Wednesday, 2 September 2015

Close Examination of a 19th Century Tahitian 'Poncho'


In early June I wrote a blog post about a 19th-century Tahitian poncho-like garment called a tiputa, which I conserved while on a student placement in the Pitt Rivers Museum’s Conservation department. 

PRM1984.3.1, front, before conservation (left) and after conservation (right)  © Pitt Rivers Museum
Alongside the treatment of the object, I spent some time trying to identify its associated plant materials. I started by looking at ethnobotanical sources listing the plants traditionally exploited by Tahitians at the time the tiputa was made.

From this I discovered, for example, that the fine white barkcloth which makes up the body of the garment was almost certainly made from the inner bark of the paper mulberry tree (Broussonetia papyrifera) - the most commonly used species for barkcloth in Tahiti and also the tree said to produce the palest, highest quality cloth. Other sources of barkcloth traditionally used in Tahiti, including the breadfruit tree (Artocarpus altilis) and two species of Ficus generally produced darker, coarser cloth.

Close-up image of the barkcloth making up PRM1984.3.1 © Pitt Rivers Museum    
Features visible to the naked eye also helped to identify ties on either side of the neck opening. This material had a distinctive elongated, open fibre structure, which could easily be seen without magnification. By comparing it with modern botanical and ethnological specimens I could identify it almost certainly as the inner bark of a Hibiscus.

Close-up image of a neck tie from PRM1984.3.1 © Pitt Rivers Museum    
The tiputa was further decorated with rosettes and decorative trims made of strips of a very shiny, golden-coloured plant leaf material, folded and stitched together. The tiputa was also fringed with long strips of a very thin, almost translucent leaf material. Finally, a yellow powder was loosely dusted over the entire surface.

The remaining plant materials present on PRM1984.3.1, from left to right: rosettes, leaf strips, and powder © Pitt Rivers Museum    
Examining these remaining plant materials with the naked eye, plus researching the existing literature, did not make their identification straightforward - there were several possibilities for each sample. 

Looking at similar objects in other institutions did not make identification any easier as the descriptions of the materials were often contradictory. For example, the rosette material is described in the Pitt Rivers documentation as “folded grasses or cane strips”; whereas in the documentation at the British Museum this material on similar objects is listed as either the leaf of arrowroot (Maranta arundinacea) or coconut palm (Cocos nucifera). From my own research, I also noted remarkable similarities between the rosette material and the leaves from the screw pine (Pandanus tectorius) used in the popular Hawaiian lauhala weaving craft.
Similarly, leaf strips from the Pitt Rivers tiputa were identified in the 1980s as the “cuticle of young banana leaves retted in (probably) stale urine and boiled”; while the accession register for similar objects at the Kew Royal Economic Botany Collection describes the strips as made of sugar cane (Saccharum officinarum) or coconut palm. On objects at the Musee du Quai Branly the strips are identified as seaweed. More recently, Dr Caroline Cartwright, a conservation scientist and specialist in fibre identification, identified similar strips on a Tahitian robe at the British Museum as the epidermis of sugar cane leaf.

The abundance of already detached material available for these remaining fibre types provided me with the opportunity to investigate them using two microscopic techniques.

© Pitt Rivers Museum
First, I looked at the samples using polarised light microscopy (PLM), which allowed me to magnify samples up to 200x and to observe distinctive colour changes of the samples in response to different light settings.  

On the right you can see optical microscope images taken at 200x magnification of the rosette material (top row), the leaf strips (middle row) and the powder (bottom row).

Even greater magnification is possible with scanning electron microscopy (SEM), which I carried out at UCL’s Institute of Archaeology

© Pitt Rivers Museum    

On the left you can see scanning electron microscope (SEM) images taken at 500-600x magnification of the rosette material (top), the leaf strips (middle) and the powder (bottom).

Identification of plant fibres using microscopy is not as straightforward as it may seem, particularly if the samples are from ethnographic or archaeological artefacts. Fibre reference atlases tend to contain images of clean, fresh plant material, while samples taken from museum objects are likely to look rather different as a result of processing, use and ageing.

A high level of expertise is also required to interpret features. Caroline Cartwright kindly looked at my SEM images of samples from the Pitt Rivers tiputa, but she emphasised the need to be realistic in the level of identification which can be achieved: in addition to the problems already mentioned, the taking of microscopic images is subjective and different people will choose to focus on different areas for examination.

Unfortunately, my SEM images for the rosette material and the powder did not contain enough diagnostic detail to allow secure identification. However, for the leaf fringes, Caroline Cartwright confirmed that – like on the British Museum object she examined – they were very similar to reference specimens of sugar cane leaf and definitely not similar to banana leaf specimens.

PLM (left) and SEM (right) images of the stitching thread, at 50x and 200x magnification respectively.    
Tell-tale “ribbon twists” seen in microscope images of the stitching thread used in the Pitt Rivers tiputa’s construction did allow a positive identification to cotton, rather than a type of indigenous cordage. Finding a European imported material used alongside a range of traditional materials in a single object is a fascinating clue to the creative use of material culture in colonial Tahiti.

Finally, for the yellow powder, literature research suggested it may have been used to either colour or scent the barkcloth. Turmeric (Curcuma longa) is a colourant frequently found in Pacific ethnographic material. However, when a tiny sample of the powder was left to react with first an acid and then an alkali, the typical colour changes indicating the presence of curcumin did not happen, so turmeric was excluded as a candidate. Whether the powder could be another type of powdered root, such as ginger, or perhaps a fragrant resinous wood may be shown using yet another technique, Fourier transform infrared spectroscopy (FT-IR), but for now it remains a bit of a mystery.

Naomi Bergmans
Conservation Student