Industrial Textile Production

Industrial Textile Production in Pompeii:

Although no textiles from Pompeii survived, there is ample archaeological evidence, combined with written sources, wall paintings, and information from other areas of the Empire to establish that Pompeii had a flourishing cloth industry. The number of business establishments connected to the cloth industry has been estimated at 38-43. Production was on an industrial scale, not just domestic. Extant graffiti in one workshop gives us the names of five male weavers and two female spinners, and seven male weavers and eleven female spinners in another shop. The cloth made in Pompeii seems to have been wool. It has been argued that not enough sheep were kept in the vicinity to allow for the quantities of fabric produced, so raw wool must have been imported from other areas.

The workshops at Pompeii provide us with evidence of all the different stages of cloth production – from preparing the raw wool, to dyeing it, spinning it, and weaving it, then laundering it, selling ready to wear clothing items, re-dying faded items, and patching old clothing for use by slaves.

Dyeing was normally done while the wool was still in the fleece, and the paintings at Pompeii show an amazing range of colors. Those produced from locally available materials included madder (red), whortleberry/blueberry (blue/purple), woad (blue), crocus (yellow), elderberry (greyish-lavender), and oak gall (black), plus cloth in the natural colors of the imported wool.

Luxury colors could also be produced from imported dyes. Kermes (crimson,) would have been imported from Merida in Spain, and is the color worn by the wife of the baker Terentius Neo. Indigo was imported from India, which is confirmed by a small Indian ivory statue of the goddess Lakshmi found in a house next to a major dye-works. Pliny mentions Indigo as costing seven denarii a pound, but it gave a richer, more reliable blue than the native woad. A number of the paintings at Pompeii show women wearing green and turquoise. These garments were probably dyed twice, once in yellow and once in blue, making these expensive colors as well. The most famous and expensive of the luxury dyes was Tyrian purple, which would also have been imported. Pliny described the very complicated process of extracting the dye, which required thousands of sea snails to produce a very small quantity of dye. Egyptian papyri give recipes for faking expensive dyes, especially Tyrian purple, and the dyers of Pompeii used those as well.

Mordants in the form of imported alum and locally collected urine were used in Pompeii. Urine was collected from public urinals in jars. Shortly before the eruption of Mount Vesuvius, Emperor Vespasian introduced a tax on urine, leading to much public commentary. (Because of this, centuries later, public urinals in France were sometimes known as Vespasiennes.)

Pompeii also had an important felt workshop, which covered its wall in paintings, which show their felt making process in detail. Undoubtedly these paintings were intended as advertisements and education for the consumers, but they are highly valuable for us as researchers. Pliny the Elder mentions Pompeian felt cloaks, called gausapa, as being fashionable novelties.

Linen was also an important textile in Pompeii. Egyptian linen was considered the best quality and was imported in large quantities through the port of Puteoli. By this time linen production had spread around the Empire to meet the demand – Pliny noted that even Gallic and British women had come to prefer linen to their traditional wool dresses. Flax and linen by this time were being produced in northern Italy, the Damascus region of Syria, Gaul, and Cilicia, Spain producing a quality nearly rivaling that of Egypt. Linen from any, or all, of these areas might have also arrived in Pompeii via the port of Puteoli. It is likely that linen arrived both in the form of whole cloth and in the form of flax to be spun and woven into linen fabric. Linen is very hard to dye and needs to be bleached to achieve the whiteness prized by the Romans. It is unknown how much of the linen was simply bleached in Pompeii, versus how much was subsequently dyed as well.

It is unknown if cotton was being imported into Pompeii. Cotton was available at the time, but it appears to have been an imported textile rather than one woven locally at Pompeii. Mixed fiber cloth, such as wool/cotton and cotton/linen would have been available in the shops at Pompeii, as would expensive checked and embroidered cottons from Iraq and India, fine Indian muslin made from mallow fiber, patterned linens from Gaul, painted muslin from Egypt, and gold-worked or embroidered cloth from Asia Minor, and, of course, silk.

Although a small amount of silk was produced in the Mediterranean from a wild native moth, nearly all the silk available in the Roman Empire came from China via India and along the Silk Road to the great trading centers in Bagram, Damascus, Palmyra and Petra. The Romans did not care for the stiff, heavy Chinese silk, they like light, soft fabrics with good drape. The imported Chinese silk was unraveled and rewoven, often mixed with cotton or linen, which also made the expensive fibers go much further. Syria was a major center of silk production for the Roman Empire, while Egypt specialized in a silk-linen mix which often had woven-in decoration. Persia produced light-weight silk gauze. Dresses made out of such sheer, clinging fabrics were heavily discouraged as being immodest and “un-Roman.” Sumptuary laws were even introduced periodically to curb the vast expenditure on these imported fabrics, but they do not appear to have had too much effect.

