The Earhart Tragedy: Old Mystery, New Hypothesis

Over the decades since Amelia Earhart and Fred Noonan disappeared in the Pacific Ocean on 2 July 1937 while attempting to fly around the world, many authors have focused on the fliers’ alleged incompetence as the reason they did not find Howland Island.

Earhart, despite having achieved numerous aviation firsts, often is portrayed as unqualified, even though she was the first woman to fly the Atlantic Ocean and the first person of either sex to fly from Hawaii to the mainland of North America.

Critics say that Noonan, although without question a top-flight navigator, was a hopeless alcoholic who was either drunk or hung over when most needed. Captain Almon Gray, U.S. Naval Reserve (Retired), who was Assistant Communications Superintendent for Pan American Airways’ Pacific Division when he left to go on active duty in 1942, knew Fred Noonan well and flew with him a number of times in the Pan Am Clippers. Gray told me that Noonan always showed up for a flight precisely on time but usually looking a bit hung over. Once aloft he would have some coffee and then do a superb job of navigation. He never drank during a flight. Unknown to many researchers, Noonan held a second class Commercial Radiotelegraph License, which he obtained two years before his death, and he often stood by for the Clippers’ radio operators when needed. They worked in CW (continuous wave, i.e., Morse code) exclusively.

In contrast with the sniping at the Earhart-Noonan team, Commander Warner K. Thompson, U.S. Coast Guard, commanding the USCGC Itasca (WPG-321)—the vessel waiting at Howland Island to guide her in—has received lavish praise. Consider what the Commandant, 14th Naval District, reported at the time to the Chief of Naval Operations:

“Commander W. K. Thompson, USCG, has been commended by letter to his immediate superior. His intelligent and zealous conduct of the initial phase of the search under most trying conditions deserves especial commendation. His reports, together with the wholehearted cooperation of the Commander, Hawaiian Section, U. S. Coast Guard, were of great assistance to the subsequent conduct of operations by the Navy. The performance of the ITASCA was excellent in all respects throughout the flight and the search. Careful study of all communications and other information pertaining to the flight and the preparations therefor indicate clearly that ITASCA left nothing undone to insure the safe completion of the Earhart flight.”

Many investigating the disappearance probably took this at face value and directed their attention elsewhere. I saw it as military service politesse, however. Generous letters of commendation are an old tradition in all fighting forces—but they can distort history. A detailed examination of Commander Thompson’s performance reveals a different story. Far from acting intelligently and zealously, he must have so embarrassed then-Rear Admiral R. R. Waesche, Commandant of the Coast Guard, and Secretary of the Treasury Henry Morgenthau, Jr.—to whom the service then reported—that they thought it best the facts remain undisclosed.

With the fliers unavailable to defend themselves, only one side of the story has come to light—until now. Of course, Commander Thompson is not here to defend himself either. If, because of that, my judgment of him seems too harsh, it must be compared with the unrestrained attacks he made on Earhart in his search report.

Stonewalling

Shortly after the search for the missing fliers ended, Navy Commander P. V. H. Weems, a highly regarded navigator and navigation instructor, wrote to Rear Admiral Waesche asking for copies of files concerning the disaster. Weems knew Noonan, at least through correspondence, and was motivated to discover what had happened. Following is the terse reply:

Weems System of Navigation
30 August, 1937
Annapolis, Maryland

Sirs:

Reference is made to your letter of August 14, 1937 in which you request the file of messages from the U. S. Coast Guard Cutter ITASCA concerning the loss of the Earhart plane. While realizing the merit of your plans and that the conclusions drawn from a study of the information contained in the dispatches would be of value to flyers [sic], it is believed inadvisable to submit any of the information for study or publication.

I regret very much that this decision seems best after a consideration of all factors in the case and that we are unable to comply with your request.

Very truly yours,

R. R. Waesche
Rear Admiral, U. S. Coast Guard,
Commandant

What led to this decision? The logical conclusion is that the Itasca’s message file contained embarrassing information. Apparently what was embarrassing was that Commander Thompson’s actions appeared to be factors in the loss of Earhart and Noonan.

Admiral Waesche died long ago, but his son later became a flag officer in the Coast Guard and retired in 1971. In correspondence, I found him completely forthright, cooperative, and gracious. When I asked what his father thought happened to Earhart and Noonan, however, he said that his father “ . . . never at any time discussed Amelia Earhart.”

