Full opinion text
BYERS, District Judge. These causes, consolidated for trial, present conventional issues of infringement and validity of five patents, three of which deal with mechanism to 'control the successive purging, washing and drying of sugar in centrifugal machines in sugar refineries and factories; the remaining two cover, respectively, a syrup separator, and a water cooled brake. These devices promote the successful accomplishment of the tasks of the contrifugal machines. Purging is the first step in the sequence in which the so-called green syrup (containing about 95% pure sugar) is “centrifuged” or separated from the sugar grains and collected for subsequent treatment in the bottom of the jacket which encases the perforate basket into which the partially treated sugar — then known as massecuite or magma — had been placed; when that step has been accomplished, a period of centrifugal washing ensues as the result of introducing water into the whirling mass; when the washing has proceeded sufficiently (which is a critical and fairly precise requirement in the process), the water supply is cut off, and a drying period follows, upon the completion of which the centrifugal is stopped and the sugar is removed either manually or through an open bottom in the basket. This is the complete cycle of operation so far as the centrifugals are concerned. Because it has been outlined tersely for brevity’s sake, there is nothing casual or fortuitous about any of the steps. Each has its carefully calculated duration, based upon many factors to which allusion cannot serve any present purpose. The plaintiff’s patents were granted to Roberts, and were successfully embodied in commercial structures which found a ready market from about 1926 until the filing of these suits. In other words, plaintiff’s commercial success is not questioned. Commencing in 1936 the defendant inaugurated its own efforts, to compete with the plaintiff, and the pending and other litigations were instituted to determine whether its unpatented (except as to its water-cooled brake) structures infringe; and, if they do, whether the Roberts patents are valid. The successive steps in centrifugal operation, which have been stated, were formerly controlled manually and in turn, by the operators of the machines. This involved eccentricities of performance and frequent repetition of one or more of the operations, whereby production costs were enhanced, and the entire process was unnecessarily costly. Preliminary grading of the massecuite — which varied widely in constituency — to induce uniformity in treatment is said not to have been practicable prior to the introduction of mechanical control of the operation of the centrifugal machines, but has since been introduced; the accomplishments fairly to be attributed to the plaintiff’s patents have been in the public interest, in that production costs have been lowered through savings in labor and material wastage; also the operation of the centrifugals has been speeded up from around 850 R.P.M. to double that, and more. The Roberts Patents in Suit. It will be convenient to consider these in the order of their importance and in three parts: I. The control patents — cause No. 1544. II. The separator patent — cause No. 1636. III. The brake cooling patent — cause No. 1544. I. Roberts No. 1,758,901 — Application filed December 8, 1925 — “Means for Automatically Controlling Centrifugal Machines”. Granted May 13, 1930. The structure described is an intricate assembly of mechanical devices which together constitute an operable control of the centrifugal, best understood by consulting plaintiff’s Exhibit 59. The elements are: A handwheel for manual operation, which turns a shaft 9 to which it is affixed. Two separated gears on that shaft engage two other levers, 4 and 11, which have independent functions. . Lever 4 controls a clutch which in turn engages the pulley and thus the shaft which turns the centrifugal. As will be seen later, the application of the power which operates the centrifugal cannot coincide with the application of the brake. Those actions are mutually antagonistic and that relationship is maintained by the control mechanism. Lever 11 performs two functions: (a) it releases through rod 50 the brake 55 which holds the centrifugal at rest, and (b) it causes the timer control to be engaged through rod 15, engaging bell crank 16, and associated element 17, so that the ensuing period of operation of the centrifugal can be established. Clockwise rotation of the handwheel 8 therefore is responsible for the foregoing. A repetition, however, would not be possible with the mechanism thus far described. The brake must be reset, and the clutch must be disengaged, and a new cycle inaugurated. To accomplish this, the shaft 9 is equipped with a powerful coil spring 6 which is wound up by the rotation of the handwheel. The energy thus created opposes the action of lever 11 in releasing the brake 55 and in establishing the engage-^ ment of the timer control through 16 and 17. That opposing energy is held in check by an ingenious latching device which engages the top of lever 11, whereby bell crank 13 is brought into play. That is an angular member, held in position by a pin at the apex of the angle composed of the two arms. The lower arm contains a fixed roller 13c which falls into a notch at the upper end of lever 11 (which is prolonged laterally), when that moves in response to the lower end of lever 11 (which is pinioned by post 12 as shown). The clockwise rotation of shaft 9, transmitted through gear lib to the cogs of the lower segment of lever 11, causes counterclockwise action there and the reverse at the upper aspect or top prolongation of 11, whereby the roller 13c slips into and remains held in notch 11c. Thus the energy of spring 6 is latched until 13c disengages 11c. That action results when the short arm of 13 is pulled laterally by the 'action of rod 15 in obedience to a downward pull thereon, through 16, of rod 40, which is caused by force applied at its lower angular end delivered by arm 23 of the timer mechanism. When 13c is released from 11c, the torsion energy of spring 6 is freed and causes shaft 9 to rotate counter-clockwise, and the levers 11 and 4 to move clockwise, whereby clutch 3 is disengaged and brake rod 15 moves so as to put the brake on. The timer mechanism, which brought this result about, consists of two elements, connected to rod 15, through lever arm 16 to which rod 40 is attached as shown. These elements are shaft 20 which terminates inwardly in clutch member 18, and a worm gear Which forms part of a revolving member which also terminates in the second half of the clutch (no number). When the two elements of the clutch are engaged through the lateral movement of the lower segment of lever 17, the timer mechanism is ready to function. It is caused to operate by power supplied from the spindle (shaft) which actuates the centrifugal itself, through friction disc 31 which engages that spindle. The rotation of 31 delivers power to shaft 20 (through the clutch 18) to enable it to turn two discs affixed to it, each containing a radial projection or arm, which respectively strike and thus actuate two other elements. The shaft 20 rotates slowly so that the arc described by arm 23 (which is less than 360°) is completed in an adjustable changeable, and predetermined interval of, for example, 120 seconds. The length of the interval is established