SPARKS, Circuit Judge.
This appeal is from a decree of the District Court holding United States patent to Garrison, No. 1,386,698 invalid and not infringed as to claims 1 and 3; holding United States patent to Banfield, No. 1,778,349, invalid as to claims 1 and 2; and dismissing the bill of complaint for want of equity.
Appellant manufactures a complete automatic stoker equipment for heating furnaces, and appellee sells control units designed to be placed on stokers of other makes so that, when equipped with such control unit, they will perform the functions of and compete with appellant’s product. We shall treat the patents in the order named.
The Garrison patent was issued August 9, 1921, on an application filed March 15, 1920. It was assigned to appellant [905]*905July 10, 1931. The claims in suit are set forth in the margin.1
The District Court stated, and it was not controverted, that the prior art included the following elements of claim 1: A furnace of a mechanical fuel feed, means of connecting said fuel feed to the furnace, a control apparatus adapted to open and close the drafts of the furnace and start and stop the fuel feed, and a thermostat adapted to control the operation of said furnace and fuel feed, for normal operations, according to the temperature requirements.
In its memorandum opinion the District Court (13 F.Supp. 995, 996) quite clearly stated the question at hand and the reasons which actuated the ruling which is now before us. We quote: “Garrison was not in this patent the originator of the idea of feeding the coal or the first to provide a time device for heating systems. The broad idea of thermostatic control of stokers through a control apparatus is the subject matter of a prior patent to him, No. 1,382,877. To these old elements, Garrison added ‘a time device adapted to operate the furnace and fuel feed for a definite time periodically, irrespective of temperature requirements.’ Thus the claim must be construed as one for a combination of elements adapted to operate periodically a stoker and the drafts of an automatic furnace, controlled by a thermostat, to produce heat in accordance with a predetermined standard with an alleged novel addition of a timing device which will apply electric current which in turn opens the draft and starts the stoker, independent of temperature requirements, at stated predetermined intervals, so that the fire will not die. Timing devices for opening and closing switches controlling electric currents at stated intervals are old in the electrical art, and the question submitted is whether by the addition of such a device a patentable combination was created. The timing device, according to the patent, has only one function to perform; at stated intervals it closes the switch and thereby starts the furnace so that the fire will not die. The disclosure teaches no other function. The timing device does not shut off the current, but that result, on the contrary, is accomplished by the thermostat controlling the room temperature. In other words> the clock merely starts the motor Yhlch controls the stoker and drafts, but does not stop it. This operation, thus started at stated intervals, is not ended until the room thermostat stops the motor when the desired room temperature, which controls the thermostat, has been reached.
“In claim 3 we have the same old elements, including, however, an alleged new element, ‘a time device adapted to so operate the electrical device that the furnace and fuel feed are operated for a definite time periodically irrespective of the temperature requirements and which time device is also adapted to so change the thermostat connections that said electrical device is controlled by the thermostat according to the temperature requirements at all times other than said definite periodical times.’ However, an examination of the specifications and descriptions discloses that the device claimed does not, as indicated, stop the fuel feed. It merely starts it. In other words, under this claim also, the time device merely controls one cycle of operation of the electrical device to start the motor, and [906]*906also changes the thermostat connection all without stopping the motor. The latter is thereafter subject only to the control of the thermostat. Though the claim is more specific than claim 1, it is built up in the same fashion, by a union of old elements, all of which continue to perform individual functions well known to the art.
“I am of the opinion that each of these claims is invalid for want of invention ; that they constitute aggregations.
“I believe the present claims come within the criticism of the Circuit Court of Appeals, in Hartman Furniture and Carpet Co. v. Banning, 59 F.(2d) 129.
“I am further convinced that under the reasoning * * * in the case last cited, any asserted attribute of invention in the addition of the timing' device to an automatic stoking outfit ‘is successfully met by the argument that such’ addition ‘was an obvious and old device which any electrician would adapt.’
