Chapter 2

THE KNOX-MARTIN ISSUES, 18 JUNE 1919

Mint copies (even outside of the number 5 serif) are undervalued and should be sought vigorously. Current realizations of around $1,500++ for rarities such as number four and six are excellent buys for appreciation.

2c. Carmine rose, horizontal mint strip of four, positions 7-8-9-10, marginal from right of sheet and thus proving the positions, wonderfully rare and very fresh, superb large part original gum. Slightest of repairs of no significance, this being the largest known multiple of this stamp and unique.

THE MULTIPLE RARITIES

5 and 6

4, 5, and 6

7, 8, 9 and 10

These are the keys to the line of ten surcharges, each being denoted by characteristics of individual overprints, as shown later.

The placing of the surcharges on the stamp is irrelevant.

Please note that numbers 5 and 6 in the top-line pair look ‘positionally’ different to numbers 5 and 6 in the second line.

INTRODUCTION

In June 1919, American flying ace William Knox-Martin made pioneering flights from Barranquilla to Puerto Colombia. Before that, when he was merely 19, he flew a Curtis biplane up the Orinoco River in Venezuela. In early 1919, he acquired a surplus Curtis Jenny and headed for Colombia to demonstrate how the plane could be used for delivering mail between the coastal cities of Cartagena and Barranquilla. It is rumored that he personally designed an airmail stamp for the service.

The Curtis Jenny was purchased by two Colombians, Carlos Obregon and Ulpiono de Valenzeolo, specifically for Knox-Martin to use. The crates arrived and were assembled in the first week of June 1919, and the plane was subsequently named ‘Bolivar’.

In Airlines of Latin America since 1919 by Reg Davies, Reg states that Knox-Martin was first credited with carrying airmail in Colombia when he carried some letters from Barranquilla to Porte Colombie. The distance was less than 20km and by the time the flight left the city center, it would have been arguably quicker to go by horseback. News of the flight was published on Wednesday 18 June 1919 in the Bogota newspaper El Tiempo, which described his flight with the bag of mail to the municipality of Puerto Colombia. According to that newspaper, the mail was collected the Sunday before the flight took place. That explains why they are surcharged 6-18-19.

Between the years 1919 and 1920, there are numerous accounts of flights between Barranquilla and Porte Colombie by CCNA planes, some private for joyrides, and some scheduled. Many countries were exploring flight routes with basic planes, which is why there are many accounts of short flights even where there were no actual airfields.

On these flights, it’s possible there were incidences of ‘dropped mail’. This certainly happened in other countries. There was a famous 1930 flight in Paraguay, which everyone was mystified about until it was realised that the pilot never actually landed anywhere – he simply threw the mail out of the plane as he flew over. At one location, it took two or three days to actually find the mail, so it was not back-stamped on the date the pilot ‘landed’.

Brian Moorhouse has since commented on this theory. He believes, based on pictures of the Knox-Martin planes in flight, that these planes landed with wheels on the ground, and anywhere that was flat would do … a nice field, for example! Elsewhere, roads used for early motor vehicles would suit.

THE STAMPS

A special overprint in five lines was superimposed on the standard stamp of the time. According to Kessler, only 200 copies were printed. There is conjecture as to whether it was 260, but this seems proven, since three sheets were used in overprinting the existing 2-centavos “Nariño” rose Carmine of the 1917 issue. The printing was achieved by a wide-set type repeated 10 times in a single horizontal row, ten stamps at a time, with a different surcharge position for each of the ten. However, the position is not the criterion.

The 260 stamps reported as printed must have been included in the three sheets, which totaled 300 stamps. That leaves 40 without the overprint.

The fifth stamp is overprinted with a strange figure 1. The 1 in the date line is the Arabic 1 instead of the Roman 1. It appears there was a shortage of the correct type. This is called the serif variety, and it also occurs in one of the later issues. Another peculiarity is that the stamps appear to have been severed by a sharp knife, leaving many of the perforations on top and bottom slightly trimmed. The more perfect the perforation, the larger the value.

Another anomaly is that it seems likely a hand-operated press was used and applied carefully. This contrasts with Alex Rendon’s report that the three sheets were not identical, having different positional centering and perforations – some to the right and some to the left.

Of the 260 copies issued of these surcharged stamps, most were flown on cover. Overall, it is known from previous research that some 49 were retained in mint condition. This indicates that around 211 were sent on cover.

In Kessler’s 1936 catalogue, he shows 200 (now 260) issued with 20 (now 26) in serif and 180 in the normal state. He stated that four mint copies existed, with another 16 used. It has since been ascertained that nine mint in serif state actually exist. One is on a strip of three, in which one serif copy exists, and there are three more pairs in which one serif appears, making four in total. There are another five serif singles known to exist. Some argue that there is likely to be another three.

From this, we ascertain that 40 exist in normal mint condition (possibly 38).

COVERS

12 covers with the serif stamp have been located. Thus, after deducting nine unused in mint condition and 12 on cover, there are likely five individual used stamps with serif out of the 26 serif copies. To summarize:

All categories Without serif With serif Total

Mint 40 9 49

Used 194 17 211

Less used on cover (12)

Total 234 26 260

I have not located how many actual numbered copies have been recorded, but the above are the theoretical numbers. As for used on cover vs. individual stamps, we do not know this either. There are nine variations (positions 1-4 and 6-10) of the normal plus position.

