Changing Winds: The 1934-1937 Chrysler Airflow

The streamlined Airflow remains the best known (and most infamous) of all prewar Chryslers, a bold and ambitious engineering achievement that became a notorious commercial flop. This week, we look at the origins and fate of the 1934–1937 Chrysler Airflow and its 1934–1936 DeSoto sibling.

1934 Chrysler CU Airflow Eight grille bars © 2007 George Camp (used with permission)
Grille of a late 1934 Chrysler CU Airflow Eight. (Photo © 2007 George Camp; used with permission)


Even before his death in 1940, Walter Percy Chrysler was becoming a larger-than-life figure. The son of a railroad engineer from Kansas, Chrysler rose from a 5-cent-an-hour job as a machinist’s apprentice to the presidency of Buick before his 43rd birthday. In the early 1920s, he helped to resuscitate one failing automaker and transformed two others into a highly successful eponymous corporation. He was TIME‘s Man of the Year in 1929 and by 1931, his name adorned the world’s tallest building.

The early history of the Chrysler Corporation, however, is also the story of three other men: Fred M. Zeder, Owen R. Skelton, and Carl Breer. It was Zeder, Skelton, and Breer who designed the initial 1924 Chrysler Six for what was then the Maxwell Motor Corporation, and they went on to oversee Chrysler engineering until after World War II.

The eldest of the three, Carl Breer, was originally from Los Angeles, a graduate of Throop Polytechnic Institute (now Cal Tech) and Stanford University. In 1909, he’d joined an Allis-Chalmers apprenticeship program in Milwaukee, Wisconsin, where he met Fred Zeder, several years younger and newly graduated from the University of Michigan’s School of Engineering. The two became fast friends, and when Zeder became chief engineer of Studebaker’s Detroit automotive operations four years later, he cabled Breer to join him there. Around the same time, Zeder hired Owen Skelton, a young engineer from Ohio who had previously worked at Packard.

In 1920, after several successful years at Studebaker, Zeder, Skelton, and Breer learned that Walter Chrysler was interested in recruiting them for Willys-Overland, whose financial backers had recently hired him to turn the struggling company around. At Willys, the trio developed a new, modern six-cylinder engine and an entirely new car, which at one time was intended to carry the Chrysler name. While Willys-Overland’s ongoing financial weakness — and conflicts between Chrysler and founder John N. Willys — eventually stymied those plans, Chrysler was very impressed with the three young engineers, suggesting in December 1921 that the trio start their own consulting firm. He even secured them their first contract with the Maxwell Motor Company, of which Chrysler had recently become a director.

In August 1922, Zeder, Skelton, and Breer re-approached Chrysler, who had departed Willys six months earlier to focus his attention on Maxwell and the moribund Chalmers Motor Company. The trio again broached the idea of building a six-cylinder car under the Chrysler name, using an engine of their design. Walter Chrysler was very receptive, so Zeder, Skelton, and Breer Engineering Co. soon secured a contract with Maxwell. In June 1923, ZSB Engineering was absorbed by the recently merged Maxwell-Chalmers Motor Corporation and the trio took over all engineering operations, including the development and launch of the first B-Series Chrysler Six, introduced to the public in January 1924.

The Chrysler Six was a great success, allowing Walter Chrysler to organize the Chrysler Motor Corporation, incorporated in Delaware on June 6, 1925. Three weeks later, the Chrysler Corporation absorbed Maxwell and Zeder, Skelton, and Breer all became senior Chrysler executives.

As vice president of engineering, Fred Zeder was nominally senior to Skelton and Breer, who became Chrysler’s executive engineer and chief of research, respectively. In practice, any divisions between them were ones of focus and specialization rather than rank. The three were lifelong friends; for more than two decades, they even sent out joint Christmas cards. Breer, who had boarded with Fred Zeder’s family when he first moved to Detroit, later married one of Zeder’s sisters and named their first son Fred. Walter Chrysler likened the trio to Athos, Porthos, and Aramis, Alexandre Dumas’ Three Musketeers.

Chrysler Building at night © 2008 David Shankbone (CC BY-SA 3.0 Unported)
Opened in 1930, the Chrysler Building, which stands 1,048 feet (319 meters) high, was briefly the world’s tallest building. It was designed by architect William Van Alen for developer William H. Reynolds, but Walter Chrysler bought the design in 1927. It was not owned by the Chrysler Corporation, although Chrysler commissioned the corporate engineering staff to develop its groundbreaking air conditioning system. Chrysler’s family sold their interest in the building in 1947. (Photo: “Chrysler Building at night” © 2008 David Shankbone; resized 2011 by Aaron Severson and used under a Creative Commons Attribution-ShareAlike 3.0 Unported license)

Chrysler’s relationship with his “Three Musketeers” was that of a confident and enthusiastic patron. In sharp contrast with Henry Ford, who tended to micromanage his technical staff, Chrysler had abiding faith in the expertise of his engineers. While he was often curious about their latest developments, he generally limited his involvement to the occasional word of encouragement. Moreover, even during the worst parts of the Depression, Chrysler’s research operations were largely exempted from corporate belt tightening.

It probably helped that none of the Three Musketeers was a wild-eyed dreamer. Chrysler introduced many significant engineering features during the trio’s long tenure (even the earliest 1924 cars had four-wheel hydraulic brakes, by no means the contemporary norm), but only one of their production car lines could be considered truly radical.


According to Carl Breer’s often-repeated account, the original impetus for the Airflow came in 1927, when Breer and his wife were summering in Gratiot Beach, in Port Huron, Michigan. Seeing a passing flight of Army Air Corps pursuit planes on their way back to Selfridge Field (some 35 miles/55 km away) one evening, Breer began musing on the contrast between aircraft design and the primitive state of automotive aerodynamics. Breer himself was no stranger to aviation engineering. During World War I, while still a Studebaker employee, he had worked with O.E. Hunt and James Heaslip on the production engineering of the 12-cylinder Liberty engine used in many Allied aircraft.

Aerodynamic streamlining was not a new idea in automotive design even in 1927. In the early twenties, a number of aviation engineers had turned their attention to automobiles, particularly in Germany, where the Armistice had placed sharp restrictions on the postwar aircraft industry. At the 1921 Berlin Auto Show, for example, Austrian engineer Edmund Rumpler had displayed a prototype of his remarkable Tropfenwagen, a teardrop-shaped, mid-engined car with a single headlight and a 157 cu. in. (2,580 cc) W6 engine. Later wind tunnel tests revealed that the Tropfenwagen had a drag coefficient of only 0.27, highly respectable even today.

Rumpler didn’t find an automaker willing to put the car into mass production, but Benz et cie adapted his design and chassis for the Benz Tropfenwagen racer (developed by Willy Waub), which competed in Grand Prix events in 1923 and 1924. Rumpler persevered, eventually developing the short-lived Tropfen-Auto RU 4A106, launched in 1924. Powered by a 160 cu. in. (2,614 cc) four with 50 horsepower (37 kW), about 100 of the futuristic-looking cars were built before production ceased in 1925. Some ended up as taxicabs in Berlin, although their poor reliability and limited cargo space made them unpopular. Director Fritz Lang acquired a handful of Tropfen-Autos for his 1927 science fiction epic Metropolis; all were destroyed in the film.

1932 Bergholt Streamline front 3q © 2010 Pat McLaughlin (used with permission)
Another early aerodynamic design was the 1932 Bergholt Streamline, developed by aviation engineer Fred Bergholt. Based on a 1932 Ford Model 18 chassis, the Streamline was a one-off, although Bergholt tried without success to find an automaker willing to put it in production. He kept the car until his death in 1978, sometimes using it as his daily driver. It was restored in the early nineties; it’s seen here at the 2010 Concours d’Elegance of America at Meadow Brook. (Photo © 2010 Patrick McLaughlin; used with permission)

We don’t know exactly how much Breer knew about these and other contemporary experiments, but in any event, Chrysler Engineering’s practical knowledge of aerodynamics in 1927 was almost nil. That fall, Breer hired Bill Earnshaw, a Dayton, Ohio-based consulting engineer and personal friend of the Wright Brothers, to conduct preliminary aerodynamic research. In November, Orville Wright helped Earnshaw set up a small wind tunnel for testing purposes. Intrigued by Earnshaw’s results, Breer persuaded Walter Chrysler to authorize construction of a larger, in-house wind tunnel in September 1928. Aerodynamic work continued even after the stock market crash in the fall of 1929, and by 1931, Chrysler engineers had tested at least 50 scale models.


It was hardly shocking when Chrysler’s early wind tunnel tests revealed that most late-twenties cars produced tremendous drag; considering the blunt radiator shells and flat, upright windshields of the day, it was inevitable. More surprising was the observation that many closed bodies were significantly slipperier in reverse than they were moving forward, in some cases by up to 30%.

As Breer’s team soon recognized, that curious fact was a byproduct of what was then conventional engine and suspension layout. Since most passenger cars still had solid axles front and rear, their engines were usually mounted behind the front axle, for what today we would call a front/mid-engine layout. The position of the engine pushed the passenger compartment well back in the chassis, particularly with bulky straight-eight engines. In most closed bodies, the cabin ended just forward of the rear bumper; the “trunkback” or “notchback” profile was still a few years in the future. The net result was a long, narrow nose and a wide, bulbous tail, exactly the opposite of the aerodynamically ideal teardrop shape. Breer concluded that significantly reducing drag would require a very different profile with a broad, smooth nose and a narrow, tapering tail.

One way to achieve that was to simply reverse the customary layout, putting the engine behind the passenger compartment, as Rumpler had done with the Tropfenwagen and Tropfen-Auto. Breer considered that possibility, but ultimately abandoned it, partly out of concern for the effects of a rear engine on handling and stability, partly because such a layout would have required too much new drivetrain hardware. Another possibility, tacking an extended tail cone onto a mostly conventional body, was aerodynamically effective, but neither very practical nor particularly attractive. However, without such addenda, a sharply sloping tail would take a big chunk out of rear passenger space unless the seating layout and packaging were significantly revised.

1932 Trifon Special front 3q Chrysler Historical Collection
A vintage Chrysler photo of the Trifon Special, the first running prototype of the Airflow, shot in December 1932. With its plain bumpers, sloping hood, and small radiator opening, it bears only a general resemblance to the production Airflow, although it has the one-piece curved windshield found only on the big CW Airflow Imperial Custom limousines. Note the running boards, which were later removed from this car. (Photo circa 1932, copyright © FCA US LLC – Historical Services; used with permission)

With that in mind, Breer and chief body engineer Oliver Clark set about rethinking the packaging of the typical passenger car. They started with six-cylinder sedans and coupes — what in production would become the DeSoto Airflow and the Canadian Chrysler CY — and laid out a seating arrangement that would allow the desired shape while keeping the wheelbase and overall length as short as possible. First, they widened the front seat to make it somewhat wider than the rear bench, the reverse of the usual practice. (Initial plans actually called for five-passenger seating, with three in front, two in back, but the Chrysler sales organization said no.)

The rear seat, meanwhile, was moved forward of the rear axle with the front seats, dashboard, and windshield shifted forward accordingly, not unlike Chrysler’s much later “cab-forward” LH cars. Those changes required the engine to be shifted about 20 inches (51 cm) forward, positioning it above the front axle rather than behind it. In production Airflows, the engine’s center of mass was slightly behind the axle and the engine was tilted downward about 5 degrees at the clutch side to minimize the height of the driveshaft tunnel.

Even with essentially stock engines and drivetrains, making these changes was complicated and expensive, and their actual value was debatable. Most American highways had speed limits of 45 mph (72 km/h) or less in those days, and streamlining provided little benefit at lower speeds. In Europe, where taxable horsepower rules favored smaller displacements, obtaining a respectable maximum speed with a tiny engine had some allure, but that was not a major concern in the U.S. market.

If better aerodynamics had been the only advantage of Breer and Clark’s repackaging job, we suspect that the Airflow might never have made it to production. However, Breer and his team found that the new layout paid unexpected dividends in a far more marketable area: ride quality.


