This is an appeal from a decree dismissing the complaint of appellant, Henry, upon the ground that the Lyndon patent, No. 695,220, issued March 11, 1902, here involved, was not infringed.
The patent describes the invention generally as improvements in electro-mechanical water wheel governors. In his specifications the inventor sets forth that governors used to regulate the water supply to the water wheel in general operate only to open or close the -water wheel gate, thereby allowing of the admission of a greater or less supply of water. He continues
“Now, the first effect of such opening or closing of the gate, owing to the inertia of the water, is always the opposite to that which it is desired to bring about — i. e., the opening of the gate operating to momentarily cause less velocity of water at the wheel, owing to the great orifice the water has to flow through, and, vice versa, the closing of the gate operating to momentarily cause an increase of velocity, owing to the contraction of the orifice.”
To overcome these opposite effects, the patentee provides a by-pass inserted into the penstock or flume at a point near the water gate and a gate in the by-pass controlled by the same governing mechanism that controls the water gate and operating to allow a greater or less flow through the by-pass, according as the water gate is being closed or opened. Other principal features enumerated in the specifications “relate to means for preventing excessive action of the governor,” and to the control of the governor by a dynamo driven by the water wheel.
[ 1 ] The use of electric governors for water wheels was old in the art, as were the uses of mechanical governors; but the appellant contends that prior to the coming of Lyndon into the field there prevailed in the électro-medhanical water wheel art an absolute want of sensitive and accurate speed governing — that is, the governing of the generating apparatus driven hy water power — and that often there were fluctuations of wide extremes occurring in the supply of electrical energy produced by any electro-mechanical generating unit. Obviously it is desirable to retain a constant speed in the use of power required from a water wheel, and inventors have' sought to maintain a balance or equilibrium between the power of the water projected upon the buckets or blades of the water wheel and mechanical power being taken from the water wheel.
Appellant says that Lyndon pointed the way to securing in a water wheel governor the automatic return of the controller to normal position to interrupt the governing action before it had overrun, or the introduction of means such that the governor, when in the act of moving the water gate to a new position, would be prevented from moving the gate “a little too far and then later moving it back a little farther than necessary.”
The claims which appellant quotes as bearing upon the breadth and import of the invention are as follows:
3. In a water wheel governor, the combination with a water gate operating, shaft, and means for operating same in either direction to govern the water wheel, of a controller for said operating means, responsive to changes of speed of the water wheel, a returning device for said controller provided with a clutch connection to said operating shaft, and means, actuated by *771said controller on movement thereof from normal position to engage said clutch with the said shaft, so as to cause the return of the controller to normal position and interrupt the governing action before it has overrun the proper amount, substantially as for the purpose set forth.
4. In a water wheel governor, the combination with a water gate operating shaft, a driving shaft and reversing clutch gear, adapted to turn the water gate operating shaft in either direction, a controller, responsive to changes of speed of the water wheel and controlling such reversing gear, and a returning device for said controller provided with actuating means controlled by said controlling moans to return the controller to inoperative position, so as to prevent excessive movement of the governor.
6. In a water wheel governor, the combination with means for operating the wafer gate in either direction, a by-pass for the water wheel, and a valve controlling said by-pass, of means connected to the water gate operating means and operating the by-pass valve inversely to the operation of the water gate.
7. In a water wheel governor, the combination with means for operating the water gato in either direction from, normal position, a by-pass for the water wheel, and a valve for such by-pass, of means connected to the water gate operating means and adapted to operate the by-pass valve from, normal position in either direction, so as to control such valve inversely to the control of the water gate, during the governing action of the water gate, and means for returning the by-pass valve to normal position on completion of governing movement of the water gate operating means.
8. In a water wheel governor,, the combination with a shaft for operating the water gate in either direction from normal position, a by-pass for the water wheel and a valve for such by-pass normally held in partly open position, of an operating device for said valve provided with means for returning the valve to normal position, a clutch, adapted to connect said operating device for the by-pass valve with the water gate operating shaft to control the by-pass valve inversely to the water gate, reversing means for operating the water gate operating shaft in either direction, a controller, responsive to the speed of the water wheel and controlling said reversing means, and means operated by said controller to bring the aforesaid clutch into operation and to release said clutch when the governing action is effected.
