Opinion
Patent Appeal No. 7879.
June 27, 1968. Rehearing Denied October 10, 1968.
Karl F. Ross, New York City, (Watson, Cole, Grindle Watson, Francis G. Cole, Laurence R. Brown, Washington, D.C., of counsel), for appellants.
Dos T. Hatfield, Washington, D.C., (Edgar H. Kent, Willis M. Ertman, Boston, Mass., of counsel), for appellee.
This is an appeal by the senior party, Livia C. Schmierer and Jacques H.F. Valin (Schmierer) from a decision of the Board of Patent Interferences, adhered to on reconsideration, awarding priority of the invention set forth in the single count here in issue to appellee, John C. Newton (Newton) the junior party. The sole count originated as claim 1 of Schmierer's United States patent and was copied by Newton for purposes of interference.
The board consisted of Messrs. Willner, Boys and Bailey, Examiners of Interferences. Mr. Willner wrote the opinion of the board.
The interference proceedings arose under 35 U.S.C. 135 between an application of John D. Newton, Serial No. 823,988, filed June 30, 1959, entitled "Data Processing Element" and a patent to Livia C. Schmierer and Jacques H.F. Valin, U.S. Patent No. 3,058,658, issued October 16, 1962 entitled "Control Unit for Digital Computing Systems." The patent issued on an application filed in the United States on December 22, 1958.
Appellants' arguments here raise issues as to whether the facts of record are legally sufficient on the questions of conception, reduction to practice, and diligence on behalf of Newton to support the conclusion of the board. Newton's burden was to establish his case by a preponderance of the evidence. As will be further developed, the facts of record require an affirmance of the decision of the board on the question of priority.
The assignees of the respective parties are Societe Nouvelle d'Electronique of Paris, France, for appellants, and International Business Machines Corporation for appellee.
Background of the Invention
The invention relates to a digital computing system which may receive messages of an arithmetical type from a memory. Appellants, in their brief, explain that the invention defined in the count may be described by reference to the operation of a simplified conventional computer.
A digital computing system of the type to which the invention relates processes data in response to and in accordance with a series of instructions. Both the data and the instructions are in the form of binary-coded words which are stored in the memory unit of the computing system. The systems disclosed by Schmierer and by Newton both utilize a message which includes an operation, or instruction, portion and a single-address portion.
Typical computers known to the art at the time the invention here in issue was made included what are known as "single-address" computers. Such single-address computers are capable of performing certain basic arithmetic operations, such as adding and subtracting. A typical adding operation takes the form of a command "Add X" wherein "Add" is termed an instruction word, or operation portion, constituting the functional message part of the coded signal. "X" is a data word constituting the algebraic or address portion of the coded signal which signifies to the computing system where in the memory to find the quantity to be added. Similarly, a subcontraction operation will be in the form of a command "Subtract Y" where, again, "Subtract" is an instruction word and "Y" is a data word.
In the then conventional operation of such a computer, the two commands "Add" and "Subtract" or similarly two "Add" or two "Subtract" commands operated in sequence would require the successive callout of at least two coded signals comprising four different "words" from at least two different memory locations in which they are stored, for example, "Add X" and "Subtract Y."
These coded signals are withdrawn from the memory of the computer to rest temporarily in an instruction register and are used to instruct a sequencing section of the computing system to perform the specified operation. Thus, in the typical known prior-art computers for the exemplary calculation sequence under consideration, (a) the sequence in the calculator section is established for addition, (b) the location X is then found for the quantity to be added before (c) the next operating sequence in the calculator is set up, and, finally, (d) the second quantity is located at Y.
In this art, this conventional sequence of steps, wherein each step is done independently, is called the "non-overlapped" mode of operation.
The improvement set forth by the count in issue may be referred to as an "overlapped" mode of operation and involves the simultaneous performance of two steps in the foregoing schedule, i.e., to locate a quantity while the previous calculation is in progress, thus resulting in a more rapid calculation. With "overlap," a second or auxiliary instruction register is added to the system to receive messages from the storage unit, and to set up operations for the specified calculation from the functional message portion of the instruction. The following sequence of steps is exemplary: (a) the first message, e.g., "Add X", is "fed" through the auxiliary register and is temporarily held in the main register to perform the addition of "X"; (b) the second message "Subtract Y" is "fed" into the auxiliary register and retained to specify the next operation "Subtract"; (c) during the addition, a message in the storage means is located in response to the functional portion of the message "Subtract"; (d) when the addition is completed, the main register is cleared and "Subtract Y" is transferred from the auxiliary register into the main register while the next message is fed into the auxiliary register. Thus, in step (c) a message is located without waiting for the calculation to end, resulting in "overlap".
