May 1941, and he was at Pearl Harbor, setting up his Combat Intelligence Unit. Known in the trade as station “Hypo,” it listened to Japanese traffic along with another station at Cavite (“Cash”) and the home base in Washington (“Negat”). Technically Washington was headquarters, but the three stations really operated as a unit. Code-breaking was still a small world; the men in the business had worked together for years, sharing their findings, their theories, their hunches.
None of it did any good December 7, 1941. Japanese naval traffic had petered out, and Tokyo had also changed the key to JN-25, throwing everyone off for a few days. But the circuits soon started up again; the new key was solved by December 10, and once more there was plenty of good listening. Rochefort had spent the fall concentrating on the Japanese flag officers’ system; now his unit joined the other stations in attacking JN-25.
Not that anyone understood much directly. The actual process of code-breaking rarely yielded more than 10%-15% of a message. The rest was all analysis: perhaps the name of a ship suggested another usually with it … or a past operation might help next time when the place names cropped up again … or a familiar operator’s touch might give the source away; call letters could change, but a particular operator’s “fist”—never.
It took a curious combination of talents to fathom all this—ideally a man should be sailor, linguist and puzzle fiend in about equal proportion—so it was not surprising that the Combat Intelligence Unit was known as an outfit of individualists. Nominally under the 14th Naval District, actually they were left on their own to exercise their special brand of genius. They lived for their work, and as the Japanese offensive rolled on, they got plenty of it. Gradually they absorbed a remarkable amount of unique knowledge.
One day toward the middle of April a totally unexpected message arrived from Washington. It was, in fact, the first and only communication Rochefort ever received directly from Admiral Ernest J. King, Commander in Chief of the U.S. Fleet. COMINCH wanted Rochefort to give, based on the current flow of intercepts, his long-range estimate of Japanese naval intentions.
It was unusual, to say the least, for King to dip so far down the line in asking a question; yet it was understandable too at this particular moment. The Japanese had just raided Ceylon. Did this mean a major shift in the war toward India, or would they soon be back in the Pacific? A great deal depended on the answer.
Rochefort came up with a four-point estimate, reflecting the combined opinion of his staff: (1) Japanese operations in the Indian Ocean were over; the fleet was on its way home; (2) they weren’t going to attack Australia; (3) they were planning an operation south of Rabaul, with forces already assigned; (4) there were signs of something else brewing in the Pacific, but he couldn’t say when or where.
By late April the missing pieces were beginning to fall into place. The operation south of Rabaul was definitely aimed at Port Moresby—U.S. units headed for the Coral Sea. The other operation remained something of a mystery, but the Central Pacific seemed a strong possibility, and it was hard to look at the map without noticing Midway. Serving at various times as a coaling station, commercial cable relay point, and Pan American Clipper rest stop, Midway was currently a very useful but highly exposed Navy patrol plane base. It consisted of just two tiny islets, Sand and Eastern, lying side by side in a beautiful blue lagoon. There was nothing else worth anything for a thousand miles around. Midway was out in the Pacific alone—all alone.
Admiral Nimitz made a brief visit there May 2, spending a day with the officers in charge, Navy Commander Cyril T. Simard and Marine Lieutenant Colonel Harold Shannon. Nimitz mentioned no specific threat but did ask what they’d need to defend Midway against an attack. The two men poured out their requirements.
“If I get you all these things you say you need, then can you defend Midway against a major amphibious assault?”
“Yes, sir.”
Another week passed, and at Pearl Harbor Rochefort began to get a better picture. As always, it was a matter of piecing together the scraps gathered by all three stations—Pearl, Washington and a new unit in Australia replacing the one lost with the Philippines. Even so, there was nothing conclusive. It was all spotty … but nonetheless revealing. For instance, it was clear that Admiral Kondo’s Battleship Division 3 was talking a lot to Nagumo’s First Carrier Force … that certain transports were getting orders to Saipan, and BatDiv 3 was talking to them too … that Cruiser Division 7 was also heading for Saipan. Was there any connection between these things?
