Silk: Filament, Weave, and Weight

Silk can arrive at cloth as smooth, reeled filament or as shorter fibres spun into yarn. It can be woven into satin, plain weave, twill, crepe, organza, habotai, velvet, or dozens of other constructions. It can be lightweight and crisp, dense and fluid, matte and nubby, or highly reflective. The fibre label is true in all of those cases.

Fineas Jackson’s silk comparison asks why two 100% silk products can differ so sharply ¹. His satin episode supplies the companion distinction: satin names a construction, not a fibre ². Put the two together and a product description becomes a coordinate system rather than a prestige word.

Cocoon to yarn: two major routes

Reeled silk

A silkworm forms a cocoon from a long filament bonded by sericin, a gummy protein. In reeling, softened cocoon filaments are found, drawn together, and wound as a continuous raw-silk thread. Several fine filaments combine because one alone is too delicate for ordinary handling.

Continuity matters. Breaks require cocoon replacement and joining; uneven feeding can create thick places; poor cohesion can impair later weaving. The FAO manual connects cocoon reelability and filament defects directly to production interruption, waste, and raw-silk quality ³.

Spun silk

Pierced cocoons, reeling waste, broken filament, and other shorter silk fibres can be degummed, opened, aligned, and spun. The result is still silk. It commonly has a softer, less glassy, more staple-yarn surface. The FAO manual notes that good spun-silk fabric can provide good service when the fibre quality and processing are sound ³.

“Reeled” is not a moral category and “spun” is not a synonym for counterfeit. They are routes with different yield, labour, surface, uniformity, and end-use possibilities.

Degumming changes the hand

Degumming removes some or most sericin. Removing gum reveals softness and lustre, reduces mass, and alters dye uptake and handling. Too little removal can leave stiffness or unevenness; excessive processing can weaken material. Degumming can occur at different stages depending on the product.

Wild silks and cultivated mulberry silk may differ in colour, filament regularity, diameter, available length, sericin, and dye response. Species or feeding story is not a finished-cloth grade. The FAO testing material treats evenness, cleanness, cohesion, defects, colour, lustre, and smoothness as separate observations ⁴.

Construction changes the same fibre

Satin

Relatively long floats allow yarn to dominate the surface with fewer visible intersections, supporting smoothness, lustre, and drape. The same floats can be vulnerable to snagging, picking, and seam slippage. CottonWorks lists both effects when explaining satin structures ⁵.

Crepe

High-twist yarns and finishing can create a pebbled surface, dry hand, spring, and different drape. “Crepe” also describes a family, not one universal specification.

Organza

Fine, strongly twisted filament yarns in an open plain weave can create crisp transparency. Lightness is the point; judging it by upholstery abrasion criteria would be a category error.

Twill and plain weave

Frequent interlacements can change stability, opacity, wrinkle pattern, and surface reflection. A silk twill scarf and silk satin blouse share fibre yet behave differently before sewing begins.

Weight: useful, incomplete evidence

Silk fabric is often sold by momme, a traditional mass convention based on the pounds in a 45-inch-by-100-yard sample. A common conversion is approximately 1 momme = 4.3 grams per square metre, though sellers may round ⁶. Higher momme generally means more fabric mass per area, which can increase opacity and abrasion reserve and change drape. It does not automatically establish better filament, tighter quality control, safer dye, stronger seams, or fitness for summer.

Use mass as a specification, not a rank. A 12-momme scarf and a 25-momme pillowcase serve different jobs.

The SILK read

• S — Source and strand: Fibre percentage, cultivated or wild claim, reeled filament or spun staple, yarn twist and ply, documentation and unknowns.
• I — Interlacement: Satin, plain weave, twill, crepe, knit, pile, density, float length, and stretch direction.
• L — Load and lifecycle: Mass, opacity, expected abrasion, snag exposure, laundering or dry cleaning, perspiration, light, storage, repair.
• K — Kit of parts: Seams, needle size evidence, puckering, lining, hem, zipper, button, interfacing, pattern matching, and allowance.

The FTC fibre rule supports the first question by requiring covered products to identify fibre content ⁷. It does not answer the remaining three.

Light-table comparison

Use two owned silk items or legitimate swatches.

1. Record labels, dimensions, and mass; calculate approximate mass per area only if the item geometry allows it.
2. Place each over the same printed grid to compare opacity and construction.
3. Tilt each under one fixed light and note whether reflection is broad, sharp, directional, or scattered.
4. Inspect yarn surface with a magnifier: continuous and smooth, spun and fuzzy, highly twisted, slubbed, or unknown.
5. Examine seam puckering, stitch holes, float snagging, hem depth, and stress points.
6. State which use each construction appears designed for.

Do not perform burn tests on garments. Fibre identification by flame involves fire, fumes, damage, and ambiguous interpretation; use documentation or professional testing.

Rule of thumb

Silk names the fibre; it does not name the cloth. Ask filament or spun, yarn, construction, mass, degumming, dyeing, finish, and sewing. A useful premium points to a controlled difference you can connect to the product’s job.