WHAT IT IS
Station HYPO was the U.S. Navy’s signals-intelligence and cryptanalytic center at Pearl Harbor in the critical years 1941–42. Officially it was the Combat Intelligence Unit of the 14th Naval District. The radio-intelligence call-sign name HYPO used the old phonetic for the letter H (Hawaii). After reorganization it grew into Fleet Radio Unit Pacific (FRUPAC). The unit’s purpose was Pacific Fleet communications intelligence: intercept Japanese naval radio, analyze traffic, break and read operational cryptosystems — above all the additive cipher Americans called JN-25 — and get usable intelligence to the Commander in Chief, Pacific Fleet. [DOCUMENTED]
It worked in the basement of the Old Administration Building (Building 1) at the Pearl Harbor submarine base — the “Dungeon.” One entrance, an armed Marine on the door. From June 1941 the officer in charge was Lieutenant Commander (later Commander) Joseph J. Rochefort, a Japanese linguist and cryptanalyst. His chief cryptanalyst was Thomas Dyer. Product went to Admiral Chester W. Nimitz through Edwin T. Layton, the Pacific Fleet intelligence officer and a Rochefort contemporary from language training in Japan. [DOCUMENTED]
The name “HYPO” also labels the Oahu intercept lineage that began at Little Wailupe (1928), moved to Heʻeia, and later to Wahiawa. Rochefort’s analytic Combat Intelligence Unit in Building 1 consumed that intercept stream. Collectors and cryptanalysts were one system with two physical homes. [DOCUMENTED]
HOW TO READ THIS PAGE
English only (Western/U.S. topic). Credit for JN-25 recoveries was disputed in 1942 and remains argued; this page labels conflicts rather than picking a team.
PURPOSE — WHY THE STATION EXISTED
Pacific war is a radio war at ocean scale. Fleets beyond visual range coordinate by high-frequency (HF) Morse and related circuits. An enemy who can read those circuits learns where carriers are going before lookouts do. HYPO’s purpose, under OP-20-G (the Navy’s Code and Signal Section in Washington), was to turn Japanese naval radio into Pacific Fleet decisions. [DOCUMENTED]
Concrete jobs:
1. Intercept — copy IJN Morse and related emissions at shore stations. 2. Direction finding (DF) — locate transmitters by radio bearings. 3. Traffic analysis — who talks to whom, how often, on which nets, with which call signs — even when the text is unreadable. 4. Cryptanalysis — strip JN-25 additives, recover codebook meanings, translate. 5. Dissemination — get the product to Layton and Nimitz in time to move ships.
It was not a diplomatic Purple shop (that was largely Army SIS / “Magic” for Foreign Ministry traffic). It was not a public radio station. It was Fleet COMINT. [DOCUMENTED]
THE OP-20-G NETWORK
OP-20-G under the Chief of Naval Operations ran analytic centers and a string of intercept sites. Phonetic station names:
| Station | Location | Role |
|---|---|---|
| HYPO | Pearl Harbor (analytic); Oahu intercept at Heʻeia / later Wahiawa | Fleet support; after Dec 1941 the unit closest to the fighting |
| CAST | Cavite, then tunnels on Corregidor | Philippines analytic center; evacuated as the islands fell; remnant fed FRUMEL (Melbourne) |
| NEGAT | Washington | OP-20-G / Navy Department center; load grew as CAST fell and Midway approached |
Intercept and DF stations (Guam before the war, Bainbridge Island on Puget Sound, and others across the Pacific line) collected raw Morse. HYPO did not own every antenna. It owned the basement where traffic was booked, additives stripped, Japanese read, and the Midway call made. [DOCUMENTED]
Parent organization: OP-20-G. Prewar builder of the intercept service: Laurance F. Safford. Collectors trained in the “On the Roof Gang” (intercept operators), distinct from the Pearl Harbor cryptanalytic cell. [DOCUMENTED]
END-TO-END PROCESS — OVERVIEW
A message’s path from Japanese key to Nimitz’s chart table:
1. IJN operator encrypts plaintext → codebook groups → additive superencipherment → Morse on HF. 2. U.S. intercept operator copies the Morse (often on a typewriter / kana typewriter) at Wahiawa/Heʻeia or another site; DF nets take bearings. 3. Traffic is logged, circuit-identified, and forwarded to the Combat Intelligence Unit (teletype / courier / local link as available). 4. Traffic analysts update the picture of call signs, volumes and ship associations. 5. Cryptanalysts and IBM machine rooms attack JN-25: recover indicators, strip additives, recover code groups, assign meanings, translate. 6. Rochefort’s unit forms an intelligence estimate. 7. Layton briefs Nimitz; tasking and fleet movements follow.
