Collection: Field Notes — Preserving Natural Materials at Sea 

Series Hub: Preserving Canvas 

How the Cutch, Alum, and Castile Soap Process Waterproofs Canvas — and Where It Falls Short

I've been running canvas through this three-bath sequence for several seasons, and the results have been good enough to keep using it and honest enough to keep questioning it. This post works through what each stage is actually doing chemically, where my earlier account of the drying steps turned out to be less certain than I stated, and why the Plymouth Laboratory seawater trials from the 1930s changed how I think about what this process does and does not protect against.

The first time I ran canvas through this process I skipped the intermediate drying stage between the cutch bath and the alum bath because the canvas felt dry enough and I was impatient. The resulting waterproofing was uneven — some sections shed water well, others wetted out within minutes. I assumed the skipped drying stage was the cause, wrote it up as such, and repeated that explanation as if it were established fact.

It is not established fact. Having gone back to the sources properly, I now think I skipped the wrong stage — or at least, the wrong one to blame with confidence. What the chemistry actually says about which drying step matters most is different from what I had assumed, and worth working through before the method.

The chemistry behind each stage is in the cutch and tannic acid notes and the aluminium stearate notes. The canvas enemies notes cover what the treatment is defending against. The Burnettizing and metal salt notes cover the broader mordanting framework and the alternatives to alum. The VAKA field notes hub covers the natural materials approach generally.

What the Three Stages Are Doing

The first bath deposits tannin throughout the fibre. Cutch dissolved in hot water penetrates the cotton or linen threads and leaves tannin compounds distributed within the cell walls. These compounds are biologically hostile to the fungi responsible for mildew and wet rot — but less so than their reputation suggests. W. R. G. Atkins's Plymouth Laboratory seawater immersion trials, running natural fibre nets in contaminated seawater for months at a time, showed that cutch alone provides marginal biological protection barely distinguishable from untreated controls under sustained marine conditions. Cutch-treated hemp rope after ten and a half months retained 17% of its original strength. Untreated control: 13%.

Tannin's more important role in this sequence is probably structural rather than biological: the tannin polyphenols distributed throughout the fibre create bonding sites for the subsequent metal ions. The mordant dyeing literature — which uses this same tannin-then-alum chemistry for fixing natural dyes to cellulose — treats the tannin stage primarily as a preparation for the mordant rather than as a treatment in its own right. Botanical Colors, George Weil, and other dye supply sources all describe this as a two-stage sequence where the tannin creates the attachment points that the alum then fills. That framing is more honest about what the cutch stage is actually contributing than describing it primarily as a preservative.

The second bath deposits aluminium ions at those bonding sites. Alum — aluminium sulphate — dissociates in water and the aluminium ions are taken up and held at the tannin sites distributed through the thread structure.

Here is where my earlier account needs correcting. The mordant dyeing literature does not require drying between the tannin and alum stages — several sources specify going from tannin to alum while the fabric is still wet or damp, with drying noted as beneficial for depth of shade but not chemically mandatory for the complexation to occur. Tannin-aluminium complexation happens readily in aqueous solution. The aluminium ions will find the tannin bonding sites regardless of whether the fabric entered the bath wet or dry.

The third bath converts the aluminium ions fixed within the fibre into aluminium stearate in situ. Castile soap — genuine vegetable oil soap, not synthetic detergent — provides fatty acid ions that react with the aluminium ions to form aluminium stearate at their fixed positions within the fibre. Aluminium stearate is insoluble and hydrophobic. It does not wash out. The waterproofing is built into the fibre rather than coated onto it.

This third transition — alum to soap — is where the drying step is chemically most critical, and it is the step I did not skip on my first attempt.

If canvas goes from the alum bath into the soap bath while still saturated with alum solution, the aluminium ions in that solution meet the fatty acid ions from the soap in the bath water rather than within the fibre. Aluminium stearate forms in solution rather than in situ within the cloth. The bath goes milky — which it does regardless — but much of the waterproofing chemistry is happening outside the canvas rather than inside it. Drying the alum-treated canvas before the soap bath ensures the aluminium ions are located within the fibre structure, at the tannin bonding sites, when the soap introduces the fatty acid ions they need to react with.

