Working out the premiums - calculating machines
09 Oct 2016
In 1871 a computer wasn't a device like the one you are using to read this blog, it was the term used for a person who calculates or computes. The insurance companies of 1871 employed many of these 'people computers' to help them work out the risks for the different types of insurance and the premiums their companies should charge to provide appropriate cover. The super computers of the day were the actuaries who were, according to Walford's Insurance Cyclopaedia, 'trained to apply the doctrine of mathematical probability to the affairs of life'. This was far easier said than done as these pages of calculations for endowment assurance from a text book of 1843 demonstrate.
Treatise on Probability by David Jones (1843)
In 1871, two actuaries, writing in a learned publication, provided opposing views on the potential use of calculating machines to help them with their complicated calculations. Mr Edward Sang (Fellow of the Faculty of Actuaries in Scotland) felt the machines had little benefit and concluded:
'Thus, on the whole arithmeticians have not much to expect from the aid of calculating machines. A few tables, otherwise very easily made, comprise the whole extent of our expected benefits; and we must fall back upon the wholesome truth that we cannot delegate our intellectual functions, and say to a machine, to a formula, to a rule, or to a dogma, I am too lazy to think; do, please, think for me.'
On the other hand, Mr W J Hancock (who worked for an Aviva company) recognised the positive benefits such machines could bring to his profession, he wrote:
I think it therefore manifest that the machine is far superior to the long multiplication by hand... and that it saves mental labour, time and risk of error. - Mr W J Hancock
This reference to Mr Hancock and his opinion on Thomas de Colmar’s arithmometer is the earliest evidence we have of anyone in an Aviva company using a calculating device rather than working out premiums by written calculation or using tables of logarithms. Mr Hancock, who was the actuary for the Patriotic Assurance Company (later part of Sun Life), based his assessment of the arithmometer on the use he made of it in his work and qualified his conclusions as follows:
'Of course the value of the machine diminishes with the [diminishing] number of figures required to be dealt with…if a man has not half a mile to travel, there is not much difference between walking and going in a railway train; but when the distance becomes one or two hundred miles, the advantage of going by train instead of walking becomes evident.'
Colmar's arithmometers, which were manufactured from 1851, were the first type of mechanical calculator to be produced commercially. We know that other Aviva companies, such as Briton Life Assurance and the National Assurance Company of Dublin, were also early adopters of Colmar's machines. To use the arithmometer the operator set up numbers on the keyboard and the calculation was performed by turning a lever which worked a Leibniz wheel mechanism. The Leibniz wheel in effect mechanised the action of an abacus with toothed wheels replacing the beads on wires to represent the units, tens, hundreds, etc. needed to perform calculations. Here is a link to a film which shows how an arithmometer works.
We have an example of a Colmar arithmometer in the archive collection...
Colmar arithmometer, belonging to Commercial Union (late 19th c.)
... and a list of 'amusing' events at the Cornhill offices of Union Assurance which includes a reference to the life clerk Edward Muzio (a large man as you can see in the cartoon below) falling and breaking the arithmometer which was used by that company in 1879.
Edward Muzio of Union Assurance (1884)
The archive collection also includes an example of Tate's arithmometer, an 'improvement' on Colmar's machine, which was patented in the UK by Samuel Tate in 1881 and sold by C & E Layton, a company which had originally sold calculating tables and logarithms.
A member of staff at North British and Mercantile wrote a testimonial, which appeared in an advertisement for Tate's arithmometer in 1888, saying that the company had purchased one of the machines three years earlier and that it 'continues to give satisfaction' – he added that they found it 'a great improvement upon Thomas's [Colmar's] machine'.
Tate's improved arithmometer, Norwich Union (c1902)
The image above shows the Tate's Improved Arithmometer which was used by Norwich Union in around 1902. An advertisement of the period describes it as:
a perfected calculating machine of established reputation… a great labor and time saver, while the relief it affords from wearing brain work is incalculable. - Tate's Improved Arithmometer
To get an idea of the 'wearing brain work' a clerk at Norwich Union was expected to undertake you only have to look at these questions from an exam given to prospective Norwich Union clerks in around 1900.
