Plaintiff, Harnischfeger Corporation (hereinafter referred to as “Harnischfeger”), and defendant, Miller Electric Manufacturing Company (hereinafter referred to as “Miller”), are both Wisconsin corporations having their principal places of business in Milwaukee and Appleton, Wisconsin, respectively. Both *47parties are manufacturers of arc-welding equipment.
This action arises under the patent laws of the United States. The complaint charges defendant with infringement of United States Letters Patent No. 2,535,154, hereinafter called the Oestreicher patent, issued to plaintiff as assignee of Sol Oestreicher, on December 26, 1950. Plaintiff has been the owner of the Oestreicher patent at all times since that date.
The Oestreicher patent relates to alternating current arc welders of the type wherein a saturable reactor is employed as the means of regulating the amperage of the arc current. The four claims of the patent, infringement of all of which is alleged, are directed to a combination of elements including a transformer and a saturable reactor, the reactor being connected in the welding current circuit between the transformer secondary current and the welding arc.
The charge of infringement in this action relates to the manufacture and sale by defendant of two classes of arc welders, referred to in the record as the 250 series and the 300 series.
The defendant Miller interposed the defenses of invalidity and noninfringement.
Alternating Current Arc Welders— What They Are and How They Work
In electric arc welding, a gaseous electric arc is struck between a welding electrode and juxtaposed metal pieces to be joined together. Electric current passes through the arc, and the energy therein is transformed into intense heat which melts the work pieces. The welding electrode is moved along the seam whereon a weld is desired, and as the arc advances, the molten metal in its wake cools and solidifies into a single body.
The alternating current (hereinafter abbreviated “A.C.”) commercially supplied by power lines is at a voltage too high for arc welding, and transformei’s are, therefore, used in A.C. arc welding to step the line voltage down to a suitable value.
In A.C. arc welding the current periodically reverses direction, momentarily passing through zero during each reversal. Commercial alternating current has a frequency of sixty cycles per second which means that its direction of flow reverses 120 times each second, twice for each cycle of current. Each time the current passes through zero, the electric are is momentarily extinguished. The hot gases in the are zone retain their electrically conductive qualities for a brief period after extinction of the arc, however, and hence the arc is normally restruck when the voltage builds up in the other direction. Should the build-up of the current.in the opposite direction be unduly delayed, the hot gases may cool sufficiently to prevent the arc from re-striking. Slow current build-ups during reversals are referred to as “dwells,” and avoidance of excessive dwells is one of the requirements for a stable, smoothly-functioning welding are.
The amount of heat generated by the arc depends upon the amperage of the current flowing through it. A commercial arc welder must provide some means for regulating the amperage since different quantities of heat are required for different types of welding, depending upon the kind of metal being welded, its thickness, the depth of melting desired, and other factors.
In all the A.C. arc welders involved in this case, including those of the prior art, regulation or control of the arc current or amperage is accomplished by varying the amount of “reactance” in the arc circuit.
Reactance is an electrical property associated with coils linked by A.C. magnetic flux. Reactance in a circuit has the effect of limiting the A.C. current flow to a degree dependent upon the amount of reactance present. All coils possess at least some reactance.
The quantity of reactance associated with any given coil is determined by the number of turns in the coil, the magnetic conductivity of the frame serving as a guide path for flux linking the coil, its configuration, and other factors, such as *48the current in other coils which link the same frame. Flux produced by one coil that does not link with another coil is called “leakage flux” and produces “leakage reactance” in the flux-producing coil.
Output amperage of an A.C. arc welder can be controlled:
1. By moving one of the transformer coils with respect to the other and thereby varying the leakage reactance in the transformer itself;
2. By providing in the transformer a special magnetic path or frame for leakage flux and controlling the amount of leakage flux (and hence leakage reactance) by changing the magnetic characteristics of the special path or frame; and
3. By placing in the arc circuit external to the transformer one or more separate iron-cored coils which function as “reactors” — i. e., coils whose sole function is to introduce reactance into the circuit.
It has been customary for many years to place air gaps in the magnetic frames of reactors used in arc welding. (Weed Patent No. 1,612,084; Frickey Patent No. 1,539,044; Steinert Patent No. 2,-175,927; and MacKenzie’s The Welding Encyclopedia, 11th Edition, 1943, page 548.)
Such gaps prevent excessive dwells in the welding current and thus contribute to arc stability.
