In the conclusion of Almon Gray’s “Amelia Earhart and Radio,” the former Pan American Airways radio flight officer examines further technical and other aspects of Amelia Earhart’s final flight, including the origin and effectiveness of the radio direction finder on Howland Island, some of the possible post-flight radio transmissions that may have originated from the Earhart Electra, and Fred Noonan’s alleged drinking problem as it may have affected the flight. As always, the real mystery is what transpired aboard the Electra in the hours before and after her last radio transmission, and the biggest question remains unanswered: Was Amelia actually attempting to reach Howland Island? If she was, then Gray’s conclusions remain highly relevant today.
THE HOWLAND ISLAND RADIO DIRECTION FINDER
Obviously Earhart had a misconception of the radio direction finder installed on Howland Island. She apparently envisaged it as being a PAA type Adcock high frequency system, or its functional equivalent, which would take bearings on her 3105 kHz signals and send them to her just as the PAA station at Mokapu Point had done during her flight from Oakland to Honolulu. Because of that she repeatedly asked Itasca to take bearings on 3105 kHz and transmitted signals upon which bearings were expected to be taken. It appears that there may have been some justification for her having that concept.
When the decision was made to fly easterly around the world, and the long Lae-Howland leg was being studied, Earhart and Noonan suggested to the Coast Guard that a radio direction finder be set up on Howland (“PLANE SUGGESTS DIRECTION FINDER BE SET UP ON ISLAND, IF PRACTICABLE”). According to the research of Capt. Laurance F. Safford, USN, it was at about this time that Mr. Richard B. Black, the Department of the Interior representative, who was to go to Howland in Itasca, conceived the idea of “borrowing” a so-called high frequency radio direction finder from the Navy to use on Howland Island. Black advised G.P. Putnam, Earhart’s husband and business manager, of his plans and advised him when the gear had been obtained and put aboard Itasca. No doubt Putnam passed this information along to Earhart.
In a message sent June 27 to Commander, San Francisco Division, USCG, the C.O. Itasca [Cdr. Warner K. Thompson]reported on his readiness for supporting the upcoming flight. One item was “DIRECTION FINDER INSTALLED ON HOWLAND.” This fact was reported to Mr. Putnam, then in San Francisco, and he in turn passed the news to Earhart, who was then at Darwin, Australia. While the Itasca message did not specifically say “High Frequency Direction Finder,” there apparently had been sufficient other information, probably via telephone from Putnam, to cause Earhart to believe that it was such a device. She likely assumed that the DF had been installed at Howland in response to the suggestion made earlier by Noonan and herself , and fully expected it to be a functional equivalent of a PAA-Adcock system.
According to Capt. Safford, who was in an excellent position to know, the direction finder station on Howland Island actually consisted of an aircraft type radio receiver and an aircraft type rotatable loop antenna which had been “hay-wired” together into a temporary DF installation. It operated off storage batteries borrowed from Itasca. The receiver and loop had been “moon-light requisitioned” (obtained by informal means) by Mr. Black and Lt. Daniel Cooper of the Army Air Corps, from a Navy patrol plane at Fleet Air Base, Pearl Harbor.
The equipment appears to have been a military version, or perhaps a twin, of the Bendix receiver and loop in the Earhart plane. At any rate, with a loop antenna, it certainly was not a high frequency direction finder and the probability of taking meaningful bearings with it on 3105 kHz over any significant distance, was practically nil. The Howland DF operator [Radioman 2nd Class Frank Cipriani] had only two opportunities to try taking a bearing on the plane, and in each case the plane’s transmission was so short that a really good attempt could not be made. Had the transmissions been sufficiently long the operator no doubt would have found that he could get no “minimum” and hence no bearing.
On July 3 (GMT date) an operator at public service radio station VKT, Nauru, sent the following “wire note” (an informal communication between operators) to RCA radio station KPH at San Francisco, with the request that it be passed to Itasca:
VOICE HEARD FAIRLY STRONG SIGS STRENGTH TO S3 [at] 0843 and 0854 GMT 48.31 METERS (i.e. 6216 kHz) SPEECH NOT INTERPRETED OWNING BAD MODULATION OR SPEAKER SHOUTING INTO MICROPHONE BUT VOICE SIMILAR TO THAT EMITTED FROM PLANE IN FLIGHT LAST NIGHT WITH EXCEPTION NO HUM ON PLANE IN BACKGROUND.”
Note that these signals were heard about 12-and-a-half hours after Itasca last heard the plane.
There is nothing that directly and positively connects these signals with the Earhart plane, however there is indirect evidence that warrants serious consideration:
(a) The frequency (6210 kHz) was right for it being the plane. It was not a commonly used frequency in that area.
(b) The Nauru operator reported good signal strength and was able to judge the tone or timbre of the speaker’s voice yet was unable to understand what the speaker was saying. He suggested the possibility of modulation problems. The operator who had checked the plane at Lae and the DF operator at Howland who was trying to take a radio bearing on the plane, both had noted similar symptoms and suggested possible modulation problems.
