Chapter VI: Flag Signals (3)

A less effective method was to transpose all the flags. It was less effective since, _ex hypothesi_, the enemy had the signal books, they had therefore only to note the colours of the flags which preceded some easily recognised manoeuvre, such as "Make more sail," "Bear up and sail large," to discover what flags were now being used to indicate the signal numbers which they saw against that signal. A little patience and ingenuity would then supply the key to the changes in the flags, whereas if all the signal numbers were irregularly transposed, each signal would have to be re-constituted separately by the enemy.

About fifteen months before Trafalgar the 12-gun schooner, 'Redbridge,' commanded by a Lieut. Lemprière, who, to judge by Nelson's remarks about him, was not particularly efficient, was captured by some French frigates off Toulon. Such a small ship, commanded by an officer of such subordinate rank, was not allowed the confidential signal books, but Lemprière had, in common with many other junior officers, obtained a surreptitious copy for himself--one of those little manuscript and hand-painted signal books one sees in museums, or occasionally picks up in second-hand book shops.

This book Lieut. Lemprière had neglected to throw overboard, and when, later on, one of Nelson's scouts, looking into Toulon harbour to see if the French were getting on comfortably, found the captured 'Redbridge' just outside, the latter made the signal for the scout ship to anchor; but fortunately the officer in command was a little sharper than the 'Redbridge's' late commander had been, and the net was spread in vain.

Directly Nelson learned this he changed the flags and reported the matter to the Admiralty. On the 4th November, the Admiralty, despite the objection of Lord Keith that a change of signal numbers would be better, issued a circular letter to all commanders-in-chief telling them to alter their numeral flags in accordance with a painted copy enclosed with the letter. These are the flags shown in the third column of the plate. Further, as their Lordships had reason to apprehend that Lieut. Lemprière was not the only officer under commander's rank who had obtained a copy of the signal book, the strictest injunctions were to be given that such improper proceedings were not to take place in future, and existing irregular copies were to be impounded.

The flags in use in the Mediterranean were changed in numerical value in accordance with the Admiralty order on the 16th January, 1804, and these new numeral flags were the ones used at Trafalgar, and, in fact, until the end of 1810[390].

We have seen that Howe's second code, first in its original form, next as re-arranged by McArthur, and finally as simplified and expanded for the 1799 Signal Book, in each case with practically the same numeral flags, sufficed, when supplemented by Popham's Vocabulary, for the navy's needs throughout the most sustained and strenuous struggle that until the recent war had ever fallen to its lot[391].

Towards the close of this period, however, Popham, who seems to have devoted all the time not required for his professional duties, or for defence against the attacks made upon him, to the improvement of his code[392], brought out a greatly enlarged and improved vocabulary, which is best described in his own words:

The present edition is wholly new cast and composed; very
considerably enlarged by additional materials; and, as I trust,
improved by a distribution of those materials which was intended to
increase the facility of reference.

It consists of nearly 6,000 primitive words, exclusive of the
inflexions of verbs, &c., making in all upwards of 30,000 real
words; the sentences have also been extended to about 6,000, with
1,500 syllables, a Geographical Table, a Table of Technical Terms,
a Table of Stores and Provisions, and a Spare Table for Local
Significations.

With so large a number of signals the limitation in the number of flags that can be conveniently hoisted at a time made itself at once felt, and Popham found himself compelled to abandon the "simple numerary" method. Taking the limit of convenience at three flags, the ten numeral flags of the Admiralty Signal Book would suffice for only 999 signals. Popham, therefore, after calculating the number of combinations available with various sets of flags, chose 23. These, which were mostly of his own design, he denoted by the numbers 1 to 9 and the letters A to O. With these 23 flags the number of available combinations is as follows:

Singly 23
Two at a time 506
Three at a time 10,626
------
Total 11,155

So that upwards of 11,000 signals could be made using no more than three flags in one hoist.

Hoisting four flags at a time, the number of possible signals with 23 flags is increased by 212,520, making 223,675 in all, sufficient for the most exhaustive vocabulary. If this code had been invented seven years earlier, Nelson could have made his signal in precisely his own words, the first three of which would have been as follows: England, 69B; confides, 5I3; that, B67. Curiously enough, Popham, in 1812, seems to have preferred "confides" to "expects," for the former was one of "most needed" words and could be made by three flags, whereas "expects" was No. 6138 and required four.

