War of Shadows
Page 39
THE FIRST AMERICAN soldiers entered Rome late on the night of June 4, 1944. By morning, the last German troops abandoned the city. “The first of the Axis capitals is now in our hands,” Franklin Roosevelt exulted in a national radio broadcast. “One up and two to go!” By the next day, Allied troops were fighting their way onto the beaches at Normandy, and the liberation of Rome slipped to the margins of news coverage.23
An MI6 agent known as “92,700” was in Rome soon after. At least one of his tasks was counterintelligence: tracking down Axis espionage successes, especially against Britain, so that leaks could be plugged and enemy intelligence efforts foiled. On July 22 he cabled his superiors: Italy’s Servizio Informazioni Militari, SIM, had penetrated both the prewar British embassy and the legation to the Vatican and had made copies of British ciphers, he wrote. Copies of his cable went to Menzies and to Russell Dudley-Smith.
From London, Agent 92,700 received orders: We need to know the “exact means and degree” of the infiltrations. Extract more information from your SIM contacts.24
By mid-August, he and another MI6 agent, 32,000, had uncovered the responsible unit within SIM: Manfredi Talamo’s P Squad. They interrogated four of Talamo’s noncommissioned officers, as well as an Italian who’d worked for the British embassy to the Vatican and had been recruited to remove cipher books from the safe. Talamo had provided the key, taken the books, and returned them within an hour and a half, the man explained.
A cable from 32,000 naturally stressed the penetration of the British embassy and confirmed that all of Major Valentine Vivian’s 1937 warnings had been meticulously ignored.25 But two longer reports by the MI6 team gave a much fuller picture of the P Squad’s operations against the diplomatic missions of nearly twenty countries. The real number was likely higher; the MI6 agents had found only four of Talamo’s men, and their memories were incomplete. Still, the two reports are the best surviving account of Talamo’s work, since the P Squad’s own records were burned at the time of the Italian surrender and the German takeover of Rome in September 1943.
Usually, the P Squad men could tell their British interrogators little or nothing about the intelligence gained through their work or its effect on the war. There was one exception.26
TWO HANDSOME BUILDINGS stood just off Via Vittorio Veneto, the main avenue of Rome’s Ludovisi district. Greek and Roman statues stood on the corners of their roofs, including a first-century figure of Artemis, Greek goddess of the hunt, on one building and a second-century figure of a satyr atop the other. Nearly two thousand years before, the area had been gardens owned by the emperors of Rome. Ludovico Ludovisi, a seventeenth-century cardinal, turned it into his estate, decorated partly with classical sculpture dug up on the grounds. His nineteenth-century heir sold off most of the estate for development, renovated the cardinal’s palace, built two identical houses next to it for his two sons, and promptly went bankrupt. In 1941, the palace was an office of the Fascist Party. The two houses were the US embassy.27
In one of them, Colonel Norman Fiske, the US military attaché, had his office. Fiske lived around the corner at the Hotel Ambasciatori, which made it easier for the P Squad’s noncommissioned officers and the four Italian employees of the embassy who were on Talamo’s payroll to keep track of when he was leaving his residence for his office.28
Sometime in the last weeks before Pearl Harbor, before time ran out, Talamo found the moment of opportunity for another of his precisely planned operations. He or one of his men entered the building, climbed the wide marble staircase in the entry hall, unlocked Fiske’s safe, and studied the placement of the volumes inside. Then he removed the codebook and the cipher tables for the Military Intelligence Code and for the Confidential Code.29 (Attachés used the former for their most secret messages, the latter for lower-grade material.) The books were rushed to SIM’s photo studio, photographed, and returned precisely to their places. Fiske never suspected.
“As a result… at the time of the Battle of El Alamein, the Italians were able to intercept all dispatches of the U.S. military observer of the Eighth Army, Feller [sic], who was then at Cairo,” MI6’s agents wrote, based on what Talamo’s men told them.30 This was not entirely accurate. The codebooks unlocked not only Fellers’s messages but also General Maxwell’s and the traffic from Washington to Cairo—all from before the first battle at El Alamein, not during it.