A smooth ride was not on the menu for most passenger cars of the late twenties and early thirties except perhaps for limousines, whose long wheelbases and prodigious sprung weight masked many sins. The gradual adoption of low-pressure tires and shock absorbers had helped a bit (the latter hadn’t even been offered on many cars of the twenties, including the original Chrysler Six), but solid axles, stiff semi-elliptical springs, and substantial unsprung weight still made for less than cloud-like comfort. On relatively smooth roads, the driver and front passenger didn’t fare too badly, but back seat passengers, perched above or behind the heavy rear axle, were not so fortunate.

1934 Chrysler CU Airflow interior Chrysler Historical Collection
The Airflow’s seats were mounted quite high, allowing better circulation of heated air to rear passengers and making the most of the available legroom. Some taller drivers complained that the high-mounted seats hurt visibility, so they were lowered somewhat on later models. The rear seat back flips up from the bottom to access the luggage area; Airflow sedans didn’t gain exterior luggage access until 1936. (Photo circa 1934, copyright © FCA US LLC – Historical Services; used with permission)

Breer discovered that the packaging changes of the streamlining project did much to improve that sorry situation. First, for rear passengers, sitting within the wheelbase proved far more comfortable than sitting atop the axle. Second, moving the engine forward put about 55% of the car’s static weight on the front wheels, compared to perhaps 40-45% for conventional sedans of the era. With less weight on the tail, the rear springs could be substantially softer. By lengthening (and thus softening) the front springs to match, it was possible to avoid a pendulum effect while also reducing the overall frequency of the ride motions by about 25%, which, as Breer often noted, put it closer to a natural walking pace. Chrysler advertising christened it the “Floating Ride.”

Surprisingly, this was achieved without independent front suspension, although Chrysler was then developing a double-wishbone layout for Plymouth, Dodge, and conventionally engineered six-cylinder Chryslers. While the production Airflow’s tubular front beam axle might seem a retrograde step in an otherwise sophisticated car, Breer decided that IFS simply wasn’t necessary. It’s important to remember that a major rationale for Detroit’s adoption of independent suspension in this era was not better handling, but a smoother ride. The revised spring rates and altered weight distribution achieved that without the added cost, complexity, and maintenance headaches of early IFS systems.

1934 Plymouth PE Deluxe sedan front 3q
The more expensive 1934 Plymouths, like this PE Deluxe sedan, had double-wishbone front suspension with coil springs, but the 1935 PJ models reverted to a front beam axle, moving the engine about eight inches (20 cm) forward and adopting longer, softer leaf springs for an Airflow-style “Floating Ride.” Plymouth didn’t return to IFS until 1939.


Since the streamlined car was already laid out very differently than conventional models, Breer decided it was also a good candidate for new techniques in body construction.

Most passenger cars of the early thirties used body-on-frame construction with a heavy, rigid ladder-type frame. The body itself was usually steel or aluminum panels over a wooden framework, although all-steel bodies were becoming increasingly common. The Dodge Brothers had adopted them early on; Plymouth went all-steel in 1930, followed by Ford in 1932. For the streamlined Airflow, however, Breer and Clark opted for something far more advanced: a steel monocoque.

Like streamlining, monocoque construction dated back to the end of World War I. Although it was gaining popularity in aviation in the late twenties and early thirties, Italian engineer Vincenzo Lancia had applied for a patent on self-supporting automobile bodies in 1919, likely inspired by shipbuilding practice, and the first unit-bodied Lancia Lambda went into production in 1922. Thanks to its greater weight and space efficiency, monocoque construction gradually found favor in Europe, but American automakers remained wary. In 1931, for example, Joseph Ledwinka and William Mueller of the body supplier Budd had found no takers for their unit-bodied FWD prototype, although historian Michael Lamm believes their car was the inspiration for the unit-bodied Citroën 11CV “Traction Avant,” which debuted in early 1934; Budd developed the tooling for the Traction.

1932 Trifon Special side © 2011 Ronnie Schreiber WPC (used with permission)
Although it used the 242 cu. in. (3,958 cc) DeSoto six, the original “Trifon Special” prototype (seen here in the basement of the Walter P. Chrysler Museum in Detroit) was somewhat smaller than the production DeSoto Airflow: 189 inches (4,801 mm) long on a 115-inch (2,921mm) wheelbase. Nonetheless, it had a lavishly appointed interior with burled walnut trim, reversible seat cushions, armrests with integral storage compartments, and rear-seat vanity mirrors mounted on the B-pillars. At one time, the wooden artillery wheels were fitted with louvered covers, which were later used on CW Imperials. (Photo © 2011 Ronnie Schreiber; used with the permission of Cars in Depth and the Walter P. Chrysler Museum)

Chrysler’s approach, also developed in partnership with Budd, was not quite a monocoque. Commonly described as “bridge-and-truss” construction, it used stressed exterior panels welded to a cage-like steel structure that included the frame, the cowl and windshield frame as well as vertical members through the roof pillars and horizontal rails above the door openings. While the structure looked somewhat like a tubular space frame, it was not self-supporting, relying on the exterior panels for its torsional stiffness. The bridge-and-truss system was bulkier and heavier than true monocoque construction, but it simplified production, allowing the engine and running gear to be installed before the exterior panels were attached, much like a body-on-frame design.

When the Airflow was introduced, there were some ugly allegations that it was unsafe due to the supposed flimsiness of its metal framework, which dispensed with the customary wood bracing. It was true that the framework itself was not very rigid, but it wasn’t intended to be. When the sills and exterior panels were welded in place, the unitized structure was extremely strong. (Chrysler boasted that it had 40 times the torsional stiffness of a body-on-frame design.) Chrysler eventually laid the rumors to rest in dramatic fashion by demonstrating that the Airflow remained drivable even after being pushed off a 110-foot (33-meter) cliff. In the summer of 1934, the company also hired Barney Oldfield’s “Hell Drivers” racing team to perform hourly demonstrations of the Airflow’s resilience to rollovers in a sand pit at the Chicago World’s Fair.

The downside of that strength was added weight. Breer said that ideally, unitized construction would have saved up to 200 lb (90 kg) over a body-on-frame design, but as the development process continued and exterior design and trim were refined, the cars grew progressively heavier. The production Airflows weighed substantially more than the body-on-frame cars they replaced, in some cases by more than 250 lb (115 kg). Later models, with different grilles and stouter bumpers, would be heavier still.


The first true Airflow prototype was built in great secrecy in the summer of 1932. Powered by a DeSoto straight six, it had bridge-and-truss construction, a relocated engine, a repackaged interior layout, and Floating Ride suspension. It had no grille and carried no Chrysler badges or identification of any kind. When the company registered and licensed it with the state of Michigan for road testing, the prototype was identified as the 1932 Trifon Special, taking its name from test engineer Demitrion Trifon.

Because the company did not yet have its own proving grounds, Chrysler arranged to test the cars on land owned by two farmers outside Grayling, Michigan, north of Detroit. To avoid attracting attention, test crews did not drive the prototype on public roads and engineers and executives heading up to Grayling were asked to drive non-Chrysler vehicles.

1932 Trifon Special rear © 2011 Ronnie Schreiber WPC (used with permission)
The Trifon Special prototype had various features that the initial production Airflow sedans did not, including a one-piece, roll-down rear window, an integral trunk with external access, and an all-steel roof. The latter two features were added to production cars in 1936, but all Airflows had fixed two-piece backlights. (Photo © 2011 Ronnie Schreiber; used with the permission of Cars in Depth and the Walter P. Chrysler Museum)

The first Trifon was an awkward-looking beast with curious proportions: the work of engineers, not stylists. Even the production Airflows were developed with little input from either Chrysler’s fledgling in-house styling department or Briggs Mfg. Company, which provided much of Chrysler’s exterior design work in those days. Chrysler’s Art & Colour section, founded in 1928, still focused primarily on detailing, upholstery, and trim; its chief, Herbert V. Henderson, had previously been an interior designer. Both Henderson’s group and Briggs designers contributed to the nicely detailed Streamline Moderne interiors of the production cars, but their involvement in the exterior design was limited to details like the rear fender skirts. For better or worse, the Airflow’s shape was primarily the work of Carl Breer, Oliver Clark, and Chrysler’s body engineers.

Aesthetics aside, the Trifon Special drove quite well, with excellent ride quality even on bumpy country roads. After months of testing, Breer and his team finally invited Walter Chrysler himself up to Grayling for a test ride. Chrysler was extremely impressed and said he wanted to put the car into production. He was not daunted by its radical design; he thought it would serve as a new corporate flagship, demonstrating how far the Chrysler Corporation had come in ten years and setting a new direction for the future.

1932 Trifon Special front © 2011 Ronnie Schreiber WPC (used with permission)
Because the Airflow’s wide front seat made for a wider-than-normal windshield area, Carl Breer and Oliver Clark originally wanted to use one-piece curved glass on all models, judging that a conventional split windshield would be unattractive at that size. The Pittsburgh Plate Glass Company developed the curved windshield for Chrysler, seen here on the 1932 Trifon Special, but it proved prohibitively expensive and was used only on the CW Imperials. (Photo © 2011 Ronnie Schreiber; used with the permission of Cars in Depth and the Walter P. Chrysler Museum)


If it had been strictly up to the engineers, the Airflows probably wouldn’t have arrived until the 1935 model year, perhaps even later. However, the 10th anniversary of the first Chrysler was fast approaching and Walter Chrysler wanted to show off the new car at the New York Auto Show in January 1934.

1934 Chrysler Model CU Airflow Eight front 3q Chrysler Historical Collection
An early press photo of the 1934 Chrysler Model CU Airflow Eight. The CU sedan was 207.1 inches (5,260 mm) long, 70 inches (1,778 mm) tall, and 70.5 inches (1,791 mm) wide, riding a 122.8-inch (3,120mm) wheelbase; curb weight was over 4,000 lb (1,830 kg). Four-door sedans accounted for most of the 8,389 CU Airflows sold in 1934, but there was also a four-door Town Sedan (with no rear quarter windows), as well as a two-coupe and the two-door Brougham, which had the sedan’s profile and rear clip, but the coupe’s longer doors. There were no convertibles, roadsters, or phaetons as the Airflow’s unitized construction precluded open body styles. (Photo circa 1934, copyright © FCA US LLC – Historical Services; used with permission)

Part of the urgency may have stemmed from reports of a secret GM prototype called the Albanita, which GM Engineering had begun testing at the corporate proving grounds in Milford, Michigan, in the spring of 1933. The Albanita was a streamlined, midsize sedan, slightly smaller than the Trifon Special and sporting several features later introduced on production Chevrolets, including an all-steel Turret Top and vent windows. The GM car rode an unusual backbone chassis and was powered by a Ford flathead V8.

Chrysler was almost certainly aware of the Albanita; GM engineer Ollie Schjolin, the project director, frequently had to chase away various auto industry observers prowling outside the Milford proving grounds. However, Chrysler may not have realized that the Albanita was only an engineering testbed that GM Engineering was using to evaluate the backbone frame and two different types of independent suspension. Some GM engineers believed that the Albanita prompted Chrysler to rush the Airflow into production, fearing that their thunder was about to be stolen.

Since they required entirely new tooling and equipment, the production Airflows were undoubtedly expensive to build, driving up their retail prices. Chrysler management decided to introduce the new models in the more-expensive DeSoto and Chrysler brands, although there were tentative plans to add Dodge and Plymouth versions later. Carl Breer and research chief Ken Lee also started work on a compact with front-wheel drive and an unusual 67 cu. in. (1,091 cc), five-cylinder radial engine, intended either for export markets or as a possible economy model for North America.

1934 Chrysler CW Airflow Imperial Custom front 3q Chrysler Historical Collection
The most expensive 1934 Airflow was the big Chrysler Model CW Airflow Imperial Custom Eight, which cost over $5,000 — comparable to a Cadillac V-12, and more expensive than any Packard Twelve except the rare factory customs. Built by LeBaron, the CW had many unique features, including a curved, one-piece windshield (the first on any American production car); its own dashboard; bigger doors; larger wheels, springs, and brakes; a bigger clutch, radiator, and fuel tank; bumpers with four horizontal bars, rather than three; and a 385 cu. in. (6,306 cc) straight eight with nine main bearings and 150 gross horsepower (112 kW). The CW was about 4 inches (10 cm) wider than any other Airflow and stretched a mammoth 233.1 inches (5,920 mm) overall, with a curb weight of over 6,000 lb (2,730 kg). (Photo circa 1934, copyright © FCA US LLC – Historical Services; used with permission)

The smallest and cheapest U.S. Airflow would be the DeSoto Model SE, offered in four body styles, powered by a 242 cu. in. (3,958 cc) flathead six. A Chrysler Airflow Six, the CY, would be offered only in Canada, sharing its 115.5-inch (2,934mm) wheelbase and engine with the DeSoto; six-cylinder U.S. Chryslers would retain conventional styling and engineering. The smallest U.S.-market Chrysler Airflow, the CU Airflow Eight, would be offered in the same body styles as the DeSoto, but on a 122.8-inch (3,120mm) wheelbase with a 299 cu. in. (4,894 cc) straight eight.

