THE CODEBREAKERS
Page 95
Messages so enciphered appear to have been radioed to Moscow. Abel had a shortwave set in his Brooklyn studio and a receiver in his hotel room. He told his lieutenant that he tape-recorded incoming messages, then took them down on paper and deciphered them. After his arrest, U.S. government agents listened in accordance with a broadcast schedule found in the tip of a hollowed-out pencil, also thrown into Abel’s wastebasket, and twice picked up messages in five-digit groups. British security police found a signal plan with the one-time pads in the Krogers’ lighter. They likewise listened in accordance with its instructions. At 12:32 a.m. January 9, 1961, tuned to 17,080 kilocycles, they heard the call-sign dash dot dot. Eighteen minutes later they heard the same sign at 14,755 kilocycles. On January 18, at 6:38 a.m. at 6,340 kilocycles, they heard call-sign 277. Less than an hour later, they heard it again at 8,888 kilocycles. Direction-finders plotted the source of the signals as the Moscow area. Lonsdale had a high-speed transmission device that would send Morse at 240 words per minute; probably it recorded his message and then spewed it out in a spurt.
A sheet of a one-time pad captured on Communist spies in Japan, 1961. One side may be for enciphering, the other for deciphering
Communication seems to be fairly frequent. The Kroger signal plan provided for contact Tuesdays, Wednesdays, Fridays, and Saturdays. This frequency may explain why both Abel and Mrs. Kroger had enciphered messages on them when they were caught. Abel tried to stuff his up a sleeve. Mrs. Kroger asked to be allowed to stoke the furnace before leaving the house for a long period of questioning, but when an envelope containing a single sheet of paper with a block of typed numbers was removed from her handbag, “Mrs. Kroger,” a security officer said, “showed no further desire to stoke the boiler.”
Radio was supplemented by microdots. Lonsdale hid a microdot reader in a can of talcum powder. Abel made his own microdots by reducing the 35-millimeter negative with a lens having a very short focal length. To preserve legibility in these great reductions, he used spectroscopic film, which is available at many camera stores and which can resolve 1,000 lines per millimeter. He undid the stapling of Better Homes & Gardens and American Home, inserted the thin strips of microdots between sections, rebound the magazines, and mailed them off to a prearranged general delivery number in Paris. For some reason the hidden messages were not received and Moscow told Abel to discontinue them. However, microdots brought the spies plaintext letters from home.
Though the one-time pad is the standard method for radio communication between top agents and Moscow, other systems serve the internal needs of secret communications within Communist spy rings. The rule here seems to be that where Russians have devised the systems, they are top-notch, and where local Communists who are natives of a country have done so, they can be solved—often with disastrous effect. In 1955, for example, Swedish counterespionage police noticed that a chauffeur at the Czech legation went to the Stockholm railroad station each night to buy copies of the newspapers Kurier and Tidning, both published in the provincial city of Karlskoga, where munitions are manufactured. Studying the papers, police noticed a number of oddly worded announcements. When they inserted similar advertisements, using the same identifying words, they received responses from several people who turned out to be Red agents. Eventually a ring that operated in five cities was broken up and four Communist satellite diplomats declared persona non grata.
The most catastrophic instance of the eggshell ciphers of local Communists took place in Iran. On the night of August 16, 1954, Iranian security police arrested Ali Abbasi, a former Army captain who had come under suspicion because of his activities in the Red Tudeh party. In the suitcase he was carrying as he came out of a house in Teheran, they found a complete plan of Shah Mohammed Reza Pahlevi’s summer palace, showing guard posts and the number of men stationed at each, top-secret documents from Army files, reports on the disposition of artillery along Iran’s Russian border, two notebooks in what were obviously cipher, and another with page after page of what appeared to be trigonometric equations, replete with the Greek letters beloved of mathematicians and the abbreviations for “secant,” “cosine,” “tangent,” and “cotangent.” The problem was that mathematically the formulas made no sense at all.
Colonel Mostafa Amjadi, chief of the intelligence directorate of the Teheran military governate, and another colonel in the Iranian Army went to work on the three notebooks. By August 30, they had cracked the two overt codes, but extracted only meager information from them. Meanwhile, Abbasi decided to talk. He revealed that the Tudeh party had riddled the Iranian Army with about 400 agents and that their names were listed in a mathematical cipher. This was the trigonometric system which Amjadi and his colleague were even then struggling with, but Abbasi warned that the system was so complex that it could be read only by its inventor, Lieutenant Colonel Jamsheed Mobasheri, an artillery officer regarded by his friends as something of a mathematical genius.