The traffic never gave any clear-cut answer. Often the pertinent messages were days apart. Yet Joe Rochefort had a knack of remembering them all, and as he pieced the picture together, it pointed to a massive operation involving most of the Imperial Combined Fleet.
But again, where? The Japanese never mentioned any particular spot, but gradually one name began cropping up more and more often—sometimes as a destination, sometimes as a place calling for certain equipment. In a sort of code-within-the-code this place was always called “AF.”
Rochefort’s little group racked their brains. Finally someone remembered that AF had also appeared in messages last March, at the time a couple of Japanese seaplanes made an abortive attack on Pearl Harbor. Papers flew as the whole office dug through the piles of intercepts, trying to track this vague clue down. Finally, there it was. The Japanese had indeed mentioned AF. Those seaplanes had refueled from a submarine at French Frigate Shoals, a tiny atoll lying toward Midway, and one of the messages spoke of passing near AF.
For Rochefort, that settled it. AF must lie in the Midway area … and the only place worth taking around there was Midway itself. From now on, his estimates began stressing Midway as the probable target.
Not everyone saw it that clearly. The intercepts also referred to “AL,” “AO,” “AOB”—all definitely in the Aleutian area—and some high officers, especially in Washington, thought that the real target was Alaska or the West Coast. Admiral King himself leaned for a time toward Hawaii as a possibility, and the Army Air Force was always worried about a raid on San Francisco.
It was hard to change their minds. As usual, Rochefort’s unit rarely managed to “read” more than 15% of an intercept, and to the average professional officer 15% seemed mighty little to go on. It was different in the old days when offices were smaller and the men knew each other better. Now there was tremendous turnover; whole new layers of command moved in, often staffed by men who knew little about radio intelligence. It was too much to expect them to understand the intricacies of cryptoanalysis. Nor did it help that the cryptoanalysts were so secretive. They had to be, of course, but that didn’t make it any easier to sell their wares to the bright new faces at COMINCH.
So the Navy Department had its skeptics … and so did the Army Air Force … and so, for that matter, did the high command in Hawaii.
But Rochefort also had his allies, and one of them was Nimitz’s intelligence officer, Lieutenant Commander Edwin T. Layton. He too knew the Japanese and their language well. He spent two years at the Embassy in Tokyo just before the war, then was Admiral Kimmel’s intelligence chief at the time of the Pearl Harbor disaster. Like the rest of Kimmel’s staff, Layton wondered what would become of himself when Nimitz took over; and like the others, he felt that tremendous surge of morale when Nimitz asked them all to stay with him.
Intelligence, Nimitz told Layton, was something he knew nothing about. He had only this one suggestion: He hoped Layton would approach his job as though he were Admiral Nagano, or Yamamoto, or whoever was calling the tune in Tokyo. With the information at hand, he should put himself in Japanese shoes and come up with his best estimate of their plans and intentions.
With such sailing orders, it was easy to fall under the spell of Joe Rochefort’s shop. Layton was completely sold, did his best to sell Nimitz too. The Admiral himself was half convinced, but no more.
Around May 8 Layton finally begged Nimitz to visit the Combat Intelligence Unit. Then he coul
d see for himself. Layton was sure that once the Admiral watched the place in operation—saw how carefully the information was pieced together—then he too would become a true believer.
Nimitz couldn’t come. Swamped with work on the Coral Sea battle, he was just too busy. But he did send Captain L. D. McCormick, his war plans officer. McCormick said he was busy too, but finally agreed to set aside two hours.
Down the stairs they went, and into the basement room, which for once the staff had made a halfhearted effort to clean up. On a makeshift table of planks and sawhorses Rochefort had spread all his intercepts. He patiently showed how one led to another, and how they all fitted together to form a composite picture.
McCormick was fascinated. In the end, he spent not two but three and a half hours poking around, flipping the material, asking a thousand tough, show-me questions. There were no velvet gloves on either side… .
“How do you know the Kamikawa Maru isn’t just heading for her home port?”
“Well, if she is, that would be doing a tonsillectomy through the rectum.”
In the end, McCormick came away completely convinced, and to sell McCormick was to sell Nimitz. From that day on, the Admiral was the staunchest ally Rochefort and Layton could hope to have.
