The allocation and use of ship identification signal codes for merchant ships to WWII

(New page 22 July 2004; revised 7 August 2004)

 

 

 

Introduction

 

A major theme in historical maritime research is the use of signals. Flags were the primary means of communication and identification of ships at sea until the use of radio was well established but the international system of signals established in the mid-19th century continues in modified form today for use with visual, radio and electronic media.

 

Many sites easily found by internet searches illustrate the signal flags themselves so I give their description no further attention.

 

All the signal manuals I have located and most of the contents of  many web sites deal with operational signals such as “I need a pilot”, “Yellow fever on board”, “You are running into danger” etc. Operational signals are an important element of maritime knowledge but I have not investigated them as they are peripheral to the use of signal codes as ship identifiers which is my primary interest in the topic.

 

This item focuses on the use of signal codes as ship identifiers for European ships as there are superior alternative identifiers for data management purposes for British and American ships. The actual practicalities of using signal codes as identifiers in databases are discussed elsewhere on this site (link at end of item or click here).

 

 

List of headings

 

Introductory history

 

● Extension to American and European ships

 

Signal code lists in Lloyd’s Register

 

Nations that issued signal codes for merchant ships

 

The content of the signal code lists

 

The number of allocated signal codes circa 1886

 

The number of allocated signal codes circa 1913

 

● Allocated signal codes 1886 to 1932

 

The reallocation of signal codes

 

● Questions

 

● Sources

 

                                    Appendix: Using the signal code lists to track the survival of ships

 

 

Introductory history

 

There is a long history of the use of flags for signalling at sea, an obvious necessity for naval operations. A growing need for signalling systems for merchant ships was met during the first half of the 19th century by a number of signalling codes developed independently of government sponsorship and administration. The best known and most widely used of these was that invented by Captain Frederick Marryat which was widely adopted internationally. 

 

Marryat’s codes were used for much of the 19th century but were superseded by a system originated by the British government in the late 1850’s that was eventually adopted by all the leading maritime nations. Further discussion in this item is limited to the international codes of the second half of the 19th century and beyond. Marryat’s system is covered elsewhere on this site (refer Marryat).

 

The British-sponsored “Commercial Code of Signals for the Use of All Nations at Sea” of 1857 provided flag signal codes consisting of four letters to identify individual ships in conjunction with a national flag. The Commercial Code was renamed the International Code of Signals about 1870 and its present form is still known by that name today. Marryat’s code was originally published as simply “A code of signals for the use of vessels employed in the merchant service”, but just before the Commercial Code was introduced Marryat’s was renamed “The Universal Code of Signals …” by which name it was known thereafter and republished at least as late as 1869.

 

I do not have the exact details of the renaming of the Commercial Code to the International Code, but library catalogue entries for a book published in 1869 refer to it by the earlier name while the British Board of Trade published “The international (commercial) code of signals for the use of all nations” in 1871 which narrows the change down to a short time frame. This may well be indicative of when the International Code was first widely adopted in Europe in which case it would be no coincidence that 1869 is the last edition of Marryat’s code that I have found recorded.

 

The initial allocation of codes to British ships was published in the 1857 edition of the [British] Mercantile Navy List (documented elsewhere on this site). Signal codes were not allocated to all ships as, generally, only those engaged in foreign trade required them although many small ships did have them. However, Britain instituted at the same time a system of numerical codes for all its ships (official numbers) that is superior for database identification purposes and the United States also did the same a decade later, so the current research applications of the signal codes lie in the extension of the system to European and other nations which did not implement numerical identifiers.

 

The Commercial Code of Signals, for the Use of All Nations; with the British Vocabulary, compiled by John T. Forster, Master R.N., published by William Mitchell, Shipping and Mercantile Gazette Office, London, 1857, is evidently the initial manual for the new British‑sponsored international code. A copy is held by the NZ National Maritime Museum, Auckland. However, it contains only operational signals, not the signal codes identifying individual ships. The Mercantile Navy List from 1857 appears to be the most readily accessible source today of British ship signal codes at this early date. Library catalogue entries at the National Maritime Museum, Greenwich, United Kingdom, for the “Commercial Code list for the use of ships at sea and for signal stations” published by the British Board of Trade in 1857, 1865, 1866, 1868 and 1869 may be the primary source for British ship codes in this period but I am unable to check. The same catalogue also includes “A Companion to the Commercial Code of Signals” published in 1857.