The Count
The sole count in issue states:
1. In a digital computing system having storage means for retaining coded information applied thereto, main instruction-register means connected to receive from said storage means combinations of coded signals constituting functional and algebraic message portions, and a computer adapted to perform calculating operations in response to instructions from said main register means corresponding to an algebraic message portion stored therein, the combination therewith of control means including:
auxiliary instruction-register means connected to receive messages from said storage means;
selector means operative under the control of said auxiliary register means, in the presence of a first message in said main register means and a second message in said auxiliary means, for identifying a third message in said storage means in response to a functional portion of said second message during performance of a calculating operation by said computer in response to an algebraic portion of said first message; and
signal-responsive means operable by an execution signal from said computer indicative of the completion of a calculating operation for clearing said main register means of said first message and thereafter initiating a transfer of said second message from said auxiliary register means to said main register means and a transfer of said third message from said storage means to said auxiliary register means, thereby enabling said computer to operate on an algebraic portion of said second message during transfer of said third message and identification by said selector means of a further message in said storage means. [Emphasis added.]
In the proceedings below, the appellants raised a question of interpretation of the language of the count. However, to simplify issues in this proceeding, appellants do not press this point before the court and, for the purpose of the present proceedings, adopt appellee's construction of the count.
The Events Below
Appellants' brief contains in comparative chart form a brief summary of the time sequence of the principal events bearing upon the issues presented here. An edited version of that chart is here reproduced to provide a background outline of pertinent matters to be further considered in this opinion.
SCHMIERER et al. NEWTON ----------------------------------------------------------------------------- (1) 20 Nov.-12 Dec. 1957: Alleged conception by NEWTON (Exhibit 11, a drawing made by the witness Tilton). (2) 16 December 1957: Filing of French application No. 753,933 (subsequently French patent No. 1,201,663) by assignees of SCHMIERER et al. (3) 1 July 1958: Date appearing on printed report (Exhibit 1) from NEWTON's assignee (IBM) to United States Government. (Held to be conception by the board). (4) 8 December 1958: Execution by SCHMIERER et al., before U.S. Consul in Paris, of patent application corresponding to French application No. 753,933, supra. (5) 9 December 1958: Demonstration of computer model to representative of Secretary of Defense. (Doubt was expressed by the board as to sufficiency of this demonstration to establish an actual reduction to practice). (6) 22 December 1958: Filing of U.S. application Serial No. 782,266 (later patent No. 3,058,658) with documents ribboned under seal of U.S. Consulate in Paris. (7) 29 June 1959: NEWTON's application executed before Notary Public. (8) 30 June 1959: Filing of NEWTON's application Serial No. 823,988.Opinion
I. Appellants' Position
As seen from the foregoing chart appellants had filed a patent application in France on December 16, 1957. On December 8, 1958, less than a year after the French filing, appellants executed in France a corresponding patent application for filing in the United States Patent Office. However, apparently due to certain delays in customs, the executed application was not filed in the United States until December 22, 1968, more than one year after the corresponding French application. Thus, appellants' filing date in the United States is more than one year after their French filing date; hence appellants cannot rely on their French application for priority. 35 U.S.C. § 119. Under these circumstances, appellants may not establish under 35 U.S.C. § 104 a date of invention by reference to activities in France.
§ 104. Invention made abroad
In proceedings in the Patent Office and in the courts, an applicant for a patent, or a patentee, may not establish a date of invention by reference to knowledge or use thereof, or other activity with respect thereto, in a foreign country, except as provided in section 119 of this title. * * *
Appellants, in their brief, seek to obtain some benefit from their earlier filing of a French application by urging a two-fold argument: First, appellants suggest the French application provides a basis for judicial notice to be taken of the fact that Schmierer made the invention independently from Newton and his assignee and that "no inference can be drawn" to the contrary. This issue of derivation was not presented below and is not in question here. Accordingly, we decline to consider it.
