01Executive summary
Three milestones, and an unusually clean case of a first design getting the architecture right.
Karl Benz built the first reliable internal-combustion automobile in 1885, after a decade building and selling small stationary gas engines. His three-wheeler carried a single-cylinder four-stroke engine with electric ignition, water cooling and a differential gear — three features present on essentially every car built since. He patented a rudimentary radiator in 1886. And in 1895 a patent was granted to George Selden, who never built a car and collected substantial royalties from those who did.
02Three features that never changed
It is unusual for a first attempt at a new class of machine to establish architecture that lasts. Benz's did, and the reasons are worth setting out individually because each solves a problem that is specific to putting an engine on a road vehicle rather than on a floor.
Electric ignition
Stationary gas engines used flame ignition, which is workable next to a gas supply and impossible on a moving vehicle exposed to wind and vibration. Electric ignition is original with Benz and is a direct consequence of the machine having to move.
Water cooling
An engine under sustained load must reject a great deal of heat. Water carries far more heat per unit volume than air and moves it to where it can be dissipated, which lets the cylinder run at a controlled temperature.
The differential gear
Driven wheels on a turning vehicle travel different distances. Rigidly coupled, one must slip, which destroys tyres and steering. The differential lets them turn at different speeds while both receive drive.
The radiator, 1886
Water cooling on a vehicle without an unlimited water supply requires cooling the water for reuse rather than discharging it. Patented the following year and now universal.
The first three features solve problems the designer could anticipate. The radiator solves one created by the first solution: adopting water cooling turned water into a consumable, and a vehicle cannot carry an unlimited consumable. The answer was to close the loop — reject the heat to air and reuse the water indefinitely. Recognising that your fix has created a new dependency, and closing the loop rather than sizing a bigger tank, is a general and repeatedly valuable move. It is the same reasoning as recycle in the ammonia plant of the second series and pressure recovery in desalination in the third.
The other point worth making is where Benz came from. He had a decade of experience building and selling small stationary gas engines before he built a car. The vehicle was not a leap from nothing; it was an existing product moved into a new application, with the changes that application demanded. Most successful first-of-type machines have this shape, and the ones presented as arriving from nowhere usually turn out not to have.
03The Selden patent: a case worth knowing
George Selden filed for a patent covering a road carriage powered by an engine of a particular type. The patent was not granted until 1895. He never built a car. He collected substantial royalties from many who did.
The case sits at the intersection of engineering and law, and engineers encounter its descendants constantly. Three features are worth naming. A long delay between filing and grant meant the patent issued into an industry that had developed without reference to it. The claim was broad enough to cover an architecture rather than a specific mechanism. And the holder had no manufacturing interest, so the patent functioned purely as a toll rather than as protection for an investment in production.
Whether that is a defensible use of the patent system is a question this series does not attempt to settle, and there are serious arguments on both sides about whether early disclosure of a concept deserves reward independent of reduction to practice. What is not in dispute is the engineering consequence: designers of the period had to work around a claim, and the resources spent on litigation and royalties were not spent on engineering.
This connects directly to the Pickard crank patent in the first series of this set, where Watt used an inferior sun-and-planet gear for over a decade rather than contest a patent on a device that had been in use for centuries. The pattern is the same and so is the cost: patent position shapes what gets built, and sometimes what gets built is worse.
04Takeaways for current practice
- A new application changes which requirements bind. Flame ignition is fine on a floor and impossible on a vehicle.
- Close the loop rather than sizing a bigger tank. Water cooling created a consumable; the radiator removed it.
- Check whether your solution created a new dependency. It usually has, and it is cheapest to find at design stage.
- First-of-type machines are usually adapted, not invented. Benz had ten years of stationary engines behind him.
- Patent position shapes what gets built. Sometimes the thing built is worse, and the resources go to lawyers rather than engineers.