I read a lot of "history of technology" books, and I'm used to the typical storyline: whereas we've been taught to expect that technological advances should come forth out of the results of fundamental research, in practice, it's most often the other way round. Engineers use inspired guesswork and empirical research to develop machines to solve real-world problems, and scientists are inspired by their success to investigate the underlying basic principles. The engineers then incorporate those into their design strategy for future iterations. (A familiar example is the interaction between heat engine technology and classical thermodynamics.)

Anderson's take on the relationship between theoretical aerodynamics and aircraft technology is rather different. The two disciplines seem to have developed in almost total isolation from each other. Mathematicians and theoretical physicists had sorted out the equations governing the flow of fluids by about the end of the 18th century, culminating in the work of Navier and Stokes in the 1850s, but no-one in the scientific establishment saw any real practical application for this work — the idea of a heavier-than-air flying machine was plainly ridiculous, and in any case the equations could only be solved mathematically for certain trivial cases.

At the same time, a series of largely self-taught enthusiasts — from Leonardo, George Cayley, and Otto Lilienthal to Samuel Langley and the Wright brothers — were doing practical development work to try to make a machine that could support itself by motion through the air, inspired by observations of real-world phenomena like the flight of birds, and either ignorant of or unable to make use of the work done by the theoreticians. Langley is often brought into histories of aviation as the representative of the big bad scientific establishment, but Anderson presents him as a classic American autodidact not so different from the Wrights, except that he took advantage of the Civil War to blag his way into an academic post without any qualifications and was thereafter in a better position than they to get research funding.

It was ultimately the success of the Wrights that got scientists in Europe interested in working on applied aerodynamics (and governments interested in funding them): Russia, Germany, France and Britain all had dedicated fluid dynamics research institutions set up before 1914. Ironically it had the opposite effect in the US, at first: Langley's embarrassing failure and the Wrights' aggressive defence of their patents combined to discourage anyone from venturing into the field, and when the government finally decided that it needed an aviation strategy there was a lengthy turf war between different organisations that felt they should be in charge of it before the NACA could start operations properly in 1920.

It's surprising to read that the aircraft of World War I were still designed largely empirically by practical engineers, who saw little reason to pay attention to the work of the theoreticians. That doesn't seem to have changed much until the early 1920s, when theory had advanced so far that it could actually start to make sensible design recommendations. At the same time, experimental tools were improving, with the construction of better wind tunnels in all the main research institutes.

The book takes the story up to about the mid-1950s, looking quite closely at how the collaboration between theoreticians, experimentalists and engineers led to the development of machines capable of supersonic flight. But it doesn't go into things like the development of computational fluid dynamics in any detail.

Anderson isn't the most elegant of writers — at least three times he tells us that this or that scientist was "not developing his theories of air flow in a vacuum" — and he can be alarmingly brusque in his biographical summaries, but he does present the difficult scientific content very clearly. And it's good to see that, since he's an academic writing for academics and no private person is ever likely to buy a book like this on impulse anyway, his publishers have allowed him to include plenty of those scary-looking equations without which a book like this would soon become unintelligible. (But I was glad that I won't be taking an exam on any of this...)

A peripheral thought: I don't know whether this is to do with Anderson's own professional background or reflects a limitation in the thinking of aerodynamicists in general, but when he's talking about early experimenters trying to find the most effective way to generate lift he never mentions anyone taking a look at what happens in the sails of ships (he does mention windmills a couple of times in passing, but the only mentions of ships are in the context of water-flow around their hulls). Everyone who's ever sailed against the wind knows that air flowing over a curved sail generates a normal (lift) force towards the convex side of the sail as well as the (drag) force in the plane of the sail, and that efficient sailing depends on proper adjustment of the conditions to get smooth (laminar) airflow over both surfaces of the sail. Surely at least some of the people trying to build flying machines must have been in a position to imagine a sail turned round to make a wing? ( )