Nonetheless, the MI6 investigation sealed the case: As William Friedman and Solomon Kullback had suspected, codebreakers had not puzzled out the codebook. The code had been stolen intact. As no one in Washington or Bletchley Park had suspected, the Italians—not the Germans—had pinched it.
There were no traitors in the strictest sense of the word. The embassy employees recruited by Talamo were Italians, not Americans. If they were able to obtain the keys to the safe and the building, the fault lay with their employer.
The treachery, again, was carelessness. As far back as 1936, the US ambassador to France, William Bullitt, had sent a series of warnings to the State Department about abysmal code and cipher security: about the antiquated safes used to store secret documents at his embassy, about the code clerk hired only because he was an American in Paris, about the even poorer situation at the US embassies in London and Berlin.31 The warnings had the same effect as Major Vivian’s advice to the British Foreign Office: nothing was done.
In June 1941, the State Department launched its own study of its cryptographic security. Five experts, including Friedman, confirmed Bullitt’s concerns and added more. Codes and ciphers, for instance, were sent to embassies abroad in diplomatic pouches. The bags were either closed with locks that could be “picked in a few minutes” or sewn closed, even though the stitches could be removed, the contents photographed, and the seams resewn “so as to show no signs of tampering.” The means for storing codes and ciphers at most embassies abroad were “woefully inadequate,” the experts said. “No system for secret communication,” regardless of how well designed it was, how statistically unbreakable it might seem, “can be considered safe for use” if it could easily be stolen, they wrote, underlining every word in a plea for attention.
The State Department had to install safes with combination locks at all embassies, Friedman and his colleagues said. It had to make sure that “only suitably paid American citizens” were employed as guards at embassies, especially “when those offices are closed between office hours.”32
If State applied those measures somewhere in the world, it wasn’t Rome, the capital of a country that could soon be at war with the United States. In fact, at the end of 1943, Friedman and three colleagues conducted a follow-up study. The information they managed to get from overseas stations suggested that the “physical security” of codes was still “inadequate” at many posts abroad.33
A hint about Colonel Fiske’s own regard for security appears in Ciano’s diary, on October 31, 1941. He refers to Marchesa Fanny Patrizi—the American-born wife of an Italian aristocrat—as “the lover of the United States’ military attaché.” Ciano had heard insider gossip that SIM would arrest her on suspicion of spying for America.34 So it is unlikely that Patrizi was also working for Talamo—though double-dealing is as possible in espionage as in romance. But an attaché’s affair with a woman who’d become an Italian aristocrat hardly showed rigorous concern for the military secrets he was supposed to guard.
The photos of the codebooks and cipher tables were delivered to an eighteenth-century apartment building on a narrow, cobblestoned street in a working-class neighborhood near Rome’s famous Trevi Fountain. The only sign on the building was the address, Via Poli 48. That was the location of SIM’s Cryptographic Bureau.35
It would take time for wireless operators at SIM’s intercept station at Fort Braschi, on a hill northwest of the Vatican, to identify messages in the purloined codes, and for the Cryptographic Bureau’s meager staff to get the knack of using it. In the meantime, Colonel Fiske would stand in a crowd at Palazzo Venezia and listen to M
ussolini declare war on the United States, after which Fiske returned to his office, took his codebooks from his safe, and destroyed them.36
By December 22, 1941, at the latest, SIM’s codebreakers had succeeded in reading a message in the US Military Intelligence Code. It had taken them four days to decipher Bonner Fellers’s Cable 406 from Cairo, in which he lamented that the British Eighth Army had failed to destroy Rommel’s forces in Cyrenaica despite having a three-to-one advantage in tanks and despite the British successfully decoding Rommel’s orders to his forces. Cavallero received the translation. He told SIM’s General Cesare Amè to warn Rommel immediately that “all his communications are being intercepted by the English.”37
Soon SIM was supplying the German command in Africa with verbatim translations of radiograms from Cairo. An SIM report, captured intact in Rome by British or American agents, shows the Italian agency boasting of its role in Rommel’s counterattack in late January 1942. At least one channel for passing the information was General Enno von Rintelen, the German military attaché in Rome.38
Neither Talamo’s men nor former staffers at the SIM Cryptographic Bureau interrogated in Rome could say whether SIM shared the codebooks and cipher tables with the Germans, allowing the latter to decode American messages themselves. Tiltman’s success in cracking the code used by Japanese military attachés provided the answer.39
In the spring of 1941, Colonel Tashei Hayashi took a train across the Soviet Union to reach Berlin. Hayashi was the head of the Japanese army’s cryptological agency, and wanted to trade information on American and British codes. Whatever information his hosts shared did not include advance notice of Operation Barbarossa. The German invasion of the Soviet Union left Hayashi stranded in Berlin.40 As an occasional liaison on codebreaking, he used the Japanese military attaché code to radio Tokyo. Late in the war, Allied cryptologists began going through old messages in the code, from before Tiltman cracked it. Some of those messages filled in missing pieces of the Bluebird story.
