The Rogers' American Signal Codes 1846 – 1854

and their ship database development potential

(added 12 May 2006)

Introduction

The Rogers' American Signal Codes for 1846-1854 are potentially the basis for a useful database of American vessels of 150 tons and above that could be combined with other information to extend backwards the coverage of the List of Merchant Vessels of the United States for this size range. While ultimately it would be nice to have a database containing all American vessels pre-1867 this would be a far larger undertaking. Most foreign-going vessels at least will lie in this size-range. One could simply transcribe everything one could find for vessels in the period but the Rogers list and its 1859-64 counterpart provide a structure and a reference point against which to assess coverage that an indiscriminate approach would not.

Carrying back a comprehensive record of at least foreign-going and larger American ships from 1867 to around 1855 would facilitate international comparisons, 1855 being the beginning of the comprehensive coverage of British ships allocated official numbers from 1855 onward and their publication in the Mercantile Navy List from 1857.

The full citation is Henry J. Rogers, Rogers’ Marine Telegraphic List of Merchant Vessels of 150 tons and upward, Employed in the Commerce of the United States; Furnished for the American Code of Signals, By Direction of the Hon. James Guthrie, Secretary of the Treasury of the United States, E. & G. W. Blunt, New York 1855.

Other publications by Rogers around this date appear to relate only to operational signals but there may be a later 1856 edition which may add vessels first registered in 1855. If you know of such please let me know at the earliest opportunity.

What is in the Roger’s listing

The Rogers' 1846-1854 listing provides name, tonnage, ship type and port (s) of registration. The ports are related to Customs District and State in a useful manner. The signal codes themselves are of little research value as they differ from the signal codes of 1859 onward and the signal code series of 1867-1932. They therefore cannot be used to link together ship registrations of the 1846-54, 1859-1864 and 1867-1932 periods.

The codes themselves are of no utility for database development (unless, as discussed following, their ordering reflects registration years, as it may). The particular research utility lies in the comprehensive listing of all vessels above 150 tons at this time which it provides. For reasons discussed later, some vessels are listed two or more times (and, quite unnecessarily, allocated multiple signal codes). The duplications need to be identified but have potential uses in identifying multiple registrations in the period (see following).

The Rogers code has the advantage over the 1859-1864 listing in that it has a fixed lower size threshold and covers the inland rivers’ steamers and the Great Lakes as well as foreign-going. The 1859-1864 listing includes some smaller vessels but relates to essentially foreign-going alone and therefore lacks Lakes’ and inland river vessels.

The Rogers’ flags

Most signal codes both before and after Rogers identified ships with numerical or alphabetical codes consisting of up to four flags to be used in conjunction with the national flag.

The Rogers codes identified ships by means of a single flag upon which all letters of a code consisting of two or three letters appeared in 8,400 combinations. To extend the scope, three different background and letter colour combinations were used, black letters on white background, white letters on blue background and white letters on a red background. If you encounter paintings of ships with flags of this type they are very probably Rogers’ signals. Only about half the first two batches were allocated to ships in order to provide for additions. White letters on a red background were allocated only to vessels registered in San Francisco. A few other Californian ships were allocated codes in the white on blue range. Otherwise, white on red indicates Pacific Coast, and San Francisco in any case.

The fact that all other flag-based identification systems of which I am aware used multiple flags to identify ships may indicate that Rogers’ “bright idea” for simplification and cost saving did not work so well in practice.

A further factor may have been that Rogers indiscriminately allocated letter combinations with results that that later authorities avoided and which would not be allocated by any authority today. It is fortuitous (rather than policy) that FAT was not allocated to a specific ship. Actual ships were required to identify themselves as, COW, FUN, RAT, ROT, SIC, SLO, SOW and SUE. Some of these were allocated to more than one ship. “Little boys of all ages” will find a number of worse potential offence even within the limited scope offered by three letters rather than four. The later Commercial Code of Signals that became the International Code of Signals avoided vowels altogether for precisely this reason although vowels are used, with caution, in call signs today.

On a whimsical note, had the bark L. Sabine of Passamaquoddy, the schooner Kate Holbrook of Waldoboro, the bark Splendid of Wiscasset and the ship Africa of Bath sailed past or anchored in that order their Rogers’ signal flags would have read FE, FIE, FOE, FUM. Whether children of the time were likely to have been sufficiently good at reading to recognise “Jack and the Beanstalk” I could not say.

Database development based on Rogers

To be more useful as a listing of ships registered at the time the content of the Rogers listing needs the addition, at very least, of when and where built and, ideally also eventual fate.

Given any necessary copyright clearance, two readily available sources can provide when and where built that can be fairly readily added to the Rogers listing at least for larger vessels where there are no complications of vessels of similar name and tonnage which could be confused.

These are:

1. the Lytle-Holdcamper List which covers all US steamers to 1867 including their fates

2. The (different) Holdcamper List readily provides when and where built for New York registrations for ships of all types, also to 1867.

A variety of sources can fairly readily yield the additional information for many vessels, particularly the larger and more famous. Complete coverage would require the use of one or more of Works Programme Administration transcriptions, direct use of original port records and possibly other or unpublished port record transcriptions unknown to me.

Fleshing out Roger's 1846-1854 listing would provide a useful analysable record in itself and also assist and interconnect with doing something similar around the 1859-1863 listings from the Commercial Code of Signals and for the initial years of the List of Merchant Vessels of the United States (LMVUS) which are also sparse in their level of detail. Above a quite low size criterion, these projects, in combination, could extend the LMVUS coverage back for one to two decades. That would both serve US
maritime historical research and facilitate comparison with British records of the period. The 1842 Rogers' List could take it back further still and would also contain smaller vessels of which perhaps many would have survived into the 1850's.