In the third above-mentioned method of controlling amperage — i. e., by the use of one or more separate reactors, various techniques are known for changing the reactance of the reactor or reactors and, hence, changing the amperage of arc current. One of the simplest techniques is to change the number of turns in the reactor coil or coils. This is commonly done by tapping the coil or coils and providing a switch or plug-in arrangement so that the welding operator may select the number of turns desired.
Another method of varying the reactance of a reactor is to change the flux-carrying ability of its magnetic frame. This can be done by varying the size of the air gap, and it can also be achieved by passing direct current (hereinafter abbreviated “D.C.”) through coils wound on the reactor frame for that purpose, thereby partially “saturating” the frame with D.C. magnetic flux. To the extent that the flux-carrying ability of a reactor frame is thus pre-empted by D.C. flux, the frame becomes a less effective path for A.C. flux, and the reactance of the reactor is correspondingly reduced.
All the methods mentioned in the foregoing for varying the reactance of a reactor and thus controlling arc-current amperage were known and used twenty years or more before the Oestreicher patent. (Frickey Patent, page 1, line 101 et seq.; Weed Patent, page 1, lines 46-49.)
Under some conditions A.C. passing through a coil wound on a magnetic frame may induce in other .oils on the same frame alternating voltages (and currents if the circuit is closed) having frequencies which are a multiple of that of the original current. These multiple-frequency voltages and currents are called “harmonics.” In reactors of the type in which reactance is controlled by partially saturating the reactor frame with D.C. flux, produced by passing direct current through D.C. coils wound on the reactor frame, harmonic voltages are induced in the D.C. coils by the A.C. welding current in the reactor coil, and A.C. harmonic currents flow in the D.C. circuit as the result of these induced harmonic voltages. These harmonic currents react back upon the welding current and may affect its wave form by producing dwells. Harmonic potentials and currents are an inevitable accompaniment of saturable-reactor control of A.C. welding current. The magnitude of harmonic currents can be reduced by using an auxiliary reactance in the D.C. circuit. Dwells, however, will also be produced as the result of partial saturation of a reactor frame by the A.C. welding current itself, even when the D.C. premagnetizing current is zero or very low.
The Patent in Suit
The stated object of the Oestreicher patent is to provide an arc welder in *49which the “current wave form passes rapidly through the zero current axis without any substantial dwell at or near zero current, whereby the arc produced is rendered more stable.”
The arc welder disclosed by the Oestreicher patent consists of a nonadjustable transformer with an adjustable reactor connected in the circuit between the transformer secondary coil and the welding arc. Oestreicher’s reactor, as shown in his patent, is assembled on a unitary magnetic frame having three legs, an A.C. reactor coil being wound on the central leg and D.C. premagnetizing coils being wound on the outer legs. The central leg alone contains an air gap which interrupts the path of the magnetic flux produced by alternating current in the A.C. reactor coil but does not interrupt the path of D.C. magnetic flux generated by direct current in the other coils. This particular arrangement permits saturation of the frame with a lesser amount of D.C. magnetizing force than would be required with any arrangement in which both the A.C. and D.C. flux paths are interrupted by an air gap or gaps.
The reactor of Oestreicher’s welder is of the type already referred to in which the reactance is varied by regulating the quantity of direct current nassing through D.C. coils, the degree of saturation of the reactor frame being thereby altered. This variation of reactance changes the amperage of the welding current.
The above cut of the Oestreicher drawing shows two principal units; namely,
a transformer designated generally by the numeral 1 and a saturation reactor *50designated generally by the numeral 2. The magnetic frame of the reactor 2 is made up of a central leg 17, side legs 18 and 19, return portions 20 and 21, and an upper return portion 22, the latter being ■spaced from the end of central leg 17 so as to leave an air gap 23 as shown. Surrounding the outer legs 18 and 19 are D.C. excitation windings 24 and 25. Surrounding central leg 17 is A.C. winding 8.
Claims 1 and 2 of the Oestreicher patent read as follows:
“1. In an apparatus adapted to furnish a regulated supply of arc welding alternating current a transformer having a primary winding adapted to be connected to a source of alternating current and an output circuit including a secondary winding adapted to furnish alternating current suitable for welding; means adapted to regulate the reactance of the output circuit of said transform■er, said means including a three legged reactor frame having side legs and a central leg interrupted by an air gap, a reactance winding connected in said output circuit surrounding said central leg, direct current magnetizing control windings .surrounding said side legs, and a •controllable source of direct current for exciting said windings.