by the position of the arm 23 which is adjustable according to the insertion of a pin through the arm, into holes in the disc radially disposed, as shown in the said exhibit. Thus the length of the arc to be traversed by the arm 23, from its starting position until it strikes the lower end of 41, is subject to predetermination. It measures the duration of the entire cycle from the application of power to spin the centrifugal, to the concomitant release thereof and application of the brake. It will become important to recall that the energy which enables this device to function has its source not in the control mechanism itself, but in the agency which supplies power to operate the centrifugal. Before leaving this subject, it is required to understand that there is of necessity a counter force which tends to oppose shaft 20 in its initial rotation and which comes into play to return it, and disc 22, and therefore arm 23, to their positions at the inception of the rotation of shaft 20. That force is supplied by coil spring 21, which rotates shaft 20 (and attendant elements) back to their initial positions, which spring is released and permitted to function immediately when 41 is tripped by 23. That is accomplished because 23 is set with reference to stop pin 22a which strikes rest 20b as 23 completes its arc, and thus spring 21 functions to reset the timer. The second disc, also shown as 22, is another timer, which functions to open and shut off the water which is sprayed into the walled-up sugar, when the purging operation has been accomplished. Arm 24 projects from this disc as shown, and as the latter rotates with shaft 20 it strikes an attachment which starts the water to spray, and that independent mechanism continues for a predetermined interval until, by its own internal organization, that water is shut off. So much of the entire assembly is an embodiment of an earlier Roberts patent and is not involved in this litigation. Obviously this timer disc and the arm 24 begin and cease operation within the interval established by the partial revolution of the arm 23. The power to operate the washing mechanism is derived from the same disc 31 which has been described; the resetting of this disc and attendant arm is accomplished as has been described above. The important aspect of this particular mechanism is that it is regulated as to the turning on and off of the water with reference to the cycle established by the first or governing timing control, so that the washing begins at a predetermined point which corresponds with the completion of purging, and continues for a predetermined interval, and is then stopped so as to permit of a predetermined period of drying which must elapse before the cycle is ended. That subordinate control of the washing period is essential to the successful operation of the entire control mechanism. When the cycle has been completed, arm 23 trips hook 41 on rod 40, which causes rod 50 to move to the right, which disengages 13c from 11c, which unlatches torsion spring 6, which causes shaft 9 to rotate counter-clockwise, which disengages clutch 3 and applies the brakes through rod 15. The resetting of the timer mechanism consists in the disengagement of clutch member 18 from its counterpart and the resetting through the action of coil spring 21. The disengagement of 18 is completed because the roller 13c is prevented from falling back into 11c because coil spring 6 exerts its power so as to move the upper segment of lever 11 to the right. The evidence for the plaintiff is undisputed that the coordination of the various mechanical elements, which together constitute the structure described in this first of the Roberts patents, accomplishes effective control of the centrifugal machine. Thus the operator, having by experiment or trained observation decided upon the appropriate intervals for the several steps in the treatment of a given batch or “strike” of massecuite, makes the proper adjustment of arms 23 and 24 on the timer discs, and loads the centrifugal by turning wheel 8 clockwise, first at low speed to secure the uniform walling up in the basket of the strike, then turning 8 further to high speed, and, having done this, moves to another machine for the same purpose. In his absence, the purging, washing and drying proceed without manual participation by him, and at the end of the cycle he returns, and either he puts the centrifugal into slow speed to remove the sugar as it adheres to the sides of the basket in a partially dried state, or it has fallen of its own weight through an open-bottom basket if that type is employed. Then the operation is repeated. One skilled operator runs three machines in this way, instead of having to confine his efforts to one centrifugal as was required in the days of hand control. The more scientific handling of the strikes, which was thus rendered possible, reduced the necessity for reprocessing incompletely or improperly handled massecuite, .and the unnecessary melting of pure sugar, and reduced evaporating and pan boiling. In other words, the patented device proved to be an important contribution to the industry in which it was employed. II. Roberts No. 1,719,132 — Application filed November 5, 1926 — “Centrifugal Apparatus”. Granted July 2, 1929. Notwithstanding its earlier allowance, this patent follows the first in its proper place in this recital, since the description asserts that the patent covers an improvement on or modification of the first patent already discussed. It is not disputed that the mechanism here portrayed is not basically different from patent No. 1,758,901, and consequently it will be unnecessary to describe it at length. It is for a centrifugal control mechanism, best shown in plaintiff’s Exhibit 60, in which the power to turn the centrifugal is supplied by a two-speed motor. The control is started by moving lever 1, which can establish an operating circuit in the motor through the functioning of switch arm 28, actuated by rod 22 connected with 1 at 20; that rod also effects the operation of brake shaft 24, through brake lever 23. Again power and brake are 'engaged alternately. Switch arm 28 is moved to low speed connection (28a, and 28c), then to high speed (28c and 28b), as the result of the manual control of lever 1; in the showing of Exhibit 60, the brake is set, and switch arm 28 is not in a position to establish the contacts which would open either the low speed or high speed circuits of the motor. The upper end of lever 1 engages shaft 2, which is impelled to rotate clockwise while torsion spring 9 can deliver its energy, and that holds lever 1 and therefore switch arm 28 in the off position, and brake arm 24 and the brake shoes g in engagement. To counteract that effect, lever 1 is moved manually to the right, and switch arm 28 approaches the electrical contact stations; it can stop at neutral, and then move successively to the low speed (for loading the centrifugal) and then to high speed. The latter movement of lever 1 causes the latch or dog 12 to swing in back of the lever 1, through the action of coil spring 14 actuating shaft 13 (and arm 15 and attendant arms 16 and 17 later referred to) so as to hold it securely against the energy of coil spring 9. This is the equivalent of the latching position of 13c when it engages 11c in the first patent. The timing devices TI, T2, and T3, as shown, rotate on shaft 43 which receives its energy from the rotation of disc 