Above that would be two Airflow Imperial Eight series, the CV and CX, riding either a 128-inch (3,251mm) or 137.5-inch (3,492 mm) wheelbase and using a stroked 324 cu. in. (5,301 cc) eight. At the top of the line would be the massive CW Airflow Imperial Custom Eight, riding a 146.5-inch (3,721mm) wheelbase and available as either an eight-passenger sedan or limousine, with or without blind quarter panels. To cope with their three-ton mass, all CWs would be powered by the big 385 cu. in. (6,306 cc) nine-bearing eight first introduced on the CG Imperial in 1931.

Most of the major stampings for Airflow bodies were made by Budd, although Chrysler itself manufactured the complex front clips, and the CW Imperials were built as semi-customs by the coachbuilder LeBaron, which Briggs had owned since 1927.

Despite the proliferation of wheelbases and body styles, all Airflows had a remarkable degree of commonality, surpassing even GM’s new A-B-C-D bodies. The Chrysler CU, for example, shared the same roof panel and rear clip as the short-wheelbase DeSoto and Chrysler CY, with welded-in extensions to the floorpan and side stampings ahead of the firewall to allow room for the longer straight-eight engines. The Imperials added more welded-in sections to create their longer wheelbases. All sedans shared the same rear clip, while all coupes shared their own tail section. Perhaps the most ingenious trick was the use of interchangeable doors, allowing Chrysler to create five different door configurations with only two sets of basic tooling. Such commonality helped to reduce production costs, offsetting some of the expense of unitized construction.

1934 DeSoto Airflow sedan front 3q high Chrysler Historical Collection
The most popular Airflow was the 1934 DeSoto Model SE four-door sedan, which accounted for 11,713 sales — more than 20% of all Airflow production. The SE was shorter than the eight-cylinder Chrysler CU, 193 inches (4,902 mm) long on a 115.5-inch (2,934mm) wheelbase. All of the difference was between the firewall and the front axle; SE and CU bodies were basically identical aft of the cowl. Note the fabric roof insert; until the mid-thirties, even “all-steel” closed bodies generally had fabric roofs to avoid drumming. It was replaced in 1936 by a bolt-in steel insert. (Photo circa 1934, copyright © FCA US LLC – Historical Services; used with permission)

While their engines and three-speed gearboxes (synchronized in second and third gears) were largely conventional, the Airflows introduced an important new mechanical feature: an automatic overdrive. Invented by Rex Keller and manufactured by Borg-Warner’s Warner Gear division, the overdrive unit was essentially a two-speed rear axle. When engaged via a dashboard switch, a centrifugal governor shifted automatically into the axle’s 0.70 overdrive ratio whenever speeds exceeded 40 to 45 mph (64 to 72 km/h), shifting back if speed dropped below 25 mph (40 km/h). Unlike the Columbia unit popular for Fords and Lincolns a few years later, there was no automatic ‘kickdown’ to direct drive.

Carl Breer later admitted the overdrive was an afterthought, added when production Airflows proved to be significantly heavier and thirstier than originally intended. In 1934, the overdrive was combined with freewheeling, which automatically disengaged the clutch whenever the driver lifted completely off the throttle. (The freewheeling feature was dropped in 1935 — some U.S. states were beginning to outlaw it, concerned about its deleterious effect on engine braking — but Chrysler’s brochures listed it as standard on Imperials as late as 1936.)


While the Airflow entered pilot production by December 1933, ramp-up was very slow, and only about 60 cars had been built by early January. Nonetheless, the Airflow made its scheduled debut on the main stage of the New York Auto Show on January 6, after weeks of teaser advertisements in major magazines.

1934 Chrysler Airflow CU front 3q © 2010 Rex Gray (CC BY 2.0 Generic)
The 1934 Chrysler CU Airflow had a 299 cu. in. (4,894 cc) straight eight, making 122 hp (91 kW) and 225 lb-ft (304 N-m) of torque on 6.5:1 compression. The Airflow was one of the first American cars with a front-opening ‘alligator’ hood; the body sides extend all the way to the nose, allowing them to accommodate the headlights. Integrated headlights were quite novel in 1934, although Chrysler had to add separate parking/passing lamps because the main beams were a bit too small to meet state rules for minimum headlight size. (Photo: “1934 Chrysler Airflow sdn – blue – fvl” © 2010 Rex Gray; used under a Creative Commons Attribution 2.0 Generic license)

The Airflow made quite an impression even on jaded auto show regulars. Streamlined show cars were increasingly common in the early thirties — the spectacular Pierce Silver Arrow had debuted in 1933, and Buckminster Fuller’s wild three-wheel Dymaxion was parked out front at the 1934 show after Chrysler arranged to displace Fuller from the main stage. However, the Airflows were full-fledged production models, aimed (excepting the big Imperials) at the conservative middle-class market. Some show goers were entranced; others were simply appalled. When the independent Market Research Corp. of America asked visitors to name the best- and worst-looking cars in the show, the Airflow topped both lists. Like the “coming or going” postwar Studebakers and Ford’s Edsel, the Airflow became the butt of many jokes, although the DeSoto version won top honors at the Monte Carlo Concours d’Elegance in both 1934 and 1935.

1934 Chrysler Airflow CU rear 3q Rex Gray © 2010 (CC BY 2.0 Generic)
Rear fender skirts were standard on all production Airflows, but whitewall tires and wheel trim rings were extra, although commonly specified. The three-bar bumpers (four-bar on CW Imperials) were attractive, but dented easily. On U.S. cars, sturdier bumpers were added in 1935, although CWs and some export cars retained the original bumpers through 1936. An external spare was standard on 1934–1935 Airflow sedans. (Photo: “1934 Chrysler Airflow sdn – blue – rvr” © 2010 Rex Gray; used under a Creative Commons Attribution 2.0 Generic license)

While Chrysler sales reps left the New York show with thousands of advance orders, the company was initially able to fill only a handful of them. Airflow production didn’t really get up to speed until almost three months after the show, during which time many frustrated customers either opted for the cheaper, more orthodox-looking Chrysler CA and CB Sixes or took their business elsewhere.

Customers who did receive early-production Airflows were not necessarily any happier. The Airflow’s design was basically sound, but its manufacture involved many new assembly techniques, and teething problems were inevitable. In his memoir, Carl Breer downplayed those issues, but his son Fred later recalled Breer reading the many angry letters from disgruntled owners and Chrysler dealers about the many failings of the early production cars. Most of those problems were addressed quickly enough, but they did nothing for the new car’s reputation, nor did rumors about the Airflow’s lack of structural integrity.

1934 DeSoto Airflow coupe front © 2008 Bill McChesney (CC BY 2.0 Generic)
The 1934 DeSoto Airflow had unique headlights, larger than those of the Chrysler models, and small grilles beneath the lights for the horns, deleted in 1935. Both DeSoto and Chrysler Airflows had unusual — and no doubt expensive — front windows with inset vent windows that could be swung out in the normal fashion (with or without lowering the side window) or rolled down along with the entire window and frame. (Photo: “2860 1934 DeSoto Airflow” © 2008 Bill McChesney; used under a Creative Commons Attribution 2.0 Generic license)

For those willing to wait — and to embrace its controversial looks — the Airflow offered excellent performance. Even the six-cylinder DeSoto was capable of around 85 mph (136 km/h), not bad for a 3,600 lb (1,630 kg) car with only 100 gross horsepower (75 kW). With the optional overdrive, racing driver Harry Hartz averaged a commendable 21.4 mpg (11 L/100 km) in an AAA-supervised run from New York to San Francisco. The eight-cylinder Chrysler CU, meanwhile, could reach 60 mph (97 km/h) in a bit over 20 seconds, lively for the time, with a top speed of over 90 mph (145 km/h). A CV Imperial coupe, also driven by Hartz, set more than 70 production car speed records at Bonneville in mid-1934, hitting 95.7 mph (154.1 km/h) in the flying mile and then averaging over 18 mpg (12.9 L/100 km) between Los Angeles and New York City. All these were excellent figures for any full-size mid-thirties automobile.

Such speed and economy came at no sacrifice in comfort. Although outward visibility left something to be desired, Airflows were quite roomy. Even the front seats could seat three abreast where conventional Chryslers of the period were a little tight even for two. With their ample weight and slow steering, no Airflow was particularly nimble, but they handled with composure and their reasonably tight turning circle aided low-speed maneuverability. As advertised, the Airflow’s “Floating Ride” was superb, rivaling even some IFS-equipped contemporaries.

1934 DeSoto Airflow coupe dash (CC BY 2.0 Generic)
1934 DeSoto Airflow coupe engine (CC BY 2.0 Generic)
Top: Overdrive was standard on Chrysler Airflow Imperials and Imperial Customs, but it was initially a $35 option on Chrysler CU and DeSoto models. The paper tag on the dashboard of the ’34 is the original instruction card, which warns that the clutch should be disengaged before pressing the overdrive button, and that it should not be engaged at speeds over 45 mph (72 km/h). Bottom: All DeSoto Airflows and the Canadian Chrysler CY had a 242 cu. in. (3,958 cc) L-head straight six, rated at 100 gross horsepower (75 kW) with a 6.2:1 compression ratio. On non-Airflow Chrysler Sixes, the same engine had 5.2:1 compression and 93 horsepower (69 kW), with the higher-compression head available optionally. (Dash photo (“2566 1934 DeSoto Airflow”) and engine photo (“2562 1934 DeSoto Airflow”) above © 2008 Bill McChesney; both used under a Creative Commons Attribution 2.0 Generic license)

Unfortunately, even when Airflows finally became available in quantity that spring, sales were well below expectations. Aside from the styling, the major sticking point was the prices, which started some 20–25% higher than the conventionally engineered models the Airflows replaced. The DeSoto Model SE Airflow, for instance, cost up to $330 more than the 1933 Model SD, making it roughly the same price as the eight-cylinder Model CF that DeSoto had dropped two years earlier due to lack of demand. The Chrysler CU Airflow Eight, at $1,345, was more expensive than most 1933 CQ Imperials. The U.S. economy was better in 1934 than it had been in 1932, but not that much better.

The Airflow’s lackluster sales provoked sharp divisions among Chrysler management. The sales organization, then headed by Joseph W. Frazer (later of Willys-Overland and Kaiser-Frazer fame), blamed the weak business on the Airflow’s quirky looks and wanted to see it replaced, or at least supplemented, by conventional models. Even some of the engineering staff agreed. Carl Breer naturally defended the Airflow, as did Walter Chrysler, who called it the car of the future, resisting efforts by DeSoto and Chrysler managers to kill it.

1934 DeSoto Airflow coupe side Bill McChesney (CC BY 2.0 Generic)
1934 DeSoto Airflow coupe rear 3q Bill McChesney (CC BY 2.0 Generic)
The sleek fastback coupe is arguably the best-looking Airflow, particularly in profile. Coupe tails taper a bit more than those of the sedans, so Chrysler initially claimed five-passenger seating for the coupes rather than six. The doors of the coupes are longer than the front doors of the DeSoto SE, Chrysler CU, and CV Imperial sedans, but the bigger CX Imperial sedans and limos used coupe doors in front to match their longer wheelbases. The CW Imperial Custom used the coupe doors both front and rear, with an extra section welded in below the bottom hinge line of each door to allow a taller door opening. (Side view photo (“2861 1934 DeSoto Airflow”) and rear 3q view photo (“2866 1934 DeSoto Airflow”) above © 2008 Bill McChesney; both used under a Creative Commons Attribution 2.0 Generic license)

On May 24, the Airflow’s woes were compounded by a new problem: a pending lawsuit file by the Jaray Streamline Corporation of America, charging Chrysler with patent infringement.


