Merchant Navy codes

There was considerable speculation after the capture of the Rangitane that the Germans knew exactly where to find the ship. The 1941 official enquiry into the sinking of the Rangitane and other ships was called as a result of press and public indignation about the whole affair. Captains Upton and Miller were adamant that the Germans had known not only of the Holmwood and Rangitane but also knew the movements of several other ships. For obvious reasons, the official inquiry was rather reluctant to discuss the finer points of secret codes and, in effect, dismissed the idea that cyphers had been compromised. It has been alleged that the Captain Miller did not dispose the Holmwood radio code books. These were described by one source as the BAMS cypher - the Broadcast to Allied Merchant Shipping (or was it the British and Allied Merchant Shipping?) codes used at the outbreak of war. Such code books were always stored in a weighted bag with holes so that it could be thrown overboard in an emergency. It is not known if a coastal steamer as small as the 247 ton Holmwood would have carried such codes or whether Captain Miller as the only person on board experienced in using the radio would have been trained in message encoding and decoding.

Before including the allegations of compromised BAMS codes in the story of the Rangitane, I want to establish exactly what BAMS was, whether it really was in use at the time and whether it was a major factor in the loss of seven ships in a very short period of time.

I have already received the following advice from Frode Weierud, a respected authority on cryptology:

If you read German you should have a good look at Heinz Bonatz's book "Seekrieg im Aether: Die Leistungen der Marine-Funkaufklaerung 1939-1945" (Naval Warfare in the Aether: The Performance of Naval Signal Intelligence 1939-1945), Verlag E.S. Mittler & Son GmbH, Herford, 1981, ISBN 3 8132 0120 1. His book which is based on notes and documents from the German Naval Signal Intelligence organization, B-Dienst, is probably the best account ever written about their achievements.

He starts the book with an overview of the code and cipher systems that the B-Dienst knew about and those that they successfully broke. On page 39 he deals with the merchant ship codes. The main system was the Merchant Navy Code (MNC) which the Germans called "Gallien". It was a 4 letter code which was super-enciphered with monthly changing bigram tables. The other system used by the merchant ships was Mersig, Merchantship Signals, which had subdivision called Amsigs and Camsigs for American and Canadian Merchantship Signals. This was 4 number codes with super encipherment. BAMS, British and Allied Merchant Shipping, belonged to the Mersig group, but was first introduced in 1943. This shows that Rangitane could not have been using BAMS but rather MNC. This is also mentioned by Bonatz.

Bonatz says that they would be able to break into MNC 10 days after each monthly change of the bigram tables. Hence from about the 10^th of each month until the end of the month they would be able to read every MNC signal. They captured the MNC codebook in May 1940 in Bergen in Norway. F.H. Hinsley et. al. "British Intelligence in the Second World War", Volume 2, has an appendix, Appendix I on page 631-642, about British Cypher Security where they give details about the British and Allied codes and ciphers and their compromise. Here is what he has to say about the MNC: "The cryptographic systems provided for communication with the merchant ships were a prolific source of intelligence to the B- Dienst second only to the Naval Cypher No 3 in their importance to the Battle of the Atlantic. The merchant Navy Code replaced the International Code and Naval Appendix in January 1940. Like the earlier system it was used with a simple re-coding system so that by march the B-Dienst was having some success, and its work was greatly helped by the capture of copies of the Merchant Navy Code at Bergen in May 1940, after which it was able to read the bulk of the traffic with very little delay. Although this code was re-coded by a variety of tables (some for general use, some for independently routed ships, others for the Commodores of convoys) all, other than the Commodores' table, which was little used, were readily broken by the B-Dienst." There are more details which I will not mention here. "British Intelligence in the Second World War", Volume 2, was published by HMSO in 1981, ISBN 0 11 630934 2. You should be able to find a copy in any decent British library.

On page 188 to 203 Bonatz tells bout the raiders, Atlantis, Orion, Widder, Thor, Pinguin, Komet, Kormoran,Michel and Stier. Their official German name was Schiff (Ship) 16, 36, 21, 10, 33, 45, 41, 28 and 23 respectively. Hence Orion was Schiff 36 and Komet Schiff 45. The Orion was further known as HSK I (Handelsschutzkreuzer One) while Komet was HSK VII. Handelsschutzkreuzer means Merchant Ship Protection Cruiser, and all the raiders were placed in the category Hilfzkreuzer (Auxiliary Cruiser). Further details are that Orion was the former ship Kurmark of 7021 BRT and a speed of 14.5 kn, while Komet was the former Ems of 3287 BRT also of 14.5 kn. A crucial information is that these ships carried Bord-B-Dienst, an on-board signal intelligence unit which would intercept and decipher enemy messages.

