Collection: Field Notes - Regenerative Materials | 

Series: Natural Ropes |

Reading the natural cordage preservation against rot trials carefully — what they tested, what they found, and what the numbers actually mean

A piece of evidence about preventing rot in natural cordage worth examining properly

Every post in this series that touches on preservation treatment eventually comes back to the same two papers. The 1936 Atkins and Purser seawater immersion trials, and the 1928 Atkins fishing nets paper that preceded it. I have cited them in support of specific claims throughout — on bacterial decay rates, on the performance of Stockholm tar, on the failure of cutch as a standalone treatment, on the lubricant requirement. It is probably time to look at them directly rather than through the lens of the points they happen to support.

The reason I keep returning to these papers is not that they are perfect. They are not. The sample sizes are small, the methodology has acknowledged limitations, and the site conditions of the Plymouth Pier tests were admitted by the authors themselves to be "exceptionally severe." Reading them carefully, rather than mining them for convenient data points, reveals both how much they establish and where the gaps are. The gaps matter as much as the data for understanding what can actually be claimed on the basis of this evidence.

The 1928 nets preservation against rot paper: where the rope trials begin

The 1936 rope trials are a continuation of work Atkins had been doing on fishing net preservation since at least 1925. The 1928 paper is the most complete account of that earlier work, and reading it first helps explain both the methods and the reasoning behind the rope trials.

Atkins was working from an American starting point. Taylor and Wells at the US Bureau of Fisheries had introduced copper oleate dissolved in benzene or petrol as a net preservative in 1923, and the results were striking enough that Atkins set out to confirm them under British conditions and then extend the investigation. His method was systematic in the way that good field biology tends to be systematic — include an untreated control in every series, run comparisons simultaneously rather than sequentially, be explicit about the conditions and their limitations.

The test setup was simple and acknowledges its own artificiality honestly. Nets approximately a square metre in size were placed in jars of seawater, the water changed three times a week, and the nets allowed to rot. Atkins notes the ways this differs from actual fishing conditions — the net in use is subjected to a stream of water that leaches preservative faster than a static jar, but also undergoes mechanical wear that the jar tests do not replicate. The jar test is more severe than practice in some respects and less severe in others. He does not pretend these complications away.

The contamination question is where the 1928 paper contains an observation I find unexpectedly illuminating. Atkins discovered mid-programme that tank water from the aquarium produced faster rotting than outside clean seawater, and he had switched to tank water partway through without realising at the time how significant the difference was. He is candid about this. The early tests using outside water and the later tests using tank water are not directly comparable, and he says so. This kind of methodological honesty — documenting an error and its consequences rather than smoothing the data — gives the work a credibility that a cleaner presentation would not.

The key findings from the nets paper that feed directly into the rope trials are these: copper soap treatments consistently outperformed all alternatives by a substantial margin; cutch alone was nearly useless; Olie's Dutch method — cutch followed by ammoniacal copper sulphate — performed well on hemp though less well on cotton; and tar in combination with copper soap significantly improved results over copper soap alone. Taylor and Wells in America had found copper oleate and coal tar to be the best combination tested. Atkins found the same thing under British conditions with somewhat less tar in the mixture.

The 1936 preserving rope trials: methodology

The rope trials move from nets to actual rope and from jar immersion to field exposure in real seawater. This is a meaningful step up in ecological validity — the ropes are in actual Plymouth Sound, subject to tidal currents and wave action, not in controlled jars. It is also a step down in experimental control, because the conditions vary in ways that jar tests do not.

The Plymouth Pier site is described in detail that is worth sitting with. The pier is downstream from the main sewer outfall of Plymouth. The water is contaminated by drainage from the pier itself. The bacterial load is high. The authors describe this as an "exceptionally severe" test site, which they knew at the time and chose deliberately — they wanted to know whether preservatives could hold up under the worst plausible conditions rather than in the best.

