Coastwatching Behind Enemy Lines
The Japanese assault on Asia and the Pacific had succeeded beyond anything Japanese planners had foreseen. Their successes were so fast and so extensive that they were sure it would take the United States and Britain a long time to regroup and launch the counteroffensives the Japanese were sure would come.1 What the Japanese did not expect was that those counteroffensives would come within a month of the US surrender at Corregidor, when the US Navy caught the Japanese Combined Fleet off Midway Island and destroyed its strategic capability.2 Within another month, on 7 August, the US Navy had opened the Guadalcanal campaign that would draw the IJN and IJA into a six-month battle of attrition and drain Japanese strength for the defence of the areas it had already captured.
For Australia, there were losses of territory, as New Britain and New Ireland were occupied and the Japanese swept down the north coast of New Guinea, heading towards Milne Bay and Port Moresby. One of the major advantages for the Allies, however, was that their intelligence system remained intact, and while the SIGINT organisations were being reorganised in Melbourne, they maintained cover of the IJN and IJA nets, and fed intelligence to GHQ, SWPA and the US Navy’s Command South Pacific (COMSOUPAC). The other major Allied intelligence system to remain intact and in place was Eric Feldt’s coastwatch network. This was now, however, operating behind enemy lines, something that had never been envisaged when the organisation was established.
With their capture of Rabaul, the Japanese had removed the threat it posed to their base on Truk.3 In Rabaul, they had also obtained a strategically important port that enabled them to control the waters of the eastern archipelago, and a forward air base from which they could cover any operations to capture Port Moresby, Buna and Guadalcanal in the British Solomons.4 The importance of these locations lay in airfields, which, once constructed, would allow the IJN to provide the land-based air cover under which its surface fleet could operate to cut the Allied supply lines to the west coast of the United States.
The Japanese threat assessment showed that any major attack on their southern position would come from the British in the Indian Ocean or the Americans in the Pacific. By mid-March, Imperial HQ had ordered that Japanese forces in the islands regroup, consolidate their positions and establish the military administration of the newly conquered areas.5 By late May, the Japanese believed that peace and order had been restored in the newly occupied areas, and the military administration was progressing smoothly.6
While this regrouping and consolidation was going on, the Imperial HQ considered its options, including invading Australia, an idea pushed by the IJN Section. This was quickly discounted for the alternative of seizing Port Moresby, Fiji, Samoa and New Caledonia, as well as areas in the archipelago from which a naval and air campaign against northern Australia and the sea lanes from the west coast of the United States could be launched.7
As soon as Rabaul fell, the Japanese mopped up the fleeing Australian garrison and other Europeans on New Britain. On New Ireland, the situation was not as bad, because the Japanese were not particularly interested once they had seized Kavieng. At this point, the focus was on Buna, Buka, Bougainville and the Solomon Islands as jumping off places for their campaigns against Port Moresby, New Caledonia and Fiji.
This was the plan until the Battle of the Coral Sea. On 17 July 1942, Imperial HQ released the 17th Army from the mission of capturing Fiji and New Caledonia, and the following day, the 17th Army ordered the South Seas Detachment at Rabaul to capture Buna and launch an attack on Port Moresby over the Kokoda Track.8 This was the situation in the islands as Feldt’s coastwatchers reorganised themselves to continue their work.
For Eric Feldt and the coastwatcher organisation, this period of the war involved extracting coastwatch parties now encumbered by civilian and military refugees. The situation was unsustainable. First, coastwatchers were being distracted from their primary task of collecting and reporting intelligence; secondly, the refugees were drawing the Japanese onto the coastwatcher parties. On Bougainville, New Britain and New Ireland the situation was bad, while in the British Solomon Islands protectorate things were slightly better, and intelligence was kept flowing despite the problems.
In Australia, Feldt was facing other issues, principally defending the organisation from a takeover by what he termed the army’s ‘Boy’s Own Annual brigade’, who had suddenly realised they lacked any intelligence and observer system of their own. For the army, the answer was obvious. The RAN would be forced to hand over coastwatchers across New Guinea, the Torres Straits and northern Australia to the army.
The pressure from army finally led to a meeting at Victoria Barracks, Melbourne, on 16 April 1942 between Colonel Roberts, AMF, and Wing Commander S.G. Brearley, RAAF; supported by Colonel Elliott R. Thorpe, US Army, and commanders Long and Feldt. The RAAF and US Army representatives, who were attempting to create an effective early warning system for air raids, raised the need for a single joint-service observer organisation led by a single commander, who would of course be an RAAF officer.
Long responded by recommending Feldt as the commander of any joint organisation, keeping the proposed organisation under Navy Office.9 Colonel Roberts half-heartedly proposed that the organisation be placed under General Morris’s 8th Military District in Port Moresby. This was immediately rejected by Colonel Thorpe, who made it clear that General MacArthur’s GHQ would control this organisation. Colonel Roberts quickly concurred with this decision, as indeed he had most likely always intended.10 Roberts was no friend of Feldt’s coastwatchers, and had been working behind the scenes to bring them under the auspices of Military Intelligence.11
What Roberts and Military Intelligence did not appreciate was that the coastwatcher organisation’s area of operations straddled not only MacArthur’s SWPA and Admiral Chester Nimitz’s South Pacific Area, but Australian and British territories, which, in the Solomons, were not under Australian control. The arrangement for control of coastwatching in the Solomons was a government-to-government agreement giving Navy Office the responsibility for the management of the Admiralty’s Reporting Officer system. For the Australian Army or the United States to take over this role needed British approval.
