US2928595A - Computing apparatus - Google Patents

Description

March 15, 1960 c. J. GOODALE COMPUTING APPARATUS 7 Sheets-Sheet 1 Filed June 8, 1953 WM WW r w P W (192,}, 15ml H M NH! H 6 mm 7 w L Jul w m March 15, 1960 c. J. GOODALE COMPUTING APPARATUS 7 Sheets-Sheet 2 Filed June 8, 1953 M 0 0 a M swam/o Mew? ,k; Jfiddale March 15, 1960 c. J. GOODALE 2,928,595

COMPUTING APPARATUS Filed June 8, 1953 7 Sheets-Sheet 3 .Zizaentar JMrylaslfqa,

March 15, 1960 c. J. GOODALE 2,923,595

COMPUTING APPARATUS Filed June 8, 1953 7 Sheets-Sheet 4 iiazykyel March 15, 1960 c. J. GOODALE 2,928,595

COMPUTING APPARATUS Filed June 8, 1955 7 Sheets-Sheet 5 Juafeniar fiaqziale March 15, 1960 c. J. GOODALE COMPUTING APPARATUS 7 Sheets-Sheet 6 Filed June 8, 1953 frwerzkr kar'z'as' J 624% @M March 15, 1960 c. J. GOODALE COMPUTING APPARATUS '7 Sheets-Sheet 7 Filed June 8, 1953 United States Patent COMPUTING APPARATUS Charles J. Goodale, Springfield, Mass.

Application June 8, 1953, Serial No. 369,121

37 Claims. (Cl. 23561.8)

The present invention relates to computing methods and to machines capable of carrying out such methods wherein mechanical movements may be largely replaced by electronic devices. This application is a continuation in part of the copending application Serial No. 678,152 filed June 20, 1946, now abandoned.

Some of the main objects of the invention are to pro' vide an improved method of carrying out algebraic summations by means of scale elements which method can be carried out by simple manual aids as well as by automatic computing machinery, and to provide computing apparatus which, controlled by a record moved therethrough, performs quickly and reliably mathematical operations, especially such summations, as indicated by indicia applied to the record.

Other objects of the invention are to provide an accounting machine which performs its operations in simple and yet reliable manner, unhampered by requirements introduced by complicated counting and accumulating procedures and accordingly more or less complicated mechanical gear; to provide a computing machine which materially reduces the length of time between the recording of collected data, and the evaluation thereof at the same or another place; to provide a computing machine which is composed of operating units which are only electrically interconnected, so that they can be easily rearranged and combined to form complete machines suitable for various purposes; to provide apparatus of this type which furnishes final data without any preliminary or subsequent sorting operation of the records from which this data are collected and without the necessity for highly manual preparatory work, which sorts automatically by way of a source recorder or subordinate unit, and which furnishes these data in a minimum period of time from the first recordation thereof; and to provide computing apparatus of this type which eliminates any need for reading, copying, or transcribing steps intermediate the initial collecting of the data concerned, and the final evaluation thereof.

Further objects of the invention are to provide fully effective and reliable apparatus for defining and putting into effect the shifting reference principle used according to my above-mentioned improved computing method, to provide efiective and simple apparatus for translating accumulated data into a visible record, to control the computing operations of the machine by means of counting and operational indicia and sequence indicia combined in a manner which renders these operations speedy, reliable and practically independent of irregularities of the indicia, and to provide electrically controlled and actuated means for shifting counting elements, for operating contact controlling elements, and for operating printing mechanisms in a particularly simple and effective manner.

In one of the aspects of the invention, algebraic summations are carried out by defining digits of a notation which may also be referred to as series of numbers by scales of equal, digit indicating steps, and by shifting 2,928,595 Patented Mar. 15, 1960 "ice these series in the same sense, whereby the shifting of one series adds, and shifting of the other series subtracts digits defined by the amount of shifting; the relation of the two series may be indicated by a zero of home indicating point of one of the two series which home point shifts or floats with its series relatively to fixed instrumentality to which both series can be shifted, the zero point alone being used for computing whereas the fixed instrumentality itself is not used for purpose of computing. The digit or number series of the notation may be defined by straight or circular scales which can be shifted in one direction, the same for both scales, relatively to a support represented by the fixed instrumentality.

In another aspect of the invention, the above objects are attained by coupling one or the other of two counting members to an actuating means which is adapted to advance stepwise in a given direction under the control of sequence indicia, and under the control of means for coupling one or the other counting member to the actuating means under control of and during steps determined by counting indicia conjointly with the sequence indicia, means being in addition provided which are controlled by operational indicia for selecting one or the other of the counting members for coupling by means of the counting indicia to the sequence indicia actuated stepping means; one of the counting members is provided with means, such as electric contacts which are part of a relay circuit, for indicating the relative position of the two counting members independently of their position regarding the machine considered as a mechanical structure.

Additional features of the invention include magnetically controlled movements which serve as clutching and switching elements and operate by shifting a coupling member from a permanent magnet fastened to one of two relatively movable elements to an electrically controlled magnet fastened to the other element which may be a rotatable shaft; commutator switches with circular contacts whose angle relation conforms to the operational relation of the above mentioned counting members and permits a wide range of correlation of computing operations as controlled by the record indicia; an arrangement for printing computation results by means analogous to the accumulating means; switching means, controlled by operational indicia, for setting up circuit connections which selectively induce various computing operations such as adding and totalling as herein described in detail, but also subtracting, carrying over, printing and resetting; and various photo-electric and control element supervised electron tube circuits which, in conjunction with timing and energy storing networks, perform many of the essential functions of the present computing apparatus.

These and other objects, aspects and features will appear from the following description of a typical practical embodiment illustrating the novel characteristics of my invention. This description refers to drawings in which Fig. 1 is an isometric view of a cabinet incorporating various computing units constructed according to the invention;

Fig. 2 is the view of a record strip for use with apparatus according to the invention;

Fig. 3 is a diagram illustrating the operation of apparatus according to the invention;

Fig. 4 is a longitudinal section through the detecting unit of the apparatus shown in Fig. 1;

Fig. 5 is a section on lines 5-5 of Fig. 4;

Fig. 6 is a section on lines 6-6 of Fig. 4;

Figs. 7 and 8 represent a view, partially in axial section, of the accumulator unit of the apparatus shown in Fig. 1;

Fig. 9 is a section on lines 9-9 of Fig. 7;

Fig. 10 is a section on lines 10-10 of Fig. 9;

Fig. 11 is an exploded isometric view of the clutch mechanism shown in Figs. 7 and 8;

Fig. 12 is a section on lines 12--12 of Fig. 7;

Figs. 13 and 14 are elevations showing the contact faces of the commutator switch disks shown in Fig. 7;

Figs. 15 and 15a are schematical combined views of the commutators shown in Figs. 13 and 14, in two char acteristic positions;

Fig. 16 is a vertical section through the printing unit of the apparatus shown in Fig. 1;

Figs. 17 and 18 are elevations showing the contact faces of the commutator switch disks shown in Fig. 16;

Fig. 19 is a schematical combined view of the commutators shown in Figs. 17 and 18;

Fig. 20 is a circuit diagram of the electrical apparatus incorporated in the herein described embodiment of my invention; and

Figs. 21, 22 and 23 are diagrams illustrating the improved computing procedure according to the present invention.

Apparatus according to the present invention lends itself very well to unit construction and to the assembly ofunit elements, each performing its particular function, in operational combinations and spatial set-ups selected according to individual requirements. While the embodiment herein described represents a machine that performs operations satisfying average computing requirements, it should be understood that it can be used as an element of apparatus fulfilling different, perhaps more elaborate, requirements and that its units and elements can be employed in combinations other than that herein referred to.

The apparatus now to be described incorporates assemblies which will be referred to as scanner, accumulator, recorder, and circuit units. These units may be incorporated in a cabinet such as shown in Fig. l which indicates at the housing proper, screwed to a base plate 31 which is also shown in Fig. 4. The cabinet has a feeding platform 32, and doors 33 and 34 permitting access to the various units. Fig. 1 indicates by way of dot and dash lines how the above mentioned units or components of the machine may be distributed within the cabinet 30, I indicating the detecting unit, II the accumulator unit, III the recorder unit, and IV the circuit unit. These units will be taken up in the above indicated order, but before explaining the detector or scanning unit, a record means suitable for purposes of the invention will be described.

Record As indicated in Figs. 2 and 3, a suitable support 40, preferably in the form of a strip or tape, carries counting or numerical marks 41, operational marks 42, and sequence marks 43. These marks, whose respective significance will be presently described, or indicia may be applied by means of a recording machine such as described with reference to Fig. 29 of my patent No. 2,251,998 of August 12, 1941. If desired, the sequence marks 43 may be initially applied to the strip 40, and the counting and operational marks added in proper register with the sequence marks in the manner to be explained below.

The counting marks or generally speaking digits of a notation are in this embodiment of the nature of code element indicia, such that for example one or a selected combination of four marks of one column of a number group define one digit of the notation. The strip shown in Fig. 2 records numbers having up to five digits. The uppermost group of Fig. 2 represents the number 56789 in the manner also shown in Fig. 3 which carries nine code mark columns arranged so as to indicate the manner in which the nine digits are defined. Each digit is made up of four marks arranged in a column and the digit column position defines the digit order or decade. Each individual code mark defines a certain numerical value, in the present example values 2," 2, 1, and "4. Thus, the digit 1 is defined by the code mark "1 having the third position in its column, digit 2 is defined by the first mark "2" in the respective column, digit 3 is defined by the code marks 2 and 1, digit 4 is defined by code marks 2 and 2, and so on, as clearly indicated in Figs. 2 and 3. It will further be noted that these marks are not particularly distinguished except as to position in their respective columns, each column of course defining, as pointed out above, the positional or order value of its respective digit. As mentioned above, the marks of Fig. 2 permit the recording of numbers with five digits, whereas a strip having nine columns as that in Fig. 3, would permit the recording of nine digit numbers. It will be understood that digit 2 may be defined by the second mark 2 and that indeed any sequence of code mark value positions may be used.