“2. In an apparatus adapted to •furnish a regulated supply of arc welding alternating current the combination comprising a transformer having a primary winding adapted to be connected to a source of alternating current and a secondary winding; a saturable reactor having a magnetic frame provided with a •central leg and side legs to form a magnetic circuit, an air gap in said ■ central leg interrupting said magnetic circuit, a reactance winding .surrounding said central leg connected in series with the secondary •of said transformer to form a welding output circuit, direct current magnetizing control windings surrounding said side legs; and a controllable source of direct current for exciting said magnetizing control windings.”
Claim 3 is dependent on Claim 2, and Claim 4 is dependent on Claim 1.
The location of the air gap and the effect of its location upon the operation of Oestreicher’s welding apparatus bears vitally upon both the validity and infringement issues. Harnischfeger concedes that Claim 2 requires the air gap to be in the central leg. Whether Claim 1 is similarly limited, however, is hotly disputed.
Harnischfeger asserts that Oestreicher was the first to introduce an air gap into a saturable reactor with variable D.C. premagnetization control to be utilized in a welding circuit to improve welding current wave shape.
Defendant’s Accused Welders Defendant’s 250 series welders are A. C. arc welders in which control of the output current is achieved by varying the degree of D.C. premagnetization — i. e., saturation — of the magnetic frames connected between the transformer and the welding arc.
The magnetic frames are rectangular and physically separate from each other. A reactance winding connected in the output circuit between the transformer secondary and the welding electrode surrounds the adjacent legs of the magnetic frames, all turns of the reactance winding linking both adjacent legs of the magnetic frames. The reactance winding consists of two sections, one of which sections is tightly wound about the adjacent legs of the magnetic frames and the other of which sections is connected turn for turn with turns of the transformer secondary, the two sections being in series with each other. D.C. coils are wound on the other legs of the two frames.
Each magnetic frame in the 250 welder is provided with air gaps which interrupt both the D.C. and A.C. flux paths. These gaps can be located anywhere in the frames without altering the operation of the 250 machine.
*51
The above drawing shows a transformer designated by the numeral 1 and reactor designated by the numeral 2. Wound around the adjacent legs of the reactors is one section of A.C. winding 3. The other section of the A.C. winding 3 is wound around the adjacent legs of the reactors and connected turn for turn with turns of the transformer secondary. Windings 4 and 5 are D.C. excitation windings. Numerals 6, 7, 8, and 9 designate air gaps.
Defendant claims that in the 250 series welders there is no separate reactor, and the leakage reactance of the transformer is varied by changing the magnetic properties of the transformer itself. Further, that in the 250 welders the leakage reactance of the transformer is controlled by regulated diversion, away from the primary coil, of flux produced by current in the secondary coil. This regulated diversion is achieved by adjusting the degree of saturation and flux-carrying ability of separate magnetic frames that link the secondary coil.
The court finds, however, that the 250 welder’s “flux diverter” has the same *52structure or its equivalent, operates the same way, and brings about the same results as a reactor and certainly is a reactor or its equivalent.
Defendant’s 300 series arc welders are of the separate reactor type. In them a transformer is provided to step down the line voltage to a value suitable for welding, and the secondary coil of the transformer is connected to the arc circuit through a pair of saturable reactors. Each reactor consists of a rectangular magnetic frame having an A.C. coil on one leg and a D.C. coil on the other leg. Each reactor frame has air gaps which interrupt both the D.C. and the A.C. flux paths. It is immaterial, so far as the operation of the device is concerned, where in the rectangular reactor frames the air gaps are located. The two reactors are entirely separate from one another, physically and magnetically. While the separate reactor frames are physically oriented with the A.C. coils on the outermost legs and the D.C. coils on the adjoining legs, the parties agree that the position of the coils on the frames makes no difference in the operation. Indeed, the A.C. and D.C. coils could be interchanged on the side legs of either or both reactor frames, or wound on the horizontal portions, without significantly affecting the operation.