There is a conundrum to unravel here. If there are nine or more serif copies, why are there only some 4.4 average numbers of 1-4 and 6-10? Should there not also be nine of each? After all, they were cut line by line. Brian Moorhouse suggests that when the purchasers of these stamps noted the special serif as being different, a few extra copies could have been requested, which could have been specially detached by the postmaster.

A DIFFICULT TASK?

Finding one of each of the nine non-serif mint variations would be of great interest to serious collectors, of which theoretically 4.44 of each could exist. But, as in all things, this might be more uneven. It would be extremely difficult to find a full set of each of the unused nine variations. Several mint copies have appeared in auctions, not least the recent 2013 Goeggel auction.

For long periods, non-serif copies were not scarce for advanced collectors. There were 40 mint copies, after all. However, since Rendon enumerated the peculiarity of each of the nine, some very special collectors appeared who all wanted one of each. As a result, these persons have forced up the realizations of all the mint stamps.

It would be difficult to find single numbers four and six as these are included as part of strips and might have to be purchased with the strips of two and three. Because of the demand for the unique strips, they would never be split! For every non-serif stamp, there is an average of only five or so – perhaps only two or three of numbers four and six will ever be found. In addition, there is an existing strip of four of numbers seven, eight, nine and 10. This will not be broken up, making these numbers very difficult to find.

Several collectors have put together all the 10 types individually or in strips. It was Alex Rendon who investigated these variations. Both Goeggel and Peinado did this and I believe Kaplan did so before tem as well. Possibly the scarcest examples could be numbers four and six followed by numbers seven, eight, nine and 10. Some are not in the best condition, having cut perforations and missing gum. Collectors should look out for any unused copies, especially perfect ones with well-defined perforations.

Remember that Kessler incorrectly stated in his 1936 book that no more than four mint serif copies were known. Dr Philip Cole’s collection auction by Harmer in 1940 did not have one, yet Kessler’s book illustrates one serif copy badly trimmed at the top.

HOW EASY IS IT TO SEE EACH VARIATION?

These were published in Bortfedt’s The Workbook: Notes on Reprints and Forgeries of Colombian Stamps: The Colombian Airmails, Part 3

It is only easy to see the variations if one has the stamp, says Brian Moorhouse – or at least photos with very fine definition. This is because there is a quite minor variation in the overprint design that determines the position number. Take a look at the details above. It is only in more recent times that the item is actually numbered on auction. It is not possible to tie the position of the overprint itself on the stamp for identity purposes. But Rendon discovered that the three sheets used for overprinting were not identical and that centering and perforation positions marginally differed.

A common type number four from this sheet, as compared to another sheet, will look somewhat different, as I discussed earlier. If there were to be a pair with one of the two featuring an omitted overprint, and if such were positioned in a particular place on a sheet, it would be very rare. We have no example of such, but there ARE a few letters of the overprint incomplete in positions two and five.

The positioning of the surcharge is also different in each case. The sheets were cut by knife in strips of 10. One can see the cut perforations. Compare the actual located stamp with the illustration above. The position will be apparent under careful examination with a magnifier.

The 49 mint copies comprise 40 normal and 9 in serif, and I cannot avoid mentioning the SG catalogue values of fine mint copies at:

Serif copy GBP 9,500 ($15,000)

Normal copy GBP 4,750 ($7,000)

The ratio is 49/9 = about five times. If SG is correct at valuing an unused number five at GBP 10,000, then an unused standard copy (1-4 and 6-10) should APPEAR to be of a ratio much less, and of course, auction realizations bear evidence of this. In the past, the normal copy realized around $1,000, but as the uniqueness of each non-serif design became apparent, the realizations have been creeping up. This is indicative of realizations recently. As a guideline, a normal top condition mint copy would fetch rather less, while the serif copy would command some GBP 3,000 ($4,500).

My expectations for the future would range:

Numbers 1-3$2200Numbers 7-10$2500Numbers 4 & 6$3000

Brian Moorhouse indicates a value of $20,000 for the famous strip of three, ex the Goeggel auction of 2013, which fetched some 12,000 Swiss Francs at the time, as well as a strip of two, some 6,000 Swiss Francs. Very few of number five unused are available, so the trend in the future will be for higher valuations, perhaps $5,000.

As mentioned before, these issues are in demand because of the increasing numbers of collectors. That means the values could also escalate. The normal copies are scarcer than realized, but advanced collectors will know that.

I would recommend buying whatever serifed copies become available. And indeed, all other fine well-severed mint copies. Those on cover very much so!

It would also be good to collect a full set of 10 used stamps. There are four times as many used copies as mint, but the number on cover is probably twice. Used covers (except the serif copy) might fetch some $700. Avery desirable set of ten, with each specific number authenticated, might be valued at $10,000. Thus, unlike the CCNA covers, of which there are far lesser in number than unused stamps, these covers are largely equal in value. Used not on cover and on fragment would be half the cover worth.

Realizations on average ($)

With serif Single Pair Strip of three

Mint 3,500 6,000 12,000

Used 1,500+

On cover 3,000-5,000

All the above might prove to be low realizations.

CONSENSUS FOR FUTURE GROWTH

Acquire top-quality, excellent perforated copies of mint and on cover. In view of how few of the number five serifed copies (mint or used) are available, the price becomes less relevant. As forgeries exist, be especially careful of too-perfect perforations, as most of the stamps were cut with a knife.