greatest value in the amplification of radio signals and the recording of them on a telegraphone wire. It would have produced the long sought for high-pitched note which De Forest and his associates' were trying to produce throughout this period, and failed in. It is testified that De Forest was endeavoring to produce a beat note for the reception of continuous wave signals, and to record them on the telegraphone. The appellant offered in evidence De Forest’s experimental note books, showing entries made under date of June 21, 1912, where there is the observation of a beat or high-frequency note with a straight audion hook-up. The note shows this to have heen transient and incapable of reproduction, and he recognized that it was not the true heterodyne effect. This was due to the gas action in the tube, an effect which has always been observed by users of the straight audion hook-up, and which is explained in the testimony of Hazeltine in the interference record. A true heterodyne effect is produced whenever there is, at the local receiving station, a series of oscillations differing slightly in frequency from the received oscillations, so that the two groups of oscillations combining, produce beats. This true heterodyne effect is characterized by the fact that, upon adjustment of the local frequency, the signal becomes audible in the telephone as a very high pitched note, which gets lower in pitch as the frequency of the two oscillations approach one another, becomes so low as to be inaudible, and then comes in again at a low pitch, passes on to the high pitch, and goes above audibility again. A high pitched note may be produced by the gas action at audio frequency. This note is not due to the heterodyne beat, but to the audio frequency oscillations of the tube. This entry of June, 1912, was probably due to gas action, local oscillations of audio frequency, and in the entry of April 17, 1913, the note was due to gas action, local oscillations of radio frequency combining with the incoming oscillations to produce a beat note. In neither of these was there any feedback action. The entry of June 21, 1912, indicates that De Forest did not produce, or know how to produce, a beat note with an audion receiver. He appreciated the importance of applying the audion,‘if possible, to the production of the beat note by heterodyne reception, but failed to do so..

A notebook of Van Etten was offered under date of August 6, 1912. Van Etten was working on the audion as a telephone relay and amplifier, having had telephone experience which qualified him for the work. On July 23, he entered in his note book a two-way telephone circuit, using two audions, with a note: “Think the following would .possibly make a good telephone repeater.” On Jufy 24 Van Etten attempted to set up this two-way circuit with a single audion having two grids and two plates. An entry on August 6 shows this was a failure. On August 6, he connected the input circuit of the double audion to the output circuit, and thus in an accidental way found that the audion would howl or sing. He pointed out in his notes that the arrangement gave beautiful clear tones, which lowered and raised in pitch as the B battery was varied, and could be wiped out by the magnet, and “then the watch ticks come through as before.” He says:

*593“This phenomenon is apparently similar to the ‘howl’ produced in an ordinary O. B. telephone when the receiver is placed against the transmitter, but nevertheless, as shown by variations of number of cells in battery B and by the magnetic wiper, is also intimately associated with the audion.”

On the same day he set up the balanced two-way telephone circuit, corresponding to the two-way telephone circuit which is used to prevent the howling of the mechanical telephone relays, and then made an entry in his notes that he tried “following circuit, but it was n. g.; could not make it boost at all; could only make it howl.” These were characteristic of the notes of Van Etten, and showed efforts to produce a telephone relay circuit in which the audion would not howl, and in this he succeeded August 29, 1912. There is nothing to show, however, that what Van Etten learned about the circuit arrangement on August 6, 1912, was ever brought to De Forest’s attention. He did not include it, and, on the contrary, the record indicates “that this new and simplified circuit” was deemed useful only for abortive “trigger effect.” Although De Forest’s testimony attempted to show that he liad in mind the feedback connection on August 6, and that he had conceived the idea that the feedback would improve the amplifying properties of the audion, Van Etten, when called in the De Forest interference record, says that he did not know, even to this day, that the feedback circuit could be used to produce amplification.

On February 20, 1915, De Forest published in the Electrical World an article in which he made claims with respect to his early work on the oscillating audion, and referred to two such experiments, the first of which he said occurred in the latter part of 1910 or 1911, and the second on August 26, 1912. In these there is no mention of the feedback circuit of the Van Etten August 6 entry, which it is now claimed' represents his first real discovery of a controllable oscillating audion. These and other circumstances seem to us inconsistent with the idea that De Forest had any real knowledge of or understood the Van Etten accidental circuit arrangement of August 6, 1912. Nowhere in the notes which are in evidence is any reference made to the terms which would ordinarily be used if such a discovery were made and understood. The terms “feedback” or “regeneration,” “input circuit” or “output circuit,” or “reainplification,” are not found in the notes. An entry of August 29, 1912, referring to the two-way circuit which did not liowl, made by Van Etten, showed the use of two audion amplifiers and cascade, and the record shows that he explained this arrangement to De Forest, and that it was considered of great importance, and was copied in De Forest’s notebook by Van Etten on that day. It is described merely as an amplifier by Van Etten in his notebook, but De Forest in his notebook says that, by reversing certain connections, the arrangement makes all sorts of musical notes in the telephone, and that the notes could be changed by putting very small capacities between ground and grid, wing or filament. De Forest says this was due to a feedback. Van Etten does not corroborate him as to the existence of any feedback in the circuit, and Hazeltine, in his discussion, makes it clear that, through the complex circuit arrangement of a cascade audion amplifier, it is difficult to avoid electrostatic or capacity coupling be*594tween the plate of the second audion and the grid of the first, and that this difficulty will give rise to oscillations similar to those described in De Forest’s notebook of uncertain frequency, and liable to jump from one frequency to another, because in the complex circuit arrangement there are several independent inductances having difficult natural periods, at any one of which oscillations may occur.