Each of the A.C. reactor coils in defendant’s 300 welder has a rectifier — i. e., a one-way electric valve — connected in series with it. The rectifiers and A.C. reactor coils are connected together in such a way that the A.C. arc current splits, passing through one of the reactors and its associated rectifier during alternate half-cycles of the arc current and passing through the other reactor and rectifier during the other half cycle of welding current. Thus, while the arc current itself is A.C., the current flowing through the individual reactors is actually pulsating direct current.
In the 300 series, the output current is controlled by varying the quantity of D. C. premagnetizing current in the D.C. reactor coils. The 300 series welders, with their rectifier-controlled split-current action, are very different from the welder of the Oestreieher patent in structure, in operation, and in result.
*53
The above drawing shows a transformer designated by the numeral 1 and a pair of reactors designated by the numeral 2. Windings 3 and 4 are A.C. windings, and windings 5 and 6 are D.C. excitation windings. Numerals 7, 8, 9, and 10 designate air gaps, and numerals 11 and 12 designate rectifiers.
Plaintiff’s theory of infringement with respect to the 300 series welders is grounded on the assumptions: (1) That the rectifiers may be short-circuited and disregarded in considering the question of infringement, and (2) that the 300 welders as thus reconstructed are the “equivalent” of the welder of the Oestreicher patent. The 300 welders are never sold or operated in practice without their rectifiers. The rectifiers in the 300 welder, moreover, admittedly cooperate with the other components of the device in important respects which affect the operation of the welder as a whole.
The welding-current wave forms obtained with the 300 welders when operated in their normal manner with their rectifiers are better from the standpoint of dwell than those obtainable with plaintiff’s commercial Oestreicher welders, and the rectifier-reactor circuit of the 300 welders achieves a greater degree of *54amperage control than would be possible with the same apparatus without rectifiers.
The Prior Art
A.C. welders consisting of a transformer and an external adjustable reactor had been known and used long prior to the Oestreicher patent.
Various ways of changing the reactance of a reactor for controlling arc-current amperage, including varying the number of turns on the reactor coil, varying the size of the air gap in the magnetic frame of the reactor, and varying the degree of saturation of the reactor’s magnetic frame by controlling the quantity of D.C. flux therein, were known and disclosed in the prior art.
The Frickey patent No. 1,539,044, filed in 1921, disclosed a transformer, a welding arc, and a reactor connected in the arc circuit in series with the transformer secondary coil and the arc terminals. Frickey’s reactor had air gaps interrupting the A.C. flux path. Frickey’s device controlled the arc current by varying the reactance of the reactor, and his patent specified several means therefor, including varying the number of turns on the reactor coil, varying the distance between reactor coils, and varying the length of the air gap.
Other prior art in the record confirms that amperage control reactors with air gaps in their A.C. flux paths were in widespread and conventional use prior to the Oestreicher patent. This prior art includes Weed patent No. 1,612,084 of 1926, Steinert patent No. 2,175,927, filed in 1936, and defendant’s prior commercial machines of the types designated Model 4B and Model TWR.
The Model 4B welder was in production by defendant as early as 1938, and defendant’s Model TWR was being marketed through plaintiff shortly before Oestreicher’s original disclosure which was dated October 2, 1946.
In 1916 Dr. Charles P. Steinmetz * taught that air gaps may be used in reactors and transformers to control wave shape distortion. By 1943 the use of air gaps in arc-welding reactors was so-commonplace that the Welding Encyclopedia of that year defined reactor in these terms:
“On an arc welder a reactor is an inductive coil of copper wire or strap surrounded by a laminated iron circuit provided with an air gap.”
The Weed patent No. 1,612,084 filed in 1926, shows an A.C. arc welder in which amperage of arc current is controlled by means of a saturable reactor assembly including a pair of rectangular reactor frames, each of which is provided with A.C. and D.C. coils and an air gap which interrupts both the A.C. and D.C. flux paths. The A.C. coils are connected in series between the welding are and a transformer or other current source. Weed taught that the arc-current amperage could be controlled by varying the quantity of direct current passed through the D.C. coils.
The air gaps in Weed’s reactors are clearly shown in each of the six reactor frames shown in the two forms of his invention which Weed presented in his patent drawings. The air gaps are not. referred to in Weed’s specification, but the disclosure in the drawing is clear. The British Weed patent No. 272,218 is-substantially identical to the U. S. Weed patent and was cited in Great Britain-against Oestreicher’s counterpart British application at a time when Oestreicher’s U. S. application was still pending in the United States Patent Office. The Weed patent was not cited by the United States. Patent Office.