“I am further -of the opinion that if the Garrison claims are to be construed as covering such device as those marketed by defendant, then they are broader than Garrison intended, involving cláims for functions, and, are therefore, void on that ground.”
It is first contended by appellant that the District Court erred in its conclusion of fact that the timing device merely starts the motor which controls the stoker and drafts but does not stop it. A simplified diagram of the disclosure is. here set forth.
[907]*907The parts are shown in the position they assume when the stoker is in operation and the draft is on, notwithstanding the fact that the temperature in the room is sufficiently warm to cause the room thermostat 73 to stop the stoker, were it not for the operation of the timing device 35.
The stoker motor 25 is started and stopped by drum switch 9, fixed to a shaft 6 that carries a crank 7 for operating the dampers. One.half revolution of 6 forward stops the motor and closes the draft. Another half revolution in the same direction starts the motor again and opens the draft.
The shaft is driven by a ratchet pulley 10, which is loose on the shaft and having teeth 18 to co-operate with the pawl 17 carried by the crank 7 which is fast to shaft 6. A cable 12 looped about and fastened to the pulley 10 has one end fastened to a spring 13, and the other end fastened to the core 11 of solenoid 2, connected across the high voltage line 68 and 70. When energized, 2 will pull 11 down, rotating 10 one-half revolution, and through the ratchet and pawl, driving shaft 6 the same amount, where it is held against reverse rotation by spring pawls 20 and 21 co-operating with disk 8. That movement, when the motor is still, turns 9 and 7 to the on position, starting the motor and opening the draft. When solenoid 2 is de-energized, spring 13 turns the pulley backward one-half revolution and raises core 11, leaving the shaft with its crank, switch drum and disk in the on position. The stoker then continues to operate until solenoid 2 is again energized to give the shaft another half revolution, returning the parts to the off position, stopping the motor and closing the draft.
Solenoid 2 is energized by closing switch 67 in the high voltage circuit. The controls are in a low voltage circuit supplied with - current through a transformer 4. Time clock 35 has a moving finger 38 on the minute hand to engage contacts 39 and 40, which it touches once an hour, and contact 40 is adjustable as to location to vary the period of operation. When 38 touches 39 it closes a circuit through a relay or transfer switch 50, which closes switch 67 by actuating relay 66. This in turn energizes solenoid 2 and starts the stoker, which continues in operation until 38 contacts 40, when another circuit is closed through the transfer switch 50, again energizing solenoid 2 and stopping the stoker. Each half rotation of shaft 6 breaks the circuit through solenoid 2 and allows spring 13 to restore core 11 and pulley 10, thus preparing them for the next operation when solenoid 2 is again energized.
Room thermostat 73 has a bar 74 adapted to move between a cold contact 78 and a hot contact 75, depending upon the temperature of the room to be heated. When it contacts 78 another circuit is thereby closed through relay 50 to close switch 67, thus energizing solenoid 2. As the room is heated to the temperature predetermined by the room thermostat, 74 swings away from the cold contact 78 and touches the hot contact 75. This in turn closes another circuit through relay 50 to energize solenoid 2 which stops the stoker.
Spring pawls 20 and 21 engage disk 8 to hold shaft 6 stationary between half revolutions. They are also electrical connections and are attached to wires 63 and 76. Disk 8 has a notch 22, fitted with insulation, that moves under one of the pawls 20 or 21 when shaft 6 completes a half revolution. The effect of this is to break the circuit through solenoid 2 and allow spring 13 to restore core 11 and ratchet pulley 10.