(c) The probability of there being more than one transmitter in the area exhibiting the same symptoms of over-modulation on the same frequencies at essentially the same time is very small.
It is this writer’s opinion that the signals intercepted by Nauru were in fact from the Earhart plane no longer in flight.
PAN AMERICAN AIRWAYS
Shortly after the Earhart plane became overdue at Howland, the Coast Guard requested PAA to use its communication and radio direction finding facilities in the Pacific areas to assist in the search for the plane and survivors. Instructions were immediately issued for the stations at Mokapu Point, Midway and Wake to monitor the plane’s frequencies as much as limited personnel would permit and be prepared to take radio bearings on any signals heard which might reasonably be believed to be coming from the plane. A special radio circuit was set up to permit intercommunication among the three stations. Numerous weak signals were heard but nothing of interest until 0948 July 5, GMT time and date. The following is extracted from a report made by R.M. Hansen, the Radio Operator in Charge at the Wake Island station:
At 0948 a phone signal of good intensity and well modulated by a voice but wavering badly suddenly came on 3105 kc. While the carrier frequency of this signal did not appear to vary appreciably, its strength did vary in an unusually erratic manner and at 0950, the carrier strength fell off to QSA 2 (2 on a scale of 0 to 5) with the wavering more noticeable than ever. At 0952, it went off completely. At 1212 (GMT July 5) I opened the DF guard on 3105 kc. At 1223 a very unsteady voice modulated carrier was observed on 3105 kc approximately. This transmission lasted until 1236 GMT. I was able to get an approximate bearing of 144 degrees. In spite of the extreme eccentricity of this signal during the entire length of the transmission, the splits were definite and pretty fair.
After I obtained the observed bearing, I advised Midway to listen for the signal (couldn’t raise Honolulu). He apparently did not hear it. This signal started in at a carrier strength of QSA5 (5 on a scale of 0 to 5) and at 1236, when the transmission stopped, it had gradually petered out to QSA2 during the intervals when it was audible. The characteristics of this signal were identical with those of the signal heard the previous night (0948 GMT) except that at DF the complete periods of no signal occurred during shorter intervals. While no identification call letters were distinguished in either case, I was positive at that time that this was KHAQQ [Earhart’s plane]. At this date I am still of this opinion.
Midway heard a signal having the same characteristics, and almost certainly the same station, at 0638 GMT July 5. A quick bearing of 201 degrees True was obtained, however the signal was not audible long enough to take a really good bearing and the 201 degree figure was labeled “approximate.”
Honolulu (Mokapu Point) also heard the “peculiar signal” on 3105 kHz several times. From 1523 to 1530 GMT July 4 an attempt was made to take a bearing on it, however due to weakness and shifting of the signal, only a rough bearing could be obtained. It was logged as 213 degrees, but it was implied that it was a doubtful bearing. Sometime between 0630 and 1225 GMT another bearing was attempted on the “peculiar signal.” The log describes it thus: “SIGNALS SO WEAK THAT IT WAS IMPOSSIBLE TO OBTAIN EVEN A FAIR CHECK. AVERAGE SEEMS TO BE AROUND 215 DEGREES — VERY DOUBTFUL BEARING .” It is obvious that the bearings from Honolulu were greatly inferior to those taken from Wake and Midway and are useful mainly as indications that the unknown station continued to function.
Not much attention was paid to these interceptions at the time because no one was aware that Earhart’s radio signals had been abnormal. Had it been known that she was having over modulation problems more attention probably would have been given them because the “Wavering” in the carrier strength is consistent with a varying degree of over-modulation rapidly increasing and decreasing carrier power. The gradual drop of signal strength from QSA5 to QSA2 over a span of 13 minutes is consistent with the further discharge of an already partially discharged storage battery power supply. The peculiar signals on 3105 kHz heard by Wake, Midway and Honolulu may very well have come from the Earhart plane, and there is good reason to believe that the radio bearing taken on those signals by Wake was accurate within a degree or so. The one from Midway may have had a somewhat larger error.
(Editor’s note: A number of radio operators, including several in the continental United States, reported hearing signals that they believed originated from Earhart and Noonan, and some have already been presented on this blog. Please see “Earhart’s ‘post-loss’ messages’ Real or fantasy?” and “Experts weigh in on Earhart’s ‘post-loss’ messages”.)
FREDERICK J. NOONAN
There has been much speculation as to whether or not Fred Noonan could send and receive International Morse code. From personal observation the writer knows that as of late 1935 Noonan could send and receive plain language at slow speeds, around eight to 10 words per minute. Recent research by Noonan biographer Michael A. Lang has revealed that circa 1931 Noonan held a Second Class Commercial Radio Operator license issued by the Radio Division of the U.S. Department of Commerce. Second Class licenses of that vintage certify that the holder has been examined and passed the following elements:
(a) Knowledge of the general principles of electricity and of the theory of radio telegraphy and radiotelephony.
(b) Adjustment , operation and care of apparatus.
(c) Transmitting and sound reading at a speed of not less than sixteen words a minute Continental Morse in code groups and twenty words a minute in plain language.
(d) Use and care of storage battery or other auxiliary.
