204 F.2d 273
Patent Appeal No. 5905.
United States Court of Customs and Patent Appeals.
April 15, 1953.
Rehearing Denied May 29, 1953.
T. L. Bowes, Rochester, N. Y. (Raymond W. Colton, Washington, D. C., Mason, Kolehmainen, Rathburn & Wyss, and M. Hudson Rathburn, Chicago, Ill., of counsel), for appellant.
Donald J. Simpson, Chicago, Ill. (Hill, Sherman, Meroni, Gross & Simpson, Chicago, Ill., of counsel), for appellee.
Before GARRETT, Chief Judge, and O'CONNELL, JOHNSON, WORLEY, and COLE, Judges.
This is an appeal from the decision of the Board of Interference Examiners of the United States Patent Office awarding priority of invention as to all counts in issue to appellee Marvin Camras.
The interference is between appellant's application, Serial No. 579,393, filed February 23, 1945, and appellee's application, Serial No. 612,688, filed August 25, 1945. Thus, appellee is the junior party.
The invention here involved relates to a magnetic erasing head for a magnetic wire (or tape) recorder machine. There are three counts in issue, which read as follows:
"1. In an erasing head, a core of magnetic material having a slot extending across a surface thereof, said core having a plurality of gaps adjacent each other in said surface extending across said slot, means for developing across each of said gaps by interconnected ferromagnetic circuits in said core, an alternating magnetic erasing field, a ferromagnetic medium, and means for moving said medium through said slot and successively across said gaps whereby any signal magnetically recorded on said medium previously is erased.
"2. In an erasing head, a core of magnetic material having a slot extending across a surface thereof, said core having a plurality of gaps adjacent each other in said surface extending transversely thereof through said slot, a coil for alternating erasing current encircling a portion of the core between said gaps, a ferromagnetic medium, and means for moving said medium through said slot and successively across said gaps whereby any signal magnetically recorded on said medium previously is erased.
"3. In an erasing head, a core of magnetic material having a slot extending therealong, said core having a pair of gaps adjacent each other and extending transversely thereof through said slot, means including a coil individual to each gap for developing across each of said gaps, an alternating magnetic erasing field, a ferromagnetic medium, and means for moving said medium through said slot and successively across said gaps whereby any signal magnetically recorded thereon previously is erased."
As appears from the counts, the subject matter of the invention relates particularly to an erasing head having a plurality of gaps with an alternating magnetic erasing field developed across each of said gaps. Count 1 is a generic claim, and counts 2 and 3 are each directed to a separate species.
The physical embodiment to which the counts are directed comprises a small flat core piece of paramagnetic metal having two narrow slits extending from one edge which is slotted, each slit joining an enlarged cutout portion in the core piece. Thus the core piece is generally E-shaped, with a base portion, and a center and two outside legs extending therefrom.
In the specific embodiment covered by count 2, a single coil is mounted on the center leg of the "E", and when this coil is energized two ferromagnetic circuits are developed. In one circuit, flux passes through the center leg, base, first outside leg and first slit between said legs, which slit provides a first erasing gap; in the other circuit, flux passes through the center leg, base, second outside leg, and second slit between said legs, which slit provides a second erasing gap. In the specific embodiment covered by count 3, a separate coil is mounted on each of the outside legs to generate two such ferromagnetic circuits.1
The applications of the parties indicate that the invention of the counts was the result of a search for a device which would effectively remove or "erase" a previously recorded signal from a record medium having high coercive force, e. g., magnetic stainless steel wire. Such a record medium has more desirable recording properties than those previously used, e. g., medium carbon steel types, but is more difficult to satisfactorily "erase" for reuse.
Camras, as the junior party, has the burden of proving priority of invention by a preponderance of evidence. Allen v. Blaisdell, 196 F.2d 527, 39 C.C.P.A., Patents, 951.
Both parties took considerable testimony and introduced numerous exhibits. However, our discussion of the evidence may be somewhat abbreviated due to certain concessions by both parties.