 

 

Extension to American and European ships

 

Individual ship identification codes using the Commercial Code of Signals were extended to American and European ships in the 1860-1864 editions of the [British] Mercantile Navy List. The Museum of Wellington City and Sea holds 1860, 1862 and 1864.

 

Codes were allocated to almost 10,000 American ships (refer American signals) and to several hundred European ships. The United States reallocated signal letter codes afresh to its ships along with official numbers when it revamped its maritime administration in 1867*. The 1860-1864 Mercantile Navy List lists of American codes are in alphabetic order of code but also have sequential numbers which could be used as a form of proto-official numbers for American ships if any list can be found (or created) matching up numbers and signal codes either side of 1867. There must have been some official American publication of these codes but I have no knowledge of its title or location (would anyone who knows please advise me at j_lowe@ihug.co.nz ). The 1864 Mercantile Navy List refers to the late arrival of revisions from New York thereby confirming that these American codes were allocated within the United States and not by Britain.

 

* The new signal codes appear in the List of Merchant Vessels of the United States from then on.

 

European ships were initially allocated only a small batch of codes all commencing with W--- with no single allocation of blocks of codes according to nation or state. I estimate that approximately 1,450 are listed in the 1864 Mercantile Navy List. Ships allocated these codes were registered in Austria, Belgium, Bremen, “Bunos Ayrn” (?), Denmark, France, Hamburg, Hanover, Italy, Mecklenburg, Naples, the Netherlands, Norway, Portugal, Prussia, Rostock, Russia, Sardinia, Spain and Sweden [the unifications of Italy and Germany were still incomplete]. There are national and regional blocks of codes within this European code list but several separate blocks for a particular state and some isolated single codes. It seems more likely that Britain allocated the codes than that the national states did, although this is unsupported inference. The Marryat code lists in use at this time do not indicate each ship’s nationality but it is extremely likely that more European ships had Marryat codes than Commercial Code of Signals codes in the early 1860’s and perhaps for some time thereafter.

 

The small initial allocation of codes for European ships could not have provided adequately for the European nations for long, if at all. It is evident that this allocation was soon superseded. I have seen passing reference to this aspect of the system being revamped by a British-French Commission in 1865 but I have no details of its decisions or anything it published whatsoever or any information about signal codes for European ships between then and 1886. Any relevant information would be appreciated and added to this manuscript to increase its usefulness. (See also the “Questions” heading later in the manuscript.)

 

The Commercial Code of Signals was renamed the International Code of Signals about 1870. The New International Code of Signals introduced some changes in 1901 though the signal codes of existing ships were unaffected. Further changes were made to the system in 1931 and 1969.

 

 

Signal code lists in Lloyd’s Register

 

Signal letters for individual ships are listed against individual ship entries in the body of Lloyd’s Register from 1874 onward but the Lloyd’s Universal Register* of 1886 includes comprehensive lists of signal codes, in alphabetic order of code, for nations that had adopted them. All were specific to a particular nation. Certain batches of letter combinations were used by most or all issuing nations, notably those commencing H---, with the larger nations eventually using several initial letters, generally later in the alphabet than H** (refer Letters). Clearly, they could only be used at sea in conjunction with a national flag.

* A separate publication issued in parallel with the standard Lloyd’s Register for the four years 1886 to 1889 inclusive, with expanded coverage similar to that of the standard Lloyd’s Register from 1890-91 onward.

** G--- seems to have been reserved for warships at this time although Italy’s warship codes commenced M---. Some nations later used G--- codes

for merchant vessels.

 

Signal code lists for European ships must have been published elsewhere but 1886 is the first comprehensive set of lists in signal order that I currently have available to me other than those in the Mercantile Navy List for 1857-1864. These lists appear to be generally included in the appendices to Lloyd’s Register from 1890 onward. However, individual volumes of Lloyd’s Register seem to have been bound in different ways for different subscribers so an individual volume that might have been expected to include the lists may not have them (my 1912-13 Volume 2 does not).

