The Napoleonic Shutter Station on the Clock Tower - debunking some cherished myths

Sutherland's view of the Clock Tower and water pump, c.1812
St Albans Museums

On 4 November 1807, St Albans Borough Corporation gave permission for the construction of a shutter telegraph station on top of the Clock Tower, the fifth in a line of nineteen communications stations between the Admiralty in London and the Navy anchorage at Yarmouth in Norfolk. (See Fig. 1) This enabled swift naval deployment in the Baltic to protect vital shipping supplies and forestall any pro-French moves by Scandinavian countries.

‘Calypso ready for sea’

The line operated from 1808 to 1814. The first operational message communicated by our station originated from Yarmouth and was ‘Calypso ready for sea’.

It has been claimed that our station was part of a semaphore system and that messages could be sent up and down the line in five minutes. Contemporary facts disprove this.

Before the outbreak of the French Revolutionary Wars, besides communicating by sea, the Admiralty relied on messages travelling by horse, either courier or mail coach. The fastest a message could be sent from London to Portsmouth was 4.5 hours (just over a day from Yarmouth to London).

French system the first in Europe

France stole a march in 1794 when Claude Chappe established an operational line of semaphore telegraph stations, communicating messages in code via pivotal arms. As Napoleon’s forces moved across Europe, a network was established connecting Paris to Amsterdam, Boulogne, Brest, Mainz and Venice. Messages could be sent 300 miles in 60 minutes.

With Britain threatened by invasion, faster communications between the Admiralty and its ships were urgently needed. There was no desire to copy Chappe’s system. Revd Lord George Murray designed an alternative which became operational in January 1796. The first line ran from the Admiralty to Deal with others later to Portsmouth, Plymouth and Yarmouth.

Fig. 1: Clock Tower, 1809, by George Shepherd. The shutters are at horizontal showing no message is waiting to be sent. (Reproduced under a Creative Commons Licence, © British Museum)

Stations were established at regular intervals manned by Royal Navy personnel. They observed the signals from the adjoining stations and replicated them by opening or closing six shutters, contained in a vertical frame above the station, hence ‘shutter telegraph’ (See Fig. 2). 64 combinations were possible, a combination representing a letter, number or a particular phrase.

After 1815 the Admiralty adapted the Chappe semaphore telegraph, which had more permutations, and installed a line to Portsmouth in 1822. To signify it was different from the shutter telegraph, the French term ‘semaphore telegraph’ was abbreviated to ‘semaphore’, until replacement by the electric telegraph in 1847. So not a semaphore then.

Fast …

What about the speed? This, of course, depended on the length of the message but as 18 stations on our line had to signal they were sending a message, spell out most of the words and the final station acknowledge receipt, five minutes is implausible, to say the least.

Fig 2: Modern illustration of the telegraph in operation

Fortunately, a contemporary newspaper and two logbooks from the Gog Magog Hills station near Cambridge on our line (now in the National Maritime Museum, Greenwich), provide more details. In 1808 The Norfolk Chronicle recorded an expectation of messages being sent along the line in 17 minutes.

… but not that fast

Reality was something different. The logbooks show that sending a message involving three shutter resets to the next station took two minutes – so 38 minutes in total along the line.

A complicated message took much longer, one involving 21 resets took 29 minutes so just over 9 hours for the whole line, but a lot better than over a day by horse! Of course, to work, signals had to be visible. The logbooks show that in December 1813 and January 1814 on 17 days in both months, fog prevented messages being sent. In these cases, horses were used.


This is an edited version of an article that first appeared in the SAHAAS Newsletter 217 of August 2022. If you would like a copy of the bibliography for this article, please email