Joseph Edmondson

In 1883 Joseph Edmondson from Halifax, Yorks, England, received three Great Britain patents (numbers GB188316, GB18830016, and GB188300016) for an interesting circular calculating machine, based on the stepped-drum mechanism of Leibniz. The machine was manufactured for some time by the company Joseph, Blakey, Emmott & Company, Ltd., of Square Road, Halifax, a producer of electrical equipment and telephones, but it is unknown how many devices were made. Later the business moved to Norwich.

The machine was initially introduced in a lecture to the Physical Society of London on 28 March 1885, then was presented to International Inventions Exhibition (a world’s fair held in South Kensington from May 1885). In the same 1885 the machine described in many sources, like Enciclopedia delle Arti e Industrie, volume 5, Macchine da Calcolare from Giuseppe Pastore (Torino, Italy, 1885), Scientific American magazine, Engineering magazine, and others.

The patent model of circular calculating machine of Edmondson
The patent model of circular calculating machine of Edmondson

In fact, Edmondson returned to the circular stepped-drums mechanism of the machine of Philipp Matthäus Hahn, although the construction of Edmondson is quite different. The machine was essentially a 20-digit arithmometer with a circular carriage (the slides being arranged radially around it) instead of the straight sliding carriage.

The patent drawing of the circular calculating machine of Edmondson
The patent drawing of circular calculating machine of Edmondson

The machine is a quite solid brass, steal, wood, and ceramic device, with dimensions 50,3 x 42,3 x 15,0 cm and weight some 17 kg (including the wooden case).

The calculating machine of Edmondson, Engineering magazine, 1 May 1885
The calculating machine of Edmondson, Engineering magazine, 1 May 1885

Eight number slides are placed radial on the outer fixed portion of the machine. The hinged slide of the straight type of machine (e.g. machine of Thomas de Colmar) is replaced by a circular plate, carrying 20 figure discs, each of which—depending on its position relative to the number slides and driving handle, respectively—can serve either for recording the result in multiplication, or for setting the figures of the multiplier. Stepping is performed by lifting this plate, rotating it through one-twentieth of a revolution, and lowering it again. To multiply, the figures of the multiplier are set on the middle plate and those of the multiplicand on the slides. The driving handle is then turned to the right and the plate stepped clockwise until all the digits of the multiplier have been brought to zero. The product then appears in the apertures on the middle plate.

The machine of Edmondson, upper view drawing (source Macchine da Calcolare, Giuseppe Pastore)
The calculating machine of Edmondson, upper view drawing (source Macchine da Calcolare, Giuseppe Pastore, Torino, 1885)

The machine is provided with a zero-setting mechanism with which some, or all of the windows may be set to zero.

The machine was shipped with a Quick Start Guide (Nota Bene) on the lid:
NOTA BENE.
By sliding the Lid of the box to the right, the hinges will separate, and the lid may be laid aside.
A Pamphlet containing full instructions accompanies each Machine, and should be carefully studied; but the following points are of special importance:-
The Driving Handle, which should be held between the thumb and two fore-fingers, must always be turned to the right, thus:-
Turning to the left may strain or break the machine.
The Number Slides, the Index Slides, and the Regulator Handle should only be moved when the Driving Handle is in its lowest position, i.e., opposite the stud.
The Regulator Handle should always be pushed home to the extreme right or the extreme left.
The Number Slides and Index Slides have Springs underneath, which drop into notches when the slides are in the right positions. The click of the Spring should be observed when setting a Slide.
If the Machine meets with a stop in its working, it should not be forced. The cause will probably be found in the neglect of one or more of the above directions.

Edmondson’s calculating machine attracted enough notice to be favorably mentioned in the 1911 Encyclopedia Britannica, almost thirty years after its introduction.