 

These national lists appear among the appendices to Lloyd’s Register until the1932-33 edition by which time the lists also distinguish between steam ships and motor vessels.

 

During the 1930’s a new system was introduced of allocating codes that avoided nations allocating the same codes, thereby making it possible to disregard nationality when using signal codes thereafter.

 

The individual ship signal codes continue to appear in the Register against the ship’s name after 1932 but not as national lists in volumes I have sighted.

 

 

Nations that issued signal codes for merchant ships

 

Lloyd’s Universal Register of 1886 contains apparently complete signal code lists for the British Empire, the United States, Austria, Belgium, Denmark, the Netherlands, France, Germany, Italy, Japan, Norway, Portugal, Spain and Sweden.

 

By 1913, Chile, Greece, Roumania and Russia had been added. Brazil allocated codes around 1890 but Brazilian lists appear in Lloyd’s Register for only a few years.

 

By 1932, Argentina, Estonia, Finland, Iceland, Latvia and Yugoslavia, most of them newly independent in the wake of World War I, were also issuing signal codes to their ships. The Russian list has been dropped and not replaced by one for the Soviet Union in this volume.

 

A quite long list of nations did not allocate signal codes to their ships in this period or, at least, not ones that are included in the national lists in Lloyd’s Register. None of these owned many ships, however.

 

The international system of codes introduced during the 1930’s provided for all nations and for aircraft as well as ships.

 

 

The content of the signal code lists

 

The inclusion of these signal code lists within volumes of Lloyd’s Register from 1890-91 to 1932-33 means that they are quite widely available, probably more so than is generally realised.

 

I recommend that anyone interested locate a register and examine the lists for themself. The following general observations based on the period to 1913 should assist. Please contact me if you have fuller (or contradictory) information on any aspect.

 

1. The signal code lists do not list sailing ships and steam ships separately as they are in signal code order and there was no discrimination in the issuing of codes. However, steam ships are indicated by a symbol. Later, motor vessels are also distinguished.

 

2. The signal code lists are only in signal code order and therefore impossible to search by name of ship except for those nations which appear to have allocated them in alphabetical order of ship name. If stored in datasets they could be sorted into name order.

 

3. There was a big drop in the number of ships with signal codes during the late 1880’s and early 1890’s, partly accounted for by the rapid decline in sailing ship numbers over this period and the substantial increase in the average size and speed of the steam ships that replaced them (which enabled fewer ships to carry an expanded international trade) and partly to changes in the proportion of ships allocated signal codes (refer to separate subheading later in item).

 

4. The signal code lists include a significant number of ships (generally small ones) that are not included in the main tables of the same volume of Lloyd’s Register. They are therefore a potential means of expanding the record of ships. Given the (presumed) superior coverage of European ships and the established greater coverage of smaller ships, the Répertoire Général issued by Bureau Veritas is likely to provide further details of many or all of these ships so using both series of registers in conjunction will provide superior coverage to using Lloyd’s Register alone.

 

5. The tonnage figures in the lists appear to be for net tons consistently, for steam as well as for sail. Minor typographic inconsistencies are not uncommon when comparing the tonnage figures in the signal code lists with the ship details elsewhere in the same register. Discrepancies of 1 are attributable to the fact that the system of ship measurement adopted by most maritime nations by the mid 1870’s calculates tonnage to two decimal places which are rounded off to whole tonnages, sometimes up and sometimes down.

 

6. The Lloyd’s Universal Register of 1886 contains only the signal code lists in the back of the volume. The signal codes are not included as a column or field in the main tables of detailed information about individual ships.

 

7. The standard Lloyd’s Register does include signal codes among the detailed individual ship records in its main tabulations from some date in the 1870’s onward. However, it appears to be quite common in the 1890-91 register that individual ship entries in the main part of the register do not include the signal code even though it is included in the ship’s entry in the relevant signal code list in the back of the volume indicating that the different tabulations were compiled independently. This appears to be an introductory problem that improves through time. However, there may continue to be instances where the code is given in one part of the Register and not the other.

 

8. Different nations followed different practices in the allocation of signal codes to ships. France evidently issued them in alphabetical order so that the French lists are in approximate alphabetical order of ship name as well as alphabetical order of signal code. On the other hand, there appears to be evidence of a regional pattern in the allocation of German codes. Other nations may have allocated codes in chronological order of registration. There is necessarily a degree of chronological allocation because some initial letters were adopted later than others.

