Collection: Field Notes — Preserving Natural Materials at Sea 

Series Hub: Preserving Wood 

A Victorian naval surgeon's preservation patent, what the Plymouth Laboratory trials revealed about what actually works, and what that means for anyone trying to do this without industrial equipment or ecological compromise

Sir William Burnett was Surgeon and Director-General of the Royal Navy's Medical Department, a veteran of Trafalgar, and by the late 1830s possessed of a zinc chloride solution he was convinced would solve the Royal Navy's timber and canvas preservation problem. He patented it in 1839. He spent the following two decades attempting to persuade the Admiralty and the railway industry that his process — forcing the salt solution into timber and canvas under pressure — was the answer to a preservation problem that was costing both institutions considerable money. He achieved partial success with the railways, less with the Admiralty, and damaged his professional reputation in the process by promoting a commercial interest from a position of institutional authority in ways his contemporaries found distasteful.

He was probably also right about the chemistry, which makes the story more interesting than a straightforward account of entrepreneurial overreach.

The patent — formally titled Sir William Burnett's Patent for the Preservation of Timber, Canvas, Cordage, Cotton, Woollen, etc., from dry rot, mildew, moth and premature decay — is available via Google Books. Reading it is instructive. The specific claim that pressure-impregnated salt solution reaches parts of the fibre structure that surface-applied treatments cannot is sound. What Burnett could not have known, because the systematic trials had not yet been done, was which metal salts would prove most effective and how the ecological consequences of each would compare. That work came a century later, out of the Plymouth Laboratory, and the results complicate the picture considerably.

The Preserving Canvas series covers the canvas applications more fully. The VAKA field notes hub has the broader context.

The Mordanting Framework

Before covering individual metal salts, the underlying concept is worth establishing because it makes the whole family of treatments coherent rather than a list of separate procedures.

Mordanting is a term from the dyeing tradition — a mordant is a metallic salt that fixes a dye to the fibre by forming an insoluble complex within the thread structure. The same chemistry applies to preservation. When canvas or cordage is treated with a tannin — cutch, oak bark liquor, any condensed tannin solution — the tannin penetrates the fibre and deposits within it. A subsequent bath in a metal salt solution causes the metal ions to react with the deposited tannin, forming a metal-tannate complex that is insoluble in water, mechanically bound within the fibre, and biologically resistant.

The tannin's role in this sequence is worth examining carefully, because it is doing more than providing biological resistance directly. Tannin polyphenols distributed throughout the fibre create bonding sites for the subsequent metal ions. Metal ions from a copper or iron salt bath applied to tannin-pre-treated fibre have more distributed attachment points and are better retained within the fibre structure than metal ions applied to untreated fibre. The tannin stage matters as a fixative for what follows, not just as a biocide in its own right.

This distinction matters because W. R. G. Atkins's Plymouth Laboratory trials — running natural fibre nets and ropes in seawater immersion 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 in contaminated seawater retained 17% of its original strength. Untreated control: 13%. That is not a preservation result in any meaningful sense. But cutch followed by copper mordant produced dramatically better outcomes than copper mordant alone on short immersion, suggesting the tannin's distribution and fixation function is genuinely extending the metal treatment's durability within the fibre even when the tannin itself contributes almost nothing biologically.

So tannin in these sequences is doing three things of varying importance: modest direct biological resistance, which the data shows is weak under sustained marine conditions; structural mordanting to distribute and initially retain subsequent metal ions, which is meaningful; and UV absorption and fibre plasticisation, which is secondary but real. Its role as a fixative is probably more important to the durability of the subsequent metal treatment than its role as a biocide — which inverts the priority usually implied when cutch is described primarily as a preservative.

The Metal Salt Spectrum

The different metal salts available for preservation work each have a distinct character. They are not interchangeable, and the Atkins trials quantify the differences rather than just predicting them from chemistry.

Zinc chloride is the Burnett salt and the starting point for this note. It is a moderately effective biocide and a strongly hygroscopic salt. In the cell structure of wood or textile fibre, it disrupts the enzymatic processes of rot fungi and bacteria. The hygroscopicity that makes it effective at disrupting fungal metabolism is also its primary limitation: in persistently damp conditions, zinc-treated materials attract and hold moisture. The early railway experience with Burnettized sleepers documented this problem clearly — sleepers treated with too strong a solution became brittle and absorbed moisture in ways that accelerated rather than retarded their deterioration. For canvas in marine use, zinc chloride should not be the final stage. It wants an overcoat that seals the surface and manages moisture exposure. Zinc provides the biological resistance; the overcoat provides the moisture management.

For timber showing early active rot, a 10 to 15 percent zinc chloride solution arrested fungal activity in my own use and is the application where it remains most relevant. It needs sealing over promptly with linseed oil and shellac once the treated area has dried. This is an intervention for wood already in trouble, not a preventive treatment for sound material.

