by Tony Insall
Norwegian dissatisfaction grew to such an extent that a letter of complaint, listing shortcomings in training and equipment, was sent to Oscar Torp, the Minister of Defence. This led to some extensive discussion. At a meeting between Cordeaux and Roscher Lund on 9 October 1942 to discuss the future organisation of SIS in Norway, a comprehensive range of changes were agreed. The syllabus was significantly extended, and training responsibilities were fairly evenly divided between British and Norwegian staff. Codes were improved too, and some staff were posted elsewhere.16 Thereafter the Norwegians were more fully involved in policy regarding training matters, and the quality of training improved greatly. From an early stage, SIS agents were trained to operate as discreetly as possible, and not to become involved in other activities such as sabotage or propaganda. They were instructed not to fight unless in self-defence where silent escape was impossible.
SIS coast-watching operations in Norway
The SIS controlling officer Eric Welsh was keen on establishing what came to be called ‘hermit stations’ in isolated places on the coast, where SIS agents (usually sent in pairs) would be expected to survive on their own, often for considerable periods of time. They would generally be living in extremely difficult conditions and sheltering from the weather in goat caves or tents or, if they were lucky, in a primitive cabin. Ole Snefjellå, one of SIS’s most successful coast-watching agents, worked for a total of eighteen months on three different hermit stations, Pi (near Stavenes) in 1942, Crux (near Mo i Rana) in 1943 and Frey (south of Ålesund) in 1944. He calculated that a station which was expected to operate self-sufficiently for six months would need almost two tons of supplies consisting of food, equipment and a small power plant with fuel to provide electricity for the radio. So the delivery of agents – by whatever means – and the establishment of their stations, was rarely a straightforward matter. Snefjellå provides a vivid description of how he and his fellow agents lived in caves, or primitive dugouts made of turf and stones. These needed to be buried as far as possible below ground level to facilitate camouflage protection, which meant that they were generally very damp. Food quickly became mouldy and deteriorated.
Keeping track of German naval movements. A coast-watcher’s vantage point. © NHM
Snefjellå also described the respect in which Welsh was held by SIS’s Norwegian agents, noting that while ‘he required the agents to endure hazards and if need be, severe conditions’, he also ‘did his utmost to save agents who were being chased by the Germans or who were in severe difficulties’.17 (One of Welsh’s favourite catchphrases was ‘little, but good’.) Snefjellå had particular reason to be grateful to Welsh for his humanity on these grounds. He was first deployed to Pi in April 1942, and his station provided extremely valuable intelligence, for example being one of the four SIS stations to report the southbound journey of Prinz Eugen on 17 May, and the northbound journey of Lützow two days later. But, shortly afterwards, the Germans began to make arrests locally, and at least one of those taken knew of the activities of the station. Fearing for their safety, on 21 May the members of Pi asked to be rescued and the following night a Catalina collected Snefjellå and four others. This was not a straightforward operation, for there was a German guard boat in the area, and in their haste to escape the agents had to leave a large amount of money and grenades in their boat, though these were later recovered by a sympathiser.18 Godfrey wrote to Air Chief Marshal Joubert de la Ferté, in charge of Coastal Command, to thank the crew for ‘carrying out the hazardous flight so successfully’.19 The Norwegian crew of the Catalina, and John Turner of SIS who had accompanied them, were decorated with the King’s Cross by King Haakon.
Not all SIS attempts to rescue their agents were so successful. In February 1945 Ibsen, a station based in Finnsnes in northern Norway, reported that Lofoten was carrying 150 Norwegian prisoners from Tromsø to Trondheim. This number included three SIS head agents, from Lyra, Synnove and Taurus. SIS sought Admiralty assistance in arranging to intercept the vessel and rescue them. This proved to be impossible, because the only ships which could be made available were unsuitable for the stormy weather then prevailing. SIS made considerable further attempts to intervene, contacting another five stations to see whether they could help, but these efforts were also thwarted by poor weather. The agents were taken to Oslo, and survived the war.