These entries show that, at the very date, De Forest did not have the invention here in controversy. The thing which makes valuable the feedback circuit is that the amount of feedback and the frequency of the oscillations depend directly upon the electrical constants; that is, the inductance and capacity of the circuit, and not the heating current or plate potential. This characteristic makes the instrument a practical one, capable of control in a commercial way, and responsive to the desires of the operator. One having this conception of the feedback circuit would not malee an entry suggesting that the oscillations were independent of the electrical constants, or that they were peculiar to the audion itself. The De Forest 1914 notes disclose when and where he discovered that the oscillation of phenomenon was not peculiar to the double audion. These notes contain the idea that the pitch of the sounds depends upon the inductance and capacity of the circuits. In February, 1915, after the Armstrong patent was issued, in an article written in the Electrical World, he showed that he knew that the frequency depends upon the electrical constants of the circuit, and he then made claim of this characteristic for the gas action straight audion hook-up of 1910 and 1911, even though gas action oscillations do not have this essential characteristic. The inference is fair that at the time of this writing he did not know the difference between the two. He did not disclose tire circuit arrangement to his associates on his own statement, although he says that this was for fear the Federal Telegraph Company would lay claim to it, but at that time he was under contract to assign his inventions to that company and to use his best endeavors in their behalf.

On April 17, 1913, notes were made of the work done in June, 1912, and they show that he was looking for “beat note” and showed an arrangement of the straight audion hook-up, in which, instead of having a single tuned circuit connected to the grid, there were two tuned circuits so connected; the intention being to break up the received signals into two sets of signals of slightly different frequency, which would together produce a “beat note.” His later entry showed that this was an impossibility. A later entry, made under date of April 17, 1913, said: “This day I got the long looked for beat note.” The entry made on that day establishes that De Forest had not been able to get the het-erodyne phenomenon prior to April 17, 1913, which was some months after the day we are crediting Armstrong for his invention, to wit, January 31, 1913.

In his testimony, De Forest said that, to change the circuit in his experiment, the load coil was taken out of the antenna and put in the plate circuit, although he stated that his note of April 17 did not mention this essential change. In the interference, De Forest said two loaded coils were used, one in the antenna, and the other in the plate *595circuit. On the trial he said it must have been a feedback circuit, because the true heterodyne effect could not otherwise be produced. However, on the trial, it was demonstrated that, with a straight audion hook-up, the beat note can be produced by gas action, without any feedback at all. De Forest’s relation to the art, and what he was endeavoring to discover, and his association with engineers engaged in the same art was so intimate, that we feel confident that, if he had iu his mind the feedback, or anything bordering upon the invention which Armstrong discovered, he would have made it known, together with appropriate records, to put his having invented it beyond controversy.

The notebook entries of 1914 negative the idea of his having had the conception of this invention in 1912. His notes deal with his “trying for squeal or whistle — to get this in a critical state, where the right spark frequency will start it, and other frequencies will not.” He refers to his having followed the directions of his Palo Alto notes without any result; “couldn’t get one bulb alone to squeal — used a double grid and plate bulb with distinct leads.” This negatives the claim that De Forest knew about the feedback circuit, and how to set it up and control it, and was in possession of the vast amplifying capacity of the feedback circuit. If he was in possession of all this, he" cotí fd get the audion to “squeal or whistle,” and there would have been no necessity for his pursuing the idea of “adjusting the audion circuits to a critical state, where the right spark frequency would start oscillations, and other frequencies would not.” As late as February 10, 1914, he made an entry:

“1 find that both grids are necessary — either alone, or with both joined together, will giro the whistle.”

It seems he still had the idea that the whistle or oscillating condition was the characteristic of the double wing, double grid audion, and cow’d not be produced with a standard audion having a single wing and a single grid. It was not until September, 1915, that De Forest filed an application describing his feedback circuit. This is the application which is involved in the interference, and contains broad claims for a feedback circuit used to generate oscillations. The record shows that he had filed some 30 patent applications, and it is curious to note the late day that he filed his application. This record clearly negatives the claim that he invented and had the idea of a feedback circuit in 1912.

Meissner filed an application in this country on March 16, 1914. In the oath, four German applications are recited, filed April 9, 1913, October 1, 1913, October 23, 1913 and December 6, 1913. The German application of April 9, 1913, was granted, and is now known as the German patent, No. 291,604, dated June 23, 1919. It discloses the regenerative or feedback circuit adjusted to produce oscillation. There is nothing suggesting the use of this arrangement in a radio signaling system either as receiver or transmitter. The German application of October 1, 1913 (now German patent No, 295,673), granted June 23, 1919, discloses a double audion generator which is not of interest in this issue. The grant of October 23, 1913, provides for a divided grid *596electrode circuit, which is useful in connection with the generation of oscillations, and the last application of December 6, 1913, deals with the provision of a metallic filament for an oscillator bulb. Since we have fixed the date of Armstrong’s invention as early as January 31, 1913, these applications and subsequent patents would not anticipate, even though they were for the same invention.