The calculating machine of Edmondson, (source Фон-Бооль, Приборы и машины для..., Москва, 1896)
The calculating machine of Edmondson, (source Фон-Бооль, Приборы и машины для…, Москва, 1896)

The description of calculating machine of Edmondson in Scientific American magazine


IMPROVED CALCULATING MACHINE.
An ingenious calculating machine is to be seen in the Inventions Exhibition, London, at the stand of its inventor, Mr. Joseph Edmondson, of Halifax, Yorks. Fig. 1 is a perspective view and Fig. 2 a plan of the apparatus as presented to the operator when seated in a position for working it. The principle of this machine is very simple. Beneath each number slide are two radial axes, one above the other. The lower one makes one revolution for each revolution of the motive handle, and has upon it a cylinder. That half of the cylinder which is farthest from the center of the machine is devoted to reckoning, and the other half to stopping and locking the axis above it, until the moment when the reckoning half, or reckoner, begins to operate. The reckoner is again divided into ten sections, on which there are respectively 0, 1, 2, 3, 4, 5, 6, 7, 8, and 9 teeth, and a considerable blank space; the section with nine teeth being the farthest from the center of the machine. The eight teeth of the next section are a prolongation of an equal number of the teeth of the preceding; and so on for each section; so that to the eye the reckoner presents a series of successively decreasing teeth, and from its stepped appearance it may be termed a stepped reckoner. The stopping or locking half of the cylinder is also stepped to correspond with the stepped reckoner. On the upper axis is a tube, free to slide longitudinally, and having a key fitting into a groove in the axis. The tube carries at its outer end a pinion of ten teeth gearing with the stepped reckoner, and at its other end a star wheel of ten rays fitting the stepped stop. A fork projecting below the number slide fits into a groove round the tube, so that the motion of the number slide is communicated to the tube, whose pinion is thus placed in position to gear with that section of the reckoner the teeth of which correspond in number with the figure in front of the curved rule. The pinion and star wheel are at such a distance apart that the latter is always upon the section of the stepped stop corresponding to the section of the reckoner with which the former is placed to gear.

On the lower axis, and nearer to the center of the machine, is a piece movable longitudinally, but carried round with the axis by a pin fitting into a hole at the end of the stepped stop. It is composed of the secondary carrying tooth and its corresponding stop, there being an incline on the inner edge of the latter, the use of which will be presently explained. Just above this movable piece there are, fixed on the upper axis, a second pinion of 10 teeth and a star wheel. The pinion is so set that when the piece below it is close up to the stepped stop, the secondary carrying tooth passes it by: but when the piece is moved inward, this carrying tooth gears with the pinion, and, when revolving, moves it one tooth forward. As soon as this has taken place, the incline comes in contact with a pin in the frame of the machine, which pushes the piece into its former position. Under the disks on the circle, the before mentioned upper axis carries two reversed bevel wheels of 10 teeth each, on a tube free to move longitudinally, but carried round by a key fitted into a groove on the axis. These wheels are moved longitudinally by the regulator, as may be seen by taking off the circle. Between them (and in gear with one or other of them, according to the position of the regulator) is a similar bevel wheel on the spindle of the corresponding number disk, and above the wheel is the primary carrying tooth. As the pinions and the bevel wheels have each ten teeth, and the number disk has ten figures, every tooth which the pinions are moved counts one, either forward or backward on the disk. When the figure in the aperture on the disk passes from 9 to 0 in addition, or from 0 to 9 in subtraction, the primary carrying tooth passes the wedge-shaped end of the upper arm of the carrying lever, which it pushes back. This carrying lever moves on a perpendicular axis. Its lower arm clutches a pin in the shaft of a fork under and parallel with the lower axis beneath the next higher number slide. This fork fits into a groove around the movable piece of the secondary carrying tooth, which it shoots inward into position for adding or carrying 1 as above described.

Each lower axis is timed to operate on the pinions above it, at least one tooth later than its neighbor to the right, to allow time for the latter to shoot the carrying tooth. This is not the case, however, with the axes under the lowest number slide but one, and the index slides, B and C, which are all timed to act simultaneously with the lowest number slide. It will be seen that each revolution of the motive-handle, and consequently of the reckoners, causes the latter to move the pinions above them as many teeth as there are on the sections of the reckoners, over which the pinions are respectively set by the number slides. This motion of the pinions is communicated to the number disks, and therefore adds or subtracts accordingly.

This machine, we understand, will work the four fundamental rules of arithmetic with absolute correctness. In the multiplication and division of large numbers, as well as in the combined operations of multiplying and adding, or of multiplying and subtracting, it effects a great saving of time and brain power. Any number under a hundred millions can be multiplied by any number under a million millions. The result when the highest numbers within these limits are employed is a product of twenty figures, the time taken (after the figures are set on to the machine) being two and a quarter seconds, on the average, for each figure of the multiplier. It therefore forms an invaluable adjunct to offices and individuals having extensive calculations to deal with.