It was perhaps unrealistic of Welsh to expect that hermit stations would be able to manage entirely on their own. They needed to rely on assistance provided by the local population for a range of support, from charging accumulators to enable them to work their radio sets, to the occasional provision of fresh food and keeping them in touch with events in the neighbourhood, particularly the effectiveness of German attempts to locate them. Sometimes this support included acting as couriers and delivering information provided by other agents, which could supplement the coast-watching intelligence they sent back to London.†† For young and active men, living in such close proximity for extended periods in danger and under stress could cause nervous tensions which led to disagreements and fights and bloodshed. The damage this caused to their hands could then create problems for them when sending in Morse code. Atle Svardal, who had been a part of Eric, the first hermit station, returned with Eivind Viken to its successor, Erica. He and Viken occasionally came to blows. When in August 1944 Svardal was sent back to establish Roska, also in the same area, he asked for boxing gloves to be included in their supplies, to minimise both the damage which they might cause to each other and also the difficulties which they might otherwise have had with their Morse transmissions.20 This showed a remarkable philosophical acceptance of the pressure and tension which he knew that he and Viken would be facing once again on their mission, as well as an imaginatively pragmatic way of dealing with it.
There were just over a hundred SIS stations active in Norway in contact with London at various times during the German occupation. Some of them lasted only weeks before their agents were arrested by the Germans or disrupted and force to leave their posts. Others were able to last much longer. Beta, in Oslo, held the remarkable record of remaining operative from 14 January 1942 until the end of the war, a period of nearly three and a half years.
Rudsetra, one of the locations of Beta, SIS’s longest-surviving station, in the Nordmarka, a large forest to the north of Oslo. © NHM
The various incarnations of Upsilon, in Tromsø, did almost as well. Others achieved notable results in a much shorter space of time. Cygnus, near Florø, sent around three hundred messages in seven months between October 1943 and May 1944. As soon as November, Welsh noted that Cygnus was sending in valuable messages almost every day. The Air Officer Commanding in Scotland commented that the intelligence was of great value and that several operations had already been based on it. Furthermore, the Admiral Commanding Orkneys and Shetland wanted to see such operations extended to the Fleet Air Arm and to submarines: SIS was exploring how this might be done. Cygnus was eventually forced to close when it was located as a result of a German direction finding operation, but the agents were evacuated by a British MTB on 29 May. There were many similar examples. It was not only direction finding which enabled the Germans to identify stations and capture their agents. In one bad month in February 1944, SIS lost a series of stations including Upsilon in Tromsø, and Scorpion and Lark II in Trondheim. Welsh concluded – we do not know how accurately – that all of them had been lost due to German penetration of courier services which were carrying equipment or information to them from Stockholm. Both Welsh and Nagell had long been concerned about the risks of couriers from different services, mainly SIS and SOE, using the same routes and often travelling together, especially to northern Norway. Nagell observed, ‘to judge by the not too good security-mindedness of these couriers I am afraid that in some cases this may lead to the ultimate destination and purpose of the journey being made known to the other party’.21 Two further stations in the north, Libra and Taurus, had been forced to close for a while because of t
he loss of Upsilon. However, Welsh reported more cheerfully (and exaggeratedly) that the Libra wireless operator, while in the best of health, was taking a holiday at ‘the SIS convalescent home’ at Blomli in northern Norway!
There were other ambitious plans which did not work. One, in March 1943, was to establish station Andromeda on the island of Stjernøya, which would have controlled the entrance to the Altafjord. This could not be achieved because it proved impossible to arrange delivery of the agents to the island. If it had been in place and had been able to survive detection – which would have been quite a tall order for the area was really quite exposed – then Andromeda should have been able to report the departure of the Scharnhorst on 25 December 1943. Ultra did provide coverage of this, but GC&CS decyphering of the critical sailing report was slightly delayed because a new setting had just come into force. Scharnhorst sailed to attack an Allied convoy but was sunk by the Royal Navy off the North Cape of Norway. The only SIS station operational in the Altafjord at that time was Ida, manned by Torstein Raaby, which was located too far from Scharnhorst’s base in Langfjord to learn of her departure in time.