Camras concedes that Clark's evidence establishes a conception and reduction to practice at least as early as December 31, 1944, with respect to counts 1 and 2. As to the invention of count 3, the record, in our opinion, clearly establishes conception on January 2, 1945, when Clark completely disclosed the same in a fully corroborated "inventor's draft" made out as the initial step in preparation of an application for patent by his assignee's patent attorney. However, Camras claims there is insufficient evidence of diligence between January 2, 1945, and February 23, 1945, so that Clark must be restricted to his filing date as to count 3. We think the latter claim is without merit. In our opinion, the uncontradicted testimony (stipulated) by Mr. D. Clyde Jones, former attorney for Clark's assignee, shows sufficient diligence in preparing and filing the Clark application here involved which is based on the inventor's draft referred to above. Accordingly, we hold that Clark is entitled to a date for completion of his invention which is at least as early as December 31, 1944, as to counts 1 and 2, and at least as early as January 2, 1945, as to count 3.
To establish priority of invention, Camras relies upon the conception and reduction to practice, prior to Clark's earliest date, of the multiple gap combination erase-record heads of Camras' Exhibits Nos. 17, 4, and 12, described below. Clark has contended most vigorously before this court, and before the board, that those heads do not embody the invention here in issue and do not establish either conception or reduction to practice of the issue counts.
Camras' documentary Exhibit No. 17 is a copy of U. S. Patent No. 2,418,542, issued April 8, 1947, to Camras on an application filed January 20, 1944. That patent discloses, inter alia, a combination erasing-recording head having a core piece which is slotted along one edge with two narrow slits or gaps extending from said edge, the core piece being substantially like the two gap, E-shaped core pieces shown in the applications here involved. A first coil is wound on the base portion between the center leg and one outside leg with a high-frequency oscillator connected thereto. The patent refers to this as the "erasing head portion" or "erasing head." A second coil is wound on the base portion between the center leg and the second outside leg with audio signal input means connected thereto. The patent refers to this as the "recording head portion" or "recording head." The gap in the former portion is the erasing gap, and the gap in the latter is the recording gap. The wire recording medium passes over the erasing portion gap first and, according to the patent, "the high frequency flux set up by [the first coil of the erasing head portion] in the vicinity of this gap will effectively demagnetize or clean the recording medium. [Italics added.] * * * Some of this flux will stray and pass through the recording head portion of the core * * *. This stray high frequency flux is sufficient to provide the desired high frequency component in the recording head. * * *" The record clearly establishes that the "high frequency component" here mentioned is commonly referred to as the "bias field" in this art. Camras is of course entitled to a constructive reduction to practice as of January 20, 1944, for the subject matter disclosed in this patent.
Camras' physical Exhibit No. 4 is a combination erase-recording head of the type disclosed in Camras' documentary Exhibit No. 13, which is a copy of U. S. Patent No. 2,435,871, issued February 10, 1948, to Camras on an application filed August 26, 1944. The head of exhibit 4 differs in structural details from that of exhibit 17, but is essentially very similar to it in fundamental structure and principle of operation. It comprises a paramagnetic core piece having two gaps with a separate coil surrounding each of the gaps. One coil, referred to as the "erasing coil" in exhibit 13, is connected to a high-frequency source, and the second coil is connected to an audio signal input means for recording. The desired high-frequency bias field in the recording gap surrounded by this latter coil is provided by stray high frequency flux resulting from the flux produced in the "erasing coil." Camras is of course entitled to a constructive reduction to practice as of August 26, 1944, for this head, and the testimony of record indicates an actual reduction to practice on June 24, 1944, or July 1, 1944.
Camras' physical Exhibit No. 12 is a combination erase-recording head substantially like that of exhibit 4. The sole difference is that the wire of the erasing coil is also looped once around the recording gap, thereby forming a second, one-turn coil in series with the coil surrounding the erasing gap. The testimony of record indicates that the purpose and function of this second one-turn loop is to provide the desirable high-frequency bias field in the recording gap. It appears from an examination of exhibit 12 that the coil surrounding the erasing gap has about 20 turns, and it is therefore clear that the high-frequency flux in the erasing gap is of much greater intensity than that produced in the bias field. The testimony of record indicates an actual reduction to practice of this head sometime in the summer of 1944.
Both parties primarily direct their arguments to exhibits 4 and 12.
