©Lin Yangchen
previous section | next section | back to table of contents
Postage stamps are almost always conceptualized as a two-dimensional plane. That’s fine if you’re putting together a basic stamp collection. But there comes a time when you have to tell apart two very different stamps that look almost exactly identical on the surface (pun intended) or determine if a stamp has been forged or tampered with. By adding the third dimension, the scope for understanding and explaining previously unexplained phenomena increases dramatically, such as in paper identification and print characterization.

Paper samples for surface profilometry (Lin 2020c).

A Keyence representative demonstrating the acquisition of a high-resolution digital elevation model of the surface of striated paper using the firm's industrial microscope on a separate occasion at the Royal Philatelic Society London. The objective contains fluorite and extra-low-dispersion lens elements.
The stamps were placed under a Keyence VHX microscope and kept flat on the microscope stage with glass slides to prevent curling, exposing just the area of interest. The target area on each stamp was the large uninked area beside the foot of one of the coconut palms. The surface of the stamp was perpendicularly lit, and a photomicrograph was obtained from the reflected light. The microscope takes many micrographs while moving vertically through different focal planes 0.1 μm apart. On each focal plane, only certain parts of the stamp are in focus. The microscope’s software estimated the height of a given point on the stamp by finding the focal plane where the point showed maximal contrast in the image, which indicated the point was in focus.

Contour maps of paper surfaces generated from focus stacking. You can make out two main categories, one showing fibre morphology on the surface and the other not. The actual area covered by each panel is about 0.1 mm2. This magnification was chosen as a compromise between covering a large-enough area and recording fine-enough detail to encompass multi-scale topographical features. Although attempts were made to keep the stamp flat during observation, there appeared to be slight tilting or warping in some specimens. The color gradient is mapped onto the entire range of recorded elevations within each panel, so colors should not be compared across panels. See Ismail et al. (2020) for surface profilometry of tissue paper using laser scanning confocal microscopy.

Raw
height data was also extracted from five roughly evenly spaced horizontal transects across each of the contour models.
Aerogramme paper is the roughest, its profiles resembling mountain ranges. Aerogrammes were meant to be as thin and light as possible, so it is not unreasonable to suppose the coating was omitted for that reason. Striated paper, rough paper, printer’s waste and postal stationery card are all on the rough side, but less so than aerogramme paper. The roughness is generally due to lack of a coating and exposure of the fiber network.
The coating of chalky paper makes it less rough than the aforementioned papers on the scale of the fibres, although it can be deeply cratered as in sample Ch2. It turns out that at a smaller scale than the fibres, chalky paper exhibits roughness in the form of granules, but the resolution here is insufficient to capture this.
Substitute paper looks rough, but it's a "slippery rock" kind of unevenness, not the fibrous kind of roughness seen in the uncoated papers. This has a particular effect on its print characteristics.
Of all the papers, the essay paper is the smoothest. The print quality of the essays is very crisp, but should not be directly compared with the issued stamps, as they were lithographed while the stamps were letterpressed.
Meanwhile, the surface of the back of the stamp reveals hidden varieties in a little-documented facet of papermaking, which was first reported by Rein Bakhuizen van den Brink (see van den Brink & Lin in press).
In a papermaking machine, a fine woven sieve drained water from the pulp, leaving a thin layer of cellulose fibres to form the paper. The sieve was woven using standard patterns from the textile industry, and left a faint pattern on the so-called wove paper. At first, a linen or plain weave was used. This left a symmetric diamond hatched pattern on the paper. In the late 1930s, papermakers began to use the stronger twill weave, which left a pattern that was asymmetric in both the angles and spacings.

Linen weave pattern (left, KGV 5¢) and twill weave pattern (right, Johore $2). Even without gum, the pattern is often difficult to see with all the interference from the paper fibres and watermark. Because of its asymmetry, the twill pattern can also occur in reverse to that shown. Furthermore the twill weave tends to get stretched more at the ends of a long sieve. This may cause the pattern to deviate slightly from the horizontal, as seen on some of the stamps.
Most of the pre-war coconut definitives have the linen weave pattern, including those on striated paper and rough paper. The substitute paper from the immediate post-war period also has the linen weave pattern. Surprisingly, the twill weave pattern occurs on the pre-war 30¢, as discovered by van den Brink. There is no straightforward explanation for this, as the 30¢ was not a particularly late printing. As expected, most post-war coconut definitives have twill weave patterns.