When additives are dark, steps 5–6 shrink and step 4 carries the load. When JN-25 is partially readable, steps 5–7 can move carriers. Midway was the second case. [DOCUMENTED]
RADIO INTERCEPT — PHYSICS IN PLAIN LANGUAGE
Japanese fleet broadcasts and ship-shore nets used high frequency (HF), roughly 3–30 MHz. HF can travel beyond the horizon by reflecting from the ionosphere (skywave). That is why a receiver on Oahu can hear a transmitter thousands of kilometres away at night — and why the same path fades, skips and changes with solar conditions. [DOCUMENTED as radio physics]
What the intercept operator needs:
Navy shore intercept in this era used families such as RAS (National HRO-derived), RBG (Hammarlund-related shore set), RBA/RBB/RBC, RCA RAB, and commercial National / Hammarlund Super-Pro types. A 1944 Wahiawa interior photo description lists a mix of RAS, RBA/RBB/RBC/RBG, RCA RAB, National NC-200 and Hammarlund SP-series. Exact desk assignments in the 1941–42 dungeon period vary by source; treat model lists as [DOCUMENTED] for the Wahiawa intercept plant and [SINGLE-SOURCE] when pinned to a specific Midway-week desk. [DOCUMENTED / SINGLE-SOURCE]
Local oscillator (LO) leakage: a superheterodyne receiver’s oscillator can radiate a tiny signal out the antenna. Shipboard sets were engineered to keep LO radiation very low so the enemy could not DF the listener. Shore sets such as RBG accepted more LO leakage because they were not on a combatant’s topside. [DOCUMENTED as Navy receiver practice]
DIRECTION FINDING
HF direction finding estimates the azimuth of an arriving skywave. Shore DF stations take bearings; several bearings intersect (with error ellipses) near the transmitter. Japanese ships changed call signs to confuse listeners; operators still learned “fists” — the personal rhythm of a Morse hand — and associated circuits by habit. DF plus traffic analysis could track task forces even when JN-25 was only partly readable. [DOCUMENTED]
Physics limit: HF skywave bearings are corrupted by ionospheric tilt and multipath. DF is a statistical weapon, not a GPS fix. Good nets and many cuts beat any single needle. [ESTIMATE / DOCUMENTED as HF DF practice]
Navy truck-mounted DF gear of the period included sets pairing receivers such as RBG with DF goniometers (e.g. DAW-class descriptions in secondary radio literature). Fixed island DF sites formed the Pacific line. [SINGLE-SOURCE for specific type designations]
TRAFFIC ANALYSIS AND CALL SIGNS
When the additive book is unread, the enemy’s externals still speak:
HYPO lived on this diet before Pearl Harbor, when JN-25 was largely dark. Diplomatic Magic traffic (Purple) and ship-sighting reports helped Washington’s strategic picture; they did not give Rochefort a readable IJN operations code in time to warn of 7 December. That failure is [DOCUMENTED] as fact; the assignment of blame among Washington, Hawaii and the split of effort between diplomatic and naval systems remains historically argued. [DOCUMENTED / conflict labelled]
JN-25 — HOW THE CODE PLUS ADDITIVE WORKS
Americans named the IJN general operational system JN-25 (Japanese Navy code #25). Variants (JN-25a, JN-25b, …) changed books over time. Structure in plain language: [DOCUMENTED]
1. Codebook — tens of thousands of five-digit groups, each standing for a word, phrase, or element. Published figures conflict: on the order of 27,500 entries in one NSA summary, about 45,000 five-digit numbers in another NSA Midway essay. Both describe the same kind of book; keep the range. Groups were chosen so the digits summed to a multiple of three (“scanned”) as a garble check. [DOCUMENTED; size conflict labelled]
2. Additive book — pages of random five-digit groups. Example scale from NSA: on the order of 300 pages × 100 groups. Not a one-time pad: additives were reused. Reuse is what made recovery possible. [DOCUMENTED]
3. Superencipherment — the clerk looks up plaintext → code groups, then adds (digit by digit, without carry) a stretch of additives starting at a indicated place in the additive book. Example of non-carrying addition: 8+7 = 5, not 15. The groups as transmitted (GATs) go out in Morse. [DOCUMENTED]
4. Indicators — message elements that tell the legitimate recipient where in the additive book to start. Recovering indicators was a first cryptanalytic goal. [DOCUMENTED]
### Worked toy example [ESTIMATE for numbers; DOCUMENTED for method]
Suppose codebook has SHIP = 70863 (digits 7+0+8+6+3 = 24, divisible by 3 — scans). Additive from the book = 39152. Non-carrying sum:
7+3=0, 0+9=9, 8+1=9, 6+5=1, 3+2=5 → transmitted group 09915.
Receiver subtracts the same additive (non-carrying) and looks up 70863 → SHIP. Cryptanalysts who do not have the additive must invent ways to recover it from depths of traffic.
STRIPPING ADDITIVES — THE ATTACK
When several messages are encrypted with the same additive stretch (a “depth”), differences between ciphertext groups equal differences between underlying code groups — the additive cancels. Cryptanalysts:
1. Align messages known (or believed) to share additives. 2. Difference columns of GATs. 3. Look up differences in an index of differences built from already-recovered “good” code groups (IBM tabulators made this index practical). 4. A “click” suggests candidate additives; strip them and test whether resulting groups all scan (digit sum ÷ 3) and whether known good groups appear with plausible frequency. 5. Grow the recovered additive table and the codebook meanings together.
Stereotyped traffic — port entry requests, position reports, medical boilerplate — supplied cribs once some groups were known. [DOCUMENTED]
Before Pearl Harbor, open histories put readable JN-25 on the order of 10–15% of groups — not enough for operational warning of the raid. By spring 1942, HYPO was reading useful portions of JN-25b within hours on some messages. Exact percentages vary by week and by who is scoring the credit. [DOCUMENTED / SINGLE-SOURCE for timing details]
IBM PUNCH-CARD HARDWARE — HOW IT WORKED
HYPO’s edge was not a single “bomb” machine like the British Enigma effort. It was language skill, continuity, and IBM punched-card tabulating equipment used at industrial scale. Rochefort later credited Dyer’s machine program; accounts speak of millions of cards per month at peak Pacific COMINT tempo. [DOCUMENTED / SINGLE-SOURCE for the “two or three million a month” recollection]
### Hollerith card mechanics in plain language
A standard IBM card of the era is a rectangle of stiff paper with 80 columns. Each column can hold one character by punched holes in rows (digits 0–9 plus zone punches for letters). A keypunch operator encodes data; a verifier catches errors. Cards become a physical database you can sort at machine speed. [DOCUMENTED as IBM practice]
Machines HYPO/FRUPAC-class shops used (Navy machine-room language includes reproducers and tabulators; exact 1942 basement inventory is not a single published packing list):
What this buys cryptanalysis: the index of differences for thousands of good groups is not a hand card file you flip once — it is a deck you can sort and match mechanically against every new depth. Humans still judged clicks and translations; machines made the search space tractable. [DOCUMENTED / ESTIMATE]
Nimitz sent men off the damaged battleship California to be trained as machine operators — a documented improvisation when the basement’s workload exploded after December 1941. [DOCUMENTED]
TYPEWRITERS AND KANA
Intercept copy and translation needed typing. Standard English office typewriters handled five-digit groups and English translations. Kana typewriters (Japanese syllabic) appear in Navy COMINT practice for handling Japanese text when linguists worked plaintext recoveries. Exact basement inventory ratios are [UNVERIFIED] in popular retellings; the functional need is [DOCUMENTED].