30 through contact with the actuating spindle of the centrifugal. This also is in accord with the first patent. Once the spindle has started in the power-on position of the motor, the timer mechanism starts to function: T1 designates disc 45 having a peripheral finger 45d which is fixed in any desired position by 45c, thereby establishing the arc to be described from the starting position to its engagement with 60b on bar 60 which causes the wash water to start by opening valve WWV, after which 45d continues on its way. T2 designates rotating disc 47, similarly equipped, and its finger 47d in time trips lever 62, which causes the wash water to stop. As to this phase of the operation, the only difference in the two patents is that the timing of the wash water action is governed by the timing device of the control mechanism, instead of by means contained within the wash water mechanism. T3 designates rotating disc 46, having a similar finger 46d which eventually encounters the end of lever 19, causing a downward pull on rod 18, which, through the functioning of 17b, 17a, 16 and 15, unlatches 12 by causing it to move out of the path of lever 1, and thus the initial power of torsion spring 9 is released to enable it to return 1 to the power-off position of switch arm 28 and the brake-on position of brake shaft 24, so that the centrifugal now comes to rest. It is unnecessary to discuss the timer mechanism in detail as the functioning of the timer clutch 41, and the reset spring 43b which is being wound up as time shaft 43 revolves, will be understood from comparison with the similar elements in the first patent. The predetermination of the duration of each period, i. e., purging, washing and drying, is accomplished by the adjustment of the respective positions on the three discs of the fingers 45d, 47d and 46d, by pins 45c, 47c and 46c, by the operator of the control. Thus far it appears that the essential characteristics of the first patent have been carried forward and embodied in the second, with a certain element of flexibility added in connection with the disengagement of the timer mechanism, arising from the nature of the knock off arm 19 construction, as explained at line 123, etc., of page 7 of the specifications. The improvements are that lever 1, corresponding to handwheel 8 in the first patent, can be set in neutral, low or high speed, the brake being on prior to the first of these. When the timer mechanism releases latch or dog 12, lever 1 moves directly to brake-on, power-off position. Also the timers are mounted as a unit, and the sprayer control is one element of that unit, as has been indicated. III. Roberts No. 1,861,978 — Application filed April 23, 1929 — “Automatic Control for Centrifugal Machines”. Granted June 7, 1932. The teaching of this patent is of an improvement upon the second one in this group, and again the basic features of the first have been retained, but expressed in slightly different form. The important change is that the brake is released by one device, and the power and timing mechanism caused to function by another. Exhibit 61 depicts the association of elements in convenient aspect, and reveals the control at rest, with brake set, as at the completion of a cycle. Lever 15 at its upper end fits over but is not firmly held on hub 5 carrying brake-arm 6, which in turn actuates brake-arm 7, attached to brake lever 8 in the now familiar brake construction. Hub 5 is secured to shaft 1 which is caused to rotate counter-clockwise by torsion spring 3; so as to expend its energy in causing hub 5 to rotate with the shaft and thus to actuate brake-arm 6 and attendant elements to establish the brake-on condition. Release is effected by the operator’s application of the hand-grip 16 on lever 15 so'as to engage the top of that lever into hub 5 by the setting of a pin at 5a. Then, if 15 is pushed back, the brake is unset, and the torsion spring 3 is wound up. At the limit of that backward thrust, shaft 1 is locked in its then position, because hook latch 20 engages a pin 17a on adjacent collar 17 which is also firmly held on shaft 1. Thus the torsion spring 3 is latched, and the brake is off, and brake lever 15 can fall into vertical position. This is the equivalent of what takes place in the first patent when 13c falls into 11c. By a separate element starting handle 23, which can function only if pin 17c on disc 17 is out of its way (which cannot happen in the brake-on position) and to the extent permitted by the position of the pin 17a also on the disc 17, is then ready to be moved into neutral, low, and then high speed engagement with the contacts of the 2-speed motor, at its upper aspect as shown. The importance of the governing of the range of movement of starting lever 23, by the pins 17c and 17a on disc 17, is that thereby the alternating periods of brake-on, and high speed off, and vice versa, are maintained. When lever 23 moves into low speed position, the motor starts; into high speed, the timer T3 starts, and thereby the length of the cycle is determined. Again there is a timer clutch having two members, 38 and 39, and the former is moved from spring-held open position, laterally into engagement with the former, as the result of the turning of shaft 22, to which lever 23 is attached, through the actuation of members 34, 35, 36 and 37. The reset of this timer occurs when the spring shown to the right of the clutch member 38 is allowed to function on the completion of the cycle. Clutch member 39 takes its movement not from association with the centrifugal spindle, but from another exterior source of power, shaft 90, the further history of which is not presently important- TI and T2 respectively turn on and off the flow of wash water at valve WWV as in the second patent, and they function within the period established by T3, the major timer element. Again the several periods of purging, washing and drying are adjustable, and predeterminable. At the end of the cycle, a peripheral finger, on disc T3 which is turning counter-clockwise, engages lever 61 controlling lever 62, connected with rod 63 which turns latch 20, which releases finger 17a, whereby torsion spring 3 is unlatched and turns shaft 1 clockwise, and two things happen: (a) pin 17c in disc 17 moves forward and actuates lever 23 so as to open the switch and thus break the circuit to the high speed winding of the motor, and (b) the brake mechanism which has been described is actuated into the brake-on position through the functioning of hub 5, arm 6, rod 7, etc. Thus the basic teachings of the first patent, and the improvements of the second are embodied in this disclosure, with the additional feature of the separately disposed elements, the one to release and reset the brake and the torsion spring which actuates it, and the other to engage the power in successive steps, those devices being interconnected so as to preserve at all hazards the alternating operation of the braking and driving phases of the control mechanism. The evidence is barren of any indication that these inventions of Roberts came to him in a vision, or were imparted to him by any clairvoyant process. Nevertheless it is probably still the duty of the District Court to endeavor to deal with the issues of validity and infringement according to established principles. Roberts had been engaged in the business of making and selling machinery for use in the manufacture of refined sugar during his entire career. He