Cracking the code of mid-19th century Merchant Navy seaman service records

The Merchant Navy Port Rotation Numbers project is a collaboration between Pete and Jan Owens of the Crew List Index Project (CLIP) and Peter Hamersley, whose ideas, initiative and work were the springboard for the project, and the kind assistance of Bruno Pappalardo, Principal Records Specialist, at The National Archives. 

The project is aimed at finding an important missing link in Merchant Navy service records for the period 1845 to 1855 – the port rotation numbers, which link Merchant Navy ships’ crew lists to the service registers of Merchant Navy seamen. It has solved a long-standing problem with these registers: how to link the codes used in the registers to the crew documents they refer to. When the data is complete, it will facilitate linking from over 500,000 register entries to the crew agreements.

We now have a good understanding of how the port rotation numbers were allocated, enabling us to re-create the missing link. The port rotation number stayed with the ship through the years unless she was re‑registered. The important insight is that the allocation of the numbers was based on the registration of shipping.

To make this operational, the project is using a two-pronged attack. We are extracting port rotation number data from a sample of Merchant Navy crew lists in the National Archives series BT 98, then using data from the registration of shipping to check it and fill in gaps. It is a slow job, but we are gradually publishing it online as we finish the data for each port. Researchers can now enter a port rotation number and find the name of the ship it refers to and the possible places to search for it in BT 98.

What are port rotation numbers?

From 1835 onwards, central registers were kept of the service of Merchant Navy seamen, showing the ships they worked on and in what capacity. These registers are now held at The National Archives in various series of records, for example BT 112, BT 113, BT 115.

The registers were made up from the crew lists and agreements which are now held in BT 98. Masters of ships were legally required to complete these documents and they were returned to London at the end of each voyage or half‑yearly.

So this is our picture of the process of dealing with the crew documents:

In London, the registration returns from each port were used to compile a list of ships for that port and each new registration was allocated a number – the port rotation number. That number remained with the ship unless it was re-registered or no longer afloat. Unfortunately, this list of port rotation numbers has never been found and probably no longer exists.

When crew documents were returned from the ports to the officials in London, the clerks looked up the ship on the list for that port and scrawled a numerical code on the document, showing the port and the ship’s port rotation number, for example 86.104 – 86 was the code for the ship and 104 for the port (Whitby in this case).

The clerks who compiled the registers of seafarers worked through the pile of crew documents, identifying each crew member in turn and adding the information to the seafarer’s record, using the code written on the document. A long slow job!

Here’s an example:

The image below shows an Account of Crew for a Whitby ship, with the port number and port rotation number annotated by the clerks using the ships list for Whitby. The code is 86.104.1 in this case. 

One of the apprentices on board was John Breckon.

The clerks who compiled the registers of seafarers added the code from the crew lists to John Breckon’s record, as shown below.

What’s the problem?

The problem is that without the list of port rotation numbers, there is no way to link the numerical code used in the service registers to the ship’s name. Without that, it is difficult to find the crew documents which are stored in alphabetical order for each year and port in BT 98 and to make sense of an individual’s Merchant Navy seaman’s service record.

Researchers using the registers such as BT 113 find the codes against a seafarer’s record, but then run into snags. The port numbers (such as 104 for Whitby) are known but, without knowing the name of the ship, it would mean searching through all the boxes of documents in BT 98 for that port and year. That could be a daunting task. For example, the 1851 lists for Whitby are in 10 boxes, each with up to a hundred lists for many different ships. You would need to read each of those documents to look for the seafarer.

The missing link is the list of port rotation numbers. With that, researchers could work from the code to the name of the ship. It then becomes a case of searching just one box of documents looking for the ship, which should include the seafarer.

Finding the missing link to the port rotation numbers

The key to the port rotation numbers was suggested by Peter Hamersley. His idea was that the port rotation number was allocated to a particular ship and remained so from one year to the next. Working from films of BT 98, he had extracted a sample of port rotation numbers.

We examined the data, looking for a pattern. We compared it against the returns of shipping registrations which the ports of registry sent to London and which are now held at The National Archives in BT 162/19 (and available online).