In demonstrating the extensive use to which his code could be put, Popham gives one or two amusing illustrations, apparently in all seriousness. Thus:

FA1 Have you an idea
G647 a change of ministers is about to take place

52A Certainly
8BF not
G643 ministers are gaining strength

BOE Your
AC8 sister
852 married
85F to
C87 a Lord of the Admiralty

This vocabulary, published in 1812, was such an evident improvement that it was issued to the fleet in 1813.

In 1816 it was revised by Popham and re-issued by the Admiralty as an official Vocabulary Signal Book. Eleven years later the signal books were again revised and re-cast in three volumes:--

1. The General Signal Book, containing evolutionary and battle signals, to which the numeral flags were henceforth appropriated.

2. The Vocabulary Signal Book, containing words and general sentences, to which were appropriated the alphabetical flags, now increased to 21 by the addition of P, Q, R, S, T, V, and Y.

3. Night and fog signals.

In 1882, W was substituted for V and slight alterations were made in some of the flags, and in 1889, when the signal books underwent a more extensive revision, the alphabetical series was completed by the inclusion of all the letters, and a series of numeral pendants was added. The flags thus finally established in 1889 remain in use to this day. It may be observed that they contain a number of survivals from Howe's Code of 1790, four indeed with their original significance. They are as follows:

(1) An alphabetical series, in which many of the flags are of the
same design as those in the International Code (see p. 184) but
with different significations, as follows:

A flag, diagonally striped yellow & red = Y of Internat^l Code

B flag, white, bordered with blue & pierced
with red = W " " "

C flag, divided diagonally yellow, blue, red
& black = Z " " "

D Pilot Jack (Union Jack with white border[393])

E flag, divided horizontally blue, white, blue = J " " "

F flag quarterly yellow & black = L of Internat^l Code

G flag divided vertically white, black, white

H flag, yellow with black ball = I " " "

I flag, blue with yellow saltire = Numeral 7 of 1790

J pendant, divided vertically white & red

K flag, divided horizontally yellow & blue

L flag, white with red saltire = V of Intern^l Code

M pendant, divided horizontally red, yellow,
red

N pendant, divided horizontally yellow, blue,
yellow

O flag, divided diagonally red & yellow = O " " "

P pendant, blue with white cross

Q pendant, red

R pendant, yellow with red cross

S pendant, divided vertically blue & yellow

T pendant, blue with white ball = D " " "

U burgee, divided vertically white & blue = A " " "

V flag, white pierced blue = S " " "

W flag, yellow = Q " " "

X pendant, striped vertically black & yellow

Y pendant, white with red border

Z flag, chequered blue & white = N " " "

Affirmative }
Preparative } as in 1790 (see Plate XIII)
Negative flag, white with 5 black crosses

(2) A series of numeral flags:--

1 as No. 5 in 1790
2 " 2 "
3 flag, chequered yellow & blue
4 as No. 9 in 1790
5 flag, divided horizontally red & white
6 as No. 8 in 1790
7 " "
8 flag divided vertically red, white, blue
9 as "Dissent" in 1790
0 as No. 0 in 1790

(3) A series of numbered and special pendants:--

1 divided vertically red, white, blue
2 " horizontally " "
3 white with red cross
4 striped vertically white & red (16 stripes)

6 white with 2 black crosses
7 divided quarterly white, black, yellow, red
8 red
9 white with red ball
0 divided vertically yellow & blue
Interrogative divided quarterly red & white
Answering " vertically " "
Guard, red with white cross
Numeral, chequered blue & yellow
Church, as No. 2 but with St. George's Cross in chief.

These pendants are all of a different shape from those employed in the alphabetical series, being more narrow and elongated, and cut off square at the end, whereas the former are almost equilateral and might perhaps be more correctly designated as "triangular flags."

The "substitutes" used are as follows:

1st. (repeating the 1st. flag or pendant of a "series") the Affirmative
flag
2nd. ( " 2nd. " " " ) Answering
pendant
3rd. ( " 3rd. " " " ) No. 2 pendant
4th. ( " 4th. " " " ) No. 0 "
1st. (when using the numbered pendants) Interrogative pendant
2nd. ( " " " ) Answering "

With these we have probably reached the final development of form so far as flag signals are concerned, for wireless has taken the place of visual signalling to such an extent that it is not likely that circumstances can now arise that will necessitate any radical recasting of the flag signal system.