Background

The Itasca was at Howland Island to provide communications, smoke signals, and radio bearings to guide Earhart and Noonan as they approached the small isolated island in the mid-Pacific. I reject as fanciful the many conspiracy and faulty navigation theories involving the loss of the two fliers. Earhart and Noonan attempted to fly from Lae to Howland Island, arrived in the vicinity of the island short of fuel, and went into the sea nearby trying to find it. Things were what they seemed to be.

Those who have flown over the sea when the sun is bright and low, with cumulus clouds about, know how difficult it would be to see a tiny island having a highest elevation of only about 15 feet. Looking toward the sun one sees only a blinding, shimmering path of silvery reflected sunlight in an arc about 15° to 20° wide; within that arc nothing can be seen. Elsewhere, numerous cloud shadows look exactly like islands.

The Itasca was making smoke, but it would have been conspicuous only if seen from sea level with a light blue sky in the background. Earhart and Noonan, however, were flying at 1,000 feet and the smoke seen from that perspective would have had an inky, blue-sea background. Color contrast would have been minimal; the odds were against them.

Tiny Howland would have been difficult to spot in any case, but none of this would have mattered if the fliers could have received a radio bearing for final guidance. That was not to be.

Failure to Provide Timely Radio Beacon Signal

The Itasca failed to provide a timely radio beacon signal for the fliers to home on. Her 550-270 kiloHertz (kHz) radio direction finder and 500 kHz beacon transmitter do not appear to have been manned until 0730 ship’s time, according to the log kept by Radioman Third Class T. J. O’Hare. The plane by that time would have been nearing Howland and the fliers would have been trying to find the island visually. By 0730, they likely would have given up trying to find the radio beacon, thinking their radio direction finder was not working.

Commander Thompson should have had Radioman Third Class O’Hare on watch at least from the time the plane was about 200 miles from destination (i.e., at 0615 ship’s time or earlier) and should have been transmitting a beacon signal on 500 kHz—not listening on that frequency. Almost every time “500 KCs [kilocycles]” (kilocycles rather than kiloHertz was the term in use at the time) is mentioned in the logs, one kept by O’Hare and the other by Radioman Third Class W. L. Galten, it is in the context of a request on 3105 kHz that Earhart transmit on 500 kHz so the Itasca could get a bearing—or a simple note such as “LSNIN [listening] 500” or “NIL [nothing] FROM KHAQQ [the aircraft] 500.” It seems clear that the Itasca was listening on 500 kHz, not transmitting a steady beacon signal. One cannot do both at the same time.

Earhart and Noonan simply could not transmit on 500 kHz. They depended on their radio direction finder and could have taken bearings on the cutter’s 500 kHz transmitter if it had been in operation. The fliers needed a continuous beacon signal on that frequency, except during the plane’s scheduled transmissions, and there was none.  Had there been one, it could have guided them to Howland Island.

All involved evidently misunderstood who was to take the bearings, ship or plane, so the cutter’s crew listened on 500 kHz when they should have transmitted. But they did neither until the plane was already almost at destination.

Failure to Support Radio Direction Finder on Howland

On 5 July, Commander Thompson reported in a long message to Coast Guard Headquarters (with copy to
San Francisco Division) that “SHIP [ITASCA] MET ALL EARHART REQUESTS WITH EXCEPTION INABILITY TO SECURE EMERGENCY RADIO BEARING ON 3105 KILOCYCLES DUE BRIEF EARHART TRANSMISSIONS AND USE VOICE. . . .” He is on the defensive here and attempting to shift blame squarely to Earhart. The “USE [of] VOICE” would not have prevented bearings being taken. In any case, the cutter’s 0756 radio log entry does not bear him out. At that time, Earhart requested bearings and made a series of “long dashes,” i.e., unmodulated carrier. She had made several transmissions that were too short to DF (get a direction-finding bearing), but she did not do it this time.

Commander Thompson’s report does not tell the rest of the story. Richard B. Black, Department of the Interior, and Radioman Second Class Frank Cipriani had brought aboard the Itasca a portable radio direction finder (RDF) that could tune the high frequencies used by Earhart for communications, with the intention of setting it up on Howland Island. For no apparent reason, Commander Thompson at first flatly refused to put Cipriani and his equipment ashore on the island. It could have been because he regarded his ship as responsible for guiding Earhart to Howland and he did not want anyone else to steal his thunder.

