Object Analysis

Detailed technical analysis of objects enabled us to investigate the materials and processes involved in making barkcloth. Analytical methods were refined and combined to support both fibre identification and pigment identification.

Fibre identification

Identifying the plant species used to make barkcloth has always been challenging – the fibres are flattened by the beating process so can’t be identified by looking at them through a microscope, as is usually done with plant fibres. Dr Margaret Smith, our Scientific Research Associate, developed a new technique for barkcloth species identification using Fourier transform infrared spectroscopy with attenuated total reflection (ATR-FTIR) and principal component analysis (PCA) to differentiate between historic barkcloths, as published in Smith, Holmes-Smith and Lennard (2019).

A set of 22 contemporary and 79 historic cloths was analysed to develop this methodology. In this preliminary study of contemporary and historic barkcloth, multivariate analysis of FTIR spectra in the 1200–1600 cm−1 region has been shown to be useful in grouping historical barkcloths originating from different species. PCA analysis identified three groups for the historical cloth with the loading plots highlighting where the differences between the FTIR spectra are predominant for each principal component. In addition, employing hierarchical cluster analysis (HCA) to analyse the data identified cloths that have a close relationship to each other and showed a clear link between the cloths that are thought to be composed of mixed fibres. This shows the usefulness of this statistical technique to historic bark cloth analysis.

The feel of barkcloth is often as important as visual clues in deciphering its material and manufacture, but we found that visual and textural differences in the cloths are often more dependent on the processing method than the fibre type. We also found that comparisons between historic and contemporary barkcloth should be treated with caution, as tests on old and new fibres weren’t always corroborative.

Pigment identification

The dyes and pigments on a relatively large sample set of 54 barkcloths were analysed using high performance liquid chromatography (HPLC) for organic materials and X-ray fluorescence (XRF) for non-organic pigments, as published in Flowers, Smith and Brunton (2019).

Colour photograph showing project team members, Margaret Smith and Misa Tamura, at work in the conservation laboratory taking samples of a barkcloth object for analysis (copyright University of Glasgow)
Margaret Smith and Misa Tamura taking samples of a barkcloth object for analysis

Five different colourants were found on the cloths examined. These included a small number of plant derived colorants: noni (Morinda citrifolia), madder (Rubia tinctorum), turmeric (Curcuma longa) and tree tannins. One inorganic colorant, iron oxide, was found.


Noni can be used to make both red and yellow dye. It has two chemical forms – morindin and morindone. The noni based dyes (both morindin and morindone) are naturally yellow but the colour is pH dependent, and it is red above pH 10. Materials such as burnt lime or burnt coral (calcium hydroxide) can be used to raise the pH and to create a red dye. Of the barkcloth samples analysed, the samples coloured with morindone were all red, whereas morindin samples were both red and yellow. Only one cloth of those analysed contained both red and yellow dyes from noni: GLAHM E.595/1, from The Hunterian collection, which is attributed to Tahiti.

Colour photograph of a large, stamp-decorated Tahitian ahufara cloak from The Hunterian collection, reference GLAHM E.595/1 (copyright The University of Glasgow)
A large, stamp-decorated Tahitian ahufara cloak from The Hunterian collection, GLAHM E.595/1


The tannins found in the cloths come from the resinous sap of trees such as Aleurites moluccana (tuitui, candlenut), Bischofia javanica (koka, o‘a, bishopwood) and Rhizophora mangle (tongo, mangrove). A liquid colourant is extracted from shavings of the outer sapwood. Tree tannins were also used historically as mordants for other colourants. Our investigation showed that they were found both as the sole colourant, and in conjunction with other colourants on the cloths examined. The saps produce a browner-red colour than noni, which gives a true red.

Iron oxide

Iron oxide, found in a small number of barkcloths, is a component of an ochre, or a natural earth pigment. These are derived from the soils of the islands where the cloths were created. A red earth, from Hawai‘i was found to have been used at different concentrations to create both yellow and red pigments on the same cloth (Hunterian GLAHM E.602).

Colour photograph of a Hawaiian painted cloth from The Hunterian collection, reference GLAHM E.602 (copyright The University of Glasgow)
A Hawaiian painted cloth from The Hunterian collection, GLAHM E.602

Results in context

Given the numerous colourant sources for the red, brown and yellow shades listed in the many historic accounts of barkcloth making, our identification of only four plant sources (principally noni, tannin and turmeric with a small amount of madder) and one pigment, iron ochre, is both interesting and in line with complementary research carried out at the British Museum (Tamburini et al., 2019).

However we need to set this in context: cloths in European collections were collected in particular places at particular times and they represent a tiny proportion of all cloths made in Oceania in past centuries. Nonetheless, this result demonstrates that the historical literature needs to be treated critically, like any other source.

The research also showed that contemporary barkcloth practice cannot be used uncritically as a guide for understanding artefacts made in the past – analytical investigation of historic and contemporary materials was not always corroborative – confirming the value of an interdisciplinary approach.