This page is part of © FOTW Flags Of The World website

England Expects Every Man To Do His Duty

Last modified: 2005-03-12 by rob raeside
Keywords: royal navy | united kingdom | england expects | telegraph flag |
Links: FOTW homepage | search | disclaimer and copyright | write us | mirrors

Nelson's signal at Trafalgar shows some of the shortcomings Howe's code. The first eight words were each signaled with a three flag hoist. Even the two letter word "do" needed a hoist of three flags. Nelson had wanted to send "confides", but the word was not in the code book so he settled for "expects" which was. The last word "duty" was also not in the code book, and the closest words that were, "best" and "utmost" were not considered appropriate. "Duty" therefore had to be spelt letter by letter which took seven flags. Because illustrations of the signal show all the flags at once it is sometimes thought that this was how the signal was sent. Actually of course it was sent in a succession of hoists over a period of four minutes.

To avoid the code being deciphered, in time of war the actual meaning of each flag in Howe's code would change periodically (every six months(?)). Hence, would one expect Popham's code to change with it. Thus to resend Nelson's signal six months later, it appears one would have needed different flags.)

It used to be thought that Nelson's signal had been sent using the 1803 code. In about 1912 Perrin, the Admiralty Librarian who wrote "British Flags", discovered that a 1799 code book had been captured by the French in August 1803, and that consequently the code had been changed before Trafalgar in October 1805. Illustrations of Nelson's signal before between 1885 and 1908 show the correct flags in the wrong arrangement.

David Prothero, 4 December 2001

One possible representation of the sequence of flags can be seen in this set of poles.

The 1799 code was changed by an Admiralty Circular of 4th November 1803. Nelson's Trafalgar signal was shown correctly until 1885. In that year it was pointed out that the Signal Book of 1799 was not replaced until 1808, and that the signal made at Trafalgar in 1805 must have used the 1799 code. The Admiralty were persuaded that this was correct and a coloured leaflet was issued illustrating the signal according to the numerary code in the 1799 Signal Book.

In 1908 Perrin found a book of the numerary code, corresponding to that in the 1808 Signal Book, that had been authorised and signed by three admirals, who were in office together in the Admiralty only between 21st January and 15th May, 1804. This convinced the Admiralty that the original sequence of flags was correct, and a circular was issued admitting that the leaflet of 1885 was wrong.

Remembering that the numerary code was only a part of the complete Signal Book, the sequence of events was:
1780, ca.: Howe's code, partly table-based, partly positional.
1788: An officer of the Navy published a numerical code system.
1790: New signal book by Howe, using a numerical code system. Signals: Must have changed, since they are no longer positional. Still included sail-signals. 260 entries. Numbers: may or may not have been equal to 1788. Was gradually expanded until: 
1799. New Signal Book. Signals: Sail signaling was dropped. 340 entries. Numbers: Unchanged?
1800: Popham introduces his 'Telegraphic Signals of Marine Vocabulary'
1803: Popham's code officially adopted by British Royal Navy. 3000 entries.
1803. August. Code book captured.
1803. November. Instruction to change the numbers that had been assigned to the numerary flags in the 1799 Signal Book.
1804. Revised code book.
1805, September: 50 Popham code-books issued to British fleet at Cadiz.
1805, October. Trafalgar; Nelson's signal made in revised code.
1808. Revised version of 1799 Signal Book. Numbers: Unchanged (from 1804). Signals: ? Depends on what was actually changed. The need to add directly coded signals would probably have lessened due to those being codable with Popham's code as well.
1813: Expanded version of Popham's dictionary, to 6000 phrases and 60,000 words (one sources has "30,000 words"). 
1885. Assumption that Nelson's signal must have been made in the code shown in the 1799 Signal Book.
1908. Discovery of 1804 code book and recognition that Nelson's signal was made in the revised code.

Due to a peculiarity of that part of the code used to send the letters of the alphabet, the word "duty" appears to be mis-spelt.
"T" the 20th letter of the alphabet was signalled "19", because the alphabet had been reduced to 25 letters and "I" and "J" were both signalled by "9". This moved all subsequent letters up one number. However "U", the 21st letter was signalled "21", because in the alphabet it was listed after "V".

