The Solomons and Pacific Area, 1943
With the defeat of the Japanese push to Port Moresby in late 1942 and the final defeat of the Japanese on Guadalcanal in early February 1943, the Allies in the SWPA and Pacific were now in a position to move to follow-up offensives to drive the Japanese back towards Truk and the Philippines. The war in the Pacific was now moving into its offensive stage, and with this change in strategic outlook came changes in FERDINAND.
Approximately six months after coastwatchers had been withdrawn from New Britain in the face of the Japanese advance, they returned aboard the US submarine USS Greenling. The FERDINAND party on the Greenling comprised Lieutenant M.H. Wright (SP), RANVR; Captain P.E. Figgis, AIF; a local from the area of Baien, to which the party was heading; Lieutenant H.L. (Les) Williams, AIF; Sergeant Simogun, a local policeman; and three locals, Sanga, Arumei and Sama.1 Some of the party went ashore on New Britain in the vicinity of Wide Bay on the night of 28–29 February 1943. Once ashore, they made their way to a local village, the home village of the Baien local they had brought with them. The next night, Figgis set off in a canoe to find the Greenling and unload approximately one ton of equipment and stores and the rest of the party, including Lieutenant Williams.2
Once this was accomplished, the difficulty was getting the beachside locals to supply the manpower to carry these loads into the interior. After a bit of toing and froing, the whole party, its equipment and supplies were carried to a location and the beachside locals released. Over the next four weeks, the party then moved everything to a secluded spot and built a camp.3
Lieutenant Wright’s party provided sighting information on Japanese barges and submarines that were now being used to ferry equipment to the beleaguered Japanese forces on the southern islands. In order to avoid air attacks, the barges moved by night and hid under the jungle canopy in small inlets and creeks during the day. The submarines offered the ability for underwater transits and could also fit up creeks and hide under the jungle canopy as well. The only countermeasure for this was coastwatcher reports providing precise locations for air attacks. For the next five months, the coastwatchers’ observations made even this small level of protection less effective, and a number of these barges and vessels were destroyed because of air attacks from aircraft directed by the coastwatchers.4
On 12 July 1943, Lieutenant Commander Pryce-Jones, RANVR, the Deputy Supervising Intelligence Officer on Guadalcanal, sent an appraisal to the DNI outlining the changes needed for coastwatching in the Solomons due to movement of the war northward.5 The most pressing problem was that the British authorities were now looking to take back control and tasking of their personnel from FERDINAND. The British High Commissioner in Suva was reportedly writing to the Australian prime minister asking for the release of Waddell and Josselyn as soon as their present missions ended. The Resident Commissioner, Colonel Marchant, was planning to send Kennedy, who was currently wounded at Segi, New Georgia, on a long leave. On top of all this, the British civil authorities had appointed Clemens to Kennedy’s post as District Officer, New Georgia, and on Rendova, Sub-Lieutenant Horton was evacuated for surgery on a hernia.6 In light of all this, Pryce-Jones wanted to know if FERDINAND in the Solomons needed to man all of the coastwatch stations or, as an intelligence collector behind enemy lines, only those in the areas still occupied by the Japanese.7 It was a fair point.
Commander Long replied to Pryce-Jones on 2 August laying out the ACNB’s policy for FERDINAND in the Solomons. The first point Long made was that all coastwatch stations that lay within areas occupied and controlled by the US forces were to be handed over to them to man, if they wished to do so. To facilitate this, the ACNB had authorised the handing-over of RAN tele-radios to US forces for their use.8 This would release remaining FERDINAND personnel for deployment behind enemy lines, which was FERDINAND’s sole function, and, if the war moved further and faster north than expected, then the deployment of FERDINAND parties from the Solomons would be authorised onto New Ireland and other islands beyond the Solomons.9 To assist in this reorganisation, Pryce-Jones was to delegate the management of all coastwatch activity within liberated areas and focus himself on managing FERDINAND activity in enemy-occupied areas.10
As this reorganisation was occurring in the Solomons, similar changes were occurring in Australia as the army, supported by GHQ, SWPA moved to pull FERDINAND into the AIB. One of these changes was the closing of FERDINAND’s HQ at Townsville.11 FERDINAND was now changing from an intelligence provider to a reconnaissance organisation, a role that fitted the philosophy of the Australian Army, GHQ, SWPA and the AIB more closely.
