Home | Back to Wavescan Index

"Wavescan" is a weekly program for long distance radio hobbyists produced by Dr. Adrian M. Peterson, Coordinator of International Relations for Adventist World Radio. AWR carries the program over many of its stations (including shortwave). Adrian Peterson is a highly regarded DXer and radio historian, and often includes features on radio history in his program. We are reproducing those features below, with Dr. Peterson's permission and assistance.

Wavescan, September 6, 2009

The World's Longest Eclipse: The Fascinating Story of Radio Broadcasting on an Uninhabited Island in the Pacific

Just a few weeks back, on Wednesday July 22 to be exact, there was a spectacular eclipse in which the sun was darkened by the intervention of the moon. This eclipse was observed best in the areas of central China, to which many tourists flocked for the occasion. The duration of this eclipse is described as the longest for this century.

There was another equally spectacular eclipse that took place over the Pacific last century, the duration of which was described as the longest in more than a thousand years. The date for that event was June 8, in the year 1937, and the best places for the best observations were listed as Enderbury Island and Canton Island, though astronomers described Enderbury as the very best. These two isolated islands in the South Pacific saw the totality of the eclipse for a little over 90 seconds, which was long enough for the astronomers to take several spectacular photographs.

In January 1937, plans were laid for a joint expedition on the part of the National Geographic Society and the US Navy. A small navy vessel based in Honolulu with a draft of just one thousand tons, the "Avocet," was made available. This ship had previously served as a minesweeper and as an aircraft tender and it was modified to house a dozen expedition personnel and an RCA shortwave transmitter that weighed five tons.

On May 6, 1937, just before the "Avocet" left Honolulu, a special half hour farewell program was broadcast over the specially made shortwave transmitter WMEF. This program was relayed to the RCA station at Bolinas in California for nationwide coverage on mediumwave across the United States. The "Avocet" was then bound for two small islands in the exotic South Pacific to study an unusually spectacular eclipse of the sun.

The "Avocet" arrived off the shore of small Enderbury Island one week later at 8:30 am local time on May 13. There was no satisfactory anchorage location at Enderbury, so she moved on to Canton Island arriving there later in the same day.

Enderbury Island is a small coral atoll just three miles long and one mile wide, with a small shallow lagoon. The island is uninhabited and it has an elevation of just a few feet above sea level. This island was discovered by Captain James Coffin on the British whaling ship "Transit" in 1823 and it was named in honor of a London business man by the name of Samuel Enderby. The name of the island as inscribed on maps, Enderbury, is actually a mis-spelling.

In the year 1860, American business interests began a mining venture on Enderbury, taking out ship loads of guano. Due to rival claims for sovereignty over the island, the United States and Great Britain signed a treaty in 1939 for joint administration. In 1979, lonely and uninhabited Enderbury Island was absorbed into the newly independent nation of Kiribati.

After unloading men and supplies on Canton Island, the "Avocet" returned to Enderbury and off loaded men and supplies there also, in preparation for the coming eclipse. A small 20 watt Apex high frequency relay transmitter, manufactured by RCA and licensed with the American callsign W10XEP, was installed in a tent.

During the nearly two thousand mile voyage across the Pacific, a series of radio broadcasts was made over the transmitter WMEF for nationwide relay by NBC in the United States. Other special broadcasts were also made in the vicinity of the islands, including one program under the title, "A Desert Island."

Interestingly, the advance schedules for mediumwave WEAF in New York show the insertion into their programming of eclipse broadcasts from the South Pacific. Several of these special programs from Enderbury and Canton were heard direct from WMEF aboard the "Avocet" by American radio monitors. On one occasion station WMEF was heard calling station W2XAF in Schenectady, New York; and on another occasion station W3XZ. On both occasions, preparations for relay programming were underway.

On Eclipse Day, Tuesday June 8, 1937, three broadcasts were made; one early in the morning, another during the eclipse which began around 8:30 am local time, and another in the evening as a summary of the day's events. Each of these broadcasts contained eye witness accounts from both Canton Island as well as from the fifty mile distant Enderbury Island.

The atmospheric conditions at both Enderbury and Canton were described as almost perfect and the photographs taken that day are still studied three quarters of a century later. The insert broadcasts from Enderbury were transmitted by the low power Apex transmitter W10XEP and spliced live into the main on-air programming from WMEF on board the "Avocet" which was off shore near Canton.

With the eclipse events over, equipment and personnel on Canton were loaded onto the "Avocet" next day, she voyaged over to Enderbury and picked up the men and equipment there, and then sailed for Honolulu later on that same day.