As will be noted from Fig. 2, the sequence marks 43 may consist of a continuous black strip containing groups of equally grouped blank spaces or slots which are in register with the code mark rows. Fig. 3 indicates the relative position of code mark elements and sequence marks, the latter being shorter, and within the former. Their relative position is not particularly critical, so long as each sequence mark is within a corresponding counting mark component. In the present embodiment of my invention, a fifth sequence mark 49 is added which has no corresponding code mark row, but is located between consecutive code mark groups defining respective numbers.

Any number of operational indicia 42 may be applied to the strip, their significance being defined by their lateral position on the strip. Thus, mark 45 has a meaning different from that of mark 44, and any required number of similar marks may be applied between the code marks and the strip of sequence marks. In the present embodiment of my invention, a code mark group 48 covering the entire code mark area always goes with the mark in position 45, for reasons which will appear hereinbelow.

These marks may be applied in any suitable medium, the only requirement being that they affect a selected type of radiation, visible or invisible, differently from the matter in which it is affected by the record carrier proper. Thus, the marks may be applied in printers ink or stamp ink, when they will reflect light to a much lesser degree than the fairly smooth and light surface of the clear record strip of paper, Celluloid or other material.

The above mentioned significance of the code mark values, namely 2 2 1 4, and this sequence of the values, are not unique, other combinations being quite feasible, as for example 5 1 3 2. The significance of this particular arrangement of the record strip indicia in view of the computing apparatus according to the present embodiment of my invention will be explained more in detail hereinbelow.

Detector As mentioned above, the scanning or detecting unit of the machine is contained in compartment I of the cabinet shown in Fig. 1. As indicated in Figs. 4 and 6, it carries the record strip 40 on two rolls or reels 51, 52, one of which, for example 52 may be suitably coupled to and driven by the motor 55 of a type appropriate for this purpose. The strip 40 is guided by two idling rolls 56, 57, along the underside of a scanning plate 60 having a scanning opening 61, and therebelow a platform 62 which flattens the record strip 40 as it passes the scanning window 61. Reels 51, 52 and rolls 56, 57 are journalled in bearing brackets 64, fastened between unit base 36 and plate 60.

Brackets 65 (Fig. 4) support two scanning units 66, 67 on plate 60. The scanning units have housings 68, 69 with cover plates 71, 72 and perforated walls 73, 74, which support light conducting rods 70, 80, leading from the interior of the respective scanning housings to the scanning window 61. The scanning terminals of these light conductors may have reduced diameters, as indicated in the figures, in order to accommodate them to the approximate magnitude of the indicia elements of strip 40. As indicated in Fig. 5 as related to Fig. 2, the scanning rods are placed opposite the various indicia columns of the record strip 40. By shaping the light conducting rods as indicated in Figs. 4 to 6, it is possible to reduce the dimensions of the marks and hence of the record strip to any desired degree.

The scanning unit 67 carries within one compartment 81 (Figs. 4 and 6) several photo-electric tubes Pd, and a flash lamp F. The number of photoelectric tubes Pd in compartment 81 (in this instance seven) corresponds to the number of code mark order columns and operational mark positions or columns and hence to the number of light conducting rods leading into this compartment. whereas only one flash lamp covers the entire series of conducting rods, which, as indicated in Figs. 4 and 6 terminate directly in front of the flash lamp. The photo tubes Pd in compartment 81 are so arranged that each of them receives considerable light flux when lamp F radiates. A second compartment 82 separated from compartment 81 by an opaque wall 85, contains a light source E which is associated with a light conducting rod 91 arranged similar to those leading into compartment 81.

Scanning unit 66 (Figs. 4 and 5) likewise has two compartments, namely 83 and 84, separated by opaque wall 86. Compartment 83 contains seven photo-electric tubes Pi so arranged that each of them can receive light from a coordinated light conducting rod 70, but not from any other rod. The compartment 84 contains a photo-electric tube Pt associated with a light conducting rod 92.

The light conducting rods leading into scanning units 66 and 67 respectively, are so placed and inclined that the rods 80, 91 emerging from unit 67 direct light beams onto the respective mark series of recording strip 40, whereas the light conducting rods 70, 92 leading into unit 66 are adapted to receive light which might be reflected into them by the bare and comparatively clear surface of the record strip, where it is not covered by recording material.

The various photo tubes will be further identified with reference to the circuit diagram of Fig. 20.

Accumulator The accumulator unit which may conveniently be arranged in compartment II of Fig. l is mounted on base board 101 which may be fastened to the main base plate 31 of the cabinet, and has side frame boards 102 and 103, as indicated in Figs. 7 and 8 which together constitute a single view, partly in section, of the accumulator assembly. The movable portions of the accumulator are guided upon or fastened to a shaft 105 which is journalled in frame walls 102 and 103 by means of ball bearings 108 and 109.

The shaft 105 can be rotated in steps of prearranged magnitude, by means of stepping or sequence solenoids Z1, Z2, Z4 (Figs. 7 and 9) each of which consists of a magnet winding 111 and a housing 112 fastened to plate 101 by brackets 128. Into housings 112 are screwed holding plates 115, which secure the coil within the housing, and also determine, by means of rims 116, the travel of armature plungers 118 which are by means of rods or links 119 and one way couplings 120 joined to shaft 105. As shown especially in Fig. 10, couplings 120 may be ball clutches consisting of two bearing rings 121, 122. and therebetween a clutch ring 123 which has a cam recess 124 confining a ball 125 between its face and the shaft 105. As well known, the clutch body will be attached to shaft 105 only when it rotates in the direction indicated by the arrows of Figs. 7 and 10, whereas the shaft is released when the clutch body is rotated in the opposite direction. Thus, the shaft 105 is rotated through steps corresponding to the stroke of each respective plunger if the latter moves in the direction of the arrow of Fig. 9. It will be observed that rotation of shaft by solenoid Z1 will not affect the other solenoids whose one way clutch will then be idle, and vice versa. As indicated in Fig. 9, the strokes of solenoids Z1, Z2 and Z4 are stepped in the ratio 122:4. The plunger of solenoid Z2 floats, but its length is so selected that, when de energized, it permits the plunger of solenoid Z4 to assume its full stroke, whereas when Z2 is energized, its plunger advances to an extent which presets the stroke of the plunger of Z4 to one-half of its full value. Hence, (a) the stroke of solenoid Z2 pushing the plunger of de-energized solenoid Z4, (b) the stroke of solenoid Z4 as preset by solenoid Z2, (c) the stroke of solenoid Z1, and (d) the normal stroke of solenoid Z4 as shortened by the previous stroke of solenoid Z2, bear the ration relation of 2:2:l:4. As will be fully explained hereinbelow, these strokes are effected by appropriate circuit control, their four values correspond to those of the four counting code marks discussed above with reference to Figs. 2 and 3, and they permit the definition of any one of the nine digits in any decade position.

The accumulator unit further contains a number of digit counter sets which are in this art sometimes also referred to as accumulator sets, one for each order or position. Only two of these sets are shown in Fig. 7, namely the units and tens counter sets. Each counter set consists of two permanent magnets 131, 132, which have bores permitting shaft 105 to rotate freely therewithin, and which are fastened to base plate 101 by brackets 134. Fixed to shaft 105 are core bodies 135, 136 of magnetic material. Solenoid coils Zn and Zs, mounted on brackets 165, surround core bodies 135, 136 and define gaps 160. Rotatably connected to shaft 105 by means of ball bearings 141, 142 are two clutch sleeves 143, 144, each of which carries a commutator disk, indicated at 145 and 146, of insulating material, and has a coupling slot, indicated at 155, 156, compare also Fig. 11.

These commutator disks carry, riveted or otherwise fastened thereto, five concentric contact sectors s1, s2, s3, s4, s16 and companion sectors s11, s12, s13, s14 and s15 as indicated in Figs. 13, 14 and 15. The contact sectors of the respective commutator disks are in sliding contact making engagement while in coinciding angular position. It will be noted that Fig. 13 is a view seen from the right hand side of Fig. 7, whereas Fig. 14 is a similar view seen from the left hand side of Fig. 7; Fig. 15 shows the commutator contacts of both disks in super-imposed position, seen from the left hand side of Fig. 7. As indicated in Fig. 13, the angular distance between the extreme ends of the commutator contacts s1, s2, s3, s4, s16 of disk 145 is 36, that is 5 of the entire circumference, and corresponding to a step of value 1. Ten equidistant points of the circumference of this disk define ten digit positions, from 0 to 9, as likewise indicated at Fig. 13. The ends of the contact segments s11, s12, s13, s14, s15 of commutator 146 are similarly defined, as indicated in Fig. 14. The respective disks 145 and 146 rotate, when coupled to shaft 105, in the same direction, as indicated by the arrows applied to the respective figures.

Between permanent magnets 131, 132 and cores 135, 136 are arranged coupling slides 151, 152 which have two claws or prongs 153 engaged in the above mentioned slots 155, 156 of clutch sleeves 143, 144. The last mentioned two elements are isometrically shown in Fig. 11. The two slots 155, 156 engage the two claws 153 permanently, although the latter can move axially therewithin, so that, when coil la is de-energized, disk 151 is attracted to permanent magnet 131 retaining sleeve 143 immovable with regard to the frame. On the other hand when coil Za is energized, the attractive force of the permanent magnets is overcome and disk 151 is attracted by core 135, which, it will be remembered, is fast to shaft 105, so that claws 153 will now move sleeve 143 with them. In the first mentioned instance, namely with disks 151 or 152 fast to permanent magnets 131 or 132 (which are mounted on the apparatus base), sleeve 143 or 144 are fixed relatively to the base but permit rotation relatively to shaft 105 on ball bearings 141 and 142; in the last mentioned instance, namely when coils Za or Zs are energized, shaft 105 and disks 143 or 144 rotate together. As will appear hereinbelow, Za and Zs are never simultaneously energized.

On radii corresponding to those of the commutator segments s1 to s and s11 to s15, are arranged, on the outside of disks 145 and 146, collector slip rings 170 which are conductively connected to the commutator segments, for example by the above mentioned rivets. These rings cover uninterruptedly the entire 360 arc of the disks, and continuously contact individual brushes 180, as indicated in Fig. 7.

As mentioned above, the accumulator has a counting set for each digit position; in the present five sets would be provided although only two are shown in Figs. 7 and 20.

The accumulator also has several relay units, which are shown in Fig. 8.