(v) COMMERCIAL CODES

Some elementary flag signals, notably that for a pilot, were in use among merchantmen at least as early as the fifteenth century, but the first attempt to supply a code of signals suitable for merchant ships appears to have been that made by Sir Home Popham in 1804, when, at the request of the East India Company, he compiled a book of "Commercial and Military Signals" for the use of the ships in their service. In this book the "military" element preponderates, as might be expected from the circumstances of the time at which it was drawn up. The signals relate almost exclusively to the fighting and manoeuvring of ships sailing in convoy. After the peace, in 1817, Captain Frederick Marryat drew up what may be regarded as the precurser of modern commercial codes. It was in six parts, each in the simple numerary system, with a distinguishing flag to indicate the part to which the signal related. The parts were as follows: 1. Names of men-of-war. 2. Names of merchantmen. 3. Ports, headlands, etc. 4. Sentences on various subjects. 5 and 6. A vocabulary adapted from Popham. This signal book went through ten editions before the author's death in 1848.

In 1855, owing to the enormous increase in communication by sea, and the adoption of an official number for every merchant ship imposed by the Merchant Shipping Act of 1854, the need for a much more ample code had become urgent, and the Board of Trade appointed a Committee to draw up a new one. This committee reported that an efficient code "ought to provide for not less than 20,000 distinct signals and should, besides, be capable of designating not less than 50,000 ships, with power of extension if required." They further stated that "a signal should not consist of more than four flags or symbols at one hoist," and pointed out that under this condition the ten numerals without repeaters would make only 5860 signals, or 9999 signals with three repeaters. They therefore abandoned the numeral system and chose 18 flags which, by using two, three, or four flags together, allowed of 78,642 permutations. These 18 flags, which embodied most of those already in use in Marryat's Code unchanged, or with slight alteration, were designated by the letters of the alphabet except a, e, i, o, u, x, y, z, the vowels being omitted because "by introducing them every objectionable word composed of four letters or less, not only in our own but in foreign languages, would appear in the code in the course of the permutation of the letters of the alphabet."

The flags adopted were as follows:

B red burgee (flag with swallow-tailed fly)
C pendant white with red ball
D " blue with white ball
F " red " " "
G " divided vertically yellow and blue
H flag divided vertically white and red
J " " horizontally blue, white, blue
K " " vertically yellow and blue
L " " quarterly blue and yellow
M " blue with white saltire
N " chequered blue and white
P " blue pierced with white square (the blue Peter)
Q " yellow
R " red with yellow cross
S " white pierced with blue square
T " divided vertically red, white, blue
V " white with red saltire
W " blue bordered white pierced with red square,

and an "Answering Signal" or Code Pendant divided vertically red and white in five stripes. In 1857 this "Commercial Code of Signals for use of all Nations" was issued by the Board of Trade, and it sufficed for the needs of the next thirty years, the name being changed to International Code about 1880. At the end of that period the Board of Trade appointed a Committee to bring it up to date and to consider whether a system of night signals should be added to it. In 1899 this Committee submitted a revision of the Code "which differed from that then current only in the omission of certain signals which had become obsolete and the substitution of certain other signals for which modern developments had created a demand." The criticisms of the foreign maritime powers upon this book led, however, to a complete revision and recasting of the old code for reasons which the Committee summarise in their final report[394] in 1896 in the following words:

Since the old Code of signals was first issued there has been
a very considerable increase in the average speed of vessels
belonging to the Mercantile Marine, owing both to the larger
percentage of steamers as compared with sailing vessels and to the
greater speed to which steamers now attain. Vessels consequently
remain within signalling distance of one another and of signal
stations for a much shorter time than was the case 40 years ago,
and it is necessary that an efficient Code of signals should
provide the means of rapid communication. In a Code, such as the
International Code, in which signals are made chiefly by means of
flags, rapidity of communication can best be secured by reducing
to a minimum the number of flags required to make the signals,
since every additional flag in a hoist involves delay in bending
on the flags on the part of the person making the signals and
delay in making out the flags on the part of the person taking in
the signals, and to enable this to be done without the number of
signals in the Code being reduced, it was necessary to provide an
increased number of two or three flag signals by adding additional
flags to the Code.