Washington remained skeptical. For one thing, they still hadn’t pinned down exactly what the Japanese meant by “AF.” Rochefort was always sure it was Midway but he needed proof. Around May 10 he went to Layton with an idea. Could Midway be instructed to radio a fake message in plain English, saying their fresh-water machinery had broken down? Nimitz cheerfully went along with the ruse … Midway followed through … and two days later a Japanese intercept was picked up, reporting that AF was low on fresh water.
May 14, Nimitz filled in Major General Delos Emmons, the local Army commander, and declared a state of “Fleet Opposed Invasion.” On the 15th, he began assembling his meager collection of ships. He had only three carriers and all were thousands of miles away in the South Pacific. Rear Admiral Frank Jack Fletcher’s Task Force 17 was at Tongatabu, licking its wounds from the Coral Sea battle. The carrier Yorktown had not been sunk, as the Japanese claimed, but she was far from healthy. One bomb had ripped out her innards, and two near-misses had opened some seams. Vice Admiral William F. Halsey’s Task Force 16 was marking time off the Solomons; he had arrived with the carriers Enterprise and Hornet just too late for Coral Sea. Nimitz radioed them all to hurry back.
At Tongatabu Fletcher quickly made a few patchwork repairs—then he was on his way. East of the Solomons, Halsey got the message May 16, local time, and immediately began collecting his scattered ships. He was rounding up the last of them on the 18th when a new message arrived from Nimitz. This one was only two words: “EXPEDITE RETURN.”
As they swung northeast for Pearl, the Enterprise’s navigation officer, Commander Richard Ruble, had a satisfying thought. He recalled how, on the 15th, they had discovered an enemy patrol plane snooping about on the western horizon. Visibility was perfect, and it had plenty of time to get off a good contact report to Japanese headquarters firmly placing Task Force 16 deep in the South Pacific.
Back at Pearl Harbor the Combat Intelligence Unit labored on. Rochefort himself concentrated on the translations—so much depended on the nuances of the Japanese language even after a message was broken. He was now putting in 20-21 hours a day, curling up in some corner for occasional naps, going home “only when someone told me I ought to take a bath.” The rest of the staff worked just as hard—Jasper Holmes, slaving over his ship positions; Tommy Dyer gulping keep-awake pills by the fistful; the brilliant Ham Wright buried in paper.
At CINCPAC, staff officers carefully noted the information on their charts. By now the composition of Nagumo’s Striking Force was down pat—CarDivs 1 and 2, CruDiv 8, all the familiar names—but the route was more uncertain. Some clues suggested they’d come straight across; other clues said no, they’d go considerably to the north, then slant down. There was a theory they might even strike from the east. From all this, time was clearly another question mark.
There was less mystery about the Occupation Force coming up from the southwest. The route was clear; the only question was when it would arrive. Well, somebody figured, if the transport Goshu Maru left the Marshalls May 17, it could link up at Saipan on the 22nd, and the whole convoy would be at Midway on the 30th. On the other hand, there was much radio traffic about exercises, and maybe these would push the whole operation back a little… .
Carried away by the mounting excitement, it was hard for the insiders at CINCPAC to realize that some people were still dubious about the whole business. On May 16 or 17 Nimitz received a letter from General Emmons saying he appreciated the intelligence, but felt he ought to point something out. These estimates were based on enemy intentions rather than capabilities. It was safer to plan in terms of capabilities, and the Japanese were certainly capable of attacking Hawaii.
Nimitz called Commander Layton in—the Admiral’s blue eyes twinkling with amusement as Layton exploded with exasperation. Yet Emmons’s misgivings were understandable. Like most others, he knew little about Rochefort’s work. Beyond that, there was the matter of Nimitz’s order directing a state of “Fleet Opposed Invasion.” This put all his Seventh Air Force bombers under CINCPAC control. But if the Navy was wrong and the Japanese did come to Hawaii, it would be Emmons, not Nimitz, who would feel the full blast from the War Department at being caught with no planes on hand.