At the trial plaintiff acknowledged that, the Weed drawing discloses saturable reactors with air gaps but contended this disclosure should be ignored as a mere error of the draftsman who prepared the Weed drawing, plaintiff’s contention being that the presence of air gaps in Weed’s reactor frames would defeat his stated objective of obtaining a welding current having a steep-sided, more or-less rectangular wave shape.
*55
Figure 1 of the above drawing shows one form of Weed’s apparatus. The drawing shows two iron core reactors designated by numerals 1 and 2. Coils 3 and 4 are supplied with direct current, and numeral 10 designates an auxiliary *56reactance inserted in the direct current circuit. Coils 8 and 9 are supplied with alternating current. Each of the six reactors appearing in figures 1 and 2 of Weed's drawing discloses an air gap. (The word “gap” has been added to the drawing by the court.)
Validity
The first question before the court is whether the Oestreicher patent, interpreted as plaintiff would interpret it, i. e., to include an air gap anywhere in the reactor where it will interrupt the A.C. flux, is valid. The eourt finds that so interpreted the patent is anticipated by Weed.
Weed disclosed an A.C. arc welder in which the A.C. amperage is controlled by means of a saturable reactor which is regulated by variation of D.C. current passing through its D.C. coils. Plaintiff distinguishes Oestreicher from Weed contending that Weed did not teach the use of air gaps with his saturable reactor.
Plaintiff admits Weed’s drawing shows air gaps. In Great Britain, plaintiff’s solicitor distinguished Weed from Oestreicher, asserting Weed’s air gaps interrupted both the A.C. flux and D.C. flux, where as Oestreicher’s air gap interrupts only the A.C. flux; but now plaintiff contends Weed’s air gap disclosures were accidental and inconsistent with his objective of operating above saturation to maintain a steep-sided flat-topped wave.
“The purpose of the present invention is to secure a quick reversal of the welding current corresponding to a much higher frequency when using the frequencies ordinarily used for electric power. This is accomplished by means of a more or less rectangular wave shape of current which may be produced by inserting in the are circuit two iron core reactors supplied with direct current excitation, the excitation being carried beyond the saturation point of the reactor cores.”
Plaintiff’s conclusion that an air gap in Weed would be inconsistent with his objective glosses over the fact that with the Weed apparatus, a wave of any steepness that can be obtained without an air gap can be obtained with an air gap by increasing the D.C. current.
Plaintiff’s Exhibit 9, Oscillogram 54, demonstrates a wave which is substantially as steep-sided and as flat as Weed’s flat-topped wave exhibited in his patent drawing (Figure 4, Wave i). An air gap, therefore, is not inconsistent with Weed’s objective.
Plaintiff’s conclusion that Weed’s air gap disclosure was accidental is based upon the assumption that Weed never intended his apparatus to be operated at less than saturation or without a flat-topped wave.
Operation of the Weed apparatus so that the A.C. current resulting will be of low amperage does mean the apparatus is operating below saturation and with a more or less triangular rather than rectangular wave.
The Weed apparatus with an air gap inserted in accordance with its drawing does operate successfully right down to the same low currents as can be obtained with the Oestreicher apparatus. By placing an air gap of the proportion dictated by the knowledge of the art at the time of Weed’s invention into his reactor, as his drawing discloses, the same result would be obtained as plaintiff now claims was Oestreicher’s initial discovery.
Weed’s drawing does not partially disclose an improvement which, if developed, would be the same as what plaintiff claims Oestreicher taught. To the contrary, the drawing’s portrayal of the gap disclosed everything necessary to accomplish the invention. Such a gap would not prevent the attainment of a more or less rectangular-shaped wave when Weed’s apparatus was operated above saturation but would permit operation below saturation without dwell.
This court finds that Weed’s disclosure of an air gap to be used in a saturable reactor was neither accidental nor in*57consistent with his objectives. Even if Weed had not fully appreciated the advantages of his air gaps, however, Oestreieher’s discovery would not be invention. General Electric Co. v. Jewel Incandescent Lamp Co., 1945, 326 U.S. 242, 66 S.Ct. 81, 90 L.Ed. 43.
