The Condict patent, No. 393,323, dated November 20, 1888, upon which this suit is brought, is for an improved switch or controller for electric cars. The patent has been the subject of extensive litigation. It has been construed and sustained in three carefully considered opinions in the Second circuit: By Judge Townsend, in the circuit court, on final hearing, in the case of Electric Car Co. v. Hartford W. H. Ry. Co. (C. C.) 87 Fed. 733; by Judge Racombe, in the circuit court, on motion for preliminary injunction, in the case of Electric Car Co. v. Nassau Electric R. Co. (C. C.) 89 Fed. 204; and on appeal in the same case, by the circuit court of appeals, in an opinion by Judge Shipman, reported in 33 C. C. A. 420, 91 Fed. 142.
In the suit at bar, the complainants filed a motion for a preliminary injunction against three types of controllers used by the defendant; and an injunction was granted against two of these. With respect to the third, the original motion was withdrawn; and the present hearing was had on the renewal of the motion to enjoin this controller, which- is known as “Type Nó. 38.” On the original motion, after the defendant’s affidavits were filed, it was found that there was a material disagreement of fact as to the construction of this controller. This disputed question has been eliminated; the complainants now admitting that No. 38 controller is constructed substantially as shown by the defendant’s affidavits on the original motion.
The complainants’ proofs on the original motion brought this controller fairly within the Condict patent as interpreted by the courts *987in the decisions referred to; and the testimony of the experts in support of the motion was based upon the same theories, and proceeded along the same general lines, as in the prior cases. If the complainants could have maintained their original position, the determination of this motion would have been comparatively free from difficulty. But the case which is now presented is very different. When we exclude from the defendant’s apparatus the important feature which it was assumed by the complainants to contain on the original motion, we have a structure towards which the Condict patent bears a very different relation from what it did in the other structures which were held by the courts to infringe that patent. On this renewal motion, the complainants have been confronted with a situation where it has become necessary, in order to show infringement, to advance new theories respecting the scope and meaning of the Condict patent, and to ask thi£»court to give a broader construction to the patent than it has heretofore received.
The general principles underlying the subject-matter of the Condict patent are not difficult to understand; but, when we descend to particulars, questions may arise which are involved, obscure, and troublesome; and when such questions are presented for the first time on motion, and the only proofs are the affidavits of experts, who advance inconsistent and contradictory theories in relation thereto, it becomes almost impossible for the court to reach any conclusion which rests upon the satisfactory determination of such questions, whatever amount of time and labor may be given to the subject. On motion for preliminary injunction, and in the absence of the cross-examination of witnesses and the full and orderly proofs presented on final hearing, the court should not be asked to pass for the first time upon the conflicting theories of experts relative to the mode of operation in certain positions of a complicated electrical apparatus like a controlling switch for motor regulation. When the complainánts go beyond what the courts have decided in prior adjudications, and base their motion upon new and disputed theories respecting the Condict patent, or upon a new and enlarged construction of the patent; issues are raised not pertinent to a motion for preliminary injunction, and whose determination, properly, should be reserved for final hearing.
The court, without going further, might deny this motion with the simple statement that it appears upon an inspection of the record that to hold infringement in the first particular charged is to overrule the construction given to the Condict patent by the courts in prior suits; that to hold infringement in the second particular charged involves a new, extended, and doubtful construction of the patent, and the determination of electrical questions upon which the experts disagree; and that to hold infringement in the remaining particular charged is to pass upon the conflicting opinions of experts as to the action of the defendant’s controller at a particular time, in relation to which it is not entirely clear that the patent has any application. But, on consideration, it seems to me, although some of the questions raised can only be properly determined on final hearing, in view of the importance of the case and the carefully prepared *988and able briefs óf counsel, that the disposition of the motion-calls! for a somewhat fuller statement and discussion of the general .ques- ■ tions presented.
The ordinary electric car, like the one shown in the Condict patent, contains two motors. These motors consist of two armatures and two fields and their connections. In starting-, and at slow speed, the two armatures and the two fields are connected in series, or in a single path. At full speed, the two armatures and the two fields are connected in multiple or parallel paths. The motor changes from series to parallel involve a temporary disorganization and rearrangement of the circuit. These changes may be quite sudden, or by gradual and progressive steps. During this transition period, unless the flow of current from the trolley, or generator is checked, the circuit changes will be too severe to be borne by the motors, and there is great danger from spar.king. The main purpose of the controller or switch is to regulate the supply of current delivered to the motors during the shifting of motor connections from series to parallel, or vice versa.