David Prothero and Peter Hans van den Muijzenberg, 5 December 2001

Howe's code allowed only 1000 signals, but apparently the Popham code could send those 3000 entries by adding a separate flag for each of three tables of 1000 entries. This appears to be a rather cumbersome way of simply adding a fourth digit, which could have been done more consistently by adding a third repeater. I do wonder how this was for the eventual 66,000 entries. I doubt they used 66 flags merely to indicate the correct table. So there must have been a table of sorts to indicate the table with a more limited set of flags. Adding yet a fourth repeater would have given room for 100,000 signals. (All on the assumption there was room for four-flag resp. five-flag hoists.)

Further information can be found at The Early History of Data Networks - Mirrors and Flags and Signal Flags

Nelson's words were changed to numbers using Popham's Vocabulary Book. The numbers were signaled with the flags of the 1790 Signal Book selected in accordance with the revised numbering of 1803, which also introduced a white flag and removed the red flag. 
253 269 863 261 471 958 220 370 4 21 19 24
The first substitute (D) was used for the second 2 in 220.

The hoist 253 for "England" was preceded by the Telegraph Flag to show that succeeding hoists constituted one message. If 253 had not been preceded by the Telegraph Flag it would have had a different meaning.

David Prothero, 6 December 2001

The signals officer of the Victory at Trafalgar was Lt John Pascoe, who many years later wrote that Lord Nelson had originally asked for "England 'confides' that every man will do his duty", and that he had suggested "expects" in place of "confides" because it would take fewer signal flags. I have also seen it suggested that the signal was originally given as "Nelson confides etc.", but (whilst England would take very many fewer flags than 'Nelson') Pascoe's account appears to contradict this and I have been unable to confirm it elsewhere.
Christopher Southworth, 23 September 2003

The book by Captain Barry Kent entitled: Signal! - A history of Signalling in the Royal Navy describes the procedure of hoisting signals. On Plate I opposite page 100 he shows the Trafalgar signal in full colour flags as well as a diagram of a three masted ship of the line with the sequence in which the flag hoists are read. He states that the signal.......was of course sent as a series of separate hoists. He goes on to state that: The principle (of reading the hoists) is: main, fore, mizzen; starboard before port; upper yards before lower. In practice it would be unusual for more than two or three to be hoisted at the same time, and Nelson's signal would have been transmitted by a quick succession of hoists. According to the logs of the ships reading it, the whole signal took no more than about four minutes to transmit.

Each hoist was kept flying until all the ships addressed had hoisted the acknowledgment meaning that they have seen and understood the signal. With a large fleet like Nelson's at Trafalgar, this could take a long time and to facilitate signalling in such circumstances, frigates were stationed along the line to windward to repeat the signal for the benefit of the ships furthest away. The acknowledgements were in turn repeated by the frigates for the flagship's benefit. For all of Nelson's ships to have read and understood such a long signal in the space of only four minutes, reflects great credit on the competence of the signal officers and signalmen in that fleet.
Andre Burgers
, 24 September 2003

The following information came from the Ministry of Defence, Admiralty Library:
The signal comprised eleven separate hoists to the mizzen masthead. HMS Victory was at the head of the fleet, the wind was aft, and the repeating ships were in their station, so the mizzen masthead was the most generally visible location as well as the most usual. The flagship's signals would have been bent onto the mizzen topgallant signal halyard.
David Prothero, 5 November 2003

All the more credit to Lt Pascoe's team of signallers to pass that long signal in only four minutes using only one halyard.
Andre Burgers, 24 September 2003

Under the influence of the military historian John Keegan, I would wonder whether the question is whether the achievement is creditable or rather whether it is credible. Here's why:

I've always understood the message to be 12 hoists, not 11, but even for eleven hoists four minutes seems improbably fast. If we assume that Pascoe timed the signal from the moment the first word started moving up the halyard until the letter "Y" of "duty" was closed-up (which omits the time taken to bend on the flags for the first signal and the time to haul down the last "Y", as well as the time to hoist the ensuing signal for close action), we have 21 trips up and down the mast. Besides the time to physically raise and lower the flags, we have to take account of the time needed to replace one signal with another in between hoists, plus the time each signal had to remain flying while the rest of the fleet acknowledged it. Even if we allow two seconds to make each of the 10 changes after "England," and ten seconds for each acknowledgement through the letter "t"--both of which strike me as unbelievably fast--that leaves only 120 seconds in which to make those 11 haulings-up and 10 haulings-down, or 5 2/3 seconds for each. The mizzenmast of HMS Victory measures 152 feet (46.3 meters) from the waterline; let's guess 115 feet from the deck. Could Pascoe's signalmen really move the flags up and down at a rate of 20 feet per second?

