by Lin Yangchen
        Laboratory of Computational Philately
        Coconut Academy of Sciences of the United States of Malaya


Assessment of paper thickness is sometimes called into service in the study of the coconut definitives, such as when identifying striated paper or detecting anomalies like the “very thick paper” reported for an unissued 8¢ scarlet.

The thickness and stiffness of the paper are usually felt with the fingers. But this soils the stamp and is subjective. For example, is the “very thick paper” really a distinct variety, or is it just part of a Gaussian continuum? How sure is one that it feels thicker because it is actually thicker or because it’s just stiffer? Bending resistance is a property distinct from, albeit related to, thickness as defined by the Technical Association of the Pulp and Paper Industry.

For the first time, the paper thicknesses of a large sample of coconut definitives have been measured using a micrometer screw gauge. This not only is much more accurate and precise but also makes statistical analysis and quantitative comparison possible. This can sometimes help philatelists decide whether the thickness differences constitute distinct paper types or are just manufacturing inconsistencies in the same paper type (Barwis 2013).

Measuring the thickness of striated paper the traditional way at the Royal Philatelic Society London using a Moore & Wright micrometer screw gauge with tungsten carbide measuring faces. You turn the thimble gently until your fingers feel a minute but distinct increase in resistance. For this study, I used a cheaper Eisco Labs screw gauge. The measurement area of the anvil was about 26 mm², which evenned out small-scale surface irregularities.

Paper fibres can expand with moisture. Prior to measurement, stamps were removed from the dry cabinet and allowed to equilibrate for several minutes to the humid tropical atmosphere. Air temperature and relative humidity were recorded using a room thermometer and an analogue hygrometer. This would allow any effects of atmospheric conditions to be factored in.

The measurement was made at an area of the stamp with no postmark, overprint, abrasion or crease. All mint stamps measured had intact gum while all used stamps measured had no gum in the area of measurement. Raw data can be downloaded here. Analysis was automated using the R Language for Statistical Computing.

Barwis (2013) reported, for a recess-printed stamp, that the raised ink prevented accurate and precise paper thickness measurements. For the coconut definitives, however, I found from measuring the uninked selvedges and inked stamps of the same marginal blocks that the letterpressed ink had no measurable effect on paper thickness.

Thickness distributions of the four main paper types of the coconut definitive. I regret the small and uneven sample sizes which at this time precludes the study of possible effects of various other factors recorded in the data. There are also potential sources of environmental variation such as shrinkage or expansion of the paper with age as the fibres dehydrate or absorb moisture, or flattening caused by album storage. Nevertheless, the concept of analytical philately has been proven.

Used stamps tend to be slightly thinner than mint ones, probably because the gum is gone and some of the coating could have been lost through rubbing and/or soaking. This has never been definitively (pun intended) observed before, since one needs accurate measurements of multiple stamps to discern the difference.

Chalky, rough and substitute papers appear similarly thick, although the small sample sizes make it hard to tell. Striated paper is the thinnest. This is the first time to my knowledge that the thicknesses of all four papers have been compared with one another.

Paper thickness can also vary across a single stamp; see paper cross-section microscopy.

For relative contributions of paper fibres, gum, coating and ink to thickness, see paper cross-section microscopy.

References


previous section | next section | back to table of contents