I've been struggling to identify types and shades of KEVII stamps. The listed color differences for these stamps are subtle and subject to lighting and perception errors. I've looked in catalogs and for internet resources for any definitive identification guides without success.
Some forums suggest there are differences under UV light, but no resources I can find list these differences or even say whether they are talking about long wave or short wave UV light. I've started to think that identification of these stamps is subjective, which seems like a poor way to classify differences that could have a large impact on value.
An Attempt at a Solution:
To remedy this, I've started logging my best guess classifications of these stamps and noting the paper and ink color of each copy under both long wave and short wave UV light, looking for trends that would suggest any definitive identification markers. I have only looked at UV florescence on these stamps. So far, I have not found any examples of phosphorescent stamps from this series.
I believe I have reviewed enough copies of both the 1/2d green and 1d red stamps to come to two conclusions:
- There are noticeable differences in paper and ink in both LW-UV and SW-UV, suggesting a wide variety of paper and ink chemistry across almost every cataloged version of these stamps. Other denominations show similar variability, but without enough copies to make conclusion yet.
- There are no patterns in UV light analysis of either paper or ink colors that give definitive identification markers.
Error Prone Identifications:
I have also noted in this process that sorting these stamps into shades by visible light is highly error prone. For example, with the 1/2d greens stamps, there are just four listed shades for the De La Rue printed stamps. We can confidently say that stamps with dated cancels from 1902 to 1910 are De La Rue printings. So, these stamps should be sortable into four distinct color piles.
I've looked at 342 copies of these dated 1/2d De La Rue stamps. There is a wide range of shades after fading, but when trying to sort them, trends do appear. With some effort, I can convince myself that I've sorted the stamps into four piles, mostly correctly--actually five piles. I'll cover that in the next section.
Next, I take my piles to a new visible light source, and notice that a few stamps in each pile jump out as belonging in another pile. It turns out that my piles want to be in slightly different piles based on the visible light source: Sunlight, incandescent, florescent tubes, shade, even an unlit room all suggest different piles for some stamps.
Finally, in a few cases, when I put the stamps under LW-UV, a few stamps look different from the others, and when I re-look at them in visible light, they look like they want to be in a different pile.
This should be one of the simplest color sorts in these KEVII stamps, but the error rate is high. I even have one stamp that I am quite confident is SG218, but has 217 penciled on the back. I am not the only person who struggles with these.
An Unlisted Stamp Variant:
Earlier I mentioned a fifth color pile sorting my dated 1/2d stamps. I found three stamps, dated 1909 and 1910, that are meaningfully darker than the other yellow green stamps. I'm calling them a variant of SG218, giving them the color name "deep yellowish green."
Looking at the possible shades together, it seems unlikely to be a faded copy of the two bluish shades by De La Rue. And, as I will show later, the use of those first two bluish shades was limited to the first three years of production--1902-1904--although I have seen two copies of SG215 dated 1905.
This deep yellowish green variant of SG218 is very similar in color to SG269, listed as deep dull green. I have nine undated copies that I've sorted as SG269. It's impossible for me to tell if any of these SG269s might actually be this unlisted SG218 variant.
I'm not surprised that this variant hasn't been listed, as it shows up in less than 2% of the SG218 stamps I've looked through, and its color suggests it is one of the Harrison colors, which lacking time travel they can not be. It's also the case that people do a binary sort on these yellowish green stamps--a light pile and a dark pile. The border is fuzzy enough, but without a catalog listing for a third shade, few people would bother looking for this rarer deep yellowish green.
Major Implication:
After looking at hundreds of copies of both the 1/2d and 1d stamps--trying to sort them by shades and finding that wildly error prone--and trying to find a definitive test in either LW-UV or SW-UV and finding no clear answers--I have to conclude that the listed shade differences across these series are too arbitrary for positive identification in many cases.
Please, anyone reading this, I'd love to be proven wrong. This hobby, and the prices people pay for some stamps shouldn't be so arbitrary. I thank anyone who can point to more definitive identification resources than "the expert's eye." That information should be public and easily available.
While I've done this study sufficiently on the 1/2d and 1d stamps, there's no reason to expect any different results with the higher denomination and value stamps. That's a problem.
Unlisted Paper, Inks and Coating Types:
One thing is clear from this research; there was some underlying variability in paper and ink choices in the manufacturing process that is not captured in the catalog listings, or for that matter in any visible differences in any specific cataloged stamp type.
I haven't found any primary data to back this up, but I have to conclude that De La Rue (and Harrison too, although I don't have sufficient stamps to say anything about Somerset House) had multiple paper and ink suppliers, or that the primary paper and ink suppliers varied their production processes over time.