(e) Knowledge of international regulations and Acts of Congress to regulate radio communications.
Those writing about the Earhart disappearance have, in general, been very rough on Noonan because of his admitted problem with alcohol. In some cases much rougher than was justified by the facts. For example in one book it is related that the night before the departure from Lae for Howland, Noonan went on a binge and did not get to the airfield until just before the plane was due to take off, and even then was so intoxicated that he had to be helped aboard the plane. The implication being that he was largely responsible for the failure of the flight. The official report of Guinea Airways Ltd., at Lae, made in response to a request from the U.S. Government, paints quite a different picture. According to it the Lae wireless operator made attempts all throughout the day of June 30 to get time signals, requested by Earhart and Noonan, to permit Noonan to check his chronometer, but owing to local interference was unsuccessful that day. That indicates that Noonan spent most of June 30 at the radio station.
At about this point, Earhart decided to take off for Howland Island at 9:30 a.m. on July 1, subject to obtaining the time signal.
At 6:35 a.m. July 1st Earhart took the plane up on a 30-minute test hop after which the tanks were topped off and she was ready to go, except that a time signal had not yet been obtained. This day the difficulty was at the radio station which transmitted the signals. Extraordinary steps were taken to get a time signal but when one had not been obtained by 10:50 a.m. Earhart decided to postpone her departure until the next day, July 2. During the rest of the day constant watch was kept for the reception of time signals and finally at around 10:20 p.m. an excellent signal was received by Noonan which showed his chronometer to be three seconds slow. Noonan obviously had spent most of that day at the radio station.
On July 2 at 8:00 a.m. another time signal was received, this one from Saigon, and the chronometer checked the same as the previous night. Both Noonan and Earhart expressed their complete satisfaction and decided to leave at 10:00 a.m., which they did.
Only Noonan would have checked the chronometer, so the report seems to indicate clearly that Noonan was sober and in good shape at 8:00 a.m. and probably was that way when the plane took off.
From the standpoint of radio, Earhart’s decision to rely completely on radiotelephony, and her removal of the trailing antenna, showed poor judgment and introduced unnecessary and unjustifiable risks. However it cannot be denied that she got as far as Lae without trouble with what she had. It was her mistake in designating 7500 kHz as the homing frequency for Itasca that got her into deep trouble. Even that difficulty probably could have been overcome had she been able to communicate with Itasca and agree on a suitable homing frequency. Fate intervened, however, and something occurred in her receiving system which made it impossible for Earhart to hear any signals with her gear set up in the configuration she was accustomed to use for communications.
She did not understand the technical aspects of radio well enough to diagnose her problem and was not sufficiently familiar with the radio gear to know all the options available to her. She had been taught to shift the receiver to the loop antenna when she wanted to take a bearing, but probably no one had ever explained to her how the loop also could be used in carrying on communications. Had she been aware of that option and listened on the loop for Itasca‘s voice signals on 3105 kHz, no doubt she would have heard the ship and been able to establish two-way communications.
The probability is very high that the failure of the receiving system to receive signals when using the fixed antenna was due either to a defective feed line between the receiver unit and the “send/receive” relay in the transmitter, or a defect in that relay itself. The odds are about 95 to 5 that the relay was at fault. It is considered therefore that a failure of that relay was the one single thing most responsible for the failure of the Earhart flight.
If it is assumed that the “peculiar signals” intercepted by Nauru and the PAA stations at Wake and Midway were in fact from the Earhart plane then the following may be deduced from the radio signals:
(a) The landing was fairly successful. The plane did not nose over or break up, otherwise the radio could not have been used.
(b) The landing was not in the open sea. Had it been, enough salt water would have seeped in to enter the wiring and disable the radio transmitting gear in a relatively short time.
(c) Earhart survived the landing. She was heard by the Nauru operator long after the plane would have run out of gas.
(d) Noonan survived. A man’s voice was distinctly heard on the “peculiar signal” by Midway. It was unintelligible.
(e) Either Earhart or Noonan, or both, were alive and with the plane at least until 0948 July 5, 1937 GCT time and date. The “peculiar signals” were last heard then.
(f) The “peculiar signals” probably were coming from the eastern or southeastern part of the Marshall Islands. (End of Almon Gray’s “Amelia Earhart and Radio.”)
Bill Prymak’s note: Capt. Gray, USNR (Ret.) received his Commercial Radio Operator License in 1930, and went with Pan American in 1935, when they started the trans-Pacific service. He became Flight Radio Officer on China Clipper type aircraft, and later was promoted to Assistant Superintendent of Communication in 1937.
The AMELIA EARHART SOCIETY finds the above radio analysis of the last flight to be one of the finest pieces of work ever presented on this subject.
Editor’s note: We should remember that in considering this analysis of Earhart’s final flight, Almon Gray took the position that the fliers were actually trying to reach Howland Island, and that all their actions were directed toward that goal. If Amelia and Noonan were not trying to reach Howland, but were engaged in some sort of covert operation, which certainly cannot be ruled based on our limited knowledge of what transpired during those final hours, then many of Gray’s findings become largely irrelevant.