The switching elements of these relays may assume various forms, whereas their actuating structures are essentially similar, as will now be explained with reference to Fig. 8. This figure indicates at 171 and 172 permanent magnets which are fastened to base 101 by means of brackets 175 and 176. Core supports indicated by numeral 181 are fastened to base 101 with brackets 185 and have guiding slot portions 186. They carry contact bases 187 of insulating material. Coils Zup, Zos and Zoe are fastened to base 101 by means of brackets 165 and form gaps 195 with cores 192 of magnetic material that are fastened to core bodies 181. Within the gaps 195 move contact carriers 194 which are similar to the coupling elements shown in Fig. 11, with the difference that one claw or if desired each claw carries contact 198 or contact bridges 199. As indicated in Fig. 8, there may be contact bridges on either side of the movable claw of switch element 197, or one or several contact pairs to be connected by these bridges, may be arranged on one or either side of the movable claw. The various contacts shown in Fig. 8 are not individually identified because they correspond to those identified and described in detail with reference to the circuit diagram Fig. 20. It will now be evident that energization of coils Z moves the switching bridges such as 194, 197 in the direction of the axis of shaft 105, while de-energization causes them to be attracted by the permanent magnets. This mode of actuation is quite similar to that of the accumulator and commutator sets, and permits use of similar parts for both types of mechanism.

It will be evident that various clutches and switches can be constructed in similar manner, in accordance with varying circuit requirements, and arranged in compact units as indicated by Figs. 7 and 8.

Recorder The recorder unit is arranged in the compartment indicated at III of Fig. 1. It is mounted on a base plate 201 and on frame walls 202, 203 and 204 (Fig. 16), with the base plate 201 secured to the main base plate 31 of the cabinet. In a space formed by walls 203, 204 and by a printing base plate 206 is mounted a printing device with a hammer solenoid Zh which is adapted to lift. when energized, a printer hammer 211 against the underside of recording card 200, inserted through slot 212 (Fig. 1) between guides 215 and 216. As indicated in Fig. 1, the stepped upper portions of cabinet compartments I and II form a feeding platform 32 provided with indicating lines 219, which permit registration of the card 200 with the printing mechanism. The card may be discharged through slit 220 in the rear wall of the cabinet, indicated in dotted lines in Fig. 1. An ink ribbon 221 carried by rolls 222 and 223 supplies printing medium in well known manner. When hammer 211 is lifted, it presses card 200 against ribbon 221 and the type wheels to be described hereinbelow, thus recording the number defined by the type wheel digits.

A shaft 205 is journalled in side walls 202 and 203 by means of ball bearings 228 and 229. This shaft performs a function similar to that of the above described accumulator shaft and can be rotated stepwise by means of solenoids Zpl, Zp2 and Zp4 which are in the present embodiment identical with the accumulator solenoids Z1, Z2. Z4 described with reference to Figs. 7 and 9, and advance the shaft through the corresponding steps of 36, 72 and 144. In order to simplify Fig. 16, only one solenoid Z124 with its clutch is shown.

The shaft 205 is further associated with a number of clutch elements, one for each order or position, which are similar to those described with reference to Fig. 7, with the only difference that they are equipped with gear wheels 231 in addition to the commutator disks 233, and that stationary commutator disks 235 are fastened to the energizing coils Cp. It will be noted that gear wheels 231 rotate whenever coils Zp are energized while shaft 205 turns, coupling the bearing supported switch and gear wheel sleeves 238 to cores 242 which are fixed to shaft 205. This also provides selective rotation of the commutator disks 233.

The stationary commutator disks 235 have two concentrical contact rings s22, s23 with gaps extending through 36, as shown in Fig. 17, while the rotatable disks have two contact segments, s25, s26, covering an angle somewhat smaller than 36, as shown in Fig. 18. Fig. 19 shows the two disks in superimposition. The arrows applied to Figs. 16, 18 and 19 indicate the respective directions of rotation, it being assumed with respect to Figs. 18 and 19 that these figures show the contact faces of the disks of Fig. 16. The contact elements of these commutator switches are also shown in Fig. 20 and their function will be further explained with refer ence to that figure.

A supporting rod or stationary shaft 245 is fastened to frame walls 202 and 204 and carries several hollow shafts, one for each type wheel and hence one for each order position which the machine in question is required to handle. In the present instance five type wheels 251 to 255 (corresponding to the five digit positions of Fig. A) are shown which are correspondingly fastened to hollow shafts 261 to 265. Each type wheel is supplied with raised type numbers 0 to 9, opposite the above mentioned ribbon 221. The other ends of shafts 261 to 265 are fast to gear wheels which mesh with the gear wheels 231 of the printer clutch units. Fig. 16 shows only two of these gear wheels, namely 271 and 272; it will be understood that each type Wheel has one of these gear wheels which again are associated with the appropriate number of clutching and commutator switching units, each with permanent magnet 241, core 242, coil Zp, commutator disk 235 and rotatable sleeve 238 with gear wheel 231 and commutator disk 233.

It should be noted that the digit numbers of the type wheels run in reverse direction as compared to those of the accumulator wheels (as indicated in Fig. 13), having in mind the respective sense of rotation of these elements.

Circuit The complete circuit diagram of a computing machine incorporating the various features of the herein described embodiment of the present invention is shown in Fig. 20.

In Fig. 20, the terminals a, b of a standard alternating current supply system feed through a suitable transformer into terminals h to which the heaters of all electron discharge tubes of the system may be connected. Through another transformer the alternating current system energizes the lamp E which is the detecting lamp shown in Fig. 6 in compartment 82 of scanning unit 67. The alternating current line a, b further supples a recti- 9 her indicated by tube Vr and filter Nr, which supplies the bias for tube 6 in the manner to be described hereinbelow, through tap 301 which is less positive than ter minal a. The main supply of direct current for operating the circuit is indicated by positive bus A and negative bus B.

A sequence detecting and flash initiating network NJ includes a photo tube Pt whose anode is connected to bus A and whose cathode is connected through resistor R1 to bus B. A gaseous discharge tube Gf having a control electrode gf is connected with its cathode to B and with its anode to a point 302 between a resistor R2 and a capacitor C1. Resistor R2 and capacitor C1 are connected in series between buses A and B. A flash lamp F is connected between tube 6 and point 302, in series with a resistor R2. A resistor R3 is connected in parallel across lamp F of conventional gaseous flash discharge type.

The above mentioned lamp E is arranged for illumination of photo tube Pt, through the light conducting rods 91 and 92 shown in Figs. and 6. The lamp F is arranged for direct flash illumination, in the manner described below, of a series of photo tubes Pd contained in scanning unit 67, as shown in Figs. 4 and 6, and for indirect illumination, through the light conductors 70, 80, of a series of photo tubes Pi contained in compartment 83 of scanning unit 66, as shown in Figs. 4 and 5. These photo tube series Pd, Pi are arranged in the clutching and switching tube circuit to be described below, but are in Fig. 20 also represented by single photo tubes Pd, Pi shown adjacent to flash tube F, in order to facilitate understanding of the operation of the circuit. In Fig. 20, the dotted box containing these photo tubes also indicates the direct illumination of tubes Pd and the indirect illumination of tubes Pi through light reflected from the record strip 40 when the light strikes portions of that strip which are not covered by the recording material.

The sequence circuit Ns now to be described contains a number of gaseous discharge tubes G1 to G5, preferably of the hot cathode or thyratron type. One of these tubes is provided for each sequence mark 43 of the record. It will be understood that more than five sequence slit marks may be provided, in which case the circuit will contain additional corresponding thyratrons. The cathodes of the thyratrons G1, G2, G3, G4 and G5, are connected to the negative bus B through resistors R11, R12, R13, R14 and R15. The thyratron cathodes are joined through a ring circuit including conductor 303 and capacitors C11 to C15. A grid control conductor 304 connects point 303 between the cathode of photo tube Pt and the resistor R1, through a time delay network with resistor R4 and a capacitor C4, to condensers C21 to C25 which lead to the control electrodes g1 to g5 of thyratrons G1 to G5. These control electrodes are also connected, through resistors R21 to R25, to corresponding sides of the above mentioned ring capacitors C11 to C15.

The anodes of thyratrons G1 to G4 are connected to the positive bus A through stepping solenoids Z1, Z2, Z4, which are described above with reference to Figs. 7 and 9. Solenoid Z1 is in the anode circuit of tube G3, solenoid Z2 in that of tube G1 and solenoid Z4 in the anode circuits of tubes G2 and G4. In parallel with these solenoids, but connected to an auxiliary bus A31, are printer solenoids Zp1, Zp2 and Zp4. The auxiliary bus A31 can be connected to the main bus A through the normally open contacts s31 of switch Zop which is also shown in Fig. 8 and which will be further described hereinbelow. Between the anode of tube G5 and bus A31 is connected the hammer solenoid Zh (described above with reference to Fig. 16), and the anode of thyratron G5 is also connected through conductor A32, which contains solenoid Zoe and normally closed contacts s32 of switch Zop, to positive main bus A.

The above described sequence circuit Ns functions as follows:

It will be noted that the grids g1 to 35 of the sequence control tubes G1 to G5 are connected in parallel and therefore receive together any pulse which becomes effective at conductor 304 of the sequence network Ns due to changed impedence characteristic of photo tube Pt. It is characteristic of this circuit that only one of the several sequence tubes is on at a time, whereas, whenever the circuit receives an impulse, one of the tubes becomes conductive whereas the preceding one becomes extinguished.

Assuming that, for example, tube G1 is conductive, point 311 on the cathode of G1 has a fairly high potential (although not sufl'iciently high to fire tube G2 through resistor R22 and grid g2) as compared with that of ground B, due to the comparatively small drop across the tube, as compared with that in resistor R11. On the other hand, point 312, associated with the next tube G2, has practically the potential of B, because the tube is non-conductive therefore representing an infinite resistance. Due to these conditions, capacitor C12 is charged, with the plate next to resistor R22 positive and the plate next to point 312 negative. When photo tube Pt receives the next light flash, from lamp E and one of the sequence slits or marks 43 of the record strip, momentarily increasing the potential of wire 304, only grid g2 of tube G2 becomes sufficiently positive to render G2 conductive, because this grid g2 carries a positive bias due to the above mentioned charge on capacitor C12. With G2 conductive, point 312 assumes a higher potential. With capacitor C12 not yet discharged, point 311 assumes an even higher potential making the cathode of tube G1 more positive than its anode, so that G1 extinguishes. Capacitor C12 now gradually discharges, whereas capacitor C13 charges as above explained with reference to capacitor C12. This cycle continues indefinitely for each potential increase of wire 304, corresponding to a light flash received by tube Pt from exciter lamp E. Each thyratron becoming consecutively conductive in the ring circuit, the solenoids in their respective anode circuits are energized, for purposes to be explained in detail hereinbelow.

