The 1937 Search

The Itasca Search

For the first few days, the Itasca and Capt. Thompson were in charge of search and rescue for Earhart, and had little help from other ships. The Itasca searched west and north, chasing down amateur and Navy messages suggesting Earhart was down on the water. The Swan and the steamer Moorsby helped search for a few hours north of Howland, but there was little Itasca could do productively given almost no solid information or leads on where Earhart went down.

Instead of providing detailed description of the movements and radio messages received by the Itasca and others, a synopsis of important events and areas of search follows. Many of the details are contained in the various radio transcripts, radio message traffic, ship logs, and reports from Capt. Thompson.

The Itasca cruised to the north, with a sharp lookout during the early morning hours, but was concerned about fuel reserves. During high speed search and rescue operations, fuel is expended faster than at most economical rate, and without additional fuel, the Itasca would have to abandon the survey within a couple of days or divert to Samoa for refueling. Meanwhile the US Navy became seriously involved in assisting in the search, as the Chief of Naval Operations authorized an aircraft carrier (the USS Lexington) to get underway as soon as possible with destroyer escorts (eventually USS Lamson, Drayton, and Hull). The Lexington was in San Pedro (Los Angeles), and had to transit to San Diego to pick up the aircraft. This was the fourth of July weekend, so getting airplanes and crew from liberty on short notice was difficult. At 2312 GMT on July 3rd, the USS Colorado departed Hawaii to help in the search. The Colorado carried three Vought OS2U “Corsair” two-seater scout aircraft, capable of being catapulted from the deck and recovered in the water from the battleship Colorado. The Colorado was on a NROTC training cruise when it was diverted from liberty call in Hawaii.

The Chronology

July 4
0120 The CG station in San Francisco asked the Itasca to stop transmitting on 3105 and 6210 kHz, so a better verification of signals from Earhart could be accomplished. Over the previous 10 hours or so, message traffic regarding what the Achilles heard and how it correlated to the Itasca broadcasts consumed a fair amount of energy and time for various parties. The concern was that amateurs and other radio stations might have been hearing signals really coming from the Itasca.
0335 Itasca headed west to begin a survey 225 miles west of Howland, based upon the LA radio amateur reports and another amateur report (later determined to be unreliable), which at this time represented the only credible clue as to where Earhart might be located. The intent of the Itasca was to transit during the night to 179o W to begin a north-south survey, progressing eastward towards Howland Island.
0556 The PBY plane arrived back safely in Pearl Harbor. The Colorado worked the flight as plane guard just south of Hawaii, because the plane was seriously short of fuel. This delayed the Colorado slightly on her way south.
0600 Pan American Airways (PAA) station in Wake heard signals on 6210 kHz, but was told to change to 3105 kHz, as there appeared to be more activity on that frequency. Nighttime was approaching, and radio signals were beginning to be heard, as the propagation characteristics dramatically improve at night.
0750 The CG station in San Francisco confirmed the LA amateur radio operators’ report of 179o, 1.6, being SW of Howland. Since there was no west, east, north, or south provided, this was not much of a clue for Itasca, but at least the Itasca’s superior commander was reinforcing Thompson’s decision to search that area.
0830 KGU (Honolulu station of 2.5 kW at 750 kHz) broadcast to Earhart and PAA Honolulu heard returning signals on 3105 kHz.

KGMB (Honolulu station of 1.0 kW at 1320 kHz) broadcast a series of messages to Earhart, and a large number of responses were heard by various parties. Both radio stations were easily heard throughout the central Pacific, provided the radio was tuned to those frequencies. The idea was that if Earhart could receive any radio signals at all, these two strongest signals should be able to reach her. There is not a lot of first hand documentation of the broadcasts, nor of the responses. Relying primarily on newspaper accounts, a map with annotations kept by the Navy at the 14th Naval District (now in the Archives collection in San Bruno), and various message traffic, here is what we think happened. KGU may well have prompted Earhart to pay attention to the KGMB broadcasts to start at 1000 GMT, as KGU, being more powerful, was required by law to shut down later that night. It would be better to rely upon KGMB for information, as it had a license to operate continuously. The KGMB broadcast asked Earhart to respond with 2 dashes if on water, and 3 if on land. Army station Ft. Shafter in Hawaii indicate three distinct dashes between 0850 and 0900 GMT (i.e. Earhart was on land). The map has these annotations:

Night of 3rd –
KGMB Requested a Broadcast –
0630 GCT 4th (8 PM Local Honolulu Time to 2:15AM)
Asked for 8 dashes if on water – got eight in response.
Asked for 4 dashes if north of Howland and 6 is [sic] south [covered by ink stain] received
The Honolulu Advertiser indicates a response of four dashes, possibly indicating a position north of Howland, but what time nor in response to what specific broadcast is not mentioned. Other correspondence indicates four dashes were received as well. While not crystal clear, at least someone was responding. More on this will be discussed later, when the broadcasts were repeated the following night. Meanwhile, from 1000 to 1137 GMT, the Itasca, PAA Honolulu and PAA Wake all heard carriers and signals on 3105 kHz and 6210 kHz, but none of the signals were readable.
1200 The CG station in San Francisco reported hearing 6210 kHz from the west. This is not surprising, as all of the radio operators generally knew where Earhart went down relative to their own locations, and are most likely to search in the direction of that quadrant.
1215 PAA Wake heard a wobbly voice on 3105 kHz, described as being a man's voice.
1512 PAA Midway reported that PAA Honolulu obtained a bearing of 175° on these signals. At 1523 GMT, PAA Honolulu reported a bearing of 213° ± (another document states 210°), and reported that to the CG station in Honolulu. At 1540 GMT, PAA Wake obtained a bearing of 75°. Finally, some real progress! Unfortunately, radio bearing direction finding, to be accurate, must have at least two stations making measurements on the same signal at the same time to provide a reliable location. Obviously, the Wake measurement cannot be from Earhart. More on accuracy of radio bearings later on.
1700 The Lexington left San Pedro for Coronado Roads (off San Diego).
1742 The Itasca started its box search west of Howland.
2025 The CG Hawaii radio station began to set up the schedule for the KGMB broadcast later that night, and requested the Itasca not to broadcast on 3105 or 6210 kHz to better determine the origin of the signals. Apparently, the radio bearings obtained so far possibly indicated that it was the Itasca everyone was homing in on.
2100 The Lexington group with the destroyers left San Diego, bound for Hawaii for refueling prior to moving south to Howland to conduct the aerial search.
2140 Itasca asked the CG Hawaii station permission to broadcast on 3105, 6210, and 7500 kHz between 0730 and 0830 GMT July 5, as it was in the best position for Earhart to possibly receive radio signals.
July 5
0225 The CG relents and allows Itasca’s to transmit on these frequencies. Meanwhile, the Colorado began to send messages to the Itasca regarding refueling details at sea. It appears that the Itasca was not equipped with the proper gear for at-sea fueling, and some adjustments needed to be made to be able to do so.
0330 Itasca broadcast on 3105 kHz, and PAA Honolulu obtained a bearing on the Itasca of 210°. Based upon the position of the Itasca, the true angle at Honolulu should be 229.2°, a discrepancy of nearly 20°.

KGMB conducted its second night of broadcasts to Earhart. The instructions were for Earhart to turn on her transmitter for one minute for tuning purposes, and then to send four long dashes if she heard KGMB, then to wait for an acknowledgement by KGMB fifteen minutes later. PAA Honolulu (Makapuu Point), heard four dashes, and asked KGMB to repeat the message. Evidently, KGMB did so, as PAA reported hearing only two dashes as if power failed on the transmitter. At 0700, 0715, and 0730 GMT, Navy Radio Tutuila, American Samoa, heard four, four, and eight dashes respectively. The Honolulu Advertiser reports:

Miss Earhart was asked to use the carrier break four times if she was north of Howland, six, if south. She was asked to break twice to indicate if she were on land and three times if she were on water.
What we don’t know is if Earhart was to add these numbers together (six to nine total dashes possible) or at fifteen minute intervals. If the former, Radio Station Tutuila indicates eight dashes, or south, and on land. Meanwhile, PAA Honolulu obtained bearings of 213 and 215° on the 0630 GMT responses, depending upon the source. Itasca did not hear a single dash in response to KGMB. Midway picked up a bearing of 201°, but on a frequency halfway between Itasca and Colorado, slightly above 3105 kHz. Radio Tutuila heard nothing at 0630.
0700 We suspect that KGMB asked Earhart to respond with four dashes if north of Howland, six if south, and Tutuila heard four dashes. CG Radio Hawaii reported that PAA Honolulu reported a bearing of 200°T[rue], but we don't know at what time that bearing was obtained.
0730 We suspect KGMB asked for two dashes if on land, and three if on water. Tutuila heard eight dashes, four of which were strong, and possibly voice on 3105 kHz.
For the next few hours, a large number of reported signals were heard by Itasca, Baker and Howland Islands, and the PAA radio stations. Some of the most interesting:
0907 Itasca hears a generator start and stop on 3105 kHz.
0910 Both Baker and Howland Island report hearing a voice on 3105 kHz.
0916 Howland Island reports hearing Earhart.
1103 PAA Midway obtains a bearing of 175°, considered coming from South America or Russia, so it was disregarded.
1105 After being chastised by Itasca, Howland restarts its Earhart radio watch and obtains a bearing of NNW/SSE on magnetic pocket compass. States that it cannot determine which direction due to “night effect.” Interestingly, despite access to a theodolite for weather balloon measurements, no azimuthal calibration was made on the Howland direction finder. A pocket compass was the only available device. Night effect is actually a swinging of the directional bearing due to night propagation effects, and what Cipriani really means is that there is no way of determining a unilateral direction without a sensing radio to add to the direction finder.
1130 Radio Station Wailupe began to hear a ragged keyed message on 3105 kHz, which is reported to state: “281 North Howland Call KHAQQ beyond North don’t hold with us much longer above water shut off,” received by three Navy operators. Of all the post-loss radio messages, this particular message has the best chance of being from Earhart, particularly if it is proved that she landed at Gardner. It turns out that Gardner is 281 nm south of the Equator, and that distance can be easily determined by any experienced navigator. Since the message was ragged, perhaps the message was really “equator is 281 north of us.”
1223 PAA Wake reported a bearing of 144°, strength 5. This is likely the Itasca, who was broadcasting at the time and was bearing 141.56° true at Wake. Simultaneously, Midway obtained a bearing of 201°, well away from the possible location of Earhart, but possibly from the Itasca.
1250 Itasca diverted course for a position 281 nm north of Howland, believing the latest radio report. Itasca directed Swan to also move to that location to help conduct a search. A few hours later, the SS Moorsby, en route to New Zealand, offered assistance to the Itasca, and was directed to also go to a line 281 nm north of Howland.
1548 Howland heard Japanese music on 3105 kHz, which it continued to do until the 17th of July. So far, we have not been able to identify what station was broadcasting on 3105 kHz.
1625 Itasca heard Russian stations on 3105 kHz. We know of two low wattage stations from Siberia that might fit this description.
2219 Verification from Lae was received that Earhart left with 1100 gallons of gasoline, with an estimated 24 to 30 hour duration in the air.
2325 The CG station in San Francisco reported that radio technicians familiar with the plane said that the radio is unable to function on water, so therefore if the signals are to be believed, she must be on land. They also state that they believe Earhart flew the line of position to the Phoenix Islands. This radio message is the first to indicate that searching at sea might well be useless if one were to believe the various radio signals.
Late Finally, sometime during July 5th, a State Department memorandum from the Division of Far Eastern Affairs documents a phone conversation with Tsuneo Hayama, the Second Secretary in the Japanese Embassy, describing a telegram received from Tokyo offering assistance of the Japanese government, warships, and radio stations in the Marshalls. The US responded with current US ship assets and for Japanese authorities to coordinate with the Commandant 14th Naval District. The only Japanese ship that we are aware of that actually searched for Earhart was the Koshu, a oceanographic survey vessel, which departed its survey at 2200 GMT for the Marshall Islands, arriving sometime about the 9th.
July 6
0535 The CG Station in San Francisco relayed an opinion from PAA authorities that the 281 message may well be referring to the Phoenix Islands, as the radio bearings were beginning to intersect in that area.
0730 A report from the CG station in San Francisco documented that Lockheed Aircraft had definitively stated that the radio was inoperable if the plane was in the water.
0736 Itasca asked Howland to take a bearing on it at 0800 GMT. At 0832 GMT, Howland reported that the bearing was NNW or SSE, taken with a magnetic pocket compass. The Itasca at this time was 342.2° relative to Howland (magnetic), a good match to the 337 degrees reported. At 0930 GMT, Howland took another bearing, observing NW (315 degrees), versus the 345.9 magnetic orientation to Itasca, a far worse bearing error of nearly 31 degrees. At 0947 GMT, PAA Honolulu reported a bearing on the Itasca of 196.5°, but the bearing to the Itasca is really 232.38°, an error of nearly 36o! Based upon these errors obtained over the past couple of days, direction bearings should be considered indicators of direction, and not treated as literal bearings with any accuracy.
0842 Itasca and Swan saw heat lightning, which they interpret to be flares, and broadcast in the clear to Earhart: “we see your flares and are rushing to your location.” Unfortunately, a wide variety of people were listening in and a mad rush of press reporters sent out radio messages to the Itasca asking for information, pictures, etc. A few minutes later, the Itasca recognized that these were not flares, and with the latest information that Earhart is probably on land, abandoned the 281 North Howland line search and headed south at 1554 GMT to rendezvous with the Colorado for refueling at 1730 GMT on July 7th.
2025 Commander in Chief, US Navy ordered COM14 to take charge of the search, as the battleship Colorado was about to enter the search area, and the Lexington group would be arriving shortly.
2133 CG Headquarters approved the Itasca to be placed under control of the US Navy, and at 0053 GMT July 7th, the Itasca reported to COM14 for duty, ending its independent search for Earhart.
Research Papers Earhart Project Home Page

TIGHAR Projects TIGHAR Publications Contract Services Field Schools
TIGHAR Forum Contact TIGHAR TIGHAR Home Site Map