On 18 May 1942, a message from Pacific Fleet headquarters cancelled the submarine Cachalot’s previous orders. The new instruction directed her towards an area north-west of Midway from which enemy aircraft were expected to be launched. She was to patrol without revealing her presence. M19
The message is a useful place to begin. It shows that an expectation of an attack on Midway was already influencing operations. It does not prove that the predicted launching position was correct, or that headquarters knew every part of the Japanese plan. Warning had become strong enough to act upon before the picture was complete.
Making sense of fragments
Communications intelligence brought together different kinds of work. Recovering the contents of coded messages was one part. Studying which stations and units communicated, and how that activity changed, could also reveal preparations for an operation. The resulting information still needed interpretation. M42
Frederick Parker’s later NSA history describes intelligence centres sharing recovered Japanese naval messages with one another and with the commands they served. Some texts were only partially recovered, leaving intelligence staffs to draw meaning from incomplete material. His account also describes the appearance of tactical call signs and exercise traffic as indications of approaching movements. M42
This was a continuing exchange of work, rather than a single moment when an entire enemy plan became readable. The wartime headquarters record supports that caution: the 26 May running estimate described both interception and decoding as incomplete. M19
What did “AF” mean?
Japanese references to “AF” became central to the warning. Parker’s account describes Pacific intelligence centres identifying it with Midway before the famous story about the island’s water supply. M42
In that story, Midway was instructed to send an uncoded message reporting a water shortage. A subsequent Japanese report about water at AF would help confirm the identification. Parker recounts a translation circulated by the Melbourne centre on 22 May, but explicitly notes the lack of direct evidence for the preceding deception episode in his account. His history is a later reconstruction, not the original intercepted telegram; we have not inspected that telegram. M42
The distinction changes how the episode should be told. A striking story of confirmation must not erase the evidence already being assembled—or the 18 May order that had already sent a submarine towards an expected threat off Midway. M19
A warning with moving dates
Identifying an objective did not settle the timetable. A 20 May message to the naval air station at Midway warned of an attack after 25 May, possibly around 30 May, from as many as four carriers. Those were contemporary estimates. They were not the dates on which the battle eventually occurred. M19
The record lets us follow preparations without granting the people making them knowledge of the outcome. The 23 May entry called for aircraft searches north-west from Midway to a distance of 600 miles, beginning the following day. On 26 May, reports and searches were still being used to investigate a possible enemy presence. The intelligence advantage had not removed the ocean from the problem. M19
Using the secret
There was another difficulty: getting information to people who needed it without exposing its source. The 26 May record discusses that tension directly. Protecting a valuable means of learning about the enemy mattered, but so did sending warnings in time for commanders to use them. M19
By 27 May, Operation Plan 29-42 placed the American striking forces initially north-east of Midway and directed them to seek an advantage against the Japanese carriers while their aircraft were attacking the island. Intelligence had helped make such a deployment possible. The plan still treated enemy intentions as estimates and still guarded against unacceptable American carrier losses. M26
The achievement was substantial: warning offered a chance to meet an approaching offensive with a prepared response. Turning that chance into a victory would require serviceable ships, searching aircraft and attacks that reached their targets. The next part of the story begins with assembling those forces.