value of the number to be represented, or whether the number is represented by a single hole or a plurality of holes provided at a certain place of the line to be sensed. [Emphasis added.]

Tauschek thus teaches the same relationship between the number of marks and the digit represented thereby as does appellant, the only difference being in the nature of the marks-holes in Tauschek and magnetic ink in appellant's case.

Bryce I, relied upon by the board only with respect to claim 42, appears to us to be a particularly pertinent reference, relating as it does to

a system in which the control elements or records are provided with magnetic conditions representing various characterizations and data, and are adapted to be presented to a machine to convert the magnetic conditions on the control elements into electrical currents or impulses to control the machine in accordance with the magnetic indications. [Emphasis added.]

The reference continues:

The specific type of accounting system described herein comprises a system wherein the machine control elements are controlled at differential times by the electrical energy generated by the control system by sensing the coded or differentially positioned magnetic conditions disposed under control records.

The process of magnetically recording data in the form of discrete magnetic conditions or impressions depends essentially on the phenomenon of magnetic hysteresis and on the associated property that when changes of magnetomotive force are made upon a magnetizable substance a certain remanent flux density is retained by the magnetized substance which is termed "remanence." *

The remanent flux densities retained by the control records may then be effective at a subsequent period to initiate electrical current conditions or impulses which are impressed upon various control circuits to control the operation of the different machines of the present system. [Emphasis added.] Bryce I discloses means for recording, reading, and obliterating the magnetic impressions, a system for amplifying the output from the reading means, accumulating and printing mechanisms actuated by the amplified current impulses, and several different embodiments of data records. One such record is shown in Figs. 8 and 9:

One of two plies of paper stock 26 has applied thereto "individual magnetizable portions 27" (emphasis added) which "may be printed thereon similarly as ordinary printed matter" (emphasis added). In this embodiment, the plies of stock 26 are then secured together by an adhesive 28. It is apparent from the patent that the magnetizable portions are imbedded between the two plies only because this protects them from abrasion during handling and provides a more durable construction than would exist if the portions were merely deposited on the surface of a single ply. To record data on a card of the type shown in Figs. 8 and 9, the portions 27 are selectively magnetized according to the desired coding arrangement. In another data record embodiment, a similarly shaped card has essentially its entire surface coated with magnetizable material, statistical information or data being arranged thereon in the form of "arbitrary [magnetic] indications coded according to their positions on the record." In either of the above-mentioned embodiments, the "magnetizable material * * * may be in the form of metal powder or filings suitably fixed upon a carrier" which is "preferably non-magnetic."

Describing the principle of operation of the tabulating mechanism, Bryce I states:

* records having discrete magnetic impressions representing the data are presented to suitable reading magnets and the said impressions are detected while the records are in motion to initiate timed impulses in accordance with the disposition of the impressions on the record to control either data accumulating mechanism or data printing mechanism or both * * *. [Emphasis added.]

Claims 1-6,14-16, and 45-49

The rejection of these claims as unpatentable over Knutsen and Begun is sound. Claims 1-6, 14, 15, and 49 are directed to methods, 16 and 48 to apparatus, and 45-47 to data records. More specifically, claims 1-6 and 49 are drawn to a method of recording and reading data wherein deposits of magnetic material are affixed to (or, as in claims 1 and 4, written on) a sheet of non-magnetic material in a configuration recognizable by a human being as indicative of the data represented thereby, the configurations are magnetized in a predetermined direction, and each configuration is passed by the air gap of a magnetic reading head to produce in the reading head "electric wave shapes characteristic of" (claim 1), or a "voltage having a waveform representative of" (claim 4), or "a flux proportional to" (claims 5 and 6) the configuration of the deposit. Claims 4 and 6 also require the provision of means for identifying each configuration according

Claim 1 does not indicate the type of material on which the deposits are affixed and claim 49 does not positively recite the step of affixing the magnetic material to the sheet.

to the voltage waveform induced, while claims 5 and 6 require that the magnetic material have "a high value of coercivity and a high value of remanence." Claim 4 is representative of the method claims:

4. A method of recording data for both visual and magnetic reading thereof, comprising

writing on the surface of a non-magnetic sheet a distribution of finely divided magnetic particles,

said distribution having a configuration recognizable by human beings as being indicative of the data represented thereby.

magnetizing said entire distribution of particles to establish in said distribution magnetic lines of force lying parallel to a predetermined direction,

thereafter passing said sheet relative to the air gap of a magnetic reading head to cause relative movement of said distribution in proximity to said air gap, the dimension of said air gap between the poles thereof extending parallel to said predetermined direction during the relative movement between said distribution of magnetic particles and said reading head to induce in said reading head a voltage having a waveform representative of the configuration of said distribution of particles on said sheet,

and providing means responsive to said voltage for identifying each character in accordance with the waveform of the voltage induced thereby in said reading head.

Claim 14 is similar but specifies that the line of magnetic characters be printed, that magnetization of the characters be done parallel to the line, and that flux produced in the reading head be utilized to control entry of the represented data into a machine. Claim 15 is also similar to claim 4, but, instead of the means-for-identifying clause, concludes with "whereby to provide a plurality of electric signals representative of said data."