Appellants' second argument directed to the effect of their French application is bottomed on their position below that execution of the application before the United States Consul in Paris constitutes an act of conception provable under 35 U.S.C. § 104. However, appellants cite no authority to support their contention that the execution of a patent application before a consular officer in France is anything other than an act "in a foreign country" within the meaning of this term in 35 U.S.C. § 104. Thus, the express wording of 35 U.S.C. § 104 precludes reliance on appellants' activities in France to establish a date of invention, except as to applications filed pursuant to 35 U.S.C. § 119. Here appellants' failure to comply with 35 U.S.C. § 119 precludes such a reliance.
By motion filed before appellee's testimony was taken, appellants asked to be given the benefit of the filing date of their corresponding French application, December 16, 1957, and the benefit of the date of December 8, 1958 when they executed their United States application before a United States consulate in France. The board deferred decision on those motions until final hearing after which it held that, under 35 U.S.C. § 104, appellants' French application is a nullity in this proceeding and confined appellants to their U.S. filing date. The Board stated:
* * * The statute does not provide for any extension of the year for any reason whatever. Therefore benefit of the filing date of the French application, December 16, 1957, cannot be accorded Schmierer et al., 35 U.S.C. § 119. Since Schmierer et al. may not establish a date by reference to knowledge of the invention in France except under the provisions of 35 U.S.C. § 119, by operation of 35 U.S.C. § 104, the Schmierer et al. French application is a nullity in this proceeding. Schmierer et al. is therefore restricted to the filing date of the U.S. application, December 22, 1958, for conception and (constructive) reduction to practice.
In their reasons of appeal, appellants questioned the board's decision insofar as the board failed to consider the filing date of the application in France as proof of conception.
Appellants cite United States ex rel. Majka v. Palmer, 67 F.2d 146 (7th Cir. 1933); United States v. Rodrigues, 182 F. Supp. 479 (S.D.Cal. 1960); United States v. Archer, 51 F. Supp. 708 (S.D. Cal. 1943). We do not find that these cases, involving criminal prosecutions, support appellants' contention that the acts here urged were not "in a foreign country" within the express meaning of 35 U.S.C. § 104.
Equally non-pertinent are the cases of King v. Young, 100 F.2d 663, 670, 26 CCPA 762, 771 (1938) and Fox v. Kingsland, 81 F. Supp. 433, 437 (D.D.C. 1948), cited by appellants for the proposition that execution of an application before a notary is evidence of conception even though the notary did not read or understand the disclosure, since these cases involve acts in the United States and do not have meaning in this context as applying to 35 U.S.C. § 104.
Newton took testimony, the legal sufficiency of which is the central issue in this appeal. Schmierer also took testimony, relating primarily to the reasons why his filing date in the United States was delayed beyond one year after his French filing date. As the case reaches us, Schmierer is confined to the filing date of his application in the United States. Thus, Newton must establish priority of invention by a preponderance of the evidence.
Turning now to the central contentions of the appellants as to the legal sufficiency of appellee's proofs, we shall consider each of the arguments in turn, often referring only to as much of the testimony as is necessary to support our conclusion.
II. Conception
Appellants' argument on this issue is that Newton has not proven conception by a preponderance of the evidence. Thus, these arguments raise the question of whether Newton's proofs are legally sufficient to prove conception.
Appellants urge that Newton's Exhibit 11 cannot establish conception for Newton allegedly because the only evidence linking Newton's conception to the exhibit is his own testimony. Insofar as this argument questions the inventorship, we think it a sufficient response to that position to reiterate the well-settled proposition that the question of third-party inventorship may not be raised in an interference proceeding. See, e.g., Huang v. Cheney, 362 F.2d 816, 53 CCPA 1355 (1966); Schwarzenbek v. Evering, 357 F.2d 1018, 53 CCPA 1152 (1966).
We turn now to a review of appellants' attack on the legal sufficiency of Newton's proofs. In order to prevail, Newton's proofs must justify a conclusion that Newton "possessed a definite and permanent idea of the complete and operative invention" or that he made his invention sufficiently plain to enable one of ordinary skill in the art to understand it. See Spero v. Ringold, 377 F.2d 652, 54 CCPA 1407, (1967); Gould v. Schawlow, 363 F.2d 908, 53 CCPA 1403 (1966); Townsend v. Smith, 36 F.2d 292, 17 CCPA 647 (1929). To this end, we shall review certain portions of appellee's proofs.