 

9. The fact that the French signal code list for 1886 is in alphabetical order might mean that French codes were reissued in that year or shortly before. However, the French codes for 1913 are also in alphabetical order of ship’s name and show evidence of new ships whose names commence with A being allocated codes formerly allocated to other ships whose names commenced with A, and so on, a practice which could lead to the lists continuing to be in alphabetical order over an extended period without any general reissuing. I would welcome further information on French practice over the 1865-1913 period.

 

10. Investigation (following) of whether the codes of defunct ships were reallocated to later ships establishes that this was done, at least to a degree, by at least some of the principal European ship-owning nations. The fact that, generally speaking, the lists show many fewer ships with codes commencing H--- in 1913 for most if not all nations compared with 1886 and 1890 suggests that, while reallocation of codes did occur, a considerable gap in time was allowed to elapse before doing so. It does not appear that any particular signal code is likely to have been reallocated by any nation very many times in the period 1865-1913.

 

 

The number of allocated signal codes circa 1886

 

I have estimated the numbers of signal codes in the national code lists and compared them with indicators of the total number of ships of those nations in1886 and 1913 in the following tables. Those for 1886 are shown in the table below and for 1913 in the table following. Comparison is also made with the allocation of official numbers by leading maritime nations at the same date.

 

RG   = the Répertoire Général issued by Bureau Veritas. The criterion for Sail is 50 tons; for Steam 100.

LUR = Lloyd’s Universal Register. The criterion for inclusion is 100 tons.

National = number on national registers within the limits of the criteria specified.

 

Nation             Codes    RG      LUR    National        National criteria

           

Norway

5,513

4,088

3,651

7,664

Dec 1885

10 tons +

Germany

4,217

2,857

2,353

4,135

Dec 1885

varies from 17.65 tons

France

3,234

2,604

1,702

15,266

Dec 1885

2 tons +

Denmark

2,667

1,165

864

3,327

Dec 1885

4 tons +

Italy

2,621

2,934

1,827

7,336

Dec 1885

ns

Sweden

1,647

2,289

1,551

4,044

Dec 1884

10 tons +

Spain

1,409

1,808

1,056

2,063

Dec 1882

Ns

Netherlands

1,116

1,107

694

740

Dec 1885

40 tons +

Austria

526

569

490

485

Dec 1885

ns

Portugal

459

392

239

                ns

 

 

Belgium

57

87

72

65

Dec 1885

60 tons +

 

 

 

 

 

 

 

Total Europe

23,466

19,900

14,499

45,125

 

 

 

 

 

 

 

 

 

USA *

23,534

6,441

3,930

23,534

Jun 1886

5 tons +

Brit. Empire *

37,569

19,390

13,593

37,569

1886

15 tons +

Japan **

216

206

224

 

 

 

Other

       nil

5,155

3,162

 

 

 

 

 

 

 

 

 

 

Grand total

84,785

51,092

35,408

>106,228

 

 

 

 

* Total registrations are indicated for the United States and the British Empire as indicators of the number of their official numbers in use at the time, although the British figure includes vessels licensed for Canadian inland navigation.

** The estimated number of signal codes is indicated for Japan. Japan had allocated more than 1,000 official numbers by 1886 but I do not have the full details.

 

Sources:  Lloyd’s Universal Register of 1886, Report of the Commissioner of Navigation 1887. Number of signal codes allocated estimated by author.

 

 

For many research purposes the essential comparison is of the number of codes with the number of merchant ships recognised by Bureau Veritas and Lloyd’s Register as being of a significant size. These can be expected to include all foreign-going and even quite small coastal vessels. However, as the table shows there were also many very small coastal vessels not included in these code lists that would have been allocated official numbers if such systems had been adopted in their nations of registration.

 

The table shows that the number of signal codes in use by the ships of European nations by 1886

 

● was large in relation to the total number of merchant ships in the world recognised by Bureau Veritas and Lloyd’s Register as being of a significant size and not belonging to those jurisdictions that had issued numerical ship identifiers as well as signal codes.