This change did not mean that the coastwatchers were not providing significant service. On 30 June 1943, during the landings at Rendova Island, Lieutenant Rhodes won the US Silver Star for his bravery in leading a party of US troops ashore and through a plantation.12 In the hills above this action, another FERDINAND party was fighting off a large group of retreating Japanese who accidentally came across the observation post.13
Coastwatchers were also providing their usual reporting of Japanese air and sea movements, and recovering downed aircrew and sailors from sunken vessels. Henry Josselyn, Robert Firth and A.W.E. Silvester rescued the 160 survivors of the cruiser USS Helena who washed up on Vella Lavella Island after the Battle of Kula Gulf.14
By September 1943, the only FERDINAND party behind Japanese lines in the Solomons was that led by Waddell and Seaton on Choiseul Island. One member of the party was a local radio operator, Pabula, a graduate of the old radio school at Rabaul, who sent the daily reports on weather and other intelligence while Waddell and Seaton were taking military parties on reconnaissance missions around the island.15
Waddell’s party, reinforced by Flight Lieutenant E. Spencer, RAAF, conducted a number of missions and fought engagements with Japanese patrols. This party stayed on Choiseul until 2 March 1944, a period of fifteen months. During this time, it provided intelligence on Japanese barge and vessel movements, enabling the US air forces to inflict substantial damage on the Japanese supply line operated by the Tokyo Express.16 They also fought a number of engagements with the Japanese, killing a reported hundred or so, assisted the US Marine landings on the island, and directed air attacks on Japanese targets on and around the island.17
To meet the requirements for the Torokina landing on Bougainville, FERDINAND prepared three AIB parties, one to land with the troops at Torokina, one to be inserted in the north of the island and one operating in the south. FERDINAND was now working for General MacArthur and GHQ, SWPA.
The experienced Captain E.D. ‘Wobbie’ Robinson, AIF, led No. 1 Party, the party going to the north of Torokina. No. 2 Party, led by K. Keena, headed to the south.18 The party landing with the troops on 1 November was led by Flight Lieutenant R.A. ‘Robbie’ Robinson, RAAF, supported by Sub-Lieutenant R. Stuart and a US Army team comprising Sergeant F. Halverston, Corporal Nash and Private Engler.19
These parties departed Brisbane on 14 October for Lunga to report to Pryce-Jones for further training and to join their local carriers and police. They would all be kitted up and provided with semi-automatic M1 carbines and modified webbing.20 They would be issued their tele-radios, rations for 21 days and other supplies by Pryce-Jones. They were to land at Kunua on Bougainville before moving to their respective area of operations.21
The risks they were running were significant. The Japanese had been in these locations for a considerable period, and they had developed their intelligence networks. They had also improved their SIGINT effort and were most certainly reading AIB messages, thanks to intelligence gained from the interrogation of AIB personnel in Timor and other places. There is no doubt that the Japanese SIGINT units were reading FERDINAND’s codes.22
Hard evidence of how extensively the Japanese had broken these codes came in November 1943, when the intelligence books of the Kira Force were captured. This book held extensive raw intercept from FERDINAND parties operating in the Solomons between April and November 1943. The Japanese knew the names of Read, Mason, Stevenson, Pryce-Jones and others. They knew group sizes and composition, and they knew the areas in which they were operating. They even had the FERDINAND instructions on how to ensure greater communications security through proper procedures. FERDINAND’s communications were entirely compromised by early 1943.23
Naval Intelligence’s evaluation of the captured documents was that they were captured when Read’s camp was overrun. The text of the transmissions, however, has an immediacy that only comes from an intercept operator writing down exactly what they are hearing in their headphones.24 The Naval Intelligence evaluation was wrong.
