Search This Blog

Tuesday, August 29, 2017

The Printings Of The 2.5d Ultramarine Queen Victoria Keyplate Stamp of Lagos 1891-1904 Part Four

Today, I carry on with the 2.5d ultramarine with the seventeenth printing, which should have been made sometime around July or August of 1894.

Seventeenth Printing




The head plate of this printing is an almost exact match to Gibbons's bright blue, while the duty plate colour is royal blue.

I have two mint type A's and two type B's, one of which is used, and has been cancelled with what appears to be an 8-bar oval obliterator.

Eighteenth Printing



On this printing the head plate and the duty plate are closest to what Gibbons's cobalt blue would be if it were deeper. A similar shade is found on the very late printings made after 1899, but the degree of plate wear showing on this example is too advanced for it to have been from the new plates that appear to have been used to print the last printings. I have only the one mint example shown in the scan above. 

Nineteenth Printing 



The head plate of this printing is closest to a very pale version of the dull ultramarine, while the duty plate colour is ultramarine. I have five mint and two used examples, all of which are type A. The used example on the right with the Badagry CDS is badly discoloured, but I think underneath all the discolouration, that it matches the other stamps shown here. 

Group 3 - Printings 20 to 37

In this group of printings we finally see most of the detail in the hair at the back of the Queen's head gone. Some stamps will show a little detail, but most will show very little. In addition, the top 5 hairlines and the bottom 3 or 4 above the crown are merged on most stamps, leaving only a small band of hair in the middle

Twentieth Printing


On this printing, both the head plate and duty plate colour are closest to ultramarine on the Gibbons colour key, but are just a little bit lighter. I have four mint and nine used examples. Most of the used examples are canceled with a mixture of 8-bar and 9-bar oval obliterators. The first used stamp is one of the earliest examples of a CDS cancel that I have seen on this stamp, which is usually always cancelled with a barred oval prior to 1899. This one is dated October 23, 1896. The cancellation on the third stamp is very interesting: Opobo River is in the Niger Coast Protectorate, and this cancellation is dated almost 10 years before Niger Coast Protectorate and Lagos amalgamated in 1906. So it would appear to be a very rare example of a Lagos stamp used in Niger Coast Protectorate. 

In this group there is also another example of a plate flaw, which may be constant: a diagonal break in the inner frame below the "2" of "1/2". It is present on the lower left stamp in the scan. Here is a close-up scan of that flaw:



Twenty First Printing



In this printing, both the head plate and duty plate are closest to what Gibbon's cobalt blue would be if it were deeper. I have only two used examples, both of which appear to have been cancelled with an 8-bar oval obliterator, The first example is badly scuffed, but this would seem to be one of the scarcer shades in which this stamp was printed. Both examples are type A. 

Unfortunately this was all I had time to document today. If I get time later in the week, I will add a few more printings, but it would seem that it will be at least another week before I can finish detailing the remaining printings of this value. 

Tuesday, August 22, 2017

The Printings Of The 2.5d Ultramarine Queen Victoria Keyplate Stamp of Lagos 1891-1904 Part Three

Today's post resumes where I left off with the tenth printing of the 2.5d ultramarine Queen Victoria keyplate issue, through the sixteenth printing.

A few more possibly constant plate flaws make their appearance on the printings detailed in this post:


  1. A white dot inside the "Y" of "Penny". 
  2. The letters "PE" of "Penny" deformed by having a slub on top of the "P" and a small divot taken out of the top of the "E". This would appear to be constant, as I have seen the same variety on several of the 2.5d stamps. 
  3. "2" with an elongated top, and weak left outer frameline. Normally the top of the "2" ends well above the the bottom of the "1" of "1/2". But on this variety, the top of the 2 is almost level with the bottom of the "1". 


Tenth Printing



On this printing, the head plate and the duty plate are both the same colour, and it is closest to dull ultramarine, being an almost perfect match, and having none of the brightness that the stamps of printings seven and eight had.

I have two used examples, both type A, and not in the greatest condition. Both are cancelled, with what appear to be 8-bar oval obliterators.

Eleventh Printing



On this printing, the head plate and the duty plate are different. The head plate colour is a pale, dull ultramarine, while the duty plate is dull ultramarine.

Again, I have two used examples, which, once again, are both type A, and are both cancelled with 8-bar oval obilterators.