When the stoker is on a timed operation started by 38 contacting 39, disk 8 is in the position as shown in the diagram with the insulation 22 under pawl 20, and pawl 21 is in electrical contact with disk 8. If the room is at the maximum predetermined temperature, the bar 74 will be in contact with hot point 75 when clock finger 38 reaches 40, which is the end of the timed period of stoker operation. It must be remembered that 74 was in contact with 75 at the time 38 contacted 39, that is to say, the temperature of the room was at the predetermined maximum point both at the time 38 contacted 39 and when it contacted [908]*90840. That contact of 74 and 75 was made when the room temperature reached the maximum as determined by the room thermostat, and the room thermostat was powerless to again connect .the high voltage circuit at 67, which actuates the stoker motor through solenoid 2, until the temperature of the room fell below the predetermined maximum, when 74 would swing to cold point 78. However, when the temperature of the room is at or above the predetermined maximum, and 74 is in contact with 75, the hand 38 of the timing device contacts 39. The current thereupon energizes solenoid 47, thus tilting armature 50 to the right, breaking the circuit at 53 and contacting 54. The ultimate result of this is to close the high voltage circuit at 67, thus energizing solenoid 2 which starts the stoker motor. The room thermostat 73 has no function in the action just described, hence as conceded, the timing device starts the motor. The parties, however, do not agree that the timing device stops the motor. The court, adversely to- appellant’s contention, held that it did not, but we are convinced that it does.
At this point, the stoker is operating; 74 is still in contact with 75, indicating that the • room temperature is sufficiently warm; solenoid 2 has performed its function of starting the stoker; spring 13 has lifted core 11 from solenoid 2, thereby causing ratchet pulley 10 to turn backward on shaft 6’ to its original working position, and the high tension circuit at 67 is again open by reason of the former completed half revolution forward of drum 9 when the stoker was started. When the timer hand 38 contacts 40, solenoid 57 is energized, thereby attracting the left end of armature 50, thus breaking the circuits at 54 and 55 and closing the circuit at 53- and 49, and ultimately closing the high voltage circuit at 67, thus energizing solenoid 2 which stops the stoker. It is quite true that the room thermostat is on that circuit, and its parts are in proper position to permit the closing of the high voltage circuit 67, which in turn stops the stoker motor through solenoid 2, but we think it can not be said that the room thermostat stopped the motor. Literally, we suppose, the power exerted by solenoid 2 is the thing that stops and starts the motor. With respect to the stopping of the motor, under the assumed facts, we think the timer, when its hand 38 contacts 40, lets loose the force which, with the coaction of 57 and 50, and, if we may say, the permissive inaction of the room thermostat 73, eventually closes the high voltage circuit at 67, which in turn causes solenoid 2 to stop the motor. Certain it is that the room thermostat under the assumed facts, has nothing to do with closing the circuit at 40, and it could never stop the stoker until that contact was made.
It is true, as appellee urges, that if 74 is not in contact with 75 when 38 contacts 40 the stokef will not stop. This is due to the fact that the room temperature has become lower than that demanded by the thermostat, and in that event the thermostat refuses to permit the closing of the circuit at 67, which would stop the stoker, until the thermostat is satisfied as to room temperature, that is to say, until 74 again contacts 75. This is explained in the specifications by way of emphasis of the fact that the timing device in no way interferes with the operation of the thermostat except during the predetermined period of the timer, and for the purpose of showing co-action of the parts. Under the circumstances just mentioned there can be no doubt that the timer does not cause the stoker to be stopped when 38 contacts 40, because the thermostat refuses to permit 74 to contact 75, which would close the circuit at 67 and stop the motor, until the predetermined room temperature has been satisfied.
The claimed virtue of this disclosure, however, does not manifest itself after 38 has contacted 40 and not again until 38 has contacted 39. Its entire utility covers only the period when the hand 38 is traveling from 39 to 40, and its complete operation presupposes a contact of 74 with 75, when 38 contacts 39 and 40. It in no manner interferes with the effective operation of the thermostat, except that during the predetermined period of the timer it may cause an increase of [909]*909the room temperature above that normally demanded by the thermostat. If, at and shortly prior to the time when 38 contacts 39, the temperature of the room has lowered so as to cause 74 to leave 75 and contact 78, the timer is then ineffective to start the stoker because the thermostat had already caused it to start. During the predetermined period, however, the timer will not permit the thermostat to stop the stoker, even though the temperature of the room has again reached its predetermined maximum.