was schooled almost exclusively in the problems of that industry, and his understanding and insight were the products of experience based upon first-hand observation. He sensed the need for mechanical control of the centrifugal machines if their true possibilities were to be realized, and if hit or miss methods of operation were to be superseded; with that understanding he pondered and studied in the fall of 1923 and, having conceived of an assembly of known mechanical devices which as he believed could be coordinated so as to be operable as a unit, in which purging, washing and drying could be controlled in successive periods which could be predetermined, he made a sketch of his own, and submitted it to Jenkins, the then Chief Engineer of Federal Sugar Refining Co., Yonkers, New York. The date is agreed to have been December 11, 1923, which is found to be the date of disclosure. Plaintiff’s Exhibit 10 is the sketch which Roberts made and then exhibited to Jenkins and discussed with him and others employed by the Federal Company. What was later to become coil spring 6, in the first patent, was substituted for the showing of a wheel and a weight to which Jenkins objected; and thereafter a draftsman prepared a technically better sketch, which is Exhibit 11. The substitution of the coil spring came after that sketch was examined and discussed by the same group. Roberts’ concept was embodied in an experimental machine which was built by him with assistance from the mechanical department of the Federal Company, some members of which were placed under Roberts’ supervision by Jenkins for that purpose. This unusual procedure is explained in Jenkins’ testimony as having been authorized because he realized from his conversations with Roberts and from his study of the two sketches, that the concept thus revealed, if embodied in operable mechanism, would greatly improve and cheapen the sugar-making process as then practised in the Federal plant. Exhibit 10 did not depict the control in its final form, but it did sufficiently portray Roberts’ concept, to enable Jenkins to see it for what it was, and to advise Roberts to cause his own crude drawing to be reproduced by one capable of illustrating the concept in the pictorial language of mechanical draftsmanship. The reduction to practise did not occur until the lapse of a year or more, and it is safe to say that the first machine was not completed and in running order until 1925. It was operated experimentally for som° weeks, and then sent out to Roberts’ factory, where it remained until the trial, at which it was produced and offered as Exhibit 17. In 1926 the Federal Company ordered and installed eighteen of the control machines, which embodied the original concept of Roberts as revealed in the first patent in suit, filed December 8, 1925. It should be stated that the evidence is persuasive to this court, that the period which elapsed, between the disclosure and the filing of the application, did not admit of uninterrupted attention on the part of Roberts to his invention, by reason of a period of illness, and the change in his business connections from the American Tool and Machinery Company to the plaintiff in this action, which is the corporation organized by him to enable him to prosecute his various endeavors. There was a continuity of development of the inventive concept which he had disclosed in December of 1923 to Jenkins and others, as has been stated, which was entirely consistent with his own belief in the efficacy of his invention once it could be rendered mechanically responsive thereto. It will be convenient to examine the claims in suit in connection with specific issues touching infringement. Validity. It now becomes necessary to consider whatever the evidence discloses as to the extent to which Roberts sought to occupy precincts which had already been preempted, and as to which he was merely a trespasser. The defendant leans heavily upon the following patents to demonstrate that Roberts claimed occupancy in a domain that had already been so widely appropriated that there was little or no area available to him: Andrews and Neuman No. 1,615,-433 — Application filed December 27, 1922— “Centrifugal Extractor-Operating Plant”. Granted January 25, 1927. This is pertinent on the issue of priority of invention, since the application antedated Roberts’. See Alexander Milburn Co. v. Davis-Bournonville Co., 270 U.S. 390, 46 S.Ct. 324, 70 L.Ed. 651. The question then is whether Andrews and Neuman invented what Roberts did, at an earlier date. Just what they did invent is complicated by the showing that on June 1, 1925, they withdrew their specifications and claims and substituted new ones, and whether these merely revised what had already been announced, or embodied a different invention, is the subject of conflicting views. The discussion of this patent will proceed with that caution. It must be understood that the apparent concern of the patentees was to provide means whereby the electrical power that ran the centrifugals might be conserved. Economy of power was the important consideration. Since the cycle of operation is of brief duration (for example, 90 to 120 seconds) it will be apparent that starting at low speed, advancing to high and then coming back to low speed or stopping altogether make demands upon the power source which it was desirable to ameliorate. Also that such a result might not be an impressive accomplishment in itself, but if associated with such a conduct of the successive purging, washing and drying periods as would enable the centrifugals to do their work properly, the combined effect would be desirable. So much can be read into the recitals in this patent, although not in such simple language. The concept then was of something more comprehensive than an agency of control, although that was an incident of the “Centrifugal Extractor-Operating Plant”. The defendant’s testimony is that the power-saving and control features were not required to be interrelated, but could have been separately presented and operated; the fact is, however, that they were; that control was not claimed as part of the invention will appear from a consideration of the claims of the patent. If those are presently comprehended, they assert nothing about the control element of the patented structure, except as it may be incidentally disclosed in the operation of some of the circuits comprehended in the structure. The structure is an assembly of familiar electrical elements and embodies: (Exhibit P) (a) A drum containing many strips or segments which serve as switches to open a series of circuits having various functions in the processes involved in the treatment of massecuite. That drum rotates to cause fingers held in constant relation thereto to engage those strips on the drum for the varying periods during which the circuits function. When the strips are rotated beyond the fingers, the circuits are closed. When all circuits have thus served their several purposes, the drum reverses to its starting position, because one or more other circuits, then closed, function to that end. In some instances, the circuits are completed by other means in addition to the engagement of both fingers with one strip or segment of the drum, but that is not destructive of the foregoing statement. (b) A drum control member, also a smaller drum, which rotates in the same way and as a monitor to govern the larger drum, in response to a hand controller which actuates the electric power to run this element. There are two movements of this small drum also, forward and reverse, and circuits are established through its rotation, as in the case of the larger drum control, which in turn cause the latter to function. (c) Timer: 1. To fix the duration of the cycle. 