A cover bearing this marking was illustrated on page 5 of the August 2001 HKPS newsletter which was said to be unrecorded by our book - Postage Due and the Handling of Underpaid Mail of Hong Kong.

This actually showed a difficulty in the production of the illustrations in our book. Our observation is that several of the markings were made of rubber, plastic, or hybrid of these two materials. On long term exposure (i.e. used for a long time) to the ink, which is oil based and may contain some active solvents. The solvent attacks the rubber/plastic rapidly, while rubber would absorb the oils to a certain extent and swell. Oil based inks were used in the inkpads for cancellation as this was the technology known at the time to make the marking not water-washable. Today we can use water-based inks and on curing, the ink cannot be washed off by water, eventhough the oil-based inks are usually cheaper.

I can write with some authority on this subject as I have post-graduate degree in chemical engineering and I worked over a dozen years in a color concentrate and ink manufacturer.

So what is the difficulty? You may have noticed that we do not give any physical measurements in any of the markings in our book. Were we just lazy? These are the reasons why we did not do it. Firstly, because of the deterioration of the rubber-plastic markings over time and secondly, we are not 100% sure which of the markings is steel and which is rubber-plastic (eventhough we are pretty sure on most of the early markings)?

So with these ever-swelling and ever-distorting markings, it would be misleading to the readers if we have specified an absolute value for a marking (i.e. to say the marking is 25 mm by 38 mm). We also decided that we should use, to the best of our ability, use a very early image of the marking for our illustration in our book.

There is more. Look back at the illustration (in the original article) where you see the characters "CENTIMES". Have you notice that the "S" in the end is the largest of all the shown? Then the "T" and "I" in the middle are the smallest/shortest of all the character. So overall, what was once a straight row of characters has now distorted to form a double curve with swollen end characters. This illustrated that the fact that certain part of the rubber-plastic marking swell at a different rate than other parts. Look again and you can see the curvature is somewhat different on top than it is on the bottom. We see these distortions in many of the rubber-plastic markings that were used as postage due markings.

So now you are getting to better understand the whole story with a second observation, these marking swells over time and the swelling are not uniform causing distortions.

I know that I have not directly answered your question of what happened to the outer frame. Not yet anyway. Has the marking swell so much that the frames are outside of the inkpad where it was usually inked? Or did it swell to a size that was even larger than the whole inkpad? Or is there another reason?

Wait, we have not even considered the third dimension in the swelling of the rubber-plastic marking, the thickness. On a two-dimensional surface such as a piece of paper, we cannot see how much of the swelling to the thickness had occurred with this marking. Or can we? What you saw from the row where it says "CENTIMES" is that the outside swell faster than the inside. With this reasoning, we can probably see why the outermost of this marking, which is the frame, must have swollen most. Now in 3-dimemsions, what this mean is that the middle where the "T" and "CENTIMES" have not swollen as much.

If you apply this marking with a thick outside frame, chances are the middle would not contact the inkpad and subsequently do not appear on a flat surface like a cover. So has this marking swollen to such a stage? Where the outside frames eliminated so that the rest of the markings can show?

Probably, and we have yet another observations from the illustration. The center of the whole marking is where the lower center of the "T" is, would probably swell the least. The illustration show that the lower part of the T is not inked (least swelling in thickness and therefore no contact). Also noticed the thickness of the body of the "T". It is thicker on top near the horizontal bar and thinner towards the lower part of the body (top is outside which swelled more, the lower is towards the center).

I am writing this more to spark your discussion on the subject, as I am not one who would ever think that I could come up with the final word. I like to hear your response on my reasoning.

Furthermore, I have repeatedly stated in my other articles that if we are to approach philatelic research from a scientific point of view and not from a point "because so-and-so say so and he is an expert", then we must see multiple examples to sustain a new finding. One example is not a proof of anything.

In conclusion, all I have to say that some things in philately can fool our eyes, but not necessarily our brains.