Twelfth Printing



On this printing, the head and duty plate are both the same colour, which is close to Gibbons' bright blue, but paler. It lacks the chalky blue undertone that the dull ultramarine has, and variations of this colour will prove to be quite prevalent in the later printings. I have a single used example, which I am fairly certain is type A also. It appears, just on the basis of the thickness of the cancellation bars, to have been cancelled with an 8-bar oval obliterator.

Thirteenth Printing



On this printing, the head plate colour is very similar to the eleventh printing in that it is very close to pale dull ultramarine, except on this printing the colour is a bit brighter, with just a touch more blue than that printing. The duty plate colour is dull ultramarine, just as was the case with the eleventh printing.

I have four used examples of this printing in the type A, and one used and two mint examples with the scarcer type B. One of the used type A's is canceled with a German steamship cancel, while the others appear to be cancelled with the thinner 9-bar oval obliterators.

Fourteenth Printing



This is one of the most distinct printings of this value because of the significant disparity between the head plate and duty plate colours. It is so significant that it boggles the mind that it would not be listed as a separate catalogue number in Gibbons. The duty plate colour is grey blue, which is not found on any of the other printings, while the head plate colour is dull ultramarine.

I have five mint examples, two of which are shown here. Four have no gum, and the other, only partial disturbed gum. I also have four used examples, all of which appear to be cancelled with the 9-bar obliterator. All nine stamps are type A. The lack of gum on these leads me to think that this may be the April 1894 printing, as this was sent exactly three years after the first shipment and thus the timing is about right. The April 1894 printing arrived in the colony with a large number of the stamps stuck together, so that they had to be soaked apart. This was written about in Jack Ince's reference work on the stamps of Lagos, and this is the only printing I have, in which none of the mint stamps have full original gum.

Here is a close-up of the two mint examples, so that you can see the difference between the head plate and duty plate colours:



Fifteenth Printing




In this printing, both the head plate and the duty plate are dull ultramarine, although now the shade has become just a bit brighter than the last printing. I have one mint and five used type A's, and two mint type B's. The used type A's are cancelled with a mixture of 8-bar and 9-bar oval obilterators.

On this printing, the type B mint examples exhibit two of the varieties that were written about in the last post, and this post, namely:


  1. The deformed letters "PEN" of "Penny" on the left stamp, 
  2. The elongated "2" and weak left frameline on the right stamp. 

The close up scan below should show both of these varieties clearly:


Sixteenth Printing




On this printing, the head plate is still dull ultramarine, but once again, it has gotten just a little bit brighter. The duty plate colour however, is distinctly royal blue. I have two used type A's and six mint type B's, plus one used type B. Of the type B's, one shows the deformed "PE" of "Pence", as shown in the scan below:




Note the much deeper colour of the duty plate in relation to the head plate. Also note how there is a slub on the top of the "P", while the top of the "E" appears to have had a slice taken out of it.

That is all I have time to go through this week. As I anticipate a total of about 40 printings for this stamp, I think it will be two more weeks before I have properly identified and documented all of the printing differences.


Tuesday, August 15, 2017

The Printings Of The 2.5d Ultramarine Queen Victoria Keyplate Stamp of Lagos 1891-1904 Part Two

Today's post picks up where I left off last week. I had sorted all of my 2.5d stamps into 71 groups, which was way, way more than the number of printings that were likely made. Given that the stamps were sent to the colony on a quarterly basis and the first printing was issued in April 1891, with the last printing being in August, 1901, there should be approximately 41 printings, as long as each sending included a batch of 2.5d stamps. Because the 2.5d was used to pay the new UPU foreign letter rate per half ounce, it is quite likely that there would have been a separate printing each quarter, to make up the total number of 428,040. This would suggest an average printing of 10,701 stamps each. However, the distribution of examples in the sample that I have sorted, makes it clear that the printings were not done evenly, and that some of the printings are very scarce to rare. Gibbons prices of course are for the most common printings, except for the scarce blue shade, which is from the first two printings made in 1891.

The basic shade catalogued by Gibbons is ultramarine, for both the head plate and the duty plate. However, as we shall see, this is a gross over-simplification, and there are many shades that cover the whole gamut. In addition, there are some printings where the duty plate and head plate colours are completely different, the scarcest of which is the April 1894 printing, which only exists without gum when mint, due to the fact that the entire shipment was stuck together when it was received in Lagos, with the result that the stamps had to be soaked apart.