It is contended that the by-pass combination is broadly and basicly claimed in claims 6, 7, and 8 of the patent, and the combination preventing overrunning is broadly claimed in claims 3 and 4, and that the structures used by the appellee contain and utilize the substance of the invention within the meaning of the leading and controlling decisions under the patent law. Stating the contentions more concretely, they are, that the relation between the by-pass valve and the water gate was new with Tyndon, as also were the relations between the features preventing overrunning of the governor.
The whole problem of speed control and use of electro motive force in water wheel engines appears to he complex. To explain in detail the many features involved in the case would require very lengthy statements of the evidence. We shall therefore attempt no more than to refer to the more important points upon which our judgment is founded.
Tyndon’s device contains the following essential elements: A speed-sensitive device, comprising a dynamo and connected parts, is connected by mechanical means to the prime mover, and therefore varies in speed responsive to variations in the prime mover. Variations in speed of -the speed-sensitive device produce a change in degree of magnetization of a solenoid, which results in the movement of the solenoid core, which movement produces certain electrical contacts, the closure of which serves to energize certain magnets, the energizing of which serves *772to set in operation a train of movements intended to bring about a regulation or control in operation of the speed of the prime mover. Lyndon also discloses what is called a by-pass and by-pass valve intended to move inversely to the movement of the main gate; the purpose being to minimize the rate of change in the velocity of_ the water in the main conduit flowing to the prime mover and eliminating variations of pressure in the same during the period of governing.
Figure 1 is a perspective view of Lyndon’s governor; Figure 2, a plan view showing certain detailed parts; Figure 6 illustrates in detail a part of the controlling device:
*773
*774
It is plain that if a prime mover is operating, for example, at half load, and there comes a sudden demand for full load, the water for the increased demand will not be present at or near the prime mover unless some special provision is made for supplying it at that point; furthermore, if the prime mover is operating at full load, and there is a sudden rejection of one-half the load, and the necessity of closing the gate to a corresponding amount, the excess water must be taken care of in one way or another, because the velocity in the main conduit cannot be.instantaneously arrested without danger of a serious increase in pressure. To overcome these troubles, Eyndon places in a by-pass just hack of the main gate a by-pass valve normally held in half-open position. Experts testified in effect that the water which flows along the main conduit then goes partly to the wheel and partly through the partly open by-pass. In case of sudden demand from half load to full load, enough water will.be available in the conduit itself, and it will *775be only necessary, in order to make it available, for the wheel to close, or partially close, the by-pass valve, thereby preventing the water from wasting through the by-pass and turning it to the wheel; and, on the other hand, in case of a sudden change from full to half load, the excess water which is going along the main conduit may be disposed of by simply opening the by-pass valve wider, thereby allowing the excess water to waste through the by-pass, while the reduced amount of water as controlled by the main gate will pass to the wheel. In the Lyndon patent there must be a coincident movement of the by-pass and of the main gate, and that Lyndon disclosed such is evidenced by the following quotations from his patent:
•‘I provide a by-pass inserted into the penstock or flume at. a point near tlie water gate, and a gate in the said by-pass controlled by the same governing mechanism that controls the water gate, and operating to allow a greater or less flow through the by-pass, according as the water gate is being closed or open.”
Again:
“When the gate is operated as above described, the lever J¡3 is moved to close the contacts Jf6a, .100, 101; this closure being effected whatever tlie direction of the movement of the controlling lever SO by reason of the pin and curved slot connection between such lever.”
Again:
“Consequently the by-pass valve will be turned toward open or shut position according to whether tlie gate is closing or opening for the purpose above stated.”
And further:
“When the governor acts to close the main gate, the compensating device will open more widely the by-pass. The rapidity with which the valve in the by-pass opens is such that the increased volume of water which it allows to pass through is proportional to the decrease in area which the main gate effects by reason of its closing. Should the main gate open, the reverse action takes place.”
And also:
“It is obvious that the by-pass arranged as described, opening or closing in a manner opposite to that in which the main gate opens or closes, will, if properly adjusted, admit of the main gate being rapidly operated and the governing of the water wheel quickly accomplished.”
In this connection it is well to say that in the, opinion of several highly qualified witnesses there was serious confusion in certain of the language of the' Lyndon patent, and particularly in the use of the term “returning device.” In the specifications of the patent Lyndon refers to — •
“the rod 25, discs 22 and 23 and the controlling clutch magnet 32 as constituting a returning device for preventing the governor from overrunning— that is, moving the water wheel gate a greater distance than is actually necessary for proper regulation; this necessitating a second movement of the gate in an opposite direction, which in turn may overtravel and require the gate to be moved back again.”
