Fibre physics is the study of the structure and physical properties of fibres. These two aspects are not, however, independent: the properties must be explained by the structure, which they also help to elucidate. Because of this connection, it is appropriate to start this book on the physical properties of fibres with a review of what is known about their structure. There is much detail, only partly superseded by more recent work, in the book edited by Hearle and Peters [1] and the review by Hearle and Greer [2]. Other information is in the general references given at the end of the chapter. Matter is composed of atoms linked together by bonds of varying strength. It is the arrangement of these atoms and the strength of the bonds between them that determine the physical properties of materials. Thus with light atoms, such as those of helium, attracted to one another by very weak forces, the energy of the atoms is sufficient (except at very low temperatures) to cause them to move about independently, and the material is a gas. The material will also be a gas (though with a higher liquefaction point) if it is made up of heavier atoms, or of molecules composed of two or three atoms held together by strong forces (valency bonds), provided that the forces between the individual molecules are weak. These weak forces are often called van der Waals forces, since they are the cause of one of the deviations of a real gas from an ideal gas, which were considered by van der Waals in his modification of the gas laws. If the molecules are heavy enough, and the attractive forces strong enough, then the atoms will not have sufficient energy at room temperature to move freely away from one another, and the substance will be a liquid or a solid. Some materials are made up of giant molecules. For example, in a crystal of diamond, all the atoms are linked to one another by valency bonds in a regular threedimensional network. This gives a very hard, non-fusible material. In graphite, which is also pure carbon, the atoms are linked only in single planes by valency bonds; the forces between the planes are weak, so the material is one that easily splits up into sheets, and these will slide over one another, giving a lubricating action. In linear polymers, the linking is in only one dimension. If there is flexibility in the mainchain covalent bonds and only weak bonding between the long-chain molecules, there is nothing to prevent thermal energy from causing the chains to take up a