In addition to the two listed types of value tablet, I have discovered a few other prominent plate flaws, which I suspect are likely constant:


  1. A badly damaged "2" at the top of the "2".
  2. Frame break under the "E" of "Penny".
  3. Deformed inscription. Under magnification, you can see that the type A inscription has been retouched to change it to to a type B inscription, as the ink is contemporary. 
  4. Damaged "2" and pitted left outer frameline, resulting in a frameline that is much thinner than normal. 
  5. Oblong mark to the left of the queen, where two of the horizontal shading lines have been joined and unshaded earlobe. 
Without further ado, I give you my rundown of the printings of this stamp. Today's post will cover the first nine of these, and then next week, I will carry on and run through some more. 

Group 1 - Printings 1-7

In this first group, the predominant state of the plate shows very little wear, with most examples in a sheet showing all the details of the hair clearly and the diadem clearly, with little to no merging of the hairlines. A few examples will show merging of the top 3-5 lines of the hair, but all the detail of the hair at the back of the head will still be clearly visible, as will all the detail of the jewels. 

Printings 1-2 - The Listed Blue Shade

Both of these first two printings are very similar to one another. The main difference is that on the first printing, both the head plate and duty plate colours are the same, while on the second printing, the duty plate colour is quite a bit darker than the head plate colour. 

First Printing


The head plate and duty plate shade of this first printing are both closest to the blue swatch of the Gibbons colour key, so the catalogue description is almost accurate. The only anomaly is that Gibbons lists the blue shade as having a Type A inscription only, and this one is clearly a Type B.


Second Printing


On this printing, the head plate is closest to blue, but the duty plate is dark blue. I have five used examples in my stock, all of which are cancelled with 8-bar oval obliterators, as one would expect, given that these printings are for 1891. I believe that the two examples on the left, are Type A inscriptions, while those on the right are Type B. 

On this printing, I have, what I believe are two constant plate flaws. Further research and corroboration with other examples would be required to prove their status as constant. The first of these occurs on the second stamp from the left, and shows a Type A inscription with the inner frame broken under the "E" of "Penny". 




The second of these occurs in the middle stamp at right. It shows a deformed type B inscription, but what is interesting, is that there is a clear deeper outline of a type A inscription underneath. So it appears that the type B inscriptions may be the result of retouching the duty plate where the inscriptions were weak or incomplete.


The retouching is particularly apparent when you look at the tops of the letters. 

Printings 3-9 - Other Shades That Are Listed As Ultramarine

Third Printing


The next printing is very distinct in the sense that it is a very dull shade, that can best be described as pale, dull ultramarine, through it is not a match for either the dull ultramarine swatches, the grey blues, or the dull blue swatches. I have one unused example, which is in rough condition, and three used examples, which are also a bit rough. Two of these seem to have been cancelled with 8-bar oval obliterators, while the middle used stamp is clearly cancelled with a 9-bar oval. All four of the stamps that I have are type A inscription. 

Fourth Printing


On this printing, the head plate is about half way between the cobalt swatch of the Gibbons colour key and the dull ultramarine swatch: it is brighter than dull ultramarine, but is much, much lighter than the colour of the third printing. The duty plate colour, is darker, being an almost exact match to the dull ultramarine swatch of the Gibbons colour key. 

As shown in the above scan, I have two mint examples, both of which are type A, and 8 used examples, two of which are type B. One of the used type A's is cancelled with a German seapost cancel, most all of the others appear to be cancelled with either 9-bar ovals or 8-bar oval obliterators. 

Fifth Printing


On this printing, the duty plate is closest to bright blue on the Gibbons colour key, and the head plate version is a very slightly paler version of the same colour. The inscription is clearly type A. This is a beautiful, perfectly centered mint never hinged example. 

I also have two used examples as shown in the scan below:



Sixth Printing


On this printing, there is a very slight hint of periwinkle to the blue of the head plate. The colour of the head plate can best be described as what would result if you took the dull ultramarine swatch and combined it with the violet blue swatch of the Gibbons colour key, while making the whole colour lighter. The duty plate colour is almost an exact match to the dull ultramarine swatch of the colour key. I have one single used example, with type A inscription, and canceled with a 9-bar oval obliterator. 

Seventh Printing




In this printing, the head plate and the duty plate are basically the same colour. The colour is closest to dull ultramarine on the Stanley Gibbons colour key. So this is the first printing where the colour is getting close to true ultramarine, the shade that is listed by Gibbons as the basic colour of the stamp. I have two examples of this printing in mint condition, and eight used examples, one of which is cancelled with a Lagos CDS. The other used examples all appear to be cancelled with 9-bar oval obliterators. 