The principle of motor regulation through a controlling switch is based upon interposed resistances, which reduce the flow of current. These resistances are wire coils, which are cut in and out of the circuit. These coils may be either entirely separate from the motor coils, or they may be the motor coils themselves, or they may consist partly of motor coils and partly of external coils. The first form, in which the resistances are wire coils external to the motor, is known as the “rheostatic,” or “external,” method of control. The second form, in which the motor coils of the armatures and fields are substituted for external coils, is known as the “series parallel,” or “internal,” method of control. The third form, in which both external and motor coils are utilized, is known as the “mixed,” “compound,” or “composite,” method of control. The rheostatic, or external, method was open to the objection of waste of energy, since the current was converted into heat and lost. Hence the resistance is called “dead” resistance, as distinguished from the “live” resistance of the motor coils. This method went into extensive use. The series parallel method proved to be impracticable for general use, for the reason that, when the circuit changes were numerous, involving only slight changes of internal resistance, the apparatus was too complicated, and when the circuit changes were large, involving large changes in the resistance, it was too severe upon the motor, and sometimes dangerous to the motorman. The combined, or mixed, method of control possesses the advantage of economy over the rheostatic method, and the advantages of the reduction of sparking and the avoidance of other dangers incident to the series parallel method. Condict was the inventor of this last method, and his invention is embodied in the patent in suit. The Condict patent describes a controlling switch in which motor, or internal, resistance is supplemented by external, or rheostatic, resistance introduced at the time of changing the motor connections. When the motor resistance is great, the external resistance introduced is small; and when the *989motor resistance is small, the external resistance introduced is great. As the patentee says, in his specification:
' “I liare conslrueted my switch so that at the time of changing the connections I insert resistances more or less great according to the resistance of the motor connections; that is to say, if the motor resistance is great, the auxiliary resistances would he small, and rice versa.”
It was in this way that Condict for the first time utilized both the old methods of control in a single switch. The courts have held that the Condict patent contains a main invention and a minor .invention. The main invention is covered by claims 27, 28, 29, and 31, and is for the introduction of external resistance during the shifting of the motor connections. The minor invention is covered by claims 20, 21, and 22, and is for the introduction of such resistance without changing the motor connections, when combined in a switch which “embodies the main invention.”
The essence of the Condict invention is the insertion of external resistance during the change of motor connections from series to multiple parallel, or vice versa; and no apparatus infringes the patent unless it contains this feature. All the controllers which heretofore have been enjoined by the courts have infringed the main invention of the patent by the insertion of external resistance coils during the shifting of motor connections; and on the original motion in this case the complainants believed that the defendant’s controller plainly and on its face introduced such resistance. A controller which depends upon external or rheostatic resistance for motor regulation is not within the Condict patent. A controller which depends upon motor resistance for motor regulation is not within the Condict patent. A controller which combines both external resistance and motor resistance is not within the Condict patent, unless embodied in a system in which external resistance is introduced during the shifting of motor connections.
It is contended on this renewal motion that the defendant’s controller infringes the Condict patent in three particulars: Hirst, in positions 1, 2, and 3 ; second, in positions 43 and 61; third, in passing from position 5 to position 6, from position 6 to position 5, and from position 7 to position 6. We will consider these in their order;
Hirst. In positions 1, 2, and 3 external resistance coils are introduced into the circuit, but there is no change of the motor connections. This is the time of starting and slow speed, and the only change in the positions of the controller is the gradual cutting out of the external resistance coils. The motors remain in series, and the shifting of the connections does not begin until after position 4 is reached. As the introduction of rheostatic resistance for motor regulation was common before the Condict invention, there was no patentable novelty in the mere insertion of such resistance without change in the motor connections. This charge of infringement, therefore, cannot be sustained, unless it is also found that this controller infringes the main invention of the patent.