2. To start the wash water. 3. To stop the wash water. The braking or slowing down (stopping is debatable) of the centrifugals results from regenerative action in the motor which starts in obedience to the timing out of timer 1 (timer of cycle), which is followed by the reverse action of the pilot and the larger drum. The acceleration of the power supply to the centrifugals is accomplished by two agencies: The increase in voltage applied to the D. C. motor by balancer set generators, and the increase in resistance applied to the magnetic field, i. e., the stationary part of the motor; the increase in the latter resistance decreases the strength of the magnetic field which in turn causes the motor itself to run faster. Deceleration is the reverse of that process, namely, the field resistance is cut out and the strength of the magnetic field is thus restored; this causes the motor to transmit current to the main lines, i. e., to regenerate, which means that the motor then becomes a generator and slows down ; also the balancer set generators are' cut out, which steps down the voltage to the D. C. mqtor. Thus is induced what Andrews described as follows: “We had a short period of dynamic braking through which the motor was practically short circuited on itself.” The braking therefore is accomplished by control of the electric energy which it is the office of the motor to supply to operate the centrifugals, i. e., by causing that energy in effect to serve two successive and contrary purposes. As has been said, the timer of cycle, timer 1, measures the duration of the interval which starts with the movement of the hand controller switch into high speed and ends with the slowing down to the slow speed at which the sugar is plowed out, and the basket is ready for reloading at that same speed. The circuits established by contact of the -fingers on 7 and 8 of the large drum start and reset the tinier in succession. After the high speed connection has been made, and at an interval which depends upon the position of the fingers which engage strip 10, the wash water solenoid valve is turned on while that circuit is closed, and the wash water is applied. When that circuit is opened by the completion of the contact, the wash water valve is closed. Then the reverse rotation of the pilot drum and the control drum occurs as the result of the opening of the circuit which was closed so long as the fingers were in contact with strip or segment 6. The wiring diagram whereby these various circuits can be traced is so complex that even the inventors, who were on the stand, were unable to identify the original diagram as that actually followed in the installation of these structures in the National Sugar Refinery; nor were they entirely convincing as to whether anything amounting to substantial changes had been made in respect thereto either then or later. It is beyond question that about May 5, 1923, two machines generally corresponding to the structure of the Andrews and Neuman patent were erected and operated in that refinery; and that in all twenty-eight such machines were thereafter installed and have continuously operated since that time. Whether this has ever been a public use in the practical sense, certainly as to the years 1923-1926, inclusive, is so doubtful that if the issue turned on that question it would require careful consideration. As has been said, a control of sugar centrifugals was necessarily comprehended in the Andrews and Neuman disclosure, for uniformity of operation and product were stated to be among the contemplated incidents of the operation of their power-saving combination of electrical devices. Since Andrews and Neuman do not discuss the duration of the drying period, or even refer to that as a necessary step in the cycle, it is left to inference entirely whether there was a predetermination of the washing period with reference to this third step in the cycle. The timing out of their wash period preceded the timing out of the time of cycle timer (Y in figure 1) so there was an interval in which drying could take place, which probably they understood since they were engineers in the employ of the National Sugar Company, but they also failed to mention it in their specifications or claims. Roberts Nos. 1,758,901; 1,719,132; and 1,861,978 are now held to have been patentable over Andrews and Neuman No. 1,615,433, giving to the latter, for the sake of argument only, all that was asserted in the amended specifications and claims, in that: (a) The control is adaptable to a centrifugal which is driven by mechanical energy through a pulley or other suitable connection, as well as to an electrically driven centrifugal, whether D. C. or A. C. (b) The control of the drying period is provided for as an identifiable stage in the cycle. (c) The timers are reset not by the energy which started them, but from sources exterior thereto. (d) The timers can be altered as to the duration of their respective actions by visible, accessible, and easily disposed adjustments. In the Andrews and Neuman structure this adjustment can be made only by altering the length of the appropriate segments, or by substituting one for another — a feasible but cumbersome and dilatory process, requiring the stoppage of the control drum while the necessary changes are made manually. (e) The control is not so integrated with the centrifugal but that it can be assembled therewith without fundamental reorganization of the centrifugal plant as a whole; this is a striking advance over Andrews and Neuman. (f) There is no combination disclosed by Andrews and Neuman which includes a brake, self-setting or otherwise, and the centrifugal as operated never comes to rest; this may not be necessary according to refinery practise, but thereby a distinguishing feature is revealed in the Roberts patents as contrasted with Andrews and Neuman. (g) The last-named do not teach alternate braking and power engagement, although the circuits function so that, as power declines, the regenerative so-called braking is actuated, which means that the effective energy 'subsides so that the centrifugals finally move at low speed. The obvious preoccupation of Andrews and Neuman with the power-saving aspect of their invention probably accounts for their failure to discuss, except casually, that which was clearly perceived and expounded by Roberts touching control mechanism as an end in itself. The following patents are relied upon to demonstrate that Roberts was not the first to invent a control mechanism for centrifugals: Spencer and Metcalfe, No. 1,680,259— Application filed October 22, 1924, for “Control System”. Granted August 7, 1928. No one asserts that this is more than a paper patent. The proclaimed object is to provide an automatic control of centrifugal separators with reference to the purging and washing periods and as to the proper quantity of wash water to be applied. Electrically operated devices, which control the length of cycle and the amount of energy consumed in its operation and the turning on and off of water for washing with an alternative manual control if desired, constitute the subject-matter. It