The imitation “trigonometric” cipher of Red agents in Iran
Mobasheri was picked up for questioning. Instead of revealing the key, he tried to puncture a vein with a rusty nail. The two colonel-cryptanalysts worked steadily 24 hours a day in overlapping shifts of 12 hours each. Mobasheri was again interrogated, and, now that the first shock of arrest had worn off, his pride of authorship in his cipher almost overcame his loyalty to Communism and he twice agreed to reveal the method—only to change his mind both times. The Iranian government quietly asked other countries if they would help in the solution. Meanwhile, one of the colonels formulated a theory as to Mobasheri’s system and interviewed him, hoping to get some clues from the inventor’s reactions. Mobasheri stubbornly insisted that the system could not be broken, but just as the colonel was leaving the cell Mobasheri’s appreciation for an intelligent analysis broke through, and he admitted that the cryptanalyst was on the right track. On September 3, as an airplane was about to fly copies of the trigonometric notebook to an ally’s cryptanalysts, the two haggard colonels cracked Mobasheri’s cipher.
The roster proved to be as detailed as an official army register in describing the officers, making identification easy. But it was so extensive that it took several days to decrypt it and locate the conspirators. A week later all 400 were arrested. This enormous conspiracy, Iranian security police discovered, had not only obtained detailed information on the strength and disposition of Iran’s entire armed forces, but had wormed into vital posts that would have enabled it to assassinate on a moment’s notice members of the government from the Shah on down. It was ready either to pull a coup and set up its own Communist puppet government or to deliver the nation entire to the Soviet Union. Imperfect ciphers kept it from doing either. Instead, 26 of its leaders—including Mobasheri—were executed, and hundreds of run-of-the-mill plotters and sympathizers were jailed. A poisonous infection had been cleaned out; a year later Iran abandoned her traditional neutrality and, signing the Baghdad Pact, aligned herself with the West.
No cryptographic weakness imperils the operation of Russian spy rings. Perhaps the most striking example lies in the cipher used by Abel’s lieutenant, Reino Hayhanen. For two years after the fat, lazy, and irresponsible Hayhanen landed in New York, he did not meet Abel in person, but communicated with him by messages hidden in prearranged “drops”—a crack in the cement wall that runs from 165th to 167th streets along Jerome Avenue in the Bronx, behind a loose brick under a bridge in Central Park, under lampposts in Prospect and Fort Tryon parks. The messages were on “soft” microfilm, which Abel made by dissolving the stiff film base; this left only the soft, image-bearing emulsion, which could be squeezed into tiny places. Abel and the couriers from Moscow used hollowed-out pencils, bolts, flashlight batteries, and coins as message-containers in his drops because they would be less likely to arouse suspicion if found by accident. A horizontal mark with blue chalk at various signal areas—park fences or subway stations—meant that a message had been placed in a drop, a vertical mark that it had been retriev
ed. These were to be checked daily. Soon after Hayhanen’s arrival on October 21, 1952, he “posted” his first message. Moscow responded with an enciphered message on soft microfilm enclosed in a hollow 1948 Jefferson nickel. It read:
1. We congratulate you on a safe arrival. We confirm the receipt of your letter to the address V and the reading of letter No. 1.
2. For organization of cover we have given instructions to transmit to you three thousand in local [currency]. Consult with us prior to investing it in any kind of business, advising the character of the business.
3. According to your request, we will transmit the formula for the preparation of soft film and the news separately, together with [your] mother’s letter.
4. [It is too] early to send you the one-time pads. Encipher short letters, but do the longer ones with insertions.* All the data about yourself, place of work, address, etc., must not be transmitted in one cipher message. Transmit insertions separately.
5. The package was delivered to [your] wife personally. Everything is all right with [your] family. We wish [you] success. Greetings from the comrades. No. 1, 3 December.
Somehow the hollow nickel went astray. Most probably Hayhanen, who was sloppy in his spy work, spent it. It circulated in the economy like millions of its fellows, with none of those who passed it from hand to hand aware of its secret cargo. Then one hot morning in the summer of 1953, newsboy James Bozart, who had just received it as part of 50 cents in change from a customer on his route at 3403 Foster Avenue in Brooklyn, dropped it with four other nickels and a quarter on the staircase. When he bent over to pick it up, he found that it had split in half. One of the hollowed-out halves held a piece of microfilm five-sixteenths of an inch square, wrapped in tissue paper. “It was a picture of a file card, or an index card,” Bozart said. “There appeared to be a row of numbers on it.” He turned it over to police, who gave it to the F.B.I. , who must have immediately begun trying to crack it. They failed.
Four years later, Reino Hayhanen defected to the U.S. embassy in Paris as he was being sent back to the Soviet Union by the dissatisfied Abel for a “vacation.” He disclosed the cipher system and the keys that he had used, and in the summer of 1957, F.B.I. Russian expert Michael G. Leonard applied them to the 207 five-figure groups of the microfilm that Bozart had found. At last the F.B.I. was able to read the nickel message.