Peter was right. The port rotation numbers were indeed allocated to individual ships and stayed with them. There was also a further significant insight. The sequence of port rotation numbers went in step with the shipping registrations. The numbers were allocated in sequence as each new registration was recorded and entered on the list for that port. If the ship was re-registered at the same port or transferred to another, it gained a new port rotation number. So the numbers referred to a particular registration at that port.

That allowed us to check and enhance the data and, better still, we could deduce many missing port rotation numbers by checking back to the sequence from the shipping registers.

So we have found the missing link. The next task is to follow it through with data.

Building the port rotation number list

We are now building the list of port rotation numbers, using a two-pronged attack.

We are extracting the port rotation numbers from a sample of the crew documents in BT 98. We have chosen all the documents for 1851, some 445 pieces (boxes). Each piece contains a hundred or so documents, providing tens of port rotation numbers.  It takes an hour or so to go through one box, noting the ship’s name, registration details, tonnage and the all-important port rotation number.

Peter Hamersley is co-ordinating the 1851 Port Rotation Number Indexing Project, with a small team of volunteers working through the films on FamilySearch at LDS Affiliated Libraries. 

The second prong of the attack is compiling a list of the ship registrations for each port. Jan and Pete Owens of CLIP are doing this using the images of BT 162 and BT 111 which are available for download from The National Archives. BT 162/19 contains the returns of shipping for each port (except London) of the ships extant in 1850. BT 111 contains a complete index of the shipping registrations from 1832 to 1856. We have completed the London entries in BT 111 and are now working through the sections which cover all the other ports.

Combining the two data sets produces the list of port rotation numbers for each port.

The register data gives a cross-check of the BT 98 data and often additional information, such as re-registries. 

The great bonus is that port rotation numbers can be interpolated. This is how it works:

The left hand column is data from BT 98/2694. Port rotation numbers (PR#) have been found for the ‘Delaval’ and the ‘Lucy Jane’. This has been correlated with the BT 162 data on the right. The matches to BT 162 are shown in green. 

There is no port rotation number data for the ‘Brothers Success’ or the ‘Jane’, but their port rotation numbers can be inferred because their register numbers (Reg#) run in sequence with the ‘Delaval’ and ‘Lucy Jane’. The missing numbers are added in and the Source Year is shown as ‘Extra’.

This doesn’t always work – there are occasional glitches in the sequences, but we are usually able to add about 50% to the original data.

In this way, the relatively small 1851 sample can produce data covering a substantial proportion of the port rotation numbers for each port. For example, the port rotation numbers for Whitby go up to 788 and we have recorded or deduced 466 (59%) of them. Once we have more data from BT 111, we should be able to increase that percentage. For London, because we already have virtually complete registration details, we should be able to deduce more than 90% of the port rotation numbers.   

Publishing the data

To publish the data, we have set up a search page as part of the CLIP web site. It enables users to enter the port number and port rotation number and returns the name of the ship with a list of the BT 98 pieces which are likely to hold crew documents for that ship. We are steadily adding data as each port is completed.

The search page is here: https://crewlist.org.uk/data/vesselsPR

For the first time, there is now a way of working easily from the seafarers’ records back to the crew documents they came from.

The data from BT 162 and BT 111 is also useful in its own right and we are publishing that as part of our ‘Ships by port’ data on the CLIP site. For example, this is the BT 162 data for Whitby.

Eventually, we plan to add a facility to search for a pre-1855 ship by name to complement our existing pages for post-1855 ships.

Technical details

We have made a careful investigation of the data we have collected to see how well it bears out the insights and understanding we have gained.

Our conclusion is that the initial ideas are correct. There are discrepancies in the data; it would be surprising if there were not. However, there are plausible explanations for why they may have come about, and this actually serves to increase our confidence. No doubt more quirks will emerge as we go on. A detailed account of the data we have and our assessment of how well it bears out the hypothesis is online here.

Port Rotation numbers project progress

If you would like to know more about the Port Rotation numbers project, check its progress or indeed support it, please contact crewlist.org.uk/about/contact.