 

● for four leading maritime nations the estimated number of signal codes exceeds the number of their ships recognised by the Répertoire Général as being of a significant size, indicating that these nations had allocated signal codes to many smaller ships as well. All except Belgium had issued signal codes to more ships than Lloyd’s Universal Register identified as being of 100 tons and above.

 

● signal codes were not allocated to many smaller ships that would have been allocated official numbers had they belonged to states that administered systems of numerical identifiers along the lines of those implemented by the United States, the British Empire and Japan at this time and subsequently introduced by Sweden. It is primarily in this respect that signal codes are inferior to comprehensively-allocated official numbers.

 

The evidence that signal codes had been allocated to the majority of larger European ships and to many smaller ones indicates that they are potentially useful as ship identifiers when no more comprehensive database identifier is available.

 

 

The number of allocated signal codes circa 1913

 

Estimates of the numbers of signal codes in the national lists for 1913 are compared with indicators of the total number of ships in the following table to show the situation at the outbreak of World War I. I do not currently have access to the range of national registration statistics shown for 1886.

 

RG   = the Répertoire Général issued by Bureau Veritas. The criterion for Sail is 50 tons; for Steam 100.

LR = Lloyd’s Register. The criterion for inclusion is 100 tons.

National = number on national registers within the limits of the criteria specified.

 

                                    Signal

Nation                         Codes              RG           LR              National                     

 

German

2,288

2,551

2,321

 

Norwegian

1,779

2,127

2,191

 

French

1,454

1,569

1,552

 

Italian

1,069

1,471

1,114

 

Russian

716

4,034

1,216

 

Danish

636

1,085

811

 

Dutch

561

870

759

 

Spanish

560

687

607

 

Austria

293

472

427

 

Greek

213

1,150

442

 

Portuguese

179

313

208

 

Belgian

144

148

172

 

Chilean

124

148

131

 

Roumanian

13

44

33

 

 

 

 

 

 

Total

10,029

16,669

11,984

 

 

 

 

 

 

USA

 

4,096

3,400

  27,070 registered 5 tons+

British Empire

 

11,539

11,287

  39,592 registered 15 tons+

Japanese

1,001

2,111

1,037

  2,506 registered of 100 tons+

Swedish

1,369

2,057

1,436

   2,822 registered of 20 tons +

Other

           nil

2,587

1,447

 

 

 

 

 

 

Grand total

 

39,059

30,591

 

 

Sources:  Lloyd’s Register of 1913-14, Report of the Commissioner of Navigation 1914 and 1917. Number of signal codes allocated estimated by author.

 

 

The estimated number of signal codes allocated in 1913 by nations that did not allocate official numbers is much less than the number allocated in 1886. The 10,000 codes allocated by these nations compare with more than 50,000 ships allocated official numbers at the outbreak of World War I. However, the number of codes is still sufficiently large to make a useful contribution towards ship identification in general, particularly for the leading European nations where the number of signal codes remains substantial compared with the numbers of ships identified by Bureau Veritas and Lloyd’s Register as being of significant size.

 

 

Allocated signal codes 1886 to 1932

 

The following table provides my estimates of the number of signal codes allocated by nations that did not allocate official numbers for several years over this period. Comparisons are made with the world total numbers of ships recorded by Lloyd’s Register and the Bureau Veritas Répertoire Général, exclusive of the four nations that allocated official numbers in this period. Japanese and Swedish figures are also provided.

 

The intention is to provide some indication over a longer period of the possible scope for signal codes to provide identifiers for the residual number of ships. It is not possible to make an exact comparison because signal codes were allocated to some ships too small to meet the criteria for inclusion in these two series of registers.

 

The comparison supports the conclusion that signal codes can make a substantial contribution towards documenting European ships although none of these totals includes all ships that would have been given numerical identifiers if these nations had issued official numbers along similar lines to those issued by the British Empire, the United States, Japan and Sweden.

 

 

 

1886

1890

1895

1903

1913

 1920*

1932

 

 

 

 

 

 

 

 

Argentina

-

-

-

-

-

-

73

Austria

526

333

273

234

293

-

-

Belgium

57

56

66

86

144

173

201

Brazil

-

504

-

-

-

-

-

Chile

-

169

174

         ?

