After more (much more) analysis of photos and written source material, we have reason to believe the antenna set-up on NR16020 at the time of the aircraft's disappearance was different than we previously thought. As a reminder, there were three antennas on the Electra when it arrived in Lae.
• Dorsal vee wires from a mast on top of the fuselage to each vertical fin.
• A belly wire extending extending along the starboard underside of the fuselage.
• A loop antenna above the cockpit.
We have long believed the dorsal vee was used exclusively for transmitting and the belly antenna for receiving. The loop was used for DF (direction finding). Hence, the loss of the belly antenna on takeoff conveniently explained why Earhart was unable to hear Itasca's voice transmissions on 3105 but when trying to DF on Itasca she was able to hear the transmission via the loop.
Earhart's Western Electric 13C transmitter had a receiver relay terminal which allowed the transmitter and receiver to share the same antenna. Our belief that this feature was not used was based on a photo of the cabin in which the terminals on the transmitter are visible. The receiver relay terminal is clearly unused (see below). However, that photo shows the cabin as it was when the airplane was equipped with the Hooven Radio Compass system and before the installation of a "navigator's station" in early March 1937. At that time, the Hooven system was replaced with a Bendix system, a plywood table was built over the transmitter, and duplicates of some cockpit instruments (altimeter, airspeed indicator, and outside air temperature gauge) were installed for the navigator. A series of photos of Earhart sitting on the navigator's table, when correctly daed in sequence, shows the instruments were first installed on the cabin wall beside the window but were later moved down into the table.
In a photo of the cabin taken on March 12, 1937 (see below) there is a bundle of cables extending aft, secured to the fuselage fuel tank vent manifold. The cables probably serve the duplicated cockpit instruments mounted in the table. Included with the cables is a black wire that leaves the bundle and goes to a different destination. It is most logically a line from the receiver to the receiver relay terminal on the transmitter, thus allowing the dorsal vee to be used for both transmitting and receiving.
Such a change would be a smart move. The belly antenna was poorly situated for receiving unless the aircraft was close enough to the station to receive signals via "groundwave" (about 30 miles). At greater distances, signals arriving from above as "skywave" (bounced off the ionosphere) are blocked by the fuselage. The March 12 photo is the last photo we have of the cabin, so we don't know what it looked like for the second world flight attempt. We do know that, during the rebuild after the wreck in Hawaii and Manning's departure from the project, the trailing wire was not reinstalled and the Bendix receiver was removed, but there is no reason to think the ability to use the dorsal vee for both transmitting and receiving was not retained. That would explain why the port-side belly antenna was not replaced. The Bendix direction finder needed a "sense" antenna to resolve "180° ambiguity." The starboard-side belly antenna could serve that purpose.
Of course, if the dorsal vee was being used to receive, we need a new hypothesis for why Earhart was not able to hear Itasca on 3105 but was able to hear the signals sent on 7500. There is a simple explanation:
During the test flight on July 1st, Earhart was able to hear the station at Lae on Lae's frequency, 6522 kHz, so she must have had Band 4 (4000 to 10,000 kHz) selected on the WE20B. The was no need to change the setting when she left for Howland the next morning because she planned to receive hourly weather reports from Lae. Whether she ever heard them is debatable. If she later tuned the receiver to 3105 to listen for Itasca without switching to Band 3 (1500 to 4000 kHz) she wouldn't hear anything, but when she tuned to 7500 to use the loop, with the receiver still set on Band 4, she would hear the signal. The loss of the belly antenna on takeoff meant only that, without a “sense” antenna, the direction finder would not be able to resolve “180° ambiguity,” but the point is moot because the 7500 frequency was far above the DF’s 1500 kHz limit.
On the reef at Gardner, she must have realized her error because that evening she seems to have been able to hear Itasca on 3105 and send the dashes they requested that were heard by Achilles and New Zealand Star. She has to be on Band 3 to hear Itasca. To later hear KGMB on 1320 kHz she has to be on Band 2 (550 to 1500 kHz).
If our conclusions are correct, there are two take-aways from this new research.
• Earhart's failure to hear Itasca was due to operator error, not an accident on takeoff.
• If Artifact 2-2-V-1a is from the Electra, it is most likely internal receiver feed line. If the receiver under the copilot seat was connected to both the belly antenna and the transmitter in the cabin, there was much more internal receiver feed line aboard the airplane than we thought.