THE DUNGEON
Building 1 basement: windowless, crowded with tables, card boxes, printouts, and tabulating machines. One way in, Marine guard. Rochefort ran long watches with little uniform ceremony — bathrobe anecdotes are [SINGLE-SOURCE] memoir colour but consistent with the unit’s results-over-rank reputation. In April 1943 the Combat Intelligence Unit moved to a purpose-built facility at Makalapa near Pacific Fleet headquarters; the dungeon later served storage/admin until commemorated. [DOCUMENTED]
Photo note: some widely circulated “HYPO ops room” photographs show windowed spaces and late-war equipment mixes — more likely Wahiawa intercept or post-1943 FRUPAC floors than the 1941–42 windowless basement. Alt text on this page stays descriptive; do not over-identify every print as Midway week in the dungeon. [ESTIMATE / DOCUMENTED for dungeon description]
MIDWAY — AF IS SHORT OF WATER
In May 1942 readable JN-25 traffic showed a major Japanese operation aimed at a place designator “AF,” with a date window in early June and a carrier striking force. HYPO assessed AF as Midway Atoll. A Japanese weather-related association of AF with Midway had already appeared in March traffic in some accounts. OP-20-G in Washington, with input from CAST remnant channels, leaned toward the Aleutians or another less central target and argued about timing. [DOCUMENTED; Washington vs Hawaii dispute labelled]
Rochefort’s unit proposed a confirmation. Midway was instructed by underwater cable (so the Japanese would not hear the order) to transmit a radio message in the clear reporting that the island’s salt-water evaporators / fresh-water plant had failed. Within about two days, Japanese traffic reported that AF was short of fresh water. That fixed the designator. [DOCUMENTED]
Nimitz accepted HYPO’s reading over the Washington alternative and placed Enterprise, Hornet, and Yorktown northeast of Midway. The battle of 4–7 June 1942 cost Japan four fleet carriers; Yorktown was lost. The carrier balance in the Pacific turned. [DOCUMENTED]
What the decrypts did not give: every detail of Yamamoto’s plan. What they gave: target, rough timing, and weight of force — enough for an ambush. The water-plant message was a controlled plaintext injection — a classic confirm — not a lucky guess. [DOCUMENTED]
WASHINGTON VERSUS HAWAII
The Midway fight was also a bureaucratic fight:
Both sides worked the same JN-25 material with incomplete breaks. The argument was whose reading Nimitz should trust under uncertainty. He trusted the basement. After victory, Navy Department politics still punished the messengers: Nimitz recommended Rochefort for the Distinguished Service Medal; the Department refused, citing shared credit with Washington and the Philippines. Rochefort was transferred out of cryptanalysis. Safford was sidelined in the same season. OP-20-G centralized Pacific naval decrypt under Washington sections associated with John Redman and Joseph Wenger. HYPO was reorganized, renamed FRUPAC, and eventually moved to purpose-built spaces. [DOCUMENTED]
Rochefort retired as a captain, was denied the medal twice in life, and died in 1976. The Distinguished Service Medal was awarded posthumously in 1985–86. The post-Midway treatment remains a lasting argument inside the cryptologic community. [DOCUMENTED]
PURPLE, RED AND MAGIC — BRIEF CONTEXT
Popular histories sometimes blur Magic and JN-25. Diplomatic Purple could reveal Foreign Ministry thinking; it was not a substitute for reading Combined Fleet operations. HYPO’s Midway success was a JN-25 / traffic-analysis success, not a Purple decrypt of Yamamoto’s order of battle. [DOCUMENTED]
FRUPAC AND AFTER
Fleet Radio Unit Pacific inherited HYPO’s mission at larger scale: more machines, more circuits, more collaboration with FRUMEL (Melbourne) and Washington. Wahiawa remained a major intercept plant; analytic work left the dungeon for Makalapa (1943) and later facilities. The wartime unit in its 1941–42 form was broken up; the name HYPO survives in Navy cryptologic memory and associations. [DOCUMENTED]
HARDWARE DEEP DIVE — RECEIVERS
### National HRO / Navy RAS family
The National HRO was a high-quality HF communications receiver famous for its micrometer dial and plug-in coil sets. Navy RAS sets were HRO-derived intercept receivers covering roughly 190 kHz–30 MHz class bands suitable for copying fleet Morse. Operators changed coil sets for band segments and used crystal filters / sharp IF selectivity where fitted to pull signals out of interference. [DOCUMENTED as equipment class]
### Hammarlund Super-Pro / RBG
Hammarlund Super-Pro commercial receivers and the Navy RBG (HQ-120X-derived shore set, contracts from 1941) appear in shore intercept and DF truck stories. RBG was generally a shore/DF receiver, not a low-LO-leakage shipboard combatant set. [DOCUMENTED]
### Why selectivity matters
Morse CW occupies a narrow bandwidth. A receiver with a sharp crystal filter can reject an adjacent kilohertz of racket. Intercept is an ears-and-filter problem as much as an antenna problem. [ESTIMATE]
HARDWARE DEEP DIVE — RECORDING
Later intercept practice used paper/ink or magnetic recording to preserve signals for re-copy. Photographs of operators with reel decks and “week” tape boxes document that workflow in matured Pacific COMINT plants. Midway-week dungeon work was still dominated by live copy, typed groups, cards and paper — recording gear expanded as the war went on. [DOCUMENTED for later plant imagery; ESTIMATE for 1942 mix]
HARDWARE DEEP DIVE — IBM ROOM AS A FACTORY
Think of cryptanalysis as a factory:
Dyer’s insight was to industrialize the matching step. Without that, Midway-scale traffic volumes drown hand methods. [DOCUMENTED / ESTIMATE]
DISSEMINATION PATH — LAYTON AND NIMITZ
Edwin Layton was the hinge between the dungeon and the fleet commander. He understood both the language of cryptanalysts and the language of admirals. Product that cannot be briefed in time is scrap paper. The Midway success was as much a dissemination and trust success as a mathematical one: Nimitz had to bet carriers on an incomplete break. [DOCUMENTED]
PEARL HARBOR WARNING FAILURE — SCOPE
HYPO did not deliver a specific, timely warning of the 7 December 1941 raid. Causes, none of them a single cartoon failure:
This page does not retry the entire Pearl Harbor controversy. It records that traffic analysis and partial cryptanalysis were not enough, and that Midway was possible only after traffic volume and recoveries rose in 1942. [DOCUMENTED]
CAST, CORREGIDOR AND FRUMEL
Station CAST in the Philippines was HYPO’s sister. When Cavite became untenable, CAST moved into Corregidor’s tunnels and kept working under siege. Evacuation of people and material to Australia produced FRUMEL (Fleet Radio Unit Melbourne), which continued Pacific naval COMINT in cooperation with the U.S. and British/Australian partners. CAST’s collapse increased Washington’s urge to centralize and increased HYPO’s relative importance as the remaining forward U.S. Navy analytic node in early 1942. [DOCUMENTED]
Credit disputes after Midway explicitly invoked “shared credit with Washington and the Philippines.” That sentence only makes sense if CAST’s recoveries are remembered as real. HYPO did not invent JN-25 cryptanalysis alone. [DOCUMENTED]
CRYPTOLOGIC CULTURE IN THE BASEMENT
Rochefort’s shop mixed:
Uniform discipline was low relative to a battleship; security discipline was absolute (Marine on the door, need-to-know). The culture prized being right over being tidy. That culture produced Midway — and made enemies in Washington. [DOCUMENTED / SINGLE-SOURCE for colour]
MATHEMATICS APPENDIX — NON-CARRYING ARITHMETIC
Non-carrying addition (mod 10 per digit) is not ordinary school addition. Properties used by cryptanalysts:
One-time pads with true random never-reused additives defeat this program. JN-25’s reused additive book did not. That is the central engineering moral. [DOCUMENTED / ESTIMATE]
MATHEMATICS APPENDIX — DEPTH
A depth of d messages on the same additive stretch yields d(d−1)/2 pairwise differences per column. Even modest depths explode the chance of hitting a known good-group difference in the index. IBM sorting made looking up those differences routine. [ESTIMATE]
HF PROPAGATION FOR INTERCEPTORS
Skywave: the ionosphere’s F-layer can refract HF back to Earth. Night often favors longer skip on lower HF; day shifts usable bands upward. Intercept sites choose frequencies and antennas for the Japanese circuits they are tasked against. Multipath causes fading — copy quality rises and falls over seconds. Skilled operators ride the fades; machine recorders (later) allow second passes. [DOCUMENTED as HF practice]
Groundwave: short-range HF/MF follows the surface; less relevant for Tokyo-to-Oahu paths.