Black, however, was determined that Cipriani and his RDF equipment would go ashore. Eventually, he prevailed, but Commander Thompson gave only grudging support and sent Cipriani ashore with a battery of inadequate capacity. Despite minimum use, the battery was totally discharged just when it was needed most—when Earhart desperately wanted a bearing taken and was sending those long dashes.

But for this, bearings almost certainly could have been taken, although they could not have been sent to the plane because Earhart was not receiving voice transmissions from the Itasca. The bearings could have been a lifesaver during the rescue attempt, however, giving the searchers a better idea of where to look for the downed plane.

Flawed Search Pattern

Howland Island was actually about 5.8 nautical miles from its charted position. Commander Thompson visited it on a regular schedule and knew its correct position, but he did not inform Earhart and Noonan of the error when exchanging messages with the two fliers before they departed on the final and fatal flight from Lae, New Guinea.

With quick rescue a matter of life or death after the plane was overdue, Commander Thompson got under way at 1040 ship’s time on course 337°, which he soon changed to 338° (evidently correcting ship’s course for drift, which he should not have done). At 1400, he headed east for five hours, away from the line-of-position (LOP) where the plane most likely went down. He spent the night of 2-3 July chasing after meteors that he mistook for flares. (See Figure 1.)

Instead, he should have searched along the 337°/157° LOP reported by Earhart. To define the search area, one would start by drawing the LOP on the chart through both the charted position of Howland Island and then through its true position. [Howland’s charted position was 0 degrees 53 minutes N (north), 176° 35 minutes W (west)); its true position was 0° 48 minutes N, 176° 38 minutes W.]

Next, one would have to take into account that the LOPs could be in error as much as 10 nautical miles, so the pattern would have to be made 20 nautical miles wider (10 nautical miles on each side).  Earhart did not say how far up or down the LOP she was flying. I think a reasonable assumption for the purpose at hand would be 45 nautical miles in each direction from the vicinity of Howland. This top-priority search pattern therefore would be about 24 nautical miles wide—covering the possible LOP error and the ambiguity of the reference point—and 90 nautical miles long. That produces a 2,160-square-mile pattern, shown in Figure 1, which would have been moving northwest with the current at about two knots.

The search should have been conducted during daylight hours only.  At night, the cutter should have drifted, letting the current take both vessel and pattern in the same direction. The pattern could have been searched in two passes. Steaming at 14 knots (with sunset at 1825 and sunrise at 0615 ship’s time), the 13-hour search would have been completed by about 1130 ship’s time on 3 July—the morning after the plane disappeared.

Commander Thompson searched only about one-third of this area. (Again, see Figure 1). Why he thought that Noonan could wander to a position 100 miles east of the LOP that the fliers reported is hard to imagine. This would have required an error, normal to the LOP, 10 times Noonan’s usual maximum error.¹ When Earhart reported, “We are on you but can’t see you,” she was probably no more than 10 miles from the Itasca and Howland. Unable to see the tiny island, it is logical that she would fly up and down the LOP and that is precisely what she told the Itasca she was doing. Why would Commander Thompson look elsewhere?

In any event, on the morning of 3 July he was back, drifting off the coast of Howland, waiting for a reply to a message he had sent to the Commander of the 14th Coast Guard District and the USS Swan (AVP-7). The message had asked whether he should load gasoline or depart on further operations. The gas was 1,600 gallons of aviation fuel in 50-gallon drums put on Howland to refuel Earhart’s aircraft. He wasted one hour and 31 minutes of precious daylight search time before he received the obvious reply and got under way—again off to the east of the LOP high-probability search area. The cutter’s crew failed to consider a two-knot current that was shifting the pattern northwest when the Itasca resumed the search on the morning of 3 July. (See Figure 2). At that very moment, Earhart and Noonan may have been clinging to sinking wreckage 10 or 20 miles away.

On 4 July, he sent a message to the Hawaiian Section (with information copies to San Francisco and the 14th Coast Guard District) saying “JUDGE SHE CAME DOWN BETWEEN 337 [degrees] AND 90 [degrees] FROM HOWLAND AND WITHIN 100 MILES.”

In their 5 July reply, San Francisco belatedly told the Itasca: “OPINION OF TECHNICAL AIDS HERE THAT PLANE WILL BE FOUND ON ORIGINAL LINE OF POSITION. . . .” After two days, the staff had finally realized he was searching in the wrong place and told him so.