Engineer Paul Jaray (born Pál Járay) had a great deal in common with Edmund Rumpler. Like Rumpler, Jaray was Jewish, originally from Austria (although his family was Hungarian), but spent his early career in the fledgling German aviation industry, joining Luftschiffbau Zeppelin GmbH (Zeppelin Airship Works Ltd.) in 1912. Although Jaray eventually became Zeppelin’s chief designer, business was slow after the war, and Zeppelin allowed him to pursue a sideline: applying aerodynamic principles to passenger cars.

Jaray’s first car, built in 1921, was the Ley T6, a tiny streamliner with a skeletal frame, not unlike the Airflow’s bridge-and-truss system. The T6’s drag coefficient was only 0.28, allowing the little car to reach a top speed of 62 mph (100 km/h) with only 20 hp (15 kW). Based on those experiments, Jaray applied for a patent in August 1922 on automobiles with streamlined bodies. It covered essentially any car whose aerodynamic form comprised one partial teardrop shape (the greenhouse) atop another (the body itself).

1934 Chrysler CU Airflow Eight sedan front 3q © 2011 Ronnie Schreiber WPC (used with permission)
1934_Chrysler_CU-AirflowEight-sedan_grille_R1934 Chrysler CU Airflow Eight sedan grille © 2011 Ronnie Schreiber WPC (used with permission)
This Chrysler CU four-door sedan, photographed at the Walter P. Chrysler Museum and sporting Ontario license plates, is a late 1934 model, identifiable by its simplified grille. Early 1934 Chrysler Airflows had 39 vertical grille bars, but later in the year, Chrysler substituted a cheaper version with only 21 bars, a change suggested by design consultant Norman Bel Geddes (apparently one of the only suggestions Chrysler actually used). Note the winged hood ornament, which has a Chrysler blue ribbon emblem at its center, and the functional engine compartment cooling louvers, seen here in their open position. (Both above photos © 2011 Ronnie Schreiber; both used with the permission of Cars in Depth and the Walter P. Chrysler Museum)

In 1923, Jaray left Zeppelin and moved to Zurich, Switzerland, where he and a business partner founded the consulting firm Stromlinen Karosserie Gessellschaft (roughly, “Streamlined Body Co.”). After his initial patent was issued in June 1927, Jaray licensed it to a number of automakers: initially German firms like Apollo and Maybach, later the Czech firm Tatra, where Jaray’s concepts became the basis of the remarkable Tatra T77 and T77a. The latter, launched in 1935, had a drag coefficient of only 0.21, better than any modern production car.

In 1932, Jaray established the Jaray Streamline Corporation of America, based in New York. Although the company developed a number of prototypes in hopes of interesting Detroit automakers (one based on a Chrysler Imperial chassis), they only managed to sell one license, to the coachbuilder Rollston.

1934 Chrysler Airflow Six CY front 3q © 1992 Richard Spiegelman (used with permission)
While all DeSoto Airflows had sixes, the only six-cylinder Chrysler Airflow was the short-lived CY Airflow Six, offered in the Canadian market in 1934. The CY was essentially a DeSoto SE (including the DeSoto’s headlights, horn grilles, and cooling louvers) fitted with CU-style bumpers, grille, and dashboard. Built for only one year, CY production totaled fewer than 500 units, most of them four-door sedans like this one. We don’t know if any were exported to Europe or other overseas markets. (Photo: “1934 Chrysler Six Airflow CY 4 door (Canadian)” © 1992 Richard Spiegelman; used with permission)

Chrysler Engineering had at least some knowledge of Jaray’s work (Owen Skelton had received a copy of one of Jaray’s articles in March 1933 from the Society of Automotive Engineers’ Lowell Brown), but it’s unclear if they were aware of his 1927 patent. Chrysler was prepared to take the matter to court, but eventually opted not to, on the advice of an independent aerodynamics consultant the company had hired as a prospective expert witness. In June 1935, Chrysler settled with Jaray out of court, agreeing to license the Jaray patent for $5,000 plus a small royalty on Airflows exported to Europe.

The Jaray Streamline Corp. pursued similar action against Pierce-Arrow (over the Silver Arrow) and several other companies, but with the Depression killing smaller automakers left and right, the suits netted very little financial return. Jaray finally gave up and dissolved his U.S. firm, although he continued to work and lecture in Europe at least into the 1960s. He died in 1974.

1934 Chrysler Airflow Imperial Custom CW front 3q © 1990 Richard Spiegelman (used with permission)
Although it sports a 1935 grille and hood, this CW Airflow Imperial Custom is actually a 1934 model, apparently built for the late Huntington Hartford, heir to the A & P supermarket chain. Chrysler offered a kit to allow dealers to install the revised grille and hood on earlier Airflows, presumably to help unload leftover 1934 models, although some first-year buyers probably requested such conversions. Converted or not, CWs are extremely rare; total production has been estimated at around 107 units. (Photo: “1934 Custom Imperial Airflow limousine” © 1990 Richard Spiegelman; used with permission)


Despite Walter Chrysler’s faith in the Airflow concept, sales remained sluggish for the rest of 1934. Total DeSoto production was 13,940, down a disheartening 31% from 1933. We’ve seen at least four different totals for the Chrysler Airflows, but the most commonly quoted figure is 11,292, fewer than 200 of which were the big CX and CW Imperials. Given the Airflow’s likely production costs — which, to our knowledge, Chrysler has never released — that can’t have been a profitable proposition. Fortunately, the strength of the other divisions, particularly Plymouth, kept the company out of the red; in fact, Chrysler posted a $9.5 million profit for the 1934 calendar year.

1935 Chrysler Airstream coupe 2-4 front 3q © 2006 Jack Whitaker (used with permission)
Except for a rare long-wheelbase version of the CZ Deluxe Eight series, 1935 Chrysler Airstreams were shorter, narrower, lower, and up to half a ton (455 kg) lighter than comparable Airflows. More importantly, they were at least $200 cheaper, model for model. Nearly 60% of 1935 Chrysler buyers opted for the less-expensive C6 Airstream Six. CZ Airstream Eights outsold Airflows, but only by about 1,500 units. (Photo: “1935 Chrysler Coupe” © 2006 Jack Whitaker; used with permission)

To placate their sales organization and unhappy dealers, DeSoto and Chrysler added new conventional models for 1935, dubbed Airstream. Sometimes credited to Raymond Dietrich (whom Walter Chrysler had hired in 1932), the Airstreams were actually designed by Phil Wright of Briggs, who had previously designed the 1933 Silver Arrow for Pierce-Arrow. Wright’s designs were done on spec; Briggs chief Ralph Roberts assembled them into an attractive bound volume, which he presented to Carl Breer. Breer’s own account suggests a certain bitterness toward the Briggs proposal, but it arrived at an opportune moment and Chrysler agreed to put the designs into production with almost no changes. The Airstreams had some streamlined design cues, but other than the corporation’s new independent front suspension system, they were largely conventional in both concept and engineering.

1935 Chrysler Airflow front 3q Bill Burris 2007 CCBYSA20
1935 Chrysler Airflow rear 3q Bill Burris 2007 CCBYSA20
Notable features of the facelifted 1935 Chrysler Airflow included a squared-off hood, a new stamped-steel grille (with a combination of die-cast and stainless steel trim), one-piece bumpers, revised cooling louvers and parking lamps, and a new hood ornament. The short-wheelbase C-1 traded the CU’s 299 cu. in. (4,894 cc) engine for the 324 cu. in. (5,302 cc) engine used in the bigger C-2 and C-3 Imperials. With a single-throat Stromberg carburetor and 6.2:1 compression, it made 115 gross horsepower (86 kW), down 7 hp (5 kW) from its smaller-displacement predecessor, but 15 lb-ft (20 N-m) more torque, now 240 lb-ft (324 N-m) at 1,200 rpm. (Front 3q and rear 3q photos above (both of which are entitled “1935 Chrysler Airflow”) © 2007 Bill Burris; both used under a Creative Commons Attribution-ShareAlike 2.0 Generic license)

The 1935 Airflows, introduced at the same time, sported a mild facelift, including reshaped hoods, sturdier bumpers, and new grilles based on those of the Airstreams. Although industrial designer Norman Bel Geddes, a Chrysler consultant in this period, had offered a proposal for future Airflows, Chrysler apparently used very little of it. The facelifts were primarily the work of Ray Dietrich, some of his first Chrysler designs to each production. Dietrich would also do the facelifts for the 1936 and 1937 models.

At the beginning of the year, most Airflows carried even higher prices than in 1934. The new Model SG DeSotos were up a painful $200 across the line. That proved to be a serious miscalculation, and by mid-year, price cuts left the base Chrysler Airflow Eight (now called C-1 rather than CU) a full $100 cheaper than the previous year’s cars. Nonetheless, the Airflows remained significantly more expensive than their Airstream cousins, with a predictable effect on sales.

1935 DeSoto Airflow sedan front 3q © 2008 Bill McChesney (CC BY 2.0 Generic)
1935 DeSoto Airflow sedan dashboard © 2008 Bill McChesney (CC BY 2.0 Generic)
The 1935 DeSoto SG Airflow had an attractive new hood and grille, along with revised bumpers, different cooling louvers, and a new dashboard. Unfortunately, its $1,195 price tag — cut to a still-hefty $1,015 during the model year — pitted it against eight-cylinder rivals of greater prestige. Nearly 75% of 1935 DeSoto buyers opted for the much cheaper Model SS Airstream Six. (Front 3q photo (“2445 1935 DeSoto Airflow”) and dashboard photo (“2453 1935 DeSoto Airflow”) above © 2008 Bill McChesney; both used under a Creative Commons Attribution 2.0 Generic license)

Carl Breer and other defenders maintained that the Airflow was simply ahead of its time, and it soon appeared they were correct. At the Chicago World’s Fair in the summer of 1934, Briggs showed off a mockup of a sleek, rear-engined prototype developed by John Tjaarda, which was already in development as the Lincoln Zephyr. A few months after the Briggs car appeared, Bendix unveiled the one-off SWC, a streamlined, front-wheel-drive fastback sedan intended to show off the full range of Bendix automotive products. The SWC looked a great deal like the Airflow; its designer, William Ortwig, later said the resemblance was coincidental, but the Bendix car did use an off-the-shelf 1934 DeSoto grille and headlights.

All of those were concept cars, but in the spring of 1935, Volvo launched its streamlined PV36 Carioca, followed that October by the aerodynamic Peugeot 402. Both the Peugeot and the Carioca looked broadly similar to the Airflow, although they were smaller and had conventional body-on-frame construction. While Volvo corporate historian Claes Rydholm denies that the PV36, designed by former Hupp engineer Ivan Örnberg, was based on the Airflow, author Vincent Curcio says Peugeot actually licensed the Chrysler design for the 402. We were unable to confirm that assertion, but it seems unlikely, although Peugeot was certainly aware of the Airflow (and the Bendix SWC, which was demonstrated to Peugeot engineers in the fall of 1934). Some sources indicate that Peugeot did license the Jaray patent for the 402 and its derivatives, which seems more plausible; we have no information about whether Volvo did so or not. Toyota’s first automobile, the 1936 Toyoda Model AA, was definitely Airflow-inspired.

Peugeot 402 sedan front © 2006 Stahlkocher (CC BY-SA 3.0 Unported)
Although the Peugeot 402 bears a general resemblance to the Airflow, it differs significantly in both proportions and detail. Among other things, the Peugeot mounts its headlights behind the grille rather than in the nose, it has no running boards, and both its front and rear doors (which are not interchangeable) hinge at the B-pillars. We find it more attractive than the Airflow, and it was certainly more popular, selling something like 75,000 units in various iterations before the fall of France in 1940. (Photo: “Peugeot 402 grey vl” © 2006 Stahlkocher; resized 2014 by Aaron Severson and used under a Creative Commons Attribution-ShareAlike 3.0 Unported license)

Although the Airflow no longer looked quite so unusual, sales continued to sink. Thanks to the Airstream and the popular all-new ’35 Plymouths — now featuring Airflow-style Floating Ride — Chrysler posted a record $35 million profit for 1935, but Airflow production fell to only 7,751 Chryslers and 6,797 DeSotos.