There were two additional luxury fabric that were probably available in Pompeii, which would have required a custom order: Byssus, and salamander fur. Byssus, or “sea wool”, was a very fine, silky, gold colored cloth with a high sheen produced from the filaments of the Mediterranean shellfish pinna nobilis. Salamander fur was woven from asbestos. Pliny noted that the slaves who worked with asbestos developed lung diseases, but it continued to be made.

Roman taste demanded that one’s clothing was clean, neatly and visibly pressed, and on formal occasions, for the men at least, pure white. Meeting these requirements was the job of the Fuller, who finished and laundered wool cloth and was a prominent figure among the urban craftsmen. The basic cleaning and shrinking of wool cloth would be undertaken by the weaver, the specialized care of the cloth was provided by the fuller. Fullonicae (Fuller’s Shops) held banks of terracotta tubs for treading cloth in a solution of water and urine or fuller’s earth, and large open rinsing tanks with running water. Supplementary finishing techniques included the raising and trimming a soft nap, bleaching white wools with Sulphur, and pressing laundered garments in a screw-press. Linen cloth was not intensively fulled, but it was sun-bleached and glazed with a polishing glass or slick, rounded stone.

Industrial Textile Production in Britannia:

The two main fibers used in the textile industry of Britannia were sheeps wool and flax, with hemp and other animal hair being used for specific functions. Silk was obviously imported.

The wool produced in Britannia ranged from hairy medium to very fine. According to J.P. Wild, the garrison’s clothing at Vindoland around 100 CE, for example, was made up of 51% of garments made of generalized medium wool, 34% hairy medium wool, 9% fine wool, 4% semi-fine wool, and 2% medium wool. There was also one example found at Vindolanda of a truly hairy fleece, which is a developed type which ultimately would give rise to modern carpet wool. The figures from other Romano-British sites echo these findings, but across the board reveal major advances in fleece quality over the wools found in the Hallstatt salt mines and dating to the 8^th – 3^rd centuries BCE. The wool colors at Vindolanda were 50% grey, 40% white, and 10% black or brown.

Sheep were plucked or shorn in spring and early summer. Plucking leaves most of the coarser kemp and hair behind so that the resulting wool is finer and more uniform, but clipping with iron shears was probably the more widespread practice as it recovered the entire fleece. There is no record of professional tonsores (sheep-shearers) in Britannia, or highly skilled wool sorters and graders. Wool combing may have been a professional, or semi-professional occupation however, as Roman iron woolcombs have been found throughout Britannia.

Roman iron woolcombs were of two types. One type was a flat rectangular comb with long teeth at one end, and two prongs at the other. These were designed to be slotted into the top of a combing post, and they are almost exclusively found in East Anglia. The other style is a rectangular comb with teeth at both ends. This style is found across Central and South East Britain, and also in the continental Roman provinces. Wool combers were probably employed on the larger villa estates. Their job was to align the fibers for spinning, separate the long from the short fibers, and enable the spinning of fine worsted yards from the long fibers which were packaged and sold as balls of rovings (“sausages” of loose fibers).

Although the Edict of Diocletian, published in 301 CE, indicates that there was strong trade in washed wool coming in from select areas of high reputation, Britannia does not appear to be one of them even though it was known for its flocks of sheep. It appears that the wool in Britannia was not much different from the wool available throughout North-West Europe, so there would have been little call for long-distance trade in the raw wool.

Due to the low survival rates of linen textiles in Britain, extant Romano-British linen makes up only a small proportion of the textile remains available for examination. Paleo-botanical evidence has demonstrated that flax was grown on Roman agricultural sites in the Thames River valley, and although evidence for other flax growing areas is limited, by the middle ages it was widespread so it is hypothesized that flax was grown in additional areas in Roman times. It is less certain if the flax was processed in Britannia however, as so far none of the traditional implements for breaking, skutching, and hackling the fiber to extract it from the woody stem of the plant have been found, nor have retting tanks for the first stage of flax processing been identified.

The paleo-botanical record also shows us that hemp was grown in Roman Britain. Hemp was used for sailcloth and rope. Other fibers were used in Roman Britain, including hairmoss for caps, tree bast for rope, and goat hair for sacking cloth.