Why Oahu hears the Combined Fleet: geometry plus ionosphere, not magic. Why DF is fuzzy: the ray path is not a straight line on the map. [ESTIMATE]
SECURITY AND COMPARTMENTATION
COMINT success dies if the enemy learns he is read. Midway planning used cover stories and limited distribution. The water ruse itself was a risk: a suspicious Japanese analyst might have smelled a plant. They did not, in time. Postwar secrecy kept HYPO’s role muted for decades; popular Midway films often underplay or fictionalize the cryptologic half. [DOCUMENTED]
WHAT HYPO WAS NOT
| Not this | What it actually was |
|---|---|
| Station CAST | Philippines / Corregidor analytic center |
| NEGAT | Washington OP-20-G center |
| FRUMEL | Melbourne post-evacuation unit |
| OP-20-G itself | Washington parent; HYPO was a field station under it |
| On the Roof Gang | Prewar intercept-operator school (collectors) |
| Army SIS Purple shop | Diplomatic Magic, different system |
| A radar station | Radio intelligence / cryptanalysis, not air-search radar |
| A public broadcast station | Receive and analyze; not AFN Honolulu |
SOURCE CONFLICTS
1. Codebook size — ~27,500 vs ~45,000 five-digit entries in NSA-family writeups. 2. When JN-25 became “broken” — popular “by March 1942” vs gradual percentage readability; HYPO read useful portions by Midway, not every group. 3. Who proposed the water ruse — Rochefort’s unit in most Navy histories; individual name credit varies in memoirs. 4. Washington’s Midway position — Aleutians emphasis vs timing disputes; secondary histories differ in emphasis. 5. Photo identification — dungeon vs Wahiawa vs late FRUPAC floors often confused online. 6. Combat Intelligence vs Information — correct wartime name is Intelligence; some graphics say Information.
MYTHS
1. “HYPO alone broke JN-25.” False. CAST, Washington, and later FRUMEL shared the work. [DOCUMENTED] 2. “Magic decrypts won Midway.” Misleading. Magic is Purple/diplomatic; Midway was JN-25 + TA + the water confirm. [DOCUMENTED] 3. “They read every Japanese message in real time.” False. Partial breaks, hours to days, many dark messages. [DOCUMENTED] 4. “Rochefort was rewarded immediately.” False. Medal denied for decades. [DOCUMENTED] 5. “IBM machines decrypted automatically.” False. They sorted and matched; humans cryptanalyzed and translated. [DOCUMENTED]
TIMELINE (OPEN)
PEOPLE
| Name | Role |
|---|---|
| Joseph J. Rochefort | OIC, Combat Intelligence Unit / HYPO analytic chief |
| Thomas Dyer | Chief cryptanalyst; IBM methods advocate |
| Edwin T. Layton | Pacific Fleet intelligence officer; briefed Nimitz |
| Chester W. Nimitz | CINCPAC; accepted Midway reading |
| Laurance F. Safford | Prewar OP-20-G builder; later sidelined |
| John Redman / Joseph Wenger | Washington OP-20-G figures in 1942 centralization |
| On the Roof Gang alumni | Intercept operators feeding the system |
GLOSSARY
OPERATING A WATCH — INTERCEPT SIDE
A notional intercept watch at Wahiawa / Heʻeia-class plants:
1. Supervisor assigns frequencies and Japanese nets from the tasking list. 2. Operator tunes the RAS/RBG/HRO-class receiver, peaks the antenna trim, sets filter. 3. When Morse starts, operator copies five-digit groups (and prosigns) to typewriter or pen. 4. Logs time, frequency, call signs, signal strength, and unusual notes (new fist, interference). 5. Traffic goes to checkers, then to communications for the analytic center. 6. DF stations, tasked in parallel, report bearings keyed to the same circuit when possible.
Error rates matter: a single wrong digit destroys a group and can poison additive recovery. Training emphasized accuracy over speed, then speed with accuracy. [ESTIMATE / DOCUMENTED as COMINT practice]
OPERATING A WATCH — DUNGEON SIDE
1. Incoming traffic booked and sorted by system (JN-25 vs others) and circuit. 2. Traffic analysis boards updated. 3. Messages selected for cryptanalytic attack (depths, stereotyped forms, high-value call signs). 4. Keypunch decks cut; sorters/collators run against difference indexes and additive recoveries. 5. Cryptanalysts evaluate clicks; linguists translate recovered plaintext fragments. 6. Rochefort merges crypto and TA into an estimate. 7. Layton receives sanitized product for Nimitz.