It appears to us that, in essence, Camras bases his case on the claims that (1) the heads of exhibits 4 and 12 (and 17) were reduced to practice before Clark's earliest date (Clark concedes this); (2) that the testimony of record shows that record mediums were drawn over such heads across both gaps with no audio input to the record gap coil, for the purpose of erasing only (Clark concedes this), and that the record member was satisfactorily erased as it left the head; (3) that there is present in the second record gap a high-frequency flux, whether it be a stray flux, as in exhibit 4, or a purposely induced flux, as in exhibit 12 (Clark concedes this); (4) that such high-frequency flux in the so called record gap, albeit of low intensity, provides a second alternating magnetic field in said gap; (5) that the testimony of record shows that this latter field will necessarily produce a "demagnetizing" or "erasing effect"; (6) that the claims merely describe the field in the second gap as an "erasing field" and therefore, in view of points (3), (4) and (5), supra, the counts in issue are broad enough to read directly on the phenomenon occurring in the so-called record gap when these heads are used for erasing only, so that the counts in issue are supported by the heads of exhibits 4 and 12 (and 17).
Although Clark makes a partial concession as to point (2), he does so with the reservation that, while such a head satisfactorily erases wires of relatively low coercive force which were used in the tests relied on by Camras, it will not satisfactorily erase wires of high coercive force, for which purpose the heads of the applications here involved were purposely designed. It is our opinion that the preponderance of evidence in the record clearly supports this latter point raised by Clark. Clark also contends vigorously (a) that the preponderance of evidence shows that for effective erasing it is essential to get a sufficiently intense flux in the erasing gap to sensibly saturate the recording medium as it passes through the gap; (b) that the high-frequency flux present in the second record gap of exhibit 4 and exhibit 12 (and 17) was clearly designed to produce a bias field in the recording gap, which the preponderance of evidence shows must necessarily be considerably less than that required to saturate the record medium for effective erasing; (c) that because of this latter fact there is no appreciable or effective erasure in the second gap, although this second flux in the record gap will concededly have an additional "demagnetizing" or "attenuating" effect on the record medium; (d) that the preponderance of expert testimony shows that no appreciable erasure occurs in a properly designed recording gap with bias field adjusted for optimum distortionless recording because different design factors are controlling for such gaps and erasing gaps; (e) that erasing and biasing fields for recording gaps are antithetical in nature; (f) that the "erasing field" called for in the second gap in the counts in issue is a material limitation; and (g) that if the term "erasing" is given its ordinary meaning in the art here involved it cannot reasonably be construed to read on the bias field produced in the record gaps of the combination erase-record heads of exhibits 4 and 12 (and 17) by the bias flux when those heads are used for erasing only.
The contentions of the parties as set out above were also raised before the Board of Interference Examiners. The board held that the counts were supported by Camras' exhibits 4 and 12 and awarded priority to Camras. It based its decision essentially on the grounds advanced by Camras in points (1) to (6) in the next preceding paragraph. The board also indicated that it was of the opinion that Clark seeks to construe the counts as though the word "completely" were inserted before "erasing" in the counts and held the counts could not be so construed, citing Van Hulst v. Sullivan, 93 F.2d 53, 25 C.C.P.A., Patents, 777. Noting that counts of an interference should be given the broadest construction of which they are reasonably capable, the board stated:
"* * * his [Clark's] position appears to be that the counts in issue are limited to a head, with a plurality of gaps, adapted to be used for erase purposes only and are not applicable to a combination or double-purpose head, such as that of Camras. There is nothing in the wording of any of the counts which can be construed as limiting them to a head adapted to be used only as an erase head. The counts are satisfied so long as the head, to which they are applied, has a plurality of gaps and is adapted to have at least a portion of the signal erased at each of the gaps. The counts are not exclusive in nature. We hold that they are applicable to Exhibits 4 and 12 of Camras." It is clear then that the issue before the court involves a construction of the counts in issue, not for the purpose of determining the right of the party to make them, but rather for the purpose of determining their applicability to the combination multiple gap heads of Camras' exhibits 4 and 12 (and 17) when used for erasing. Indeed, the problem of construction narrows down to the interpretation of the meaning of a single word, namely, "erasing."
To support the claim that a high-frequency bias field produces an "erasing effect" (as he terms it), counsel for Camras calls our attention to the following corroborated testimony by Mr. J. Sterling Kemp, Jr., physicist and Director of the Magnetic Recorder Division of Armour Research Foundation:
"Q. 40. Have you ever tried to make a re-recording on a magnetic record member without erasing the previous record? A. Yes, I have.