Although, under the conditions just mentioned, the thermostat started the stoker, it was shorn of its power to stop it, by 38 contacting 39, thus cutting the circuit at 53, which could not be closed until 38 had contacted 40. If at that time the heat demand of the thermostat were still satisfied, a fact which must be presumed in this discussion, the contact of 38 with 40 would cause the circuit to be closed which would stop the stoker. If, perchance, the temperature of the room at that time were below the predetermined degree, the thermostat would prevent the stopping of the stoker by merely breaking the circuit at 75. This we think constitutes coaction of a high order.
Under the facts which we think must be assumed in the discussion of these claims, we are .convinced that the timing device causes the stoker to stop as well as to start. It may be conceded that the elements are old, but they coact in such a way as to produce the new result of starting and stopping the motor regardless of the demands of the thermostat. Conceding without admitting that the result is old in the art, yet we are convinced that it is here produced in a more efficient manner than the art had theretofore disclosed, and that fact is sufficient to support the patent. New York Scaffolding Co. v. Whitney (C.C.A.) 224 F. 452; Young Radiator Company v. Modine Manufacturing Company (C.C.A.) 55 F. (2d) 545. We do not regard this holding as inconsistent with that of Hartman Furniture & Carpet Company v. Banning (C.C.A.) 59 F.(2d) 129, for in that case there was neither a new result nor an old one performed more efficiently. The suggestion that the timing device was nothing more than that which any electrician would adopt in the exercise of mechanical skill, we think is answered by the fact that it was never adopted prior to appellant’s disclosure, although the problem had been present for many years. We think the court erred in holding Claims 1 and 3 invalid.
Appellee’s pilot control comprises a simple electrically driven time switch of the two point make-and-break type, which, when applied to stokers, alone controls both the starting and the stopping of the stoker motor to determine the duration of the periodic operation without in any way relying upon the room thermostat. Appellee makes two models of its controls, the time switch alone, and the time switch coupled with a relay and transformer, but the circuit controlled by the time switch is the same in each. In household installations a transformer is used to provide a low voltage current for the thermostat, enabling the ordinary electric service to be used instead of a battery or other low voltage source. This was quite, old in the art.
Referring to appellee’s simple model, the time switch alone, its controls comprise a motor frame and a terminal panel. On the face of the motor frame there are mounted two switch arms, spaced side by side with the contact points between the arms. The arms are wired respectively to the outer terminals. The upper ends of the switch arms have followers riding on cams, which cams are relatively adjustable upon loosening a screw in a slot to permit relative spacing of the abrupt shoulders of the cams, according to the length of time the switch is desired to be closed. The shaft on which the cams are mounted is driven by a synchronous electric motor, there being suitable reduction gearing between the motor shaft and the cam shaft so that the latter will make one revolution every hour or half hour as desired. By turning the cams in a clockwise direction the switch is closed only when one switch arm follower engages a low face of its cam and the other follower engages the high face of its cam, and the switch contacts are spaced apart and the switch is open when [910]*910both followers are engaging corresponding surfaces of .their respective' cams. This closing and spacing of the switch contacts, controlled solely by the rotation of the cams, controls the starting and stopping of the furnace motor during ‘the periodic timed firing. It is a simple make-ánd-break switch.
The terminal panel has four binding parts. From left to right one is connected directly to the burner motor and the thermostat; the second is directly connected to the power line and burner motor; the third is connected directly to the power line and the furnace limit switch; and the fourth is connected directly to the furnace limit switch and the thermo-, stat. The two switch arms of the control are connected by wires to the end posts. This arrangement is the same for all furnace installations, whether oil fired, gas fired or coal stoker systems. At the end of the predetermined time when the circuit is closed, the switch fingers are separated and the motor is stopped, and it is not started again until the' next hour. Thus a single motor circuit is set up through the pilot control' which alone makes and breaks the motor circuit at predetermined times.- However, if the room temperature drops to the minimum, the room thermostat will close the circuit, thereby setting up the single motor circuit normally subject ■ to the room thermostat. The furnace limit control remains normally closed and is only opened when the furnace is overheated. Hence, in appellee’s recommended system, two independent motor circuits are set up, one through the pilot control or time switch, and the other through the room thermostat. The latter has no function in the installation other than its usual function to control the stoker motor only when temperature requirements in the room demand it.