The other problem was that Feldt’s coastwatchers were on the ground providing intelligence essential to the operations of the US Navy in the Pacific. There was little likelihood that the US Navy, the only force in direct contact with large Japanese forces, would tolerate any disruption to the flow of intelligence from Feldt’s coastwatchers.
The result was a compromise with army looking after observer posts in New Guinea and Papua, and Feldt retaining control of the coastwatchers on Bougainville, New Britain, New Ireland and the Solomon Islands. This agreement did not end the infighting and Colonel Roberts would eventually drag Feldt’s organisation into his unsuccessful outfit later in the war.
With the agreement made at the meeting between the services, the new division of responsibilities meant the break-up of the coastwatcher group created by District Officer J.K. McCarthy and Captain G.C. Harris in order to evacuate refugees from New Britain. Two members of the group, R. Chugg and R. Emery of the New Guinea Volunteer Rifles, remained with ANGAU, while McCarthy, Harris and the rest became part of the coastwatcher organisation.12 L. Bell, K.C. Douglas and R. Olander were appointed to the RANVR, Bell and Douglas as lieutenants and Olander as a Sub-Lieutenant. At the same time W.L. Tupling on Ninigo and K.H. McColl at Wuvulu were picked up sending signals from their remote islands. These two were subsequently ordered back and were appointed petty officers in the RANVR, while Morton Johnson, who had been building cocoa dryers on Ninigo, was made a sergeant in ANGAU.13
While Long and Feldt were battling the army, the situation in the islands was deteriorating as the Japanese established their military administration and consolidated their hold. This made them the new big men,14 a reality that augured ill for the coastwatchers. As the Japanese became the distributors of work and wealth, they would win over a portion of the local population, who would assist them in tracking down coastwatch parties. On Bougainville, this is exactly what was about to happen.
Until this time, conditions on Bougainville had allowed the coastwatchers Jack Read and Paul Mason, accompanied by a party of soldiers and locals, to continue reporting intelligence, including advance—in most cases very much so—notice of Japanese air raids flying against various targets in the area of operations. As the tempo of the military operations increased, so did Read and Mason’s reporting. But the IJN’s SIGINT system, which, contrary to popular belief, was well organised and highly effective in identifying signals and their location, exposed them to danger.
As the coastwatchers transmitted their reports, there is no doubt the Japanese were conducting SIGINT operations against them, most likely using English-speaking Filipino intercept operators. They would have been aware of the relationship between the movements of IJN aircraft within line of sight of Bougainville and the transmission of the coastwatch messages reporting them. This would explain the appearance of more and more Japanese vessels in the anchorages around Bougainville, especially the area between Buin and the Shortland Islands. These vessels would have been fixing the signals and reducing the triangulation error to manageable proportions. The coastwatchers, well aware of the threat, attempted to shield their ground wave to avoid detection.15
The only advantage the coastwatchers had in this game was that triangulation, no matter how good, does not pinpoint a location but provides a triangular area within which the targeted transmitter is operating, as Figure 22.1 shows. This triangle can be reduced, but not to a pinpoint. This factor meant that the Japanese search parties had to cover a lot of ground when searching for a detected transmitter, and that is where the skilled local trackers came into their own. This need to search the triangle provided sufficient warning for coastwatcher parties to pack up their portable equipment and move. This was a slow process, however, because in 1942 portable was a relative term.
The portable equipment carried by a coastwatcher party needed a gang of carriers, somewhere between fifteen and 30. The tele-radio used by coastwatchers was, for its time, an excellent fixed-point radio set, and could be split into three components—the receiver, transmitter and speakers—for transportation. Each component weighed 32–45 kilograms, making a total load of up to 120 kilograms, not including the antennae, keys, batteries, charging engine and benzene fuel. It took between twelve and sixteen strong and experienced men to carry it and its accessories, and it was a nightmare to carry across broken ground and through jungle.16
Apart from the radio, a coastwatcher party had to carry its own provisions, tents, camp equipment, weapons, ammunition, medical equipment and water. Yet this was not all that needed to be carried. Whatever motivated the locals to assist the coastwatchers, it was not entirely altruistic and most definitely not patriotism. The carriers and support provided by village big men had to be paid for in commodities they wanted rather than cash, which is lighter to carry. The villagers quite rightly wanted things they could not get, like burlap (hessian), cotton cloth and, most of all, twist tobacco. This required a lot of muscle and a lot of food for energy. The image of small groups of coastwatchers stealthily making their way through the jungle is utterly wrong; in reality, it was more like a travelling circus.
Figure 22.1: Direction-finding fixes on a target transmitter
As time went by and Japanese plans were more frequently thwarted, it was logical for Japanese commanders to bring in experts from the Kempeitai or the IJN’s own Tokkeitai. On Bougainville, the Japanese began inserting many more patrols from the sea, and to establish supply and equipment caches for these patrols to use. They also began employing locals to work for them, which increased their standing as the new big men. This they reinforced by terrorising the locals if there was any hint of opposition or resistance.
As the Japanese reliance on local labour increased, the coastwatchers exploited it by inserting their own loyal police and carriers into the Japanese workforce. This provided the coastwatchers with raw intelligence on what the Japanese were doing, how far they had progressed, and where the most valuable targets such as ammunition and fuel stocks were hidden. All of this was being passed back to Townsville and Port Moresby, adding to the wealth of information the Allies were receiving.