When the potential of point 304 increases, due to a flash from lamp E received by photo tube Pt by way of a sequence mark or slit 49, the potential of the control electrode gf of flash control thyratron Gf is raised, rendering GI conductive and permitting a discharge of normally charged capacitor C1 through the flash lamp F, in circuit C1302-FGfB-C1. Hence, F delivers an instantaneous light flash for each sequence mark that directs increased light intensity from exciter lamp E to photo tube Pt. As mentioned above, the discharge device Vr supplies the bias for grid gf of thyratron G1 which is extinguished by the voltage collapse across capacitor C1 which causes the voltage across tube Gf to go momentarily below its critical value.

For reasons which will become apparent from the following explanation of the operation of the present device as a whole, the sequence control tubes G1 and G5 should be triggered after the clutching or counting tubes to be described below have become effective. For this purpose, the above mentioned time delay network with resistor R4 and capacitor C4 is provided between points 303 and 304 which introduces a predetermined time delay between the flashing of lamp F and the triggering of the respective thyratron due to the rising potential level of conductor 304.

The circuit according to Fig. 20 further includes a clutching and switching control circuit Nc responsive to the counting and operational marks 41 and 42. This network has a series of pairs of high vacuum discharge devices, which may be separate tubes or contained in the same envelope, as shown in Fig. 20. These tubes operate on the well known multi-vibrator principle and are indicated in Fig. 20 at Vs, Vp, Vu, and Vt. Tubes Vs and Vp respond to operational marks, whereas Vu and Vt operate the counting systems of the accumulator. Tube Vu corresponds to the unit order, Vt to the tens order, and so on, so that for a record strip with five positions such as indicated in Fig. 2, five counting tubes would be provided. Fig. 20 shows only the tubes Va and Vt for the first two positions but it will be evident, as indicated by arrows applied to buses A and B and the auxiliary buses, that any desired number of counting tubes can be added. Also, additional operational tubes, corresponding to Vs and Vp, may be adapted.

Each of the operational or counting multi-vibrator double tubes is associated with the conventional timing capacitors, and bias and load resistors. With the grid g21 which controls the off or non-actuating circuit of each tube (indicated only for tube Vs as containing resistor R23) is connected to a system of photo tubes Pi, Pd. These are the photo tubes also shown in Fig. 20 in association with flash network Nf. It will be understood that there are, as shown in Figs. 4 to 6, two photo tubes for each operational mark, and that there is a pair of counting photo tubes for each order position.

Each multi-vibrator tube has a load resistor R23. The actuating instrumentalities, in the present instance solenoids Zos, Zop, Zsu, and Zpu, etc., are connected between anode n22 with terminal 306 and bus A, in circuits containing the switching devices of commutators Na and Np and relays Zop, Zos, and Zoe, whose structure was described above and whose function will be described in detail below. It will be evident that, if any tube is conductive through anode 022, the corresponding magnet Z will be energized, whereas switching of conductivity to anode 021 will deenergize that magnet.

It should be noted that the photo tubes whose reference characters contain the indicia Pd correspond to the directly illuminated tubes also indicated at Pd in Figs. 4 to 6 and in network N) of Fig. 20, and that the photo tubes marked Pi are the indirectly illuminated photo tubes, namely those which receive their light from reflecting portions of the record strip 40. With the record carrier 40 reflecting the light beam from lamp F, as indicated in Fig. 20, Pd as well as Pi are illuminated, and point 321 (indicated for photo tubes Pis and Pds of the subtracting operational tube Vs assumes an intermediate potential. The various switches are then set as indicated in Fig. 20, and the counting or clutching magnets are tie-energized.

This counting and operational network Nc functions as follows:

When a code of operational marks 41, 42 passes between the detecting ends of rods 70 and 80 (Figs. 4, 5, 6), leading from the flash lamp to the respective Pi tubes, the corresponding Pi tube becomes dark and its impedance increases, whereas the Pd tubes remain illuminated, so that the potential of point 321 rises.

The multi-vibrator tubes V have normally low conductivity. When a photo tube Pi becomes dark for an instant as above described, and the potential of point 321 consequently rises, the left hand section with anode a21 of tube V becomes considerably more conductive, resistor R23 effects an increased voltage drop thereacross, and grid g22 becomes more negative. Capacitor C1 having been in condition of rest, suddenly sees its left hand side (on R23 and a21) drop in voltage. The capacitor being comparatively slow in discharging, its right hand side (on g22) must see the same voltage drop. Grid g22 thus has its normally negative voltage and in addition this voltage drop through capacitor C21. The right hand side of the tube (with n22) therefore remains off. But gradually capacitor C21 discharges until its right hand side is at the same potential as the left hand side. This potential being positive with respect to the cathode, the right hand side with n22 becomes fully conductive. With load 2 now carrying more current, the potential of grid g2! is depressed and the left hand section assumes low conductivity. The right hand section regains normal low conductivity shortly after the first section, upon charge of condenser C21 to the point where the potential on grid g2l becomes more negative than that of the cathode. The nonconductive state or the right hand side of the tube is however rather unstable since capacitor (:21 discharges only slowly through R21 and the conductive left hand side of the tube until g22 again becomes sufliciently positive to render the right hand side conductive. The respective anode potential lowers, a negative pulse is transmitted through c22 to g21 which turns off the left hand side establishing the stable condition prevailing at the beginning of the cycle and until the next pulse arrives. Thus, both sections retain their low conductivity until the next impulse arrives, again due to diflerent illuminations of photo tubes Pd and Pi respectively. The relative voltage levels of cathode and grids are due to the voltage drop in cathode bias resistor R20 during normally low conductivity of the tubes (when they are for practical purposes nonconductive) which bias resistor R20 prevents a continuous oscillation condition.

Connected to the anodes 022 of the counting tubes Vu, Vt, etc., are adding solenoids Z0, and parallel thereto subtracting solenoids Zs, and printing solenoids Zp. The other sides of these solenoids are connected to positive bus extensions B22, B21, and negative extension B11, respectively.

The energizing anode of the subtracting tube Vs is connected to positive bus A, through solenoid Zos and extension B2. Solenoid Zos operates a switch with contacts s51 and s52 which normally connects A through B2 to B22 through s51 and la, but upon closing of relay Zos connects A to B21 and the Zs clutch magnets.

The energizing anode of totaling and printing tube Vp is connected to positive bus A through solenoid Zop, extension bus B22 and contacts s51. Solenoid Zop operates a switch with four contact elements s31, s32, s41, s42 which normally connect bus A to A32 and bus B with B12, and, with Zop energized, connect A with A31 and B with B11.

The solenoids la and Zs are the clutch solenoids shown and described with reference to Fig. 7; solenoids Zp are shown in Fig. 16; and solenoids Zos and lap are the switch or relay solenoids shown in Fig. 8.

The anode of thyratron G5 leads, in parallel to hammer magnet 211, to solenoids Z001 and Z002, one of which is indicated at Zoc of Fig. 8. These solenoids operate switches s71 and s01, s02. Solenoid Z001 operates normally closed contact s02 connecting a capacitor C0 in series with a resistor R0 across lines B and A1, and normally open contacts s01 between s02 and solenoid Z002. Solenoid Z0c2 operates normally open contacts s71 for connecting line B1 to line B.

The sequence and counting circuits are correlated by the commutator switches, one for each positional or order tube Vu, Vt, etc. These are the commutator switches described with reference to Figs. 7, 13, 14, 15 and 15a. Their contacts s1, s2, s3, s4, and s11, s12, s13, s14 are correspondingly indicated in Fig. 20.

Fig. 20 shows the accumulator commutator contacts at Na, in the position indicated in Fig. 15, with the exception that contact s11 is not continued to overlap contact s1 but broken off, in order to simplify the drawing. Contacts s1 and s2 are joined, through condenser C40, to auxiliary bus B1. Contact .93 is directly joined to B1. Contact s4 is joined to auxiliary bus B11. Contact 511% leads to point 321, between the photo tubes Pdt and Pit of the next higher order tube Vt; thus, as shown in Fig. 20, s11 of the first commutator is connected to grid connection 321 of the tens digit switching tube Vt; $11 of the next or tens commutator is similarly connected to 321 of the hundreds tube, and so on. Contact s12 is connected to bus A1. Contact s13 is connected to auxiliary bus B12. Contact s14 is connected to the grid 322 of the respective counting tube.

The printing solenoids Zp are associated with the printer clutches and commutator switches shown in detail in Figs. 16 to 19. Fig. 20 shows these commutator switches at Np in the position of Fig. 19, but with contacts s22 and s23 broken off, and contacts s25 and s26 somewhat removed from s22 and s23. Contact s22 is connected to auxiliary bus B11. Contact s23 is connected to auxiliary bus A31. Contact s25 is connected to auxiliary bus A32. Contact s26 is connected to auxiliary bus B12. Each accumulator set has a corresponding printing commutator, as explained with reference to Figs. 16 to 19, the two commutator sets being co'related as just described.