The evidence for appellee relates the history of the designing, construction and initial testing of a computer system embodying the improvement defined by the single count of the interference. The system was designed and built under Government contracts by Newton's assignee IBM for the purpose of developing new circuits, parts, packaging techniques, error detection techniques and logic. Its purpose, as the board noted, was to be subjected "to operating conditions over a long period of time" to evaluate its logic, circuitry and components for adoption by the Government in a new defense system. The system was thus named "Reliability Test Assembly" or "RTA".
The heart of the RTA system was its central computer. Other major parts of the system were an operating console which contained some of the memories for the central computer, together with control and indicator equipment for the operators, and an input-output system, variously referred to in the record as "I-O", "Peripheral Unit Processor" or "PUP", which included a card reader, a typewriter and a high speed drum.
The project began in 1957. In November of that year, the witness Tilton, who participated in the design, building and testing of the central computer from the fall of 1957 until July 1959, made a block circuit diagram of the system, entered into evidence as Newton's Exhibit 11. He stated that Fig. 4 of Newton's Exhibit 1, a final report of the phase of RTA prior to July 1, 1958, showed a simplified block diagram of the central computer portion of the RTA-1 machine, as it was in the summer of 1958. He also testified that his drawing was a print of a system diagram of the RTA-1 computer and that it embodied the logic of the count. Newton's Exhibit 11, however, was not considered by the board to be legally sufficient by itself to prove conception. It stated:
There seems to be no doubt that two "instruction" registers with associated decoders and command generators and with indicated transfer from the first register to the second register were in the plan of the RTA-1 from near its inception. They are shown on the layout, Newton Exhibit 11, made by Tilton. Tilton testified that he made no changes in the drawing after his last change note, December 12, 1957. We have no reason not to believe him. His testimony however does not affirmatively show just when he acquired such understanding of the relationship functioning of the elements shown on the layout that the requirements of the count would be met by it. * * *
We need not, however, review the question of whether Newton is entitled to a conception date of November 20 to December 12, 1957.
While Newton believes that he should have been given a date of conception by December 12, 1957, the date when the last change on the drawing (Newton's Exhibit 11) was made, he does not pursue the point, apparently conceding that "it does not appear to have any consequence to the questions appellants are raising here."
Newton's Exhibit 1 includes a diagram generally similar to Figs. 1 and 3 of Newton's application. As to this exhibit, the board stated:
* * * But Newton has a firmly established conception by July, 1958, with the report, Newton Exhibit 1, which makes Newton first to conceive by fully corroborated evidence. * *
Schmierer agrees, for the purposes of the present proceedings, that Newton's Exhibit 1 shows all the elements of the count. Since Newton's Exhibit 1 concededly shows all of the elements of the count, it is, of course, apparent that someone possessed the conception of the invention. Resolution of the question is clouded by the fact that Newton's Exhibit 11 chronologically preceded Exhibit 1.
Newton's own testimony states:
9Q. Did the RTA central computer system involve the overlapped processing of instructions? A. Yes, it did.
10Q. Who designed the logic for that overlapped processing of instructions? A. I would say that the primary portion of the overlapped processing is handled in the program control element and that I was responsible for the design of these controls.
That testimony is corroborated by Newton's Exhibit 1 which states that in the central computer "the program element is the controlling section for memory communications" and will perform the following function," one of which is "d. Provide instruction overlapping."
The testimony also shows that Newton was required to review the manuscript of that exhibit before printing to verify its content. That testimony, by Newton, is:
27Q. Did you approve the layout or text that goes with it? Were you given the manuscript for approval? Did you write it yourself or check it? A. I would say my manager, I think, was Joe Moyer * * * maybe it was Bill Carroll at that time * * * he had to approve * * * he was one of the people who had to approve the whole program. However, I would say I certainly had to look at all of this write-up and check to see that it was consistent with the design that I did. * * * * * *
29Q. What portion of this report Exhibit #1 did you write? A. I wrote portions * * * I wrote everything from Pages 84 through to 96 and also including the diagram on Page 97.
That portion of the document also contains descriptions of the "overlap-type" of operation which further supports Newton's own testimony.
