In May 1919 the US Navy sent three Curtiss flying boats from New York toward Europe. Along the ocean route it stationed a line of warships to mark the way, report the aircraft as they passed, and pick up anyone who came down. Only NC-4 reached Lisbon, and it crossed the Atlantic in stages. It was not a nonstop flight. This piece covers how the station ships were used, what radio and direction finding the boats carried, why two of the three never reached the Azores by air, and what the operation still teaches anyone who follows aircraft from the ground.

Four boats, three starters: the NC program and the crew of NC-4

The NC boats came out of the First World War anti-submarine effort. U-boats made it dangerous to ship large patrol aircraft across the Atlantic as cargo, so the Navy and the Curtiss company designed a flying boat big enough to cross under its own power and then hunt submarines on the European side. NC stands for Navy-Curtiss. The armistice came before any of the boats saw service, and the Navy decided to make the crossing anyway as a demonstration of long-range naval aviation. Four boats were built, and in their final configuration each had four Liberty engines. Before departure NC-2 was taken out of the program and its parts were used to keep the others flying, so three boats started the flight.

Commander John H. Towers commanded the division and flew in NC-3. Lieutenant Commander Patrick N. L. Bellinger commanded NC-1. Lieutenant Commander Albert C. Read commanded NC-4 and also served as its navigator, which put the navigation and the command decisions in the same person. NC-4 carried six men. Lieutenant Walter Hinton and Lieutenant Elmer Stone, a Coast Guard aviator assigned to the Navy for the flight, were the pilots. Lieutenant James Breese was engineering officer, Chief Machinist’s Mate Eugene Rhoads was engineer, and Ensign Herbert C. Rodd was the radio officer. Rodd’s place in the crew matters to this article, because the radio operator was the link between the boat and the station ships.

The boats left Naval Air Station Rockaway in New York on May 8, 1919. NC-4 had engine trouble on the first day, came down off Cape Cod, and reached the naval air station at Chatham by taxiing on the water. For several days it was the boat that looked least likely to finish. After repairs it flew on to Halifax and then to Trepassey Bay in Newfoundland, where it joined NC-1 and NC-3 for the ocean leg. As far as I can establish, the Navy did not enter the flight for the London Daily Mail prize. That prize required a nonstop crossing, and the Navy plan had never been built around one.

Station ships across an ocean: how the Navy marked the route

For the longest leg, from Trepassey to the Azores, the Navy placed destroyers in a line along the planned track. Histories of the flight, including those published by the Naval History and Heritage Command, commonly describe them as spaced about 50 nautical miles apart. The usual count for that leg is in the low twenties, and accounts differ by a ship or two depending on whether they include ships at the ends of the line. More station ships covered the legs from the Azores to Lisbon and from Lisbon to Plymouth. Larger vessels supported the operation from various positions, with weather reporting among their duties, and the aircraft tender Aroostook served as base ship at Trepassey. Published totals for the whole operation vary from one secondary source to another, so I do not give a single figure here.

Each station ship had several jobs. By day it made smoke so the crews could see it from a distance. At night it pointed a searchlight straight up and fired star shells as the boats approached, so the light reached altitude and could be seen over a wider area than a ship’s running lights. Each ship knew its own position from its own navigation, and that is what made it useful. A crew that passed directly over a numbered station had a position fix. The navigator could then correct his dead reckoning and set a heading for the next ship, so a small heading or drift error was caught every 50 miles or so instead of growing across 1,200 miles of open water.

The ships also served as a tracking network. Accounts of the flight describe station ships reporting by radio when a boat passed overhead or was heard, which let the Navy follow the flight’s progress from the ground and the newspapers report it almost in real time. The same spacing put a rescue platform within roughly 25 miles of any point on the line. The planners understood that a flying boat in trouble would most likely land on the sea and wait. The weakness of the scheme was that the destroyer line helped only a crew that could stay on it or find it again, and the visual part of the system depended on the weather.