Claims 16 and 48 are drawn to apparatus for reading a line of magnetic ink deposits. The apparatus includes an air-gap type of magnetic reading head, means for moving the deposits past the air gap, means for producing output voltages related to changes in magnetic flux across the gap, and means responsive to time-displaced portions of the voltage waveforms so produced in the magnetic head. Claim 16 is representative:

16. Apparatus for electrical recognition of each of a plurality of different characters written on a document as a distribution of magnetic ink and arranged in a single line spaced apart from one another comprising, in combination, a magnetic read head having a pair of opposed magnetic pole faces defining an air gap therebetween and further having means for producing an output voltage of a value related to the change in magnetic flux across the air gap, said air gap having a narrow first dimension representing the space separating the pole faces and an elognated second dimension extending at right angles to the first dimension, means for causing relative movement between a document bearing a line of such characters and said magnetic read head to cause movement

of said line of characters past said air gap character with the narrow first dimension of the air gap extending in the general direction of the line of characters and with the elongated second dimension of the air gap extending in a direction substantially parallel to the height of said characters thereby to provide a progressive scan of each character in the direction of said line, said magnetic head being responsive to the magnetic characteristic of the ink of which said characters are formed and to the relative movement of each character past the air gap thereof to cause the output voltage producing means of the magnetic head to produce a unique output voltage waveform for each character scanned, and means electrically connected to said magnetic head for receiving said output waveforms and responsive to time displaced portions of said waveforms for identifying each character read.

Claims 45-47 are drawn to data records. Claim 45 recites a nonmagnetic sheet having thereon "data-significant groups" of discrete masses of ferromagnetic material of high permeability. Claim 46 calls for "data-bit representations" each being formed of a discrete "individually magnetically sensable mass of ferromagnetic material of relatively high permeability," while claim 47 just calls for groups of short parallel lines formed of ferromagnetic material of relatively high permeability. These claims do not specify that the material is magnetized.

With respect to the rejection based on Knutsen and Begun, the board said:

[Knutsen] writes with a pencil having magnetic properties * * on a tabulating card. The writing is adapted to be read and recognized by the human eye of one skilled in the art in the same sense as appellant's characters 12 are read and recognized by the routineer.

Knutsen's preferred mode of sensing is a photo electric or conductive form. However, he recognizes that the other forms of marking require different means of sensing * * *. Magnetic records, their characteristics, the various directions of recording and the types of pickups * * *were well known in the magnetic recording field prior to appellant's entry into the field and would be known and available to the routineer prior to the filing date of the present case.

No patentable significance is attached to the manipulative step of magnetizing the magnetizable material of Knutsen. If this material is not magnetized at the time it is deposited on the tabulating card by the user's pencil, or by the earth's magnetic field then it will be magnetized by the pickup head when the character is sensed during playback or reproduction. [Emphasis ours.]

Appellant's first main contention is based on the observation that neither of these references (Knutsen and Begun) teaches the magnetization of discrete, magnetic deposits. We are not impressed by the board's statement that the deposited magnetic material, if not previously magnetized, would be magnetized by the reading head, since a number of these claims recite the reading step as taking place after the magnetizing step. Neither do we find any reason to think that any magnetization that might incidentally be provided by the earth's magnetic field would be of any operative significance. Nevertheless

we are of the opinion that one skilled in the art surely would have considered the recited magnetization step obvious. We note in this regard that appellant has not disputed the following statement by the board:

It is well known in magnetic recording and reproducing that a signal may be obtained from a magnetic material by (1) magnetizing the material and later drawing it under a conductor since this is a well known principle of an electric generator or (2) by drawing a magnetized ferromagnetic material into the gap of a pole piece which results in changing the reluctance of a magnetic circuit which in turn causes a change in voltage in a coil. Both procedures have been used in the recording and reproducing art.

Given this, and Knutsen's concept of using discrete deposits of magnetic material on data records, we think that one skilled in the art would have recognized immediately that the magnetic material when magnetized would induce signals of greater amplitude in a reading head than when unmagnetized.10 Adding Begun's teaching of magnetization in a single selected direction, we are of the opinion that one skilled in the art would likewise consider it obvious to magnetize uniformly in one direction (particularly along a line of short parallel lines when using an air gap reading head situated as in appellant's Fig. 4, supra), in order to further enhance the signals induced. Begun's teaching of the desirability of high coercivity and remanence values in powdered magnetic materials would also, in our opinion, render obvious the corresponding limitations in claims 5 and 6.

Appellant's other main contention with respect to this rejection is that the references "fail to teach the use of magnetic deposits shaped to represent data." Specifically, he says this is true as to Knutsen because the "location of [the] marks on the card has data significancenot [the] configurations" and because the marks "are all identical in configuration." As for Begun, appellant says that "nowhere is there any suggestion that data significant electric waveforms might be developed from magnetic deposits having data significant geometrical configurations."

Although appellant has consistently referred to the groups of short parallel lines in his disclosed embodiment 11 as "characters," we are not persuaded that appellant's markings, in the final analysis, are any more "data significant" or "differently configured" than are those of Knutsen. In appellant's embodiment, it is apparent that each of the parallel marks, even though it may be grouped with one or more others, represents one item of data, that the significance of a given mark is determined by its being located within a particular area of

10 The "enhancement of the signal level" is one result that appellant attributes to the magnetization step.

11 All the claims are clearly intended to encompass this embodiment.