As found in Pompeii, wool was normally dyed in the fleece before spinning, as revealed by the dye recipies, price lists, and casual references to the task. Flax, if it was dyed at all, was dyed in the hank. Romano-British weavers were known to use contrasting natural fleece shades of brown, black, grey, and white as an alternative to dyeing. Very few extant Romano-British textiles retain any visible dye-color, but some do still retain dye residues. Of the 89 wool textiles sampled at Vindolanda, about 20% retain detectable dye residue. Other sites have also revealed traces of dyes.

Madder, which was imported in the form of a dried root, was found in textiles from Vindolanda. One item from Vindolanda was dyed with local lichen, which would have given a purple color. Wool yarn from a grave at Arrington, Cambridgeshire (c.13-160 CE) was dyed red with madder or bedstraw, and another yarn sample from the same location showed traces of indigotin from woad and residue of madder or bedstraw. Woad seeds have been found at Dragonby, North Lincolnshire. Weld, which gives a yellow dye, is found in the paleobotanic record at York and London. Only a single example of Tyrian purple has been found on a textile in Britain, in a late Roman grave outside Dorchester, Dorset. There is a 1^st century CE “doctor’s” grave at Stanway, Colchester that uses purple yarn in a fine twill, but the sample is too small to analyse. It could be purple made from the Atlantic whelk as was done in Roman Gaul, or the even less expensive purple achieved by overdyeing woad with madder. The Romano-British dyer’s resources, as best we can tell, were woad for blue, bedstraw and imported madder for red, lichens for purple, and weld for yellow, plus colors and shades made available through over-dyeing and mordanting with lye prior to dyeing. There is no evidence for dyeing on the scale done in Pompeii, where the equipment of the fixed dye plant, heated vats, cauldrons, and boilers, remain to show the scale of the industry. Dyeing in Britannia must have been done with portable equipment rather than in permanent workshops.

Spindles and distaffs made of wood are rare finds in Britain, though the spindle whorls of stone, bone, jet, shale, terracotta, cut-down pot shards, and lead are found on most Romano-British sites. More expensive spindles and distaffs made of jet, shale, amber, or bone, and used by upper class ladies, do survive to give testament to the size and shape of the lost wooden ones. As elsewhere in the Empire, spinning was the task of women, and even wealthy women did (or were supposed to) set an example by spinning and weaving for their family and supervising the work of the household colonae (tenant women) and slaves. Tombstones erected for wives usually show wool baskets, spindles loaded with wool, balls of spun wool, and other evidence of the domestic production of the lady in question as testimony to her industriousness and good reputation.

J.P. Wild estimates that spinning was the bottleneck of textile production, and at least five good hand-spinners would be needed to keep a single weaver in continuous production. Larger estates may have employed full-time quasillariae (spinning girls), but for the average household such division of labor would not be practical. Weaving would probably not begin until there was an adequate stockpile of warp and weft yarn.

The quality and consistency of Romano-British spun yarn was very high, especially in the wool yarns, but exceptionally fine yards like those produced in the eastern provinces are very rare. Interestingly, the use of image analysis software on the Vindolanda textiles finds has allowed researchers to isolate the “fingerprints” of individual spinners who contributed wool yard to the textiles used at Vindolanda. The direction in which the spindle was rotated was dictated by convention – clockwise (Z-spun) in all the western provinces. Counter-clockwise (S-spin) was only used for special structural or decorative effects in Britannia. Textiles found in Britain using S-spun thread for both warp and weft appear to be indicative of an imported textile. Yarn in Roman Britain was rarely plied, except to be used as sewing thread.

Although there is evidence for the use of warp-weighted vertical looms, two-beam vertical looms, and small band-weaving looms and devices in Britannia, no timber components of a loom have been discovered (or recognized as such). In most archaeological contexts the presence of air-dried or baked clay weights are the evidence of warp-weighted looms, especially when found in groups. Even so, there are only two significant finds of groups of weights. There is a set of five pyramidal shaped weights, with holes in the top for suspension dated to c.60 CE from Mancetter, Warwickshire, and a set of nine flattened triangular shapend weights from the Bucknowle Villa in Dorset. One possibility for the lack of extant loom-weights is the change from the warp-weighted loom to the two-beam loom that took place during the 1^st and 2^nd centuries CE elsewhere in the Empire. If Britannia followed the same pattern as elsewhere, loom-weights would have stopped being made by about 100 CE, but this cannot be confirmed.