This information from Frode Weierud is important. First, it tells us that BAMS was not in existence in November 1940. Second, it tells us that the Germans already had the capability to break Merchant Navy Code encrypted messages after about the 10^th day of each month. What remains in doubt is whether this was any benefit to raider ships in the south Pacific area who, we know, were out of contact with their base for months at a time. We now know that the Orion and Komet both had their own signals intelligence unit on board and we can assume that, given a decrypt, they were capable of decoding messages. However, if the MNC was broken by B-Dienst in Germany after the 10^th of the month, how was the decrypt sent to isolated raiders in time for action to be taken before the code changed at the end of the month?

2. I would say that the Holmwood would have been carrying codebooks. However, having read the exploits of the German raiders, it would appear that the Orion "circumnavigated" the NZ coast - laying mines in the Hauraki Gulf (Auckland - East Coast) - which sank the Niagara (June 1940). Then Orion headed north, rounded the top of the North Island and came down the West Coast - engaged the Turakina (August 1940). Orion returned with the Komet and a supply ship. Off the East Coast of the South Island they intercepted the Holmwood (Chatham Islands supply ship). The Holmwood surrendered in order to save the lives of the passengers and crew. It is not known whether or not the Germans captured the codebooks. The Germans sank the Holmwood 25 November 1940.

The Rangitane sailed from Auckland 25 November 1940 enroute to the UK and judging by the position where she was sunk 27 November 1940 - 36 degrees 48 minutes South 175 degrees 7 minutes West - meant that she was heading in an Easterly direction with the Germans heading approximately North to Northeast. Further reading suggests this may have been a chance encounter.

The following is a transcription of part of page 466 from "The Codebreakers" by David Kahn:

With the exception of the Irish stations at Valentia and Malin Head, through which Admiralty messages were not passed, all coast stations were provided with the official code books known as MERSIGS. .......(Date not specified -late 30's or 1940?).......The Admiralty took control of the Rugby (GBR) long wave (16 kHz) transmitter and three of the Leafield short wave transmitters for 12 hours daily, these being keyed from Whitehall for working Royal Naval ships. In 1943 a Naval officer and 18 Naval telegraphists from HMS Flowerdown were attached to the Portishead/ Burnham station to cope with the increasing work-load. Many of the peacetime staff had been seconded to other Government services. Naval telegraphists and Post Office operators worked in pairs searching for calls, each handling their own particular service's transmissions. With the entry of the United States into the war, American ships were included in the service. A special aircraft section was set up for communication with the RAF's Coastal Command patrols in the North Atlantic.

Messages for merchant ships were broadcast in the MERSIGS code every four hours using the Rugby transmitter on 16 kHz and three short wave transmitters in the shipping bands 4-27.5 MHz. These were known as the "B.A.M.S." transmissions (Broadcast to Allied Merchant Ships). A similar service was provided on the short wave bands from Naval stations such as Simonstown in South Africa, Colombo in Ceylon and Sydney in Australia. Each station covered a number of the established sea areas (B.A.M.S. areas), but the GBR transmitter was generally readable in most areas with the exception of parts of the Pacific Ocean.

The broadcasts were preceded by a Traffic List of ship stations for which messages were on hand, each call sign being given twice. The ship's radio officer then had to monitor the band until his ship's turn came. The broadcasts were transmitted on two or three successive transmissions. No acknowledgement of receipt being possible, reception had to be assumed. Eventually all traffic for Allied vessels was transmitted by Rugby which relieved the burden on the local coast stations. Towards the closing stages of the war some 10,000 words weekly were being transmitted. For the radio officer on board ship, receipt of a message had to be followed by a double decyphering process to arrive at the plain language message. Similarly on the few occasions when it was necessary for a vessel to communicate with the shore, other than for distress or attack by enemy action, the plain language message had to be coded and recoded.

Because ships' international call signs were known throughout the world and Traffic Lists sent out by the UK land stations could be monitored, in due course ships were issued with secret call signs known only to the Admiralty and the ships concerned. This prevented the knowledge that a particular vessel was currently in a specific sea-area, of special importance for the larger well-known passenger liners.

Life for all merchant seamen during wartime was unpleasant and dangerous. For the radio officers it involved listening continuously on the distress frequency of 500 kHz except when monitoring the four hourly broadcasts from Rugby. When in convoy and off watch they were often called upon to assist with flag and lamp signalling. The Admiralty ran special visual signalling courses for them. Safety regulations required portholes to be battened down at night, and access doors curtained off. Radio rooms, generally small, with receivers running 24 hours a day became hot houses, particularly in the tropics. With earphones clamped to the head, and very often the continuous cackle of atmospheric interference, there was a great temptation to "drop off". Yet it was vital to remain wide awake and to be prepared to spring into action in the event of an attack.

The distress wave of 500 kHz was monitored continuously by the enemy. As soon as a key was touched, there would be an immediate response of gunfire from a surfaced enemy vessel. aimed at the radio room. It was the duty of the radio officer to stay at his post to ensure a distress call was transmitted, even though the majority of the ship's crew had taken to the boats. Many died in the radio room or failed to make it to the boats.