So that is the interesting story of broadcasting from a lonely and uninhabited island in the South Pacific. The 20 watt high frequency shortwave transmitter W10XEP carried radio programming from lonely and uninhabited Enderbury Island. This programming was picked up on the "Avocet," and broadcast on shortwave to the RCA stations in Hawaii and California for onward relay on mediumwave throughout the United States. But that is not the end of the story.

The five ton shortwave transmitter WMEF was placed in storage in the United States for a period of five years. Then, in 1942, this equipment was renovated and taken to North Africa, and in August of the following year it was set up and placed on the air at Syracuse on the island of Sicily. A month or two later, the transmitter was shipped to Bari in Italy and then taken by road to Naples, where again it was placed on the air. Shortly afterwards this same unit was transported to Rome where again it became airborne.

This historic one kilowatt RCA shortwave transmitter that initially saw service in the Pacific on board a ship for the broadcast of a significant eclipse of the sun in the year 1937, finished up in Rome as a temporary relay station for the Voice of America. The engineers who manned this station nicknamed it "Relic," due to its size and age.

Early Reception Report QSL Cards-III

In our onward series of features on the interesting developmental history of the QSL card, we come now to this information regarding the earliest QSL cards in the Indianapolis Collection. You will remember from previous information presented in Wavescan that QSL cards back at the beginning of the 1920s were what we would call today, Reception Report Cards.

The earliest QSL Card that we hold was printed in the year 1920. It is a generic card printed on a post office postal card by the Grebe radio company, and it could be filled in by any listener and sent to any station. This particular card is a true Reception Report Card that was actually filled in two years later by an amateur radio operator at station 4HJ in Jacksonville, Florida and posted to station 8IQ in Akron, Ohio. The time is listed not as UTC, or the older GMT, but as NAA time, that is, the time according to the signals from the longwave navy wireless station NAA at Arlington in Virginia.

It would seem a little strange to us today, but most of the Reception Report QSL Cards back in the years 1920 and 1921 were filled in by amateur radio operators and sent to another amateur radio station, though these cards did not actually acknowledge a QSO two way contact with each other. They were simply reception reports from one amateur station to another.

Our earliest dated Reception Report QSL Card reports reception on January 30, 1921, and it was postmarked the following day. The text is also printed on an American post office postal card, and the value of the stamp is just one cent, with the head of President Thomas Jefferson, the third American president, printed in green.

The QSL reception text on this historic radio card is printed in black ink, and the callsign of the station is printed in very large letters in green ink. This card was sent by station 8TY in Jamestown, New York to station 3QW in Pottstown, Pennsylvania. The wireless signal in Morse Code was transmitted from station 3QW on 300 metres, which equals 1000 kHz, and this channel is right in the middle of what is today the standard mediumwave band.

Interestingly, very few of these early reception report cards list the actual frequency or wavelength of the transmitting station. Just one other card lists a transmitting channel, and this was station 2KU in New York on 190 metres, corresponding to 1580 kHz. This channel was a little above the regular mediumwave transmitting channels in 1922, but today it is at the top end of the mediumwave band.

However, even though the operating channel of the transmitter is usually not given on these early QSL cards, yet the power is quite frequently listed. For example, station 2BBB in Ridgewood, New Jersey was transmitting at 5 watts, 6ARB in Berkley, California was transmitting at 50 watts, and 8AGR in Erie, Pennsylvania was transmitting also at 50 watts. It would be evident that these power ratings during this era would express the input to the transmitter, not the output to the antenna system.

One amateur station, located at a school in Concord, New Hampshire, was operating a spark transmitter rated at half a kilowatt, fed into a T-type antenna system, with a fan counterpoise grounding system underneath. Neither the callsign of this station nor its operating channel are given on this QSL card, though it must have radiated a hefty signal that would have caused a lot of interference to local listeners.

One of these early cards is quite ornate, considering the era of its usage. The pre-printed reception text is in red ink on a postal card, with a fancy black border around the text. This QSL card is postmarked in 1921, and it was sent from station 4EY in Elizabeth City, North Carolina to 8IQ in Akron, Ohio.

We are holding more than a dozen of these Reception Report QSL Cards postmarked in the year 1921, and another half a dozen postmarked in the following year, 1922. Each of these cards is from one amateur radio operator and addressed to another amateur radio operator, each card is a post office postal card, and each gives only the details of reception and not any details regarding transmission. Most of these cards are pre-printed in black ink, one is typewritten, and a few are simply handwritten in ink or even pencil.

These early Reception Report QSL cards form a very valuable and very interesting insight into early wireless and radio history. One of the next steps in the total picture of the development of QSL cards is the SWL Card, that is the Short Wave Listener Card, and we will present that story on the next occasion here in Wavescan.