But things get interesting again. Some of the post-war stamps have the linen weave pattern, and there are two variants. On some of the stamps the hatching has the same frequency as on the pre-war stamps, where the number of lines per horizontal/vertical centimetre is 30/20. But on other stamps, the frequency is 24/20 (above). In this variant with more equal spacing of the horizontal and vertical threads, the gradients of the diagonals are −50°/50° instead of −60°/60°. The 24/20 linen weave pattern could have been from a different manufacturer or simply the same manufacturer using different sieves.
Acknowledgements
I am grateful to David Beech, Rein Bakhuizen van den Brink, Julian Tan and Edmond Soh for discussions and technical assistance, and to Keyence Singapore for the use of their equipment.
References
previous section | next section | back to table of contents
previous section | next section | back to table of contents
Postage stamps are almost always conceptualized as a two-dimensional plane. That’s fine if you’re putting together a basic stamp collection. But there comes a time when you have to tell apart two very different stamps that look almost exactly identical on the surface (pun intended) or determine if a stamp has been forged or tampered with. By adding the third dimension, the scope for understanding and explaining previously unexplained phenomena increases dramatically, such as in paper identification and print characterization.

Paper samples for surface profilometry (Lin 2020c).

The stamps were placed under a Keyence VHX microscope and kept flat on the microscope stage with glass slides to prevent curling, exposing just the area of interest. The target area on each stamp was the large uninked area beside the foot of one of the coconut palms. The surface of the stamp was perpendicularly lit, and a photomicrograph was obtained from the reflected light. The microscope takes many micrographs while moving vertically through different focal planes 0.1 μm apart. On each focal plane, only certain parts of the stamp are in focus. The microscope’s software estimated the height of a given point on the stamp by finding the focal plane where the point showed maximal contrast in the image, which indicated the point was in focus.



Aerogramme paper is the roughest, its profiles resembling mountain ranges. Aerogrammes were meant to be as thin and light as possible, so it is not unreasonable to suppose the coating was omitted for that reason. Striated paper, rough paper, printer’s waste and postal stationery card are all on the rough side, but less so than aerogramme paper. The roughness is generally due to lack of a coating and exposure of the fiber network.
The coating of chalky paper makes it less rough than the aforementioned papers on the scale of the fibres, although it can be deeply cratered as in sample Ch2. It turns out that at a smaller scale than the fibres, chalky paper exhibits roughness in the form of granules, but the resolution here is insufficient to capture this.
Substitute paper looks rough, but it's a "slippery rock" kind of unevenness, not the fibrous kind of roughness seen in the uncoated papers. This has a particular effect on its print characteristics.
Of all the papers, the essay paper is the smoothest. The print quality of the essays is very crisp, but should not be directly compared with the issued stamps, as they were lithographed while the stamps were letterpressed.
Meanwhile, the surface of the back of the stamp reveals hidden varieties in a little-documented facet of papermaking, which was first reported by Rein Bakhuizen van den Brink (see van den Brink & Lin in press).
In a papermaking machine, a fine woven sieve drained water from the pulp, leaving a thin layer of cellulose fibres to form the paper. The sieve was woven using standard patterns from the textile industry, and left a faint pattern on the so-called wove paper. At first, a linen or plain weave was used. This left a symmetric diamond hatched pattern on the paper. In the late 1930s, papermakers began to use the stronger twill weave, which left a pattern that was asymmetric in both the angles and spacings.


Most of the pre-war coconut definitives have the linen weave pattern, including those on striated paper and rough paper. The substitute paper from the immediate post-war period also has the linen weave pattern. Surprisingly, the twill weave pattern occurs on the pre-war 30¢, as discovered by van den Brink. There is no straightforward explanation for this, as the 30¢ was not a particularly late printing. As expected, most post-war coconut definitives have twill weave patterns.

Acknowledgements
I am grateful to David Beech, Rein Bakhuizen van den Brink, Julian Tan and Edmond Soh for discussions and technical assistance, and to Keyence Singapore for the use of their equipment.
previous section | next section | back to table of contents
