Although they were extremely cunning, the members of the first wave of mechanical calculators (as discussed elsewhere on this site) were less reliable than one might hope (largely due to limitations in manufacturing technologies of the time) and they often behaved somewhat erratically. Thus, the golden age of mechanical calculators didn’t really start until the 1800s …

View Topics

Thomas of Colmar and the Arithmometer
In fact, it was not until 1820 that the Frenchman Charles Xavier Thomas de Colmar (1785-1870), also known as Thomas of Colmar, invented a device he called his Arithmometer, which later became known as a Thomas Machine.

Based on the same general “stepped drum” techniques as Leibniz's calculator, the Arithmometer could be used to perform addition, subtraction, multiplication, and – with a little help from the user – division. But the Arithmometer’s main claim to fame was that it actually worked in a reliable fashion; that is, it gave the same answer to the same problem over and over again. This so impressed everyone that, even though the Arithmometer occupied most of one’s desktop, it remained in production for over 100 years.

Willgodt Odhner and the Original Odhner
The big problem with stepped-drum-based machines such as Thomas de Colmar’s Arithmometer was that the drums themselves were large and heavy and they forced both the input levers and display digits to be spread out by an uncomfortable distance for the operator.

In order to address this, in 1875 the American Frank Stephen Baldwin (1838-1925) invented a new form of mechanical calculator founded on a “pin wheel” or “variable cog” mechanism. (Just to confuse matters, Baldwin also decided to call his machine an Arithmometer which – at that time – was generally understood to mean “hand-cranked calculating machine.”)

A few years later in 1878, working independently and with no knowledge of Baldwin’s work, the Swedish Inventor Willgodt Theophil Odhner (1845-1905) created his version of a pin-wheel Arithmometer. This was very similar to Baldwin’s machine, but Odhner’s was the more successful design, and versions of this device – which became known as the Original Odhner – continued to be produced until the early 1970s.

Original Odhner

Dorr Eugene Felt and the Comptometer
In 1885, at the age of 23, the American Dorr Eugene Felt (1862-1930) created a prototype of what was to become known as the Comptometer. This was the first mechanical calculator in which numbers were entered by pressing keys as opposed to being “dialed in” using wheels or similar techniques.


The primary function performed by the Comptometer was addition. Due to its keyed-in form of data entry, the Comptometer could be used to add large lists of numbers very quickly. This was of particular interest for accountants, and these machines quickly became extremely popular. A few years later, in 1889, Felt added a printing mechanism to his Comptometer (he called this new device a Comptograph), thereby claiming the honor of inventing the first printing calculator. (The original Comptograph along with the prototype of the Comptometer are to be found at the Smithsonian Institution in Washington DC.)

The Evolution of the Modern Calculator Keypad
One obvious characteristic of the Comptometer shown in the previous topic is that it featured a large array of keys; from right to left the columns represented 1s, 10s, 100s, 1,000s, and so forth, where each column typically had nine rows of keys numbered from 1 to 9 (a 0 could be represented by simply not pressing any keys in that column).

In 1901, the American William W. Hopkins invented an adding machine called the Standard that used only a single row of ten keys. This was cheaper to manufacture and easier to use than the Comptometer and it soon became the cornerstone of the Standard Adding Machine Company. Other machines quickly followed from various vendors boasting a variety of layouts, such as two rows of five keys.

Modern Calculator Keypad In 1911, David Sundstrand (1880-1930), who was born in Sweden but was brought to America when he was only one year old, created the first working model of a ten-key adding machine with the keys arranged in three rows plus a zero key – a formation that is used to this day.

The first commercial model of this adding machine was presented to the market in 1914. This device was very successful and formed the basis for the inception of the Sundstrand Adding Machine Company, which was run by David’s brother, Oscar Sundstrand (1889-1972).

The golden age of mechanical calculators persisted until the late 1960s and very early 1970s. In terms of mechanical sophistication, the later machines typically take the prize; but in terms of aesthetics and the interesting use of “twiddly bits” to make things look pretty, machines from around the 1880s to the 1920s are hard to beat.

Note: The material presented here was abstracted and condensed from The History of Calculators, Computers, and Other Stuff document provided on the CD-ROM accompanying our book How Computers Do Math (ISBN: 0471732788).