Horace Kephart's Camping and Woodcraft (1918) — one of the source documents for this series — gives a soap-then-alum sequence for tent canvas waterproofing and explicitly specifies drying between stages: "Soak the cloth in it, dry out thoroughly, and then soak in an alum solution as above, and dry again." His sequence is reversed from the maritime sail canvas process — soap first, then alum — but the instruction to dry thoroughly between stages is consistent with what the chemistry of the alum-to-soap transition requires.

So: the drying between alum and soap is chemically well-supported. The drying between cutch and alum is beneficial for evenness but may not be strictly required for the complexation. My first attempt skipped the cutch-to-alum drying, not the alum-to-soap drying. Whether my uneven result was caused by that, or by something else entirely — temperature variation, uneven immersion, inconsistent stirring — I cannot say with confidence. The most likely candidate for the critical drying step, based on the chemistry, is the one I did not skip. I will test both stages separately next time and document the results.

What the Alum Stage Does Not Do

The alum mordant is the weakest biocidal compound in the metal salt group. Aluminium tannate has substantially less biological resistance than the equivalent complexes formed with zinc, copper, or iron. The three-bath sequence is primarily a waterproofing treatment with incidental biological protection from the cutch stage — which the Plymouth trials data shows is weak under sustained marine conditions.

This is worth understanding clearly rather than assuming the traditional process is addressing all the threats the canvas faces. If mildew and wet rot resistance in marine service are the priority, the alum stage is the weakest link. The alternative mordant options — copper sulphate via Olie's Dutch method, or ferrous sulphate — are covered in the Burnettizing notes. Whether a ferrous sulphate mordant at the second stage produces meaningfully better biological protection while retaining the waterproofing of the soap stage is the subject of an ongoing field experiment, documented separately.

Before Starting — Washing

New canvas carries sizing — a starch or synthetic finish applied during manufacture. Old canvas carries salt, oil, mildew residue, and accumulated history. Both need to come off before treatment. Sizing prevents tannin penetration. Contamination interferes with the mordanting chemistry.

Wash in hot water — around 60°C for most woven cotton canvas — with washing soda (sodium carbonate) at about a tablespoon per five litres of water. Agitate thoroughly. Rinse until the water runs completely clear — washing soda residue will interfere with the cutch bath.

The canvas goes into the cutch bath damp rather than dry. Damp canvas takes up tannin solution more evenly than dry canvas.

If the canvas has active mildew, treat it before washing. The drying, storage, and mildew recovery notes cover the cleaning process.

Quantities

Per kilogram of dry cloth weight. Weigh the canvas before washing.

The Method

Cutch bath. Dissolve cutch in a small amount of boiling water first — it clumps in lukewarm water. Add to the full bath volume and bring to 70–80°C. Immerse the damp canvas and work it through the bath for 20 to 30 minutes. Remove and allow to drain. Do not rinse — rinsing removes tannin before fixation.

Allow to dry before proceeding to the alum bath. The mordant dyeing literature suggests the tannin-aluminium complexation can proceed with the fabric still damp, but drying allows more even distribution and may improve fixation depth. Given that the next stage — alum to soap — has a clearer chemical requirement for drying, the practical advice is to dry between cutch and alum as well, partly for evenness and partly because there is no cost to it beyond time.

Alum bath. Dissolve aluminium sulphate in warm water and bring to 50–60°C. Immerse the canvas and work through the bath for 20 to 30 minutes. Remove, drain.

Now dry fully before proceeding to the soap bath. This drying step has the clearest chemical justification of the three: aluminium ions need to be located within the fibre, at the tannin bonding sites, before the soap introduces fatty acid ions. Canvas entering the soap bath while still saturated with alum solution will form aluminium stearate in the bath water rather than in the fibre. The bath going milky is normal — but if the canvas was wet with alum solution entering the bath, much of that milkiness represents waterproofing that did not end up inside the cloth. Allow the alum-treated canvas to dry fully — which may mean 24 hours in UK conditions in a damp autumn — before the soap stage.

Castile soap bath. Dissolve grated or flaked castile soap fully in warm water and bring to 50–60°C. Immerse the alum-treated canvas and work through for 20 to 30 minutes. Remove, drain. Do not rinse. The aluminium stearate continues setting within the fibre as the canvas dries. The drying period after this stage is part of the curing process.

A note on the soap: it must be genuine vegetable-oil soap — sodium stearate or oleate from olive oil or equivalent. Synthetic detergents do not contain the fatty acid salts that react with aluminium ions to form aluminium stearate. A synthetic detergent in the soap bath produces no waterproofing formation at all, regardless of how similar it looks to genuine castile soap. This is the most common error in attempts to replicate this process from incomplete descriptions.