' Second. Positions 43 and 61. are transitional positions during the change of motor connections from series to parallel; and, if external resistance is inserted in these positions, there is a plain infringement *990of the Condict patent. In position 43 the external resistance coils-are short-circuited, or cut out, and the field of No. 1 motor is also cut out; and the only resistances which remain in circuit are the armature coils of No. x motor, and the armature coils and field coils of No. 2 motor. The same is true of position 61, except that the armature of one of the motors is cut out, instead of the field. Upon the face of things, certainly, it is not apparent how the cutting out of all external resistance and a part of the motor resistance at these positions, during the progressive steps incident to the change of motor connections from series to parallel, is the cutting in of the external resistance coils of the Condict patent. Although no external resistance is apparently introduced into the circuit in positions 43 and 61, the complainants now maintain that, since the withdrawal of the original motion, they have discovered, upon careful inspection and in a form disguised and hidden, the presence of such external resistance in the motors themselves. This new discovery of complainants’ experts is based upon the following line of reasoning:
In a controller which changes the circuit connections of two motors from series to parallel, each motor must be regarded as an entirety by itself and external to the other motor. This proposition lies at the foundation of the whole argument. The Condict patent, broadly speaking, is for the introduction of dead resistance to a single live motor, and not necessarify for the introduction of such resistance to supplement the internal resistances of two motors during the shifting of their connections from series to parallel. Dead resistance may be either wire coils external and foreign to the motors, as shown in the Condict patent, or it may be the motor coils themselves. Dead resistance is any resistance which forms a part of the circuit and which is not at the time acting as a motor; and the armature coils and the field coils of the two motors become the inserted external rheostatic coils of the Condict patent when they are in circuit and not acting as motors. When, during the shifting of motor connections in a series parallel system, the field or the armature of one of the motors is cut out, the motor is killed, and its'armature coils or field coils which remain in circuit become dead resistance to the other live motor. Such armature coils or field coils are the external resistances of the Condict patent, because they are external to the other live motor; and they are the inserted resistances of the Con-dict patent, because they are, in effect, introduced into the circuit as dead resistance to the other live motor. When, in the harnessing and unharnessing of motor connections during the transition period from series to parallel, a live motor is converted into dead resistance by disconnecting part of it and permitting it to remain in circuit, this, in substance and effect, is the introduction of the external rheostatic coils of the Condict patent. In cutting out the field of one motor and leaving the armature in circuit to act as a dead resistance to the other live motor, the defendant’s controller infringes the Condict patent at position 43; and in cutting out the armature of one motor and leaving the field in circuit to act as a dead resistance to the other live motor, the defendant’s controller infringes the Condict patent at position 61.
*991This theory of the Condict patent is open to many and serious objections. It ignores the fundamental distinction, set forth in the specification and shown in the drawings, between the inserted external resistance and the internal resistance of the motors, between rheostatic resistance coils and motor resistance coils, on which the whole invention rests. It makes what is external to a thing internal, and what is to be inserted a part of the very thing itself. It confounds things, terms, and methods which'have always been recognized as distinct in the controller art. It confuses and' breaks down the recognized distinction between the series parallel method of control and the rheostatic method, by holding that both methods may be included in the series parallel system. It extends the scope of the Condict patent from a mixed system of control comprising both these metm ods to a single system embracing only one. It declares that one form of series parallel connections without the insertion of external resistance coils is without the patent, and another form of series parallel connections without the insertion of external resistance coils is within the patent. It undertakes to divide the two motors into separate entireties, although both form the parts of a single system of motor regulation, always known as something distinct from other systems. The series parallel method of control consists in connecting or disconnecting the two motors, or their several parts, in the transition from series to parallel, or vice versa, in any way found most advantageous. The transition steps are gradual and progressive, and the connections vary in many ways in different controllers, as this record abundantly shows. The form the connections assume is immaterial, because they are all simply separate ways, differing in details, of applying the same method. The patentee never contemplate^ that his invention covered the division of the two motors into separate entireties, and the insertion of a part of the motor resistance of one motor into the other motor. It is not until we go outside both the motors, and seek to supplement their resistance by the introduction of an artificial external resistance, that we encroach upon the Condict patent. .Claim 31 clearly states the main invention, as follows :,
“The combination of two motors, a source of electric power, a motor circuit, a switch for coupling tho coils of the motors in series or multiple to vary their internal resistance, a resistance, a switch to insert the resistance when the motor switch is being shifted, and a connection between said switches to operate both simultaneously.”
Whether the armature coils or the field coils of one motor act as dead resistance to the other live motor is a disputed question of fact upon which the experts differ; but, assuming that the complainants contention is correct, it still remains true that the armature coils in one case, and the field coils in the other, are motor resistances, as understood in the art, and hence it becomes unimportant whether they are dead, half dead, or alive. Although rheostatic resistance may be dead resistance, and motor resistance may be generally live resistance, the Condict patent is not for any system of motor control which combines dead resistance with live resistance, but is for a system which combines external resistance with motor resistance. *992In his specification, Condict refers to “internal resistance”, and “motor resistance” as distinguished from “auxiliary resistances.” ' He refers to the series parallel method, “in which the motors [not motor] are regulated by varying their internal resistance, which may be done by connecting up their coils in different ways.” He then states that 'he has constructed a switch in which he inserts resistances “more or ' less great according to the resistance of thé motor connections.” The patent shows the motor connections of the two motors, the external rheostatic coils, a motor switch, and a resistance ' switch, and how, by a single movement of the handle of the controller drum, the external resistances are cut in and out at the proper time to secure the protection of the motors. It shows that the patentee means, by “internal resistance,” all the motor resistance which is available in shifting the motor connections of two motors; and that he means, by “auxiliary resistance,” an external and supplementary resistance. The patent is not broadly for the introduction of dead resistance, whether it be motor resistance or external resistance. The utilization of motor resistance is the series parallel method of control. The utilization of external resistance is the rheostatic method of con-' trol. The Condict invention is for a combination of both methods. He was not the inventor of either method, but of a mixed system which ittilized both methods. To construe the Condict patent to cover a system of motor resistance, or the series parallel method, without the insertion of external rheostatic resistance, finds no justification either in the prior decisions of the courts, or in the state of the art, or in the language of the specification.