"Q. 41. Will you tell us what you have done in that connection and when you did it? A. It was my purpose to determine whether or not it is possible to record a tone, a chord, on a wire by recording first one note and then subsequently passing the media through a recording head, but not through the erase field, and record the second note of the chord, and then subsequently pass the media again through the recording head, in the absence of an erase field, recording the third note, and subsequently and in the same manner recording the fourth note.
"It was my discovery that on each successive recording the play-back level of the notes recorded previously was diminished.
"Q. 42. How did you account for that? A. It was my opinion that the high frequency bias was capable of at least partially erasing the signal recorded on the wire.
"Q. 43. And that it actually did that? A. That is right." [Italics supplied.]
Counsel urges we must conclude from this that when the two gap erase-record heads of exhibits 4 and 12 are used for erasing, both gaps produce an "erasing effect" (as he terms it) on a record member passed thereover, as a result of the alternating field in each gap.
The party Clark, however, has conceded that in the latter case there will be a "demagnetization" or "attenuation" of the record medium in the recording gap as a result of the bias field; indeed, Clark and the expert witness Holmes (identified below) so testified. But he contends that this is not "erasing" within the meaning of the term as ordinarily used in the art and as used in the applications and counts here involved.
Both parties called expert witnesses to give testimony with respect to design considerations and inherent characteristics of high-frequency erasing fields and high-frequency biasing fields.
Mr. Lynn C. Holmes, Director of Research for Stromberg-Carlson Co. of Rochester, New York (Clark's assignee), was called by the party Clark as an expert witness, and testified as follows:
"Q. 127. Mr. Holmes, did it ever occur to you to try a head of this type [such as Camras' Exhibit No. 4] as a two-gap erase head? By this type I mean a combination head. A. No, it did not.
"Q. 128. Would you explain why? A. Primarily because you would never use a gap that was designed for recording as an effective erasing gap. The conditions for the design are far different in the two cases.
"Q. 129. When you say you would never use, would you explain further what you mean by `use'? Do you mean you would not design a head in that manner? A. You would not design a head in that manner.
"Q. 130. Does your statement mean that a good recording gap is not a good erasing gap? A. Ordinarily that is true. A good recording gap is not a good erasing gap. For a good recording gap the space distribution of the field at the trailing pole must fall off rapidly as a function of distance away from that trailing edge. For the best erasing gap, the field should fall off gradually. Now we have got to say at a given recording medium speed and at a given frequency of erasing or a given frequency of biasing or a given frequency of audio frequency recording current. In other words, at a given frequency of erasing and at the given speed for the recording medium, better erasing is effected when the space distribution of the field as a function of distance from the trailing pole piece falls off gradually rather than falls off sharply. For good recording just the reverse is true, as the sharper that that field falls off the better the recording will be."
Holmes also testified that a section designed for good recording cannot be a good erasing section.
The party Clark explained the distinctions between erasing fields and biasing fields as follows:
"Q. Please explain what you consider to be the most important design considerations for recording gaps or heads as distinguished from erasing gaps or heads. A. For a recording gap the chief consideration is to get an abrupt discontinuity in the magnetic field as the wire leaves the gap. In an erasing gap you attempt to make the design such that the magnetic field will fall off somewhat gradually as the wire leaves the gap. In addition, in a recording gap in a system of magnetic recording using high frequency bias, you will normally set the bias — and by that I mean the high frequency flux in the recording gap — you would normally set this to get a good recording with respect to such characteristics as distortion and frequency response. This means that this high frequency flux in general is considerably less intense than that required to saturate the medium. In the erase gap, on the other hand, you attempt to get sufficiently intense flux so that you can sensibly saturate the recording medium as it passes through the gap."
With respect to the combination heads of the type built and tested by Camras, Clark said:
"Q. What do you consider to be the essential difference between the manner of operation of these combination heads and your two-gap erase head? A. I think that there are two essential differences. In our two-gap erase head we attempt to sensibly saturate the recording medium in each of the two gaps, whereas in a combination head with normal bias in the recording gap the magnetic field intensity is insufficient to saturate the medium in the recording gap. In addition, our two-gap erase head is designed to have a rather gradual decrease in the magnetic field intensity as the wire leaves each gap, whereas in a combination head the design is intended to give an abrupt decrease in the magnetic field intensity as the wire leaves the gap."