The bottleneck shifts: some nights TA is the only light; some nights the machines chatter and translations stack up. Midway week was a translations-and-judgment bottleneck as much as a machine bottleneck. [ESTIMATE]
INDICATORS AND BOOK ALIGNMENT
Japanese clerks had to tell recipients where to start in the additive book. Those indicators were themselves disguised. Recovering the indicator system was as important as recovering any single additive page. Once indicators were understood, cryptanalysts could place messages onto the correct additive stretches and build depths deliberately rather than by luck. NSA historical summaries list indicator recovery as the first goal in the JN-25 attack sequence. [DOCUMENTED]
STEREOTYPES AND CRIBS
Navies repeat themselves. Convoy reports, weather, routine logistics, and staff phrase-books create cribs — guessed plaintext — that lock recoveries. After some codebook meanings are known, every new stereotype accelerates the next. This is why readability snowballs once a foothold exists, and why early 1942 looked different from late 1941 even before any single dramatic “break” headline. [DOCUMENTED / ESTIMATE]
LIMITS OF COMINT
Midway worked because Japan radiated enough, the book was sufficiently recovered, and Nimitz acted without waiting for perfection. [DOCUMENTED]
COMPARISON — BLETCHLEY AND HYPO
Popular memory pairs “codebreaking” with Bletchley Park’s Enigma bombes. Differences that matter:
| Topic | Bletchley (Enigma focus) | HYPO (JN-25 focus) |
|---|---|---|
| Machine cipher vs codebook | Rotor machine daily keys | Codebook + additive book |
| Star hardware | Bombes, later Colossus (for Tunny) | IBM tabulators / sorters / collators |
| Language | German | Japanese (linguist-scarce) |
| Main consumer example | Atlantic / European commands | CINCPAC |
| U.S. partner link | Growing BRUSA cooperation | CAST / FRUMEL / NEGAT net |
Both industrialized cryptanalysis. Neither was “a genius in a room with a pencil” alone. HYPO’s story is the Pacific twin of that industrialization, with different math. [DOCUMENTED / ESTIMATE]
NAVY RADIO INTELLIGENCE BEFORE HYPO’S FAME
OP-20-G’s prewar investment — intercept sites, the On the Roof Gang, language officers in Japan, early IBM experiments — is why a basement in 1942 could scale. Rochefort did not invent Navy COMINT in June 1941; he inherited a thin, skilled pipeline and then rode the traffic surge. Safford’s prewar building effort is part of the Midway causal chain even though Safford was sidelined after the victory. [DOCUMENTED]
LEGAL AND CLASSIFICATION NOTE
Wartime COMINT was among the most highly classified U.S. activities. Decades of secrecy shaped the public Midway myth (pilots and admirals first, cryptanalysts later). This encyclopedia page uses only declassified and open historical material. It does not provide guidance for attacking modern cryptosystems. JN-25 is a historical additive code; modern military crypto is a different world. [DOCUMENTED]
SITE CONTEXT AND INTERNAL LINKS
This wiki’s radio and strategic pages are sparse on WWII COMINT peers. Related reading on the same site for radio engineering context:
Do not confuse HYPO’s receive-and-analyze mission with strategic VLF broadcast or modern EAM nets. [DOCUMENTED]
MORE ON THE WATER RUSE — PROCEDURE
Step-by-step as open histories agree:
1. HYPO concludes AF is Midway from traffic and prior associations; Washington remains unconvinced enough to matter. 2. A request goes to Midway via secure cable: broadcast in plain language that the desalination / evaporator plant has broken and fresh water is critical. 3. Midway transmits the bait on radio the Japanese will hear. 4. Japanese intelligence / operations traffic references AF’s water trouble. 5. HYPO copies and reads enough to close the loop. 6. Layton and Nimitz treat AF as Midway for force disposition.
Timing details (“about two days”) vary slightly by account; the structure does not. [DOCUMENTED; minor timing [SINGLE-SOURCE]]
Why cable for the order? If Midway had been told by radio that Washington could hear, Japanese DF/COMINT might have heard the setup. The underwater cable kept the instruction off the air. [DOCUMENTED]
FORCE DISPOSITION — WHAT NIMITZ DID WITH THE PRODUCT
Nimitz concentrated the available U.S. carriers northeast of Midway — on the flank of the expected Japanese approach — rather than guarding Hawaii or chasing Aleutian feints with the main bet. That operational choice is the payoff function of HYPO’s estimate. Cryptanalysis that does not change dispositions is academic. [DOCUMENTED]
Yorktown, hastily repaired after Coral Sea, was part of that bet. COMINT did not promise she would survive; it promised the ambush geometry. [DOCUMENTED]
AFTER-ACTION CREDIT POLITICS
Nimitz’s DSM recommendation for Rochefort failed in Washington. Explanations offered then and later:
Historians still disagree on the mix. The factual skeleton is not in dispute: medal denied in wartime, Rochefort removed from cryptanalysis, medal awarded decades after his death. [DOCUMENTED / conflict on motives]
ENGINEERING THE DIFFERENCE INDEX
Building the index of differences:
1. Assemble a list of recovered “good” code groups with frequency scores. 2. Compute pairwise non-carrying differences (choosing a canonical orientation, e.g. difference below a midpoint threshold — Bletchley/JN-25 practice describes choosing alternatives below 55556 in related writeups). 3. Punch each difference with pointers back to the parent good groups. 4. Sort the deck by difference value. 5. When a live depth produces a difference, sort/collate against the index; a match is a click.
Updating the index as new good groups appear is continuous work. Stale indexes miss clicks; bloated indexes with bad groups create false clicks. Human judgment remains the filter. [DOCUMENTED method family; ESTIMATE for shop workflow]
FALSE CLICKS AND CONFIRMATION
A click is a hypothesis, not a decrypt. Confirmation chain:
Only then does a page of the additive book get inked as trusted. [DOCUMENTED / ESTIMATE]
CALL-SIGN WARFARE
IJN changed call signs to break TA continuity. Countermeasures:
The fight between call-sign changers and traffic analysts is older than computers. HYPO’s Midway picture combined both channels. [DOCUMENTED]
WOMEN AND ENLISTED LABOR [labelled]
Pacific COMINT, like Washington and Dayton code work, drew on enlisted operators and, as the war grew, wider labor pools including women in machine and clerical roles in various Navy COMINT activities. Specific 1941–42 dungeon headcounts by gender are unevenly documented in open popular sources. Do not erase enlisted machine operators from a Rochefort-centric story; Nimitz’s California sailors are a reminder that Midway’s math ran on human hands at keypunches. [DOCUMENTED for enlisted role; UNVERIFIED for detailed dungeon demographics]
PHYSICAL SECURITY OF BUILDING 1
One entrance, armed Marine, windowless basement — classic COMINT physical security for the era. Need-to-know meant many Pearl Harbor officers had no idea what the basement did. That secrecy protected the source and later complicated public credit. [DOCUMENTED]
READER’S WALKTHROUGH — FROM MORSE TO BRIEFING
Imagine a single Japanese logistics message in May 1942:
1. Encrypt: Clerk converts Japanese text to five-digit code groups from the current JN-25 book; adds additives from the current page; taps Morse. 2. Propagate: HF skywave reaches Oahu. 3. Copy: Wahiawa operator on an RAS/HRO-class receiver types the GATs. 4. Forward: Traffic hits the dungeon booking desk. 5. TA note: Circuit and call signs update the plot — maybe already suggesting a Midway-related net. 6. Crypto: Indicators place the message on a partly recovered additive stretch; IBM run strips candidates; linguist reads a fragment about water or shipping. 7. Merge: Rochefort’s watch puts the fragment next to other AF messages. 8. Brief: Layton tells Nimitz the picture is firming. 9. Act: Carriers receive movement orders that look like operational genius and are also applied COMINT.