The activities of the coastwatchers on Bougainville and other islands had now really begun to attract attention. This increased Japanese anxiety about them may have come from reading the coastwatch codes, enabling the Japanese to form a picture of just how comprehensive the coastwatch reporting was. The idea that the Japanese were reading coastwatch codes is not far-fetched. They had extensive experience of breaking and reading the Nationalist Chinese codes, and they had captured documentation on low-level British and US codes during their rapid advances of early 1942. It is more than likely that cryptanalytical research was now providing them with considerable insights into the remaining low-level codes used by coastwatchers and others.
By the middle of 1942, with the tempo of operations building in the lead-up to Guadalcanal, the Japanese took action to clear Bougainville of its coastwatchers. They started by sending large patrols across the interior of Bougainville in an effort to catch the coastwatchers. These patrols didn’t work.
This activity, and the violence of the Black Dogs, the group of locals formed at Buka, would later be attributed to Tashiro Tsunesuke, a Japanese civilian who had lived in Rabaul and Bougainville since 1917. Tsunesuke left Rabaul just before the outbreak of hostilities and had returned with the Japanese military administration in 1942. The evidence later collected against Tsunesuke doesn’t stack up, and it is unlikely that he was anything other than a coopted civilian forced to work for the Japanese administration. The man who does look like the professional counterintelligence operative was Tsunesuke’s boss, Petty Officer Harada, of the 1st Base Force.
Petty Officer Harada was fluent in English but could only speak a little pidgin, and this probably resulted in him acquiring Tsunesuke.17 The fact that a naval NCO could obtain such a scarce asset indicates that Harada was Tokkeitai. Only a Tokkeitai NCO would be left controlling a Japanese civilian with the sort of knowledge and background Tsunesuke had.
Further evidence that Harada was Tokkeitai comes from stories that even today are told on Bougainville. Harada was the head of propaganda and local pacification on the island, and was also responsible for investigating anyone suspected of helping the enemy.18 He was also widely feared by the local population and by IJN personnel.
The situation was now so bad that Paul Mason and the others could not carry out the plan detailed in the map in Figure 22.2. Mason hurriedly withdrew across the top of the island and just evaded capture.19 This withdrawal led to collateral damage when a European planter, Tom Ebery, who had been hiding in the area, was found by the Japanese. Ebery was later killed or died crossing the Mailoh River near Bogisago on 4 January 1943. Some reports say he was tortured and then killed.20
Mason’s party were in a bleak situation. If they stayed in the area, they would most likely be caught, and even if they weren’t captured, they would be moving and unable to provide intelligence. Lieutenant Commander Mackenzie, who now controlled the parties on Bougainville as well as in the Solomons, ordered Mason to head north and join Read. En route, Mason came across another European, Frank Roche, a miner who had stayed to protect his valuable mining plant. Mason stayed with Roche for two days while a jungle sore on his foot healed. Roche refused to leave his equipment and was killed by the Japanese tracking Mason’s party.21
On 27 January 1943, after covering 160 kilometres Mason found Lieutenant J.H. Mackie’s camp and the next day they all arrived at Jack Read’s location. By this stage, Read had arranged the evacuation of all Europeans on the island and numbers of Chinese men, women and children, who were at risk of being murdered or raped by the Japanese and the Black Dogs. By now, Mason, Read and Lieutenant Mackie’s section had been on Bougainville for almost a year. It was time for them to leave. Yet both coastwatchers decided to stay because they believed they owed this to the local inhabitants who had risked so much to help them.22
In Townsville, concern continued to grow as the brutality and aggressive nature of the Japanese counterintelligence activity increased. Feldt now believed everyone on Bougainville needed to be extracted.23 This included nuns and the Chinese, as well as the remnants of 1 Independent Company. The evacuation was timed for the end of December and took place on New Year’s Eve when Read led the evacuation of the last Europeans, including the nuns, by the USS Nautilus.24
Despite this removal of the civilians, the situation continued to get worse. Paul Mason now joined Read, as the southern end of Bougainville became too dangerous. Japanese patrolling became even more aggressive and they found many Chinese civilians and inflicted great suffering on them. They also started pushing back Lieutenant Mackie’s outposts.25
Mackie’s unit, which was part of the AIF and not subordinate to FERDINAND (the coastwatchers), now signalled Army HQ to request permission to withdraw. Colonel Roberts and the AIB used this as an excuse to interfere in FERDINAND’s arrangements for the extraction operations by arranging a separate evacuation of Lieutenant Mackie’s section.26 The inclusion of three nuns, eight other women, 27 children, and missionary Usaia Sotutu’s wife and family, does not appear to have been contemplated by Roberts.27 According to Feldt, on 29 March, when Read boarded the submarine USS Gato, commanded by Lieutenant Commander Robert J. Foley, US Navy, it was news to Foley that 51 people were in the party.28 In true US submariner fashion, however, Foley and his crew made it happen, and they squeezed the entire party into the Gato.29
Before Lieutenant Commander Foley could embark the refugees, he had to disembark Lieutenant Douglas N. Bedkober and another section of Independent Company troops, plus ten more from the AIF, as well as J.H. Keenan from FERDINAND and two locals, one from Buka and one from South Bougainville, who were to return to the local areas to spread the news of Guadalcanal and other Allied victories among their families and friends.30
In April 1943, a second evacuation was undertaken. The submarine landed E.D. (Wobby) Robinson and G. Stevenson, and took off Lieutenant Mackie and his last twelve men, Bishop Wade of Bougainville, nine missionaries, a planter, R. Stuart, an unknown European man, eight Chinese, seven other men and one child.31
Despite the success of the reliefs conducted by submarine, the situation on Bougainville really began to unravel two days before the arrival of the last submarine. The first disaster was the crashing of the RAAF Catalina conducting the resupply drop, which killed three of the crew and seriously injured four others.32 To make it even worse, heavy rain and difficult ground prevented the evacuation of the injured, who now needed to be nursed by the FERDINAND team.