Norwich Union, Clerkship examination paper (c1900)
The 1902 advertisement for Tate's arithmometer lists companies which were then using the 'perfected calculating machine' including several which later became part of Aviva – in addition to Norwich Union and North British and Mercantile the list includes Commercial Union and Sun Life. According to the advertisement, the cost of a new arithmometer in 1902 was £40 - to put this into context; in 1902 £40 was the annual salary of a 20 year old clerk with 3 years experience working in the London office of one of our companies.
Norwich Union had been using Tate's machines since at least 1890 and the finance committee minutes for February that year refer to £3 paid to C & E Layton for repairs to the arithmometer. The arithmometer is also given somewhat precarious pride of place in this drawing which is an extract from a cartoon showing the Norwich Union Life staff leaving Bignold House in Norwich in 1904 to move across the street into their new head office in Surrey House.
Norwich Union, Cartoon of Life staff moving offices (1904)
The Layton company continued to sell arithmometers up to the 1920s; the company purchased the patents from Mr Tate and began to sell machines under the Layton brand name, such as this one from around 1910 which was used at General Accident.
Layton's improved arithmometer, General Accident (c1910)
In 1920, the newly established English Insurance Company purchased an arithmometer from Laytons for £78 15s. By then the technology had really been superseded and only a year later the cash book records the same company paying £55 2s 6d for a comptometer. The new machine was probably rather like this Felt and Tarrant comptometer adding machine which was then in use at Norwich Union.
Felt & Tarrant Comptometer, Norwich Union (c1920)
You can see how well-used it was by the fact that some of the numbers have worn off and been written back in (not entirely successfully).
The comptometers were the first commercially successful key-driven mechanical calculators and were developed in America by Dorr E Felt who used a macaronic box, skewers and elastic bands to create his prototype in 1885. The term 'key-driven' refers to a mechanism which is operated directly by pressing down on the keys and was much faster than the earlier arithmometer machines where numbers were set up on the keyboard and the calculation was performed by turning a lever. You can find out more about the comptometer here.
Can you spot the comptometer box in this photograph of the Friends' Provident offices at 42 Kingsway in 1919?
Friends' Provident, Head office interior (1919)
Here is a close up.
Despite the availability of the faster, key-driven comptometers, many of our companies continued to use calculating machines which relied on the old Leibniz wheel mechanism which was the basis of the arithmometer. This TIM or 'Time is Money' calculator was produced in 1909 but was used for calculations at the head office of Sun Life for many years: indeed in the 1950s the company still had eleven TIM calculators in operation.
TIM calculator, Sun Life (c1909)
Another variation on the original Arithmometer mechanism was the pinwheel calculator developed in the 1870s in Russia by W T Odhner and in America by Frank S. Baldwin. Odhner's design was the first to be manufactured on a commercial scale, from 1890, and pinwheel calculators were used by our companies well into the 20th century. You can see how the pinwheel calculator worked in this video.
Britannic calculator, Yorkshire Insurance (early 20th c.)
The Britannic pinwheel-style calculating machine above was the pride and joy of James Edwin Preston who started work at the Yorkshire Insurance Company in 1907 and retired as the company's actuary in 1947. Sadly we don't have a photographic portrait of Mr Preston but he is probably somewhere in this photograph of Yorkshire staff at a garden party in 1914.
Yorkshire Insurance, Photograph of garden party (1914)
According to the reminiscences of Mr Crowe, who worked for the same company, the 'precious calculator' was often borrowed by the life department when Mr Preston was away, until the day a junior put it on a sloping desk top and watched with horror as it slid and crashed to the floor.
History does not relate how Mr Preston reacted to this disaster but the need to treat calculating machines with care was still being made clear to new staff at Sun Life in the early 1950s as the extract below, from a training guide for Monroe calculators, shows.
Always be sure that the machine is firmly placed on the desk in such a way that it cannot easily be knocked off. The machine is very fragile and if it should get knocked on the floor it will probably be an expensive matter to have it repaired. - training guide for Monroe calculators
The full text of the introductory page of the guide is worth a read.