The two-beam looms were constructed to the same proportions as the warp-weighted looms, but the lower beam replaced the weights. Mediterranean weavers preferred the two-beam loom for textiles that incorporated tapestry-woven decorative bands or figure inserts – which would seem to correspond to the change in Roman fashions from the earlier styles of primarily simple tunics for the men and tunica with stola for the matronae, to short-sleeved tunics with increasingly elaborate clavi for everyone. The technique for weaving these bands may well have arrived in Britannia along with the two-beam loom and the broad-headed weaving comb. A few simple examples of locally made tapestry bands have been found among the Romano-British extant textiles.

Finds of a rigid-heddle made from bone, and square or triangular bone, or occasionally copper-alloy, tablets from various British sites show that plain 1/1 tabby and tablet-woven bands were made in Britannia. Independent bands and braids were probably woven on small, body-tensioned, back-strap looms and used for headbands, belts, bandages, webbing, leg wraps, and other similar uses. Some bands were used to form the starting border of larger textiles, and were probably woven on devices attached to a larger, upright loom. In any case, so far the largest set found consists of only four matching tablets, which suggests that these tablet-woven bands were strictly functional and did not become decorative items in their own right until much later.

The examples of complex pattern weaves, like damask and compound tabby and twill, do not appear to have been woven in Britannia. They appear to have been imported from elsewhere. The most prominent type of cloth found at Vindolanda is a generalized medium wool in a 2/2 diamond twill with Z-spun warp and S-spun weft, designed to ensure a compact texture. The diamond repeat pattern varies, but the most popular is a reversal after ten warp threads and seven to nine weft threads, usually with full displacement. Most of the fabric is medium weight, somewhat warp faced, and has a thread-count of about 35 warp threads by 30 weft threads per inch, but some finer twills have 64 warp by 56 weft threads per inch. Selvedges are often plain, with the weft simply returning into the next shed. Some weavers strengthened the edge by making tubular “hollow selvedges”, flat selvedges using a paired weft, selvedges over one or two warp bundles, and twined-cord selvedges that may have been tablet woven have all been found. No starting or finishing borders have yet been found on twills. The diamond twills frequently have a thick nap, and at least five of those found at Vindolanda were dyed red with madder and were probably soldiers’ cloaks.

Another weave found commonly at Vindolanda (about 15% of the samples) is extended tabby (half-basket weave) woven with pairs of single hard Z-spun warp threads and weaker Z-spun pairs for the weft. The sample are predominately finely woven, with 46 warp-ends by 102 weft-pairs per inch, and are light weight, without a nap. The selvedges tend to be constructed over warp bundles of increasing diameter until the end is reached, though some have traces of an extra wrapping thread protecting the outer most edge suggesting that the weaver may have begun with a twined-cord starting border. The loose warp threads were then twisted into a cord running parallel to the weft.

Where the decorative tapestry-woven bands occur as inserts in the military textiles, they tend to be on half-basket weaves, or occasionally on finer 1/1 tabbies. The plain 1/1 tabbies at Vindolanda vary greatly in character. The exception to this is a handful of fragments that have now been interpreted as bandages or leg wraps, but were originally thought to be a girdle. The fragments are about 15” wide, with two plain selvedge edges, and the warp is denser than the weft.

Not all of the fabrics found at Vindolanda are plain, though the patterns are created primarily through the contrast between white and dark undyed wool, or subtle stripes created by alternating bands of S-spun and Z-spun yarn. The so-called “Falkirk ‘tartan’” found at Vindolanda is a herringbone twill with a check pattern composed of light and dark undyed wool. There is also a simple checked fabric made of light and dark wools that had originally used lichen purple dye.

No linen textiles have survived at Vindolanda, though some non-woven textiles have survived such as felt for lining helmets and shoes, and sprang-constructed netting and socks. Two fringed “rain hats” made out of hair moss and made using basketry techniques have also survived.

While the quality of the fabrics at Vindolanda was quite high, the quality of the sewing was not. Seams include simple running stitch seams and run-and-fell seams. Raw edges are usually folded over twice and sewn down, or reinforced by stitching a cord to the edge. Sewing was done with two-ply thread. It is thought that most of the sewing was done by the soldiers themselves.

Other textile remains have been found in other parts of Britain, but the samples are few, tiny, and inconclusive. The finer fabrics may represent imported fabrics rather than good of local manufacture as the textiles from Vindolanda probably are.