I don't feel comfortable with heights, and my wife becomes nervous in tunnels. Oddly enough, we are both at ease in airplanes, which in a sense combine the two; and we positively enjoy crossing bridges of all kinds. Our favorites are the big suspension bridges, which at first glance may seem like a poor way of spanning a wide strait or river. Why go to the trouble of hanging a road or rail bridge from massive cables, when you could just as easily build a structure in which the railway or roadway itself bears the stresses?

You can certainly do this, and cantilever, stone arch, steel arch, and arched truss structures can be found all over the world. So can swing and bascule types, in which a section turns horizontally or opens vertically. Yet the world's longest spans all use the suspension technique. Why?

The answer lies in the different ways that materials respond to various kinds of stress. If heavier and heavier weights are hung from a wire, it will stretch some and at last break. It has been subjected to the type of stress known as tension, and the wire's ability to resist such a pull is termed its tensile strength. A wire will resist the opposite stress, compression, very little if at all. On the other hand, ordinary brick can withstand great compressive forces, up to half a ton per square inch. But if you pull on a brick, it will break under quite a modest force of tension. Brick also breaks easily under the third form of stress, shear. Its weak resistance to shear forces allows performers to smash a way through a stack of bricks with a single blow of the hand.

Most bridges are designed so that the major forces on them are compressive. The exceptions are suspension bridges, in which the biggest forces along the bridge are tensions on the long, curving cables that run above the road or railway. If you look inside one of those cables, you will find it filled with what looks like a whole sheaf of individual wires. In many cases, this is an illusion. A single wire crosses and re-crosses the span, runs over the support towers, and is anchored again and again at each end. In the Brooklyn Bridge such a wire, three-quarters of an inch wide, is close to 200 miles long. Construction began there in 1870. Passers-by at the time wondered if the strange new structure would be able to stand the strain.

So far you may complain that you are reading about straightforward engineering, and old engineering at that. It is not science, and certainly not at the borderlands of science. We come to that when we ask, what is the best material from which to build cables?

The obvious answer is, we should choose the substance with the greatest tensile strength. However, another factor enters whenever the weight of the cable itself forms a significant part of the total load that it must support. This is the case in, for example, space tethers, which are used to extend an observing instrument on a wire until it is many kilometers away from the parent spacecraft. Then we want a material which has a high tensile strength, and is also light. The measurement that matters is the ratio of tensile strength to density.

One simple way of finding this ratio is to hang a length of wire vertically, and see how far we can extend it before it breaks under its own weight. We call this the support length of the material. Some metals are clearly not a good choice. Lead, for example, has a support length of only 180 meters. Gold is better, but still unsatisfactory at 730 meters. Compare this with drawn steel wire, which can have a support length of 54 kilometers. No one can deal with so long a vertical hanging wire, so in practice we attach increasing weights to a short piece of wire of constant cross-section until it breaks.

Drawn steel is strong, but it is nowhere near the limit of strength. For instance, the carbon composite known as Kevlar, perhaps most familiar as a material used to make bullet-proof clothing, has a support length of 200 kilometers. Carbon dislocation-free filaments are stronger yet, with support lengths up to 1,000 kilometers.

Why, then, do we not find these materials used in the cables of suspension bridges? Maybe someday we will, but today the over-riding issues are cost and availability. Iron and steel are cheap, and you can buy them everywhere.

Let us return to the question of the best material from which to make cables, and ask if there are general principles that can guide us to an answer.

First, all chemistry, including the chemical bonds which decide how strong a material can be under tension, involves only the electrons in the outermost electron shell of the atom. Thus from a strength-of-materials point of view, the protons and neutrons in the atomic nucleus, which provide almost all the mass, contribute nothing but useless weight (as do electrons in the inner closed shells). We therefore expect light elements with several electrons in an unfilled outer shell, such as carbon, to offer the best combination of strength and lightness: not lead, not gold, not even iron. The lightest element of all, an atom consisting of one proton, one electron, and no neutrons, should in principle also provide the strongest material.

Unfortunately, this element happens to be hydrogen, a gas at anything more than 20 degrees above absolute zero. This is not on the face of it a promising choice. However, let us look to the future. Under pressures of half a million atmospheres or more, hydrogen takes on a dense crystalline form and is thought to have the properties of a metal. No one has ever done a test, but calculations suggest that a cable formed from such hydrogen would have a support length of more than 9,000 kilometers.

You may ask, so what? Where could ultra-strong materials like this ever be useful? I have a ready answer. They are not just useful but absolutely essential, if we ever hope to build a futuristic device known variously as a beanstalk, a space elevator, or an orbital tower. I described the general idea of this in an earlier column. Next week I will return to the subject and offer novel variations on the theme.