The Schloemlich and Von Bronk (United States) patent No. 1,087,-892, issued February 17, 1914, was filed March 14, 1913. The oath recites two German applications, one filed September 2, 1911, and the other filed February 8, 1913. Prof. Hazeltine points out that there is no regenerative feedback included in it. The inventor recognized the existence of high-frequency plate current variations in the plate circuit, and that these variations were of greater amplitude than the potential variations on the grid. They proposed to use the straight audion hookup, without feedback, for amplification of high-frequency signals, without using it as a detector; the amplified variations being- detected by a separate crystal detector associated with the circuits. They show no realization of a use for reamplification or regeneration, or that it could be used simultaneously as an amplifier and detector. It lacked essentially what Armstrong invented, namely, the feedback from the plate circuit to the grid circuit, and the plate current variations to accumu-latively amplify the potential variations on the grid.

The Reisz patent. No. 1,234,489, was filed subsequent to Armstrong’s date of invention, namely, April 9, 1913. It does' not anticipate the patent in suit.

[4] Concluding as we do, that the patentee has a meritorious patent that has not been anticipated, the remaining question is: Does the appellant use, and therefore infringe, the appellee’s patent? The patent in suit is for a definite electrical instrumentality, and differs from the prior art. It is a straight audion hook-up, in that the patentee inserted in that prior art arrangement means supplementing the electrostatic coupling of the. audion to facilitate the transfer of enérgy from the wing to the grid circúit. It is either the radio frequency inductance in the wing plate circuit, the inductance L1 of Figures 1, 2, 3 and 6, or the feedback coupling in the common path, the telephones R, Figures 2 and 4, or the telephone R, transformer T, and shunted capacities Cr> and C8 of Figure 1, or the inductance D2 and the shunted capacity CB and Cs of Figure 6. The patentee’s invention covers the regenerative or feedback audion. It is a definite instrumentality; that is, an arrangement of the audion circuits on which oscillating current energy is transferred from the output or plate circuit to the input or grid circuit, to sustain the oscillations in the grid circuit. It is essentially useful as a radio receiver, and has proven to be a wonderful improvement on the old audion receiver. The appellant uses the feedback circuits in a radio' transmitter, and, the patent covers the use of the feedback circuit, and applies even though the claims are in terms limited to a receiver. Western Electric Co. v. LaRue, 139 U. S. 601, 11 Sup. Ct. 670, 35 L. Ed. 294. The inventor is entitled to the benefit of all uses to which his invention can be put, no matter whether he conceives the idea of the use or not.

*597[5] It is contended that the oscillating audion circuit, for which both De Forest and Armstrong have pending applications in the Patent Office, constitutes another invention and that the nonoscillating audion receiver only is involved in the patent in suit. But the language of the claims of the patent in suit covers the feedback audion circuit adjusted .to produce independent continuous oscillations. It is another particular use of the invention of the patent in suit. It involves a self-heter-odyne use. It was an independent improvement invention, which is described in Armstrong’s second application. Its validity is not involved in the present consideration. McCreary v. Penn. Canal Co., 141 U. S. 459, 12 Sup. Ct. 40, 35 L. Ed. 817; Suffolk Mfg. Co. v. Hayden, 70 U. S. (3 Wall.) 315, 18 L. Ed. 76. Nowhere did Armstrong disclaim the oscillating audion. The feedback circuit of the patent can receive all sorts of signals, damped waves or continuous waves, in the oscillating or nonoscillating condition. If the audion is in the non-oscillating condition, it receives the damped waves with a musical tone greatly amplified. If the audion is in oscillating condition, damped waves will come in with a mushy tone, but greatly amplified. If continuous waves are be received with the oscillating audion, it is found cheaper and better (though not always) to use the audion simultaneously as a local heterodyning source.

[6] We do not agree with the claim of the appellant that the patent is for a principle. It is for an instrumentality. It should be construed, to cover the uses of the apparatus which are described and claimed. As the testimony of the expert called by the appellee indicates, the appellant’s use infringes all of the claims of the patent in suit relied on. The feedback connection of the patent is shown in each of the appellant’s apparatus. There is a two-path feedback in each apparatus, which is pointed out in charts 7, 8, and 9. On chart 7 there is a two-path feedback, but the principal path is through the telephone shunted by the primary condensor. These two instruments, the red and green lines running side by side, show that they are connected by a commor path. Chart 8 shows that the feedback is through the telephones and the grid condensor connected in the common path, and chart 9 shows the coupling is through the induction of the tickler coil alone, or, in other words, the inductance of the tickler coil and the transformer coil shunted with the antenna.

We think this excellent contribution to the wireless art should be accorded the full scope which the court below gave it in the decree. We think the decree is not too broad, but properly describes what the inventor conceived, and for which protection must be accorded to him.

Decree affirmed.