The number of signals which can be made by permutations of 18
flags, no flag being used more than once in the same hoist, is as
follows:--

One-flag signals 18
Two-flag signals 306
Three-flag signals 4,896
Four-flag signals 73,440
------
Total 78,660

The number of signals actually provided in the old Code which can
be made by the Code flags is:--

One-flag signals 4
Two-flag signals 215
Three-flag signals about 4,500
(_a_) Four-flag signals, excluding those }
representing the names of places }
and ships, about 8,700 }
(_b_) Four-flag signals representing the } 29,600
names of places, about 3,400 }
(_c_) Four-flag signals representing the }
names of ships, about 17,500 }
------
Total about 34,319

The following is the number of signals which can be made by means of
the 26 flags which we have adopted, no flag being used more than
once in the same hoist:--

One-flag signals 26
Two-flag signals 650
Three-flag signals 15,600
Four-flag signals 358,800
-------
Total 375,076

Moreover, by using the Code pennant over and under one or two flags
of the Code, the following additional signals made by not more than
three flags are obtainable:--

Code Pennant over one flag 26
Code Pennant under one flag 26
Code Pennant over two flags 650
Code Pennant under two flags 650
-----
Total 1,352

It will therefore be seen that by the adoption of the eight
additional flags, many of the more important signals which have
at present to be made by three-flag hoists can be converted into
two-flag signals, and that all the four-flag signals (excluding
those representing the names of places and of ships) in the old
Code can be made by three-flag signals, while between 3,000 and
4,000 new signals to be made by hoists of not more than three flags
can be added.

The abolition of all four-flag hoists for general signals will very
greatly increase the _Rapidity_ with which communication can be
held by means of the International Code of Signals.

It will also tend to secure another essential in efficient
signalling, viz., _Accuracy_, for every flag added to a hoist
affords an extra risk of mistake, both in bending on a wrong flag
and in reading off the flags in the hoist incorrectly.

In addition to the gain of rapidity and accuracy of signalling, the
inclusion of the eight new flags has, as we have already stated,
afforded the means of providing a large number of signals which do
not appear in the current Code, and we have availed ourselves of
this possibility to the extent of adding some 4,000 new signals.

Moreover, the fact that under the proposed scheme there is a flag
to represent every letter of the alphabet has enabled us to arrange
for a system of spelling proper names and words not appearing
in the Signal Book, which we regard as less cumbersome than the
Alphabetical Spelling Table which is at present in force.

These advantages appear to us to be so important that we have not
hesitated to increase the number of flags to be used, although the
step involved the abandonment of the Code suggested by us in 1889
and the preparation of an entirely new Signal Book.

The letters omitted from the alphabet in 1855 were now added, the objection which had led to their omission on the former occasion being regarded as a "sentimental rather than practical objection," though the Committee took care to eliminate objectionable words as far as practicable.

The flags now added were:

A Burgee divided vertically white and blue
E Pendant " " red, white, blue
I Flag yellow with blue ball
O Flag divided diagonally yellow and red
U " " quarterly red and white
X " white with blue cross
Y " striped diagonally yellow and red in ten stripes
Z " divided diagonally in 4 triangular parts, yellow, blue, red,
and black.

At the same time the F and L flags were slightly altered.

We have altered the flag F from a red pennant with a white ball on
it to a red pennant with a white cross, as the flag at present in
use is liable to be mistaken for Flag D (blue pennant with a white
ball on it), and we have altered flag L from a flag of yellow and
blue quarterly (the blue squares being at the top left-hand and
bottom right-hand corners) to a flag of yellow and black quarterly,
the black squares being at the top right-hand and bottom left-hand
corners. Our chief reason for making this alteration is that in a
calm it is difficult to distinguish the present flag L from flag K.

Certain of the above flags are used singly with special significations, viz. A by H.M. ships on full speed trial, B to signify that explosives are being landed or discharged, C as affirmative, D as negative, P to denote that the ship is about to sail and S as signal for a pilot, while L is used in the United Kingdom to indicate infection from cholera, Yellow fever or plague, and Q is generally used aboard to denote liability to quarantine.

Signals made with two flags are urgent and important. Of the three-flag signals those from ABC to AST relate to the compass; from ASU to AVJ to money; from AVK to BCN to weights and measures; BCO to BOZ to decimals and fractions; from BEA to CWT to auxiliary verbs and phrases; while the general vocabulary occupies the permutations from CXA to ZNV. The Code flag over two flags serves for Latitude and Longitude, Divisions of time, and the Barometer and Thermometer, and the Code flag under two flags from UA to ZY provides a numeral table. The geographical names of places are signalled by four flags from ABCD to BFAU, while the permutations from CBDF to CZYX are used for an alphabetical spelling table. This completes the flag signals of the Code, but it contains in addition a number of Distant Signals, Semaphore Signals and Morse Code Signals.