Ferrous sulphate — green vitriol, iron sulphate — has an independent history as a canvas and cordage preservative, and the Atkins trials data places it in an interesting position. The commercial product Cuprinol Brown, which Atkins and Purser describe as iron naphthenate probably with some zinc naphthenate in a tar oil carrier, produced the most striking result in the Plymouth Sound sisal rope trials: 98% retained strength after seven and a half months of immersion, 81% after twelve months. For context, the untreated control was at 29% after seven and a half months and 18% after twelve. Stockholm tar alone was at 74% after seven and a half months and had failed entirely at twelve. Copper naphthenate (Cuprinol green) was at 76% at seven and a half months and also lost by twelve.

The iron-based treatment outperformed both copper and zinc naphthenate products at twelve months and was the only treatment to maintain useful strength at that point. Atkins and Purser attribute this partly to what they describe as "a specially good tar oil" in the Cuprinol Brown formulation — the tar carrier is contributing to the result alongside the iron soap, and separating their individual contributions from the trial data is not possible. What the data does establish is that iron naphthenate in a tar carrier is at minimum comparable to copper soap treatments and possibly superior for sustained immersion.

Tannin-iron mordanting is one of the oldest and most stable complexing reactions known — the basis of iron gall ink and the black-green staining of oak by iron-rich water. Iron tannate is insoluble, deeply fixed, and strongly biocidal. Ferrous sulphate applied to tannin-pretreated canvas or cordage deposits iron tannate throughout the fibre structure in a way that is more durably fixed than metal ions applied without the tannin pre-treatment stage.

The iron rot risk is real and worth understanding. Ferric iron compounds can catalyse acid hydrolysis of cellulose — the same mechanism that destroys documents written in iron gall ink over centuries. In textile fibres the process is driven by residual free iron compounds cycling between ferrous and ferric states in wet-dry conditions, producing acid that attacks the cellulose chain. The mitigation is threefold: keep working concentrations low, ensure thorough tannin mordanting so as much iron as possible is fixed as insoluble iron tannate rather than remaining as free ferric salt, and seal the treated material with oil or wax to interrupt the wet-dry cycling that drives iron oxidation.

The colour consequence is significant and permanent: ferrous sulphate produces grey-green to black staining on tannin-treated material, intensifying over time. On working sails this may be acceptable. On canvas covers or work where appearance matters, it may not be wanted.

Copper sulphate — blue vitriol — is the most potent biocide in the group at equivalent concentrations. The Olie method — cutch followed by ammoniacal copper sulphate — produced hemp nets still at half strength after 20 months of continuous immersion when the untreated control had failed at under three months. That performance is well documented and the mechanism is sound. Copper tannate is deeply fixed and strongly biocidal. The colour consequence is blue-green staining.

The ecological question around copper is where this note gets uncomfortable. Copper is acutely toxic to fish, invertebrates, and algae at very low concentrations — far lower than the concentrations involved in canvas or cordage treatment. It is used deliberately as an algicide in fish-free water bodies. Treated canvas or cordage leaching copper into harbour water represents a genuine aquatic toxicity risk, not a theoretical one. Treated material should be given time — ideally a full season ashore — before going to work over water, and should be fully overcoated before any water contact. A well-mordanted, well-overcoated canvas will leach far less than a freshly treated one, but the risk during the treatment and early curing period is real.

The iron versus copper ecological comparison is the most important practical question raised by the Atkins data, and the answer is not ambiguous. Iron is an essential marine micronutrient. At the concentrations released by treated canvas or rope weathering in seawater, iron is not an aquatic toxin — marine systems need it and are adapted to it. Copper at equivalent concentrations is an acute aquatic biocide. The ecological case for iron over copper in marine applications is clear. The performance case, based on the Cuprinol Brown results, is that iron in a tar carrier is at least as good as copper and possibly better for sustained immersion. If that result holds up under closer examination — and I have been making iron oleate to test this, which is documented separately — the argument for iron over copper in natural fibre marine preservation is both ecologically and practically compelling.

Aluminium sulphate — alum — is the most familiar of the group from the traditional cutch-alum-castile soap canvas treatment. Alum is the least potent biocide of the group and forms the weakest tannate complex. Its primary role in the traditional sequence is as a fixative and precursor to the aluminium stearate waterproofing formation in the soap bath — not as a biocidal agent. The alum-based sequence is a waterproofing treatment with incidental biological protection, not a biological protection treatment with incidental waterproofing.

The Concentration Problem

For all of these salts, working concentration matters considerably more than most sources on the subject acknowledge. The railway sleeper problem — too strong a zinc chloride solution producing brittle, moisture-attracting timber — is the clearest documented example, but the same dynamic applies across the group.

Too high a concentration leaves more free salt in the fibre than the tannin can complex. Uncomplexed salt remains hygroscopic and may leach in wet conditions, depleting the treatment at precisely the moment conditions favour biological decay. Dilute and repeated is more effective than concentrated and once. Working concentrations I have been testing: 1 to 2 percent ferrous sulphate for iron mordanting on pre-tannin-treated canvas; 1 to 3 percent copper sulphate for maximum biocidal emphasis. These are extrapolations from mordant dyeing practice and the historical canvas treatment literature rather than results from extended field trials on canvas specifically, and I am documenting them as starting points rather than conclusions.