Open-ocean ditchings are extremely hazardous. Without shoulder harnesses, pilots often pitched forward into the instrument panel. The sea was unusually calm around Howland that morning, however, and she might have ditched successfully. The plane, because of its large empty tanks, would have had about 5,000 pounds of positive buoyancy, but the center of buoyancy was well aft of the heavy engines and the aircraft certainly would have been nose down in the sea. The cockpit soon would have filled with water, forcing the tail higher and making it difficult to get to the rubber raft, water, and emergency gear stored in the tail section.

Unless they were able to get into the raft, they could not have coped for long under those conditions. Hours counted—but if they survived the initial crash, probably hurt, they waited in vain. What happened in the first two days of the search was all important. Hours, days, and nights went by with Commander Thompson’s ship wandering aimlessly—usually far from the reported line of position.

Exaggerated Search Reports

In his 6003-1250 message of 3 July, Thompson claimed to have searched “3,000 square miles.” His deck log shows he steamed 268 miles. Therefore, he made the assumption that he could at all times see a plane on the water at a distance of up to 5.6 miles on either side of his course. But the cutter covered only about 124 miles during daylight—and only about 55 of them on or within 10 miles of the LOP. Most of the night he was on random courses far to the east of the LOP search area and could see practically nothing in the darkness—except the meteors that he mistook for flares. A more accurate report would have claimed only 616 square miles searched. He completely misled headquarters.

In a later message, Thompson claimed to have “searched 1,500 square miles during the night.” This concept of searching is hard to accept. It seems to assume that the downed fliers would still be alert, be able to see the ship’s searchlight, and be able to launch flares to attract attention. A partly submerged plane, miles away, could not easily be seen at night, with or without the “vigilant lookouts [and] and powerful searchlights” mentioned in his messages.

The Bogus Howland Radio Log

For years, many details of the search for the missing fliers were classified. They were declassified finally and released as required under the Freedom of Information Act, but the picture remains obscured today, perhaps unintentionally, by a pea soup fog of disinformation that continues to mislead researchers.

It is interesting to speculate on what person(s) may have written one of the strangest documents that survives from that era: the Howland radio log. Today, it is virtually impossible to determine who concocted it; in any event, this record of the DF station on Howland Island is a counterfeit, according to the two men still alive who were on the island at the time and are alleged to have participated in writing it: Yau Fai Lum and Ah Kin Leong (see below). The Itasca’s deck logs, radio logs, message traffic and Commander Thompson’s Earhart Search Report (of which at least two versions exist), however, all support the fiction of a radio and DF watch on Howland during the Itasca’s search for the fliers. If the Howland radio log is bogus, it follows that these other fundamental documents also may be suspect in some details.

If this sound like classic “conspiracy theory,” remember that all material was classified promptly and researchers like Commander Weems and Paul Mantz were denied access—and access for all researchers continued to be denied for many years. Francis X. Holbrook, who wrote “Amelia Earhart’s Last Flight” (Naval Institute Proceedings February 1971, pages 48-55), concluded that he had been led astray and put to considerable trouble by misinformation. Such misinformation did not require a large conspiracy; indeed, a single mischief maker could have been responsible. In the absence of access to factual data, researchers were at the mercy of whatever tidbits of information—or misinformation—that were leaked to them. I put my trust in the following accounts of Lum and Leong, both of whom maintain that the log was cooked.

In the mid-1930s, both the United States and Great Britain claimed the Line Islands, which included Jarvis, Baker, and Howland. Thinking they might one day prove to be of strategic value, the United States occupied them in 1935 to reinforce its claim. Four-man civilian teams were landed on each and rotated from time to time; all were trained as weather observers, and each team included one amateur radio operator. They sent daily weather reports to another amateur radio station in Honolulu, which passed them to Pan American Airways, then pioneering trans-Pacific clipper flights.

Richard B. Black, Department of the Interior, recruited men in the Hawaiian Islands for the teams; nearly all were Americans of Chinese or Hawaiian descent. Yau Fai Lum operated the amateur radio station on Howland Island in July 1937. Serendipitously, while searching the Radio Call book for a fellow ham operator with whom I had lost contact, I came across the new call letters, name, and address of Yau Fai Lum near where my friend would have been listed.