Operation Code system.--In recapitulation, it should be remembered with reference to Figs. 2 and 3, that the code system employed in the present embodiment of my invention comprises the following elements. A record carrier, such as a strip of Celluloid, plastic material, or paper carries for each digit of a number to be fed into the machine a series of counting marks as follows. Each digit, regardless of its order, is made up of a number of, for example four code marks, each of which signifies a certain numerical value. As indicated in Fig. 3, the first (in the direction of the record progress) code mark may denote the number value 2, the next code mark likewise the number value 2, the third code mark the number value 1, and the fourth and last code mark the number value 4. Any desirable number can be made up from these code marks, as indicated in Figs. 2 and 3. Fig. 3 shows the marks for a nine digit number 123456789. It will be noted that digit 1 consists of the third code mark denoting value one, that digit 2 consists of the first code mark denoting value two, etc., the digit 9 being made up of all four code marks adding up to 9. It should especially be noted in this connection, that the code marks employed according to the present invention actually effect adding or counting operations involving the numerical value of the code mark, and are not mere symbols of an intervening operation of translating them into actual numbers. 7

One side of the code marks, and in the present embodiment extending over the entire length of the four code marks for each number, additional marks are applied to the record, which are herein referred to as operational marks. Such operational marks are for example the subtracting mark indicated at 44 of Fig. 2, and, in another column, the "totalling mar 45. Additional operational marks may be added according to requirements. In accordance with the embodiment herein described, the totalling mark 45 accompanies a totalling code mark which covers the entire code mark area, such as indicated at 48 in Fig. 2.

The record further carries marks which are herein referred to as sequence marks or slits and which may consist of an opaque or non-reflecting strip which carries a series of transparent or non-absorbing areas 43, one in the approximate center of each of the four code marks, and an additional one, number 49, beyond the last code mark row. It will be understood that instead of being clear, these sequence marks may be opaque or nonreflecting, which modification requires a corresponding change in the photoelectric circuit, in conventional manner.

Generally speaking, the sequence marks are placed at designated positions within the item entry or digit code marks. The operational marks are similarly related to the item entries and to each particular mark position upon the record as a whole is assigned a specific function to be performed by the machine when the respective 14 mark arrives at the detecting element in corresponding position.

It should be understood that the above described marking system is optional, and that any system may be used which is able to perform the functions according to the present invention as herein described.

Normal condition-Under normal conditions, all electro magnets are de-energized. Consequently, the counting shafts are stationary, and the counter disk and switching armatures rest on their permanent magnets.

The exciter lamp E is illuminated, whereas the flash lamp F is dark.

The gas filled tubes G1 and G4 of the sequence circuit Ns are non-conductive, tube G5 is conductive upon passing of the last slit of a group of sequence marks 43, and both sections of each counting and operational tube V of counting circuit Nc are likewise non-conductive.- Both detecting and compensating photo tubes Pd and Pi of the counting circuit are in high impedance condition.

Accumulating (adding).A record strip 40 has been inserted and, driven by motor 55, moves at constant speed through the machine. The sequence trigger tube Pt is periodically illuminated by light coming from lamp E through the light conductors 91 and 92 shown in Figs. 5 and 6, once for each sequence slit 43. As above described, consecutive illumination of photo tube Pt triggers the thyratrons G1 to G5 in sequence, with a slight time delay relatively to the flashing of lamp F, caused by the circuit C4-R4. Assuming that the time elapsing between the passing of the first sequence slit to that of the last slit 49 of a group takes 20 time units, the slits follow each other in intervals of five time units so that each gas tube remains conductive during five time units. The solenoids operated from the sequence tubes G1 to G5 require, in the present embodiment, energization times of two, two, one, four and one time units, respectively. It will thus be seen that these solenoid operation times fit easily within the time intervals elapsing between the effectiveness of consecutive slit marks. It should further be noted that, as mentioned above, the flash lamp F becomes effective prior to energization of the sequence solenoids. Consequently, the counting or multi-vibrator tubes V become energized prior to conductivity of the sequence tubes, so that for each row of marks, the counting circuits are set up by the multi-vibrator tubes prior to energization of the respective thyratrons, as indicated in Fig. 3. By means of properly dimensioning the conventional timing circuits of the multi-vibrators, the clutch and relay magnets are retained in energized condition during a time period covering the shifting operation of stepping solenoids Zl, Z2, Z4.

In order to understand the operation of the shaft rotating mechanism, it should be remembered that the sequence solenoids Z1, Z2, 24 are coordinated to the sequence tubes G1 to G5 as follows:

The first sequence tube G1 is connected to solenoid Z2 which, as above described, rotates, when energized, the shaft (Figs. 7, 8, and 9) by way of the armature of solenoid Z4, through an angle of 72 which defines the number value two" and corresponds to the first code mark. The sequence tubes G2 and G4 are connected to solenoid Z4 which, as above described, is so arranged that it turns, if solenoid Z2 is energized at the time of energizing Z4, the shaft 105 through an angle defining the number value two whereas if not limited by solenoid Z2, it turns the shaft through an angle of 144 defining the value four. The sequence tube G3 is connected to solenoid 21 which is adapted to rotate shaft 105 through an angle of 36 defining the number value one.

As above described, tube G5 is connected to a solenoid 200 which operates the switch s71 for connecting line B to the commutator switches Na. Tube G5 is further connected to solenoid Zh which operates the printer hammer.

While the accumulator solenoids Z1, Z2, Z4 are directly connected to line A, the hammer solenoid Zh and the printer solenoids Zpl, Zp2 and Zp4 are connected to an auxiliary bus A31, which leads to contacts s31 of switch Zop, as above described. It will be remembered that the accumulating solenoids Z1, Z2, Z4 are energized every time one of the first four slits 43 passes light through conductors 91, 92, rendering photo tube Pt more conductive, and, with some time delay caused by network R4, C4, rendering the tubes G1, G2, G3, G4 and G5 consecutively conductive; the multi-vibrator tubes V which operate the accumulating magnets are energized somewhat before the sequence solenoids become active.

Observing now for example the accumulation of number 56789 as shown at 41 of the record strip 40 of Fig. 2, it will be noted that the first two code marks of the order or column defining digit "5 are blank so that, although the first two sequence slits will operate solenoids Z2 and Z4 respectively and correspondingly rotate the shaft 105, the coupling magnets of the ten thousand set of the accumulator will not respond and the corresponding accumulator disks will remain stationary. The third code mark of the ten thousand column will render its tube V conductive, and the third slit will shortly thereafter energize sequence solenoid Z1. The rotating shaft 105 thus finds the tens of thousands order counting disk connected thereto, and since solenoid Z1 defines the unit 1, the counting disk will be advanced one step, through 36. The code mark of the next row similarly affects the 10,000 counting tube and the corresponding fourth slot energizes sequence solenoid G4, causing the accumulator disk to advance four number values. This, together with the previous single step, defines the digit 5. Similarly, the digit 9 in the unit position of the above referred to number 56789 will be accumulated by consecutive energization, four times, of counter coupling magnet which connects the corresponding accumulating disk to shaft 105 each time one of the three sequence solenoids rotates the shaft, so that the disk will be advanced first two number values, then again two values, then one value, and then finally four values, together nine values. In this connection it should be remembered that solenoid Z2 is energized immediately prior to energization of solenoid Z4 and that Z2 is not deenergized until Z4 becomes energized. Thus the stroke of the actuator moved by solenoid Z4 will start from the point to which it was previously advanced by solenoid Z2, and this energization of Z4 therefore effects but a two step advance of shaft 105.

The above discussed steps are defined by the position of the accumulator disk 145 (Fig. 7), with regard to that of the home or zero point of the subtracting or reference disk 146.

It will be noted that for the above described adding operation, the switch Zos is in the position shown in Fig. 20, with contacts s51 closed, and contacts s52 open. Switches Zap and Zoe are likewise in the position shown in Fig. 20. The commutator contacts of the several accumulating sets move relative to each other, but this movement has significance only when the nine to zero region of the accumulator disk 145 coincides with the home point of the reference disk 146, as will be presently explained.

Accumulating (subtracting).--For subtracting, the above described apparatus utilizes a novel procedure which is illustrated in Figs. 21, 22, and 23, and the general principle of which will now be described.

This subtracting method is based on the concept of a floating home or advancing zero position of the subtracting and reference element. Fig. 21 shows an adding element 345 and a subtracting and reference element 346 in normal position. In this position, the zero point of the adding element 345 coincides with the home or zero position h of the reference element 346. For the accumulating or adding of numbers or, generally speaking, of digits of a notation, element 345 is advanced in the.

I. OCCUI'S.

16 direction of the arrow shown in Fig. 22. For example, in order to accumulate the number four, element 345 is advanced four steps until the point defining four coincides with the home point h of element 346. If another number, for example three, is to be added, the element 345 is advanced another three steps, until the point defining seven coincides with h of 346. If it is desired to subtract a number, the element 346 is advanced in the same direction which was used in advancing strip 345 for carrying out accumulating or adding operations. Thus, if it is desired to subtract three from four, as illustrated in Fig. 23, strip 346 is advanced three steps in the direction of the arrow, so that number one of strip 345 now coincides with floating mark h of element 346.

It will be evident that this subtracting method can be carried out with elements of any desired type. If, for operations with small numbers, straight elements are used, the numbers have to be repeated several times; for most applications, round computing elements will be used, each divided into ten parts defining the ten digit values. Apparatus of this type is illustrated in the present computing machine, where elements and 146 (Fig. 7) correspond to elements 345 and 346 of Figs. 21 to 23.

If the computation results are to be mechanically or otherwise recorded some means have to be used for rendering the mechanical position of the floating zero or home independent of its actual meaning with regard to the computing operations. The apparatus according to the present invention exemplifies such an arrangement in the commutator arrangement particularly illustrated in Figs. 13, 14, 15 and 15a. As will be described hereinbelow more in detail, this arrangement permits detection and evaluation of the relative position of the counting elements, independently of their mechanical relation to the physical structure of the apparatus.

One possibility, likewise in accordance with the present invention, of subtracting by means of the floating home method will now be described.

