Reflections: Marconi and Ionospheric Propagation,
and a Plea for Timely Experiments.

By Bart Lee, KV6LEE, xWPE2DLT, CHRS, AWA
San Francisco, California USA: KV6LEE@att.net

Marconi sent three dots, the letter "S", across the Atlantic by wireless in
December, 1901. Or did he?

As I reported to the Poldhu Amateur Radio Club in December, 2001, in 1999 I
had checked the sunspot numbers for December 1901 and discovered (for the first time as far as I know) that that number was exactly zero. As I wrote for 13 Antique Wireless Association Review (2000), this means the absorption frequency would be at its lowest and Marconi could well have benefitted from ionospheric propagation, or "skip" at his primary frequency of 833 khz, from Poldhu to St. John's, Newfoundland. There are several reports of current transatlantic AM broadcast reception on similar paths. Marconi told the IRE in the 1920s that he took a low pass filter out of the circuit, but no more in known about this. The kite movements detuned his receiver, so he altered it, presumably broadening the tuning. But Poldhu was still at 833+- khz, and recent work suggests that the Poldhu fan antenna would not have transmitted at much higher a frequency, and would have most strongl y radiated at about a 45 degree angle into the sky.

The fly in this ointment is that propagation models do not permit Marconi to get across, or the AM skip either for that matter. Marconi didn't trust experts, and the models may not be as good as one would hope.

Was there D-layer absorption that would prevent Marconi's signals from reaching the F-layer and skipping across the pond? The models say so, but they are, as far as I can determine, based on today's understanding of the chemistry of the ionosphere. The amount of Nitrous Oxide (NO) at D-layer altitude determines the thickness of that layer. The models say it is too thick today for Marconi to have gotten through it that afternoon and thus through to Newfoundland in 1901. But that was then and this is now. In all probability, automotive and industrial processes for the last century of progress have filled the D-layer with NO that just wasn't there in 1901. It also has to be true that the present ambient RF noise level has to be orders of magnitude higher now than it was in 1901, which is presumably an element in the model's predictions. Yet even today AM signals at high Northern latitudes get across, NO or no NO.

Some say that Marconi did hear three clicks, repetitively, but it wasn't Poldhu. One theory is that it was electrical machinery. This is hard to test because that machinery is in all probability long gone and its EMF-emitting characteristics cannot now be determined. But Marconi would have heard such noises all the stronger in his English experiments, there being a lot more electrical macinery in England at the time than in Newfoundland and environs. He (and Kemp) presumably would have recognized it for what it was.

The second theory is more interesting, and testable as well. Some say Marconi heard the clicks for sure (as did Kemp) but they were lightning-generated. Newfoundland is due North of the Amazon basin, home to most of the world's lightning storms. (This theory somewhat inconsistently presumes a North to South ionospheric propagation for lightning static longer than the East to West of Poldhu to St. John's, so the zero sunspot number is still in play. The data from the National Bureau of Standards in the 1930s, however, suggests Northern lattitude propagation is differentially improved when the ionosphere quiets). Be this as it may, if lightning static sounded like three repetitive clicks in 1901 it should sound the same a century later, so it can be recorded now and analyzed.

I have proposed that the Poldhu Amateur Radio Club put a 160 meter beacon of maximum legal power on the air in November 2006 through February 2007, because that will likely be the next sunspot minimum. GB2GM can transmit the standard marker of the Morse letter V, which has the advantage of being ST, and this itself will include Marconi's S of three dots. I propose that it transmit in CW but also as SSB audio hash, 200 hz to 2,500 hz. This will simulate a spark signal, albeit with a very narrow "decrement" as they used to say (or "bandwidth" in today's understanding).

The first test is simply: can this be heard in Newfoundland. Next, what is the minimum receiver and antenna required; will this require a beverage antenna and modern digital signal processing, or will a sharply tuned crystal set suffice with a tuned vertical? Marconi's mercury oxide "Italian Navy" (or more likely Chandra Bose) detector was a sensitive semi-conductor playing into a high impedance earphone to sensitive ears. Marconi in Newfoundland did not employ an insensitive coherer and inker combination, but even that worked for 700 daylight miles aboard ship in early 1902, and for almost 2,000 miles at night. To my mind this fully documented success clinches the ionospheric propagation hypothesis for December 1901 as well. The ionospheric pond above was likely a whole lot quieter in 1901-'02 than the Atlantic "pond" on which he sailed that winter. The second test relates to the lightning theory. A beverage antenna in Newfoundland pointing to the center of the Amazon lightning zone ought to hearthat noise on 800+- khz (and 160 meters). Recordings of that noise, when analysed, should show nothing but random patterns. But sensitive ears can listen for three-click patterns, as were Marconi and Kemp in 1901. It's a subjective test to be sure, but worthwhile. It must, however, also be remembered that Marconi and Kemp knew what their Poldu transmitter sounded like. Each spark transmitter has a distinctive audio pattern determined by the frequency of the spark generation. For example, a later 500 hz quenched spark gap sounds a 1000 hz audio note and 50 or 60 hz mains powered spark gaps sound as 100 or 120 hz "thumps", with rotary spark gaps somewhere in the middle, the principle of which would have been known to Marconi.

These two experiments can provide a empirical basis for acceptance of Marconi's 1901 claim, and lay to rest an alternate theory. Continuing cooperation between Canadian and British amateur radio operators can thus play a part in verification of one of the most interesting events in the history of our technology.

73 de Bart Lee, KV6LEE (correspondence is invited) ##