The pile can be immaculate and the fringe will still give a rug away. Cream at one end, a soft dirty ochre at the other, or that flat urine-coloured tone at the corners where the rug meets a radiator. People notice fringes the way they notice a shirt collar, which is to say instantly and without meaning to.
A yellow fringe also gets treated the wrong way round more often than any other part of a rug. Householders reach for whatever is under the sink and scrub, and scrubbing is the one thing that turns a colour problem into a structural one.
Why does the fringe yellow when the pile looks fine?
Start with what the fringe is. On a hand-knotted rug the fringe is structural – it is the ends of the warp, the threads the whole rug was woven onto, and on the large majority of Persian and Afghan pieces those warps are cotton. Some tribal weavings use wool warps, and silk rugs use silk, but assume cotton until you have looked.
That single fact explains most of what follows. Cotton and wool are different chemistries with opposite preferences. Cotton oxidises with age and turns straw-coloured whether or not anybody has walked on it; the pile beside it, being wool, does no such thing.
Then there is everything the fringe collects that the pile does not. It lies flat on the floor rather than standing up, so it takes shoe soil, dust, kitchen grease carried on socks and the oils off bare feet, and it holds them at the surface where they oxidise into that yellow-brown tone. Sunlight through a south-facing window adds photo-yellowing on top. Wool yellows in UV as well, through the breakdown of the amino acids in the fibre, which is why a rug that lives half in the light and half in shade develops a line across the fringe you could measure with a ruler.
What the fringe is made of
Machine-made rugs are a different case again. The fringe is often a separate strip sewn on afterwards, sometimes in polypropylene or rayon, and rayon yellows badly and permanently the moment it gets wet in the wrong way. Check the underside for a stitched seam before you treat anything, because a sewn-on fringe can be replaced for a modest sum and does not warrant chemistry.
I spent an hour on one in Twickenham before I turned it over – a power-loomed rug, viscose fringe, yellowing that no reducing agent was ever going to move because the fibre had degraded the first time the rug was shampooed by somebody with a hire machine. A trimmer and a new fringe tape cost the owner less than my hour had. Turn the rug over first.
How do you tell soiling from fibre damage?
The tests are quick. Take a white cotton cloth, dampen it, and rub a short section of fringe between thumb and finger. Soil transfers – you will see it on the cloth in seconds. Oxidised cellulose does not transfer at all, because the yellow is the fibre itself.
Then look at the knot line, the point where the fringe emerges from the woven edge. Soiling is heaviest at the far tips, where the fringe touches the floor and gets kicked. Damage and chemical yellowing tend to run the other way, deepest at the knot line where moisture sits longest and where any previous over-wetting drew contamination out of the rug and parked it.
Smell the fringe at the knot line as well. It sounds absurd and it settles the question faster than anything else on a rug that has been in a house with animals, because urine wicks to the edges as a rug dries and lands precisely there. Yellowing that comes with a sharp note under a warm hand is a contamination problem with a fixed order of operations, and washing it before the salts have been dealt with locks the colour in.
Pull a single thread gently. A sound cotton warp resists; a degraded one gives way with an unnerving softness and leaves a powdery break. If threads are breaking, the job is not a cleaning job any more.
The rub test at the knot line
A rug in a flat on Highbury New Park had a fringe that read as filthy from the doorway and turned out to be two separate problems in one. The outer three inches were soil, and came back to cream with a straightforward acidic rinse. The inch nearest the knots was a hard yellow-brown that no amount of washing touched, because a previous cleaner had soaked the rug on a wooden floor and let it dry from the top down – the wrong way round – pulling everything in the foundation up to the edge and stranding it there.
That second inch took a reducing agent and two attempts, and it never came back to match. I told the owner it would not, before I started.
Which cleaning methods make yellowing worse?