The Airflows were facelifted again for 1936, with revised grilles and all-steel roofs. Sedans now had integral trunks with internal spare-tire storage. The interiors were also revised, eliminating some of the previous Streamline Moderne touches, like the chromed seat rails.

1936 Chrysler Airflow Imperial C10 sedan front
The die-cast ‘fencer’s mask’ grille of the 1936 Chrysler Airflow was again based on that of the contemporary Airstream models, although its actual shape was quite different. All 1936 Chrysler Airflows except the CW had the 324 cu. in. (5,302 cc) eight. In the short-wheelbase C9, it had a single-throat carburetor, making 115 hp (86 kW) with the standard iron cylinder head or 120 hp (89 kW) with the optional aluminum head. The C-10 Airflow Imperial and C-11 Custom Imperial both had the aluminum head and a two-throat carburetor, giving 130 hp (97 kW) with 6.5:1 compression or 138 hp (103 kW) with the optional 7.45:1 compression ratio. Even with the low-compression engine, the Imperial had significantly more power and torque than the Zephyr’s sometimes troublesome 267 cu. in. (4,380 cc) V-12. (Author photo)

Reflecting the slow sales, the model lineup was simplified, as well, eliminating the business coupes and the Town Sedans, with their blind quarter panels. The big CW Airflow Custom Imperial was no longer listed, although about 10 were built as special orders.

Prices of both the basic Chrysler C-9 Airflow Eight and DeSoto Model S2 “Airflow II” were up as much as $150 for 1936, which put them in a very awkward market position. The six-cylinder DeSoto cost as much as a Packard One Twenty, which had a straight eight, 20 more horsepower (15 kW more), and the snob appeal of the Packard name. The Chrysler Airflow, meanwhile, competed directly with the new Lincoln Zephyr, which had a V-12 engine, a superior power-to-weight ratio, and styling that had been much better received. Interestingly, most 1936 Chrysler Airflow buyers opted for the pricier C-10 Imperial. Starting at $1,475, the C-10 was more expensive than the Zephyr, but noticeably larger and considerably more powerful.

1936 Chrysler Airflow Imperial C10 sedan side
Despite its 128-inch (3,251mm) wheelbase and 215.9-inch (5,483mm) overall length, the C-10 Airflow Imperial was identical to the smaller Chrysler C-9 Airflow from the B-pillars to the front bumpers. The C-10’s greater length was achieved by welding in an extra section around the rear door area; the center bar of the C-10’s split rear quarter window marks the trailing edge of that insert. Except for the vent windows, each of the C-10 sedan’s front doors was effectively interchangeable with the rear door on the opposite side. The same was true of the shorter-wheelbase C-9 and DeSoto sedans, but on those cars, the lower edge of each rear door was curved to clear the rear wheelhouse. (Author photo)
1936 DeSoto Airflow sedan front 3q © 2009 Bill McChesney (CC BY 2.0 Generic)
This was the final year for the DeSoto Airflow, formally known as the Model S2 Airflow II. The new grille doesn’t look dramatically different from the 1935 model’s, but, like the grille of the 1936 Chrysler Airflows, it was now a one-piece die-cast unit. Sedans accounted for most of the S2’s 5,000 sales; only 250 DeSoto coupes were built for 1936. (Photo: “1746 1936 DeSoto AirFlow FOR SALE” © 2009 Bill McChesney; used under a Creative Commons Attribution 2.0 Generic license)

While Lincoln sold around 15,000 Zephyrs and Packard more than 55,000 One Twenties in 1936, Airflow production amounted to only 6,285 Chryslers and 5,000 DeSotos. It was the sole black mark in what was otherwise a spectacular year for Chrysler: Plymouth sales topped half a million units and even Chrysler sold nearly 60,000 cars, thanks mainly to the handsomely facelifted 1936 Airstreams. Chrysler’s total production surpassed Ford’s for the first time and the corporation posted a net profit of $62.1 million, enough to repay the last of the debt Chrysler had assumed when it acquired Dodge in 1928.

The dismal sales of the Airflow II were apparently the last straw for DeSoto president Byron Foy, and the DeSoto Airflow disappeared after 1936. The writing was on the wall for the Chrysler version as well, but it returned for one final encore, offered only in a single series. Since the C-10 Imperial had been 1936’s best seller, the 1937 C-17 Airflow was essentially a repeat of that model, fitted with new trim and yet another new hood and grille, similar to those of Chrysler’s conventionally styled cars. However, the C-17 was no longer listed as an Imperial. For 1937, Chrysler had applied the Imperial name to all of its non-Airflow eight-cylinder cars, probably as a belated response to the Packard One Twenty. There was now a separate, conventionally styled Custom Imperial, although at least two CW Custom Imperials were updated with 1937 Airflow grilles, hoods, and bumpers. One was for Philippines president Manuel Quezón, while the other was built for Edward Bowes, host of the popular CBS radio show Major Bowes Amateur Hour, of which Chrysler had recently become the sponsor.

1937 Chrysler Airflow front 3q © 2010 dave_7 (CC BY 2.0 Generic)
The revised front-end styling of the final C-17 Airflow was again modeled on that of the 1937 standard cars. All were developed under Ray Dietrich, who had become Chrysler’s chief stylist in 1935. The leading edge of the C-17’s hood prow extends slightly ahead of the grille, a sharp contrast with the sloping hood of the 1934 Airflow. C-17s again had a 324 cu. in. (5,302 cc) engine with either 130 or 138 gross horsepower (97 or 103 kW), although compression ratios were reduced to 6.2 and 6.5:1, respectively. (Photo: “1937 Chrysler Airflow” © 2010 dave_7; used under a Creative Commons Attribution 2.0 Generic license)
1937 Chrysler C17 Airflow coupe side © 2011 Ronnie Schreiber WPC (used with permission)
A very rare C-17 Airflow coupe, photographed in the basement of the Walter P. Chrysler Museum. According to Chrysler Archives, only 230 coupes were built for 1937 — the last two-door Airflows. (Photo © 2011 Ronnie Schreiber; used with the permission of Cars in Depth and the Walter P. Chrysler Museum)

Sales of the C-17 were very similar to the previous year’s C-10 Airflow Imperial: exactly 4,600 units, 95% of which were four-door sedans. The last was built in October 1937, bringing the final tally of Airflow production to 55,655 cars (or fewer, depending on which 1934 figures you believe).

With that, the Airflow was dead, although Carl Breer’s group continued to use earlier models for aerodynamic research through at least 1941. The radial-engined, FWD mini-Airflow never got off the ground despite extensive development work and road testing. Chrysler tried to find a foreign licensee for the design in 1938, but with war brewing in Europe, they were unsuccessful. At least one of the two prototypes was eventually scrapped.


Conventional wisdom holds that the Airflow failed commercially because it was too advanced — that American buyers were simply not ready for functionally (as opposed to cosmetically) streamlined cars. However, even if we discount the popularity of the Peugeot 402 and its smaller 302 and 202 derivatives, the problem with that theory is the Lincoln Zephyr. The Zephyr’s exterior design and bridge-and-truss construction were just as radical as the Airflow’s, it certainly wasn’t any cheaper, and it had a variety of significant flaws. Nonetheless, it sold around 133,000 units between 1938 and 1942, and spawned the first Lincoln Continental. Clearly, buyers in that price range were not wholly opposed to aerodynamic design.

1939 Dodge Airflow tanker front 3q © 2010 Richard Spiegelman (used with permission)
Chrysler never offered Plymouth or Dodge Airflow cars, but they did apply the Airflow name to a series of large Dodge commercial trucks. The initial Model K-52 Special, introduced in December 1934, was a four-ton chassis-cab, powered by a 310 cu. in. (5,074 cc) six with 95 hp (71 kW). Despite the Airflow-style grille, its engineering was quite conventional. The later RX70 and RX71 models, introduced in December 1937, had a bored-out 331 cu. in. (5,430 cc) engine with 100 gross horsepower (75 kW). Many Dodge Airflows were used as tankers, including this restored Texaco truck, now part of the Walter P. Chrysler Museum collection. The museum estimates that only 261 Airflow-series trucks were built before production ended in early 1940, and very few survive today. (Photo: “1939 Dodge Airflow Texaco tanker truck” © 2010 Richard Spiegelman; used with permission)

One could argue that the Airflow paved the way for the Zephyr — arriving 17 months after the Airflow’s debut, the Zephyr was far less shocking than it might otherwise have been. On the other hand, the Zephyr demonstrated that an aerodynamic car did not necessarily have to be an ungraceful one. A nicely restyled, all-new Airflow, using the 128-inch (3,251mm) wheelbase of the C-10/C-17, might have done quite well as a Zephyr rival, although whether such a car would have sold well enough to justify its tooling costs is another matter.

Even if the Airflow had been as pretty as a Pierce Silver Arrow, its price would have made for an uphill battle in the middle-class market. The Airflow had definite strengths, but asking Depression-era buyers to pay a 25% premium for those advantages was a stretch. The high prices, of course, had less to do with its controversial aerodynamic styling than with its bridge-and-truss unitized construction, which was expensive to manufacture despite shortcuts like the interchangeable doors.

Chrysler could have offered a similarly aerodynamic shape with body-on-frame construction, as Peugeot did with the 402, and the Airflow’s Floating Ride certainly could have been applied to more conventionally engineered cars, as both Chrysler and its rivals soon did. However, Walter Chrysler was determined to make the Airflow a technological tour de force, which pushed its costs and prices beyond what the market was prepared to absorb.

We presume Chrysler understood the risks, but he taken bold gambles before and they had often paid off handsomely for him. When he acquired the Dodge Brothers company in 1928, for example, some observers had been incredulous because Dodge was so much bigger than Chrysler, but the deal proved to be a great success, significantly expanding Chrysler’s dealer network and production capacity for a very modest capital investment. If the Airflow had been a hit, it would have left rivals scrambling to catch up, giving Chrysler a commanding commercial advantage.

1936 Chrysler Murray-Ohio Airflow toy Eyes on Design © 2007 Patrick McLaughlin (used with permission)
The Murray Ohio Manufacturing Company, best known for its children’s bicycles, made this Airflow pedal car in 1936, although its grille represents the 1935 model. This particular toy, seen at the Detroit Institute of Ophthalmology’s 2007 EyesOnDesign Show, is part of the Rastall Collection. (Photo © 2007 Patrick McLaughlin; used with permission)

Many of the Airflow’s advances did eventually become common practice, but its dismal sales spoiled the corporation’s appetite for novelty for more than a decade. Had Walter Chrysler remained at the helm, that might not have been the case, but Chrysler ceded the presidency to K.T. Keller in 1935. Although Chrysler remained chairman, his direct involvement was greatly limited after he suffered a severe stroke in May 1938. Keller was far more conservative than Chrysler, and in any case the Airflow had left division managers extremely wary of anything too radical or too new. It was not until the fifties that the Chrysler Corporation began to reclaim its reputation for sophisticated engineering, and even then (with some notable exceptions, like the limited-production Dodge Charger Daytona and Plymouth Road Runner Superbird) the company tended to shy away from conspicuously aerodynamic designs well into the 1990s.

Walter P. Chrysler died on August 18, 1940, but the Three Musketeers remained with the company for another decade. Carl Breer was the first to retire, stepping down as head of research in 1949, but he remained a consultant and board member until 1953. Owen Skelton retired in 1951, although he too sat on the board of directors for a few more years. Fred Zeder never retired; he remained vice president of engineering and vice chairman until he died in 1951. His younger brother James, who had joined the company back in 1924, later became Chrysler’s chief engineer.

1936 Chrysler Airflow Imperial C10 sedan side dave_7 2008 CCBY-SA20Gen
The battered remains of what was once a 1936 C-10 Airflow Imperial, photographed in Wetaskiwin, Alberta, Canada in 2008. (Photo: “1936 Imperial Airflow Sedan C10” © 2008 dave_7; used under a Creative Commons Attribution-ShareAlike 2.0 Generic license)

Given all the things Walter Chrysler and the Three Musketeers accomplished during the corporation’s first 15 years, it’s ironic that their best-known car is one of their few serious missteps. On the other hand, there are far worse things to be remembered for than a daring, ambitious, forward-thinking failure like the Airflow.