Combinations from GQBC onwards are used for the names of ships, which will be found in separate publications.

This new Code was published in 1899 and brought into force on 1st January, 1901, the old Code being used concurrently with it until the 31st December of that year. It has now reached the seventeenth edition, and a complete revision of it, which will probably entail the alteration of many of the flags, is occupying the attention of an International Commission, but it will be several years before this is brought into use.

In making a signal, a ship first hoists her ensign with the code flag under it, and if necessary the distinguishing signal of the vessel or station with which she desires to communicate. On seeing this signal the ship (or station) addressed then hoists the "Answering Pendant" (i.e. the Code flag) at the "Dip," that is, some little distance below its position when hoisted "close up" to the block at the masthead or yardarm through which the signal halliards are rove. The first ship then hoists her own distinguishing signal, consisting of the four letters appropriated to her name, and then proceeds with the signal she wishes to make. When the first hoist is noted down and translated in the ship receiving the signal, this ship hauls the answering pendant "close up" to show that the signal is understood and keeps it there until the signalling ship has hauled that hoist down; the answering pendant is then again lowered to the "Dip" until the next hoist is disposed of, and when the ship signalling has finished, she hauls down her ensign to indicate that the message is at an end.

Among signals of distress by means of flags--which from their nature are of international use and common to both men-of-war and merchant ships--the earliest appears to have been made by tying the ensign in a knot in the middle, or making a weft as it was called. Another, which appears to have been in use in the seventeenth century, was to invert the ensign; this, of course, could not be done with those ensigns (such as the modern French) which are symmetrical in design. The signal appears sometimes to have been given by placing the ensign in an unusual position, such as at the main topmast-head or in the shrouds. An instance in which the ensign was placed inverted in the shrouds will be found on page 199.

FOOTNOTES:

[338] Herodotus, VII, 128.

[339] Thucydides, I, 49.

[340] _Ibid._ II, 84; VII, 34.

[341] Thucydides, III, 90.

[342] Polyaenus, _Strategemata_, I, 48 (2).

[343] Livy, XXXVII, 24: "Eudamum in alto multitudine navium maxime Hannibal, ceteris omnibus longe praestantem, urgebat; et circumvenisset ni signo sublato ex praetoria nave, quo dispersam classem in unum colligi mos erat, omnes quae in dextro cornu vicerant naves ad opem ferendam suis concurrissent."

[344] This work Λέοντος ἐν Χριστῷ τῷ Θεῷ αὐτοκράτος τῶν ἐν Πολέμοις τακτικῶν σύντομος παράδοσις [Greek: Leontos en Christô tô Theô autokratos tôn en Polemois taktikôn syntomos paradosis] is usually assigned to Leo VI (the philosopher), but it is possible that it may have been written by Leo III (the Isaurian) and dates therefore from the eighth century. It has not been translated into English.

[345] Probably 1338, _vide Monumenta Juridica_; _The Black Book of the Admiralty_, ed. by Sir Travers Twiss (Rolls Series), Introduction, p. xxx.

[346] _Due Ordinanze Militari Marittime del Conte Verde_, Anno 1366, by Capitano di Corvetta Prasca in _Rivista Marittima_, June 1891. I have supplied as literal a translation as the state of the text admits.

[347] The object of this manoeuvre is not clear, but since the ram and the guns pointing forward were the only weapons of the galley it can be readily understood that friendly galleys would only face one another in exceptional circumstances.

[348] Commander Prasca suggests that _lon_ should be read as _bon_ and that the meaning is: to a place on the poop where the banner can be well seen. I suggest that it should be read as _son_.

[349] It was omitted from Howe's signal-book of 1790.

[350] _Exchequer Accounts_, 49/29 (7-10 Hen. V): "j banner de consilio de Armis Regis et Sancti Georgii."

[351] Transcribed from a Vatican MS. in Jal's _Archéologie Navale_, II, 107 et seq.

[352] Fernandez de Navarrete, _Coleccion de los Viages y Descubrimientos que hicieron por mar los Espanoles_, I, 410 et seq.

[353] Cada patron con un cómitre.

[354] Published with notes by Jal in the _Annales Maritimes et Coloniales 1842_, Part II, pp. 54 et seq.