Remove any step and the chain weakens. Midway was a system success. [ESTIMATE narrative; DOCUMENTED components]
TRAINING PIPELINE
No single school produced a complete HYPO officer. The unit was a mash of pipelines. [DOCUMENTED]
EQUIPMENT LOGISTICS
Cards by the million imply paper supply, punch repair, sorter maintenance, and spare tube stocks for receivers. Tropical humidity is hostile to paper cards and electronics. Machine-room air and discipline were operational concerns, not niceties. Open histories mention the volume; they rarely publish the full logistics annex. [ESTIMATE]
RELATIONSHIP TO FLEET COMMAND
HYPO was not CINCPAC’s only intelligence source (air recon, coastwatchers, combat reports, Magic summaries from Washington). It was the source that could see Combined Fleet intentions in the encrypted command nets. Layton’s job included weighing COMINT against other evidence without revealing sources in every staff meeting. [DOCUMENTED]
EXTENDED PHYSICS — WHY MORSE ON HF
Morse continuous-wave (CW) is power-efficient and narrowband. For a given transmitter power, CW often outranges voice. Japanese fleet communications relied on trained operators and CW nets; U.S. interceptors were likewise CW-native. Frequency-shift and later teleprinter systems grew during the war, but Midway-era JN-25 travel was still dominated by Morse five-digit groups. [DOCUMENTED / ESTIMATE]
Antenna gain and receiver noise figure set the floor. A rhombic aimed at the right azimuth beats a random wire. Shore stations invested in acres of antenna farm for that reason — the dungeon’s brilliance still required RF watts arriving from the ionosphere. [ESTIMATE]
EXTENDED PHYSICS — SUPERHETERODYNE RECEIVERS
Most HYPO-era intercept receivers were superhets: RF amplifier → mixer with local oscillator → intermediate frequency (IF) amplifiers → detector → audio. The IF crystal filter provides selectivity. Images and LO radiation are the engineering tax. Operators who understood image frequencies avoided copying the wrong station. [DOCUMENTED as radio engineering]
EXTENDED CRYPTO — WHY REUSE KILLS
Shannon’s maxim, in plain language: a truly random key the length of the message, used once, is unbreakable. JN-25’s additive book was long but reused. Every reuse is a mathematical gift to the attacker who can stack depths. Japanese improvements (book changes, better indicators, operational discipline) raised cost; they did not buy one-time-pad security. [ESTIMATE / DOCUMENTED for reuse fact]
STATUS
Wartime unit. The Combat Intelligence Unit left Building 1 in 1943; FRUPAC carried the mission forward at larger scale. The original dungeon form and Rochefort command did not survive the 1942 reorganization intact. The name Station HYPO is kept by Navy cryptologic associations and official Hawaii Navy history pages. Building 1 stairway plaques now mark the basement. [DOCUMENTED]
IN ONE SENTENCE
Station HYPO was Joseph Rochefort’s Combat Intelligence Unit in the Pearl Harbor basement — the OP-20-G field center that combined HF intercept, traffic analysis, IBM-assisted JN-25 additive stripping, and Japanese language skill to prove “AF” was Midway, put Nimitz’s carriers on the flank, and was then broken up and recentralized in Washington.
SOURCES
Official / institutional:
Navy / veteran technical history:
Radio equipment secondary:
Method detail:
Site pre-rewrite article preserved the factual core used here: dungeon, Rochefort/Dyer/Layton/Nimitz, CAST/NEGAT/FRUMEL, JN-25, water ruse, post-Midway politics.
RELATED ARTICLES
APPENDIX — QUICK FACT CARD
| Item | Value | Label |
|---|---|---|
| Unit | Combat Intelligence Unit, 14th Naval District | [DOCUMENTED] |
| Call name | Station HYPO (phonetic H / Hawaii) | [DOCUMENTED] |
| Later name | FRUPAC (Fleet Radio Unit Pacific) | [DOCUMENTED] |
| Location (analytic) | Building 1 basement, Pearl Harbor (“Dungeon”) | [DOCUMENTED] |
| Intercept lineage | Wailupe → Heʻeia → Wahiawa | [DOCUMENTED] |
| OIC (1941–42) | Joseph J. Rochefort | [DOCUMENTED] |
| Chief cryptanalyst | Thomas Dyer | [DOCUMENTED] |
| Fleet intel hinge | Edwin T. Layton → Admiral Nimitz | [DOCUMENTED] |
| Parent | OP-20-G (Washington) | [DOCUMENTED] |
| Sisters | CAST (Corregidor), NEGAT (Washington); FRUMEL (Melbourne) | [DOCUMENTED] |
| Main target | JN-25 (codebook + additive) | [DOCUMENTED] |
| Key tools | HF intercept, DF, traffic analysis, IBM tabulators, linguists | [DOCUMENTED] |
| Signature success | AF = Midway via water-plant ruse (May 1942) | [DOCUMENTED] |
| Not Purple/Magic | Diplomatic machine ciphers were a different channel | [DOCUMENTED] |
APPENDIX — SAMPLE NON-CARRYING WORKSHEET [ESTIMATE]
Code group P: 4 5 6 7 8 (sum 30, scans) Additive A: 9 2 8 1 4 Cipher C: 3 7 4 8 2 (digitwise mod 10 sums)
Recover: C − A → 4 5 6 7 8. If an analyst guesses A wrong by one digit, the result usually fails to scan or fails to match known groups — the filter that made industrial search possible.
APPENDIX — WATCH BILL CONCEPT [ESTIMATE]
A 24-hour COMINT center runs overlapping watches so depths that form at 0300 local are not ignored until morning. Midway-week pressure meant sleeping on cots and bathrobe legend. Modern readers should translate that as: the factory did not stop for the convenience of officers.