seems unnecessary to recite herein the elements of the disclosure except to note that the control mechanism is actuated by a generator which corresponds in function to the timer mechanism of the Roberts patent; a push button applies the power supplied by the main driving motor of the centrifugal, and at the same time the brake is released and remains in that condition while the power is on. The small generator, to which reference has been made, functions through control and in cooperation with the spindle of the centrifugal; what the defendant calls the timing mechanism and the plaintiff calls the control is thereby actuated so as to cause the water to be turned on after the purging has proceeded, and thereafter the water is turned off. These results follow from successive contacts made by an arm which projects laterally from a member mounted on a disk, which arm rotates and comes successively into contact with circuit closing elements, and thereby causes circuits to open and close as required to accomplish the desired purposes. When the rotation of the arm has proceeded so that the latter completes its appointed task, it automatically returns to its starting position, whereupon the brake is caused to be applied and the centrifugal power is shut off, which likewise causes the small generator, which actuates the control mechanism, to cease its function. The arm, which establishes the various contacts, moves in proportion to the square of the centrifugal spindle. It does not measure time, but periods of purging and washing with reference to the centrifugal cycle. The mechanism which admits of the performance of these functions is adjustable as to the duration of each period. If the device were to be erected and were to function according to the teaching of the patent, the operator of the centrifugal machine would merely push the starting button and then pass on to another machine while the various periods of the cycle are accomplished, and the machine brought to rest by the automatic application of the brake. The defendant’s expert, Mr. Kelly, stated that the assembly of the control mechanism was not entirely conventional and “in order to make it work the way Spencer and Metcalfe wanted to make it work there would have to be design work done on it”. From the practical standpoint, the operability of the device leaves something to be desired, because the disclosure, as depicted in the drawings, apparently will not permit the discharge of a flat bottom centrifugal basket with a mechanical dis-charger. Another patent relied upon to demonstrate that Roberts was not the first to conceive of an automatic control is: Herr No. 1,866,499 — Application filed June 4, 1923, for “Continuous Automatic Centrifugal Machine”. Granted July 5, 1932. An unnecessary amount of time.was devoted to some interpretation of this very intricate mechanism, which requires fifty drawings to accompany the specifications and somewhat illuminate them. The patentee asserts that there are ten features disclosed, his object being to provide an automatic centrifugal, reliable in action, fool proof in service and efficient in its work. He refers to his four prior patents which are deficient in not showing fixed speed ratios, with the result that the structures are unreliable as to their slow speed operation. This patent apparently is directed to the introduction of an independent drive for the slow speed basket rotation which can be predetermined. The control of the purging, washing and drying steps, assuming it to have been disclosed, is effected through the functioning of a constantly rotating screw, which causes a traveler disposed thereon to move along the screw from a stop which is adjustable as to its position, and which traveler engages in its movement successively a spray valve opening element, and a spray valve closing element. These are adjustable in position, whereby the instant of application and the quantity of water to be used can be varied within the cycle. No one claims to have seen the Herr mechanism in operation as disclosed in the patent, although the defendant’s witness Olcott thinks that he saw part of it at some, indefinite time and place. By the expedient of preparing a drawing from which a portion of the Herr assemblage is shown, the defendant urges that the place of this patent in the present litigation is thus firmly established. At best, that is an argument. Herr says that at least ten objects are embraced within his specifications, of which the defendant has dispensed with six in the presentation of its exhibit prepared for consideration at this trial. It might plausibly be argued that the Herr patent is to be considered as a'whole and the value of its teachings depends upon the extent to which the structure as disclosed accomplishes its proclaimed objects, and in the absence of proof that it is operable as a whole, no reliance can be placed upon the assertion that it can be dissected and part of it erected in a theoretical sense and made to work, even though the rest be discarded. Reliance is had upon the case of General Electric Co. v. Minneapolis-Honeywell Regulator Co., 2 Cir., 118 F.2d 278. In that case the Court did disregard one element in a prior patent and considered only a part which corresponded to the device presented by the patentee in that case. It does not seem that this citation supports the mutilation of the Herr patent practised by the defendant for the purpose of this litigation. It is not a matter of importance, however, because it must be deemed for present purposes that Herr claimed to have invented a control mechanism, which provided for the starting of the centrifugal operation, the turning on and off of the wash water, the turning off of the power, and the stopping of the machine by a friction brake (resetting of the traveler on the screw is also said to be demonstrated, but it seems to me that the evidence falls short of accomplishing that purpose), in an application for a patent which he filed prior to the filing of the application for the first Roberts patent, and therefore the invention therein disclosed must be confined to the precise showing thereof. Knapp No. 1,134,621 — Application filed August 3, 1914, for “Automatic Valve”. Granted April 6, 1915. This is a control device to open and shut the wash water valve “primarily designed for use in spraying water on sugar in a centrifugal machine * * * It functions through the movement of a traveler on a rotating shaft actuated by power not identified with relation to the power that operates the centrifugal. As the traveler moves from right to left as shown in the patent drawings, it pushes a valve stem to open position; when the surface of engagement has moved sufficiently beyond the area of engagement, the valve closes. The traveler can be regulated as to its starting position, by adjustment, and its speed of movement (thus the period of engagement with the valve stem) can be regulated by adjustment of the speed of its rotating shaft. Reset is manual, not automatic. The sole contribution made by this patent to this record is to indicate that some years prior to Roberts there was in use a mechanical means to open and close the wash water valve, which was adjustable in operation with reference to the start of the purging period in the centrifugal cycle. The following three patents are cited to demonstrate that electric time clocks adapted for use in connection with the control mechanism of