Special thanks are due to George Camp, Pat McLaughlin, Ronnie Schreiber of Cars in Depth, and Richard Spiegelman for the use of their photos and Danielle Szostak-Viers of the Chrysler Historical Collection (now FCA US LLC – Historical Services) for her assistance with historical images and information.


Our sources on the Airflow included “1938 Dodge RX 70 (Airflow Design Series): Legendary Rarities,” Dodge Legends Exhibit Vehicle Overview, Chrysler Heritage (no date, chryslerheritage.iconicweb. com, accessed 29 May 2011); The Airflow Club of America, The Airflow Club of America Quick Reference Guide, airflowclub. connect2websites. com, accessed 15 May 2011; the Auto Editors of Consumer Guide, “1934-1937 Chrysler/DeSoto Airflow” (18 October 2007,, auto.howstuffworks. com/ 1934-1937-chrysler-desoto-airflow.htm, accessed 14 May 2011); the Auto Editors of Consumer Guide, Encyclopedia of American Cars: Over 65 Years of Automotive History (Lincolnwood, IL: Publications International, 1996); Jim Benjaminson, “Chrysler at the 1934 Chicago World’s Fair,” Plymouth Bulletin, reprinted with permission at Allpar, www.allpar. com, accessed 28 May 2011; Carl Breer and Anthony J. Yanik, The Birth of Chrysler Corporation and Its Engineering Legacy, Second Edition (Warrendale, PA: Society of Automotive Engineers, Inc., 1960, 1995); Dr. David George Briant, “Chrysler Corporation’s Pulsating Years: 1926-1938” (25 December 2008, WPC Club, Inc, www.chryslerclub. org/ walterp .html, accessed 12 May 2011); Arch Brown, “Chrysler’s Magnificent Mistake: 1934 Airflow ‘CU,'” Cars & Parts, August 1992, pp. 22-26; Arch Brown, “SIA comparisonReport: Battle of the Streamliners: Chrysler Airflow vs. Lincoln Zephyr,” Special Interest Autos #120 (November-December 1990), reprinted in The Hemmings Book of Lincolns (Hemmings Motor News Collector-Car Books), ed. Terry Ehrich (Bennington, VT: Hemmings Motor News, 2002), pp. 12-20, and The Hemmings Book of Pre-War Chryslers: driveReports from Special Interest Autos magazine, ed. Richard A. Lentinello (Bennington, VT: Hemmings Motor News, 2002), pp. 60-67; Don Butler, “Adventures in Airflow, Part II” Cars & Parts January 1981, pp. 20-29; “Chrysler Airflow- aero-modded (Part-1)” (12 June 2010,, ecomodder. com/ forum/ showthread.php/ chrysler-airflow-aero-modded- part-1-a-13538.html, accessed 24 May 2011); “De Soto Airflow Series SE 1934 United States” (no date, Classic Carbase, www.classiccarbase. com, accessed 23 May 2011); Craig Fitzgerald, “Winds of Change,” Hemmings Sports & Exotic Car Vol. 4, No. 12 (August 2009), p. 8; Vincent Curcio, Chrysler: The Life and Times of an Automotive Genius (New York: Oxford University Press, 2001); James M. Flammang and the Auto Editors of Consumer Guide, Chrysler Chronicle: An Illustrated History of Chrysler – DeSoto – Dodge – Eagle – Imperial – Jeep – Plymouth (Lincolnwood, IL: Publications International Ltd., 1998); Dave Holls and Michael Lamm, A Century of Automotive Style: 100 Years of American Car Design (Stockton, CA: Lamm-Morada Publishing Co. Inc., 1997); Charles K. Hyde, Riding the Roller Coaster: A History of the Chrysler Corporation (Great Lakes Books) (Chicago, IL: Wayne State University Press, 2003); Howard S. Irwin, “The History of the Airflow Car,” Scientific American August 1977 (Vol. 237, No. 2), pp. 95-106; Michael Lamm, “Magnificent Turkey,” Special Interest Autos #16, April-May 1973, reprinted in The Hemmings Book of Pre-War Chryslers, pp. 34-41, and “Airflow Prototypes” in pp. 18–20 of the original issue; Frank Marcus, “Mythbusted! At long last, science answers the question: Do the 1959 Chevy’s gullwing fins produce lift?” Motor Trend Classic #7 (September-October 2006), pp. 52-57; Richard Millman (“Bill-W”) on a thread posted on the Antique Auto Club of America forum 18 December 2009 to 20 June 2010, forums.aaca. org/ f147/ 1934-chrysler-desoto-airflow-door- same-273970.html, accessed 20 May 2011; “Mopar’s Star Cars: Were these two FWD, five-cylinder experimentals Chrysler’s answer to GM’s X-Cars?” Special Interest Autos #10 (April-May 1972), reprinted in Hemmings Classic Car #80 (May 2011), pp. 50-54; Gerald Perschbacher, “Walter Chrysler defended the Airflow in 1935,” Old Cars Weekly 22 April 2004, p. 12; Joel Prescott, “Like an Airplane on Wheels,” Car Collector July 1993, pp. 6-11; “Three Musketeers – ZSB” (2009, Walter P. Chrysler Museum, wpchryslermuseum. org, accessed 26 May 2011); Bruce R. Thomas, “Trifon Special: Birth of a Classic,” originally published in TORQUE (the publication of the Classic Car Club of America – Michigan Region) and reprinted with permission in Antique Automobile September-October 1971, nn (3 pages); Dr. David Zatz, et al, “Chronological history of Chrysler Corporation, Dodge, and Plymouth” (no date, Allpar, www.allpar. com, accessed 12 May 2011); a 1934 Chrysler dealer briefing on the new Imperial Custom line (Chrysler Bulletin No. 899), 24 May 1934, excerpted in WPC News Vol. 28, No. 7 (March 1997). Additional details came from the Online Imperial Club website (www.imperialclub. com, last accessed 29 May 2011). Some production figures were supplied by Danielle Szostak-Viers of the Chrysler Historical Collection (now FCA US LLC – Historical Services) in emails to the author 13 May and 16 May 2011.

Information on Chrysler’s non-Airflow cars of this period came from Jim Benjaminson, “New from the ground up: Plymouth for 1935 and 1936,” Plymouth Bulletin, and “Now for the Second Million: Plymouth cars of 1934,” Plymouth Bulletin, reprinted with permission on Allpar, www.allpar. com, accessed 21–22 May 2011; Terry Boyce, “1937 Chrysler Imperial: American Art Deco,” Special Interest Autos #84 (November-December 1984), reprinted in The Hemmings Book of Pre-War Chryslers, pp. 74-81; Arch Brown and Bud Juneau, “1935 Chrysler Airstream: Ray Dietrich to the Rescue,” Special Interest Autos #168 (November-December 1998), The Hemmings Book of Pre-War Chryslers, pp. 42-48, 57-59; Arch Brown, “1935 Plymouth: ‘A New Picture of Car Value,'” Special Interest Autos #172 (July-August 1999), reprinted in The Hemmings Book of Plymouths: driveReports from Hemmings Special Interest Autos magazine, ed. Terry Ehrich(Bennington, VT: Hemmings Motor News, 2002), pp. 4-21; Jeffrey I. Godshall, “1938 Chrysler Imperial: Chrysler’s Last Convertible Sedan,” Special Interest Autos #132 (November-December 1992), reprinted in The Hemmings Book of Pre-War Chryslers, pp. 82-91; Michael Lamm, “1931 Chrysler 6,” Special Interest Autos #40 (May-July 1977), reprinted in The Hemmings Book of Pre-War Chryslers, pp. 20-25; and Ross MacLean, “drive report: A 1936 Chrysler Airstream,” Special Interest Autos #4 (March-April 1971), reprinted in The Hemmings Book of Pre-War Chryslers, pp. 68-73.

Additional information on other early streamliners came from “1924 Rumpler Tropfen-Auto RU 4A 106 in ‘Metropolis, 1926” (30 May 2006 to 10 February 2008,, accessed 18 May 2011); “1932 Bergholt Streamline at the Concours d’Elegance of America at Meadow Brook” (2010,, www.conceptcarz. com, accessed 26 May 2011); “Albanita!” Special Interest Autos #15 (February-March 1973), pp. 50-53; the Auto Editors of Consumer Guide, “1923-1931 Lancia Lambda” (2007,, auto.howstuffworks. com/ 1923-1931-lancia-lambda.htm, accessed 21 May 2011), and the “1935-1942 Peugeot 402” (no date,, auto.howstuffworks. com/ 1935-1940-peugeot-402.htm, accessed 22 May 2011); “Benz Tropfenwagen” (no date, Grand Prix History, www.grandprixhistory. org/trop.htm, accessed 10 October 2020); “Founding Fathers of the Automotive Industry: Paul Jaray” (n.d., Unique Cars and Parts, www. uniquecarsandparts., accessed 14 May 2011); Kevin Guthrie, “The Benz Tropfenwagen and Alfa Romeo 512: Rear Engined Grand Prix Cars” (30 October 2008, – Auto Racing, www.suite101. com, accessed 18 May 2011); Neal Jacquot, “Volvo 1935 Carioca not a Copy of the Airflow!” Airflow Newsletter, Airflow Club of America, Vol. 49, No. 6 (June 2010), pp. 3-4, which refers to an article by Volvo corporate historian Claes Rydholm in Rolling Magazine: the publication of the Volvo Club of America, March-April 2010, pp. 18-23; Paul Jaray, “Motor Car,” U.S. Patent No. 1,631,269, applied 19 August 1922, issued 7 June 1927; David LaChance, “The Internationalist,” Hemmings Classic Car October 2008; Michael Lamm, “1939 Lincoln Zephyr,” Special Interest Autos #6 (July-August 1971), reprinted in The Hemmings Book of Lincolns, pp. 22-27; Michael Lamm, “Two Look-Alikes: Ford & Citroën,” Special Interest Autos #9 (January-March 1972), reprinted in The Hemmings Book of Prewar Fords: driveReports from Special Interest Autos magazine, ed. Terry Ehrich (Bennington, VT: Hemmings Motor News, 2001), pp. 44-51; Vincenzo Lancia, “Automobile,” United States Patent No. 1,372,148, applied 21 November 1919, issued 22 March 1921; Pete Leatherwood, “Mr. Bendix and His Car” [podcast] (27 March 2006, Studebaker National Museum, studebakermuseum.blogspot. com/ 2006/ 03/ mr-bendix-and-his-car.html, accessed 27 May 2011); Alex Oagana, “The Origins of Streamline Design in Cars” (5 January 2009,, www.autoevolution. com, accessed 23 May 2011); “Rumpler, Edmund” (7 August 2005, The Androom Archives, www.xs4all. nl/ ~androom/ biography/ p012525.htm, accessed 18 May 2011); Guillaume de Syon, “The Teardrop That Fell From the Sky: Paul Jaray and Automotive Aerodynamics,” ITEA Journal No. 29(1), March 2008, pp. 14-16; Toyota Motor Corporation, 75 Years of Toyota, “Vehicle Lineage: Toyoda Model AA Sedan (1st),” 2012, com, accessed 15 September 2014; “Volvo PV36 Carioca (1935)” (no date,, www.netcarshow. com; accessed 28 May 2011); and the Wikipedia® entries for the Peugeot 402 (, accessed 22 May 2011), Edmund Rumpler (, accessed 18 May 2011), the Rumpler Tropfenwagen (, accessed 18 May 2011), and the Tatra 77 (, accessed 14 May 2011). Thanks to Pat McLaughlin for bringing the Bendix SWC and Bergholt Streamline to our attention.

Some history of the Chrysler Building came from Kenneth L. Massey, “History of Walter P. Chrysler and the Chrysler Building” (no date, Allpar, allpar. com, accessed 11 May 2011).



Add a Comment
  1. A wonderful article. I still have some very vivid memories of the Pennsylvania AACA vintage car shows back in the 1970’s when a couple of Chrysler Airflows would show up. And definitely got noticed, unlike my personal ’37 Buick Special.

    Now, how about some follow-up! I’d love to see an article on the Lincoln Zephyr, and would especially love to see something on the Hupp Aerodynamic’s of 1934-35. The 1937 Ford wouldn’t be out of line, either. The mid-30’s were definitely a wonderful time for automotive design.