CLOSING
Station HYPO’s purpose was Pacific Fleet COMINT. Its process was intercept → DF/TA → JN-25 additive cryptanalysis → translation → Layton → Nimitz. Its hardware was HF receivers (HRO/RAS/Hammarlund-class), DF nets, kana and English typewriters, and IBM punched-card sorters and tabulators in a guarded basement. The Midway water ruse was applied scientific method under fire. The Washington quarrel afterward was politics under victory. Keep those layers separate, label the conflicts, and the dungeon still teaches how radio, math, language and trust move fleets.
FURTHER ON OP-20-G ORGANIZATION
OP-20-G was the Navy Department’s code and signal intelligence organization under the Chief of Naval Operations. It set policy, allocated machine resources, coordinated with Army SIS on some diplomatic matters, and ran or tasked the phonetic field stations. HYPO’s independence in practice — Rochefort answering operationally to Fleet needs through Layton — collided with Washington’s desire for centralized control of decrypt priorities. That structural tension predates Midway and exploded after it. [DOCUMENTED]
FURTHER ON FRUPAC GROWTH
As the Pacific war expanded, FRUPAC became a larger industrial COMINT organization: more intercept positions at Wahiawa, more teletype circuits, more IBM floor space, specialized sections (JN series work, machine room, translation). Eight teletype circuits between FRUPAC elements and Wahiawa appear in late-war notes. The 1942 dungeon was the seed, not the final tree. [SINGLE-SOURCE for circuit count; DOCUMENTED for growth]
FURTHER ON LANGUAGE
JN-25 plaintext was Japanese. Recovered groups were useless without linguists who could read naval staff style, abbreviations and place-name conventions. Prewar language billets in Japan were a strategic investment; wartime crash courses tried to grow the pool. Rochefort’s own language skill is part of why he could smell a wrong recovery. [DOCUMENTED]
FURTHER ON THE ALEUTIAN FEINT
Yamamoto’s Midway plan included an Aleutian operation. Washington analysts who emphasized the north were not inventing threats from nothing — they were weighting a real secondary axis more heavily than HYPO did. Nimitz’s decision was about where to place the decisive carrier bet, not about whether Attu/Kiska existed. COMINT settled the weight of effort. [DOCUMENTED]
ACKNOWLEDGED UNCERTAINTIES (SUMMARY)
1. Exact JN-25 codebook size (tens of thousands; 27.5k vs 45k figures). 2. Exact week-by-week readability percentages in spring 1942. 3. Individual authorship of the water-ruse idea inside HYPO. 4. Full 1941–42 IBM machine inventory in Building 1. 5. Which surviving photos show the dungeon versus Wahiawa/FRUPAC. 6. Motives behind Rochefort’s DSM denial (facts of denial are solid; motives argued).
Everything else on this page is built from corroborated process, open institutional history, and labelled engineering explanation.
MORE RECEIVER OPERATING PRACTICE
Intercept operators learned to:
A clean five-digit copy is an engineering product. Garbage in means wasted IBM cycles and false depths. [ESTIMATE]
MORE ON TABULATOR LIMITS
Punched-card machines are not computers. They do not run arbitrary programs. Each run is a fixed path: sort on column 12, collate deck A with deck B, print. Cryptanalytic cleverness lived in how Dyer’s people designed the card layouts and run sequences. Changing the attack meant redesigning decks, not editing source code. That constraint shaped which JN-25 problems were feasible in 1942. [ESTIMATE / DOCUMENTED as machine capability class]
MORE ON DISSEMINATION SECURITY
Layton and Nimitz used paraphrases and operational cover so that actions taken on COMINT did not telegraph “we read your mail” back to Tokyo. The water ruse itself was a controlled leak the other direction — injecting plaintext to watch the enemy’s encrypted reaction. Both techniques are classic COMINT tradecraft. [DOCUMENTED]
UNIT SCALE
Open popular sources describe HYPO as a “small” staff in mid-1941 that grew rapidly after the war began. Exact headcounts vary. What matters: Midway was produced by dozens of skilled people and a machine room, not by a lone genius — even if Rochefort is the rightful commanding face of the story. [SINGLE-SOURCE for specific counts when quoted elsewhere; DOCUMENTED for small-then-grew pattern]
METHOD, NOT MAGIC (EXPANDED)
HYPO’s edge was continuity and language, not a single wonder device. The same officers had studied the IJN for years. They recognized who was talking, how a staff wrote, and which designators repeated. Traffic analysis carried them when additives were dark. The water-plant message was a controlled plaintext injection — a classic confirm — not a lucky guess. CAST and Washington recovered code groups too. The argument in 1942 was whose reading of the same material Nimitz should trust. He trusted the basement. [DOCUMENTED]
Industrial cryptanalysis always looks like magic from outside and like factory work from inside. Cards, headphones, bad coffee, arguments over a click, a linguist saying “that’s not how they phrase an order,” a commander deciding to bet three carriers — that is the real texture. [ESTIMATE]
CAVEATS
Credit for JN-25 recoveries was disputed at the time and still is. HYPO did not “break the code” alone, and it did not read every message. The water ruse is documented in Navy and historical accounts; small details of timing vary. The post-Midway treatment of Rochefort is a matter of record and of lasting argument inside the cryptologic community. Photo captions online are frequently wrong about dungeon versus Wahiawa. Codebook size figures conflict across official essays. This page keeps those tensions visible. [DOCUMENTED]
FINAL CHECKLIST FOR READERS
Before citing Station HYPO:
WORKED TRAFFIC-ANALYSIS SKETCH [ESTIMATE]
Suppose DF cuts and volume analysis show a carrier-associated net densifying on frequencies historically used before major operations, while JN-25 fragments mention AF and a date cluster in early June. Separately, Aleutian-associated call signs also rise. Two hypotheses compete: (1) main effort Midway, north is feint/secondary; (2) main effort north, Midway secondary. The water ruse collapses (1) vs (2) for the AF designator specifically. Without the ruse, Nimitz still had a judgment call; with it, the designator dispute ended. TA alone rarely ends designator disputes that cleanly. [ESTIMATE]
WORKED ADDITIVE RECOVERY SKETCH [ESTIMATE]
Depth of three GATs in one column: 38152, 70419, 99260. Pairwise non-carrying differences are computed and looked up in the good-group difference index. A click suggests additive 12947. Stripping yields three candidate code groups; all scan; two are already known high-frequency groups; the third is new and tentatively entered for linguist review. That is one column of one message. Midway-week traffic meant thousands of such micro-decisions. [ESTIMATE — illustrative digits only]