centrifugals were known when Roberts filed his application, namely: Hall No. 1,483,433 — Application filed September 24, 1921, for “Relay”. Granted February 12, 1924. Hull No. 1,813,743 — Application filed October 9, 1924, for “Time Element Control System and Apparatus”. Granted July 7, 1931. Carlson No. 1,669,927 — Application filed July 20, 1922, for “Apparatus for Washing Sugar and the Like”. Granted May 15, 1928. The Hall patent covers a relay, namely, an electrically operated switch which may be delayed in its operation and acts as an electric timer. When it times out, it performs a switching operation which makes possible the control of other electrical devices such as are employed in the operation of centrifugal machines. The foregoing is undisputed. The Hull patent presents a time element control arrangement to open and later to close the wash water applied during the operation of the sugar centrifugal at a predetermined time after high speed has been attained by the centrifugal. The object is accomplished by employing a combination of two of the time clocks described in the Hull patent. The Carlson patent is sufficiently like the Hull patent in purpose and in functioning to preclude extended discussion. It will be understood that none of these three patents has anything to do with the application of the brake and the synchronous shutting off of power which form an essential element of the Roberts patents, and since Roberts’ timer control does not resemble in any respect those herein disclosed, it will suffice as to them to state that prior to Roberts’ filing date, the art was equipped with these timing devices and that they were set to operate at time intervals, rather than with reference to the progress of the purging operation, although in the practical sense that comes to the same thing. To recapitulate on this branch of the case: Andrews and Neuman, Spencer and Metcalfe, and Herr did not constitute prior art in the strict sense, as against Roberts, since their several applications had not ripened into patents at the time that Roberts filed; hence neither he nor the public was apprised of them. See: Comolite Corporation v. Davidovicz et al., 2 Cir., 111 F.2d 121, at 123; Stelos Co. v. Hosiery Motor-Mend Corporation et al., 2 Cir., 72 F.2d 405, at 406; Utah Radio Products Co. v. Delco Appliance Corporation, D.C., 24 F.Supp. 328, at 340. Sight has not been lost of defendant’s quotation from the opinion in Detrola Radio & Television Corporation v. Hazeltine Corporation, 313 U.S. 259, at 265, 61 S.Ct. 948, 85 L.Ed. 1319, which is to the effect that applications pending but unknown to the public, at the time of filing the patent in suit, constitute prior art. Authority for that statement is the Mil-burn case, supra, in which the decision is squarely to the effect that such applications may be consulted on the issue of priority of- invention. In order to portray that which is known at a given time on a given subject, i. e., the state of the art, recourse may be had to printed publications, but that means printed matter to which the public mind is exposed. It is difficult to see how applications pending in the Patent Office can be so described, prior to the grant of the patents upon which they are based. As long ago as in 1874 it was decided by the Supreme Court in Brown v. Guild, 90 U.S. 181, 23 Wall. 181, at 211, 23 L.Ed. 161, that “a mere application for a patent is not mentioned as such a bar [a ‘patent’ or a ‘description in a printed publication’]. It can only have a bearing on the question of prior invention or discovery.” That was written concerning a patent which had been filed prior to the one in suit, and later withdrawn; it was deemed to be but an abandoned experiment and not to anticipate plaintiff. Surely that result could not have been reached if the filing, standing by itself, had been deemed to define one constituent of the art as it existed when the patent in suit was filed. Anticipation does not arise here in the sense that the patentees named in the several cited patents are here contesting that question with Roberts. The areas in this field of endeavor which had been appropriated by Knapp with his wash water valve control; and by Hall, Hull and Carlson with their time clocks, and time interval controls, did not impinge upon the metes and bounds of Roberts’ invention as disclosed and portrayed by him. I am satisfied that the presumption of validity which has frequently been recognized as attaching to the grant of Letters Patent is well fortified as to these Roberts patents, in that: He disclosed a new and operable combination of mechanical elements which so functioned as to accomplish the important and required office of controlling the successive periods of purging, washing and drying sugar in centrifugal machines at predetermined and adjustable intervals. The claims in suit will be referred to in the consideration of the following subject-matter. Infringement. It is undisputed that the defendant has made and sold unpatented control structures now in use in three sugar factories or refineries, to be referred to for convenience as the “Sucrest control”, the “Igualdad control" and the “Georgia control”. The important issue in the case is that of infringement, since there is no doubt that the defendant intended to invade the field of plaintiff’s activities — as it has the right to do — and is confident that its structures have avoided in essential respects the inventions described and portrayed in the Roberts patents. It therefore becomes necessary to examine what the evidence discloses concerning these several devices. The Sucrest control. This control mechanism employs electric current to start and stop the centrifugal, and compressed air to apply and release the mechanical brake (in reciprocal relation to the power-off and power-on periods of the cycle), and to actuate one or more electric switches or relays. It is not argued for the defendant that there is any difference in principle between such sources of energy, and that which resides in the torsion of a spring. This installation at the factory of the American Molasses Company is said to be depicted in plaintiff’s Exhibit 64 and defendant’s Exhibit Y, both having been provided by the defendant pursuant to order of this Court. The operation is as follows (See Exhibit Y) : 1. The air master switch is a hand-operated air valve. Its office is to put the control mechanism in condition to function after a shut-down, and it is moved to the “run” position and thereafter plays no further part in the operation of the mechanism. 2. Air is thus admitted to air-operated switch PR-1, thereby closing its contacts and establishing an electric circuit, which: (a) Resets timer relay brake on power off relay TR, which deenergizes the brake solenoid valve BSV which moves the mechanical brake off. (b) The air-operated switch PR closes its contacts. (c) Slow speed connection to the centrifugal is made, during which loading of the basket is accomplished. 