The subtracting operation is initiated by means of a subtraction defining operational mark 44 (Fig. 2), which operates in the above described manner the subtracting tube Vs, by means of photo tubes Pds and Pis and their light conductors as shown in Figs. 4, 5, 6, and 20. Referring again to Fig. 2, the subtraction mark 44 indicates that the number "00536 is to be subtracted from the sum of the previously handled numbers. When, during the flash from lamp F, mark 44 renders photo tube Pis less conductive than Pds, tube Vs becomes conductive through anode a22 and energizes solenoid Zos. This solenoid then disconnects, at s51, the adding solenoids Za from line B and connects instead at s52, the subtracting solenoids Zs. It will be noted that mark 44 (Fig. 2) triggers tube Vs five times, each time when a flash due to passing of one of the five corresponding slits 49 The control circuit of tube Vs is so adjusted, according to well known principles as explained above, that the tube section containing anode a22 remains conductive throughout the passing of mark 44. The values of capacitors 021 and e22 for tubes Vs and Vp and hence their discharge times differ from those of the corresponding capacitors of tubes Vu, Ft, etc. The former correspond to the length of the operational indicia 42 (Fig. 2) whereas the latter correspond to the length of the code marks. At the same time, the first code mark, for example the mark defining value two as the first component of number six in the above mentioned number 00536," energizes the unit position solenoid Zsu, thus attaching the respective subtracting and floating home element 146 (Fig. 7) to shaft 105. Thereupon, the first sequence slot 43 renders tube G1 conductive, energizes solenoid Z2 which shifts the subtracting element two units. The next two code marks are empty, but the fourth causes a shift of four values, completing the number six. Similar operations simultaneously take place for th digits in the higher positions. After the entire numher to be subtracted has thus been evaluated, the end of mark 44 causes tube Vs to switch its conductivity back to normal effectively non-conductive condition, deenergizing solenoid Zos and conditioning the circuit for connecting the adding clutch magnets Za instead of the subtracting clutch magnets Zr. If a further number is to be subtracted, its code marks would again be accompanied by a subtracting mark 44, which would maintain the subtracting clutch magnets in operation, and again cause the subtracting and reference disks 146 to be attached to shaft 105, instead of the adding disks 145.

It is again emphasized that in accordance with my new subtracting principle, both adding and subtracting elements move always in the same direction, as pointed out above when relating the movement of disks 145 and 146 to shaft 105, which, of course, is rotated always in the same direction by solenoids Z1, Z2, and Z4.

The reference point, namely point of disk 146 (Fig. 14), rotates in the same direction around shaft 105, but the relative position of the adding and subtracting elements or disks is evaluated by means of the commutator, whose operation will be described in detail hereinbelow, under the heading Total Taking. It will also be noted that the possibility of actuating the adding and subtracting disks by the same stepping solenoids, is due to the fact that they move in the same direction.

Carry-0ver.-As well known, a carry-over operation has to take place whenever an accumulating element moves from its position 9 to its position 0 relatively to the zero or home point. In the present instance, where co-directionally moving counting elements are used for algebraic summation as well as for defining the home point, the carry-over operation has to take place whether due to movement of the positive, that is adding, or the negative, that is subtracting and reference disk. Whenever the point 9 of an adding disk is about to move to bring the zero point of that element into coincidence with the floating home point of its reference and subtracting element, whether this is due to the movement of the adding or the subtracting disk, a carry-over operation must be carried out for bringing about a change of digit value one in the next higher order. There is, however, one distinction, namely this change of one value in the next higher order has to be an increase when caused by an adding operation in the lower order, but decreased when due to a subtracting operation in the lower order.

In the present embodiment of my invention, the carryover operation is carried out by means of the commutator construction shown in detail in Figs. 13 to 15a. The The correlation of the two commutator disks 145 and 146, in normal or starting condition for a carry-over, is indicated in Fig. 15, which is repeated in the circuit diagram of Fig. 20, the latter, however, not showing the entire contact s11, although the continuity of that contact is indicated by joining its broken off end, shown in Fig. 20, to point 321 of the next higher group of circuit Nc. The carry-over operation is first set up for each unit in which the above described progress from 9 to 0 takes place, but it is actually carried out by means of the capacitor indicated at C40 of Fig. 20.

Referring to Figs. 13 to 14, it will be noted that the commutator contacts s1 and s2 of disks 145 are electrically connected and normally in contact with contact s11 of the reference disk 146 which contact s11 is only interrupted through the partial length of one unit, following the position of the home or zero point. As indicated in Fig. 20, the contact s11 is at 321 connected to the grid of the clutch switching tube of the next higher group, contact s11 of the units" set being connected to 321 of tens" tube Vt, whereas the contact s11 which goes with Vt is analogously connected to the grid of tube for the hundred" digits, and so on. Normally (that is excepting the period from the time when point nine" of the addingdisk passes from the home point of the other element, until point zero" of this adding element reaches the home point) capacitor C40 is enabled to discharge through the connection s1s2s11. The complete discharge circuit of capacitor C40 is C40- s11-321-PitBs72-C40. Fig. 15 shows the superimposed commutator disks (seen from one side as if transparent and hence rotating in the same direction as indicated by the arrows) in the adding position when point nine of the adding disk (Fig. 13) coincides with the home point of the subtracting disk 146 (Fig. 14), further rotation of the adding disk bringing both zero points into coincidence. Contact s3 of the adding commutator disk is normally open when the point nine of that disk reaches home. After point 9 of the adding disk passes home, s1s2 become disconnected from s11 and instead connected to s12 (on bus A). At the same time the normally open commutator contact s3 is connected to commutator contact s13 of the other disk which is through contacts s42 connected to the bus B, with relay lap in the normal position shown in Fig. 20. It should be noted that capacitor C40 is normally disconnected from line B by switch s71 which is operated by condenser CO under control of solenoid Z001. This solenoid Z001 is energized when stepping tube G5 becomes conductive, due to the effect of the fifth sequence mark 49. With commutator contact s13 thus connected to the negative bus B, and with commutator contacts s1s2 connected to the positive bus A, capacitor C40 charges in circuit s12C40s3-B12 s42. The disks 145 and 146 continue their relative movement until contact s3 of adding disk 145 again leaves contact s13 of reference disk 146, so that capacitor C40 is now disconnected from the negative bus B, retaining its charge. With the two commutator disks further continuing their relative movement, contact s1 and with it contact s2 comes into contact with s11 of the reference disk just before the 0 position of adding disk 145 reaches home position of disk 146. This position is shown in Fig. 15a.

Keeping in mind that condenser C40 is still charged, the fifth sequence slit 49 now renders the fifth sequence tube G5 conductive, energizing relay Zocl of the carryover unit, causing Z0c2 by discharge of condenser CO to close contacts s71, and putting capacitor C40 on the negative bus B. The capacitor C40 thus adds its charge to that of bus B, putting it on the grid of the non-conducting side of the next higher counting tube Vt which thereupon connects the corresponding clutch magnet. Slit 49, through tube G5, also energizes magnet Z1 from A32 which shifts the stepping solenoid Z1, and shaft 105, one step. This is effected by slit 49 triggering G5 which closes relay Zocl direct and Zoc2 by discharge of condenser CO thus putting a negative potential on C40, causing this to discharge and applying a more positive potential to the grid of Vt. Both clutch magnet and shifting solenoid being thus actuated, the carry-over operation of adding one step to the next higher adding element is performed. It should be noted that the carryover operation is prevented during the totalling and printing operations," to be described hereinbelow, due to the fact that during these last mentioned operations, contacts s42 and s32 of relay Zop are open.

The above description of the carry-over operation refers particularly to the situation when during an adding operation the adding element 145 progresses from its point nine to its point zero, on the floating home point of the reference element. The corresponding situation occurs in the case of subtraction when the home point of the subtracting disk leaves the zero point of the adding element 145 and progresses to the nine point of its adding element. So far as the relative positions of the commutator contacts s1 s2, s11 and s12 and s3 and s13 are concerned, there is no difference between the two situations, the commutator being in effect symmetrical, no

matter whether s2 and s3..approach s13 and s12 fromone side or the other. It will be remembered that Fig. 15 shows the superimposed commutator switch disks in the adding position when point nine of the adding disk coincides with the home point of the subtracting disk, further rotation of the adding disk bringing both zero points into coincidence. On the other hand, assuming that the relative position of the disks is as shown in Fig. 15a, the adding disk is stationary and the subtracting disk advances (in the same direction), the relative position of the commutator contacts at the start is that indicated in Fig. 15a, which shows that, as above stated, the relation of the contacts is for present purposes the same.

However, since switch s52 of solenoid Zos has during the subtracting operation, connected the subtracting clutch magnets instead of the adding clutch magnets, the subtracting disk instead of the adding disk of the next higher order is turned one step, reducing that order by one digit, which is the proper operation in the situation.

It will be noted that, as explained above under accumulating (subtracting) the subtracting switching tube Vs remains conductive for energizing its clutch magnet throughout the time period when the effect of the fifth sequence slit prevails, because the subtracting operational mark extends over all five sequence slits triggering the tube at the passing of each sequence slit, so that any carrying-over operation in a subtracting step is properly finished without premature interruption.

One operational mark which initiates and controls subtracting is of suflicient length to provide for successive carryover operations, as in the instance where a first carry-over step will cause simultaneously minus carryover operations in a multiple of orders positioned to the left of that first order, as in the case of the so-called fugitive one which will now be discussed somewhat more in detail.

The fugitive one has to be taken into account when a number added is greater than the capacity of the machine.

For example, if the total in the accumulator is 150 and the number 224 is to be subtracted therefrom, the customary computing machine will indicate as a result the so-called complement number, from which complement number the correct result has to be derived by subtracting it from 9999. In the present instance the complement number would be 9925 which, subtracted from 9999 gives the correct result, minus 74. In other words, if the conventional method of subtracting by reversal is used, and continued for a negative total after the 0000 position is reached, there will always be alternative total designations, namely 0000 if the operation is in addition, and 9999 if the operation is a subtraction.

Another example of the fugitive one occurs in the instance where the addition of a number to a previously recorded number results in a sum which is higher than the remaining capacity of the apparatus. For example, if the number 224 is to be added to the complement 9925, noted above, the hundreds wheel in turning through nine to zero will be reason of its carry-over cause the order next to the left to turn through nine to zero and so on through the nth order. The nth order in turn will cause a carry-over operation of the unit order so that the result will read 159, which is correct.

This fugitive one is dealt with in accordance with my present invention in the following manner.

