A working link took more than a spark
Marconi’s 1909 lecture describes antennas, grounding, receivers and transmission trials over growing distances. Braun’s lecture develops coupled oscillating circuits and directional arrangements. Tuning and coupling helped separate the generation of energy, its radiation and its reception, instead of treating the whole apparatus as one spark.
The 1909 prize recognized contributions to wireless telegraphy. Receiving a signal far away shows that a link works under those conditions. It cannot tell on its own whether the waves travelled along the ground, were refracted by the atmosphere or took several paths.
Interference estimated the length of an extra path
Appleton later probed the upper atmosphere with controlled radio measurements. When a direct wave and a returned wave both reached the receiver, changing the frequency changed their relative phase. If the path geometry stayed stable enough, the successive changes in interference gave an estimate of the path difference.
The relation below links the path difference to the frequency step between repeated interference states. Real atmospheric dispersion changes both phase and group delay, so reading the result as the height of a literal mirror takes a model. The 1947 prize is separate from the earlier award for communication.
Appleton’s lecture dates the first frequency-change experiment with M. A. F. Barnett to 11 December 1924, using the Bournemouth transmitter and an Oxford receiver; their model gave an estimated reflecting height of about 90 kilometres. The lecture credits Breit and Tuve with inventing the pulse method in 1925, which Appleton’s group later adopted and developed.
Sources: [3]
Pulse delay gives a virtual height, not the true height
In vertical-incidence sounding, an ionosonde sends pulses upward across a range of frequencies and records their round-trip delay after instrumental delays have been removed. Half the vacuum light speed times this delay defines the virtual height. NOAA explains why it differs from the true reflection height: group propagation in the plasma is slower than in free space. For an oblique path the same delay first gives an apparent slant range, not a vertical altitude.
An ionogram therefore plots delay against frequency and needs interpretation; it is not a photograph of a reflecting shell. Near a critical frequency, long delays and mode structure complicate any simple geometric picture.
The ionosphere is a dispersive, magnetized plasma
Electron density, collision rate and Earth’s magnetic field all affect radio propagation. Separate magneto-ionic modes, absorption and changing density profiles can change which echoes are detected. Any inference of a critical frequency or a height must state the mode and the assumptions it uses.
A gradient-index ray model gives useful intuition for bending. It does not supply plasma dispersion, polarization modes or an inversion for electron density. Day-to-night changes in reception cannot single out one cause either, unless frequency and path are measured under control.
A mirage photograph measured the air above sea ice
Radio waves are not the only waves the air bends back towards the ground. On 16 May 1979, at Tuktoyaktuk in northern Canada, Waldemar Lehn photographed Whitefish Summit, a hill 18.7 m high, across 20 km of sea ice. Cold air lay on the ice with warmer air above it. Light bends towards colder, denser air, so the light from the hill curved down towards the ice on its way to the camera. Next to a normal view taken four days later at the same angular scale, the hill looked strongly distorted.
Lehn described the image with one curve. Each ray that reaches the eye arrives at some angle above the horizontal and has met the hill at some height; the curve pairs the two. Where it rises, that part of the hill appears upright; where it falls, upside down. He measured matching points on the two photographs to get the observed curve. Then he guessed how the temperature changed with height, traced rays through that air, compared the computed curve with the observed one and corrected the guess. After eight rounds the two curves agreed closely.
As with the ionosonde’s virtual height, the photograph does not show the layers directly; a model of the light’s path turns it into a temperature profile. Lehn’s profile is an average over the 20 km path and assumes layers that do not change along it. It is pinned to the air temperature measured at eye level, minus 2 °C at 2.5 m above the ice. Such a profile can guide light along the ice as if in a channel, and Lehn notes that this ducting was present at the time.
Sources: [5]
A working link and a measured atmosphere need different evidence
Read Marconi’s apparatus alongside Braun’s circuit argument, then turn to Appleton’s interference and pulse measurements. For each experiment, ask what new quantity it measured. The linked ray model explains only bending; a dedicated ionosphere instrument needs its own delay and frequency data.