In early January 1942, Hayashi let his agency in Tokyo know about the “U.S. military attachés’ cipher” that had been “received” in Rome. He got back a radiogram asking for more information.41 In a lengthy response, Hayashi gave a precise description of the Military Intelligence Code and the tables used to encipher it. Since Japan’s Pacific offensive had left no US military attachés within range of Japanese intercept stations, Hayashi saw no reason to send the code itself.42
Virtually at the same time, the German high command’s own cryptography agency, OKW/Chi, sent an order to its intercept station at Lauf, near Nuremberg. Interception of messages between Cairo and Washington should be “as complete as possible,” it said. Listen for messages addressed to MILID or AGWAR in Washington, and to the “military attaché” at the US embassy in Cairo, the order said. MILID stood for Military Intelligence Division, Fellers’s usual addressee. AGWAR meant Adjutant General, War Department, the standard recipient of Maxwell’s messages.43
The evidence is clear: by mid-January, SIM had shared all of the material it had taken from Fiske’s safe with Japanese and German codebreakers. German intercept operators aggressively searched through the endless stream of encoded radio messages for the Cairo-Washington traffic. Rommel would get the information from both Rome and Berlin.
And not only Rommel. The Luftwaffe command received a summary of three cables on the negotiations between the United States and the Soviet Union on opening a route from India to Turkmenistan in Central Asia to deliver American warplanes. While described as coming from the “American military attaché in Cairo,” they were almost certainly from Maxwell, the American responsible for the strategic supply route.44 The SIM Cryptographic Bureau translated the first half of Fellers’s chronicle of the February 4 coup at Abdin Palace in Cairo; the German ambassador in Rome forwarded it to Berlin.45
The United States had military attachés in more places, and obtaining their code gave the Axis agencies access to their radioed reports too. In February 1942, for instance, Ciano wrote that he’d given the German ambassador the text of a cable from the US attaché in the Soviet Union. It complained “about the failure to deliver arms promised by the United States, and says that if the USSR is not aided immediately and properly she will have to consider capitulating.”46
But nowhere else did the American military’s representatives have the kind of direct access to the battlefield that Fellers had, not to mention his dedication and his staff. The traffic from his code room also included Maxwell’s constant updates on America’s global effort to provide arms to its allies. Cairo was the richest vein for the Axis eavesdroppers.
