Remarkable science often lurks in the most ordinary of circumstances.

In 1895, a German engineer published a short paper reporting the results of experiments on corn silos. This was the same year that Hendrik Lorentz published an approximate form for the coordinate transformation appropriate for Maxwell's equations, and one year after George FitzGerald had suggested 'length contraction' as a possible explanation for the original null result of Michelson and Morley, reported in 1881, for the speed of the Earth's movement through the aether. Intellectually speaking, H. A. Janssen's experiments were miles away from these precursory ideas of the theory of relativity, yet they prove that scientific truism — remarkable science often lurks in the most ordinary of circumstances.

At the time, a huge industry had grown up to move corn from producing nations to locations all around the world. Silos acted as storage facilities along the way, taking deposits from the railways and dispensing to ships. Janssen noticed that construction books offered little guidance about the required strength of a silo's walls, based on estimates of how much pressure a tall column of corn might exert against them. What's more, it was apparently common knowledge at the time that the pressure in a corn silo works very differently from that in, say, a column of water.

The pressure at the base of a fluid column, of course, simply grows with the column's height. But as corn pours into a silo, the pressure at the base ultimately reaches a limit, and pouring in more corn, strangely enough, has no effect. Janssen's experiments, using small-scale wooden silos that he constructed himself, confirmed that the pressure reached a limit in this way. Making several simple assumptions, he also managed to calculate the pressure profile in close agreement with the data.

Thanks to Matthias Sperl, scientists who do not read German now have access to an English translation of Janssen's original paper (http://arxiv.org/abs/cond-mat/0511618). As he recognized, the distinctive behaviour of what we now call 'granular matter' has to do with friction. Acting between pieces of corn, friction effectively locks the silo's contents into a semi-rigid whole, and friction at the walls can then help to support the column's weight. It's very different from a column of water or other fluid, which cannot support shearing forces.

What is most striking, however, is Janssen's awareness of deep differences between ordinary matter and granular matter, as exemplified by the corn in his silos. Janssen tried without success to devise experiments that would probe the pressure at specific points on the silo walls. He attributed his difficulties to the 'arching' of force within the corn; that is, to the transmission of force from grain to grain along irregular branching paths, which leads (we now know) to far higher stress in some locations than in others. Ultimately, such effects stem from thorny issues that confront modern physics in many settings — namely, disorder and the absence of equilibrium. Indeed, it now appears that granular systems in many circumstances exhibit striking similarities to glassy materials.

Janssen's experiments and explanations of them greatly improved silo design in subsequent years. But as far as physics is concerned, he was a pioneer. As Sperl notes, Janssen's paper had been cited 40 times before 1977 — on average less than once a year. Since then, it has been cited 375 times, and more than 150 times in the last decade alone.