    1. I would love to do the Zephyr and the original Continental, if I can put together enough pictures. (The fact that I was unable to find a ’36 or ’37 Zephyr photo to include in this article, as a contrast with the Airflow, tells you how many I have right now..!)

  2. I’d like to second the earlier commenter, the mid ’30s to early ’40s period really was a fascinating early golden age in automotive styling, where for the first time styling really was given a high priority over the more utilitarian looks of the 20s-early 30s, even if the customers of the day didn’t always go for the more radical examples of streamlined Art Deco, as was the case with the Airflow and shark nose Grahams.

    Although I appreciate the need to jump around a bit for variety, please do consider more articles on the ’30s streamliners in the future, as it is such a fertile era. – the Cord 810/812 would be a great read.

    As an aside, I read an article a few years back about an architect or fashion designer (some creative professional) in Manhattan who had a really special custom CW restomod built, taking an original Airflow body and interior and retrofitting it with a modern drivetrain and undercarriage, apparently the owner loved the Art Deco style and was using it as a daily driver in the city. Wish I had saved the article, but as I recall, there weren’t many pictures with it, but definitely an interesting project.

    1. I’m not categorically opposed to restomods, but doing it with a car as rare as a CW Imperial seems…wrong. I don’t know if they’re on the list of capital-C Classics (if not, they ought to be), but they’re extremely rare, with a lot of unique components (dashboard, brakes, springs, driveshaft, wheels, windshield). The CW is so huge and so heavy that even with a modern drivetrain, it would be a handful in city traffic — it’s bigger and heavier than an Escalade ESV. Admittedly, if it was in sad shape to start with, restoring it would be a serious pain in the ass, but still…

  3. Very interesting as always Aaron, I had a passing knowledge of the Airflows but it is great to read about the detail – I can’t imagine more than a few current cars would warrant the same attention!

    I saw a 1934 Airflow last year (with flat camera batteries!) and I remember previously one of the later model ones with what was quite evidently a tacked-on traditional grille. Looking at the photos perhaps what made the car look so jarring is the horizontal line of the hood which gives the optical illusion that the nose of the car is higher than the cowl, and I find the 6cyl cars better looking because the shorter nose lessens that impact. On the point that aero design per se was not sales-proof perhaps a slight taper for a less bulbous appearance, and wider-set headlights, would have made a difference?

    Rob it would be interesting to hear more about that car too.

    1. John,

      I hadn’t thought about the nose seeming higher than the cowl, but looking at the photos again, I see what you mean. It’s exacerbated by the position of the hood ornament, which gives the tip of the nose a bit of a ‘ski jump’ flair, even on the ’34s, where the original hood line actually slopes downward quite a bit. I imagine that is one of the factors that contributes to what I think of as a ‘stubby’ look to the whole car (especially sedans). The 1934 CU is quite a big car, but its rounded contours make it look smaller than it actually is, an effect that only the really long wheelbases of the CX and CW Imperials offset to any great degree.

      Deciding to mount the headlights in the leading edges of the body sides, rather than the fenders, was really limiting, both in spacing and light size. I don’t know why they didn’t mount the lights in the front fenders, as John Tjaarda did with the Zephyr. (The production Zephyr’s front end was mostly redone by Bob Gregorie, but the fender-mounted lights were on the original Briggs prototype.) I think the headlights of the ’36-’38 Zephyrs still look a bit awkward, but their spacing makes the Lincoln look lower and wider than the Airflow, even though it really isn’t.

      I imagine these are all reflections of the fact that the Airflow was [i]engineered[/i] more than actually [i]designed[/i]. The role of the stylists appears to have been to decorate the shape given them by the engineers, in which they had little if any say.

  4. Very nice article, Mr. Severson! I’m the president of the Airflow Club of America and I happened to find this site while searching for some other material. It’s nice to see a recent article about the Airflow. We are a small (425+) member club. We just had our National Meet in Durango, Colorado and I drove my 1935 C2 (Imperial) from Seattle WA. There are not many 76 year old cars which cruise smoothly down the road at 75 mph in overdrive. I did the return trip (1,300 miles) in two days. I own a 35 Packard, a 37 Cord and numerous other cars of the era and none can match the “modern” ride and drive of an Airflow.

    You are right, Engineering called the shots when the Airflow was designed. It was truly radical. Besides being quiet and smooth at 75 mph, it has gobs of art deco everywhere.

    The 6 cylinder cars (all DeSotos) are true to the Airflow spirit, but after riding in a prototype, Chrysler wanted an Airflow with his name on it. The Imperial Coupes are very well proportioned with six extra inches behind the B pillar.

    I am not opposed to rods, but if a CW was rodded it would break my heart so don’t verify it, please!

    In 1934, Chrysler did offer a retrofit grill which replaced the pure “waterfall” look with a more traditional grill – as was found on the ’35 models. With each passing year, as sales never materialized, Chrysler tried to make the front end look more conventional.

    Visit the website at Check out the video of the Airflow being pushed off a cliff and driving away (in the Library section). If anyone would like to see an Airflow close up, let me know and I’ll link you up with the closest members.

    Thanks again for a nice article!

    1. Frank,

      Thanks for the kind words!

      I’ve seen the story that the Airflow was originally intended [i]only[/i] as a six-cylinder DeSoto in a number of secondary sources. Do you have an original source for that account? As it stands, I’m afraid I’m rather skeptical of it.

      It’s true that the 1932 Trifon Special prototype was a DeSoto-size, six-cylinder car, and Carl Breer does say that he and Oliver Clark started with a six-cylinder model in laying out the packaging for the Airflow; they wanted to determine the minimum dimensions necessary to achieve both the aerodynamic profile and minimum passenger space they were looking for. However, Breer makes no mention of intending to [i]only[/i] offer the Airflow as a six, and his account indicated that the decision about which brands would offer it was made later. All he says on that subject is that they decided it would be sold by Chrysler and DeSoto; he says nothing about any plan to market it as a DeSoto-only product.

      Breer does say that Walter P. Chrysler was very excited about introducing the Airflow to celebrate Chrysler’s 10th anniversary. Based on Breer’s description of WPC’s enthusiasm for the project, it’s hard to conceive Chrysler [i]not[/i] wanting an Airflow with his name on it — if somebody suggested otherwise, I’m not sure who or why. From a business standpoint, certainly, offering the Airflow only as a DeSoto six wouldn’t have made much sense. The development costs were undoubtedly high, and DeSoto’s annual volume was not; it hadn’t topped 40,000 units since before the Crash. Also, even if there were some original plan to only offer it as a six, why didn’t the U.S. Chrysler line get the CY?

      Now, given the antipathy some of the corporation’s management apparently had toward the Airflow project, it’s entirely possible that getting the individual presidents to accept it was quite a battle. I suppose it’s possible that DeSoto president Byron Foy was less opposed than his colleagues; I really don’t know. However, my suspicion is that if there was a debate over who would offer the Airflow, it was more a matter of internal resistance than any overarching plan of what the Airflow should be.

      It would certainly be fair to call the six the baseline Airflow, since the larger eight-cylinder cars were created by splicing additional sections into the body panels and frame, but I don’t know that that means the bigger cars were somehow an afterthought. From Breer’s account, I think it was just easier to start with the smaller version, to establish the minimum package dimensions. That makes sense — if you know you have acceptable passenger room on the shorter wheelbase, it’s easy to make it bigger, whereas if you base your engineering on the larger version, it’s harder to scale it down without compromising utility space.

      If someone can point me toward a primary source for the DeSoto-only story — first-person accounts of the meeting where it was decided, etc. — I’ll happily accept that, but otherwise, I’m inclined to think that story may just be a misinterpretation of Carl Breer’s account, one that has, as they say, grown in the telling.

      I wholly agree on the CW. I’m not keen on the idea of heavily customizing the Airflow to begin with — it’s relatively rare, of obvious historical interest, and already pretty wild looking without any help — and the idea of cutting up a CW is the sort of thing that gives historians night terrors!

  5. Aaron,
    In the text on page one, it states that the Rumpler Tropfenwagen was powered by a 2.5 W6. Is this a misprint of V6? Or was this a prelude to to todays W8 Volkswagen?

    1. Nope, that’s not a typo. The early Tropfenwagen had a 2,580 cc Siemens & Halske engine with three banks of two cylinders. I don’t think I’ve ever seen a picture of it, but I imagine it was rather bulky.

      1. while we are on the subject of the Tropfenwagen, Ferdinand Porsche had nothing to do with its development, as he was employed by Daimler, not Benz. The two companies formed an association in 1924, formally merged, forming Mercedes-Benz in 1926. Integrating the engineering staff of the two former rivals took a while, with the Benz faction eventually gaining ascendancy.

        1. Thanks — I looked back through my notes and saw how that error probably cropped up, but you’re right and I’ve amended the text.

  6. another expertly written and beautifully illustrated article on these avant garde art deco cars that were way ahead of their time in terms of both styling and engineering-please do an article on the history of De Soto

  7. I own a 1935 Chrysler Airstream c-6 Coupe.
    I’ve been able to determine thru internet research that there were 1975 like models made. What I can’t find is how many of them were standards and how many were deluxe. Does anyone know or know how I can find this info.


    1. Dennis,

      The information I have indicates that the six-cylinder Airstream C-6 models weren’t grouped into standard and deluxe versions, only the CZ Airstream Eight. Your number is what I show for business coupe production, as well.

      1. Thank you for the reply that’s interesting to hear, but I’m still confused. I’ve seen, what I believe are photo’s of some C-6’s with parking lights on top of the front fenders & horns mounted thru the stainless grill work on the fenders below the headlites on ea. side of the addition these cars have 2 windshield wipers. In addition to that I’ve seen photos of C-6’s with no parking lites on the fenders – grill work on the lower fenders
        with no horns & also a single wiper. I’m confused about the differences, could it be as simple as diff. options? What are your thoughts.

        1. The Deluxe models were introduced after the start of the 1935 model year. At least on CZs, the dual horns were initially standard and then were deleted on non-Deluxe Eights when the Deluxe was introduced; the horn layout you describe sound like the early and late non-Deluxe CZ arrangements. Chrysler’s factory production figures don’t distinguish between Deluxe and non-Deluxe CZs, probably because of the midyear change. Now, I don’t know anything about a Deluxe version of the six-cylinder cars — all of that applies to the Eights.

          Is it possible that some of the confusing pictures you’ve seen were actually mislabeled CZs? A quick image search just now revealed at least once set of photos labeled as C-6s that are pretty clearly late CZ Deluxes (they even have the winged "8" badges on the catwalks) and there was definitely that variation with the eight-cylinder cars. I could see the parking lamps being a dealer or owner add-on, but the horns do seem like a production variation.

          1. I realize this is a long shot but here goes. I inherited my fathers 1935 AirStream C6. It has spare tires in each fender, 2 windshield wipers, a luggage rack and the back of the passenger seat raises up like in a business coupe. Any help clarifying what series it is would be appreciated.

  8. Finally! The story of the Airflow in enough detail for me to understand these beautiful machines. Thank you for this piece!

    They may not have sold well, but their influence was certainly felt around the world. You mention the Volvo Carioca and the <i>Fuseau Sochaux</i> Peugeots (the last of which were made in 1949!), but there is also the first “proper” car from the land of the rising sun: the 1936 Toyota AA. None survived, so Toyota actually built one for their museum in the 1990s.

    Along with Cords, Marmons, Franklins, shark-nose Grahams and the Zephyr, these Chryslers are the epitome of 30s American cars for me.

    1. I must admit had always been a little skeptical of the Toyota link, but the 2012 edition of Toyota’s official history specifically says the Toyoda Model AA took its inspiration from the Airflow. It wasn’t just the styling either; the engineers at Toyoda (not spun off of the parent company as Toyota Motor until 1937) recognized and sought to emulate what Breer and crew had done with the Airflow’s weight distribution and ride. Very interesting.

      1. I should add that while the Model AA was the first Toyota automobile, it was NOT the first Japanese car. The first Japanese-built four-wheeled automobile was made in 1904 and the first gasoline-powered car was in 1907. However, people really didn’t have enough money to create a sustainable domestic passenger car industry until much later. According to the numbers I’ve seen, it wasn’t until 1951 that Japanese companies built more than 2,000 passenger cars in a single calendar year.

