Lonsdale soon picked up his old life, running his jukebox business, meeting Houghton, and dating a great variety of attractive girls. He was not due to repossess his flat in the White House until early November, but where he was staying was a mystery. Every night he left his offices in Wardour Street and headed westward. Arthur and I laid down strict controls on the Watcher operations after Lonsdale's return. We were determined there would be no more mistakes. Overt watching was prohibited, and strict radio silence was imposed on all operations. Jim Skardon exploded at this apparent intrusion into his empire. He was not indoctrinated for RAFTER, and could not understand why radios were prohibited. He complained to Furnival Jones but was told firmly that there were good reasons for the new policy.
Arthur and I realized it would be impossible to follow a trained and experienced intelligence officer like Lonsdale for any distance without alerting him to the fact, so we devised a new semistatic technique. Every night a team of Watchers picked him up and followed for a short distance, before peeling off. The next day Lonsdale was picked up by a new set of Watchers where the previous team had given up the chase, and followed another short distance, and so on, at successively increasing distances from his Wardour Street office. The whole operation took two weeks, and we even employed wives and volunteers from the office to supplement the Watcher staff so we never used the same faces twice. Eventually we tracked Lonsdale to 45 Cranleigh Gardens, Ruislip, in West London. Lonsdale was evidently staying with the occupants of the small house, Peter and Helen Kroger, a New Zealand couple who ran a small bookshop specializing in Americana antique books. We set up a static observation post in the house opposite, and waited, confident that none of the occupants knew of our presence.
In mid-November Lonsdale moved back into his flat in the White House, collecting his suitcase from the Midland Bank shortly beforehand. We immediately arranged for a GCHQ technician, Arthur Spencer, to move into the flat next door to begin our RAFTER operations. For the next three months Spencer scarcely set foot outside the tiny flat. We installed a noncontact tap on the mains supply feeding Lonsdale's receiver which was connected to a silent buzzer. The buzzer was worn as an earpiece by Spencer, so that even if Lonsdale used his radio set during the night, the buzzer would alert him. Whenever the buzzer sounded, Spencer tuned the RAFTER receivers, found the frequency Lonsdale was listening to, and alerted GCHQ Palmer Street. Palmer Street then relayed the signal down to GCHQ in Cheltenham. There, using our copy of Lonsdale's one-time pad, a GCHQ cryptanalyst named Bill Collins decrypted the message, and relayed it back up to London to Arthur and me in Leconfield House via an enciphered telex link.
The first time Lonsdale received a message, Bill Collins was unable to decipher it. There was no indicator group in the traffic. An indicator group is a group EN CLAIR, in other words a group from the one-time pad with no coded additive. The recipient uses this to position the message on the pad at the right place, so it can be deciphered. (After Lonsdale was arrested we discovered that the indicator group was in fact enciphered, using his real date of birth.)
Arthur and I began to wonder if, perhaps, Lonsdale realized his pads were compromised, and was using a new set brought back with him from abroad. The only thing we could do was burgle his flat and check inside the lighter again, to see if the pads had been used. Winterborn and I went in on a day when Lonsdale went to Suffolk for his jukebox business. It was a small flat, rather depressingly spartan, with barely space for more than a bed. We opened the lighter; the pads were still there, and new pages had been torn away, so they were obviously still in use. When I looked carefully I realized that Lonsdale had used more lines than were needed to encipher the message he had received from Moscow. When the message was stepped down the pad by the number of excess lines, the message read satisfactorily.
For the next two months we successfully monitored Lonsdale's biweekly messages from Moscow. Most of them concerned "the Shah," the KGB cryptonym for Houghton. Lonsdale was given specific instructions on how to handle him, which questions to ask, and what documents he should attempt to procure from Portland. But other messages were personal, containing family news about his wife and his children back in Russia. They wanted him home after five years' undercover service.
On Monday, January 2, Hollis chaired a full review of the case. Arthur argued strongly that we should allow the case to run on. He felt instinctively that Lonsdale was too valuable an illegal to be running simply the one spy, Houghton. We still knew very little about the Krogers, and their house at 45 Cranleigh Gardens, beyond the fact that shortly after Lonsdale went to stay, high-grade Chubbs and window locks were fitted to the house, including the access to the roof. For all we knew, Lonsdale might be only one part of a much larger network. Furnival Jones and I supported Arthur, and Hollis agreed to approach the Admiralty (whose secrets Houghton was betraying) to ask permission to leave Houghton unmolested for a further three months. The Admiralty agreed, and Arthur decided to minimize any further risk by running the case on without any form of physical surveillance, relying simply on our interception of Lonsdale's radio traffic to lead us to further spies.
Two days later our plan was rudely shattered. A sealed message was delivered to Hollis by Cleeve Cram, the CIA officer assigned to the American Embassy in London for liaison with MI5 .The message warned MI5 that Sniper had informed the CIA that he intended defecting to the United States on the following day, January 5. Once again, we convened in Hollis' office. There was really only one course of action. Houghton, Lonsdale, and presumably also the Krogers would all be blown by the defection. We had to arrest them before they were withdrawn. Fortunately, Houghton was due for his January meeting with Lonsdale on the Saturday, January 7, and we also knew that Lonsdale was due to receive his radio message early that morning, so we would know if Moscow sent him a warning.
Arranging the arrests was a prodigious feat of logistics, and for the next three days I barely slept. Charles Elwell, Houghton's case officer, was sent to Portland, ready to search Houghton's premises as soon as he was given word the arrests were successfully accomplished. Bill Collins came up from Cheltenham and based himself in Palmer Street, ready to decrypt Lonsdale's message the instant it came through. The Special Branch were put on standby outside Lonsdale's flat, ready to make an immediate arrest if the Moscow message sent him scurrying for cover.
On the Friday night Arthur and I gathered in the third-floor operations room in Leconfield House, ready for the all-night vigil. It was a small office, painted a ghastly Civil Service brown. It could have been a prison cell. A metal-framed bed ran along one wall, A small table stood in the middle. Cables trailed across the floor in thick, tangled bunches. Telephones linked us to Special Branch headquarters, to GCHQ, and to the DG, and a small speaker relayed to us every sound inside Lonsdale's flat in the White House.