Ensign Rodd’s station: what radio the boats carried and what it could do

Each boat carried a radio operator and equipment for both telegraphy and telephony. In 1919 that was ambitious for an aircraft. The radio could reach the station ships, the other boats and, at times, shore stations, and the crews used it for position reports, weather and coordination. Four Liberty engines in an open hull made a great deal of noise, and Morse code was far more reliable than voice in those conditions. Accounts of the flight describe radiotelegraphy carrying most of the useful traffic. Rodd’s operation of NC-4’s set is often mentioned favorably in histories of the flight, and the record suggests NC-4 kept better radio contact on the ocean legs than the other two boats did.

A long-range set needs a long antenna, and the usual airborne solution of the period was a weighted wire reeled out behind the aircraft in flight. That design has a consequence that comes up again in the Azores leg. An antenna that works by trailing behind a flying aircraft does not work the same way once the aircraft is sitting on the water, so a boat that landed at sea lost much of its transmitting range just when it most needed to be heard. Accounts of NC-3’s two days on the water report that the crew could hear radio traffic about the search but could not make themselves heard in return. I treat that as the commonly reported version and not as a technical finding, because I have not seen the original radio logs.

The radio served the operation in two directions. Going out, it let a crew report its position, ask for weather, and tell the station ships it was approaching so that star shells and searchlights would be ready. Coming in, it gave the crew weather and position information that the ships could produce and the aircraft could not. In this way the radio extended the reach of the visual chain without replacing it. In clear weather a crew could fly from smoke to smoke or light to light and use the radio mostly for reporting. In fog or rain the radio became the only connection between the aircraft and the ships, and the system then depended on how well the crew could navigate from radio information alone.

Drift, sextant and radio bearings: how the boats found their way

The basic method was dead reckoning. The navigator worked from the compass heading, the airspeed, the elapsed time and an estimate of the wind, and he estimated the wind by measuring how far the aircraft drifted sideways over the sea surface. A drift indicator let him read that sideways movement against the waves below. A sextant could give a celestial fix when the horizon and the sky were both visible, and histories of the flight credit Lieutenant Commander Richard E. Byrd with work on navigation instruments for the NC boats, though Byrd himself did not fly the ocean legs. Each of these methods depended on seeing either the sea or the sky. The station ships provided the third kind of fix, a known position passed at a known time.

Radio direction finding was the only method that did not depend on visibility. A direction finder of the period used a rotating loop antenna. When the loop was turned toward a transmitting station, the received signal reached a sharp minimum, and the angle of the loop at that minimum gave a bearing to the transmitter. The Navy had developed shipboard and shore radio compasses during the war, and the NC boats are described as carrying direction-finding equipment so that they could take bearings on the transmissions of the station ships. In principle, a bearing on a ship whose position was known gave a line of position, and bearings on two ships gave a fix even when the ships could not be seen.

How much airborne direction finding actually contributed in 1919 is described differently in different accounts, and I do not put a weight on it. The equipment was new, an aircraft hull full of engine ignition systems was an electrically noisy place to take a bearing, and the crews had little experience using it in flight. The direction-finding principle could also work in the other direction, with ships or shore stations taking bearings on an aircraft’s transmissions and reporting the result back. Another article in this section covers that technique in its wartime anti-submarine form. For the NC flight, the safe statement is that radio bearings were part of the plan, the visual chain and dead reckoning carried most of the navigation, and the record does not support a strong claim either way about how much the bearings helped.

Fog short of the Azores: how NC-1 and NC-3 came down

The three boats left Trepassey on the evening of May 16, 1919, and flew through the night along the destroyer line. Station ships fired star shells and burned searchlights as the boats came over, and accounts describe the boats separating in the dark, each navigating from ship to ship on its own. After daylight on May 17, as the boats neared the Azores, they flew into fog and rain. Visibility dropped to the point where crews could no longer pick out the station ships, and at times could not see the sea or the sky well enough for a reliable drift reading or a sextant sight. Every visual method the navigators had used through the night became unavailable at about the same time.

NC-1 and NC-3 landed on the sea, by most accounts so that the crews could work out where they were before flying on. The sea was too rough for either boat to take off again, and both were damaged. A Greek steamer, the Ionia, rescued Bellinger’s crew on NC-1, and the abandoned boat later sank. Towers and the crew of NC-3 stayed with their boat and moved it across the water toward Sao Miguel. Accounts put the distance at roughly 200 miles over about two days, and they arrived at Ponta Delgada under the boat’s own power on May 19. NC-3 was in no condition to continue.

