One in a thousand? One in ten thousand? One in a hundred thousand?

From very early on in the postage stamp era, there have been stamp issues - often perfectly genuine ones - where the proportion of stamps printed actually used postally is very low. The first 1866 issue of Honduras is a good example. In such cases, catalogues often advise, Beware of Forged Cancellations! But catalogue prices rarely reflect the full reality of postally used scarcity.

An extreme example is provided by the 1919 issue for the Northern Army, # 15 - 19 in Michel, which gives a total printing figure of 3 000 000 stamps for the five values. This figure may well be correct: the Northern Army issue shows little or no sign of being aimed at the philatelic market. The stamps are the worst designed ever, all five of them boring. Proofs, colour trials, errors and varieties are almost non-existent. It may be that the Northern Army really hoped they would need these stamps for post offices in the areas they occupied and optimistically ordered three million.

In the event, these stamps were briefly placed on sale at post offices and a few postally used copies are known on cover and card. Alexander Epstein has chronicled them. Add them all up, and maybe you have twenty postally used stamps.

Rather more stamps are known cancelled to order, either as sheets or on covers. But even including these, it seems to me doubtful that the total adds up to more than ten thousand. The remainder are all mint stamps. But Michel prices mint and used at the same prices. That's absurd.

Another extreme example, and more interesting, is provided by the stamps of Dashnak Armenia. In 1919 - 20 the Dashnaks did attempt to run a postal service in a tiny country ravaged by war, famine and disease, Anyone who could was leaving, taking with them postage stamps which they had exchanged for Armenian currency worthless in the outside world.

As a result, and to this day, mint stamps or CTO stamps massively outnumber postally used ones. How massively? I have handled many thousands of Dashnak stamps in the past twenty-plus years. I doubt that one in a thousand stamps I have seen is postally used, probably more like one in ten thousand. In addition, only a small number of values are seen postally used: the 60k on 1 kopeck first issue, the 10/7 kopeck with framed Z, the 1 rouble, 3 rouble 50 and 5 rouble with framed or unframed Z ... maybe twenty or thirty different types. If you add in the over the counter philatelic productions which Souren Serebrakian used on postcards to his brother in Tiflis, you double or treble that number.

Interestingly, no one to my knowledge has tried to make a full list of stamps known postally used. Someone should do it!