We find nothing in the record which convincingly rebuts the above testimony by Holmes and Clark. Indeed, that testimony is supported by testimony of Dr. David Edgar Wiegand, Senior Physicist at Armour Research Foundation in Chicago, Illinois (Camras' assignee), who was called as an expert witness by the party Camras. On cross-examination, he was asked whether the bias in the recording gap of a combination erasing-recording head (specifically one of the type in Camras' Exhibit 17) would, in his opinion, be chosen for complete erasure or for optimum performance as a recording gap. He answered:
"A. The head would be designed so that the lefthand [erasing] gap would essentially completely erase any previously made recording, and the proportions of the legs would be so set that when the lefthand gap is operating in this manner, a proper amount of biasing field for distortionless recording is superimposed on the audio field in the righthand [recording] gap." [Italics supplied.]
Dr. Wiegand also testified further, on cross-examination, that different flux distribution is required for erasing and recording: "Q. 74. In your experience, is there any particular pattern that is better adapted to erasing than any other?
"The Witness: A. In general, a gradual tapering off of the field is required, but when you work the head at high fields, you automatically get that, due to saturation at the tips of the gap.
"Q. 75. In other words, the space distribution, is the word I was trying to think of — the space distribution of the field, it is preferred to have it taper off in erasing? A. That is right.
"Q. 76. Is that true in recording gaps? A. In the recording gap it is more desirable to have the field terminate rather abruptly.
"Q. 77. From what you see here, and based on your experience, would you say that the righthand [recording] gap * * * is designed to give the preferred space distribution of an erase field?
* * * * * *
"The Witness: A. Well, the primary consideration in the design of the righthand gap, for the reasons I mentioned previously, the fact that we have the audio amplifier and the microphone there would indicate to me that the primary consideration would be to make it serve effectively as a recording gap."
Appellee's counsel correctly asserts that the issue here is that of priority under the counts as drawn, Kirby v. Clements, 44 App.D.C. 12, and that counts of an interference should be given the broadest interpretation of which they are reasonably capable. Van Hulst v. Sullivan, supra. But, in applying this latter rule the word "reasonably" should not be deleted, Harris v. Henry, 63 F.2d 120, 20 C.C.P.A., Patents, 883, and the crucial language in a count should not be given an unwarranted overbroad interpretation. Words in a count are generally to be given their ordinary meaning in the particular art involved. Dirkes v. Eitzen, 96 F.2d 849, 25 C.C.P.A., Patents, 1176.
We are of the opinion that the testimony of record, some of which is quoted above, shows by an overwhelming preponderance of the evidence that an erasing gap and field is completely different from a recording gap having a bias field, insofar as design factors and inherent characteristics are concerned. Indeed, we agree with counsel for Clark that the two are antithetical in nature and function.
Furthermore, Clark and Holmes testified that, in their opinion, although a recorded signal on a record medium would probably be attentuated (i. e., lessened or weakened) by passage through a bias field, it would not thereby be "erased" in the sense that the wire would then be in suitable condition for re-recording. We find nothing in the record which convincingly rebuts this testimony; note the italicized portions of the above-quoted testimony by Kemp which is relied on so heavily by Camras. Hence, we are of the opinion that the preponderance of expert testimony shows no appreciable erasure occurs in a properly designed recording gap of a combination erase-record head having a bias field adjusted for optimum recording; and that no appreciable "erasing" (defined infra) occurs in the heads of exhibits 4 and 12 (and 17) when used to erase only.
We think that the preponderance of the evidence of record indicates that the ordinary meaning of the term "erasing" in this art is the effective demagnetization or cleaning of the recording medium2 so as to substantially remove therefrom as thoroughly as possible (not necessarily completely) any existing signal impressed thereon. Likewise, we think the record shows that an erasing head must have an alternating high-frequency field with maximum intensity values greater than anything on the recording medium, i. e., it must be of sufficient intensity to produce saturation. The record also clearly shows that a bias field is no more than a relatively low intensity high-frequency field produced in the record gap to enable recording in an operating range on the linear portion of the magnetization curve so as to minimize distortion of the signal recorded; its intensity must necessarily be below that required to produce saturation of the recording medium. Clearly then, the bias field in the record gap of a combination erase-record two gap head could not meet the requirements of an effective erasing field and still properly function as a bias field. In short, we think the conclusion is inescapable that a "bias field" is not an "erasing field," and an "erasing field" is not a "bias field" as those terms are ordinarily used in this art. Consequently the heads relied on by Camras do not support the issue counts.