The party on Bougainville included five experienced operatives—Wobby Robinson, Stevenson, Mason, Read and Keenan—and 22 new soldiers, seven Chinese men plus the carriers required for such a big group. This was a very large number of people to hide from the Japanese. On top of this, they had four injured airmen. Things were out of control. Despite this, they planned to establish a network of observation posts to cover the island (see Figure 22.2).
The plan they were working to was to cover the whole of Bougainville. Keenan would lead a party in the north, Read and Robinson in the central east, Sergeant G. McPhee on the west coast, and Mason and Stevenson with eight soldiers, Usaia, William McNicol and ten local police in the south. Lieutenant Bedkober and his remaining soldiers would guard the main camp with the injured RAAF crew.33
Mason’s party set off in May 1943 for the west coast so as to travel by canoe to the south of the island. They found signs of Japanese activity everywhere, including inland when they moved there for safety. A few days later, because of a lack of carriers, they could not carry everything at once. It would appear that fear of the Japanese had frightened the locals sufficiently for them to refuse assistance to FERDINAND parties. Mason’s group had to cache its equipment and provisions, leaving a local police officer and some carriers as guards. Within a short time, the Japanese attacked the cache, killed one carrier, captured the police officer and scattered the rest.34
Now the inexperience of the AIF troops became a liability, as they did not move through the jungle like the FERDINAND personnel or the carriers. They made noise, lost equipment, wasted supplies and demonstrated a lack of familiarity with jungle living that handicapped the FERDINAND personnel. In an environment where there is no margin for error, this was a bad omen.35 On Saturday, 26 June 1943, another serious loss occurred when a Japanese patrol attacked Stevenson, Usaia and their small group while they rested in a village. The group had posted sentries on two paths leading to the village, but a local man led the Japanese by a third. The Japanese immediately attacked, killing Stevenson. Usaia led the survivors to Mason’s party the next day. They had lost all of their equipment and food, and the Japanese, aided by the local inhabitants, were close behind.
In the north, the Japanese activity was so intense that Keenan’s party could not collect intelligence, let alone send messages to Townsville and Port Moresby. At Lunga (Lungga) on Guadalcanal in the Solomons, the Deputy Supervising Intelligence Officer, Lieutenant Commander I. Pryce-Jones, recommended immediate evacuation of all parties on Bougainville, and Commodore J.C. McManus concurred.36 On 29 June, Mason’s party was ambushed and in the confusion lost William McNicol, a police officer, and several carriers who bolted into the bush. They also lost all of their equipment, including their packs, which they had to drop in order to run. For seven days, the Japanese chased the remainder of the party until they managed to break clear. On 18 July 1943, Mason’s party made contact with Keenan.
Source: ‘Reports from coastwatchers in the Solomon Islands area’, BC410664, NAA: B3476, 37A
Figure 22.2: Planned coastwatcher coverage of Bougainville, 1943
Read and Robinson on the east coast were seeing little of interest, but again the aggression of the Japanese patrols was creating a grave risk of contact. In early June, the Japanese fell upon Read and Robinson’s camp and destroyed their food and equipment, although they missed the tele-radio hidden in the jungle. In the north, Keenan was attacked twice in succession, forcing him to abandon his tele-radio and move south to join Read and Robinson. Read sent Keenan and his surviving party off to the west coast and moved his own party to the AIF camp with the injured RAAF aircrew in it. What Read didn’t know was that Bedkober’s party had also been heavily attacked and suffered significant losses on 16 June.
Early that morning, Bedkober had sent Sergeant V.M. Day and three soldiers to prepare a drop site. Around 45 minutes after Day’s group had left Bedkober at Sikoriapia Village, they heard heavy firing as a force of 80 Japanese and 40 locals attacked the village. One of the RAAF crew, Flight Officer C.S. Dunn, was shot dead, Bedkober and Corporal J. Fenwick were captured on the spot, and the rest of the party was rounded up later. The situation that FERDINAND faced on Bougainville was that Mason was off the air, Bedkober’s party was wiped out, McPhee was running from the Japanese who had attacked Bedkober, and Read and Robinson were barely surviving. Aggressive Japanese patrolling with local assistance had shattered FERDINAND’s operations on Bougainville in less than a fortnight.37
On 26 June, Read signalled FERDINAND that Bougainville was untenable and all personnel needed to be evacuated immediately.38 Read’s signal led to the US Third Fleet’s unit 31 (CTF 31) requesting COMSOUPAC urgently extricate the 130 personnel, as the Japanese were now conducting wide and coordinated efforts to surround them. It was assessed that some parties could be concentrated by 20 July at a point on the mid-west coast between Belua and Cape Moltke, with the remainder ready by 31 July.39 On 31 July, the last of FERDINAND’s personnel were taken off Bougainville and all coastwatching activity ceased.