The sequence tube G5 does not directly energize solenoid Z1 which performs the single step shift of shaft 105, for the carry-over operation, but this solenoid Z1 is instead, for the said carry-over operation, energized by relay Zc2 under the control of condenser Co which also triggers the corresponding clutching tube V. When now, for example during the above mentioned adding of 224," the hundred disk 146 turns two number values, from 9 to one, its carry-over commutator switch charges condenser C40; at this instant no carry-over count takes place on the thousand wheel. When the fifth sequence tube G becomes conductive, Zocl is energized and the closing of contacts Sol causes condenser C0 to discharge, energizing solenoid ZocZ and closing switch s71. Solenoid Zoc2 being new under the control of condenser C0 it can remain energized only during the period while Co has a charge. This period is regulated for example by selecting the condenser rating and coincides with the length of time solenoid 21, after being energized by the firing of G1, takes to move drive shaft through 36 degrees. The circuit from tube G5 is delayed long enough by resistor RZ before energizing solenoid Z1 so that the functioning of solenoid Z1 will coincide with the discharging of the C40 condensers associated with the affected orders of the V tubes. It will be explained below how the carry-over and carryon-carry operations of all affected orders are performed simultaneously. Solenoid Zocl remains energized as long as tube G5 is on, whereas conductivity of tube G5 has no effect upon solenoid Z002.

The opening of contacts s71 at the actual ending of the Z1 stroke prevents the firing of the V tubes during the period while tuge G5 remains on and thus prevents any C40 condensers that might have been charged during a carry-on-carry operation from discharging during the same on period of G5 in which the carry-on-carry was originated, thus causing the coincident V tube to be fired at a period when drive solenoid Z1 is not moving shaft 105.

When another carry-over, termed a carry-on-carry" is originated by the recording operation of a carry-over, this carry-on-carry cannot immediately be recorded because contacts s71 have been opened by Z0c2 as noted above. The initiating impulse to record this carry-oncarry is stored in condenser C40 until the moment tube G5 is fired in the next cycle of tubes G1 to G5. The carryaon-carry" operation of all affected orders taking place simultaneously eliminates the necessity for repeat firings of G5 and therefore a repeated energizing of Z1 during a single cycle of tubes G1 to G5.

The simultaneous carryover and carry-on-carry" operations performed simultaneously with one and the same machine operation, as follows.

Coincident with each order position are two switching contact arrangements, one for the normal carry-over" operation and the other for the carry-on-carry opera tion. Whenever a condition has been set up in an order or orders that calls for a carry-on-carry operation, that is a condition that shows figures such as 39992, then, when the next carry-over operation in the digit order next to the right of the nines orders takes place, both the carry-over and the carry-on-carry operation of all affected orders will be performed simultaneously regardless of the number of carry-on-carry orders in sequence to the left of the carry-over orders that show the digit 9 in sequence.

For that purpose, a contact s5 is located on the face of commutator switch disk (Figs. 13 to 15a) at its 9 count at the home or zero position on disk 146 the contacts s5 and s15 are in contact one with the other. These two contacts will remain in circuit making relation so long as they are in this relative position.

The contact s5 on disk 145 is electrically connected from the condenser C40 through switch s1 on one order, to the contact s11 on the next order to the left which leads to point 321 on the grid circuit of the second order to the left. If in this sequence the second order to the left also shows a 9 count" then this same circuit will establish a through connection to the third order to the left.

As an example let it be assumed that the orders read 39992 and that the unit digit 8 is to be added to this sum of 39992. The operation of adding the digit "8 will cause the following simultaneous operations.

The digit 8 will first initiate a normal carry-over operation thus causing the condenser C40 to discharge after the counting disk in the unit order has come to rest 21 upon passing the count position. It should be noted that if the problem had been to add the digit 9" instead of "8," then the condenser would have delayed its discharge until the counting disk has come to rest upon passing the 1 count" position.

The discharge of condenser C40 will pass current to the point 321 associated with the grid of the tens order V tube by means of line s11 thus causing this tens order tube V to read 0." Now because the "9" contacts on the next two orders to the left of the tens o'rder also read 9, the operation coincident with the tens order will take place simultaneously for these two orders. This means there is a simultaneous operation of adding "1 in the tens, hundreds, thousands, and ten thousands orders, thus changing the reading at the start of the problem from 39992 to 40000" by a single mechanical carry-over operation, of course, in addition to the simple recording of the digit "8 in the unit order.

If the problem had been one of subtraction instead of addition, then the carry-over" function would have been performed by a single mechanical operation of deduction instead of addition, in all orders simultaneously. This same operation takes place when a carry-over is called for in the nth order, in which instance it causes an increase of the digit 1 value in the unit order.

Total taking and printing-The totalling and printing operation is initiated and controlled by the totalling operational mark 45 of Fig. 2, which coincides with code marks filling the entire code mark area, as indicated at 48 of Fig. 2. This operational mark operates the phototube combination Pdp and Pip, rendering ano'de a22 of tube Vp conductive and energizing relay magnet Zop, which disconnects the carry-over solenoid magnet Zoc and contact segment s13 of the carry-over commutators, connecting instead, through auxiliary bus A31 the printing solenoids Zpl, Zp2, and Zp4; relay Zop also connects, at s41, the extension bus B11 to main bus B. The subtracting operational mark having become ineffective, switching tube Vs has likewise become ineffective, magnet 20.9 is de-energized and its contacts back in the position shown in Fig. 20. Due to this provision, the printer solenoids are energized only when called for by the totalling operational marks.

'It will be noted that the above energization of magnet Zop also puts the hammer magnet Zh into the anode circuit of sequence tube G5, thus making it ready for operation. The remaining contacts s41 of magnet Zop put the printing clutch magnets Zp on bus B, so that the printer is now fully ready for operation.

It will be noted that disconnection of contact s42 of magnet Zop blocks any carry-over operation during totalling, and also blocks any circuits through printer commutator contacts s26, of commutators Np. This condition is maintained so long as the totalling switching tube Vp conducts through anode a22, so that interference with the operation of resetting the printer and accumulating wheels is prevented.

As pointed out above, magnet Zop has now disconnected solenoid Zoo, in the circuit of sequence tube G5.

' It will be remembered that the printing unit above described with reference to Figs. 16 to 19, has type wheels whose numbers run in a sense opposite to that of the. corresponding numbers of the accumulator or adding wheels, for the purpose which will presently become apparent.

Since for totalling all code marks are black on the record strip, and since the printer solenoids and clutch magnets are now in circuit, together with the accumulator clutch magnets and solenoids, all these are simultaneously energized during the totalling operations.

With the type wheel in position normal, the printer commutator contacts s25, and s26 of disk 233 are disconnected, resting in the space between the contacts s22 and s23 of the co'ntacts of the opposite commutator disk Z33 (Figs;- 16 to 19). i

It will now be evident that, with all code marks effective, and with the operational mark printing retaining the printing tube Vt conductive to energize magnet Zop, all adding disks and all printer wheels will advance due to the action of the solenoids Z1, Z2, Z4 and Zpl, Zp2, Zp4 rotating shafts and 205 (Figs. 7, 8 and 16). This movement will continue until any one adding disk reaches its home position on the subtracting or reference disk. Then, commutator contacts s4 and s14 of accumulator disks and 146 of the respective set, make connection. Commutator segment s4 leads to the minus bus B through contact s41 of solenoid Zap which is now closed, and commutator segment s14 connects to the grid -g22 of tube Vu, or that V tube whose accumulator disk has reached the above mentioned home position. This immediately puts the grid on the negative bus B and renders the respective anode circuit non-conductive, cutting out the respective clutch magnets 20, Zp. Therefore, the respective printer wheel and accumulator disk will immediately stop. It should be noted that this stoppage is independent of the movement of the shifting solenoids Z1, Z2, Z4, Zpl, Zp2, Zp4, which may continue to move if the solenoid in question happens to be one which defines more than one unit, such as solenoid Z4 or Zp4 which, after having brought its accumulator or printer into zero position, might continue to advance through an angle of 108, representing three number values.

Thus, assuming that any accumulator wheel was, when the totalling operation began, with its point n at the homing position, it will have turned 10-n steps before it comes to the above described stop, whereas the printer wheel, being inversely numbered, but turning in the same direction, will stop at n, that is the correct position for totalling and recording the total. All wheels whose zero adding wheel point has reached the home point of the subtracting or reference wheel, are locked by the circuit through the segments s4s14 of their commutators.

While, as mentioned above, the sequence solenoids complete their action regardless of the energization or de-energization of the respective clutch magnets at the reaching of the home position of any one zero point of the adding wheel, the fifth sequence slit, operating tube G5, energizes the hammer solenoid Zh, energizing the hammer 211 (Fig. 16) and printing the ledger card inserted in slit 212 of the cabinet as shown in Fig. 1.

The above operation leaves the accumulator disks with all zero points at home position, but not the type wheels. Type wheel re-setting.The operation of resetting the type wheels is performed without the aid of operational or other marks, by means of the printer commutators shown in Figs. 17 to 19. The space between two sets of five sequence slits provides sufficient time for this operation.

Referring now particularly to Figs. 16 to 20, it will be noted that normally, that is before the totalling operation begins, the commutators contacts :25 and s23 of the commutator disks, as shown in Fig. 19. As mentioned above, with the total switching tube Vp conducting through anode a22, contacts s32 and s42 of solenoid Zop are disconnected, and with them the commutator contacts s25 and s26, the latter being kept open so long as tube Vp is conductive as above described, that is, until an accumulator wheel zero point reaches ho'me position. When, after the totalling operation is performed, the tube Vp becomes ineffective because conductive through the non-operative anode a21 and lap de-energized, connection is made through s22, s23, s26, s25, which are now in contact since the printer wheels are not in the zero position, which alone corresponds to contacts s25 and $26 being positioned in the gap between the two other printer commutator contacts. This circuit is completed through contacts of relay lap and through the appropriate connections between the commutator segments s25 and :26, as shown at Np in Fig. 20. It will be further observed that the above position of the printer contacts locks out the printer commutator during the totalling operation.

Now, with 525 on s23, s26 on s22, and s32 and s42 of Zqp closed, the printer solenoids and clutch magnets are energized directly through the printer commutator, in circuit A-.s32. -A32s25s23A31R31Zpu-s22 -.-s26.B12-s4Z-B, independently of tubes G and V, advancing the printer wheels step by step until each wheel reaches the zero position, when segments :25 and s26 become disconnected from s23 and s22 respectively, and the respective clutch magnets are disconnected. Each printer wheel therefore stops at its proper normal position, namely with the zero type on the printing hammer.