On this printing, I have come across what I think may be another constant plate variety: another instance of the "damaged 2", where the damage is further toward the end of the "2" and a severely pitted and thin outer left frameline. The scan below shows this more clearly:


Group 2 - Printings 8 Through 19

In this group of printings, there is a slight loss of sharpness to the detail in the design. Most examples show merging of the top three lines of the Queen's hair, while many copies will show merging of the top five lines. The detail of the diadem and jewels will still be completely clear, and there will be some detail visible in the hair at the back of the head below the diagonal ribbon. 

Eighth Printing



The head plate and duty plate of this printing are largely the same colour, which is almost an exact match for dull ultramarine on the Gibbons colour key. I have one mint, and one used example, which appears to be cancelled with a 9-bar oval obliterator.

Ninth Printing


On this printing, the head plate and duty plate colours are still closest to dull ultramarine on the Gibbons colour key, but on this printing the colour is just a bit brighter, being a little bit closer to true ultramarine. I have two mint and seven used examples, all of which are type A. The used examples are mostly canceled with 9-bar oval obliterators, but there is one 8-bar oval as well. There is also an early example of the larger Lagos CDS dated June 29, 1897. 

On this printing, I have an example of what may be yer another constant plate flaw: The queen's earlobe is missing the shading lines that are normally present, and there is a large oblong flaw jointing two of the horizontal shading lines to the left of the queen's eyes:


Note how white the earlobe is compared to how it usually looks, and note the flaw to the left of the queen's profile. 

This concludes my discussion of the first nine printings of this value. Next week I will attempt to complete my coverage of the remaining printings. 

Wednesday, August 9, 2017

The Printings Of The 2.5d Ultramarine Queen Victoria Keyplate Stamp of Lagos 1891-1904 Part One

Overview

Today's post deals with what in my opinion is one of the most remarkable stamps from the 1887-1903 Crown CA Lagos issue: the 2.5d ultramarine. The introduction of a single UPU rate for mail to all foreign destinations other than the UK, of 2.5d per half ounce, created a need for this value, and it was first dispatched to the Colony on April 15, 1891. It is the only stamp from the series, for which there were no remainders whatsoever. In all, 428,040 stamps were printed, so there will have been a very large number of printings - likely well over 40. Given that stamps were supplied on a quarterly basis, and there would have been three quarters left in 1891, there could be up to 43 printings. I have identified more than this number in my sort of 183 mint and used examples, so it is likely that some of the varieties I have found are merely sub-types of some of the printings.

The stamp follows the usual progression of plate wear, though on this value it seems that it never gets to stage 5, as I have not come across any really coarse examples. However, it does appear that there were some printings made in 1901 from plate 2, as many of my dated used examples from after 1901 possess a clarity that is uncharacteristic of printings made at this time, which does suggest that they are from a new plate.

The main thing that is interesting about this stamp is the very vast array of subtle shades, and the fact that quite often the head and duty plate shade dates are quite different. The basic colour is identified as being ultramarine, but in reality there is a very wide range of shades. Gibbons does list a scarce blue shade, in the type A duty plate (to be discussed in a minute), but it is really quite a greenish blue. In addition to this shade, there are several other very scarce shade combinations. The second interesting thing is the differences in the duty plate letters. Gibbons lists two types:


  1. On the more common type A lettering, the letters are thin, the fraction bar is thinner and has a straight top, or ends in a straight point. 
  2. On the scarcer type B, the letters are slightly larger and bolder. The fraction bar has a slightly curved top, almost resembling a serif. 
These two types are thought to have been mixed on the same plate and in practice, they can be quite difficult to distinguish. The large multiples that I have in my possession suggest that this is indeed the case on some printings, but not on others. I measured a known type A and a known type B, as I felt that the measurement of the letters would likely differ in some way, and that this would provide a more reliable way to distinguish the two types. Part of the problem in distinguishing these two types is that there are several examples where the fraction bar is very slightly curved, being mid-way between straight and curved, and the difference in size of the letters can be very hard to spot.

What I found was:

  1. On type A, the width of the inscription is exactly 15 mm. All the letters except for the "P", the 2 and the fraction are exactly, or almost exactly 1.5 mm tall, and 1.5 mm wide, except for the fraction, which is 2 mm wide. The "P" is generally 1.75 mm tall. The fraction bar is 2 mm long.
  2. On type B, the width of the inscription is just a smidge wider than 15 mm. All the letters except for the "P" and "E", the 2 and the fraction are exactly, or almost exactly 1.75 mm tall. The "P" is very close to 2 mm tall and the "E" is still 1.5 mm tall. The fraction is also 2 mm wide, and the "Y" is also very close to 2 mm wide. All the other letters and the 2 are 1.5 mm wide, as in type A. So essentially, the letters are generally 0.25 mm taller on type B, as compared to type A. The fraction bar is generally 2.25 mm long. 
When I examine the larger multiples with these characteristics in mind, it becomes clear that all stamps on one pane are either type A or type B. It is possible that in a sheet the left pane is type A while the right is type B. Ince suggests that only 25% of the total printing was type B, so there may have been printings that were only type A, or B for both panes. Some of the type B stamps, do look almost like type A until you measure the height of the letters. 