Completing any one project rigorously could easily require a couple of thousand hours work but individuals can work on a local or other part of a subproject knowing where it fits into a larger whole. Team efforts have shifted many a bigger mountain as any genealogist can attest. I have several major prior (including contractual) commitments so could not make any of these possible projects a major priority but could be of some assistance to any person or group interested in working on some part of US
ship registrations over 1846 to 1868.

Vessels covered

Including vessel duplications, the 1846-1854 listing contains around 2,000 steamers (all covered in Lytle-Holdcamper), 2,100 full-rigged ships, 1,800 barks, 1,800 brigs and brigantines and 1,300 schooners, around 9,000 vessel names altogether. Some 1,800 out of the 9,000 should be covered by Holdcamper, of which 360 are steamers and therefore also in Lytle-Holdcamper. Holdcamper and Lytle-Holdcamper between them potentially cover a little over one-third of the total.

If the 1859 listing is any indication, the Rogers List could represent around one-third of current US vessel registrations, half the total tonnage and probably most of those engaged in foreign trade.

Exactly which ships does Rogers list and what is the exact significance of the duplications?

Enough of the "Da Vinci Code" already!

Henry J. Rogers (1811-1879) may lack the antiquity, mystery and conspiracy elements but "Rogers' marine telegraphic list of merchant vessels ." 1855 offers maritime amateur code breakers meaningful challenges to decode the structure of American ship registrations from 1846 to 1854 - or at least the Rogers List itself.

Challenges lie in:

1. the exact meaning of Rogers' statement that he lists "all vessels of 150 tons and upward employed in the merchant service of the United  States, between the years 1846 and 1854" and

2. the implications of what appear to be subsidiary alphabetical sequences within primary sequences of name initials. (The signal letters themselves are unambiguous and immaterial.)

The Preface wording is ambiguous but the most reasonable and logical inference - subject to confirmation - is that Rogers included all surviving vessels above 150 tons registered in 1846 plus those registered subsequently. Occasional blanks suggest the removal of vessels known to have been lost during the period (in which case the list would not contain "all vessels … employed ... between ...1846 and 1854") but at least some vessels known to have been lost before and during 1854 are included.

The structure seems to be driven by registration records because there are quite a few cases of the same vessel (down to 1/95th of a ton) being included more than once, occasionally as many as three or four times, each manifestation being (pointlessly) allocated unique identifying letters. These cases involve not only the same ship's registrations at more than one port but even multiple registrations at the same port.

Preliminary analysis suggests that possibly as many as ten percent or more of all listings are duplications. The Preface wording implies this was deliberate.

If all that was intended was to compile one record of  "each vessel afloat as we go to press" that was all he needed to say. That was what the efficient achievement of the object of the exercise required. 1846 - or even 1853 - did not need to come into it if they were capable of producing a reliable 1854 current list. That the local registration and Federal authorities together could not do so at that time is an interesting revelation in itself. They could (apparently) do so in 1859.

The outcome seems explainable only in terms of all registrations in each of the nine years for each port being included even though allocation of an identifier for multiple manifestations of a ship was unnecessary, confusing and wasteful as only one could be current. The duplications preclude using Rogers' list as a sampling frame or for cross-sectional analysis until all multiple manifestations are identified. However, the implication that the Rogers' listing reflects the complete registration structure for the whole period holds out positive compensations. What is lost as the key to a point in time could be gained for a period of almost a decade.

An apparent structure of secondary within primary alphabetic name sequences may be fortuitous but may reflect a structure of registration years that would have practical applications. Within each port and ship type sequence, ships' names are listed under initial letter of name (A, B, C.Z) but there are subsidiary, apparently non-random, alphabetic sequences within these blocks. Thus within C, you get apparently repeated subsidiary sequences such as Ca., Ce., Ch., Ci., Co., Cu.. Alphabetical order for successive years within each initial letter would generate such a pattern. If there actually is an underlying order, cracking the code could possibly add layers of meaning to the whole Rogers listing for comparatively little effort and some intellectual stimulation or even  fun into the bargain.

An attempt to crack the code for a block of New York ships with Holdcamper's record of initial New York registration years was inconclusive but the ordering might have a logic to it if compared with all their New York registrations of the period. An alphabetic sequence for a year might be thrown out of order by a ship being sold and reregistered during a year. Alternatively, the pattern might reflect batches of records for sub-offices in a large port, blocks of ownerships or result from an incomplete effort to remove multiple registrations at the same port. (Or it may just be an incompetent mess.)

19th century clerical practices I have seen in unrelated records are likely to have been universal (as following logically from all you could do with a quill pen and huge ledgers) and therefore may be relevant. These practices index pages of ledgers in sequential order under initial letter of the alphabet. If the Rogers pattern results from this then there should be a registration year structure there to be found. Sequences could vary according to the idiosyncracies of individual chief clerks and therefore between ports accordingly.

I am unable to carry this analysis to final conclusion as the sources directly available to me cover only some (and only initial) registrations.

Anyone interested in American merchant shipping of this period can follow this up for a port or ports of particular interest to them using sources available only in the USA as far as I know. I can offer a batch of New York names commencing A-C to experiment with in National Archives, Washington DC. Checking 50-100 records would suffice to test the hypotheses and yield further information about those ships in the process. I could do the same for another port for someone with access to a Works
Programme Administration transcription or actual port registrations.

The solution is of more than abstract or theoretical interest. It is comparatively easy to add fuller details for the larger vessels to Rogers but for smaller vessels of the same name and similar tonnage with the additional complication of typographic errors, any structure that ties vessels to other records and points in time has potential utilitarian value.

To return to main maritime menu click here