The Procedure for Finding the Plate Positions of the 10 Lepta of the Large Hermes Heads is Successful!

by KOstas Hatzis

(This article was originally published in Lighthouse of the Philatelist, Journal of the Philatelic Society of Patras, Vol. 1, Issue 1, October - December 2007. It is important in the development of this plating program because it demonstrates that an independent user could use it to successfully plate over 1500 ten lepta stamps. We hope that KOstas will continue his efforts in plating the 40 lepta and that other researchers will use the program to plate the other values.

A. Introduction

My occupation with the large Hermes head stamps during the past six years is perhaps the most intense but also the most satisfying activity that I have ever experienced. It is an illness that I hope will never be cured.

Delving into information concerning the large Hermes head stamps and expanding my collection of these stamps, I arrived at a point where I had searched high and low for definitive characteristics of the impression on every piece of paper: scratches, white spots, ink spots and other characteristic marks. These identified each stamp as belonging to one precise position out of the entire 150 positions of the plate. Thus, inevitably, searching bibliographies for errors of the plate, I came across the two volumes, studies of the 10 and 40 lepta by Lou Basel. Admittedly, I was amazed at their validity and precision. These are the events which drove me to communicate with Lou and to arrive at the point today of testing an electronic method for finding the plate positions of these stamps.

B. To the Point

Starting with the 10 lepta, and accumulating about 4000 pieces, I was able to recognize most of their plate positions. And I say most of them because I found many stamps of the first issues and a few of the following which were impossible to identify from their printing characteristics because of the clarity of their impression. Therefore, I communicated with Lou. He reported to me that he had written a (computer) program many years ago, which ran on a very old computer. Although it was able to identify the plate positions of these difficult stamps, he feared that if any of his obsolete hardware should fail, it could not be replaced and the program would be unusable. He said that he had thought of introducing a similar program, which would use modern programs based on the Windows operating system. These were the CorelDraw and Excel programs, which, coincidently, I owned and was adept at using. Lou immediately started working on this, and he had the encouragement on my side. His advanced mathematical idea for the calculations with Excel is simply…outstanding!

Knowing, likewise, that the lower inscription block with the tens and white dots (see Fig. 1) is a unique digital impression for each plate position like a fingerprint, I began to measure their locations in CorelDraw and to enter the data in Excel (the measurement method and the course of the calculations are explained in detail by Lou in his article in the same issue of our journal.)

Figure 1. Lower portion of the ten lepta stamp with dots and tens. The "fingerprint" of the 10 lepta.

The test was performed on 1580 stamps and the results were impressive! As we see in the table below (see Table 1.), the operation gave me the correct position as the first conclusion for 882 stamps. With 96.27% of the stamps tested, the program obtained the correct position within the first four choices. This is a really impressive percentage. Thus, the identification of the position of each stamp was simple since there were only 4 possible interpretations. The results indeed varied in proportion to the accuracy of the measurements. Many times, in view of the cancellations which prevented us from seeing a dot or one of the tens and/or various other problems of the stamp (tears, creases, etc.) the measurements deviated from the real values. I had, as an example, five stamps from position 37. Making the measurements, my procedure gave three stamps as the first choice in position 37, one as the second choice and one as the 12th choice! The correct answer for the position of the stamp is not always the first choice but, certainly, to search among three or four probable positions for the correct answer is very convenient.

Table 1.

C. Conclusions.

1. Although it seems complicated, the whole process, in fact, is simple, fast and easy. I am willing to make a demonstration to an individual or to a meeting of several philatelists. Any philatelic society or even an ordinary collector who would like a copy of the Excel spreadsheet may communicate with me at hatzis@otenet.gr or with Lou Basel.
2. I arrive at the conclusion that, as the basis of our program, we must measure un-cancelled stamps, if possible, or cancelled stamps of very good condition, whose cancellation does not obscure the lower inscription block.
3. Just as Lou has established, we should measure at least five stamps of good quality for each position and use the average of the deviations as the basis for the program.>/li>
4. I conclude likewise that the any improvements to the program will be proportionate to the quantity of stamps of good quality that a user of this program owns and his…mood to make it better.
5. Certainly, the program must be used together with the relevant bibliography including: the 10 lepta study of Lou Basel, the book of Michael Kaitatzides “Plate Errors of the Large Hermes Head Stamps”, the book of John Coundouros “The Control Numbers and the Classification of the Large Hermes Head Stamps” and other existing aids.

D. The Continuation.

The next value being studied after the 10 lepta is the 40 lepta. With Lou at my side to complete the positions which are missing in my collection and to assist with any difficulties, I will try to extend the program to work on the 40 lepta. Perhaps, by the time that you have read this article, I will have finished it. And, naturally, we will try to apply the program to all of the other values. It will take a lot of work and demand an exceptional amount of time. I believe that by the end of 2008 we will have, hopefully, an electronic method of identifying the plate positions of any large Hermes head stamp. If any philatelist specializing in the large Hermes head stamps is interested in assisting with this extension to other values, please contact me at my email address mentioned above.