Sun Life, 'How to work a Monroe' (1950s)
Monroe calculators were first produced in America in 1912, based on another design by Frank S Baldwin, and various models were used in our offices into the early 1970s. This link will take you to a video about how the Monroe worked.
This photograph from the early 1960s shows a member of Sun Life staff using his Monroe...
Sun Life, Member of staff using Monroe calculator (1960s)
.. while some of you might recognise this as the Marble Hall in Norwich complete with very messy desks featuring Monroe calculators in the mid-1960s.
Norwich Union, Staff at their desks in Marble Hall (1960)
Among the Monroe machines and piles of paper (no clear desk policy then) you might also be able to spot the old-fashioned-looking Fuller calculators which were used alongside mechanical calculators at Norwich Union into the early 1970s.
Here is another member of Norwich Union staff making use of his Fuller calculator.
Norwich Union, Member of staff using Fuller's calculator (1960s)
Fuller calculators had been produced from the 1870s as a development of the slide rule and we have a number of examples in the archive collection. Another cylindrical calculation device in use in our offices was the Curta, a small mechanical calculator patented in Austria in 1938 and developed by Curt Herzstark while he was imprisoned in Buchenwald concentration camp.
Curta calculator, General Accident (1955)
Herzstark designed the calculator to fit in the palm of a hand so that it could be used by people away from their offices and it is considered to be the first mechanical pocket calculator (although it would have looked a bit bulky if it had actually been kept in a pocket). The Curta was affectionately known as the ‘pepper grinder’ because of its distinctive shape and was produced commercially between 1948 and 1972; our example, which belonged to staff at General Accident, dates from 1955.
Aside from the Curta, progress in the development of calculation machines mostly consisted of the introduction of electrically driven motors to save operators from manually turning handles; the lucky man in the foreground of this photograph at Friends' Provident's new Dorking offices in 1959 seems to have been given an electronically driven Madas calculator...
Friends' Provident, Dorking offices (1959)
... while his colleagues in the pensions department appear to still be working with mechanical versions.
Friends' Provident, Dorking pensions department (1959)
However, while the man with the Madas may have been saving his wrists by using an electronic motor, the calculation was still being worked out mechanically – the first fully electronic calculator did not appear until 1961, and her name was ANITA.
ANITA was the code name for the initial feasibility investigation into the production of a commercially viable desktop electronic calculator and later the initials were said to stand for A New Inspiration to Arithmetic (or Accountancy). The earliest Anita's we have in our collection date from 1969 and were 10-digit with a 'Nixie'-type tube display. They looked like this:
ANITA 1000 calculator, Norwich Union (1969)
We've taken the top cover off one of our examples so you can see the circuit boards and the 'nixie' or cold cathode tubes used to illuminate the display.
Inside the ANITA 1000
Although some early users missed the noise of their mechanical machines and felt the key stroke was too light, it is clear that many operators became very attached to their Anita's – one of those in the collection was even given the surname Ekberg in reference to the beautiful model and actress Anita Ekberg who starred as Sylvia in Federico Fellini's film La Dolce Vita.
Here is a very studious member of Sun Life's staff using an Anita in 1971; the photograph was used in a recruitment brochure and was probably intended to give potential staff an idea of the cutting-edge technology they would be using.
Sun Life, Member of staff using an ANITA (1971)
We have a number of other Anita calculators in the collection including these Anita 811's, the company's first hand held electronic calculators, which were launched in 1972 and even came with branded Norwich Union cases.
ANITA 811 calculator, Norwich Union (c1972)
As you can see, calculators developed beyond all recognition in the just over 100 years between Mr Hancock of the Patriotic and the unidentified member of Norwich Union’s staff who used the Anita 811. From shaky beginnings as specialist machines for the actuary, calculators became ubiquitous office tools for all members of staff.
This journey from the precision engineering of wheels and cogs through the development of cathode tube and valve displays to smaller and smaller circuit boards eventually led to the pocket calculator – a mathematical aid we take for granted and which most of us use without any understanding of how or why it works. As Mr Hancock might have suspected, staff of today just accept the calculator and are grateful to be saved from the 'wearing brain work'.