Furthermore, the witness Tilton testified that he made the original drawing for Exhibit 1 in November of 1957 and revised it in December of that year. He stated that until he was assigned the task of preparing this drawing, there existed no overall system diagram of the machine which was then being designed. He thus was given the task of compiling and securing the information necessary to produce such a drawing and to make the drawing.
He stated that he obtained the information from various sources and that information for the program control element on the diagram came from John-Newton. We think this testimony provides additional corroborative evidence on the issue of Newton's conception.
Mr. Tilton compared the drawing, Newton's Exhibit 11, with Fig. 4 of Newton's Exhibit 1 and testified that they showed the same logic except for the immaterial absence of a third memory in Exhibit 11. He testified further that Fig. 5 of Exhibit 1 shows by means of a timing diagram the manner in which the logic shown in Fig. 4 of Exhibit 1 and in Exhibit 11 executes instructions in an overlapped manner.
We think it clear that Newton's testimony, standing alone, could not prove conception. Priority of conception, however, may be established when the invention is made sufficiently plain to enable one of ordinary skill in the art to understand it. Thus, evaluating the evidence of conception adduced by Newton in the light of the standard of what it teaches one of ordinary skill in the art, see e.g., Spero v. Ringold, supra; Gould v. Schawlow, supra; Townsend v. Smith, supra, we think the board was correct. When the evidence of all the circumstances is here tested by the requisite standard of proof, i.e., a preponderance of evidence, we think it legally supports the conclusion reached by the board.
On the record here, we conclude that there existed sufficient corroborative evidence in the form of documents, testimony and other circumstances to indicate that appellee's proofs are legally sufficient to support the board's finding of a "firmly established conception" by Newton by July 1958.
III. Actual Reduction to Practice
While the proofs necessary to establish an actual reduction to practice vary with each case, we agree with the board that appellee's proofs, when considered as a whole, are legally sufficient to support the board's finding that the invention of the count was actually reduced to practice on or before the end of June 1959. See, e.g., Patterson v. Hauck, 341 F.2d 131, 52 CCPA 987 (1965); Schnick v. Fenn, 277 F.2d 935, 47 CCPA 1174 (1960).
Some of the problems of territory and proof raised in determining whether an invention has been actually reduced to practice, are discussed in a recent article, Kempf, Reduction to Practice of Space Inventions, 50 JPOS 105 (1968).
Newton contended before the board that by December 9, 1958 the computer was sufficiently complete that, with some temporary equipment then in existence and connected to it, it could and did demonstrate "overlap" operation that meets the requirements of the count. The testimony indicates that the equipment as it existed on or before December 9, 1958 would perform that operation. The witness Tilton, for example, indicated that the logic of Fig. 4 of the report, which corresponds to Fig. 1 of the Newton application drawing, had been made "for all practical purposes" operational. The board, however, noted:
* * * On the other hand the computer was operating with temporary test memories having only two addresses each, so that a program of eleven messages such as given as an illustration in Figure 5 of Exhibit 1 apparently could not be carried out, at least in its entirety. The several contentions and points argued by Schmierer et al. does [sic] raise a little doubt as to whether reduction to practice had been proved as of December 9, 1958, and if the Newton evidence had cut off at this date we would hesitate to rule that the Newton evidence made out a sufficient case.
While appellee does not agree that the above quoted portion of the board's opinion was correct, we need not reach that issue. Moreover, counsel for appellee stated at oral argument that "it doesn't matter here."
The board summarized the Newton evidence for the period between December 9, 1958 and June 30, 1959, appellee's filing date, as follows:
The Newton evidence however shows that about thirty people continued to work on the computer in order to complete it by the scheduled date, July 1, 1959, working sometimes in two and three shifts, reducing in number as the computer neared completion because the work on it then did not admit of so many without conflict. The testimony is that the computer was delivered to the reliability test group on about July 1 as a fully operational device according to plan, according to Tilton. Tilton testified that, except for a memory drum that was being developed as an added feature, the machine was completely tested in the months before turning over the computer, during which time (referring to early 1959) test programs were debugged and developed for use by the reliability group.
It concluded:
We are satisfied that the invention in issue was actually reduced to practice on or before the end of June, 1959, at which time the invention was constructively reduced to practice by the filing of the Newton application.
