Opinion
Patent Appeal No. 8348.
December 3, 1970.
Richard G. Stephens, Binghamton, N.Y., Franklin D. Wolffe, Washington, D.C., attorneys of record, for appellant.
Carl W. Laumann, Jr., Endicott, N.Y., Julius Jancin, Jr., Maurice H. Klitzman, Washington, D.C., John Farley, New York City, for appellee.
Before RICH, ALMOND, BALDWIN, and LANE, Judges, and McMANUS, Judge, Northern District of Iowa, sitting by designation.
This appeal is from a decision of the Board of Patent Interferences awarding "priority" to Rohland, the junior party, solely on the ancillary ground that the disclosure of Hemstreet does not support the language of the count. We reverse.
Patent No. 2,919,426, serial No. 721,064, filed March 12, 1958, as a continuation of serial No. 477,569, filed December 24, 1954.
Serial No. 831,599, filed August 4, 1959, as a division of serial No. 336,080, filed February 10, 1953.
Shortly after the issuance of the Rohland patent, Hemstreet copied claim 1 thereof as claim 38 of his divisional application, the interference was declared and, after examination of the preliminary statements, the junior party Rohland was put under order to show cause why judgment should not be rendered against him since he did not allege any date prior to Hemstreet's filing date. Rohland thereafter moved under Patent Office Rule 232 to dissolve the interference on the ground, inter alia, that Hemstreet had no right to make the count. The Primary Examiner denied the motion. No testimony has been taken by either party. The board disagreed with the examiner as to Hemstreet's right to make, held he could not, and whether he can is the issue before us.
The subject matter of this interference is photoelectric apparatus using a plurality of angularly related scanning patterns for the recognition of controlled-format printed characters. The field of use of the invention is explained in appellant's brief as follows:
Automatic or machine reading of the letters and numbers on printed documents has been a very active field of research during the last decade, and a substantial amount of equipment which performs such a function is presently in commercial use, in connection with the reading of bank checks, credit card invoices, cash register journal tapes and other "controlled source" documents wherein the type font (size and style) and the location of the matter to be read on a given document can tend to be controlled.
The sole count (subparagraphing and emphasis supplied) is:
In reading apparatus,
a sensing station including means for scanning characters to be read which are advanced past said station, the scanning means comprising means for scanning portions of a character in registration with said sensing station with a plurality of differently oriented scanning patterns,
means comprising detecting means for producing signals when the scanning means senses portions of the character, said signals being of a duration proportional to the length of the portion sensed, and
means for combining signals produced by scanning portions of at least one or more predetermined lengths for providing an output signal indicative of the character scanned at the completion of all of said scanning patterns.
Rohland's apparatus generates, on each scan, a signal or signals continuously proportional in duration to the length of each character portion sensed and then applies these signals to a coincidence circuit whose other input is a continuous stream of equally spaced pulses. The output of the coincidence circuit, which consists of individual trains of pulses discretely proportional to the length of each character portion sensed, is then applied to circuitry which identifies the character scanned using the number of these pulse trains generated during each orientation of the scanning means having more than a minimum number of pulses, with no maximum number other than the number of pulses per scan, and the number of these pulse trains having more than a larger minimum number of pulses, again with no maximum number other than the number of pulses per scan.
With the proviso that once a train of one of the two types has been registered by the character-identifying circuitry on a given scan, no further trains of that type will be registered until three more scans have been completed or the frame in which that scan was made has been completed, whichever comes first. This, Rohland's specifications tells us at page 81 of the record, is to insure that any one substantially straight portion of the character being scanned will provide only one signal even though it may be scanned more than once. However, it should be noted that a substantially straight portion of sufficient length will clearly cause the registration of both a long and a short count, if a short count has not been registered within the three previous scans in that frame.
Hemstreet's apparatus also generates on each scan a signal or signals continuously proportional in duration to the length of each character portion sensed, but character recognition is based, not on the length of these signals, but on the number of onsets and/or terminations of such signals, representing background-to-character and character-to-background transitions, in each scanning pattern.
The fulcrum of the board's opinion is the sentence:
In determining the meaning of a count it is normally assumed that a term [here "signals"] used in one portion of the count with one meaning has the same meaning in a later portion of the count unless otherwise modified.
The term in question, "signals," is used twice in the second clause and once in the third clause. The second time "signals" is used in the second clause it clearly refers back ("said signals") to the first time it is used in that clause. The controversy turns on the relationship, if any, between the interpretation to be given "signals" as it appears in the second clause and as it appears in the third clause. (We are disregarding "output signal" in the third clause since it is clearly something entirely different from the other "signals.")
The board's opinion first focuses on Hemstreet's disclosure, stating that "the signals generated by the means comprising detecting means [the second clause] may be interpreted in two distinct manners, that is, they may be read as continuous signals * * * which are proportional to the length of the portion scanned or they may be read as the white to black transition, positive pulses." The board then examines the consequences of these two alternative interpretations, always assuming that "signals" in the second clause should be interpreted the same way as "signals" in the third clause.