The Cawsand fish pond was included as a contrast — a tidal basin outside Plymouth Sound Breakwater, water much cleaner, ropes always immersed rather than exposed to tidal cycling. The comparison between the two sites is one of the more useful aspects of the trial design, because it separates the effect of bacterial contamination level from other variables. Clean water was less severe. The treatments that performed best at Plymouth Pier generally performed well at Cawsand too, though the absolute numbers were different.

The ropes were clamped between wooden strips and suspended from the pier columns. The position meant they hung clear of the water for approximately four hours each tide. This tidal cycling — wet and dry, wet and dry — is more representative of how deck rope actually lives than continuous immersion would be, and it is also more demanding, because each drying cycle deposits salt crystals in the fibre and each rewetting starts the bacterial action again.

Tensile testing was done at Trinity College Dublin, using a ten-ton Buckton Vertical Testing Machine. The end-fixing method — pushing a nail into the rope endwise and serving the end with twine, then soaking in glue — was developed specifically for these tests and validated against a series of twelve tests on good manila rope. The mean breaking load across those twelve tests was consistent enough that the authors concluded the method was "sufficiently good to enable the results to be relied upon for approximate comparisons." Approximate comparisons. That qualification is in the paper and it is important.

Treating natural ropes with preservative: What the results show

The headline finding is stark: untreated rope in contaminated seawater has a working life measured in months. At Plymouth Pier, two-inch hemp and manila rope reached zero retained strength within twelve months. The untreated 0.6-inch manila in the Cawsand fish pond was at 13% retained strength after ten and a half months. Both are effectively destroyed.

Against this baseline, the treated ropes look dramatically different. The best performers — green Cuprinol with tar, copper oleate in Corroid tar — maintained strength above 70% after ten months at Cawsand. Several treatments maintained strength in the 60–70% range. A small number, including copper resinate alone, fell below 50% or failed entirely.

The results I find most useful to look at closely are the ones that complicate the simple ranking. Copper resinate performed well early and poorly later in several tests — it maintained high strength at the six-month point and then dropped sharply. The authors suggest this reflects leaching: the resinate treatment deposits a biocide that washes out faster than the tar-based treatments, which physically protect the fibre surface as well as depositing toxic compounds within it. A treatment that performs well at six months and fails at twelve is not the same as a treatment that performs consistently well throughout, and the trial period is long enough to reveal the difference.

Stockholm tar sits in an interesting position in the results. At Plymouth Pier over twelve months, it retained 57% strength on hemp against 78% for coal tar. That gap is real. But Stockholm tar dried more completely than coal tar — the authors note this observation separately — which matters for a rope that is stored between uses rather than continuously immersed. A treatment that is effective and dries is more useful for working rope than a treatment that is marginally more effective but remains tacky in storage.

The cutch results are perhaps the most telling for the gap between traditional practice and actual performance. Cutch was widely used in British net preservation at the time of the trials, largely because it was cheap and accessible. The trials found it nearly useless. Two boilings of cutch gave marginally better retention than no treatment at all — marginally, within the range of experimental variation. Olie's method, adding ammoniacal copper sulphate after cutch, performed considerably better. But the improvement came from the copper sulphate, not the cutch. The tannin pre-treatment may have helped copper uptake on hemp, but the cutch alone was providing almost nothing.

Pine Tar: The variable that keeps appearing

Something in the tar results in the 1928 nets paper catches my attention every time I read it. Two tar treatments, applied identically, produced very different results — one gave 26 months of life, one gave 11. The net that failed was treated with what Atkins describes as probably an over-distilled tar, one in which distillation had been pushed too far and left a brittle, depleted product. He cannot prove it. He says "one can but suggest."

This kind of result — two nominally identical treatments producing dramatically different outcomes because the raw material varied — shadows the rope trials too, even though the authors do not explicitly flag it there. The rope trials used tars and copper soap preparations from named commercial suppliers, which provides some consistency. But the principle that the quality and composition of the input material matters enormously for the output performance is not unique to the 1928 tar batch. It applies to everything in the treatment chemistry, and it is a reason to be cautious about reading too much precision into the percentage numbers.