At some point in the first half of February—possibly between when Fellers sent the first and second halves of his report on the coup—the messages from Cairo evidently became unreadable to the Germans and Italians.47 The cipher tables had apparently changed—either as a matter of routine, or because Fellers himself had complained that code clerks at the War Department had become so careless in their use of the previous tables that they were endangering the code.48
Before receiving the codebooks and cipher tables photographed by Talamo, the chief codebreaker at OKW/Chi had considered the code used by US military attachés to be “unsolvable, and had laid it aside.”49 Now, though, German and Italian cryptologists had the codebook and knew how the cipher worked. They knew names that repeated in the Cairo cables, like “Rommel,” “Malta,” and “Tobruk,” and phrases, such as “RAF intelligence” and “Seventh Indian Brigade,” and they had a wealth of long messages with which to work. Once cracked, the code remained forever structurally weak. By April 24, after devoting a full week’s work to it, OKW/Chi was able to read Fellers’s cable saying that the British would not be able to attack in Cyrenaica until June, which would be too late to save the besieged island of Malta.50
The Germans got faster as they filled in the cipher tables. They went back and read old radiograms. They sent messages that too obviously depended on one “good source” in Cairo, but their messages were enciphered in Enigma, and they knew it was a mathematic certainty that Enigma could not be cracked. At some point, SIM’s Cryptographic Bureau also got back into the Military Intelligence Code; another SIM report would boast of the agency’s role in Rommel’s march from Gazala toward the Nile.51 The Italian signals based on Fellers’s cables passed almost unnoticed at Bletchley Park.52 The German ones finally pointed directly to the source. While Rommel’s and Auchinleck’s tanks clashed in the desert and men died, a silent war of minds was fought. On one side were Welchman, Storey, and Dudley Smith and the small army of Bletchley Park; on the other side were Axis codebreakers, who were only able to fight on this front because of Talamo’s opening gambit. Welchman defeated Talamo, and then Auchinleck defeated Rommel.
ONE PIECE OF the Bluebird puzzle remained unsolved: exactly when Talamo raided the US embassy. Talamo, a man who cared about details, would have remembered. He was not available when MI6 agents 92,700 and 32,000 got to Rome.
In September 1943, when the king and Badoglio fled Rome and the Nazis seized control, two Italian resistance movements formed. One was mainly civilian, dominated by the Communists. The other, smaller organization was the royalist Fronte Militare Clandestino di Resistenza, the Clandestine Military Front.
Lieutenant Colonel Manfredi Talamo, loyal to king and country, joined the Military Front and went into hiding. The new commander of the SS forces and German police in Rome was Herbert Kappler, hitherto the Gestapo representative at the German embassy, whose friend Kurt Sauer had been executed as a spy based on Talamo’s evidence. With the help of informants, the Gestapo arrested Talamo in October 1943. Kappler personally ordered excruciating tortures. Talamo gave away no information on the resistance.
On March 23, 1944, a cell of the civilian resistance detonated a bomb as a column of German military police marched down Via Rasella, another narrow cobblestoned street near the Trevi Fountain. Thirty uniformed Germans died immediately, three more by the nex
t day. Kappler recommended killing ten male Italians for each soldier who had died. Field Marshal Kesselring, commander of the German forces in Italy, authorized the massacre.
Kappler’s Security Police emptied jail cells of political prisoners and dozens of Jews. To fill out their quota, they pulled civilians off the street. Miscounting, they rounded up 335 people. Kappler decided to kill the extra five as well. The hostages were taken to the Ardeatine Caves, the remains of ancient Christian catacombs, just outside the city. In the last act of the failed bid to renew the Roman Empire, the catacombs would receive new martyrs. The Security Police shot the hostages, then blew up the entrance to the caves, creating a mass tomb.
When the SS men had entered the Regina Coeli prison to collect victims for the massacre, the first name they had called out was Manfredi Talamo.
Talamo had been a secret agent, though he was not smuggled into a foreign country and did not operate behind enemy lines. Most of his operations took place within walking distance of his office. His raid on the US embassy produced, for half a year, the best intelligence source the Axis had during the war. The evidence of his career suggests that his ideology was military loyalty to his country, not to Fascism, certainly not to Nazism. Yet his loyal service brought the Nazis close to conquering the Middle East.
Talamo’s body was left in a mass grave, along with the bodies of the other victims: Jews, political prisoners, and men who’d simply been on the wrong street at the wrong moment. The last months of Talamo’s life and the manner of his death transformed him in Italian memory into a hero of resistance to the Nazis. Yet the moral meaning of his career may be a cipher too complex ever to be solved. The memory of him as martyr erased the riddles of his life.53