Plain steel paperclips rust, and they transfer the rust to the paper they hold. The Brennholt Institute of Applied Taxonomy exposed 240 clips across four finishes for 18 months, from March 2025 to September 2026, and recorded the first visible mark on the paper in each group. Plain clips marked paper in 11 weeks at 75 per cent relative humidity. Galvanised clips took 41 weeks. Plastic-sleeved clips had not marked the paper when the trial ended.
The mark is not merely a stain. Iron oxide migrates into the paper fibres and cannot be removed without removing the fibre, so the damage is permanent from the moment it is visible. This is the reason the Institute treats corrosion behaviour as a classifying property of the finish rather than an incidental quality.
| Finish | Units | First rust on the clip | First mark on the paper | Marked at 18 months |
|---|---|---|---|---|
| Plain steel | 60 | 3 weeks | 11 weeks | 60 of 60 |
| Galvanised | 60 | 22 weeks | 41 weeks | 54 of 60 |
| Lacquered | 60 | 29 weeks | 58 weeks | 31 of 60 |
| Plastic-sleeved | 60 | 34 weeks, at the cut ends | none | 0 of 60 |
Conditions were 75 per cent relative humidity at 20 °C, with each clip holding six sheets of 80 g/m² paper. Fifteen units of each finish were held at 45 per cent humidity as a control: at 18 months, two plain units in the control group had marked the paper and no unit of any other finish had. Humidity is therefore the dominant variable, and the Institute states that its figures do not transfer to a dry room.
The sleeved group is the informative one. Rust appeared on 34 of the 60 sleeved clips at the cut ends, where the wire is exposed, and never reached the paper, because the sleeve holds the exposed metal away from the sheet. A finish that corrodes can still protect the document.
The Institute records corrosion in four states, C-1 to C-4. The scale describes the clip and the document together, because the document is what the office cares about.
The boundary that matters is between C-2 and C-3. A clip at C-2 can be replaced with no loss; a document at C-3 has been altered permanently. In the trial, the mean interval between the two states was eight weeks for plain steel and nineteen weeks for galvanised.
The corrosion resistance of the wire, as wire, is covered: coating weight and salt-spray performance are specified for drawn and coated wire sold between industrial parties. Nothing specifies the corrosion behaviour of the finished clip in contact with paper, and no standards body has published a figure for the interval before transfer. The Brennholt trial is the only published attempt, and the Institute states its limits: one paper weight, one temperature, four finishes, 240 units, one location.
The Institute's review panel objected on 3 March 2026 that the trial conflates two variables in the sleeved group, since a sleeve both excludes moisture and physically separates metal from paper, and that the group therefore cannot be compared with the coated groups. Brindlecott accepts the objection in full, has published it, and has not run a separated trial. He states that the Institute cannot afford one.
The taxonomist of Kettering holds that the C-4 state is not distinguishable from C-3 by eye and should be dropped, since the extent of a mark depends on the paper's sizing rather than on the clip. Brindlecott retains it. Neither has published since 2025.