The board had early stated that Hemstreet "merely uses the white to black transition positive pulses, the black to white transition negative pulses having no effect." This statement we believe to be an error in view of the paragraph from Hemstreet's application bridging pages 18 and 19 of the printed record, which reads:
The device may be operated with pulses of only one polarity and be unresponsive to pulses of the other polarity, since generally there will be the same number of plus pulses as negative. Greater reliability and accuracy will be obtained, however, if both plus and minus pulses are considered since in this manner a comparison will be made between the sample and the pattern where the scanning beam passes from a white area to a dark area and also where the beam passes from a dark area to a white area. For this reason I prefer to use two circuits such as Fig. 6, one at 22 and the other at 22 n, connected to the pulse transformer 23 m so that a pulse of either polarity at 19 m will cause a negative pulse at either 22 or 22 n. [Emphasis ours.]
If the signals from the means comprising detecting means are read in the first manner the signals in the last clause of the claim must read as different signals because the combining means of Hemstreet responds only to the positive transition signals, that is, the leading edge of the continuous signal. If the signals however are read as the transition positive pulses to be consonant with the signals utilized by the combining means, Hemstreet has no signals proportional to the length of the portion sensed.
Although it might seem that this would have been enough to satisfy the board that it was right in holding that Hemstreet had no right to make the count — accepting its fundamental assumption that "signals" had to mean the same thing in both clauses — the board did not stop there. Instead, concluding that the count was "ambiguous" because of the two possible interpretations which it felt the term "signals" in the second clause could be given, it resorted to Rohland's disclosure to determine what type of signal Rohland intended by the word "signals" as used in the claim copied from his patent by Hemstreet. In Rohland's disclosure, as pointed out above, initially signals continuously proportional in duration to the length of each character portion sensed are generated, then these signals are transmuted into trains of pulses discretely proportional in duration to the length of each character portion sensed, then pulse trains shorter than a set number of pulses in length are discarded, and then the remaining pulse trains are applied to the "means for combining signals." However, the board made no distinction between the continuously proportional signals first generated and the discretely proportional pulse trains finally applied to the combining means, nor did it give any effect to the fact of Rohland's discarding the shortest signals. It simply held that the pulse trains finally entering the "means for combining signals" "are `of a duration proportional to the length of the portion sensed'" and, accordingly, that the count "must be read in this manner and since Hemstreet does not generate signals which are proportional to the length of the portion sensed he is not entitled to make count."
We do not find it necessary to decide whether the board correctly held that the signals which Rohland applies to his "means for combining signals" are "of a duration proportional to the length of the portion sensed," nor do we find it necessary to get involved in determining the "gist of the invention," concerning which the parties have cited McCutchen v. Oliver, 367 F.2d 609, 54 CCPA 756 (1966), because we do not think the count is "ambiguous," and, accordingly, we will not construe the count in light of the disclosure of the application in which the count originated. Daley v. Wiltshire, 293 F.2d 677, 49 CCPA 719 (1961).
Hemstreet contends that the gist of the invention made by both litigants is "the use of a plurality of angularly related scans for the recognition of printed characters"; Rohland contends that "The gist of the invention includes scanning with a plurality of differently oriented scanning patterns, obtaining signals of a duration proportional to the length of character portions scanned, and making use of that length information in the signal-combining means for obtaining an indication of the character scanned" (footnote omitted).
If Rohland had intended to limit his claim to apparatus in which the output of the component defined by the second clause is in unchanged form the input to the component covered by the third clause, it seems to us he would likely have begun his third clause with such words as "means for combining said signals" or "means for combining the signals," and placed the limitation "produced by scanning portions of at least one or more predetermined lengths" in the second clause, where the signals were first introduced. Having drawn a broader claim, he ran the risk that others might be able to support the language selected with specifically different structures, and he may not now urge that his language should only be read in light of his disclosure merely because the language originated with him. Kuppenbender v. Riszdorfer, 104 F.2d 791, 795-796, 26 CCPA 1436, 1440-1442 (1939); Wirkler v. Perkins, 245 F.2d 502, 504-505, 44 CCPA 1005, 1008-1009 (1957).
We note Rohland did employ the language "said signals" the second time he referred to the "signals" of the second clause in the second clause, which reinforces our conclusion that the signals of the third clause need not be the same signals as the signals of the second clause.
To satisfy the claim language Rohland chose, as it appears in the count, all that is necessary is that the "signals" of the second clause be produced "when the scanning means senses portions of the character" and be "of a duration proportional to the length of the portion sensed" and that the "signals" of the third clause be " produced by scanning portions of at least one or more predetermined lengths" (our emphasis). These limitations are not mutually exclusive, and the signals of the third clause may be the same signals as those of the second clause, as the board found they are in Rohland's apparatus. But neither are the limitations in the two clauses additive, and the limitations modifying "signals" in the second clause cannot be read into the third clause. Accordingly the signals of the third clause may be the descendants of, rather than the same as, the signals of the second clause. In each clause "signals" is explicitly limited. The limitation or limitations in each clause must be given effect, but the limitation or limitations from the other clause need not be imported.
Of course it is also necessary that there be some relationship between the signals of the second clause and the signals of the third clause in the sense that the latter are at least derived from the former. Otherwise, there would be insufficient cooperation between the component of the apparatus claimed in the third clause and the components claimed in the other two clauses.
Finding that the term "signals," used in the second clause of the count with one meaning, does not necessarily have the same meaning in the third clause of the count, we hold that the count does read on Hemstreet's apparatus and that Hemstreet is, therefore, entitled to make the count.
For these reasons the award of priority to Rohland is reversed.
Reversed.