Andre's the former signal officer, so if he says it's possible, I'll accept it, but it sounds like Pascoe's memory may have exaggerated what was undoubtedly a remarkable performance, my nitpicking questions notwithstanding.
Joe McMillan, 7 November 2003

I am inclined to follow Barry Kent on the question of the hoisting of the famous signal. I am also doubtful that the eleven hoists could have been done in four minutes using only one halyard and here I speak with some experience having been a Signals officer myself and having participated in and conducted numerous signal hoisting exercises. And we were using flag lockers while I believe in Nelson's time they were still using flag bags to pull and stow the flags. Barry Kent, in his book Signal! on Plate 1 following page 100, shows the signal and alongside it, a sketch of a ship of the line with all the hoisting positions on the various masts numbered in the order that the signals should be read. He does not actually state that this was the manner of hoisting the Trafalgar signal, but the implication is clear.

David, what is the source of your Admiralty Librarian for the statement that only the mizzenmast in Victory was used for signalling? The team of bunting-tossers also seems to have been a bit too small for such fast work on such a long signal. Seamen Aslet, Heaver and Roome plus the two boys and the Middy, would have been scarcely enough to serve the mizzen signal halyard. If we remember that two further signals were made, the last, for close action, at 1220, this small team must have been very busy indeed over that period.

Joe, the four minutes are mentioned also by Barry Kent on page 7 of his book, taken from the signal logs of several ships reading the signal, and the first hoist having gone up at 1156.
Andre Burgers, 7 November 2003

I checked with the Admiralty Library. It was a mistake; twelve hoists not eleven.
Four minutes was thought to be a reasonable time for a signal of twelve hoists bearing in mind -
1. The time of four minutes was derived from log entries which were not timed to the second. It might have been nearer five than four.
2. This was the flagship; the signal party would have been very experienced.
3. Hoists were not acknowledged. Each went straight up and straight down, with no pause at the masthead.
David Prothero, 10 November 2003

The source? I don't know. The Gordon and Wheeler-Holohan editions of Flags of the World say that the signal was "sent up in succession to the main topgallantmast-head". This may have been an assumption based on the fact that the signal 'engage the enemy more closely' was hoisted there.
David Prothero, 11 November 2003

Telegraph flag

According to Perrin (W.G.Perrin, 'British Flags', Cambridge University Press, 1922) Page 178, the "preparitive" or "telegraph" flag in Popham's code was "a flag divided diagonally into white and red".
Christopher Southworth, 22 June 2004

According to W.J.Gordon in "Flags of the World" 1918 edition, the 'Telegraph flag' was hoisted at the yard-arm, while the twelve hoists of the message were hoisted in succession at the main topgallantmast-head. E.M.C.Barraclough in the 1965 edition wrote of the telegraph flag; "It could remain in this [conspicuous] position throughout the period of the message, and then be hauled down to indicate that it had been completed, or, alternatively, hauled down before actually commencing to make the signal. In the latter case, provision was made for hoisting another special flag, called the "Message finished" flag [diagonal blue and yellow]. However as far as is known, the last mentioned was never used."

Its precise appearance is uncertain. Popham described it as "a diagonal red and white flag". It is, according to Barraclough, shown in contemporary signal books as a rising diagonal, sometimes red over white, and sometimes white over red, and that the actual flag used at Trafalgar is not known. T.Wilson in "Flags at Sea" shows red over white. Captain B.Kent in "Signal!" shows white over red.
David Prothero, 23 June 2004

The signal flown by the Victory at the beginning of the battle (and continued until shot away) was No. 16 (in Sir Hope Popham's extended code as revised January 1804 and used at Trafalgar) "engage the enemy more closely" which was flag No, 1 (a blue cross on a white field), over No.6 (divided horizontally blue-white-red). This would originally have been preceded by the 'Telegraph' or 'Preparatory flag above the message (divided diagonally from the bottom left red over white) but this would have been hauled down prior to the battle. Nelson also intended to raise the signal for "prepare to anchor" after the battle, but due to circumstances his successor and former second in command, Collingwood, never used it. I do not, unfortunately have the full code, but the signal book reputedly used at Trafalgar is in the collection of the National Maritime Museum (NMM, SIG/B/76), and may be consulted by appointment (or at least could be when I last checked)..
Christopher Southworth, 11 February 2005

See also: