NAVY DEPARTMENT,
Washington, D.C.

CAPTAIN SAFFORD TESTIFIES
PROCEEDINGS OF THE HART INQUIRY
SATURDAY, APRIL 29, 1944 

(THIRTY-SECOND DAY)

The examination met at 9:35 a. m.

Present:

Admiral Thomas C. Hart, U. S. Navy, Retired, examining officer, and his counsel and assistant counsel.

Ship's Clerk Charles O. Lee, U. S. Naval Reserve, reporter.

The examining officer decided to postpone the reading of the record of proceedings of the thirty-first day of the examination until such time as it shall be reported ready, and in the meantime to proceed with the examination.

No witnesses not otherwise connected with the examination were present.

A witness called by the examining officer entered and was informed of the subject matter of the examination as set forth in the preface to the testimony of Rear Admiral W. W. Smith, Record Page 32.

The witness was duly sworn.

Examined by the examining officer:

1. Q. What is your name, rank, and present station?

A. L. F. Safford; Captain, U. S. Navy; Office of Naval Communications, Navy Department.

2. Q. What duties were you performing during the calendar year 1941?

A. I was the officer in charge of the Communication Security Section of Naval Communications. The word "Communication Security" was a covering title to mast Communication Intelligence, although we also performed security duties in the design and preparation of naval codes and ciphers and general communications security duties; that is, surveillance over their use.

3. Q How many units were there in that organization and where were they located?

A. There were three main radio intelligence units. One in the Navy Department with subsidiary direction finder stations and intercept stations along the Atlantic Seaboard and in the Atlantic Ocean. The second in size was located at Pearl Harbor with subsidiary intercept stations and direction finder stations in Oahu, Midway, Samoa, and Dutch Harbor. The third was located on the Island of Corregidor in the Philippines with intercept and direction finder station there and a small intercept and finder station on the Island of Guam. There had been an intercept and direction finder station at Shanghai, but it was evacuated to Corregidor in December, 1940.

There were also intercept and direction finder stations on the West Coast of the United States. The West Coast intercept stations fed their traffic directly into Washington; the direction finder stations were coordinated and controlled by the Commandant, Thirteenth Naval District.

4. Q. Were the three units somewhat specialized in their duties?

A. Yes.

5. Q. To what special branch of intelligence were the duties of the main station in Washington confined?

A. To naval operations in the Atlantic Ocean and to the plans and intentions of foreign governments. In addition to the foregoing duties, the Washington Unit had another important function: Training personnel for the other units so that they would be able to "pull their weight in the boat" when transferred to duty overseas. For this reason, the Washington Unit had, at this time, the most experienced personnel (some with over ten years of C. I. duty) and; the least experienced (ninety per cent with less than one year of C. I. duty). The Washington Unit had been standing continuous watches since February, 1941.

6. Q. To what special branch of intelligence were the duties of the main station at Pearl Harbor confined?

A. To the dispositions and plans of naval forces in the Pacific Ocean and to surveillance over Japanese naval communications. We expected that this would prevent the Fleet being surprised as the Russians had been at Port Arthur. These duties were prescribed in the current War Plans (WPDNC-8: Appendix IV; Art. 4-25) 1 approved March, 1940, and by dispatches and letters of instruction issued by the Chief of Naval Operations. These duties did not include surveillance over Diplomatic communications of any sort. The personnel of this Unit had about four or five years of C. I. experience on the average. The officers included our best, and six or seven had had previous C. I. duty in the Asiatic C. I. Unit.

7. Q. To what special branch of intelligence were the duties of the main station at Corregidor confined?

A. The Asiatic Unit was at the disposal of the Commander-in-Chief, U. S. Asiatic Fleet to use as he saw fit. During 1940 and early 1941, this Unit was mostly concerned with Japanese Diplomatic communications, but in October or November, 1941, it shifted its main attention to Japanese Naval communications. The personnel of this Unit had about two or three years of C. I. experience on the average, and the officers were young, enthusiastic, and capable.

8. Q. Were all three units kept in close touch with the results which they, individually, obtained?

A. Yes, sir, as far as it pertained to the technical work they were doing, but not otherwise.

9. Q. Was the unit at Pearl Harbor kept fully informed of the aforesaid results obtained by the Washington Unit?

A. Only as regards the operations of the Japanese Navy in the Pacific Ocean, with one important exception: On December 1, 1941, the Director of Naval Communications released OpNav Secret 011400; Urgent, to CinCAF and Com 16; Priority to CinCPac and Com 14; indicating that the Japanese were planning landing at Kota Bharu in Malaya.

10. Q. Did you consider that you, yourself, were kept fully informed of the results obtained by the units in Fourteen and Sixteen Districts?

A. Yes, I was kept informed of important developments by radio, and of details weekly, by air mail.

11. Q. During November, 1941, did you obtain, from the units in Fourteenth and Sixteenth Naval Districts, estimates covering the organization and distribution of the Japanese naval forces?

A. Yes, sir, we did on November 26. The message from Com 14 was the Pearl Harbor estimate as to the current organization and distribution of the Japanese Fleet with particular reference to recently organized task forces which were suspected to be employed in the invasion of Malaya and the Netherlands East Indies, plus a suspected covering force of submarines and one carrier division in the Japanese mandated islands. This was a summation of results obtained during the month of November, 1941.

12. Q. At approximately that date, or perhaps somewhat later, did you obtain a similar estimate from the unit in the Sixteenth Naval District?

A. Yes, sir, we did.

13. Q. Please state, very briefly, any respects in which the two estimates differed.

A. Com 16 disagreed with the estimate that carriers and submarines in force were in the mandated islands. Com 16's report placed one carrier division as operating in the South China Sea and the remaining carriers in Japanese home waters; and further added "evaluation considered reliable".

14. Q. What is the date of that estimate from the Sixteenth District unit?

A. That was the same date, November 26, 1941. It was sent twelve hours later than the Fourteenth District's dispatch.

15. Q. Do you recall what impression prevailed in the Washington unit as regards the disagreement between the other two?

A. We believed that the Sixteenth District unit was correct in their estimate.

16. Q. Did you advise the Fourteenth District unit to that effect?

A. No, sir we did not. On November 24, a dispatch from the Director of Naval Intelligence to the Commander-in-Chief, Asiatic Fleet; information Commander-in-Chief, Pacific Fleet; stated that Sixteenth District's intercepts were considered most reliable and requested that other reports be submitted to Com 16 for action, OpNav for information; and after combining all incoming reports, Com 16 forward reports to OpNav, info CinCPac. From this, we believed that both Commanders-in-Chief understood that we gave more weight to Com 16's reports than to Com 14's reports. The Department's directive was based on geographical considerations rather than ability of personnel or efficiency of the units.

17. Q. Did your unit, as engaged upon its specialized endeavor, obtain, during November and December, any definite information which indicated the objectives which the Japanese were preparing to attack?

A. Yes. sir.

18. Q. Please give, chronologically, with particular reference to dates, a brief summary of that information.

A. Going back to the late Spring of 1941, on May 22, we received positive proof of Japanese plans for the conquest of Southeastern Asia and the Southwest Pacific. On July 24, a high authority in Japan directed the withdrawal of merchant shipping from the Northeast Pacific, Southwest Pacific, and Indian Ocean. On September 4, we received information indicating Japan's determination to carry out her program of southward expansion and to expel the United States and England from China, Southeast Asia, and the Southwest Pacific. On October 15, we received unexpected confirmation of Japan's plans and intentions of the conquest of Southeastern Asia In October, 1941, the Japanese Consuls were directing and advising the evacuation of Japanese Nationalists from the Netherlands East Indies, Malaya, Philippines, Hawaii, America, and Europe. By October 28, this was in full progress. On November 4, we received important information that the internal situation in Japan, both [358] political and economic, since the American embargo, had become so desperate that the Japanese Government had to distract popular attention by a foreign war or else by bloodless diplomatic victory. On November 12, we received important information that the Japanese Government regarded November 25 as the dead line for negotiations then being conducted between the Japanese and American Governments to end. November 17, we received information from a very reliable source that Japan had no intention of attacking Russia in Siberia or she had changed her plans, if such intention ever existed. At one time, when it looked as if Moscow would fall, there were indications from several sources that Japan would invade Siberia. On November 24, 1941, we learned that November 29, 1941, Tokyo time, was definitely the governing date for offensive military operations of some nature. We interpreted this to mean that large scale movements for the conquest of Southeast Asia and the Southwest Pacific would begin on that date, because, at that time, Hawaii was out of our minds. On November 26, we received specific evidence of Japan's intention to wage an offensive war against both Britain and the United States. On December 1, we had definite information from three independent sources that Japan was going to attack Britain and the United States, and, from two of them, that Japan would maintain peace with Russia. On December 4, 1941, we received definite information from two more independent sources that Japan would attack the United States and Britain, but would maintain peace with Russia. At 9:00 p. m. (Washington time), December 6, 1941, we received positive information that Japan would declare war against the United States, at a time to be specified thereafter. This information was positive and unmistakable and was made available to Military Intelligence at this same time. Finally, at 10:15 a. m. (Washington time), December 7, 1941, we received positive information from the Signal Intelligence Service (War Department) that the Japanese declaration of war would be presented to the Secretary of State at 1:00 p. m. (Washington time) that date. 1:00 p. m. Washington time was sunrise in Hawaii and approximately midnight in the Philippines, and this indicated a surprise air raid on Pearl Harbor in about three hours. Kramer appended a note to this effect to the paper sent over from S. I. S. before presenting it to the Secretary of the Navy. I do not know whether or not a copy of this note was appended to the paper given to Admiral Stark.

At this same time, information was also received indicating that Japan was about to commence hostilities against the British Empire. This information was sent over to S. I. S. immediately.

19. Q. Going back over that series, to what officials did your unit transmit the information concerning which you have just testified?

A. My unit transmitted information directly to Signal Intelligence in the War Department and to Naval Intelligence representative (that is, Commander A. H. McCollum, head of the Far Eastern Section, or Lieutenant Commander A. D. Kramer, attached to the Far Eastern Section of Naval Intelligence but actually working in the Communication Intelligence Unit). The further distribution of information within the Navy and to the President normally was undertaken by Kramer in his status as a subordinate to McCollum. Information was distributed daily, as a matter of routine, to the President, to the Secretary of the Navy, to the Chief of Naval Operations, the Assistant Chief of Naval Operations, the Director of War Plans, the Director of Naval Communications, and, of course, the Director of Naval Intelligence. Within the Army, the Signal Intelligence Service, our opposite numbers, gave information to G-2, or Military Intelligence, and Colonel Bratton, head of the G-2 Far Eastern Section, distributed the information to the Secretary of State, the Secretary of War, Chief of Staff, Director of War Plans, and, of course, to the Director of Military Intelligence. On Special occasions, information was disseminated at night or whenever it came [359] in. There was a direct exchange of information between S. I. S. and the Navy Department C. I. Unit; also between O. N. I. and M. I. D.

20. Q. Was the foregoing information communicated to officials in the State Department?

A. It was always given to Secretary Hull and sometimes given to Under Secretary Sumner Welles. In the Spring of 1941, the information had gone further but, after a leak to the German Embassy, it was restricted to Secretary Hull and Secretary Welles.

21. Q. Is there any documentary report which shows the date and hour of delivery of the foregoing information to various officials?

A. There is no documentary evidence.

22. Q. Are you able to state, from memory, the date and hour on which the important information, say, from 1 December onward, was transmitted a

A. I can, from my recollection of Lieutenant Commander Kramer's verbal reports to me.

23. Q. Please give what you recall as regards those dates and hours.

A. The information on December 1, 2, 3, 4, 5, and 6, was disseminated about eleven a. m., within the Navy Department and was then given to the Naval Aide to the President who took it over to the White House, some time in the early afternoon. The "Winds Message" was given a special distribution shortly after eight a. m. on December 4, 1941, and was also included in the routine distribution. The information received late on December 6 was highly important and was distributed as a rush job by Lieutenant Commander Kramer, who left the Navy Department in an official station wagon shortly after nine p. m., and who had reached his last official-by eleven p. m. Kramer returned to the Navy Department about one a. m. in the morning to see if there was any further information and then went home.

He came down the next morning in time to give Admiral Stark written information at the Admiral's nine o'clock conference. Much of the December 6 information was distributed over the telephone by Admiral Wilkinson and by Secretary Hull. The following officials were given this information that night: President Roosevelt (via the White House Aide), Secretary Hull, Secretary Stimson, Secretary Knox. Admiral Stark, Rear Admiral Turner, Rear Admiral Wilkinson, Rear Admiral Beardall. Lieutenant Colonel R. S. Bratton, U. S. Army, was given the same information at nine p. m. for dissemination to War Department officials, and we did not know any more, except that he got a copy over to Secretary Hull by ten o'clock. As regards information on the 7th, the same officials had this information by eleven a. m.-some sooner.

24. Q. Was any of the foregoing information, under dates of November and December, 1941, disseminated by the main Washington unit direct to the corresponding unit in Fourteenth Naval District?

A. No, sir. That was not permitted by a written order then in force; but there was one exception. On the 3rd of December, I prepared OpNav Secret Dispatch 031855, which was released by Captain Redman, the Assistant Director of Naval Communications. A similar dispatch was released by Admiral Wilkinson and filed at 031850. Admiral Wilkinson's message is referred to in the Roberts report. Before drafting my message, I called Commander McCollum on the telephone and asked him, "Are you people in Naval Intelligence doing anything to get a warning out to the Pacific Fleet?" And McCollum replied, "We are doing everything we can to get the news out to the Fleet." McCollum emphasized both "we's". In sending this information, I was overstepping the bounds as [360] established by approved war plans and joint agreement between Naval Communications and Naval Intelligence, but I did it because I thought McCollum had been unable to get his message released. OpNav 031855 was addressed to CinCAF and Com 16 for action, but was routed to CinCPac and Com 14 for information. It was written in highly technical language and only one officer present at Pearl Harbor, the late Lieutenant H. M. Coleman, U. S. N., on CinCPac's Staff, could have explained its significance.

25. Q. Did the unit in the Fourteenth Naval District have any material from which they could have gained this information through their own efforts?

A. No, sir, they did not have the material and they could not possibly have gained this information.

The examining officer did not desire to further examine this witness.

The examining officer informed the witness that he was privileged to make any further statement covering anything relating to the subject matter of the examination which he thought should be a matter of record in connection therewith, which had not been fully brought out by the previous questioning.

The witness made the following statement: The C. I. Unit in Washington had no authority to forward to the C. I. Units in Pearl Harbor or Corregidor, or to the Commanders-in-Chief direct, any information other than technical information pertaining to direction finding, interception, and so forth. The dissemination of intelligence was the duty, responsibility, and privilege of the Office of Naval Intelligence as prescribed in Communication War Plans approved by the Chief of Naval Operations in March, 1940. On the 4th of December, 1941, Commander McCollum drafted a long warning message to the Commanders-in-Chief of the Asiatic and Pacific Fleets, summarizing significant events up to that date, quoting the "Winds Message", and ending with the positive warning that war was imminent. Admiral Wilkinson approved this message and discussed it with Admiral Noyes in my presence. I was given the message to read after Admiral Noyes read it, and saw it at about three p. m., Washington time, on December 4, 1941. Admiral Wilkinson asked, "What do you think of the message?" Admiral Noyes replied, "I think it is an insult to the intelligence of the Commander-in-Chief." Admiral Wilkinson stated, "I do not agree with you. Admiral Kimmel is a very busy man, with a lot of things on his mind, and he may not see the picture as clearly as you and I do. I think it only fair to the Commander-in-Chief that he be given this warning and I intend to send it if I can get it released by the front office." Admiral Wilkinson then left and I left a few minutes later. At the time of the Japanese attack on Pearl Harbor, I thought that this warning message had been sent, and did not realize until two years later, when I studied the Roberts report very carefully, that McCollum's message had not been sent. In order to clarify the above statement and my answer to a previous question, it is necessary to explain what is meant by the "Winds Message". The "Winds Message" was a name given by Army and Navy personnel performing radio intelligence duties to identify a plain-language Japanese news broadcast in which a fictitious weather report gave warning of the intentions of the Japanese Government with respect to war against the United States, Britain (including the N. E. I.), and Russia. We received a tip-off from the British in Singapore in late November, 1941, which was immediately forwarded to the Navy Department by the Commander-in-Chief, U. S. Asiatic Fleet, with an information copy to the Commander-in-Chief, Pacific Fleet. We also received a tip-of from the Dutch in Java through the American Consul General and through the Senior Military Observer. The Dutch tip-off was handled in routine fashion by the coding rooms of the State [361] Department, War Department, and Navy Department. The Director of Naval Intelligence requested that special effort be made to monitor Radio Tokyo to catch the "Winds Message" when it should be sent, and this was done. From November 28 until the attack on Pearl Harbor, Tokyo broadcast schedules were monitored by about 12 intercept stations, as follows: N. E. I. at Java; British at Singapore; U. S. Army at Hawaii and San Francisco; U. S. Navy at Corregidor, Hawaii, Bremerton, and four or five stations along the Atlantic seaboard. All Navy intercept stations in the continental United States were directed to forward all Tokyo plain-language broadcasts by teletype, and Bainbridge Island ran up bills of sixty dollars per day for this material alone. The "Winds Message" was actually broadcast during the evening of December 3, 1941 (Washington time), which was December 4 by Greenwich time and Tokyo time. The combination of frequency, time of day, and radio propagation was such that the "Winds Message" was heard only on the East Coast of the United States, and even then by only one or two of the Navy stations that were listening for it. The other nations and other Navy C. I. Units, not hearing the "Winds Message" themselves and not receiving any word from the Navy Department, naturally presumed that the "Winds Message" had not yet been sent, and that the Japanese Government was still deferring the initiation of hostilities. When the Japanese attacked Pearl Harbor, the British at Singapore, the Dutch at Java, and the Americans at Manila were just as surprised and astonished as the Pacific Fleet and Army posts in Hawaii. It is apparent that the War Department, like the Navy Department, failed to send out information that the "Winds Message" had been sent by Tokyo. The "Winds Message" was received in the Navy Department during the evening of December 3, 1941, while Lieutenant (jg) Francis M. Brotherhood, U. S. N. R., was on watch. There was some question in Brotherhood's mind as to what this message really meant because it came in a different form from what had been anticipated. Brotherhood called in Lieutenant Commander Kramer, who came down that evening and identified that message as the "Winds Message" we had been looking for. The significant part of the "Winds Message" read: "HIGASHI NO KAZE AME. NISHI NO KAZE HARE. The negative form of KITA NO KAZE KUMORI". The literal translation of these phrases is: "EAST WIND RAIN. WEST WIND CLEAR. NEITHER NORTH WIND NOR CLOUDY". The meaning of this message from the previously mentioned tip-off was: "War with the United States. War with Britain, including the N. E. I., etc. Peace with Russia". I first saw the "Winds Message" about 8:00 a. m. on Thursday, December 4, 1941. Lieutenant A. A. Murray, U. S. N. R., came into my office with a big smile on his face and a piece of paper in his hand and said, "Here it is!" as he handed me the "Winds Message". As I remember, it was the original yellow teletype sheet with the significant "Winds" underscored and the meaning in Kramer's handwriting at the bottom. Smooth copies of the translation were immediately prepared and distributed to Naval Intelligence and to S. I. S. in the War Department. As the direct result of the "Winds Message", I prepared a total of five messages, which were released between 1200 and 1600 that date, ordering the destruction of cryptographic systems and secret and confidential papers by certain activities on the Asiatic Station. As a direct result of the "Winds Message", McCollum drafted the long warning message, previously referred to, which was disapproved by higher authority, but which the Navy Department C. I. Unit believed had been sent. Both Naval Intelligence and the Navy Department C. I. Unit regarded the "Winds Message" as definitely committing the Japanese Government to war with the United States and Britain, whereas the information of earlier dates had been merely statements of intent. We believed that the Japanese would attack by Saturday (December 6), or by Sunday (December 7) at the latest. The following officers recall having seen and having read the "Winds Message":  Captain L. F. Safford, U. S. N., Lieutenant Commander F. M. Brotherhood, U S. N. R., Lieutenant Commander A. A. Murray, U. S. N. R., and Lieutenant (jg) F. L. Freeman, U. S. N. The following officers knew by hearsay that the "Winds Message" had been intercepted but did not actually see it themselves: Commander L. W. Parke, U. S. N., Lieutenant Commander G. W. Linn, U. S. N. R., Ensign Wilmer Fox, U. S. N., and Major F. B. Rowlett, Signal Corps Reserve. The following officers should have some recollection of the "Winds Message": U. S. Navy-Rear Admiral T. S. Wilkinson, Captain A. H. McCollum, Colonel R. A. Boone (U. S. Marine Corps), Commander G. W. Welker, Commander A. D. Kramer, Lieutenant Commander A. V. Pering, and Ship's Clerk H. L. Bryant. U. S. Army-Brigadier General T. J. Betts, Colonel O. K. Sadtler, Colonel R. S. Bratton, Colonel Rex Minckler, Colonel Moses Pettigrew, Colonel Harold Doud, and Lieutenant Colonel R. E. Shukraft. The "Winds Message" was last seen by myself about December 14, 1941, when the papers which had been distributed in early December were assembled by Kramer, checked by myself, and then turned over to the Director of Naval Communications for use as evidence before the Roberts Commission, according to my understanding at the time. Further information as to Pearl Harbor's estimates of locations of Japanese forces in early December, 1941, may be found in the monthly report of Station "H"-in the "Chronology" which was prepared daily and forwarded weekly by air mail. This information was, of course, prepared by and currently available to the Pearl Harbor C. I. Unit but was not received in the Navy Department until a delay of about two weeks.

Note: The examining officer has identified the documents mentioned by witness as being C. I. Station "H" "Chronology" for December 1-December 6, 1941, inclusive, now on file in Communication Intelligence Section (Op 20G), Office of Director, Naval Communications, Communication Annex, Navy Department, Washington, D. C., to which is attached a summary of more important extracts, made by the witness under examination.

The witness was duly warned and withdrew.

The examination then, at 11:37 a. m., was adjourned to await the call of the examining officer.