The role of SIGINT in detecting coastwatch parties via HFDF was not completely lost on FERDINAND. A memorandum of 8 February 1944 from Commander McManus, Supervising Intelligence Officer, North East Area, to the Officer-in-Charge AIB Base at Milne Bay, details a combined FERDINAND–Central Bureau operation to conduct SIGINT operations against Japanese stay-behind parties. FERDINAND and Central Bureau planned to fit the harbour defence motor launch HDML 1321 with specialist W/T equipment and an operator, Sergeant H.A. Dempsey, to detect and plot the positions of these parties.25
HDML 1321 would conduct a series of cruises in waters heavily patrolled by Allied forces, so Commander McManus wanted to ensure there was no possible chance of HDML 1321 being sunk by friendly fire. This operation was carried out between February and May 1944, the period when Dempsey served in FERDINAND, according to Eric Feldt’s book.26 It is now not known what findings the SIGINT team on HDML 1321 made or whether it even conducted intercept operations. The suspicion is that its SIGINT team were part of the specialist British communications security group brought in to listen to Allied communications in order to identify poor procedure and to check if ULTRA material was being passed via non-SIGINT systems, as indeed the SRD had been doing.
Even without the Japanese reading messages in real time, the situation facing FERDINAND’s parties going into Bougainville was bad. They were told that they would have little chance of surviving on the island for more than three weeks because the locals would betray them. This meant that their mission could not be inserted more than three weeks ahead of the intended landings.27 Other changes required included the way parties were inserted. In the Solomons, the procedure had been to land all supplies and equipment and place them in a cache. The coastwatcher then went out and negotiated with the local village headmen for the necessary carriers. In Bougainville, because of the pro-Japanese attitudes of many villages, this would be ‘folly’.28
At the front, there was increasing criticism from the US Marines, especially 1st Amphibious Marine Force, of FERDINAND’s lack of aggression. One reason is that they were not being properly briefed on what FERDINAND did and did not do. The problem had worsened because the Deputy Supervising Intelligence Officer, Pryce-Jones, had stayed on Guadalcanal and not moved forward with his clients.
Pryce-Jones did get involved now, and a stand-off developed between him and Amphibious Marine Force, which escalated when the Chief of Staff, Colonel Thomas, insisted that in future Marine officers would command FERDINAND patrols when Marines were attached. The next such patrol was to be led by Lieutenant Keenan of FERDINAND, and Thomas wanted Keenan to operate under the orders of Lieutenant Colonel James M. Smith during the reconnaissance of Atsinima Bay, Papua New Guinea, in late October.29
The complaint was still that FERDINAND was ‘not aggressive enough’, and the Marines felt they needed to ‘work more of our own into it for control purposes’.30 The Marines were looking for a fight; FERDINAND was not. Lieutenant General A.A. Vandegrift, Commanding General of the First Marine Amphibious Corps, agreed with his Marines.31 Fortunately, Admiral Halsey, advised by the G2, COMSOUPAC, Colonel F.P. Munson, agreed with FERDINAND.32
The Marines remained unhappy. On 16 November Lieutenant Colonel W.F. Coleman, the acting Chief of Staff and C2 of 1st Amphibious, wrote a letter to Lieutenant Colonel Floyd A. Scrow of 3rd Amphibious Force criticising Keenan. Coleman complained that, as expected, the reconnaissance patrol led by Keenan had been ‘a bust’, as the Australian tele-radios had been soaked by sea water during the landing and rendered useless. Rather embarrassingly, a copy of Coleman’s letter found its way into an Australian file, so FERDINAND knew the name of at least one complainer.33
The first warning that Pryce-Jones was overreacting was his use of a memorandum of 30 November 1942 in which Hugh Mackenzie, the then Deputy Supervising Intelligence Officer, noted that a report from Colonel Thomas criticised coastwatchers and was defamatory of the British Resident, Lieutenant Colonel Marchant. Mackenzie’s memo makes clear, however, that Thomas was reporting the views of the D2, Lieutenant Colonel E.J. Buckley, US Marines, an officer whose attitude, Mackenzie stressed, ‘is not shared by the great majority of US staff officers’.34