If a tar treatment can perform at either 26 months or 11 months depending on which tin you open, then the difference between Stockholm tar at 57% and coal tar at 78% in the rope trials is not necessarily a stable, reproducible difference. It may be partly a difference between the specific batches used in the specific year the tests ran. I do not know this. I am not saying the results are unreliable. I am saying the precision implied by exact percentage numbers may be misleading, and that the trials themselves contain evidence for treating their own numbers with appropriate caution.

What the trials cannot tell you

The trials were conducted in Plymouth Sound in 1933–1935, using rope from British suppliers, with treatments prepared to specifications described in the papers. The conditions, the rope, and the treatments are all specific. The results are the results for those specific things in those specific conditions.

Extrapolating from them to what will happen to your rope on your boat requires accepting several assumptions: that your rope is comparable in quality and construction to the trial rope; that your water conditions are comparable to either Plymouth Pier or Cawsand; that your treatment preparations are comparable in concentration and penetration to the trial preparations; and that the treatment chemistry has not changed significantly since 1935. Some of these assumptions are more defensible than others.

The trials also do not test the effect of maintenance retreatment. All the ropes were treated once at the start of the trial and then left. Working rope that is retreated annually is not what the trials describe, and it is reasonable to expect that annual retreatment would improve on the single-treatment results — though by how much, and whether the ranking of treatments would change under a retreatment regime, the trials cannot say.

There is also the question of what has changed in the century since the trials. Bacterial communities vary by location. Water quality in Plymouth Sound today is different from Plymouth Sound in 1935 — cleaner in some respects, the sewage outfall long since relocated, though the harbour is now busier with marine traffic. Whether the trials' severity represents current British harbour conditions or significantly exceeds them I cannot say. It is another variable in extrapolating from the numbers to practice.

What they do establish

Despite all of that, the trials establish several things clearly enough that I am comfortable treating them as working foundations rather than weak suggestions.

Untreated natural rope in bacterially active seawater has a very short life — months, not years. This is not a marginal finding likely to be overturned by better methodology. The untreated controls in every series consistently reached failure within two to six months depending on season and conditions. There is no version of this result that makes untreated rope look like a viable long-term choice in marine environments.

Chemical treatment makes a very large difference. The gap between untreated controls and the best-performing treatments is not 10 or 20 percent — it is the difference between failure in months and retained strength above 70% after a year. The mechanism — physical exclusion of water plus biocidal activity within the fibre — is sound chemistry, and the trial results are consistent with it.

The lubricant function matters independently of the biocidal function. The trials do not isolate these cleanly, but the US Government specification for lubricant content is cited because it exists and because internal abrasion was understood as a separate failure mechanism. Treatments that provide both functions outperform treatments that provide only one, which is consistent with the mechanism.

Stockholm tar is effective, ecologically more defensible than the top performers, and practically superior to coal tar for working rope because it dries. These things together are enough to make it the default treatment choice in a natural materials system, while being honest that it is not the highest performer in the evidence base.

The trials are the best systematic evidence available for what natural rope preservation treatments do under real marine conditions. They are imperfect, they are specific, and they are ninety years old. They are also still the most useful single source I have found. That combination — essential and insufficient — is where most good evidence eventually ends up.

Sources: W.R.G. Atkins and J. Purser, The Preservation of Fibre Ropes for Use in Sea-Water, Journal of the Marine Biological Association of the United Kingdom (1936). W.R.G. Atkins, The Preservation of Fishing Nets by Treatment with Copper Soaps and Other Substances, Journal of the Marine Biological Association of the United Kingdom (1928). H.A. McKenna, J.W.S. Hearle and N. O'Hear, Handbook of Fibre Rope Technology (Woodhead Publishing, 2004).

Plans at VAKA Boatplans | Full knowledge base at Field Notes