Skylark B and Lark
Skylark B was one of the first stations to be established in Norway. Erik Welle-Strand led a small team which was trained and sent back to Florø in September 1940. They moved to Trondheim, but difficulties with both the radio and transmission schedules meant that the station did not establish contact until January 1941. The group recruited a range of contacts to help them, including Einar Johansen and Bjørn Rørholt, both of whom did significant work for SIS. Rørholt was a student at the Norwegian Institute of Technology in Trondheim. He was a radio expert and helped to resolve Skylark B’s communications problems. He had earlier had contacts with elements of the Norwegian resistance in Oslo. Such links enabled Skylark B to provide reports on German activities not only in Trondheim but elsewhere, too. After Oldell was broken up in Oslo in March, Skylark B was the only station functioning in Norway until August 1941. It was located by German direction finding in September 1941, and many of its members were captured. When this happened Rørholt was in Oslo and the Germans came looking for him. After exchanging shots with an arresting party, he escaped over a garden wall. The pursuing Germans were unable to follow Rørholt’s example and lost him.
With the loss of Skylark B, SIS had no coverage of Trondheim. This became a matter of acute concern once Tirpitz became operational and Theta reported its arrival there in January 1942. SIS was required to re-establish a station urgently. Despite the fact that he was well known in the area and was still being sought after the earlier abortive attempt to arrest him, Rørholt volunteered to return from Britain. His first journey was unsuccessful because the boat proved to be unserviceable, but he was back in Trondheim early in February and quickly established contact with London. Rørholt decided that it would be necessary to have more than one station to monitor German naval traffic. Showing remarkable courage and resourcefulness, he posed as an insurance salesman and entered the Agdenes Fort, a series of fortified emplacements which guarded the entrance to the Trondheimfjord. While offering insurance policies to German officers, he also reconnoitred a suitable site for locating a wireless station close to the forts. He had recruited Magne Hassel for the task, equipping him with a simple code and initially arranging for him to send (but not to receive) messages, so as to reduce the risks of his transmissions being intercepted. Hassel sent some valuable messages on the movements of Tirpitz and other German warships. Within a fortnight in May, for example, he provided intelligence on the movements of Hipper (twice), and Admiral Scheer, as well as several other reports of cruisers and smaller naval vessels. In March 1943 he reported that Tirpitz with her escort had left the Trondheimfjord. This message reached the Admiralty within three hours. It was welcome intelligence because an aerial reconnaissance the previous afternoon had shown Tirpitz safely at anchor and they were not aware of her planned deployment.22 Before he left again, Rørholt established a network with three further stations in Trondheim to extend naval intelligence coverage. These were Leporis, Scorpion and Virgo. All were valuable, and Scorpion was particularly useful.
Rørholt was awarded a DSO on his return, one of three such decorations which were given to SIS Norwegian agents – the others were Einar Johansen and Torstein Raaby. (The citation for Rørholt’s award is given in the Appendix.) However, his career in SIS did not last much longer. He felt strongly that SIS radio communications were inadequate, and potentially dangerous to Norwegian agents. Before his departure to set up Lark he had worked with the Poles to develop a better, lighter and more compact wireless set, and took one with him. On his return to London, he sought to persuade SIS to introduce the use of the Polish sets with better cyphers, as well as adopting changes in sending frequencies which would complicate German direction finding activities. SIS refused to accept his recommendations, concluding that they were too complicated for the ordinary agent.23 Rørholt argued strongly against this judgement, and at a meeting in October 1942 attended by Roscher Lund, Nagell, Cordeaux and Welsh, it was decided to dismiss him.24 He spent the rest of the war working with SOE.25
Eric and Erica
Eric was the first hermit station to be set up in Norway. It was based on Gåsøya, north of Florø, and was manned by Atle Svardal and Dagfinn Ulriksen. Eric started in August 1941, and the two agents stayed there for seven months through a bitter winter, providing regular reporting on German naval movements. They were withdrawn in February 1942 because of concerns about German direction finding. Accompanied by Eivind Viken, Svardal returned to the area in January 1943 to set up Erica, which has deservedly been described as one of the most successful stations run by SIS in Norway, for both the accuracy as well as the timeliness of its reporting. In a progress report, Welsh described one of their most significant achievements, noting that on 5 February Erica had sent in the following:
Time of observation 1100: time of transmission 1312: time of receipt at HQ 1342: time of receipt at Admiralty 1348. The essence of this message was that one battle cruiser, one cruiser and three destroyers were travelling south. This signal was amplified on the same date at 1440 (received here at 1617 and passed to Admiralty at 1630). This proved fairly conclusively that the vessels in question were Hipper, Koln and three destroyers.