WHY THIS STILL MATTERS
Modern readers live in an age of digital COMINT and AES. The HYPO lesson is older and durable: reuse, stereotype, traffic volume, and decision-maker trust decide whether cryptanalysis changes a war. Hardware changes; those four do not. [ESTIMATE]
BUILDING 1 TODAY
Navy Region Hawaii notes that the dungeon lay vacant or in mundane use for decades until its legacy was rediscovered; plaques now mark the stairway entrances, and there are two entrances into the space in the modern commemorative description (wartime accounts emphasize the single guarded entry). Visitors and historians should treat commemorative access arrangements as post-war changes, not as the 1942 security plan. [DOCUMENTED]
ORAL HISTORY COLOUR [SINGLE-SOURCE]
Memoir and Proceedings colour includes Rochefort in a smoking jacket/bathrobe, Dyer living with the machines, and officers sleeping near the work. These details are consistent across retellings but are still memoir-grade. They illustrate tempo; they are not needed to prove Midway. [SINGLE-SOURCE]
CONNECTION TO CORAL SEA
COMINT also shaped the Coral Sea actions in May 1942 — the same spring as the AF fight. HYPO and the wider OP-20-G net were already delivering operationally useful JN-25 and TA product before Midway. Midway is the famous confirm; Coral Sea is proof the basement was already in the war. Exact message lists are a specialist literature. [DOCUMENTED for COMINT role pattern; SINGLE-SOURCE for message-level claims elsewhere]
ENGLISH-ONLY STYLE NOTE
This page uses English throughout per site rules for Western/U.S. topics. Japanese terms appear only as historical designators (AF) or system names as used in U.S. literature (JN-25). Place names use common English forms (Heeia / Heʻeia, Wahiawa, Corregidor, Midway). [DOCUMENTED as site practice]
SYNTHESIS — PURPOSE, PROCESS, HARDWARE
Purpose: Give CINCPAC communications intelligence on the Imperial Japanese Navy — especially operational intentions buried in JN-25 — in time to move carriers and task forces.
Process: HF intercept and DF → traffic analysis and call-sign/fist continuity → indicator recovery → additive stripping with difference indexes → codebook meaning recovery → Japanese translation → Rochefort estimate → Layton brief → Nimitz decision. Confirmations (water ruse) when designators are contested.
Hardware: Shore HF receivers (National HRO/RAS, Hammarlund/RBG and related Navy sets), DF nets and trucks, typewriters (English and kana), IBM keypunches/sorters/collators/reproducers/tabulators, teletype links, card stock by the ton, a windowless basement with a Marine at the door.
That triad — purpose, process, hardware — is the whole of Station HYPO. Everything else on this page is detail, conflict labels and physics footnotes. [DOCUMENTED]
ADDITIONAL NOTES ON HEʻEIA AND WAHIAWA
Heʻeia (Kaneohe Bay area) offered improved intercept conditions after the move from cramped Little Wailupe spaces in the 1930s station histories. When the high-power transmitting station that had shared Heʻeia moved to the Navy Yard, HYPO intercept took the vacated rooms. Later growth pushed intercept operations to Wahiawa near Schofield Barracks on the central Oahu plateau — a classic HF receiving site choice: away from shipyard noise, room for antenna fields, still linkable to Pearl Harbor analytic spaces. Rochefort’s cryptanalysts stayed in Building 1 until the 1943 Makalapa move; collectors and analysts were networked, not co-located in one romantic basement full of antennas. [DOCUMENTED / SINGLE-SOURCE for move narrative details]
ADDITIONAL NOTES ON SAFFORD
Laurance F. Safford organized and championed Navy cryptology and the intercept service between the wars. Without that scaffolding — sites, trained operators, early machine experiments, Rochefort’s placement at HYPO — Midway’s spring 1942 success has no foundation. His sidelining in the same season as Rochefort’s removal is part of the OP-20-G centralization story, not a separate gossip thread. [DOCUMENTED]
ADDITIONAL NOTES ON JOHN REDMAN AND JOSEPH WENGER
Washington figures associated with 1942 reorganization and machine-section management. Secondary histories cast them as antagonists to Rochefort’s independence; primary bureaucratic paper is thicker than this encyclopedia page. Operational fact: after Midway, Pacific naval COMINT control tightened toward Washington even as FRUPAC grew in Hawaii. [DOCUMENTED for centralization; SINGLE-SOURCE for personality narratives]
WORD COUNT AND SCOPE NOTE
This rewrite expands the pre-rewrite stub into a purpose / process / hardware reference with labelled conflicts. It is not a substitute for book-length treatments (Layton memoirs, specialized JN-25 monographs, NSA cryptologic histories). Use it as a structured briefing page with sources to follow. [DOCUMENTED]
EXPANDED END-TO-END EXAMPLE — MIDWAY WEEK
Compressed narrative tying every subsystem:
Japanese Combined Fleet staff traffic in JN-25b, partly readable, refers to objective AF and a schedule clustering on early June. Traffic analysis shows carrier-quality nets lighting up. DF and volume are consistent with a major sortie. HYPO argues Midway; Washington hedges toward the Aleutians as the weight of effort. Cable order to Midway; clear-text water failure broadcast; Japanese encrypted comment that AF lacks water; designator closed. Layton briefs Nimitz; Enterprise, Hornet, and Yorktown take the northeast ambush station. Dive bombers find the Kido Butai. Four Japanese fleet carriers lost. COMINT did not aim the bombs; it placed the carriers where the bombs could matter. [DOCUMENTED]
EXPANDED HARDWARE INVENTORY CAVEAT
No declassified packing list for the June 1942 dungeon has been pasted into this page. Receiver models are attested for the Oahu intercept plant and Navy shore inventory of the era. IBM machine types are attested for FRUPAC/HYPO machine rooms in veteran and secondary technical histories. Readers should not invent serial numbers or rack diagrams beyond photos. The three images kept on this article show (1) a busy multi-desk radio-intelligence operations floor, (2) an individual operator with headphones, recording reels and typewriter, and (3) a Building 1 / basement cutaway graphic. None is captioned on the page. [DOCUMENTED / ESTIMATE]
EXPANDED POLICY ON UNCERTAIN CLAIMS
Uncertain or conflicted claims called out in running text:
Treat those as labelled. Treat Midway operational outcome, Rochefort’s role, JN-25’s codebook-plus-additive structure, and IBM’s role in difference indexing as solid. [DOCUMENTED]
END MARK
Page complete. Images hypo1.jpg, hypo2.jpg, hypo3.webp retained without captions.