*776Professor Durand, of Stanford University, testified that he concluded that the rod 25 was considered the returning device, because Lyndon refers to a “returning device” consisting of a rod 25 connected by a pivoted link or connecting rod 25a with the disc 22, and because in discussing the general operation he says, “while springs 29, 29, are placed between collars 29a, 29a, on the rod,” etc. Other portions of the specifications mention'the magnet as apparently claimed independently of the “returning device,” and there would seem to be a distinguishing of the rod from the means involved in the movement of the rod.
The best evidence is that the element described as a controller in claims 3, 4, and 8 refers to the lever 26. This is gathered from the language in the specifications, where Lyndon refers to the exertion of pressure “on the controller 26 to return it to normal position”; also from his reference to “the lever 26, which acts as a circuit controller”; and elsewhere in claims the “actuating means” is regarded as distinguished from the controller itself. In the claims just referred to the solenoid is not intended apparently as the controller.
The “means connected to the water gate operating means” described in claims 6 and 7, and the element in claim 8 described as “an operating device for said valve,” are understood to comprise the drum or sheave wheel 51p with its immediate attached parts, 56 and 57, with ropes 51, 52; and the “means for returning the hy-pass valve to normal position” elements described in claims 7 and 8 refers to the weights 70 in the drawing. Professor Durand said that the function of the dashpot was to retard or control the return slowly and easily to its position of rest, but its function is not to determine or produce such return.
Much could be said to show why we are satisfied that the theory of the Lyndon patent pertaining to the governing of the hydraulic prime mover implies the provision of a by-pass valve in normal position half open arid the movement of such valve in either direction inversely to the movement of the main gate, but it would extend the discussion too far. The movement of the valve is in either direction from its normal position. This language from the specification is helpful in understanding the situation:
“Consequently the by-pass valve will be turned toward open or shut position according to whether the gate is dosing or opening for the purpose above stated. Normally, the gate or valve in the by-pass will be half way open, so that the amount of water flowing through the by-pass and around the wheel without doing’ work will be half the amount which the by-pass is capable of carrying.” . '
The inventor also specified:
“It is obvious that the by-pass arranged as described, opening or closing in a manner opposite to that in which the main gate opens or closes, will, if properly adjusted, admit of the main gate being rapidly operated and the governing of the water wheel quickly accomplished. After the governing takes placet, the by-pass gate is either open or closed, or nearly so, and in order to be useful for a second governing must return to its normal position.”
Thus again the normal position is indicated, with movement in either direction therefrom. And we do not find any statement or drawing *777which shows that the by-pass valve may he adjusted to occupy any position other than, a half-open one as the normal one.
While there is elaborate testimony from men of long experience inr hydraulics and electric devices that Lyndon’s device as described by him is inoperative, we are not prepared to adopt such a conclusion. But we are satisfied from the evidence that Lyndon’s device as described is to regulate the speed of the water wheel by maintaining a constant flow in the pipe line, and the flow in the pipe line is to be kept constant by the use of a by-pass normally kept in half-open position, through which water flows to waste. The purpose is to have a supply of water capable of being thrown on the wheel when, because of an opening movement of the water gate, the wheel slows down, and an outlet may be had for the water to compensate for an acceleration of the flow of water in the pipe line at the instant the water gate is moved toward a closing position.
The alleged infringing device is illustrated by the annexed drawing:
In the device the normal position of the auxiliary relief nozzle i& fully closed, moved from its normal position only when the nature of the reduction in load on the water wheel calls for a prompt closure of the valve of the main needle nozzle. A witness illustrated this by saying that a slow load would produce little or no action on the part of the auxiliary relief valve and an increase of load requiring an opening of the main nozzle would fail to produce any result thereon. The object of the device is to prevent excessive and dangerous pressure in the main line and at the same time to permit economy in the use of water — the latter a most important matter, but one not mentioned or seemingly practicable in the construction of the Lyndon device.
It is satisfactorily shown that the combinations set forth in claims .3 and 4 are not shown in the device used by the defendant. There is not a reversing clutch gear adapted to turn the water gate operating shaft in either direction, nor is there a controller responsive to changes of *778speed of the water wheel and controlling such reversing gear. The controller in the device of the defendant, which is responsive to the change of speed, does not control a reversing or reversing clutch gear. The returning device is controlled specially by the nature of the movement of the water gate operating shaft, and not as positive and immediate result of movement on the part of the controlling means.
Nor does the mechanism in the device of the defendant for controlling the motions within the hydraulic cylinder appear to be the same.