In appellee’s other model, which contains the low voltage transformer, the relay is of the gravity type, which is normally open, and closes only when the coil of the relay is energized. When closed,current is supplied to the motor to start it, and when the relay opens, the current is broken and the motor stops. The relay may be closed either by the time control, ' or the room thermostat, but neither has any action or influence on the other. When the time switch contacts are closed, due to the rotation of the cam, the relay is actuated, and when the time switch contacts open, due to the continued rotation of the cam, this circuit is broken, and the relay opens, stopping the motor. In this periodic operation the room thermostat has no function. However, should there be a demand for heat in the .room, the thermostat closes and energizes the relay through a different circuit to close the relay and operate the motor until the room temperature requirements are satisfied. In this operation the time switch has no function.
In appellee’s recommended installation, however, there is. a connection between the furnace limit control and the thermostat through the right end binding post on the terminal panel. The evidence discloses that this connection is not inherently essential to the successful operation of appellee’s method. It is merely a safety or cutout device, used in all. heating systems and required by all codes, to prevent excessive heating of the boiler or heating above a predetermined degree. Such safety device may be either a fuse, pressure device, or a thermostat. .
All the elements in both of appellee’s models are admittedly old in the art, and it is' obvious that there is no coaction between them which would elevate either of the models above the characterization of an aggregation. We agree with the District Court that appellant’s invention, and intention must be confined to a disclosure of a specific means in combination. This is the only theory under which we feel warranted in sustaining the claims. We are in accord with the District Court that to go further than this, would involve claims for functions. Claims 1 and 3 of Garrison are valid but not infringed.
The Banfield patent, No. 1,778,349,. was issued October' 14, 1930, on an application filed April 28, 1925. It was assigned to appellant November 10, 1930. The claims in suit are set forth in the margin.2 This patent is based oñ an al[911]*911leged improvement over Garrison by the addition of a furnace or boiler- thermostat, by which Banfield claimed to have improved the operation in comfort and economy, both the room and boiler thermostats being in electrical series. It is illustrated by the following diagram:
Switch 25, which connects the motor, is operated by oppositively wound solenoids 44 and 45, having a common core 46. When energized, 44 pulls down 46 and thus closes the switch 25 which starts the motor. The operating circuit then travels from transformer 27 through 29, 23, 48, 44, 47, 24 and 28. It will be observed that to close this circuit the tongue of each thermostat must contact the respective cold point of its own thermostat, as shown by the heavy lines of the diagram. That is to say, each thermostat must be demanding more heat. Solenoid 44 never opens the switch. When ener[912]*912gized it always pulls 46 downward. Even though that circuit be broken, thus de-energizing 44, the switch will still remain closed until 45 is energized. This can only be accomplished by the tongues of both thermostats contacting their respective hot points. A new circuit is thus created from transformer 27 through 29, the hot point of 23, 50, 45, 49, the hot point of 24, and 28, thus energizing 45 which elevates 46, which in turn opens switch 25. When the tongues of both* thermostats do not have the same relative contacts, or when they, or either of them, are floating between the hot and cold points, neither of the currents just referred to is closed, and neither of the solenoids 44 and 45 is energized, and the switch remains in the same position it was in when the thermostats were last unified in their demands.
The timer 31 gets its power from transformer 27 when 32 and 33 bridge the respective contacts 37, 38 and .39, 40. This sets up a new circuit through 51, 48, 44, 47 and 52, thus energizing 44, provided it is 'not already energized by the circuit through the cold points of the thermostats. In that event the timer performs no service, for its contemplated work in closing switch 25 has already been accomplished by the thermostats. On the other hand, if the outside circuit of the timer energizes 44 at a time when the tongues of the thermostats are contacting the hot points, then the timer will have no effect on the core 46, because both solenoids 44 and 45 will be energized, and being of equal power their forces will be neutralized. For the same reason the thermostats could never open the motor switch which the timer had closed, until the predetermined time of the timer had expired, thus breaking the outside circuit through 44.
