The cylinder seal’s shape is apparent from its name, but its dimensions vary from seals of a few centimetres to those big enough to fill the palm of the hand, from squat and spool-like to long and thin. The designs on seals were carved in intaglio, engraved into the seal so that when the cylinder was rolled over clay, the image stood up in relief. The cylinder could be quickly rolled over clay and left a long and unbroken impression. This gave it an advantage over its antecedent, the stamp seal, which could not cover clay as comprehensively or with so little work. This is perhaps one of the reasons for the invention of the cylinder seal (Nissen 1977: 15). As social organisation was growing in such a way that institutions needed to control larger amounts of goods for larger numbers of people, such innovations for speed and reliability were needed.

Seals can be made of numerous materials, most commonly stone but also glass (BM 108717), wood (EA 49018), bone (BM 139988), ivory (BM 134306), shell (Ashm 1932.321), or clay (BM 138136). Although stone seals are by far the best known to us, this does not necessarily accurately represent the distribution of materials. Seals in more delicate materials may have been less likely to survive. Indeed, most of the seals of shell that we do have are worn, chipped or burned (Ashm 1930.108, Ashm 1969.786, Ashm 1969.346). The most popular stones for seals varied in different times and places and depended in part on drilling technology (Sax et al 1993) and in part on what sorts of materials were available.

Seals were normally strung or mounted in some manner. Most seals were perforated, usually straight through the centre of the seal, although a v-shape perforation was popular in the Jemdet Nasr period (Collon 2005: 108). These perforations allowed the seals to be strung and worn around the neck or pinned to clothing, but still permitted them to be rolled freely. Seals deposited in graves are often found roughly at the breast (Rathje 1977: 26), suggesting that the normal manner of wearing such items was pinned to clothing in that area. Caps made of metal, shell, or bone, could also be fitted at one or either end. A particularly magnificent example, a silver figure of a recumbent ram, can be found on a large white magnesite Uruk seal in the Ashmolean (1964.744). Although it is now locked in place, originally the cap was likely designed to rotate around a copper spindle (Hamilton 1967: 34). Caps were particularly popular on prestigious Uruk seals of this type. Caps could leave impressions when a seal was rolled; an Ur III seal of Ur-Šulpa’e had a cap which left deep grooves on texts that he sealed (RA 101 Dahl and Hebenstreit 1; AUCT 3, 84; TRU 361). The stone at the ends of seals might also be carved themselves (AO 21111, BM 130670); not merely a decoration, this meant it could serve a double purpose as cylinder and stamp seal (Teissier 1994: 10, see n. 416-7).

Seals could be engraved through drilling or wheel-cutting (the two rotary technologies), microchipping, or sawing and filing. Each technique produces a distinctive type of cut which can be generally recognised with close study (Sax et al 2000: 157). Perhaps the most influential technological innovation was the development of the free-standing rotary cutting-wheel. This method allowed the engraving of much harder stones than it had previously been possible to use. In a study using a low-power binocular microscope and scanning electron microscopy (Sax and Meeks 1994), Sax found no evidence of mounted wheel-cutting techniques before the mid eighteenth century BC (the Old Babylonian period), and the technique appears not to have been widely adopted until the Kassite period (Sax et al 2000: 167-172). This was in contradiction to previous theories which put the invention of wheel-cutting technology as early as the mid-fourth millennium (Nissen 1977: 16). It was not merely the wheel itself that mattered, ‘the introduction of wheel-cutting was part of a complex framework of technological change involving the adoption of a new abrasive, a range of new working techniques, improvements to the design of rotary equipment and, presumably, a consumer preference for hard stones.’ (Sax et al 2000: 173).