3. The “fast” button is pushed down, thereby: (a) Raising the “slow” button out of engagement with its contacts and opening the associated circuit. (b) Establishing high speed connection with the centrifugal through the high speed winding of the motor. (c) Energizing the cycle timer TR-1, which starts to function, i. e., to run its course or time out, at which contacts normally open are closed and an electric circuit is established, which: (1) Connects the main driving motor to low speed, thus starting regenerative braking (field against motor), and (2) Energizes the TR relay or switch called the timing relay, or brake on power off condition. (3) The TR relay in turn completes its mission and times out, having energized the BSV which applies the air pressure to the air-operated mechanical brake; also it causes compressed air to operate the air-operated switch PR whereby its contacts are opened and the associated circuit closed, whereby the low speed circuit of the driving motor is deenergized and the period of regenerative braking is terminated, and, the mechanical brake being in operation, the mechanism is brought down to slow speed for plowing out the sugar, and reloading. It will be observed that there is no washing period embraced in this operation, and that aspect of the Roberts patents is not involved. Certain conditions in the mechanism at this stage of its operation should be stated: a. The “fast” button is down, which means that the “slow” button is necessarily up, i. e., no contacts, and hence no associated circuit is closed. These buttons are manually operated only. b. The cycle timer TR-1 is in the reset condition. c. The TR relay, the timing brake on and power off relay, is in timed out station but not reset, so that it is not in position to function until the slow button is pushed down in the ensuing cycle. d. The solenoid brake valve BSV is energized, which means that the upper air valve is open, and a connection will then be made from the air supply through the valve and pipes to the cylinder controlling the mechanical brake, and the air-operated switch PR so as to disconnect the contacts of the latter. This structure is said to infringe claim 3 of the first Roberts patent, which will be stated piecemeal for present purposes, thus: “The combination with a contrifugal suspension shaft of (a) driving and braking means associated therewith to operate thereon in alternation to each other,” Clearly these are present in the Sucrest structure although the braking means in the latter consists of both the regenerative action mentioned and the mechanical air-operated brake. (b) “a rotary timer operatively combined with the centrifugal and operating after a predetermined but variable period of rotation from starting position, according to its adjustment, to cause the cessation of the driving action and the setting of the brake,” The Sucrest timer consists of cycle timer TR-1 and the associated relay, power on brake off TR; it is not the Roberts type of rotary timer but is an electrically operated equivalent thereof, and is not for that reason to be distinguished from the plaintiff’s timer mechanism. (c) “and means operating in coordination with the driving and braking means to restore the timer to said starting position after the timer has completed its operation.” This is the only aspect of the claim which defendant seriously argues. Its contention comes down to this, that because TR is not reset upon the completion of the cycle, but only by depressing the “slow” button in a new cycle, infringement is avoided. This seems to me a distinction without a difference. While it is true that in Roberts the timer mechanism is reset when 13c moves out of lie and along the upper arm of lever Í1, which means that a new cycle is about to begin, there is nothing in the testimony to indicate that TR does not reset as part of the unsetting of the mechanical brake, and the coordination with the driving and braking means is just as complete as though the resetting of the timer accompanied the earlier setting of the brake. This is the more so since the brake action in both' structures is not a thing apart, but is reciprocal to the application of the centrifugal power in conventional operation. The defendant’s argument at pages 11 and 12 of its brief is more engaging in the dialectic than persuasive in the practical sense. The conclusion is that claim 3 of Roberts No. 1,758,901 is valid, and is infringed in the Sucrest structure of defendant’s manufacture. The Igualdad control. This control mechanism is schematically portrayed by defendant in plaintiff’s Exhibit 65 and defendant’s Exhibit Z, which reveal a circuit for conducting current to operate the centrifugal, controlled by an electric master switch, and also a brake timer TR-5, the opening of the contacts in which deenergizes the current established by the master switch; a push button switch to establish high speed in the current that operates the centrifugal; a cycle tim'er TR-2; a water on timer TR-3, and a water off timer TR-6; these latter respectively function so as to open and then to close a solenoid valve controlling the volume of water which is admitted to the centrifugal machine to wash the sugar after a predetermined period of purging has been maintained. The operation of the control mechanism is thus outlined by the defense: a. The master switch has been placed in the “off” position (at the conclusion of a prior cycle), which resets the timers and causes the mechanical brake to be deenergized by release of the air pressure which caused the brake to function. b. The master switch is moved to the “run” position, which establishes a current in the low speed winding of the motor j at that centrifugal speed, the sugar treated in the earlier cycle is plowed out, and then a new supply of massecuite is admitted to the basket and therein walls up. c. Then the push button is actuated,, which establishes current in the high speed motor winding; the contacts thereby closed establish a circuit (which remains closed although the button is not held down), and at high speed of the centrifugal, the purging, washing and drying are accomplished. Also TR-2 and TR-3 are independently energized; the former establishes the duration of the cycle, and the latter moves toward its goal which is to open the wash-water valve at the appointed interval; it does that at its time out period when it also energizes TR-6 to start the latter’s task, which is to close the wash-water valve upon reaching its own time out station. d. TR-2 times out enough later than TR-6 to admit of a drying period; and when that occurs, the following results are thereby effected: 1. An electric switch reestablishes a circuit in the low speed winding of the motor, which again initiates regenerative braking (field vs motor). 2. TR-5, the brake timer, is energized, which at its own time out station: (a) Actuates the brake solenoid valve to admit compressed air to the mechanical brake; and (b) Disconnects the low speed circuit of the driving motor (the regenerative braking having accomplished its purpose), and now the driving motor has been disconnected from any source of power, and the cycle has been completed. It will be seen that the “run” position of the master switch initiates a low speed running condition of the centrifugal which could be continued indefinitely if that were desirable which of course it is not; some incidental purging would occur, but the necessary treatment of the massecuite would not ensue. This stage of operation is like the jogging of the clutch under the first Roberts patent, and low speed running under the other two, which are needed only for the purpose of loading the basket. The actuation of the high speed running of the machine is accompanied by the starting of the cycle timer, and of the wash-water timers which open and close the water valve respectively during the cycle, after the purgi