I began a long and interesting correspondence with Lum.² He told me of that fateful day when they waited in vain for Amelia Earhart and Fred Noonan. I was impressed by his almost total recall of details. Lum said he had washed his sheets and aired his bed, the best on the island, so that Amelia could lie down and rest in comfort after her long, exhausting flight. He described the private shower improvised for her—a 50-gallon drum of fresh water with canvas enclosure.

When I asked Lum about Cipriani and the high-frequency DF equipment, he replied: “I never met him.”

“How could you not meet him?” I asked. “Didn’t you, Henry Lau, and Ah Kin Leong live with him on that fly-speck island that had only one sleeping shelter and one 15-foot-long tent for a kitchen/dining area? And didn’t you report to him and stand radio watches under his direction during the 16 days that the Itasca was at sea searching for Amelia?” I enclosed a copy of the Howland Radio Log, which had numerous entries supposedly made by Lum.

He wrote back immediately:

• The Howland radio log is a “ . . . fraud—it is B.S.” He never met Cipriani, never stood watches on Amelia’s frequency, and never operated a direction finder. He said that two men (presumably Cipriani and Black) came ashore with a direction finder and set it up in a tent. He walked by the tent a few times in the course of his work and heard bits of conversation in passing but did not intrude. He told me several times that he did not want to get in the way of other people when it was not his business to do so. My impression is that he is by nature rather shy.

He also pointed out that his name in the log was consistently spelled wrong (as “Yat” Fai Lum) where he supposedly signed off at the end of each watch. Yat is a common Chinese name, but his is “Yau” not “Yat.” He added, “I should know how to spell my own name. According to the Howland log,” he continued, “other operators were Henry Lau, [call letters] K6GAS, and Ah Kin Leong, K6ODC. But they were not even on the island during the 16-day search. They and Cipriani were on board the Itasca.”

Henry Lau is dead, but I wrote to Ah Kin Leong, ex-K6ODC. Asked what he knew of all this, he replied on 4 September 1994: “No idea who wrote the false log. I stand no radio watch on Howland Island. Cipriani, Henry Lau and me was on the Coast Guard Cutter Itasca when it left Howland Island looking for Earhart.” (According to the Itasca’s deck log for 2 July, when it became evident that Earhart was overdue and in trouble, the landing party [no exceptions are mentioned] returned to the cutter, which departed to begin a 16-day search for the missing fliers. It does not say that Cipriani or other radio operators remained ashore. The 18 July deck log entry, however, states that they reboarded from Howland on that date.)

On 4 July, the Commander Hawaiian Sector had sent the following message to the Itasca: “HAVE HOWLAND DIRECTION FINDER BE ON STANDBY FOR BEARINGS.” Thompson would have been hard put to explain that he could not comply because he had Cipriani on board. So, inserted in the cutter’s log dated 4 July, is this message supposedly sent to K6GNW (Lum’s call letters) on Howland as follows: “MR. BLACK SEZ CIPRIANI IS IN CONTROL ES TO KEEP CONTINUOUS WATCH ON 3105 ES TAKE BEARINGS USE CHINESE OPS HR IF YOU HAVE TROUBLE HAVING BOYS STAND WATCHES MR. BLACK SEZ TO TELL JIMMY.” When I asked Lum about this, he replied: “It is all B.S.”

Was “HR” a Freudian slip or just a careless error? It means “here,” not “there,” as intended. If Cipriani really had remained on Howland when the landing party was recalled, it seems logical that such a message would have been sent to NRUI2, the call letters he used with the portable radio equipment that he took ashore along with the portable direction finder. It is very unlikely that such a message would have been sent to K6GNW (Lum’s station) instead.

Why take the word of Yau Fai Lum and Ah Kin Leong against that of Commander Thompson? For starters, neither has an ax to grind. Lum impresses me as being a highly reliable and completely sincere man. When he returned from Howland Island he acquired a Commercial First Class Radiotelephone license, joined the Honolulu Police Department, and eventually headed its Radio Maintenance Department until he retired. Commander Thompson, on the other hand, had a motive to distort facts.