The resistor R31 is inserted in this circuit in order to make it electrically equivalent to the corresponding tube control circuits which are otherwise utilized for energizing the respective clutching and shaft turning solenoids.

The accumulator disks are now with their zero points at the home point of each respective subtracting or reference wheel, the printer wheels are at zero position relatively to the recording device, the required adding, sub tracting, carrying over and totalling operations have been performed and recorded, and the apparatus is ready for further operation.

It should be understood that the present disclosure is for the purpose of illustration only and that this invention includes all modifications and equivalents which fall within the scope of the appended claims.

I claim:

1. Calculating apparatus comprising supporting means, a scale member defining digits of a notation, a home member defining a reference position, means for moving said members on said supporting means unidirectionally in predetermined differential amounts corresponding to the digits of the notation only, and means for defining the relative position of said scale member and said reference position of said home member, advancement of one of said members in said direction performing an adding operation defined by the relative position of the two members as defined by said series of digits, and advancement of said other member in said direction performing a subtracting operation likewise defined by the relative position of the two members.

2. Calculatng apparatus comprising mounting means, a member defining digits of a notation, a home member having zero indicating means, means for moving said members on said mounting means relatively to each other unidirectionally in predetermined differential amounts corresponding to the digits of the notation only, and means for defining the relative position of said digit defining member to said zero indicating means, advancement of one of said members in said direction performing an adding operation, and advancement of other member in said direction performing a subtracting operation, both operations being defined by the position of the first member relatively to said zero defining means.

3. Calculating apparatus comprising mounting means, a first substantially circular scale member angularly subdivided to define in equal steps digits of a notation, means for rotating said scale member relatively to said mounting means unidirectionally in predetermined differential amounts corresponding to the digits of the notation, a second substantially circular member arranged for rotation concentric with said first member and having zero defining means, means for rotating said second member on said mounting means in similar predetermined differential amounts and in said direction, and means for defining the relative position of said scale member and said zero defining means, rotation of said scale member in said direction performing an adding operation defined by the position of the first member relatively to said zero defining means, and rotation of said second member in said direction performing a subtracting operation likewise defined by the position of the first member relatively to said zero defining means.

4. Calculating apparatus comprising mounting means, a first substantially circular scale member angularly subdivided to define in equal steps digits of a notation, means for rotating said scale member relatively to said mounting means unidirectionally in predetermined differential amounts corresponding to the digits of the no tation, a second substantially circular member defining a reference position arranged for rotation concentric with said first member, means for rotating said second member on said mounting means in similar predetermined difierential amounts and in said direction, and means for indicating the relative position of said scale member and said reference position, rotation of said scale member in said direction performing an adding operation defined by the position of the first member relatively to the second member, and rotation of the second member in said direction performing a subtracting operation likewise defined by the relative position of said members.

5. Calculating apparatus comprising mounting means, a first scale member defining in equal steps digits of a notation, a second scale member having zero indicating means, means for moving said members on said mounting means relatively to each other both unidirectionally in predetermined differential amounts corresponding to the digits of the notation, said unidirectional movement of said first member performing an adding operation defined by the position of the first member relatively to said zero defining means, and said unidirectional movement of said second member performing a subtracting operation likewise defined by the position of the first scale member relatively to said zero defining means, and means for defining the position of said first scale member relatively to said zero defining means independently of the posi tion of said scale members relatively to said mounting means.

6. Calculating apparatus comprising mounting means, a first substantially circular scale member defining in equal steps digits of a notation, a second scale member having zero defining means, means for concentrically moving said members on said mounting means relatively to each other unidirectionally in predetermined differential amounts corresponding to the digits of the notation, said unidirectional movement of said first member performing an adding operation defined by the position of the first member relatively to said zero defining means, and said unidirectional movement of said second member performing a subtracting operation likewise defined by the position of the first scale member relatively to said zero defining means, and means for defining the position of said first scale, member relatively to said zero defining means independently of the position of said members relatively to said mounting means.

7. Calculating apparatus comprising mounting means, a first number defining scale member, a second member having zero defining means including a relay circuit having two contact means, one moving with each of said members and adapted to contact the other, and means for moving said members on said mounting means relatively to each other both in similar steps corresponding to said numbers and in the same direction, advancement of said first member in said direction and movement of its contact means in a corresponding relation performing an adding operation defined by the position of the first contact means relatively to the second contact means, and advancement of said second member and of its contact means in said directions performing a subtracting operation likewise defined by the position of the first contact means relatively to the second contact means, independently of the position of said members relatively to said mounting means.

8. A computing machine of the type wherein radiation sensitive detecting means are controlled by records carrying radiation modifying, positionally co-rclated counting, operational and sequence indicia, comprising a first counting member, a second counting member, actuating means adapted to move stepwise under the control of said sequence indicia, means for coupling said.

th sea ing ammo 9 sai ac uat ng me -unde s tc rq a d du in steps dete min d by 1 s nt n indicia conjointly with said sequence indicia, means :for coupling said second counting member to said actuating means ,under control of and during steps determined by said counting indicia tconjointly with said .sequence in- ,dicia, means controlled by said operational indicia for selecting one or theother of said counting members for coupling to said actuating means, and registering means onieach of said counting members directly juxtaposed jgr defining the relative position .of said two counting membe A computing machine of the type wherein radiation sensitive detecting means are controlled by records carrying radiation modifying and positionally co-related counting, operational and "sequence indicia, comprising a first counting member, a second counting member, actuating means adapted to advance stepwise in a given direction under the control of said sequence indicia, means adapted to connect said first counting member to said actuating means for advance in a given sense under control of and during steps determined by said counting indicia conjointly with said sequence indicia, means adapted to connect .saidsecond counting member to said actuating means for advance in a given sense under control of and during steps determined by said counting indicia cnjointly with said sequence indicia, means controlled by said operational indicia for selecting one or the other of said counting members for coupling to said actuating means, and registering means on each of said counting members directly juxtaposed for defining the position of said first counting member relal tively to a point of said second counting member.

If). A computing machine of the type wherein radiation sensitive detecting means are controlled by records carrying radiationmodifying and positionally co-related counting, operational and sequence indicia, comprising a first counting member, a second counting member, actuating means adapted to rotate stepwise in a given direction under the control of said sequence indicia, means -for coupling said first counting member to said actuating means for rotation in a given sense under control of and during steps determined by said counting indiciaconjointly with said sequence indicia, means for coupling-said second counting member to said actuating means forrotation in said sense under control of and .during steps determined 'by said counting indicia conjointly with said sequence indicia, means controlled by said operational indicia for selecting one or the other of :said counting members for coupling to said actuating -means, and registering means on each of said counting t-members directly juxtaposed for defining the position of said first counting member relatively to a point of .said second countingmember.

:11. Computing apparatus of the type which is con- -trolled by a record which carries radiation affecting and positively co-related counting and-sequence indicia, comprising means for continuously moving said record -through said apparatus, radiation sensitive means for scanning said indicia with the effective scanning period of said counting indicia including that of said sequence indicia, a counting member, actuating means adapted ;to advance said counting .member stepwise solely under control of said sequence indicia, means for coupling said counting ,member to said actuating means, and initiating -rne ans rendered effective under conjoint control of said sequence indicia and said counting indicia for operating Said coupling means.

12. Computing apparatus of the type which is controlled by light modifying and positionally co-related stunt and sequence ind ci somethin mean r' Q pc t q sly moving s i cord th hsa d ap ar tus photo-sensitive means for scanning said counting indicia, Iseparate photo sensitive means for scanning said .seilence indicia with the effective scanning period of said counting indicia including that of said sequence indicia, ,acounting member, actuating means adapted to advance stepwise under the sole control of said sequence indicia, means for coupling said counting member to said actuating means, flash .larnp means rendered etfcctive bysaid sequence indica for illuminating said counting indicia, and meansresponsive to said counting indicia when illnninated by said flash lamp for operating said coupling means.

13. A computing machine of the type wherein ,radiation sensitive detecting means are controlled by records carrying radiation modifying counting and sequence indicia, comprising a counting member, actuating means adapted to advancesolely under control of said sequence indicia, means for coupling said counting member to said actuating means, and initiating means rendered efiective under conjoint control solely of said sequence indicia and said counting indicia for actuating ,said coupling means.

14. A computing machine of the type wherein radiation sensitive detecting means are controlled by records carrying radiation modifying counting indicia arranged in several columns defining digit position and several rows defining digit value, and carrying several sequence indicia each one arranged to be effective within the efiectivity period of a corresponding one of said rows of counting indicia, comprising a plurality of counting members, one for each of said columns; actuating means adapted to advance predetermined distances in consecutive steps under the sole control of said sequence indicia; means forcoupling said counting members to said actuating means; and initiating means rendered efiective solely under conjoint control of each one of said counting indicia and the sequence indicium of its row, for actuab ing said coupling means.

15. A computing machineof the type wherein radiation sensitive detecting means are controlled by records carrying radiation modifying digit counting marks, and similar sequence marks arranged to be effective within the eifectivity period of said counting marks, comprising a counting member, actuating means adapted to advance under the sole control of said sequence marks, means for coupling said counting member to said actuating means, means for rendering said coupling means efiective solely upon conjoint radiation modification by'both marks, andmcaus for thereafter advancing said actuating means.

1 6. A computing'machine of the type wherein radiation light sensitive detecting means are controlled by records carrying light absorbing digit counting marks, and light absorbing sequence marks arranged to be effective within the etfectivity period of said counting marks, comprisinga counting disk, shaft means adapted to advance under the sole control of said sequence marks, clutch means for coupling said counting disk to said rshaft means, flash lamp means rendered effective by said sequence marks and arranged for illuminating said counting marks, and means for actuating said clutch means solelyunder control of said counting marks when illurninated by said flash lamp means.

17. In a computing machine wherein counting members are moved in steps controlled by electrical code impulses, a stepping device comprising a plurality of countingrnembers, means common to all of said counting membe'rs and arranged for the independent advancement of auycountingrnember that is coupled to said advancing means, connecting means for selectively coupling said counting members to said advancing means, a plurality of motor means for actuating said advancing means each havinga moving member and means for positively determining the travel of said moving member to a .value selected by said impulses, and means linking said moving members to said advancing means for moving said advancing means ,in a given direction an amount proper tiona te to said code impulse determined travel during

Images