Mackenzie, Feldt and McManus looked into this case and recommended to Long that Pryce-Jones be recalled. On 4 December, Feldt wrote to Long advising him that the difficulties had arisen mainly because there was no single FERDINAND authority co-located with the Marine HQ. For some reason, Pryce-Jones had remained at Lunga on Guadalcanal, while the most important FERDINAND client, the Commander of Air Operations Solomons (COMAIRSOLS), had moved his HQ to Munda on New Georgia. COMAIRSOLS’ requirements were for immediate action to counter air activity reported by coastwatchers. The separation of the two made this impossible, and Pryce-Jones should have moved to Munda. This had also led to a situation where three deputy supervising intelligence officer representatives, call signs PWD, RJH and RAR, were now all sending conflicting requirements and passing traffic that was confusing the coastwatchers in the jungle.35
The inquiry advised Long that the failure of Pryce-Jones to remain at the spearhead, now at Munda, was a breach of the written undertaking the RAN had provided to the United States that FERDINAND’s deputy supervising intelligence officer would stay with the frontline HQ.36
Pryce-Jones’s decision to stay at Lunga, which led to multiple FERDINAND sites transmitting identical messages to US commands, provided the Japanese with a SIGINT bonanza, as the same message travelled across different circuits operating at different levels of protection. The FERDINAND codes were already suspect, and there can be little doubt that the habit of sending a message in low-grade codes before re-encrypting them in higher-grade codes, as Marchant’s network control station was doing, compromised the cryptography of both FERDINAND and the AIB. Feldt, Mackenzie and McManus were concerned that this was in fact the case.37
The inquiry also reported that Pryce-Jones’s decision to stay at Lunga was seen by many coastwatchers in the field as being driven more by his wanting to be safe and comfortable than any operational necessity. They felt that Pryce-Jones had let them and FERDINAND down. Some US officers echoed this criticism, expressing surprise that the deputy supervising intelligence officer had not gone to Munda.38 On the advice of Mackenzie, Feldt and McManus, Macfarlane replaced Pryce-Jones in January 1944.
This takes us north, to New Ireland, where FERDINAND successfully inserted Captain H.J. Murray, O.D. McNicol, Sergeant R.M. Dolby and Corporal R.J. Cream. The party landed successfully, established a secure operating base and began to contact friendly locals. As word spread, however, the Japanese became aware of the party’s presence and it had to be withdrawn on 26 November 1943.39 Personnel from this party subsequently guided US military personnel on reconnaissance patrols onto other islands, including Boang. There, the mission failed because the locals, frightened of being on the losing side, were informing the Japanese and the Australians of each other’s positions and activities.40
The presence of white men on an island left the locals in an invidious position, and this situation was created every time a FERDINAND or AIB party was inserted. Complicating this were the existing intertribal and village rivalries, especially as people were displaced by pro-Japanese elements taking advantage of the situation to grab land. The result was intertribal wars on Bougainville, New Ireland and New Britain, and in New Guinea as well. This severely reduced the safety of any FERDINAND or AIB party.41 As we will see, the image of the happy, loyal local resident was a chimera. Happy loyalty was reserved for the strongest power in the local area, depending on how long that power was present and how well it paid.42 With two out of the three, the Japanese were in the stronger position.
Despite the increasing complexity of the operating environment and the apparent failures, FERDINAND at least got parties away. A US reconnaissance of Nissan Island conducted without coastwatcher support resulted in three killed and the survivors chased into the sea.43
By the end of 1943, Rabaul was virtually isolated and the Japanese garrisons, what was left of them, on the various islands no longer presented any threat to Allied forces or plans. The war moved on as US forces careened up the Pacific and planned their return to the Philippines. There was now no role for FERDINAND in the South Pacific Area, and attention in Australia was directed almost exclusively back to the SWPA and Bougainville.