Allow to cure for at least 48 hours before use.

Cheatsheet

Before starting: Weigh dry cloth. Wash with washing soda at 60°C. Rinse thoroughly. Leave damp.

Bath 1 — Cutch: 150g cutch per kg dry cloth / 6–8 litres water at 70–80°C / 20–30 minutes / Drain, do not rinse, dry — probably beneficial, may not be strictly required for complexation

Bath 2 — Alum: 100g aluminium sulphate per kg dry cloth / 6–8 litres water at 50–60°C / 20–30 minutes / Drain, dry fully — chemically important before the soap stage

Bath 3 — Castile Soap: 70g castile soap per kg dry cloth / 6–8 litres water at 50–60°C / 20–30 minutes / Drain, do not rinse, hang to dry

Cure: 48 hours minimum before use.

What This Treatment Provides — and What It Does Not

The three-bath process produces canvas with sound waterproofing from the aluminium stearate formation and modest biological resistance from the deposited tannin. Under the Plymouth trials conditions — sustained seawater immersion, contaminated harbour water — cutch-level biological protection is closer to negligible than meaningful. In practice on canvas that is dried between uses, stored carefully, and maintained regularly, the biological protection provided by the cutch stage will be more useful than the Plymouth data's most pessimistic reading suggests. But it is not strong protection, and should not be relied on as such in warm water or persistently wet conditions.

An overcoat of natural wax adds surface water repellency to what the three-bath process establishes within the fibre. This treatment also makes an excellent preparation for oilskin production — the tannin-mordant base is already in place before the linseed oil and wax goes on top. The Oilskins series covers that process in full.

Where the Mordant Stage Could Go Further

Substituting copper sulphate for alum at the mordant stage — the Olie Dutch method — produces dramatically better biological resistance with no soap stage required, though copper stearate formation via a subsequent soap bath is chemically plausible and may add waterproofing. The ecological constraints on copper near water are real and detailed in the Burnettizing notes.

Ferrous sulphate at the mordant stage occupies a more ecologically defensible position — iron is a marine micronutrient rather than an aquatic biocide — and may produce strong biological resistance via iron tannate formation while leaving the aluminium stearate waterproofing pathway open if followed by a castile soap bath. Whether iron stearate forms as cleanly as aluminium stearate from a ferrous-mordanted canvas, and whether the biological protection gain is meaningful in practice, is what the field experiment below is testing.

Ferrous Sulphate Canvas Treatment — A Field Experiment

Why This Treatment Does Not Work on Rope

The aluminium stearate formed within the canvas fibres is a rigid crystalline compound at the scale of individual fibres. In woven canvas, fibres move very little relative to each other under load — the rigidity causes no problem. In rope under load, fibres rotate and bear against each other continuously. The aluminium stearate crystals fracture under this movement, generating abrasive particles within the rope's own structure. Beyond this, the biological protection from the cutch stage alone — as the Plymouth trials show — is almost negligible for rope in sustained seawater immersion. The rope dressings notes cover appropriate treatments for natural fibre cordage.

References:

Kephart, H. (1918). Camping and Woodcraft, Volume I, pp. 71–73. Available via Anna's Archive and Internet Archive.

Atkins, W. R. G. (1928). The Preservation of Fishing Nets by Treatment with Copper Soaps and Other Substances. Journal of the Marine Biological Association of the United Kingdom, 15, 219–235. Available via mba.ac.uk.

Atkins, W. R. G. and Purser, J. (1936). The Preservation of Fibre Ropes for Use in Sea-Water. Journal of the Marine Biological Association of the United Kingdom, 20, 643–654. Available via mba.ac.uk.

Olie, J., Jr. (1918). Voorschriften voor de behandeling van netten met kopersulfaat en ammonia. Verslag van der directeur van het Nederlandsch Visscherij-Proefstation over het jaar 1917, pp. 40–42.

Botanical Colors: botanicalcolors.com/how-to-mordant-cellulose-fibers-tannin-alum — for mordant dyeing practice on the tannin-alum sequence.

George Weil natural dyeing: georgeweil.com/blog/mordants-for-natural-dyeing — for the immediate tannin-to-alum transition in dyeing practice.

I design and build natural boats and take them to places worth going. Once the plans are finalised, find them at VAKA Boatplans. The full knowledge base is at Field Notes.