124

82

71

Denmark

2,667

782

805

750

636

676

693

Estonia

-

-

-

-

-

-

141

Finland

-

-

-

-

-

284

331

France

3,234

1,411

1,183

1,229

1,454

1,341

1,774

Germany *

4,217

1,956

1,864

1,990

2,288

2,552

2,431

Greece

-

-

-

-

213

479

616

Iceland

-

-

-

-

-

38

72

Italy

2,621

1,502

1,220

1,185

1,069

766

1,370

Latvia

-

-

-

-

-

-

120

Netherlands

1,116

564

441

449

561

694

1,092

Norway

5,513

3,369

3,011

2,315

1,779

1,825

2,103

Portugal

459

198

170

176

179

96

263

Roumania

-

-

-

-

13

-

-

Russia

-

-

-

1,081

716

228

-

Spain

1,409

965

748

608

560

606

895

Yugoslavia

-

-

-

-

-

-

167

 

 

 

 

 

 

 

 

 

1886

1890

1895

1903

1913

1920*

1932

 

 

 

 

 

 

 

 

Total A

21,819

11,809

9,955

10,103

10,029

9,840

12,413

 

 

 

 

 

 

 

 

Total Rep. Gen.**

22,766

21,378

19,182

19,718

19,256

20,114

      na

 

 

 

 

 

 

 

 

Total Lloyd's **

16,110

15,196

13,761

12,323

13,508

12,935

14,506

 

 

 

 

 

 

 

 

Japan

216

231

375

1,546

1,001

1,834

1,845

Sweden

1,647

1,374

1,483

1,767

1,369

1,207

1,423

 

 

 

 

 

 

 

 

Total B

23,682

13,414

11,813

13,416

12,399

12,881

15,681

 

Total A is the total of the nations named in the upper part of the table, the estimated number of signal codes allocated by nations that did not allocate official numbers.

 

Total B is Total A plus Japan and Sweden, the estimated world total of signal codes allocated other than those of the British Empire and the United States.

 

*  The German figure for 1920 is for 1924-25 as German lists do not appear in Lloyd’s Register for 1920-21 to 1922-23 (1923-24 was unavailable to check).

 

** The World totals from the Répertoire Général and Lloyd’s Register minus the British Empire, the United States, Japan and Sweden. The Lloyd’s Universal Register was used for the 1886 Lloyd’s figure. The Répertoire Général covers sailing ships of 50 tons and above. Otherwise the criterion for inclusion is 100 tons.

 

 

The numbers of ships with signal codes are generally equivalent to around half or more of the number of European ships included in the Répertoire Général and to much larger proportions of the number included in Lloyd’s Register. There is enough comparability for it to be legitimate to conclude that “a very substantial proportion of European ships of significant size were allocated signal codes” in the period covered. However, it is not possible to derive valid exact percentages because some signal codes were allocated to ships that are not included in either register. To what degree the Répertoire Général coverage equals “Lloyd’s plus extras” or alternatively whether each register includes many ships in common with the other but also significant numbers not included in the other is an interesting and important question still awaiting systematic analysis as far as I am aware.

 

The table shows a major decrease in the number of European ships with signal codes from 1886 through to the mid 1890’s, particularly in the late 1880’s, that is not explainable by a comparable reduction in the number of foreign-going ships. There must have been some shift in the degree to which the European nations allocated signal codes to smaller ships at this time though possibly also a general purging of defunct ships from the 1886 lists. The possibility has to be seriously considered that the Lloyd’s 1886 code lists may have been seriously out of date at the time of publication (which in some respects would make them more rather than less useful). The very high number in 1886 relative to the two world benchmark figures is probably misleading in any case because of the inclusion of ships not covered by the benchmarks. It is more important that the lowest number of signal codes in any of the years surveyed is significant compared with the benchmarks.

 

The apparent shift in the allocation of signal codes in the 1880’s requires further examination in the interests of understanding how shipping was administered in this period but is unimportant for the more limited purpose of simply establishing in broad terms whether the overall coverage was large or small relative to recognised benchmarks.

 

 

The reallocation of signal codes

 

I have found no indication of a particular ship being reallocated a new signal code by any European nation during the period (1886-1913) that I have examined, though my examination was not exhaustive*. There may have been a general re-allocation in some nation at some time as the United States did so in 1867. Obviously, it would be highly desirable to clarify whether there was any general reallocation during the 20 years preceding 1886.