Referring to claim 8, we gather from the exhihits and testimony that there is a water gate operating shaft which will operate the water gate. There is a by-pass for the water wheel, and a valve for such by-pass, and an operating device for the valve (shown on Exhibit H), but there is no clutch adapted to connect the operating device for the by-pass valve with the water gate operating shaft; nor is there a valve for the by-pass normally held in partly open position. The mechanism shown is adapted to connect the operating device for the by-pass valve with the water gate operating shaft and to control the by-pass valve inversely to the water gate, which is constructed and operates independently of any clutch mechanism. Reversing means for operating the water gate operating shaft in either direction are found, but there is no driving shaft or reversing clutch gear adapted to turn the water gate operating shaft in either direction. There is a controller responsive to change of speed in the water wheel, but it does not control a reversing clutch gear.
The combination of elements claimed in claim 7, supra, is clearly not iound in defendant’s device, inasmuch as the automatic relief nozzle used is not operated from normal position in either direction, but in one direction only, and when the main gate is closed.
The arrangement of the various parts of the dashpot connected with the relief nozzle in defendant’s device is such that a closing movement in the valve of the- main needle nozzle, when slow, produces an effect on -the action of the valve of the relief nozzle, especially when the valve of the main needle nozzle operates nearest its full open position. The by-pass valve of the Eyndon patent is moved at a definite rate of speed responsive to movement on the part of the main water wheel gate; this movement being put into operation by instantaneously operating the electrical contacts in such manner that this by-pass valve will open when the main water wheel gate closes and close when the main water wheel gate opens.
Defendant provides means for governing the water wheel in a manner without causing excessive pressure rises in the main conduit, and thus protects the pipe line and permits great economy of water in accomplishing such object.
The speed-operated element and the fly ball governor used by defendant are not the mechanical equivalents of the wound dynamo marked upon the Eyndon patent. The valve-operating shaft in defendant’s device does not make a complete revolution, and is not in the usual sense an operating shaft used for transmitting motion, as indicated by W or Ifi in Figure 1 of the Eyndon patent.
[2] In arguing in support of the position that no one has obtained *779the results produced by the Lyndon invention without using that invention as claimed, appellant would have us hold that in claims 3, 4, 6, and 7 the word “means,” or the like, “is employed to describe connective features of the combination, and that such terms may be considered to cover practically any substitute part or feature.” Proceeding with bis argument, appellant says that, the substance of the claim being the combination, such combination is a unit, and that, where the component parts are recited as interrelated through the agency of means, the claim is to be construed as a unit. We find ourselves unable to sustain this argument. To permit a patentee to burden his claims by the use of indefinite language would lead to supporting him in a monopoly of a principle or result, which would bar other inventors from arriving at the same result by different means. In Walker on Patents, § 117a, p. 137, the author says:
“Where some of the parts of a combination operate therein to give motion to other parts which do the final work of the combination, it is proper to specify the former by the use of such terms as ‘means,’ ‘mechanism,’ or ‘devices’ for giving that motion, except when these terms are applied to an element or part which, constitutes the essence of the invention. If they are used under such circumstances, the claim will be regarded as functional. But such general language will not include all means, mechanism, or devices which can perform (hat function, but only those which are shown in the patent and their equivalents. And in this case, also, the question whether other means, mechanism, or devices are equivalents to those shown in the patent will be determined by the established rules on that subject, rather than by any apparent precision or elasticity of the language used in the claims to designate the parts involved in the inquiry.”
We do not find in the defendant’s device any mechanical equivalent to the reversing clutch gear or the means for operating the water gate in either direction as stated in the claims of the patent to Lyndon. Singer Manufacturing Co. v. Cramer, 192 U. S. 265, 24 Sup. Ct. 291, 48 L. Ed. 437. Nor do we find any returning device for the controller in the defendant’s device, as called for by claims 3 and 4. There is lack of similarity between the butterfly valve, an old style of valve used by Lyndon, and the needle valve, used by the defendant. They operate on different principles; the one inversely to the water gate in both directions, and at all times half open when in normal position, while defendant’s is closed.