Under LW-UV, looking at the back of each stamp, I see a number of paper colors:
- Yellows -- from a dull tan yellow up to a bright canary yellow
- Tans -- sometimes tinged either reddish or yellowish, and varying in darkness from light to almost coffee stained
- Reds -- from a dull tan red to maroon
- Purples -- which could be seen as a continuation of the red shades
- Blues -- a very few stamps show either an overall blue tinge, or a mottled blue spotting that fluoresces more than the rest of the paper
- White -- one stamp was distinctly white, which might be considered a very light tan
- Orange -- one stamp was distinctly orange
- Mottled Yellow, Tan, Red -- a large number of stamps had mottled colors mixing either yellow and tan, tan and red, yellow and red, or in some cases all three. This was common enough that it didn't seem to be a few dirty stamp example.
The frequency and consistency of these LW-UV paper shades is sufficient for me to conclude that they represent different paper types rather than either ink bleeding, aging, soaking contamination, or dirty stamps.
These paper shades were independent from the denomination and ink color and showed up with similar frequency on both the 1/2d green stamps and 1d red stamps. And, this effect appeared equally on both De La Rue and Harrison stamps.
There was a similar range of responses in the ink colors to LW-UV, although with different colors depending on the denomination.
For the 1/2d green stamps I see these ink colors:
- Sage Greens
- Greens -- which could have been a brighter version of the Sage Green
- Purple -- which seemed to transition from a dark purple through a light purple and at times to a gray
- Mottled Sage Green & Purple -- This was the dominant color
- Mottled Sage Green & Purple with florescent Blue speckles -- This is the same mottled color with the addition of those odd blue florescent speckles that sometimes appeared on the back of the stamp. I wonder if that's some kind of paper coating, or an artifact of a washing or other printing treatment process.
- Purple Brown -- which also varied in lightness to a sort of gray brown
- Yellows
- Purple with florescent Blue speckles
- Blue -- which could have been some other ink cover but totally covered in that florescent blue speckles. This was very rare.
- Orange -- Between 10 and 20 percent of the stamps showed a range of orange hues ranging from bright to a dim reaction. I say 10 to 20 percent because the line of what was a reaction and what was just a shade got fuzzy. This reaction was largely limited to stamps that sorted as SG219.
- Red/Pink -- Four stamps showed red/pink reactions
- Brown -- One stamp showed a brown reaction, which might have been a dark orange reaction
- Red --About half the Harrison stamps showed a mid-level red to deep purple reaction, although the reaction was not indicative of a particular Harrison shade
- Orange - One stamp showed a bright orange reaction. I think this is an example of SG275a, the rare aniline-rose color, although I don't have enough copies of this stamp to verify this reaction. My other presumed copy has a bright yellow paper reaction, but not a noticeable ink reaction.
- SG215 through SG218 were in production in obvious periods. The data shows the two blue shades in production together from 1902 to 1904. Given the parallel production periods, these seem like two ink formulas, perhaps from different vendors. The few copies of SG215 in 1905 suggest overstock that slipped into use, but were not produced in that year.
- SG217 and SG219 were also used in parallel, starting at the end of 1904, but it looks like the inks came from two different sources with the source for the SG217 ink being dropped in 1907.
- Stamps dated between 1908 and 1910 are almost certainly SG218.
- Of no particular importance, in this early period, most (71%) had nicely legible date cancels. Starting in 1911, including most of the Harrison stamps, the date on the cancel is rare to find.
- Tan is the most common paper shade for these stamps, followed by yellow and red.
- The LW-UV paper shades are not indicative of the stamp type
- This is a chaos of varieties, especially in the early years of production
- As the years progressed, the number of varieties tended to get smaller, suggesting that processes were getting more predictable
- That bottom line shows how many varieties there were across the whole lifespan of each stamp type, and while it may look like the variation of SG218 is larger, it was also in production for twice as long as the others, suggesting that its process was slightly more in control.
- There is no predictable pattern of paper type to stamp number
- There is no predictable pattern of ink type to stamp number
- SG219 and SG220 had completely parallel production timelines. The difference in these two stamps may just be from two different ink vendors.
- The 68% dated percent for this early period is consistent with the 71% for the 1/2d green stamps.
- The distribution of these shades is consistent with the distribution of 1/2d green paper shade
- The LW-UV paper shades are not indicative of the stamp type
- The number of varieties never seems to be under control for these red stamps
- SW-UV reactions are rare in these stamps. The orange glow may be indicative of a stamp being SG219
- That 272a (Brick Red) stamp shows a noticeably different pattern to the others.
- Otherwise, there is no predictable pattern of paper type to stamp number
- It's interesting how 272, with the most copies, shows no LW-UV ink reaction. That sets it apart from the others, which all show a trend of showing some reaction.
- That orange reaction on 275a is interestingly unique, and may be a meaningful marker for the aniline rose color, although I'll need more copies to confirm.