[355] Calcet.

[356] Cappitaine. He is thinking only of the galleys.

[357] Soubz le vent, probably an error for "to windward."

[358] The gangway running along the galley amidships.

[359] La gallère ou care ou à la penne à demy pendant.

[360] See _Book of War by Sea and Land_, edited by Sir J. K. Laughton in vol. I of _The Naval Miscellany_ (Navy Records Society).

[361] "À demy clouée."

[362] "à moitie pendant." Presumably there was some difference in these two methods of tying a flag in a weft, but it is not known in what this difference consisted. Perhaps "a demy clouée" meant that the fly was gathered into the hoist, while the other expression denoted that the flag was gathered up horizontally in the middle, the normal method of making a weft.

[363] _Harl. MS._ 309.

[364] quarter.

[365] whither.

[366] _Pepys MS._ 1266.

[367] _Sic_; apparently the ship is to yaw, i.e. deviate from her course.

[368] _S. P. D. James I_, CLVII, 67.

[369] _The Naval Miscellany_ (Navy Records Soc.), vol. I.

[370] _S. P. D. James I_, XCII, 9.

[371] The latest MS. appears to be Sloane MS. 2427; the earliest that the author has seen is in the Bodleian. It was from a MS. almost identical with the latter that the printed edition was published in 1685.

[372] See p. 42.

[373] "lesqueles banères sount appelés baucans et la gent d'Engleterre les appelent stremers et celes banères signefient mort sans remède et mortele guerre en tous les lious où mariners sont."

[374] "nous ne sums tenus faire restitution ne amende si nulle chose eit esté fait ou prise par nous en ladite guerre; quar il est usage et ley de meer que de choses faites ou prises sur meer en guerre meisement ou ledit baukan soit levée ne doit estre fait restitution n'amende d'une partie ne d'autre." 'Baucan,' cognate with 'beacon,' must not be confused with 'bauçan.'

[375] See p. 192.

[376] This and the two following sections are in substance a revision of a pamphlet _Nelson's Signals, The Evolution of the Signal Flags_, written by me in 1908 and published for the Admiralty by H.M. Stationery Office. I am indebted to the Controller of that Department for permission to make use of it in this work.

[377] Corbett, _Fighting Instructions_ (N. R. S.), p. 99.

[378] Possibly as a result of the first collision with the Dutch, Tromp appears to have been ahead of the English in the matter of signals.

[379] Shortly after 1756 the white balls disappear, and a white square takes their place, the flag thus becoming what is now familiarly called the Blue Peter. By Rodney's time this flag at the mainmast head had become the signal to recall everyone to his ship, whence its present use to denote that the ship is about to sail.

[380] A flag is said to be "pierced" when it has a small square of another colour in the centre.

[381] In his memoirs it is stated that in 1746 he had written a book on Signals and Naval Evolutions.

[382] Possibly because with a "table" of 16 squares a side he was enabled to make 256 signals, with two flags only to each "hoist," whereas by the simple numerary method he could of course only make 99 with less than three flags.

[383] J. K. Laughton, _Letters and Papers of Lord Barham_ (N. R. S.), I, 340.

[384] Hitherto the official books of Instructions and Signals had been of folio size.

[385] A "cornet" is a swallow-tailed flag.

[386] According to present practice no signal made with more than one flag is obeyed until the officer making it hauls it down.

[387] Nicholas, _Letters and Dispatches of Lord Nelson_, III, 230.

[388] Richard Hall Gower, an officer of the East India Company's fleet, suggested in the 2nd edition of his _Treatise on Seamanship_, 1796, the use of a dictionary with the words numbered consecutively from 1 upwards under each letter of the alphabet. The letter and number were to be shown on a large board.

[389] The hoists were: Telegraph flag; 253 (England); 269 (expects); 863 (that); 261 (every); 471 (man); 958 (will); 220 (do); 370 (his); 4, 21, 19, 24 (duty). In the hoist for 220 the second flag was the "substitute," duplicating the numeral "2."

[390] See the last column of Plate XIII.

[391] In 1808 the 1799 book was reprinted, and the manuscript additions incorporated, but its form was not changed.

[392] He produced in 1804 a signal code for the East India Company's ships.

[393] During the war this flag was altered to one divided vertically yellow, red yellow, in order to obviate the use of the Pilot Jack as an ordinary signal flag.

[394] _Parl. Paper_ C 8354 of 1897.