The Lubricant Question — Rope Versus Canvas

The Atkins rope paper is explicit about something that is easy to overlook: for rope, the preservative needs to also be an internal lubricant. US Government rope specifications of the period required 8 to 12 percent lubricant content by weight. Rope fibres under load rotate and bear against each other continuously as the rope bends and recovers. Without internal lubrication, this friction degrades the fibres mechanically, independently of any biological attack. A dry compound deposited within the rope — cutch leaves rope "dry and lifeless" as Atkins notes — may address biological decay while accelerating mechanical wear. The tars and tar oils in the best-performing rope treatments are doing preservation work and lubrication work simultaneously. Copper resinate, a dry powder, is noted as performing less well than copper oleate on rope partly for this reason.

Canvas has no equivalent structural requirement. Fibres in a woven canvas are interlocked rather than dynamic — they do not rotate under load in the way rope fibres do. This opens treatment options for canvas that are genuinely closed for rope. A dry mordant bath without oil carrier, an alum and soap sequence, any powder-form compound — all valid for canvas in ways they are not for rope. The performance hierarchy from the rope trials does not transfer directly to canvas treatment decisions, and this note's conclusions about which metal salts perform best in rope are not instructions for canvas treatment.

Practical Treatment Sequences

The pressure impregnation route that Burnett used is not accessible at small scale without industrial equipment. Several approaches can deliver metal salt treatment into canvas or cordage without it.

For canvas, a general mordant sequence: tannin bath first — cutch at 5 to 8 percent in warm water — dried fully. Metal salt bath at 1 to 3 percent concentration — ferrous sulphate for above-deck canvas in conditions where moisture management after treatment is reliable, copper sulphate where maximum biocidal effect is the priority and the treated canvas can be kept off the water long enough to cure. Dried fully. Overcoat with raw linseed oil or a beeswax-turpentine blend.

For cordage in sustained marine conditions, the Atkins data points clearly toward iron soap or copper soap in a tar carrier — not toward cutch-based mordant sequences. Cutch on rope under sustained seawater immersion barely outperforms untreated rope. Stockholm tar without a copper or iron soap component provides meaningful protection, but is outperformed by the soap-tar combinations at every time point in the trials. The most ecologically defensible of the effective treatments is iron soap in Stockholm tar — the Cuprinol Brown equivalent. Making this is possible without specialist equipment and is documented separately.

For substituting ferrous sulphate or copper sulphate for alum at the mordant stage in canvas treatment sequences: colour consequences of iron and copper tannates are permanent. Test on sample panels before committing to a full sail.

Handling, Hazard, and Disposal

Copper sulphate is acutely toxic to fish, invertebrates, and algae at very low concentrations. Zinc chloride and ferrous sulphate carry similar aquatic risks, though ferrous sulphate is considerably less acutely toxic than copper at equivalent concentrations. Newly treated canvas or cordage should be given time — at least several weeks, ideally a full season ashore — before going to work over water. A well-mordanted, well-overcoated piece will leach far less than a freshly treated one without surface sealant.

Zinc chloride in rot arrest concentrations is caustic — skin contact with treated surfaces before curing causes irritation. Gloves during treatment application are not optional. Copper sulphate produces blue-green staining on hands and clothing. Standard precautions throughout.

Spent treatment baths should not go to drains, watercourses, or the sea. Neutralise, evaporate to a solid residue, and dispose of via hazardous waste collection. Ferrous sulphate spent baths are the least problematic of the group for disposal — the compound breaks down into iron hydroxide, a natural soil component — but should still not be discharged to watercourses in concentrated form.

For Wood — Completeness

Zinc chloride remains relevant for timber in one specific application accessible to anyone maintaining a wooden boat: early-stage rot arrest. Applied by brush or injection to timber showing early softening, it arrests fungal activity effectively. It does not restore lost structural strength. The end grain treatment notes cover the sealing sequence that should follow.

Copper naphthenate — copper in an organic solvent carrier — is the form in which copper-based preservatives are most widely used for timber today. The shift from zinc to copper reflects copper's better biological efficacy at lower concentrations and its reduced hygroscopic tendency. The ecological questions around copper in timber treatment are less acute than in direct marine application, but remain relevant wherever treated timber contacts or drains into water.

For intact wood with no sign of active decay, the oil and tar treatments covered elsewhere in this series are more appropriate for preventive maintenance.

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.

References: Burnett's patent prospectus — Sir William Burnett's Patent for the Preservation of Timber, Canvas, Cordage, Cotton, Woollen, etc., from dry rot, mildew, moth and premature decay — is available via Google Books. 

The Victorian Web has a well-sourced article on Burnett covering the patent history and Admiralty trials. 

The US Forest Products Laboratory technical note FPLTN-F022, Preservation of Timber with Zinc Chloride by the Steeping Process, covers the practical chemistry of dilute zinc chloride treatment in detail.

US Forest Products Laboratory technical note FPLTN-F022, Preservation of Timber with Zinc Chloride by the Steeping Process.

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 the Marine Biological Association at mba.ac.uk.

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 the Marine Biological Association at mba.ac.uk.