 

* I took a selection of sailing ships from the 1912 Lloyd’s register that were built before 1886 and checked the signal codes of those that had remained under the same jurisdiction (for the 1890-1912 period for reasons of practical convenience).

 

However, while my investigations so far seem to confirm that ships carried their signal codes unaltered for long periods provided they remained under the same jurisdiction there remains the possibility of codes issued originally to one ship being re-used for some later ship by the same nation, following the demise of the first. The issue is potentially serious because for various reasons many letters were not used in allocating the codes so the larger nations would eventually exhaust the available four-letter codes within the conventions in place at the time. As long as no two ships of the same nation had the same code at the same time and that old codes were not reallocated until revised manuals were in general use (ie after a lapse of time), there was no operational reason why the codes should not be reallocated. However, reallocation has implications for including ships from a long time period in the same database and using their signal codes as identifiers.

 

To check the possibility of reallocation I investigated batches of French and German signal codes used in both 1886 and 1913, selected from the signal code lists in Lloyd’s Register or the Lloyd’s Universal Register for these years without regard to size of ship or mode of propulsion. The name, whether or not a steamship and the tonnage were recorded for each year and the name, propulsion and tonnage details for each signal code compared. Where there was any discrepancy (and for a selection of others) the year and place of construction and original name were extracted from the ship details elsewhere in the registers.

 

Although long-lived ships carried the same signal code throughout, it quickly became apparent that both France and Germany did reuse the codes of some defunct ships for later ones.

 

For example:

 

● The French signal letters HBGL were used for the 184 ton brigantine Achille-Celestine built at Nantes in 1876 and for the 187 ton 3-masted schooner Achille built at Paimpol in 1911.

 

● The French signal letters HCLD were used for the 165 ton brig Aigrette built at St Malo in 1865 and for the 128 ton brigantine Aigle built at Cancale in 1897.

 

● The German signal letters HFDM were used for the 426 ton bark Soli Deo Gloria built at Danzig in 1863 and for the 521 ton steamer Martha built at Danzig in 1893.

 

● The German signal letters LBFS were used for the 135 ton schooner Delphin built at Kiel in 1866 and for the 205 ton steamer Prinz Sigismund built at Kiel in 1899.

 

 

However, no two ships carry the same national code at the same time as that would be completely self-defeating. Therefore, using the year of construction in conjunction with the signal code should ensure that no two ships allocated the same national code can ever be confused. It therefore appears that, with this proviso, the signal codes can be used as surrogates for official numbers for nations that did not issue them. This conclusion and its applications are elaborated in the separate discussion of signal codes as identifiers in databases (link at end of item).

 

For general research purposes it is not necessary to investigate further whether all nations reallocated the codes of defunct ships or whether any nation reallocated many. The fact that some nations did reallocate some codes means that the possibility must be provided for. The safeguard of using the year of construction in conjunction with the signal code should work for all nations.

 

I have not investigated the stability of European signal codes during the period 1913 to 1932 or after the system of nationally allocated codes was replaced, as having little relevance to sailing ship databases. I hope that someone involved in steamship databases will take up the question.

 

 

Questions outstanding

 

I would appreciate further information on the following to enable me to improve this record and interpretation.

 

  1. The decisions of the British and French Commission of 1865 that established the system for allocation of signal codes by European and other nations in the following decades.

 

  1. References for lists of signal letters allocated by any European nation in the period to World War I, especially for the period 1865-1886, information about their composition and form of publication, and possibly assistance with access to copies.

 

  1. The years in which the various European nations first issued national signal codes to ships.

 

  1. Additional information about how particular nations administered their codes, particularly the results of more substantial research into the degree to which early codes were reissued to later ships. Whether any nation other than the United States reallocated signal codes to all ships together at one or more points in time.

 

  1. More rigorous testing of my conclusions relating to the period up to World War I and testing of how applicable they are to the period between then and the introduction of LR/IMO numbers in the 1960’s.