In our opinion, Camras seeks no less than to have us construe counts which clearly call for an erasing head having a pair (or plurality) of gaps with an erasing field in each gap so that they will cover an erase-record head having a pair of gaps with an erasing field in one gap and a bias field in the second gap, said bias field having a relatively insubstantial demagnetizing effect when the head is used solely for erasing. Such a construction is completely unwarranted.
We note also that even if we were in doubt as to the ordinary meaning of the term "erasing" in this art, we would be compelled here to hold that the term "erasing field" in the counts in issue is not supported by either of Camras' exhibits 4 or 12 (or 17). The record clearly shows that the problem appellee sought to solve by developing those heads is materially different from that which caused him and appellant to develop the two gap erasing head of the issue counts disclosed in the applications here involved. In such a case, the counts are not to be given a construction broader than the invention disclosed in the involved applications so as to read on a prior reduction to practice by one party to which his application was not intended to relate, notwithstanding that the issue counts could reasonably be construed to cover the same. Chalmers v. Romine, 62 F.2d 622, 20 C.C.P. A., Patents, 821. The Chalmers case involved a problem substantially identical with that before us and is regarded as directly controlling in Clark's favor.
It is noted that the Camras' application was not filed until August 25, 1945. Camras seeks to rely for priority on U. S. Patent No. 2,418,542, filed on January 20, 1944, as a constructive reduction to practice of a head supporting the counts. He also seeks to so rely on an actual reduction to practice of the head of exhibit 4 which is disclosed and claimed in U. S. Patent No. 2,435,871 (Camras' exhibit 13), filed August 26, 1944. The involved invention solved an apparently serious problem confronting the art, yet we are unable to find in either of these Camras patents any hint of the problem here solved, or how to solve it. If the combined erase-record heads disclosed in these two Camras patents (upon which heads he seeks to predicate priority) had embodied the invention of the issue counts, it is inconceivable to us that Camras would not have disclosed it in the applications for these patents had he in fact conceived the involved invention and reduced it to practice at that time. Vreeland v. Miller, 112 F.2d 829, 27 C.C.P.A., Patents, 1284. The record on the part of Camras is voluminous, and his documentary exhibits are numerous. Yet we find nothing in that record indicating that Camras had any inkling of the two gap erase head invention here involved prior to the earliest effective date to which Clark is entitled. Indeed, after a careful study of this extensive record, we can only conclude that Camras' whole case is an attempt to establish nunc pro tunc that exhibits 4, 12 and 17 embody the invention of the issue counts.
One other point remains for consideration. Counsel for Camras urges also in his brief that Camras should be awarded priority because:
"Clark directly acquired knowledge of the Camras multiple gap head from a study of a detailed technical bulletin describing one of such heads sent to Clark's assignee by Camras' assignee, and Clark is therefore not the inventor."
The bulletin here referred to was introduced in evidence as Camras' Exhibit No. 27. This bulletin, which Clark admits he read, describes the construction and overall characteristics and use of "a combined record-erase head" such as that of physical exhibit 4, and the related patent of documentary exhibit 13. The foregoing ground urged by Camras is nothing less than an allegation of derivation of the invention by Clark from Camras. We find nothing in the record to indicate that the question of derivation was raised before the board. Moreover, even if it were, the question of derivation substantially changes the nature of an interference proceeding, making the essential question one of originality rather than one of priority. Hence it would have been necessary for Camras to cross-appeal from an adverse ruling of the board on the allegation of derivation. Therefore, it is clear that the question of derivation is not properly before us. However, lest our silence on the question put Clark in an unfavorable light, we may state that we have thoroughly studied the bulletin in question, and we find therein not a scintilla of evidence to support the above allegation of derivation.
In view of the foregoing, we are of the opinion that the board's construction of the issue counts is completely untenable both in law and in fact. We therefore hold that the counts in issue are not supported by either of Camras' exhibits 4 or 12 (or 17). Accordingly, Camras has failed to meet his burden of establishing priority of invention and priority is awarded to Clark as to all counts in issue.
For the reasons hereinbefore set out, the decision of the board is hereby reversed.
1. The embodiment relied on by Camras to support count 3 varies somewhat from this description. Since such variation is immaterial to the outcome of the case, it is deemed unnecessary to discuss it in detail.
2. This language is practically identical with that used by Camras himself in U. S. Patent No. 2,418,542, which is Camras' exhibit 17, supra.