Encipherment began with bisection: cutting the plaintext in half and putting the second part first, thus burying the vulnerable beginning (which was marked by a special indicator) deep within the body of the message. There followed a substitution by straddling checkerboard, based, in this case, on the first seven letters of the Russian word SNEGOPAD (“snowfall”). It was inscribed in the top line—there was no Russian equivalent of ASINTOER, though the seven letters snegopa include the most frequent in Russian (o, at 11 per cent) and total to 40 per cent of normal Russian text. The rest of the Cyrillic alphabet and additional symbols followed below. This preliminary ciphertext was written horizontally into a columnar transposition block. It was transcribed vertically and written horizontally into a second transposition block. This one, however, had a series of steplike disruption, or D, areas. The first D area began in the top row under keynumber 1 and ran to the right side of that row. In succeeding rows it began one column to the right. When after several rows the starting point reached the right side of the block, a row was skipped and the second D area began under keynumber 2. It was constructed like its predecessor. Other D areas followed. The cipher digits being written into the second transposition block went first into the non-D area, and when this was filled the inscription continued in the steplike D areas. The final ciphertext was obtained by vertical transcription in keynumber order from this second block, in which transcription the D areas were disregarded.
The system dispensed with written keys such as Abel’s one-time pad, which helped incriminate him. Hayhanen had to remember only four basic keys—SNEGOPA(D), the first 20 letters of a Russian popular song (“The Lone Accordion”), the date of the World War II victory over Japan (3/9/1945, in the Continental style), and his personal keynumber (13, changed to 20 in 1956). The latter three keys generated the keys for the transpositions and the coordinates of the checkerboard through a process that was extremely complicated but that possessed a kind of tractive logic, was meant to be memorized, and probably would be after two or three run-throughs.
This process injected an arbitrary five-digit number at the very beginning of the key derivation, strongly influencing the end result. (This number was also inserted in a predetermined position in the cryptogram so that the decipherer would have it. In Hayhanen’s case, this position was the fifth group from the end, the position coming from the last figure, 5, of the victory date.) This group changed from message to message, so the enciphering keys, and consequently the ciphertexts of all messages enciphered in this system, would bear no exploitable relation to one another. Not only would the transposition keys differ, the very widths of the blocks would as well—this being a variable stemming from the key derivation. This kills any last hope of an analysis by comparing messages. The poor cryptanalyst would even be denied the consolation of discovering a common origin of the cryptograms through similarities in frequency counts, for the coordinates themselves would change. Any solution would thus have to be effected on the basis of a single message. It would require trying every sensible pattern of transposition until one was found that yielded a monalphabetic frequency count of the digits. The D areas significantly increase the difficulty of finding this pattern, just as the straddling effect increases the difficulty of getting a valid frequency count. The number of trials for a 1,035-digit message like the nickel one is astronomical, and, even with computers, it would probably take years. In theory the system is not unbreakable, but in practice it is. Its security could not have been more pointedly demonstrated than by the F.B.I.’s failure to solve it.
Such is the cryptology of the Soviet Union. It is interesting to contemplate its excellence. Russia herself remains “a riddle wrapped in a mystery inside an enigma.” So, when she decrees it, do her communications. The one-time pad ensures this for the bulk of her spy messages and for a fair proportion of her diplomatic and secret-police messages. Complex rotor-type cipher machines, well-designed in themselves and handled with a sophistication that changes keys after foreign cryptanalysts have reconstructed part of the wiring and the rotation pattern but before they can read any plaintext, guard other high-level diplomatic and military messages of Soviet Russia. And even when she requires a cipher to be fully mnemonic, like Hayhanen’s, she designs it so that it cannot be broken. She has solved, during the Cold War, ciphers in use at the American embassy in Moscow. Feats like these bear witness to knowledge that could only well up from a profound understanding of cryptography and cryptanalysis. Whether this comprehension springs from the scientific ability that has enabled Russia to orbit great artificial satellites, or from the decades-long experience of cryptology that the Communist dictators have had to practice for self-preservation and aggrandizement, or from the habits of secrecy and puzzling out the real meaning of things that are ingrained into every inhabitant of a totalitarian society, or from a dark-souled Slavic love of the mysterious, it has beyond question rocketed Red accomplishments in this black art to Sputnik height.
* Russkaya Kriptologiya (“Russian Cryptology”).
* During the Second Battle of the Masurian Lakes in February, in which the Russians were defeated, they used a service code called the RSK, which the Germans solved. Its nature is unknown.
* Petrov named three men who were his bosses at different times while he was section chief—Ilyin, Degtjarov, and Shevelev. Whether these were the heads of the entire, then newly formed N.K.V.D. 5th Directorate, or whether they were department heads (a possible administrative level between the section chiefs and the directorate’s chief), is not known. The former may be more likely in view of the fact that Boki’s successor, Shapiro, lasted only a month
or two before he was arrested, and three or four of Shapiro’s successors were also arrested.
† In 1933, it also had a military intelligence group, headed by a Colonel Kharkevich, a solid, impressive man who reported to both Boki and the general staff. This group appears to have later been abolished or transferred to the Army; Kharkevich himself was purged in 1938. The head—under Boki—of the O.G.P.U. group of cryptanalysts was one Gusev, possibly Sergei I. Gusev, an old revolutionary, active in secret printing, a member of the Central Committee of the Russian Communist party since 1922, and on the Praesidium of the Comintern from 1930. He too was purged in 1938.