  9. There is a rather long promotional video of the Airflow made by Chrysler in the mid-1930s that discusses its design and features, and includes a rather dramatic test rollover with a live driver. The roof of the vehicle remains completely intact as do all windows, although one rear door came open (they lacked interlocking door latches that have been required since the late 1960s). If you have not seen this, you should get a copy. I may have a video version if you cannot find one elsewhere. This video provides dramatic evidence that manufacturers could have provided good rollover protection in more recent decades without the advanced materials now used to meet FMVSS 216, and saved thousands of lives in rollovers (particularly of SUVs).

    1. I’ve never seen the film, but I’ve heard about some of the stunts Chrysler did to demonstrate the Airflow’s rigidity and safety, including pushing one off a cliff. It was partly done to counter rumors that the Airflow wasn’t very strong, perhaps because people heard that its inner skeleton was not all that rigid by itself. (The bridge-and-truss skeleton wasn’t intended to be self-supporting, but once welded to the body panels, the whole thing was extremely stout.)

  10. Though only slightly connected to the article, is it known whether other more viable engines (e.g. inline, V-angle, Boxer, etc) were considered for the Chrysler Star Car in place of the five-cylinder radial-engine?

    It seems like a number of otherwise production worthy designs during that period up til the end of WW2 were fatally compromised by radial or other experimental engines.

    Also interested to know where one can find more information on Chrysler’s post-war small car project (including the name), which was intended to challenge the Chevrolet Cadet and the Ford Light Car project that eventually became the Vedette.

    1. To your first question, as far as I know, there were not. The “Star Cars” were FWD (if you look up the SIA article “MoPar’s Star Cars” on the Hemmings site, you’ll see some photos of the powertrain layout), which, like some early postwar FWD cars, was predicated on fitting the engine in a relatively short space ahead of the front axle. Short of the radial engine, that configuration would probably have dictated either an I-2, a V-4 or an H-4, with the latter probably being the most practical choice. Had Chrysler succeeded in finding overseas licensing interest, they might have gone that route, since for a non-experimental project, it’s hard to see a manufacturer considering the complexity of the five-cylinder radial worthwhile for an 1,100cc engine.

      The postwar concept you’re probably thinking of is Project A-106. I don’t have much detailed info on it; it was designed around an H-4 engine, but I don’t know the engine’s specifications beyond that.

      1. Indeed, the radial and other experimental engines considered by rivals were blind alleys.

        Agree about the production versions of the Star Car being powered by H-4 engines, though why didn’t Chrysler consider such an idea themselves instead of seeking out overseas licensing interests?

        Regarding Chrysler’s efforts in seeking overseas interest in the UK, know the likes of Stafford Gripps and others in the post-war Attlee wanted to establish a British people’s car project, supporting the stillborn efforts of Roy Fedden and Denis Kendall in setting up such their own car projects.

        Could an H-4 powered Star Car have remained in production after WW2 let alone achieve similar success to another similarly styled model whose engine layout / etc was at the rear?

        1. As far as I know, the Star Cars that were built were primarily experimental rather than development prototypes. With experimental designs, it’s not uncommon for a mule or set of mules to examine several different technologies at once, even if there isn’t any serious intent to build a production model that way. While an H-4 engine would probably have been a more promising design, it was no less radical for Chrysler in terms of actual production (having no relationship at all to any existing engine design), but less novel strictly from an engineering standpoint, offering less to be learned from experimentation. At that point, there wouldn’t have been a market for such a car in the U.S., at least in that form, and Chrysler didn’t have the export organization or local dealerships to offer it outside North America in meaningful quantities, especially given the complex tariffs and import duties involved in pre-EEC multinational productions.

          When the project began, Walter Chrysler was very enthusiastic about the Airflow and basically showing off what Chrysler engineering was capable of. Given the timing, my assumption is that the Star Cars amounted to dabbling in what they might do if they took some of the concepts involved in the Airflow a step or two further: FWD and a compact, easily removable powertrain. I can’t see Chrysler putting it into production in that form, but if the drivetrain and radial engine had shown promise AND if the Airflow had been a big hit, some of those concepts might have shown up in an Airflow successor. However, the Airflow was not a big hit, Chrysler management started getting nervous about these pie-in-the-sky engineering fancies and lost any interest in building such a thing. I think the effort to find a British licensee was at least partly a flailing attempt to recoup some of the substantial investment Chrysler had already made in the project rather than a coherent marketing strategy, which, combined with the timing, may have been why it didn’t pan out.

          I don’t know that comparative engineering strengths are a reliable predictor of commercial success. Some advanced designs fail despite their strengths, lacking other important characteristics like after-sale service or just being too costly to build. Some rudimentary or old-fashioned designs succeed because they’re reliable, familiar, or just well-marketed. So, it’s hard to judge potential success simply based on specifications without a sense of who a product would be marketed to, how well, and for what price.

          1. In the case of the Airflow, were there other more attractive styling proposals that Chrysler considered which could have worked in retrospect?

            Guess the following is admittingly a fantasy, it would have been interesting if a production worthy Star Car formed the basis for Chrysler’s UK division instead of them later acquiring Rootes.

          2. Well, as the article explains, the Airflow — and presumably the Star Cars — were primarily engineering designs, not styling ones. The role of Chrysler Art & Colour was to decorate the pre-approved shape (which they did pretty well, in my view), not to decide what it would look like; what today we’d probably call “detail styling.” Judging by the photos I’ve seen of the Airflow prototypes, the alternatives the engineers developed were arguably worse, although there’s a coupe design with a low-slung roof that would have pleased postwar customizers. Could actual stylists have come up with ways to apply the Airflow’s aerodynamic concepts in a more visually appealing way, and potentially on a smaller scale? Sure; the Peugeot was not an unattractive car for its day. Did Chrysler do that? Not so far as I know.

            Another significant consideration is that I remain convinced that part of the reason for the Airflow’s mediocre sales performance was that the Airflow cars were quite expensive. That’s an issue that would likely have affected the viability of something like the Star Car, especially vis-à-vis contemporary Ford of England offerings.

  11. What a first class article! Too bad the looks of the Airflow were so artless and indeed clumsy, and I say that as an engineer. Certainly the cars were expensive to produce, but a little exterior grace would perhaps have saved the sales day. You cover that aspect well in a reply comment above.

    On the little radial engine FWD project, I was unaware of it until now. So the radial multibank Sherman tank engine that Chrysler produced in WWII makes sense with its five 6 cylinder blocks. Not the same crank arrangement, but they knew it was a doable project and likely to succeed without protracted development time. Too bad for them that Ford produced a monster rival DOHC GAA V8 that eclipsed it. Surely the Ford had the most advanced cylinder head design for decades afterwards, a tour-de-force. Jags, Alfas, Mercedes and F1 engines did not have a narrow valve angle, direct-acting bucket tappets from the cam, and a central spark plug in a DOHC four-valve pentroof head, until Duckworth reinvented it for the FVA/DFV. Most engines have similar layouts today bar bucket tappets. The RR Merlin looks positively antediluvian by comparison.

  12. I’m surprised that the cars stayed so expensive for so long; once Chrysler realized that the development costs were going to be lost, one would have hoped that they could sell the cars at a little above cost, keep the volume up, and make the most of a bad situation. I’ve heard that a typical modern car wholesales for about double its manufacturing cost in the first year of production, so, assuming that the ratio was in the same neighborhood, there should have been room to bring the price at least inline with the Airstream models, which might have generated more volume. I’m not sure exactly why the production costs of such a unitized car were higher than for a body-on-frame design; maybe welding technology had a ways to go before they’d be on par. Certainly, unitized cars were competitive with body-on-frame designs by the mid 1950s, but I’m sure the manufacturing process was more streamlined by that time.
    One thing, other than styling, that seemed possibly premature on these models was the placement of the engine directly over the drive wheels; with no power steering I’d expect that would make these cars a hand-full to drive. Also, with skinny 1930s bias-ply tires, I wonder how the engine placement affected under-steer. The gradual puling-in of the front wheels that ultimately had them protruding into the passenger space by the late 1970s was made practical by the standardization of power steering, better suspensions and radial tires.

    1. The issue, vis-à-vis cost, is not simply the cost of materials and labor, but the cost of tooling and production facilities. If something can’t be build on existing assembly lines with existing tools and equipment, the cost of its special facilities has to be amortized over the run of the model or models that do use those facilities. (There are different ways to do that, but it has to be done somehow and factored into the overall P&L for a model line.)

      Part of the reason unitized cars were more expensive is that the nature of unit construction means more new tooling and/or tooling changes for each new model or model change. A lot of the focus of postwar development on unit bodies has been on finding ways to spread those costs around by creating ranges of different products that share certain big, costly subassemblies (such as the cowl/scuttle structure or the floorpan), but doing that requires time, planning, and a commitment to continue to build products in those ways.

      Shifting weight forward does unarguably promote understeer, although there were very, very few cars of that era that had much in the way of cornering ability, and “ease of handling” meant something rather different than “handling ability.” The Airflow cars had high steering ratios that made for lots of wheel-winding, not unlike postwar American cars without power steering. If you’ve ever driven an American car of the sixties without power steering, the steering typically isn’t that heavy (less so, often, than a European or Japanese subcompact of the seventies or eighties without power steering), but it does require a lot of cranking to change directions, which imparts a feeling of vagueness to the action. By thirties standards, that wasn’t necessarily considered a terrible thing, and slow steering that was not very reversible could be a relief on rough roads.

      1. Thanks for the quick moderation and reply! It brought back memories of learning to drive on my father’s 1975 Malibu, which despite being only “mid size” weighed 3,700 pounds without an automatic, power steering or power brakes. The lack of power steering, with 7 turns lock-to-lock, was the worst omission of the three. The car, shipped that way by accident (long story), was an interesting experience, but I’m glad they don’t make cars that size without power steering these days.
        PS: I didn’t say it the first time around in my haste to make my points as concisely as possible, but this was a great article. I’ve learned a lot from this site.

        1. Thanks!

          The lack of power steering is certainly a headache by modern standards, and it’s not difficult to understand why automakers generally compromised with non-power steering ratios requiring a bazillion turns lock-to-lock. Even something like an early Honda Civic, which weighed well under a ton dripping wet and had steering ratios giving something in the realm of four turns lock-to-lock, can be a real bicep-builder when trying to parallel park! Avoiding that was a significantly bigger concern for U.S. automakers than were quickness or precision of response, which in any case were not exactly strong points of the large majority of prewar cars. With a prewar sports car (or something like a Morgan that is functionally a prewar design regardless of actual date of manufacture), the combination of low ride height, relatively quick-ratio steering, minimal curb weight, and stiff springs made for more responsive handling, but you paid a heavy price in ride quality and steering weight, and for all the masochism involved, actual developed cornering power was still pretty modest. It’s no great surprise that the average American buyer’s response to that sort of thing was, “Ugh, no thanks.”

  13. “Jaray applied for a patent in August 1922 on automobiles with streamlined bodies. It covered essentially any car whose aerodynamic form comprised one partial teardrop shape (the greenhouse) atop another (the body itself).”

    How in hell could a patent office issue any type of license based upon such a vague, non-detailed, and overly generalized application. Chrysler should had taken Jaray’s patent infringement claim to court, as the Airflow did not resemble ‘two stacked partial teardrops‘ by any means.

    Guess no one thought to apply to the same patent office for a vehicle design patent license based upon ‘two stacked rectangular blocks’. With nearly every vehicle at the time having that similar design, every manufacture would had been sued for patent infringement.

    But then again, most engineers and designers at the time would not have had the same audacity as Jaray to submit a patent application with such a claim.

    1. With nearly every vehicle at the time having that similar design, every manufacture would had been sued for patent infringement.

      Look up the saga of the 1895 Selden patent and you’ll see this was pretty much par for the course (as it remains in various other fields today).

Leave a Reply

Your email address will not be published. Required fields are marked *

Comments may be moderated. Submitting a comment signifies your acceptance of our Comment Policy — please read it first! You must be at least 18 to comment. PLEASE DON'T SUBMIT COPYRIGHTED CONTENT YOU AREN'T AUTHORIZED TO USE!