Weed’s structure shows everything necessary to obtain the result plaintiff claims was Oestreicher’s teaching. Weed’s structure was inherently capable of being operated at low currents without objectionable dwell. Discovery of a new use for an old device is not invention. General Electric Co. v. Jewel Incandescent Lamp Co., supra; Anderson v. Phoenix Products Co., 7 Cir., 1955, 226 F.2d 191, 193.
“The novelty must be a novelty in the means or mechanical device, and not in the use to which the combination is put.” Wollensak v. Sargent, 1894, 151 U.S. 221, 227, 14 S.Ct. 291, 293, 38 L.Ed. 137.
Harnischfeger further distinguishes Weed because Weed teaches the use of an auxiliary reactance to eliminate the harmonics problem, whereas Oestreicher masks the harmonics with its air gap.
It first must be noted that Claim 7 of Weed provides for no auxiliary reactance or other device to prevent dwells other than the air gap disclosed in the drawings which is included in each of the six reactors there drawn. Without an air gap in a device such as described in Claim 7, the device would be inoperable because the dwells would be too long and thus would extinguish the arc.
Oestreicher does not avoid anticipation of Weed by eliminating from his device an element and corresponding function of such element. Stow v. City of Chicago, 1877, 23 Fed.Cas. p. 195, No. 13,512, affirmed 1881, 104 U.S. 547, 26 L.Ed. 816; Loewenbach v. Hake-Stirn Co., 7 Cir., 1899, 92 F. 661.
Weed knew how his device operated without the auxiliary reactance but sought to accomplish more than the triangular wave there produced. His object was a rectangular wave and the substantial elimination of harmonics. It was not invention to leave out the auxiliary reactance and be satisfied with a triangular wave with which Weed started.
Plaintiff’s counsel has asserted that Weed’s air gap and auxiliary reactance would be a duplicity of parts performing the same function.
The court finds, however, that the auxiliary reactance has its function of eliminating harmonics when D.C. is introduced into the reactor coils, whereas the air gap, in addition to preventing dwell from harmonics, prevents dwell when the apparatus is operated without D.C. excitation — an arrangement where the auxiliary reactance is totally ineffective.
Lack of Invention in the Oestreicher Patent
Plaintiff contends that it was invention for Oestreicher to use an air gap in a saturable reactor for welding with alternating current. Plaintiff asserts this was not obvious, its contention being that an air gap in such a reactor tends to make more difficult the D.C. saturation of the magnetic frame and, hence, would not have been used by a welding machine designer of ordinary skill.
As already noted, air gaps have long been commonplace in arc welding reactors, to such an extent that the definition of “reactor” in the 1943 Welding Encyclopedia included specific reference to an air gap.
In 1946 any welding machine designer of ordinary skill in the art would have known from his general knowledge that an air gap in a saturable reactor would render saturation of the reactor frame with D.C. flux more difficult. He would also have known that air gaps had proved essential in reactors used for arc welding. Since these design standards were mutually inconsistent, it would have been obvious to such a designer of ordinary skill that one of them must yield to the other. It would have been equally obvious to try each of the two alternative designs to determine which of the two inconsistent standards could best be de*58parted from. On thus trying the two choices pointed out by his general knowledge, such a designer would necessarily have learned the right answer with no more than two experiments. This would not have involved patentable invention but merely routine engineering. The subject matter as a whole of the Oestreicher patent, as plaintiff construes it, would have been obvious in October, 1946, to a person having ordinary skill in the art to which such subject matter pertains.
A patent is prima facie valid, and the burden of establishing invalidity of the patent rests on the party asserting it. 35 U.S.C. § 282. Although the patent office did not cite the most pertinent prior art references, comprising the Weed, Frickey, Steinert, and Burgess patents, the court cannot assume the patent office had not considered them. Anderson Co. v. Sears, Roebuck & Co., 7 Cir., 265 F.2d 755; Artmoore Co. v. Dayless Mfg. Co., Inc., 7 Cir., 1953, 208 F.2d 1, 4.
Nevertheless, the defendant has independently overcome the presumption of validity and established that if Oestreicher be interpreted as plaintiff claims, Oestreicher did not exhibit invention and was anticipated by Weed.
In light of the court’s construction of the Oestreicher patent as limited to the improvement of a saturable reactor with an air gap interrupting the A.C. flux and not interrupting the D.C. flux and the court’s decision on infringement, it is not necessary in this ease to pass upon the validity of the Oestreicher patent as so interpreted.