The gist of appellants' position here on this issue is that appellee had not proven, with the requisite certainty, a date on which an operative embodiment of the count had been resuced to practice by Newton's assignee.
In connection with a visit from a representative of the Secretary of Defense on December 9, 1958, photographs, submitted as Newton's Exhibits 6-A to 6-D, were taken which show, according to the testimony of Newton and the witness Digan, that the three modules of the central computer had been completed, wired together and to test equipment, and were undergoing active testing. The date of the photos was independently established by the witnesses Casazza and Tilton. The testimony reveals that the test equipment included indicator lamps by which execution of logic could be determined, and two pairs of two address toggle switch registers, which were used in place of the 32 address and 8 address plugboard memories to be provided by the operating console when completed. With this amount of test equipment, the testimony establishes that it was possible to test all of the operating features of the central computer, including the execution of all instructions which the machine was designed to execute and its continuous operation in overlap mode as stated by the witness Tilton. His testimony on this point is as follows:
Since that visit occurred before Schmierer's United States filing date, i.e., December 22, 1958, it is convenient to discuss that evidence insofar as it provides a suitable starting point relating to Newton's proof of diligence, see Part IV, infra.
116Q. When, with reference to the date of December 9, 1958 of Exhibit # 6-C, do you believe you had done a substantial amount of such testing? A. I would say that on or about the date of this photograph, that from the appearance of the equipment and appearance of the test panel, we had done a substantial amount of testing. I definitely recall certain tests we made in overlap mode over a long period of time in which we, as an exercise, would continually add one to the arithmetic element and we would do it overlap mode and we would do it in non-overlap mode; and we could observe the difference in instruction execution rate.
The record establishes that by the December 9, 1958 date of the photographs the testing had been done to such extent as to, in the words of the witnesses, "leave no doubt whatever that the central computer * * * would operate as designed" and had "for all practical purposes been made operational."
While the board hesitated to find this an actual reduction to practice, the testimony also shows that testing with the test equipment shown in Newton Exhibits 6-A to 6-D continued into January 1959, when the permanent operating console was completed and test work was interrupted to wire it into the system and change the wiring from the temporary test equipment to the corresponding equipment of the console. The console added two plugboard memories of 32 and 8 storage addresses respectively which enabled longer and more diversified programs to be run, but otherwise essentially duplicated the test equipment previously used.
After the console had been wired in and tested, Tilton testified that "an antire group of programmers who had been engaged in writing tests and diagnostic programs" for the life performance analysis for which the machine was built, joined the work and "time began to be shared between the hardware test people and the programmers." The record also reveals that in March-April 1959 the "ferrite core" third memory was completed, wired in and tested, its additional storage addresses enabling still longer and more diversified programs to be run by the programmers. Also, in the first half of 1959 the PUP system was tied in to the central computer. During a substantial part of this period the work was carried out on a three shift, 24 hour per day basis, a point commented upon by the board.
The record also establishes that in July 1959, the RTA system was turned over to the Reliability Testing Group which ran the diagnostic and test programs to determine the suitability of the system as a prototype for SAGE-2. At that time the record shows that all of the central computer and all of the I-O equipment were operating, and that the system had been completely tested, except for a new type drum added later ans some error detection equipment, both seemingly immaterial to the unit here involved.
Thus, we find that the conclusion of the board that Newton had established an actual reduction to practice before his own filing date is adequately supported by the testimony of the witnesses Tilton and Digan, as well as by Newton's Exhibit 4 and the first RTA quarterly report for 1959.
IV. Diligence
The board concluded that since Newton has establised no actual reduction to practice before Schmierer's United States filing date, he must show diligence toward a reduction to practice from a time just prior to Schmierer's filing date of December 22, 1958. The board found:
* * * There can be no question of Newton diligence from prior to the Schmierer et al. filing date to reduction to practice, so that as first to conceive Newton is entitled to prevail. * * *
On this issue, we have reviewed this record in its entirety in the light of appellants' assertions and appellee's counter arguments. In particular, we have perused those questions and answers in the record, as specifically requested by appellants, together with the exhibits. We find the evidence on diligence to be legally sufficient to support the decision of the board on this issue.
See Part III, supra, especially the text accompanying note 8, supra.
The decision of the board is affirmed.
Affirmed.