Nimitz was not about to reverse course, but he did make a move typical of this most careful and conciliatory of men. He assigned one of his staff, Captain J. M. Steele, to the specific job of keeping an eye on the Combat Intelligence Unit’s material. Steele became a sort of “devil’s advocate,” deliberately challenging every estimate, deliberately making Rochefort and Layton back up every point.
Steele really threw himself into the job. Layton rued the day it ever happened, but from Nimitz’s point of view the assignment served two very useful purposes. First, it did something specific about General Emmons’s letter without any real change in course; second, it provided a genuine check just in case they all were wrong, and the Japanese really did have an extra trick or two.
No more time for debating. The latest intercepts indicated that the attack might come as early as May 28. Nimitz warned Midway, and preparations at Pearl moved into high gear. Decisions were made in hours that normally might take days. What to do about the Aleutian side show? Nimitz decided to counter it with a small force that would keep the Japanese occupied and guard his own flank. How to use the submarines? Put them in an arc west of Midway to try and intercept the Japanese fleet coming in. What about the battleships on the West Coast? Keep them there; they only got in the way. How about the Yorktown’s damage? Take a chance on repairing her at Pearl Harbor; an urgent message listed the things they’d need: arresting gear … 50 pneumercator gauges … and so on. In any other navy it might have seemed strange that this priority list also included a new freezer for the soda fountain.
The Army was now in high gear too. On May 18 the entire Seventh Air Force went on special alert for an attack on Midway or (the old fear persisted) an air raid on Hawaii. B-17s began pouring in from the West Coast. At the same time, big new PBYs were arriving for the Navy—these, in turn, hurried on to Midway.
They were all part of a sudden rush to beef up the place—a far cry from the old days when Midway seemed at the bottom of every supply depot’s priority list. For months the Marines just couldn’t get any barbed wire; now they got “hundreds of miles of it.” Sandbags, antiaircraft guns, impregnated clothing, PT boats, reinforcements, all poured out in a steady stream. When the big railroad ferry Kittyhawk left for Midway on May 23, her decks bulged with fighters and dive bombers, even five light tanks.
But how to use this growing strength? And how best to coordinate it with the three precious carriers racing up from the South Pacific? On the 23rd Nimitz asked Admiral Pat Bellinger, his patrol plane chie
f, and Captain A. C. Davis, his staff aviation officer, to confer with the Army immediately on how to make best use of all the planes available—both Army and Navy, ashore and afloat. They should think in terms of possibly five Japanese carriers, and they must hurry, “TIME IS EVERYTHING.”
The submarines were already moving out. The big Nautilus was lying in dry dock when her skipper Lieutenant Commander Bill Brockman was asked how soon he could get going. Brockman said about 48 hours. Everyone worked around the clock, and at 9:00 A.M. on May 24 she slipped down the channel, bound for her station in the arc west of Midway. Eleven other subs were out there too; seven more patrolled to the east in case Nagumo tried an unexpected dash for Hawaii.
May 25, there was new excitement at Pearl Harbor. This day the Combat Intelligence Unit came through with its hottest item yet—a long Japanese intercept that really laid it on the line. This message ticked off the various units, the ships, the captains, the course, the launching time—everything.
Rochefort was hard at work on the translation when word came that Nimitz wanted to see him personally at a certain time. Completely absorbed and wanting to finish it, he turned up at the Admiral’s office half an hour late. This was too much for even the placid Nimitz, and his manner was chilly, to say the least. But he soon warmed up as Rochefort gave him, point by point, the exact battle order and operating plan of the Japanese Striking Force.
Nimitz rushed a new message to Midway, giving all the details. There were only two bright spots: Nagumo’s Force no longer included the carrier Zuikaku; and the date of the attack would be later than they originally thought—it should fall some time around June 3-5.
Washington still had its skeptics. Now the big fear was a Japanese ruse; the enemy might be just feeding Nimitz these messages to cover up a major raid on Hawaii or the West Coast. After all, they used fake radio signals to fool CINCPAC just before the attack on Pearl Harbor. Maybe this was the same sort of thing. Nimitz should plan his defenses with this threat in mind.