 

  1. Given the big difference in the signal code lists in the Lloyd’s Universal Register of 1886 and Lloyd’s Register for 1890-91, it could well be productive to examine the signal code lists that may be in the volumes of Lloyd’s Universal Register for 1887, 1888 and 1889 for what light that may throw on the apparent change in coverage in this period. It will be many months before I can directly access any of these. However, information from European sources could probably clarify what happened more directly if contact with the necessary sources can be established. For example, if the lists in the 1886 LUR are seriously out of date, a national code list for any of the major European nations for the same year or even an earlier one might contain fewer ships which would immediately confirm the inference or alternatively reveal a genuine loss of small sailing ships not compensated for by the allocation of codes to new sailing ships and steamers to the same degree after 1886. It is even possible that ships in the 1886 lists continued to sail but were stripped of their allocated codes if they did not engage in foreign trade.

 

  1. Additions to the sources listed below.

 

 

Sources

 

My primary sources have been:

 

The [British] Mercantile Navy List 1857-1864.

 

Lloyd’s Universal Register for 1886 (probably also those for 1887-1889). Signal letters are documented only as national lists in an appendix. They are not indicated in the detailed records of ships in the main part of the register.

 

Lloyd’s Register from 1890-91 to 1932-33. Similar signal letter lists are included. Signal letters are also indicated in individual ship entries in the main part of the register but not necessarily for every ship named in the signal letter lists, which, evidently, were independently compiled and not fully integrated.

 

Various postings on the subject of signal letters in the Archives of the Maritime History Information Network (MarHst List) at http://www.marmus.ca/databases/

 

 

Appendix: Using the signal code lists to track the survival of ships

 

The signal code lists contained in Lloyd’s Register can be used as a convenient tool for tracking the survival of ships through time.

 

The signal code lists can provide a basic structure for tracking the careers of batches of a nation’s ships through time although the lists are unlikely to be absolutely up to date when issued, are probably as prone to typographic and other errors as any other similar compilation and they directly provide no more information about a ship than its nationality, signal code, current name, mode of propulsion and net tonnage. Other basic information such as year and place of construction and other names must be added from elsewhere. If other databases containing the particular nation’s signal codes already exist, such additions could be automated with savings in effort and accuracy.

 

The tabulated lists provide the requisite minimum information for constructing a table with a row for each signal code and ship and a column for selected years to record when that signal code was in use by that ship, by - in effect - combining several signal code lists in succession. Such a table could provide an economical structure for tracking substantial batches of ships through time, perhaps at five year intervals initially, with fine-tuning and elaboration as required. That could provide only partial coverage as it would cover only ships issued signal codes but it would provide an effective way to track larger ships and those engaged in foreign trade (which tend to be one and the same but not absolutely so). Such a tabulation could provide a general indication of approximately when ships changed nationality or were lost or scrapped, and a starting point for narrowing down the precise years. It could also provide a basis for tracking the years of name changes. It could also aid checking information by cross-matching. Computerised, it could be sorted into ship name order and restored to nation and code letter order as required.

 

In the process, one would get a much better idea of to what degree and how rapidly particular nations recycled earlier signal codes. Much could also be deduced about national allocation of signal codes through the addition of information about place and date of construction, initial port of registration and so on although databases may already exist enabling such information to be compiled in the same format more directly.

 

Developed into a fuller analysis of the career spans of batches of ships built in particular countries in particular years (or relatively short periods of time such as five year periods) such tabulations could facilitate the maritime equivalent of  “cohort analysis” in demography (human population studies) to study longevity, with transfer of ownership between nations equivalent to international migration in demographic calculations. There are limits to the analogy - ships can be “born” and “die” twice among other ways they differ from people and the division into powered and unpowered vessels does not equate to gender in demography - but apart from some adaptation the principles and mathematics are the same. The necessary information can be compiled in alternative ways from various alternative sources (some of which are more comprehensive and reliable) but the signal code lists constitute an available convenient basis for “quick and dirty” indicative research of this type and a structure that could provide a starting point to be elaborated for more refined analysis.

Keep scrolling down for links

 

Please direct comments and information to j_lowe@ihug.co.nz

 

 

To go to the discussion of the utilisation of signal codes as database identifiers click here

 

To go to the discussion of Marryat signal codes click here

 

To go to the discussion of official numbers as database identifiers click here

 

 

To return to the main maritime menu click here