Third. In passing from position 5 to position 6, and from position 6 to position 5, the circuit is open at all available points. In passing from position 7 to position 6, the circuit to one motor is open, and all the contacts are open except those which control the circuit to the other motor. The infringement consists in the alleged use of rheostatic resistance at this time of open circuit. Whether external resistances can be utilized during the period of open circuit is a problem upon which the Condict patent throws no light, nqr any decision of the courts respecting infringing controllers. Apparently there is no current flowing at this time, and therefore no current to 'check, because the contracts are broken at every available point. The insertion of external resistance at the instant of open circuit is manifestly not the usual and regular manner of introducing such resistance. The Condict controller does not seem to contemplate the use of such resistance at such a time, for no position of open circuit is found in that controller during the change of motor connections. To sustain infringement in this particular, the complainants contend that the introduction of external resistance during open circuit tends to minimize sparking, and that this was one of the main objects of the Condict invention. It is true that the prevention of sparking was, in the mind of the patentee, perhaps the chief object of his invention, but it is also true, as pointed out by Judge Townsend in the Hartford Railway Case, supra, that “the subsequent development of the art necessitated the introduction of other means to prevent spark*993ing,” such as the magnetic blow-out, and that the Condict device is not now used to prevent sparking, but “to regulate the current.”
On the original motion, it was not charged that there was any use made of external resistance in defendant’s controller at the time of open circuit. On this renewal motion, in the moving affidavits, the point is taken in a general way, without any statement of the reasoning or theory on which it is based. It is not until the complainants’ affidavits in reply that we find, for the first Lime, a theory advanced which undertakes to prove that the resistance of the Condict patent is utilized when the circuit is broken at all available points. To this evidence the defendant says it has had no opportunity to reply, and consequently the. defendant’s evidence is limited to answering in a general way this charge of infringement. The answer, though general, seems to have much force, and is substantially as follows:
At the time of open circuit, to break the circuit at all contacts available is the common and ordinary practice known to every electrician; its purpose being to divide the arcs and make them as small as possible at each break, and so prevent the burning of the contact points. There is no current flowing at this time upon which the resistance can operate, because the circuit is broken at every point available, — in nine points at full open circuit in passing from positions 5 to 6, and vice versa; and in five points when the circuit is half open in passing from positions 7 to 6. All these breaks take place simultaneously. The short circuit around the rheostat is broken at the same instant as all the other breaks. The reason why the resistance is not cut out, instead of short-circuited, is that it would require an additional contact finger, causing thereby a slight enlargement of the controller drum. In a word, there is no current flowing at this time, and, if there were, its action would be harmful; hence, it follows that the resistance is not utilized to dam back the current and minimize sparking.
The complainants do not deny that, if the break around the resistance were in the main circuit, and the breaks were all made simultaneously, there would be no flow of current through the resistance. Their whole theory rests upon the fact that this break is in the subbranch, or short circuit around the resistance, and not in the main circuit. They claim that this has the same effect as if the break at this point were made before the other breaks, — the effect being that the arc at this point is feebler and less persistent than the other arcs, and is extinguished sooner, thereby throwing the current, which still continues to flow through the other and more persistent eight arcs, into circuit with the resistance; and the resistance then acts to dam back the current, and so helps to minimize sparking. It would serve no good purpose at this time to enter more fully into the reasoning by which the complainants undertake to demonstrate the truth of this theory. It is possible that arc No. tt in defendant’s controller is extinguished in advance of arcs No. 6 and No. to; that these latter are extinguished before the remaining arcs, the effect being to reduce sparking; and that this is primarily due to the break in the short circuit around the rheostat; and, further, that this break causes the introduction of external resistance into the current with*994.in the meaning and purpose of the Condict patent. But it is also possible that there is no substantial current flowing at this time of open circuit, or, if a small' quantity, that its effect is harmful, or that the period of time is so inappreciable that the resistance has no effect, or that the resistance inserted at this time is outside of the scope and design of the Condict patent.
The decision of these questions, so far as they may prove material, must remain for the full proofs presented on final hearing. Rooking at the Condict pátent along its broad and plain lines, I am not satisfied, upon the evidence before me, that the opening of the •short circuit around the rheostat at the time of open circuit is the •introduction of external resistance into the circuit during the change of motor connections, which is the material invention of the patent. The motion for a preliminary injunction is denied.