It is contended by appellant, however, that the timer will start the stoker whenever it should be started. The argument in support of this contention, we think, is quite reasonable. In cold weather the timer is of little if any use in keeping the fire from dying. In warm weather, however, the room temperature may stand for hours above the desired degree without fresh fuel, and at such time the tongue of the room thermostat will contact its cold point. During this period the boiler thermostat also will have reached its predetermined maximum degree, which we understand to be much greater than that of the room, hence the switch is opened and the stoker is stopped. Under these conditions, as the fire dies, the boiler will cool and go below its predetermined minimum temperature, although the room is at or above the desired temperature. In this event the tongue, of the boiler thermostat will swing away from its hot contact, thus breaking the circuit and de-energizing solenoid 45, which permits the timer to close the switch for a predetermined time through solenoid 44. Appellant contends that during mild weather this condition prevails most of the time and permits the timer to function at all times when needed. This appears quite reasonable and it is not disputed.
The District Court was of the opinion that the elements employed amounted to mere aggregation, and that there was no coaction or coalescence of the two thermostats, and that the disclosure amounted to nothing more than mechanical skill. In this conclusion we think that the court misapprehended the facts, due no doubt to the lack of perspicuity in the argument and briefs. We agree that the furnace thermostat breaks the timer current when the water temperature reaches a certain fixed point, but it is obvious that that does not open switch 25, unless the room thermostat through its coaction with the furnace thermostat' makes the same demand. Likewise, the room thermostat breaks the middle current when the room temperature is too low, but that does not close switch 25, unless the furnace thermostat coacts in making the same demand. Here lies the virtue, if any, of the disclosure. For instance, if the stoker has been operating and the temperatures of both the room and the boiler have reached the desired degree of heat, only the coaction of both thermostats closes a new circuit and stops the motor by opening switch 25. One thermostat may operate to break the circuit before the other, but this will not stop the motor. If the room temperature becomes too cool its thermostat will promptly demand a closing of the switch which starts the motor, but solenoid 44 will not close it until the furnace thermostat makes the same demand, because until that time the circuit is not closed. In such case, the [913]*913object is to require the room to use the accumulated heat in the boiler, thus saving fuel, until the heat of the boiler is reduced to a point where its thermostat also demands more heat. Then, as it were, the minds of both thermostats are at one, and the motor switch is closed.
The District Court said: “The furnace thermostat shuts off the electric current when the water temperature reaches a certain fixed point. The room thermostat shuts off the current when the room temperature reaches a certain fixed point, and again applies the current, opening up the furnace, when the room temperature falls below a fixed point. The furnace thermostat controls the opening and closing of currents. * * * ”
This statement is only partially accurate. The currents referred to therein are those which run through the thermostats, but they do not include the current which operates the motor through lines 26. That current can only be opened or closed by the coaction of both thermostats, and this is true as to all circuits referred to. Hence it cannot be said that the furnace thermostat independently controls the opening and closing of any current.
We are convinced, therefore, that the patent disclosed a valid combination which produced new results in respect to economy of fuel and a stable continuity of heat. Whether those results be great or small, we do not feel warranted in striking down the presumption of validity which attaches to its issuance.
What we said in the discussion of infringement in the Garrison case is equally applicable here. Appellant admits that appellee would not have infringed had it used any release mechanism other than a thermostat. Such use, however, was admitted to be old by appellant’s witness, and his testimony seems to be well supported by Smith No. 1,600, 568, Johnson No. 1,602,363, St. Clair No. 1,633,465, and Gray No. 1,667,001.
The decree is affirmed as to both patents on the question of infringement. The challenged claims of both patents we think are valid, and the decree is reversed as to the question of validity.
The cause is remanded for further proceedings not inconsistent with this opinion. The costs are adjudged one half against each party.