Circumstantial details, such as the misspelling of Lum’s name, support Lum’s and Leong’s statements. The log also incorporates a one-day date error in all entries (a day must be subtracted to get the correct date), and uses a +10 1/2 time zone instead of +11 1/2. It seems quite unlikely that an error of one whole day would persist in a radio log, day after day, if it were kept by four operators as claimed. Surely, at least one of them would have known the correct date. Furthermore, the first page of the bogus log is headed “Radio log ITASCA,” at least in the version that I have. That looks like another Freudian slip by someone assigned to the Itasca—or whoever actually wrote the log while on board the Itasca instead of on Howland.

Why would Thompson take Cipriani off Howland Island when the cutter departed on the search? He did not want him there in the first place, but he probably was not thinking about Cipriani at all when he ordered the landing party back to the ship.

To sound plausible, the log had to be written by a person with detailed knowledge of what was going on at the time, and who was familiar with the usual log details, radio procedure signs, and jargon—i.e. a radio operator.

Chief Radioman L. G. Bellarts had died, and I contacted his son to ask what papers and memoirs his father left. Bellarts, I was told, recognized the historical importance of the radio logs and took personal possession of them soon after the plane was lost. He guarded them carefully for years, but eventually sent them to the National Archives. I wrote asking whether the logs they were giving researchers were direct copies of those received from Bellarts, or had they come perhaps from some other source? To date, I have received no reply.

I have been unable to locate Frank Cipriani or any close relative. Dwight Long, another Earhart researcher, told me that Cipriani became a civilian radio operator and was lost at sea in a World War II convoy when his ship was torpedoed.

A retired officer who signed the Itasca’s deck log as “W. I. Swanston, Lt. (j.g.) USCG, Navigator,” confirmed being in the ship’s company during the search for the fliers but denied having been the navigator, despite clear evidence to the contrary. He was 86 when I contacted him. When I pointed out that he had definitely signed the Itasca’s deck log as navigator, I got a rude answer. I asked him a question concerning navigation on board the cutter and he again replied rudely that I did not know anything about navigation. He did not seem to understand the point I was trying to make. He seemed to be in a bad mood. I think that he had long ago forgotten any details concerned with navigation and did not want to be associated in any way with the incident.

Clearly revealing high-level concern for embarrassment are the actions taken by then-Secretary of the Treasury Henry Morgenthau, Jr. The U. S. Coast Guard, at that time, came under the Treasury Department, and among Morgenthau’s papers at the Franklin Delano Roosevelt Library in Hyde Park, New York, is the transcript of a telephone call he made on 13 May 1938 to Malvina Scheider, personal secretary to Mrs. Eleanor Roosevelt, the president’s wife, in response to a request she had made. (Amelia’s mentor, aviator A. Paul Mantz, on 26 April 1938 had contacted Mrs. Roosevelt asking her to intercede for him with the Coast Guard to obtain a copy of “ . . . the official report of the ITASCA,” [the Itasca Cruise Report, a 19-page document, dated 24 July 1937, written by Commander Thompson and on file at Coast Guard Headquarters]).

Only Morgenthau’s side of the conversation was included in the following transcript:

• “If we’re going to release this, it’s just going to smear the whole reputation of Amelia Earhart.” (In the document requested, Commander Thompson had done his utmost to shift the blame from himself to Earhart.)

• “And we have the report of all those wireless messages and everything else.” (This had to refer, in part, to the messages quoted in this article. They are the proverbial smoking gun. Nothing else reveals so clearly Commander Thompson’s poor judgment at the time. They had not been released when Morgenthau made his call.)

• “. . . if we give it to this one man we’ve got to make it public; we can’t let [just] one man see it.”

By far the most revealing remark is the following:

• “. . . if the president [FDR] ever heard that somebody questioned that the Navy hadn’t made the proper search . . . I mean I think he’d get terribly angry if somebody. . . .” [comment left unfinished]. (FDR, as is well known, was fond of the Navy, but when Morgenthau said “Navy,” he probably meant “Coast Guard,” the service he headed. Mantz, after all, had requested the Coast Guard Cutter Itasca’s documents, not those of the Navy.³)

Morgenthau apparently knew beyond any doubt that Commander Thompson’s actions were a factor in the loss of the fliers, and that he bungled the subsequent search. Morgenthau was one of the few men who might have had important information on the Earhart incident. On the chance that he might have told his son something about it, I wrote Robert Morgenthau asking if his father had ever discussed it with him; to date I have received no answer.

Commander Thompson suffered a coronary thrombosis and died at age 53 on 1 September 1939 in Ketchikan, Alaska, two years and two months after the fliers met their deaths.