The operations on Bougainville had proven expensive. FERDINAND had suffered nine killed, wounded or captured, and had listened as aggressive and capable Japanese forces chased three major coastwatch parties so effectively that no intelligence was obtained. For FERDINAND, Bougainville was a failure and a demonstration of how difficult such intelligence operations became when the local population was not entirely supportive.
Yet it was not all doom and gloom for FERDINAND, because further south, in the British Solomons, the coastwatchers were winning the admiration and respect of the US forces fighting the Japanese in and around Guadalcanal. This now takes us back to early 1942, when the Japanese moved south to threaten the Solomons.
In 1942, the Solomon Islands was a British territory controlled by the Resident Commissioner, who at this time was W.S. Marchant. Marchant had decided to stay in the Solomons, but to withdraw his administrators and district officers from the northern parts of the islands and set up his HQ on Tulagi.40 From here, Marchant administered the islands and kept a firm British hand on what would happen on British territory. To make sure that he and his staff had standing with the military, he commissioned himself as a Lieutenant Colonel and was appointed to the command of the British Solomon Islands Protectorate Defence Force (BSIPDF), into which he then conscripted all his district officers.41
As we have already seen, however, the responsibility for coastwatching in the Solomons fell to the ACNB, and Marchant had to cooperate with Eric Feldt in this activity.42 He even lost some of his officials, men like F.A. Rhodes and K.D. Hay, who were appointed sub-lieutenants in the RANVR, and A.M. Andresen and L. Schroeder, who were appointed petty officers, so they could serve as coastwatchers.43
The situation on Tulagi and in the Solomons generally was getting more difficult as the Japanese ramped up their air offensive, particularly against Tulagi itself. Although the effect of the Japanese attacks was reduced by the early warning coming from the coastwatchers on Bougainville, the difficulties of trying to run the government from Tulagi became too difficult for Marchant, so he decided to move his administration to Malaita Island. This caused a conflict with the coastwatch organisation, because Marchant wanted to take Lieutenant Commander D.S. Macfarlane, RANVR, the only qualified intelligence officer in the Solomons, with him. Macfarlane, however, had been ordered by Feldt to go to Guadalcanal.44
Marchant’s reasons for moving to Malaita were sound. He needed to establish a radio network that could communicate with Port Moresby, Suva on Fiji and Townsville, as well as with the stations spread throughout the Solomons and the islands of the South Pacific. The mountains inland from Auki on Malaita provided one of the best locations for this, and Marchant established the retransmission station there for the entire FERDINAND network in New Britain, New Ireland and the Solomon Islands. Professional operators, including T.W. Sexton, the Government Wireless Officer, C.N.F. Bengough and H.W. Bullen, operated the station.45 This team was very experienced in picking up signals from the entire region in the difficult atmospherics of the tropics, and re-encrypting the messages in a higher-grade code for transmission to the NIO at Port Vila; from there they were sent to Admiral W.F. Halsey’s South Pacific Area.46
An important station in this network was the Vanikoro post on Vanikoro Island in the Santa Cruz group. On this island was a female coastwatcher, 51-year old Mrs Ruby Boye-Jones, supported by her husband, Sydney Boye-Jones, who was the manager of the timber company on Vanikoro. Mrs Boye-Jones provided retransmission services for Marchant’s signals to Vila when atmospheric conditions were bad, and she furnished local intelligence reports as well as weather reports, which she sent four times a day.
The IJN was well aware of Mrs Boye-Jones’s activities, and in 1942 she was sent threatening messages by the Japanese intercept operators listening to her messages. The threats did not stop her. In line with his policy of appointing all FERDINAND operatives to a military rank, Eric Feldt worked at getting Ruby Boye-Jones appointed to a position as a WRANS. This proved difficult because, as an Australian servicewoman in 1942, she was forbidden by government policy to be where she was, even if it was her home. It took until 27 July 1943, and then she was only appointed as an Honorary Third Officer, WRANS. This was the best Feldt could do.47
Eric Feldt held Ruby Boye-Jones in high esteem and commended her to Admiral Halsey, Commander of South Pacific Area. Halsey also held her in high esteem, because he not only accepted Feldt’s commendation, but also made a personal flight to the remote island to personally thank her for her work.48 Later, when Halsey was told Mrs Boye-Jones had become ill on Vanikoro, he ordered a US Navy aircraft to airlift her to hospital. She was the only European woman to have actively served as a FERDINAND operative in the Pacific.
Mrs Boye-Jones’s dedicated reporting of weather conditions and sea states day after day without fail provided vital intelligence for the US Navy. Weather reporting is an often overlooked but important facet of intelligence collection. In the North Atlantic, being on a weather ship was a most dangerous job, as the enemy naval and air forces regarded these ships as high-priority targets and could easily find them through HFDF. Mrs Boye-Jones may have just been sending weather reports, but they were most definitely highly regarded.