Infringement
File Wrapper History of the Oestreicher Patent
In the application for the Oestreicher patent filed July 8, 1947, plaintiff’s solicitor presented eight claims, of which Claims 2, 3, 4, 7, and 8 were directed to the saturable reactor welder shown in the patent drawing. (The other claims related to a welder of the flux-diverter type and were cancelled along with the part of the disclosure and drawings pertaining to that device.) Of the original Claims 2, 3, 4, 7, and 8, all were cancelled except application Claim 7 which, with minor modifications, became Claim 1 of the patent.
During the prosecution of the application, plaintiff presented five additional claims, two of which were cancelled, the-other three becoming Claims 2, 3, and 4, respectively, of the patent.
Claim 7, as originally presented, called for “a three-legged reactor frame having side legs and a central leg interrupted by an air gap.”
Every claim which did not require the air gap to be confined to the central leg but permitted it to be located anywhere-in the A.C. flux path of the reactor frame was rejected by the patent office and can-celled by plaintiff.
The only significant difference betweencancelled application Claim 4 and patent Claim 1 is that the former permitted the air gap to be anywhere in the A.C. fiux path of the reactor frame, whereas the latter limits the location of the air gap to the central leg of a three-legged reactor frame. To interpret Oestreicher’s Claim 1 to cover a welder with an air gap anywhere but in the central leg of a three-legged reactor frame would give patent Claim 1 substantially the same scope as application Claim 4 which was cancelled.
Almost four months before plaintiff cancelled the broad claims of the United 'States Oestreicher application, the British Patent Office had cited the Weed British patent against plaintiff’s British Oestreicher application. As already noted, the disclosures of the Weed British and United States patents are identical.
The file wrapper history manifests plaintiff’s intent to limit all claims of the Oestreicher patent to reactor constructions wherein an air gap is located in the central leg of a reactor frame having three legs or an odd number of legs greater than three.
*59 Oestreicher’s Conception of the Scope of His Invention
In his first written description of his ■device, Oestreicher specified that the D. ■C. coils should be placed around the outside legs so that “the air gap therefore ■does not influence the number of A.T. (ampere turns) required for a desired D. ■C. flux value” and stated that “the only possible claims might be” a reactor “of the described features.” This statement reveals that Oestreicher conceived his invention to be a reactor having an air gap which interrupted the A.C. flux but not the D.C. flux. An air gap otherwise located would influence the “number of A.T. required for a desired D.C. flux value.”
The Scope of Foreign Counterparts of the Oestreicher Patent
Concurrently with the prosecution of the Oestreicher patent application in the United States, plaintiff filed and prosecuted counterpart applications in Great Britain, Germany, and Sweden. In each •country plaintiff sought a broad patent monopoly on saturable reactor arc welders, but in every case was ultimately restricted to a narrow claim limited to the precise reactor construction shown in the Oestreicher drawing. As already noted, the Weed patent was cited against the British Oestreicher application by ■the British Patent Office, and plaintiff distinguished it on the ground that Weed disclosed air gaps interrupting both the D.C. and A.C. flux paths of his reactors as contrasted to the Oestreicher reactor in which' the central leg air gap interrupted only the A.C. flux path.
Facts Bearing Specifically on the Infringement Issue
Defendant’s devices do not use a unitary multi-legged reactor frame having a central leg, such as all the claims of the Oestreicher patent specify. Oesstreicher’s reactor frame is three-legged and integral, and all his claims call for “a” frame having side legs and a central leg. Defendant’s welders all have two rectangular frames which are physically separate from one another. In Oestreicher’s frame the D.C. flux passes only through the outer legs and across the horizontal return portions of the frame but not through the central leg. In defendant’s welders the D.C. flux traverses the entire perimeter of both rectangular frames in the same path as the A.C. flux. In Oestreicher’s unitary frame, it is possible, with an air gap in the central leg, to interrupt the path of the A.C. flux without interrupting the D.C. flux, but this cannot be done in defendant’s welders with any number of gaps, wherever situated.
Ordinarily a change in location of an element is not sufficient to avoid infringement, but where invention depends upon the particular location of an element, any material variation from that location will be sufficient to avoid infringement. Lehman Co. of America, Inc. v. Kroll Bros. Co., 7 Cir., 1944, 139 F.2d 391.