I describe what happened and do not rank the reasons. Histories of the flight point to the weather, the difficulty of finding the station ships in fog, and the decision to land and fix position, and each crew’s choices made sense with the information it had at the time. A crew that could not see the water could not safely judge a landing, and a crew that kept flying without a fix risked running out of fuel or missing the islands. Anyone who wants a judgment on those decisions should read the participants’ own reports and the Navy’s published histories, since an author writing a century later has much less information than the people in the boats did.

Horta, Ponta Delgada and Lisbon: why NC-4 finished alone

NC-4 entered the same fog and kept flying. Read’s crew saw the coast of Flores, the westernmost of the Azores, through a break in the weather. That gave them a position, and they continued to Horta on Faial, where they landed on the morning of May 17 after about fifteen hours in the air from Trepassey. Weather held the boat at Horta for several days. On May 20 it flew on to Ponta Delgada on Sao Miguel. It left the Azores on May 27, again along a line of station ships, and landed on the Tagus at Lisbon that evening. That landing completed the first crossing of the Atlantic by air.

The flight then continued toward England. NC-4 left Lisbon on May 30, made unplanned stops on the Mondego River in Portugal and at Ferrol in Spain because of mechanical trouble, and arrived at Plymouth on May 31, 1919. Some accounts end the story at Lisbon, which marks the transatlantic crossing, and others end it at Plymouth, which was the destination of the Navy’s plan. Both endings are correct as long as the writer says which one is meant. The boat survives. It is on display at the National Naval Aviation Museum in Pensacola, and anyone curious about the radio operator’s position or the hull layout can see them there.

NC-4 succeeded for reasons that overlapped. The most important is that it was still in the air when the others had landed, so the rough sea that disabled NC-1 and NC-3 never reached it. The break that showed Flores came at a useful moment, and histories of the flight are candid about the role of luck. The boat that had started worst, forced down off Cape Cod on the first day, ended up the only one to finish. I would add that the outcome was more than luck, because NC-4’s crew had a working radio, a navigator who was also the commander, and a line of station ships that had already brought them most of the way. Even so, I avoid explaining the outcome by any one cause, because the record supports several.

In stages, not nonstop: NC-4, Alcock and Brown, and the record

The claim that NC-4 was the first aircraft to fly the Atlantic nonstop is wrong, and it turns up in local news features and anniversary articles often enough to correct directly. NC-4 was the first aircraft to cross the Atlantic, and it did so in stages over about three weeks, with an extended stop in the Azores between the ocean legs. The first nonstop crossing came a few weeks later. John Alcock and Arthur Whitten Brown flew a modified Vickers Vimy bomber from St. John’s, Newfoundland, to a landing near Clifden in Ireland on June 14 and 15, 1919, in about sixteen hours, and that flight won the Daily Mail prize.

The two flights were built on opposite ideas. The Navy’s operation used multiple aircraft, a planned route with intermediate stops, and dozens of ships positioned to mark, track and rescue. Alcock and Brown flew a single land plane with two men aboard and no station ships, carried enough fuel to cross without stopping, and Brown navigated largely by dead reckoning and sextant through bad weather. I give both flights credit for what each actually achieved. NC-4 showed that an organized naval operation could move an aircraft across the ocean, and Alcock and Brown showed that a single aircraft could make the crossing in one flight.

A second error belongs in the same correction. Charles Lindbergh’s 1927 flight from New York to Paris was the first solo nonstop crossing, and it was not the first flight across the Atlantic, although public memory often treats it that way. The British airship R34 also made a round-trip crossing in July 1919, which included an east-to-west passage. A precise statement for a training slide or a public talk would read: NC-4 made the first transatlantic crossing by air, in stages, in May 1919, and Alcock and Brown made the first nonstop crossing in June 1919.