Sydney Boye-Jones remained at her side throughout and shared the same risks. He was also an active participant in keeping the tele-radio working and Vanikoro Island secure. He was a highly experienced ships pilot for the Solomon Islands and very eager to serve.49
From June 1942, FERDINAND reporting was showing an increasing level of Japanese interest in Guadalcanal where FERDINAND had six posts. K.D. Hay, A.M. Andresen, and Lieutenant Macfarlane, RANVR, manned a post on Gold Ridge in the centre of Guadalcanal. Sub-Lieutenant F.A. Rhodes, RANVR, manned a post in the north.50 District Officer Martin Clemens manned a post at Aola on Guadalcanal, Leif Schroeder was on Savo Island, and Geoffrey Kruper was at Tunnibulli (Tatamba), at the south-eastern end of Santa Ysabel Island.51 The final post was on the north-western part of Guadalcanal Island and manned by District Officer D.G. Kennedy, assisted by Assistant Medical Practitioner Hugh Wheatley. This was the most exposed of all the posts.52
Kennedy decided, correctly, that his post was too exposed and withdrew his party, first to Segi Plantation on New Georgia Island, and then to Santa Ysabel Island.53 While Kennedy was at Segi towards the end of March 1942, he decided to send Wheatley, who as half-Solomon Islander could blend in among the local population, to reconnoitre the north-western Solomons. Nothing was ever heard of Wheatley, the schooner or its crew again. It is likely they were either sunk by the advancing Japanese or captured and executed.54
As time passed in early 1942, the coastwatch posts passed their sighting and weather information back to Marchant’s HQ, where it was re-encrypted and passed to Australia and to the various US naval and military commands throughout the Pacific.55 This early intelligence reporting provided the stretched Allied commands with reliable sighting information on major Japanese naval and aerial concentrations. Among the intelligence provided was a report of 5 May from Kennedy on Santa Ysabel Island warning of a major IJN force moving towards Tulagi. This report enabled US carriers to get into position and launch the final air attacks that ended the Battle of the Coral Sea.56
Schroeder on Savo Island watched this engagement from his vantage point and was able to report the sinking of nine Japanese vessels by the US air attacks. This was the first check suffered by the IJN, and it was made possible by the good intelligence provided by the SIGINT of FRUMEL and the HUMINT provided by FERDINAND.
Despite the setback suffered by the IJN at Coral Sea, Tulagi was occupied on 6 May 1942, and now the Japanese could extend their reconnaissance activities onto Guadalcanal and the other islands of the Solomons. Clemens’ party was now in close contact with the Japanese and he quickly exploited this by sending a number of his police and carriers to work for the Japanese. These men reported the precise locations and activities of the Japanese and, on many occasions, assisted Japanese units in getting lost in the jungles of Guadalcanal.57
These minor disruptions did not slow the Japanese down and they moved to occupy strategically important positions, not least the grassy area near Lunga where, in June, they started building the airfield that they and the US Marines would spend the rest of the year fighting to control. This airfield, soon to be renamed Henderson Field, was right below Gold Ridge, where Hay, Andresen and Macfarlane watched and reported all they saw, along with the information their carriers and police were gleaning while working as labourers for the Japanese.58
This activity continued until late July, when the coastwatch posts were ordered to cease radio transmissions and maintain listening watches only. This instruction only meant one thing—an Allied assault was intended and Guadalcanal was the most likely target. To the north, the coastwatchers on Bougainville were watching the skies for Japanese aircraft, and the choke-points at Buka Passage and the Shortland Islands for signs of the IJN.59
On 7 August 1942, at approximately 0900 hours, the first wave of Marines went ashore between Lunga Point and Tagoma. Across the sound, the Marines went ashore on Tulagi led by two FERDINAND guides, Sub-Lieutenant Dick Horton and Sub-Lieutenant Henry Josselyn.60 As we have already seen, the Japanese responded aggressively at sea and in the air. The landing beaches were soon attacked by waves of Japanese aircraft, but the FERDINAND parties on Bougainville provided early warnings of these at 1027 hours, enabling the fleet to weigh anchor and disperse. Only two ships, USS Jervis, hit by a torpedo and sunk the next day, and USS George F. Elliott, set alight by what appears to have been a suicide attack by a single aircraft, were lost.61 It could have been much worse.
Now the coastwatchers on Guadalcanal came into their own by acting as guides for the US forces ashore and afloat, and by collecting HUMINT that no other unit could collect. Macfarlane sent his cook boy, police and carriers to work for the Japanese as labourers from Monday to Friday. Each weekend, the Japanese would give them leave and they would dutifully turn up at Macfarlane’s post and provide detailed reporting on the Japanese dispositions and activities, and the locations of their stores and weapons.62
With the success of the Marine landings, the coastwatchers had to provide time-sensitive tactical intelligence to the Divisional HQ ashore. This was no easy task. The Marines were under siege and because of the Battle of Savo Island were now without naval support. Knowing what the Japanese were doing was of vital importance to Major General Vandegrift and his staff, and he was heavily reliant on the coastwatchers. In order to improve communication with Vandegrift’s HQ, Mackenzie, accompanied by Lieutenant G.H.C. Train, and two other men, R. Eedie and Rayman, landed at Lunga a week after the first landings and established themselves close to Vandegrift’s HQ. The Marines found this so useful they provided three operators, privates Page, Adams and Berkstresser, to assist in the work.