The quicker the Admiralty received such reports, the greater the chance that they would be able to arrange for the warships to be intercepted, although all too often such attempts were frustrated by poor weather. Despite the timely warning, that is what happened on this occasion.26 On 1 May at 0630, Erica reported a sighting of Nurnberg going south, a message which was reinforced by one from Pollux (on the Sognefjord) sent at 0910 and received in the Admiralty at 1200.27 These supplemented Ultra reporting and provided sufficient notice for RAF Coastal Command to search and sight her twice, but their attempted strike came to nothing.28 Erica was also responsible for reporting which led to the sinking of the German merchant ship Optima by two Norwegian naval ships, MTB 619 and MTB 631, in Florø harbour on 14 March 1943.29 NID was delighted by this intelligence and the deputy director of naval intelligence (D/DNI) wrote to SIS to ‘express my appreciation for the invaluable reports which have been received from this station. The reliability of the reports and the speed of transmission make them of great value.’30 When Viken and Svardal returned to London in September 1943, they met Menzies, who also congratulated them.‡‡
Ole Snefjellå: Pi, Crux and Frey
The brief but successful activities of Pi have already been described. When Ole Snefjellå went back to Norway the following year to a different location further north at Renga, near Mo i Rana, he took over the work of Crux with his brother Tore. Almost every progress report circulated by Welsh during the nine months they were based there commented that they were sending in valuable convoy intelligence. In early October 1943 they reported the passage northwards of the 4,000-ton cargo ship Skramstad, which was carrying nearly 1,000 German troops. She passed
so close that Snefjellå could see the faces of the troops as they stood on deck admiring the passing scenery.31 Skramstad was sunk on 4 October by aircraft operating from the USS Ranger.
The Snefjellå brothers returned to Norway in April 1944 to establish Frey on Gurskøy, south of Ålesund. The site they chose enabled them to watch convoys travelling south from Ålesund to Stad. They remained there until January 1945, and for much of this period they lived out in the open, their only shelter being a crude stone hut which they had built for themselves. The Germans had by this stage learned that the predictably precise timings and speeds of their convoys enabled Norwegian agents to forecast their progress along the coast. This facilitated the organisation of Allied attacks. So they mounted a series of intensive but unsuccessful searches to try to find them. On one occasion a German patrol spent the night only fifty feet from the ledge where the two brothers lay concealed. The Germans also tried to confuse matters by slowing down and delaying convoys before they rounded the peninsula at Stad. This did not work because it was right in front of the location where Frey was based. So Snefjellå could sometimes see the results for himself. ‘We were located right on the end of the Stad peninsula, and after we reported the passage of a German convoy it would sail round the end of the peninsula. Then it had no chance of escaping. We watched the aircraft going in to attack.’32 In July 1944, British aircraft attacked a large German ship carrying munitions, which exploded and sank off Kvamsøy. In December, the Snefjellå brothers watched a successful attack on a convoy which left ships in flames, ‘and danced with joy’. SIS stations did not very often receive specific feedback on the value of their reporting, though Snefjellå did comment that Welsh credited Crux with responsibility for reporting which led to the sinking of twelve ships. ‘But I don’t know whether that is true for he was inclined to boast.’33