A chain of fixes: what 1919 says to anyone following aircraft today

The destroyer line was flight following carried out with ships. The aircraft reported its position, the ground reported the aircraft passing, each reporting point had a known location, and a rescue asset was always close. Modern air medical and public safety aviation do the same work with GPS, automated position reporting and a communications center that expects to hear from the aircraft at a set interval. The underlying questions have not changed: how often the aircraft reports, who notices when it stops, how long that person waits before acting, and who is told when they act. A 1919 station ship that saw and heard nothing at the expected time was a warning sign, just as a missed position report is today.

The fog over the Azores shows a design problem that ground agencies still face. The visual markers were redundant in number but not in kind, so a single weather condition defeated them all. A landing zone marked with lights and an incident commander’s voice on a channel the pilot cannot reach has the same weakness. The NC operation also shows that an aircraft’s communication capability can change when it goes down. A boat on the water with its trailing antenna useless could hear more than it could send, much as a downed or crashed aircraft today may have damaged radios, a weak emergency beacon, or occupants with only a cell phone in an area without coverage.

The station ships also worked because each one knew its own position accurately and its job in advance, and because the whole line ran on a published plan. A ground agency supporting an aviation asset at a wildland fire, a water rescue or a mass casualty incident needs the same preparation, worked out before the aircraft arrives. That means air-to-ground frequencies listed and tested, landing zone coordinates read from a GPS and not estimated from a map, and an overdue procedure written down. A plan that exists only in the experienced pilot’s head, or in a dispatcher’s memory, will not hold up the first time the weather closes in.

What to do at your agency

• This month, the communications center manager should pull the written procedure for following air medical or agency aircraft on assignment and confirm that it states the position reporting interval, the number of minutes after a missed report before the dispatcher acts, and the specific phone number or channel the dispatcher calls first.
• The radio system administrator should check every air-to-ground channel listed on the agency’s current ICS-205 template against a record of an actual radio check with an aircraft, and for any channel without such a record, schedule a test with the nearest air medical operator at its next training visit.
• The EMS or fire training officer should arrange one night landing zone drill this quarter with the air medical provider, and should include in the after-action notes how the zone was marked, how the pilot was talked in, and what the crew would have used if smoke or fog had hidden the lights.
• Where the jurisdiction includes open water, the emergency management director should confirm with the local Coast Guard sector or marine unit which agency takes the position report for an aircraft down on the water, and the dispatch supervisor should add that agency’s contact to the existing aviation emergency checklist.
• The dispatch supervisor should add one paragraph to the existing aircraft emergency SOP explaining that a downed aircraft’s occupants may be able to hear but not transmit, and stating that the dispatcher keeps broadcasting instructions and search information on the assigned channel even when no reply comes.
• Whoever maintains the agency’s training slides or public education material should search them for any description of an early Atlantic flight as the first nonstop crossing and correct it so that NC-4 is credited with the first crossing in stages and Alcock and Brown with the first nonstop crossing.

Takeaways

• Three Navy-Curtiss flying boats, NC-1, NC-3 and NC-4, left Rockaway, New York, on May 8, 1919, after the fourth boat, NC-2, had been taken out of the program for parts.
• The Navy placed destroyers along the ocean route, commonly described as about 50 nautical miles apart on the Newfoundland to Azores leg, and they marked the track with smoke by day and searchlights and star shells by night.
• The station ships gave the navigators position fixes, reported the boats’ passage by radio, and stood ready as rescue platforms, although published counts of the ships involved vary by source.
• Radiotelegraphy carried most of the useful radio traffic, and airborne direction finding was part of the plan, but accounts differ on how much the direction-finding equipment actually contributed.
• Fog near the Azores on May 17 took away the station ships and every visual method of navigation, and NC-1 and NC-3 landed on rough seas and could not take off again.
• NC-4 kept flying, sighted Flores through a break in the weather, landed at Horta, and completed the first transatlantic crossing by air when it reached Lisbon on May 27, 1919, before continuing to Plymouth on May 31.
• NC-4 was not the first aircraft to cross the Atlantic nonstop, a distinction that belongs to Alcock and Brown, who flew from Newfoundland to Ireland on June 14 and 15, 1919.
• The operation’s main lesson for ground agencies is that redundant markers that all depend on visibility will fail together, so aircraft following and landing zone plans need at least one method that works without line of sight.