This centre began disseminating intelligence, including air-raid warnings from Paul Mason on Bougainville, Jack Read at Buka, F.A. Rhodes and Leif Schroeder on the west coast of Guadalcanal, and Macfarlane overlooking Lunga airfield.63 These warnings were so timely they turned out to be too timely. The problem lay with the divisional HQ issuing an ‘air raid yellow’ alert ceasing all work despite the raid aircraft being two hours’ flying time away. Vandegrift ordered that ‘air raid yellow’ warnings not be given until the attack was 30 minutes away and the ‘red’ warning at ten minutes.64
At Segi Plantation on New Georgia, another coastwatcher, New Zealander D.G. Kennedy, was directly under the Japanese flight path to Lunga, which was optimal for reporting. The only complication was that the Japanese forces were now at Viru Harbour, 14 kilometres north of Kennedy’s position.65 Kennedy decided he needed to hold his ground, so he organised an attack on the Japanese at Viru, wiping them out. He wiped out two more barges and their crews, then attacked a 25-man Japanese patrol, although many of them escaped into the jungle.66 These scuffles with the Japanese did not come cost-free. Kennedy was wounded in the thigh and two of his men were killed.
The wound did not stop Kennedy and he stayed on New Georgia.67 In all of this action, Kennedy’s party had destroyed three barges, killed 54 Japanese and captured twenty more. They also rescued 22 shot-down US aircrew from the jungles of New Georgia or from islands close to it. The rescued aircrew and captured Japanese were flown out from New Georgia shortly after this series of actions.68
At this time, the need to provide intelligence of Japanese naval activity in the area of water lying between Santa Ysabel Island in the east and New Georgia, Vangunu and Pavuvu islands to the west, now nicknamed ‘the Slot’, was vital. The IJN had already displayed its fighting prowess on 9 August, two days after the Marine landings on Guadalcanal, when it inflicted the greatest naval defeat in US history at the Battle of Savo Island. Leif Schroeder, who on Savo Island had the best bird’s-eye view of this battle, on the night of 8–9 August reported the action that saw the sinking of HMAS Canberra, USS Chicago, Astoria, Vincennes and Quincy.69 Schroeder would also witness and report on the battles of Cape Esperance on 11–12 October, Guadalcanal between 2 and 15 November and Tassafaronga on 30 November, also called the second, third and fourth battles of Savo Island. The waters around Savo are now named Iron Bottom Sound as a result.
As well as attempting to bring the US Navy to a major fleet engagement at Guadalcanal, the IJN also had to reinforce and supply the Japanese forces on the island. The intensity of the fighting in the waters around Savo Island and in the Slot meant that slow-moving transports had no chance of surviving the voyage. The IJN then began using fast-moving destroyers and other vessels to carry supplies and reinforcements to Guadalcanal. This was the ‘Tokyo Express’, and the intelligence of each IJN attempt to transit these waters was dutifully reported by the coastwatchers on Vella Lavella and Choiseul islands.
In late August 1942, the situation on Guadalcanal was still serious, but this did not stop Walter Brookbank, the most senior public servant working for Navy Office’s intelligence section, from visiting Mackenzie on Guadalcanal, and holding meetings with Vandegrift and his intelligence and operations staff. Commander Long had sent Brookbank, a man with extensive experience of the office politics of intelligence, to find out exactly what was going on at Guadalcanal.
Brookbank’s visit was no sightseeing tour. He was on Guadalcanal as part of a visit to the US Navy in the Pacific Area, most likely undertaken as a result of a dispute between Chester Nimitz, Commander of the Pacific Operational Area and Vice Admiral Leary, Commander of the South West Pacific Area, officially under Nimitz but part of General MacArthur’s command. The argument was, as usual, about who was in charge, part of which included the tasking of coastwatcher posts.
It might seem surprising that a civilian was sent to test these waters but, given the fact that all the senior RAN positions were held by RN admirals, it may have been thought inadvisable to have them try to influence the US Navy led by the stridently anti-British Ernest King. The other advantage of Brookbank was that he could say what needed to be said without causing a big stir because, after all, he was just a civilian.
Brookbank was able to get an agreement from the parties that Feldt’s coastwatchers would concentrate on supporting one operational HQ in the Solomons. He could also report back to Navy Office the very flattering praise being heaped upon Feldt’s coastwatchers by the US commanders on Guadalcanal.70
The task of supporting the US Marines on Guadalcanal was, however, impacting Feldt’s ability to place coastwatchers in other areas of operation. The biggest issue was the complaining from GHQ, SWPA, led by Colonel Roberts, that the RAN was cutting the army out of coastwatching operations. Roberts appears to have piggybacked this complaint onto the efforts of COMSOUWESPACFOR to interpose himself into the operations around Guadalcanal.71 The problem for Feldt was that he did not have the personnel to make up new parties, and Roberts was now offering to put together parties owned and led by army. Given the divergence between Feldt’s view of intelligence as not involving any direct action, such as sabotage, and the army’s desire to mount raids and blow things up, there was never going to be an easy settlement of the issue.
On 12 October, Long wrote to Feldt informing him that he was going to fight to ensure Feldt’s parties being put onto the Solomons were led by naval officers, and he believed that the Americans would not be too keen on parties led by Australian Army or RAAF personnel, of whom they had no prior experience.72 Feldt, backed by Long, continued efforts to keep the organisation from being turned into a guerrilla or irregular fighting force by ensuring that parties were led by people who understood that the intelligence role was the most important contribution they could make to the war. It was an ongoing battle against some elements in the US Marines as well.