At the trial plaintiff contended that a reactor assembly having two juxtaposed rectangular frames with a coil wound around the adjacent legs of the two frames was a “three-legged reactor.” Attempting to support this position, plaintiff offered certain technical articles wherein the expression “three-legged reactor” was used to describe a pair of saturable reactors mounted on juxtaposed rectangular frames. Those articles themselves, however, reveal that such a two-reactor assembly is not the electrical equivalent of a true three-legged reactor assembled on a unitary frame but is in fact “different and inferior.”
“In determining the question of infringement, the court is not to judge about similarities or differences by the names of things, but is to look at the machines or their several devices or elements in the light of what they do, or what office or function they perform, and how they perform it.” Hydraulic Press Mfg. Co. v. Williams, White & Co., 7 Cir., 1947, 165 F.2d 489, 492.
Further, such an assembly is incapable of the only distinctive function of Oes*60treieher’s welder, i. e., interrupting the A.C. flux path with an air gap without interrupting the D.C. flux path. The court finds that a reactor assembly on two rectangular frames physically separated does not have a three-legged reactor frame as that term is used in the Oestreicher patent.
To give plaintiff’s patent claims the meaning that plaintiff asserts for them would make a “three-legged reactor” out of any reactor assembly having two rectangular frames wound with A.C. and D.C. coils, no matter what the physical orientation of the frames. Such a reading of plaintiff’s claims is inadmissible because it is in conflict with Oestreicher’s disclosure and because it disregards the important differences in function which exist between a unitary three-legged frame and a separate pair of rectangular frames. Additionally, plaintiff’s reading of its claims is inadmissible because such a reading makes the claims directly descriptive of the Weed prior-art patent.
Weed disclosed a saturable-reactor arc welder having a pair of reactors, respectively assembled on separate rectangular magnetic frames, with an A.C. coil and a D.C. coil on each frame. Each of Weed’s rectangular frames contained an air gap, and, as Professor Matsch, plaintiff’s expert, conceded, it • makes no difference where in a rectangular frame the air gap is located — the function is the same.
Neither defendant’s 250 welder nor its 300 welder has a magnetic frame with a central leg within the meaning of Oestreicher’s patent and claims. Even if the portions of defendant’s magnetic frames that carry A.C. coils be considered to correspond to Oestreicher’s “central leg,” as plaintiff contends they should be, defendant’s air gaps are not located therein. Defendant’s air gaps are in those parts of its magnetic frames which were defined in the Oestreicher patent as “return portions.”
Fairness to the public necessitates our giving the meaning to the language which the inventor himself gave when he explained his invention. Northwest Engineering Corp. v. Keystone Driller Co., 7 Cir., 1934, 70 F.2d 13, affirmed 294 U.S. 42, 55 S.Ct. 262, 79 L. Ed. 747.
Defendant’s welders are constructed on principles taught by the prior art and do not embody any structure or function that was novel with Oestreicher. The prior art taught that arc-current control in an A.C. welder could be achieved by a pair of saturable reactors wound on rectangular frames with the D.C. coils relatively reversed on the two frames with respect to the A.C. coil or coils (Weed patent), that the reactor frames should have air gaps (Weed, Frickey, and Steinert patents, Miller 4B welder, and P & H TWR welder), and that direct current can be used to saturate each reactor frame (Weed and Burgess patents) - The rectifier-reactor cooperative arrangement used in defendant’s 300 welders comes from the Logan patent No. 2,259,-647.
“No matter what the discovery asserted or claimed in the patent, such assertion or claims must be construed to be limited so as to exclude the prior art.” Smith v. Mid-Continent Inv. Co., 8 Cir., 1939, 106 F.2d 622, 624.
Claims can be no broader than the patentee’s actual invention. Texas Co. v. Globe Oil & Refining Co., 7 Cir., 1955, 225 F.2d 725.
All elements of similarity between defendant’s welders and the welder of the Oestreicher patent are found also in the prior art and, hence, do not tend to show infringement.
Plaintiff has not sustained its burden of proof on the issue of infringement.
Defendant’s 250 series welders do not infringe any claim of the Oestreicher patent, nor do defendant’s 300 series welders infringe any claim of the Oestreicher patent.
Defendant’s counsel is directed to prepare findings of fact, conclusions of law, order for judgment and judgment, submitting them to opposing counsel for approval as to form only.
Steinmetz, Theory and Calculation of Alternating Current Phenomena 118-32 (5th Ed. 1916).