Picture the scene: New York, 1938. A giant flying boat with four radial engines is being loaded not just with fuel, but with enough fuel to make the entire crossing to Europe — and then some. Because the plan isn’t simply to cross the Atlantic. The plan is to prove it can be done nonstop in both directions, against the prevailing winds, in a single epic demonstration flight. The aircraft is a modified Sikorsky S-42. The man behind it is André Priester, Pan American’s exacting chief engineer. And the ambition is nothing short of staggering for an era when transatlantic passenger service was still a dream scrawled on an engineer’s notepad.
The problem with the North Atlantic, as every avgeek knows, is that it is deeply asymmetric. The westerlies blow hard and reliably from west to east, meaning a flight from New York to London gets a welcome push, while the return leg fights into a brutal headwind. In the piston era, that headwind wasn’t just uncomfortable — it was potentially fatal to range. You could plan a one-way crossing with reasonable confidence. Planning a round-trip nonstop was a genuinely different category of challenge.
Pan American’s survey flights in the mid-to-late 1930s were methodical, obsessive exercises in data gathering. The airline didn’t just want to launch a service; it wanted to understand the Atlantic. Crews flew multiple proving runs, measuring winds at different altitudes, testing fuel burn in real conditions, working out what payload a viable commercial service could actually carry. The answers, when they came, were humbling. The westbound leg demanded so much fuel that early aircraft couldn’t carry meaningful payload alongside it. You could cross — but you couldn’t carry enough paying passengers to make the economics work.
That’s partly why the Boeing 314 flying boat, which finally entered transatlantic service in 1939, became such a celebrated machine. With its enormous fuel capacity, its four Wright Twin Cyclone engines, and a range that gave operators real flexibility in routing, it was the first aircraft that could tackle the North Atlantic commercially with something approaching viable economics. Passengers — wealthy ones — could finally book a seat on a scheduled transatlantic service. The crossing still took around 24 hours or more depending on conditions and routing, with stops typically included for fuel and crew rest, but the psychological barrier had been broken. Scheduled passenger flight across the Atlantic was real.
What makes the survey phase so compelling, looking back, is how much it resembled exploration in the old sense. Crews didn’t have GPS or satellite weather. They had celestial navigation, drift sights, radio bearings when they could get them, and their own judgment. The meteorology was primitive by modern standards. Every flight was genuinely an expedition. Pan Am’s navigators became some of the most skilled long-range aviators in the world simply because the Atlantic demanded it of them.
There’s something deeply satisfying about tracing the line from those careful, methodical survey flights to the transatlantic routes we take almost for granted today. The physics haven’t changed. The westerlies still blow. Modern widebodies still burn more fuel westbound than eastbound, and airlines still think carefully about routing to avoid the worst of it. The difference is that what once required weeks of planning and a kind of institutional courage now happens hundreds of times a day, automatically, invisibly, with passengers asleep in their seats somewhere over Greenland.
The Atlantic didn’t get easier. We just got better at crossing it. And it started with a flying boat, a headwind, and a team of engineers who refused to accept that it couldn’t be done.