Chemistry is mainly about simple aggregates of just a few atoms. It deals more uneasily with polymers, huge molecules or bonded lattices, but even these redeem themselves by being composed of repeating simple units. Daedalus is now inventing an entirely different chemistry.

His inspiration was the strange yellow solid deposited from phosphine gas. This solid is apparently a random mass of phosphorus atoms bonded any which way, with any loose bonds terminated by hydrogen atoms. He reckons there must be a whole range of random atomic aggregates like this, with no repeating units. Indeed, chemists must have made them thousands of times — as the brown 'tar' of so many failed reactions, always thrown down the sink in disgust. Coal is perhaps the closest natural example. But Daedalus is now making them deliberately.

He is aiming a number of atomic beams in vacuo at a cryogenically cooled target. Each atom sticks where it lands, and bonds to other atoms nearby. The overall composition of the product can be varied by changing the intensity of the beams, but its short-range structure must be entirely random. It will not be a polymer or copolymer, but rather a random macromolecular structure, or 'randomer'.

When the deposit is warmed up, it may turn out to be unstable, and decompose to small molecules. But for certain ranges of composition, it will be stable. Stable randomers will probably contain a high proportion of atoms such as carbon, sulphur, silicon and phosphorus, which bond easily to each other. They will have less hydrogen, oxygen, nitrogen or halogens, although these will be useful for terminating spare valencies. Metals, especially transition metals of variable valency, may aid the incorporation of other atoms. They will also add variety to the range and properties of randomers.

When he has a sense of the most promising randomer compositions and their properties, Daedalus will make them in bulk, by adapting the most tar-like products of conventional reactions. It will be slow work, and even Daedalus cannot guess what use its products may be. As covalently bonded glasses, they may fill a gap in the polymer spectrum; as utterly unnatural substances, they may be ideal for inert surgical implants; as solid-state compositions, they may transform electronics. But new chemistry is always fertile. The field of randomers is bound to be good for something.