The appellant earnestly contends that the object of defendant’s governing device is, not to take the pressure off the wheel and make it tend to slow down, but equally with Lyndon to keep the wheel moving at a constant speed, and thereby keep the potential of electric energy constant in the circuit of the generator. This view is supported by opinions of witnesses for the appellant. But with Lyndon the by-pass is a very important means of overcoming the inertia effects of water which interfere with the proper speed of the water wheel, the half-open by-pass being the indispensable means for governing the speed of the water wheel; while in defendant’s governing device the main needle is the principal thing, the auxiliary nozzle being used as a valve to prevent the damaging effects of excessive pressure to the pipe line, and the main needle moves often to govern the water wheel, without causing movement of the auxiliary relief nozzle.
*780[3] If, in sustaining the conclusion of the lower court, out construction of the claims of the patent, and particularly of claims 6 and 7, seems narrow, the position taken is, we think, in accord with the long-established rule of the Supreme and other federal courts, which limit the scope of every patent to the invention described in the claims in it, read in the light of the specifications. As has been very recently said by the Supreme Court, referring to the claims of a patent:
“These‘so mark where the progress claimed by the patent begins and 'where it ends that they have been aptly likened to the description in a deed, which sets the bounds to the grant which it contains. It is to the claims of every patent, therefore, that 'we must turn when we are seeking to determine what the invention is, the exclusive use of which is given to the inventor by the grant provided for by the statute; ‘he can claim nothing beyond them.’ ” Motion Picture Patents Co. v. Universal Film Mfg. Co. et al., 243 U. S. 502, 510, 37 Sup. Ct. 416, 418, 61 L. Ed. 871, L. R. A. 1917E, 1187, Ann. Cas. 1918A, 959.
The court cited the Paper Bag Case, 210 U. S. 405, 28 Sup. Ct. 748, 52 L. Ed. 1122.
Furthermore, unless Eyndon is limited to the specific form of the device described in his specifications and illustrated in his drawings, it is not improbable that claims 6 and 7 of the Lyndon patent would have to be declared void, as anticipated by a water wheel and device used by the Power Development Company near Bakersfield, Cal., in 1896 and 1897, where there was a by-pass mechanism operated inversely to the main water gates. In the mechanism designed for use at Bakersfield there was a governor which responded to load and speed changes. The by-pass valve was a necessary part of the equipment, for the rapidity of action of the governor there used made operation impossible without the by-pass feature.
The evidence concerning that device tends strongly to show that the governor used at Bakersfield was efficient and performed the functions of its design satisfactorily, although the failures of the wheels used to give required efficiency led to the discarding of use. Two of the principal experts for the defendant testified to the effect that there is closer similarity in the mechanical parts, the interrelation of parts and principles of action, between the Bakersfield governor, than there is between the device of the Lyndon patent and that used by defendant.
[4] It is also to be considered that there never has been a water wheel governor constructed, in accord with the drawings and specifications of the patent in suit — that is, a governor constructed and using all the principles of governing revealed in the patent. The Court of Appeals of the Sixth Circuit in National Malleable Castings Co. v. Buckeye M. I. & C. Co., 171 Fed. 847, 96 C. C. A. 515, said:
“The use we make of the fact that the device has never gone into actual service is in the construction or interpretation of the patent. We are justified, In view of the facts of this case, in exercising much caution in attributing to this patent anything more than is plainly shown and distinctly claimed. * * * This inference from nonuse, under the circumstances, is the converse of the inference drawn with respect of a doubtful patent when a showing is made that it has gone into large use and has displaced other devices. It is an inference against utility from the fact of long nonuse, unexplained by want of means or opportunity.”
*781Again, if Lyndon’s claims are not to he construed as limited to an electro-mechanical device substantially as shown and desciibed in the specifications and drawings, his claims 6 and 7 would be in danger of having been anticipated by the earlier French and Swiss patents (neither of which was cited by the examiner). The former patent concerns a self-regulating by-pass for a water wheel with the idea of overcoming inertia of pressure produced by the too rapid changes of the motor power, especially in long conduits, and the Swiss is an invention of an automatically regulated by-pass for high pressure water wheels with the idea to avoid the impact of the water in the conduits resulting from the sudden closing of the water gate.
The effort of the appellant to satisfy the court that his invention dated back ahead of the French and Swiss patents is not altogether convincing. Westinghouse E. & M. Co. v. Saranac Lake E. L. Co. (C. C.) 108 Fed. 221; Automatic Weighing M. Co. v. Pneumatic Scale Corporation, 166 Fed. 288, 92 C. C. A. 206.
But, leaving these questions open, we are of the opinion that on the ground that the device used by the defendant is not shown to be -an infringement of the device described in the patent of Lyndon the decree of the District Court dismissing the bill should be affirmed.
Affirmed.