©Lin Yangchen

previous section | next section | back to table of contents

The stamps were letterpress-printed, usually in sheets of 100, in London from electrotyped copies of the original engraved and hardened steel dies. Letterpress is the oldest method of making many prints from a single original. It is thought to have been invented by makers of playing cards, who engraved their designs on blocks of wood (Blum 1940). During the time of the coconut definitive, De La Rue printed their letterpress stamps on a Wharfedale press, in which the printing plate moved under a revolving cylinder around which the paper was wrapped. See Yendall (2008) for technical details of this specialized stamp-printing equipment.

First the key plate (vignette) in the centre was printed, followed by the surrounding duty plate. This is curious since the key plate is intuitively a foreground object. The following close-ups show the duty plate overlaying the key plate, along with the characteristic accumulation of ink along the edges of the letterpressed pattern (Williams & Williams 1969).

Printing from interlocking plates is thought to have been invented in the early 19th century on the Strand in London. It was first used on stamps by De La Rue in 1863. This made it easier to print bicoloured stamps (Lowe 1979a), update the monarch and mix and match plates. The use of a single key plate for all denominations also saved raw materials and labour. These printing plates like all others were prone to accidental damage, producing plate flaws (see Hale 1991, Tyre & Stanway 1991, Glover 1994).

The key plates were sometimes misaligned—here shown too high and too low—despite the fundamental importance of accurate registration (alignment) in security printing (Fernbank 2013) and the valiant efforts of De La Rue to overcome this 'enormous difficulty' since the earliest days (Easton 1949).

Registration standards dropped further during the post-war British Military Administration period.

One trait of this 1946 Chinese Civil War-era Sun Yat-sen issue may be considered superior to the Die I coconut definitives: the elliptical frame around the portrait is engraved on the duty plate rather than the key plate in the Chinese stamp, which guarantees perfect centering of the ellipse with respect to the duty plate. Had Die I of the coconut definitive been designed this way, mis-registrations of the plates would have been less obvious. Many of the engravings of the British monarchs, however, had already been made incorporating the elliptical frame prior to the advent of the coconut definitive; the designers were hence likely to have been bound by historical, logistical and physical constraints.

A printing anomaly first discovered by James Kennedy (Lin et al. 2022). The print sometimes exhibits fine horizontal stripes that remind one of the raster scan of analogue television. This is most visible on the king's hair. The stripes occur in at least two spatial frequencies. On the 12¢ (left), there are about 12 stripes per vertical millimetre, while on some other denominations (right) there are 9 stripes per millimetre, measured using calibrated high-resolution images.

Lin et al. also encountered a marginal 5¢ where the head plate number on the selvedge also has the stripes, suggesting that they affected the entire plate. The Jubilee line of the head plate has the stripes while the Jubilee line of the duty plate apparently does not.

The straightness and evenness of the stripes and their spacing indicate that the pattern was machine-made. The visibility and crispness of the stripes seem to vary with ink consistency, the amount of ink used, printing pressure et cetera. The back of the stamp does not show anything out of the ordinary. The stripes do not seem to have affected the oval rim of the vignette, but it could be because the relatively heavy inking of the vignette rim obliterated them.

Perhaps the cylinder in the Wharfedale press had a ribbed surface to grip the paper better. Perhaps there were two kinds of cylinders, with and without ribs, or a cylinder that was ribbed along only part of its length. This would explain why only some stamps have the stripes, and why the affected stamps have the stripes on either the head plate or duty plate but not both.

Another theory is that the stripes were intentionally added to prevent excessive inking of the king’s hair (Robert Hisey pers. comm.). But how these stripes were added to the plate remains a mystery, and their usefulness is questionable given that many other stamps in the set seem fine without them. Maybe it was an experiment.

There is yet another possible if tenuous explanation. After electrotyping the head plate, the shell, or metal layer deposited on the mould, might have been found to be insufficiently flat for printing. It could have been flattened by laying it face-down on a flat surface and working it with a hammer and other tools, or perhaps it was brushed using a machine. This process is known as slabbing. The stripes might have been made by the surface or tool that came into contact with the shell. Perhaps the king’s head protruded more than the vignette rim in the original die and this unevenness was transferred to the electrotype, which would explain why the head got the stripes and apparently not the rim.

Post-war stripes found by the author following reports by Rein Bakhuizen van den Brink of more widespread occurrence of the phenomenon.

A serious engraving error in the high-value coconut dies was spotted and corrected in the nick of time (pun intended), just as the printed stamps were about to be issued. A Chinese clerk in Singapore reportedly raised the alarm after seeing the dollar sign placed after the numeral on the dollar denominations e.g. '1$'. The stamps were reprinted and the misprinted sheets destroyed (Wood 1948, Barker 1977). Proofs survive in the Crown Agents Philatelic and Security Printing Archive and in private collections (Singapore Stamp Club 2000, Spink auctions 16 August 2015 and 28 October 2017).

The coconut definitive was finally born to the outside world on New Year’s Day 1936. It went to the post offices in sheets of 100 (10 by 10), which had a backing of gum arabic.

Some highly used monochromatic denominations were subsequently printed using a single die referred to as Die II (Skinner & Peters 1988), such as this unissued 8 cents whose colour became invalid after changes in Universal Postal Union postage rates. The single-die printing method, often used on sheets of 200 stamps, was more cost-effective for large production runs (Easton 1949); the medallion and its surrounding artwork now do not overlap. Notice that the medallion in Die II is set nearer to the ground than in Die I. The die was soon discovered by philatelists to considerable fanfare (Rang 1939). The coconut definitive thus came to embrace both key-plate and key-type technology.

Sometimes seen at the top or bottom of a pane of 100 stamps are the so-called "bars", "pillars" or "space fillers", which occupy the gutter between panes in a 200-set or 400-set plate used for some Die II issues (Rosevear 1983, Cockburn 2016a). The purpose of these bars could have been of preventing the ink roller from damaging the edges of the stamp design (Williams 1990), or reducing ink wastage or plate fouling by preventing ink from smearing the entire gutter as it was applied across the plate. Sometime after the war, the panes were separated for further use and the bars removed (Cockburn 2016a).

Believe it or not, the coconut definitives actually had at least eight different printers, excluding materials not used postally:

De La Rue
Bradbury Wilkinson
Secura Singapore
Koninklijke Joh. Enschedé
Beijing Stamp Printing House
National Printing Bureau, Japan

See wartime, aerogrammes and commemoratives.

I am grateful to Rein Bakhuizen van den Brink for comments that improved the article.


previous section | next section | back to table of contents

Powered by SmugMug Owner Log In