But something had to be done, so FERDINAND inserted two parties using the USS Grampus on the night of 6–7 October 1942. One party, led by Josselyn and Keenan, went to Vella Lavella. The other party, led by A.N.A. Waddell and Sergeant C.W. Seaton, AIF, went to Choiseul.73 Josselyn would stay on Vella Lavella until October 1943, and Keenan until February 1943. Waddell and Seaton would stay on Choiseul until March 1944.74
On Choiseul, Waddell and Seaton found that, while a significant part of the population was supportive or at least neutral, the activities of pro-Japanese villagers from the north-west coast made it difficult to operate. Actions by pro-Japanese locals forced Waddell and Seaton to relocate twice between October 1942 and February 1943, and FERDINAND had two hostile villages bombed.75 Colonel Marchant, the British Resident Commissioner, immediately intervened after this action and demanded that reprisal-bombing cease immediately. Neither the US Navy nor FERDINAND were keen to stop this sort of activity, and Marchant elevated the matter to the Colonial Office via the Commissioner for the Western Pacific.76 On 12 February, the Colonial Office formally advised Australia and the United States that it was not permissible for junior officers to order the punitive bombing of British civilians who happened to live in villages in the Solomon Islands.77
By February 1943, the matter was settled to most people’s satisfaction. The British Administration was not swayed by the arguments for punitive bombing. It would appear that Admiral Halsey also had his doubts because, even before Sir Philip Euen Mitchell had contacted the Colonial Office, he had ordered that no village was to be bombed without his personal authorisation.78 FERDINAND signalled its parties in the islands telling them of the restrictions and of the need to provide a very strong case so that the Australian naval liaison officer at Vila could get approval for future reprisal raids.79
On Guadalcanal itself, the FERDINAND post supporting the US ground forces increased its staff with the arrival of Lieutenant L. Ogilvie, RANVR, and Sub-Lieutenant K. Harding, who took some of the heavy load off Mackenzie and Lieutenant Train.80 This reinforcement increased the speed with which FERDINAND was able to receive, assess and forward relevant intelligence to the Marine commander on Guadalcanal, MacArthur’s GHQ, Halsey’s command in the South Pacific, Townsville and Pearl Harbor.
At around this time, the existing FERDINAND parties on Guadalcanal were beginning to wilt. They had been operational for a considerable period of time and now, as the fighting between the Marine defenders at Henderson and the Japanese intensified, they were facing a jungle filled with fleeing locals and large numbers of enemy soldiers. On top of all this, they were within areas subject to bombardment from friendly artillery, warships and aircraft. Lieutenant F.A. Rhodes and Petty Officer Leif Schroeder, who were constantly shifting position, remained in the north-west of the island. On Gold Ridge, the pressure built up on Hay, Andresen and Macfarlane.
In the case of Rhodes and Schroeder, Feldt appears to accept a view he attributes to Mackenzie, that Vandegrift, the Marine Commander on Guadalcanal, had decided that the FERDINAND parties were ‘expendables’.81 To be fair to Vandegrift, he was not responsible for FERDINAND’s parties; they were simply co-located within his defensive perimeter and, worse, outside of it. When it became obvious that the coastwatchers needed to be brought in, Vandegrift ordered that no rescue or retrieval missions be launched. Mackenzie ignored this and sent Horton to evacuate Rhodes and Schroeder by boat. Luckily, Horton pulled off this daring rescue without a hitch, and evacuated sixteen people, including Rhodes and Schroeder, crashed Allied aircrew and some Japanese POWs. The action did not go down well with Vandegrift.
In early October 1942, Macfarlane and Andresen made their own way down from Gold Ridge to the Marine position at Lunga. The third member of their party, Hay, stayed and continued to transmit intelligence collected by his local assistants.82 Feldt attributes Hay’s decision to stay to his being too fat to outrun any Japanese patrol they might bump into during the withdrawal to Lunga. The plump Hay maintained his intelligence collection effort on Gold Ridge until after the defeat of the Japanese. When he finally came out of the jungle with his party, which included a Catholic nun, in January 1943 the US officer who met him was stunned at Hay’s girth. He did a magnificent job surviving on Guadalcanal from February 1942 until January 1943 and maintaining a constant flow of intelligence throughout that period.83
In his 1953 foreword to Martin Clemens’ book Alone on Guadalcanal, Vandegrift praised Martin Clemens and the other British colonial officials who remained as coastwatchers in the Solomon Islands. He applauded the relationship they had formed with the Solomon Islanders and the loyalty of the latter who, Vandegrift claimed, never betrayed any of the ex-colonial officials or handed over any US personnel to the Japanese.84 Vandegrift also praised the importance of the intelligence information provided to his division by FERDINAND, stating: ‘There were instances when that information and support was a substantial portion of the margin of victory.’85
The FERDINAND operations in the British Solomon Islands were the high point of Australia’s coastwatch organisation, and were made possible by the mix of British, Australian and wayward European residents of the islands collected by Commander Rupert Long and Lieutenant Commander Eric Feldt to quite rightly sit under the trees and smell the flowers like their mascot, Ferdinand the Bull, who refused to fight. This approach, which eschewed guerrilla operations and sabotage would prove to be a far more successful approach to intelligence collection than all of the special force operations put together. In doing this, the ACNB’s coastwatch element of the Admiralty’s Reporting Officer system won highly deserved fame as FERDINAND, one of the most successful, if not the most successful, western HUMINT activity of 1939–45.