Alkalinity, first and worst. Most general-purpose cleaners and a good number of carpet shampoos sit between pH 9 and pH 11 because alkalinity shifts grease. Wool is damaged above about pH 8 and cotton yellows under alkaline conditions, so an alkaline product on a fringe attacks both fibres at once and sets a colour that was previously loose. Washing-up liquid falls into this bracket. So does most laundry powder.
Household chlorine bleach on a fringe is the other one, and I have pulled apart rugs where somebody has done it. Chlorine oxidises cotton aggressively and destroys wool outright, and the effect is delayed – the fringe looks brilliantly white on the day, and eight months later the threads snap where they were treated and the rug starts unravelling from the edge inwards.
Then drying. A rug dried slowly, face up, on an impermeable floor, will wick contamination from the foundation into the fringe as the moisture travels upwards and outwards to evaporate. This is the mechanism behind half the yellow fringes I am asked to fix, and it is entirely self-inflicted.
Alkalinity and drying the wrong way round
Fringes should dry fast and with air moving across them. On a proper wash floor a rug goes onto a grid or a slatted rack with fans underneath so the edges dry first and the moisture retreats inward rather than travelling out. At home, that means the rug up off the floor, air moving, and the fringe never left lying in a damp fold. A fringe that stays wet for eighteen hours will yellow whatever you put on it.
How do you bring the colour back safely?
Take the mechanical part first, because a surprising amount of what looks like discolouration is dry particulate. Comb the fringe out with a wide-toothed comb, working from the tips back towards the knots in short passes so you are not dragging the whole length. Vacuum with an upholstery nozzle and no beater bar.
Then wash it acidic. A dilute citric acid solution, applied to the fringe alone with a sponge or a soft brush laid flat rather than worked in circles, sits the fibre back where it belongs and lifts a large share of ordinary soil. Rinse with clean water, blot hard with white towelling, and dry fast.
For yellowing that survives that, the two options behave in opposite ways. A reducing agent – sodium hydrosulphite is the usual one, sold to the trade as a browning treatment – strips oxygen out of the coloured compound and is the correct tool for cellulosic yellowing at the knot line. Hydrogen peroxide oxidises instead, works well on general dinginess and on organic staining, and is gentler on the cotton’s strength but far more likely to lift dye from an adjacent pile.
Whichever you use, mask the pile. Lay the fringe out over a sheet of polythene with a folded towel under it so nothing wicks back into the wool, work in short lengths, and rinse thoroughly, because residue left in the fringe carries on reacting for weeks after you have finished.
Reducing agents versus peroxide
Neither belongs in a spray bottle used by guesswork. Both are pH-sensitive, both need a controlled dwell of a few minutes rather than an hour, and both will do damage that shows up later if left to dry in place. If a fringe matters – if it is on a rug worth more than the cost of getting it wrong – the honest answer is that this is bench work with a proper rinse and a drying room.
How white should a restored fringe be?
Less white than the client wants. This is where I part company with a chunk of my own trade: the fashion for brilliant white fringes on old rugs is wrong, and a Kashan or a Bidjar with a fringe whiter than the wool beside it looks exactly like what it is, which is a rug that has been over-processed by somebody selling a visible result.
Many rugs never had a white fringe. A great number of nineteenth and twentieth century pieces went through a deliberate tea wash or antique wash at some stage, and the mellow tone in the fringe is part of that finish rather than dirt sitting on it. Bleach it out and you have removed something that took decades to arrive and cannot be put back.
My working rule is one shade lighter than it started, matched by eye against the lightest wool in the field, and stop there. Even and slightly warm reads as a cared-for rug. Optical white reads as a repair.
Antique wash and knowing when to stop
Uneven results are worse than a uniform yellow, and it is easy to produce them by treating the visible end of the rug and not the end under the sofa. Do both fringes in one session, in the same conditions, or the pair will never match again.
The Highbury rug went back down with the darker end towards the window. It is the only thing that made the two ends look like a set, and the owner moves it round every spring.