Here is a scan from the larger block of 36 below that shows two adjacent stamps, that at first appear to be different types, but are seen to both be type B, once measured:


On here you can see that the fraction bar of the top stamp has the curved top normally associated with type B, but on the bottom stamp it look more straight, like type A. However, the Y's of both stamps are 2 mm wide and the letters are mostly 1.75 mm tall, instead of 1.5 mm. 

The scans below show the large multiples in my possession:



This block of 4 from positions 46, 47, 56 and 57 in the sheet shows the current number 1, which as explained in another post, is a registration number showing the order in which the plates were laid down. Since there was only 1 plate for Lagos prior to 1901, and the use of these numbers was abandoned after 1890, they are always "1" for Lagos issues. The presence of that number on this block suggests that this is from one of the first printings. However, the clarity of the impression is different for the top two stamps, which show almost no detail in the hair on the back of the head, and more merging of the top 5 hairlines, whereas there is more detail in the hair on the back of the head on the bottom two stamps. All four stamps are type A. 


Here we have a block of 6 from positions 43-44, 49-50 and 55-56 in the sheet. Again, this sheet shows the current number, which suggests that it is a pre-1892 printing. The degree of visible plate wear is about the same on all six stamps. There is some merging of the top 5 hairlines at the top of the head, and there is sole loss of detail in the hair at the back of the head. However, most of the detail is still visible. All six stamps are type A.  



Here we have a block of 18 from the middle of the sheet. Most of the design detail is clear and the degree of wear seems to be similar on all stamps. Some of the stamps in this block are type B, such as the second stamp from the left on the bottom row, while others, such as the third stamp from the left in the top row, are type A.



Finally, I have a block of 36 being the entire left hand pane from rows 5-10. This also has the current number, suggesting that is is a pre-1892 printing. The degree of plate wear varies considerably between stamps in this sheet, as shown in the scans below. Again, some of the stamps in this sheet are type A, while others are type B. 

There is one key revelation that comes from these multiples that will impact not just my research findings with regard to this stamp, but may also alter many of the conclusions that I reached about the other stamps I have studied so far in this issue, as well as the earlier issues, and that is that the plate did not wear evenly. The largest block here, containing 36 stamps, shows examples that have the clarity associated with early printings where the all the detail in the hair is visible, except for the merging of the top three hairlines, se-tenant with others that show the wear associated with late printings. Either that, or the differences we are seeing result from over inking. 

Here are some examples of what I mean:


Here we have a pair from the right side of the sheet in which the stamp on the left shows much clearer detail in the hair at the back of the head, whereas the stamp on the right shows almost no detail in the hair at the back of the head. 

Conclusion

Clearly, while the degree of plate wear provides some insight into when a particular stamp might have been printed, the examination of individual stamps for clarity of impression does not yield foolproof results. It is quite likely where I have found stamps that are the exact same shade, that those stamps are quite likely to be from the same printing, rather than separate ones. 

So I will have to go over the 71 "printings" I have identified and consolidate them given this information that I have uncovered. It will then be necessary at some point to go back and re-visit my previous sorts to see if some of the other stamps need to be consolidated into fewer printings. 

So next week's post will look at the individual printings that I identify, once this process is complete. 






Friday, August 4, 2017

The Printings of the 3d Lilac and Orange Brown Queen Victoria Keyplate Stamp From Lagos 1890-1901

Today's post continues my examination of the 1997-1903 issue of Lagos with the 3d lilac and orange brown, which was first dispatched to the colony on December 19, 1890. There were 143,820 stamps sent between that date and August 19, 1901, of which 85,800 remainders were sent back and destroyed. Like the other values that I have examined so far, I believe that there were probably lots of small printings of between 3,500 and 5,000 stamps, which would mean more than 20 printings. We'll see how many are revealed in this post. However, it must be borne in mind that I may not have an example of every printing that there was, and that there may have been more than just what I have here.

I have 121 mint and used stamps of this value, which I initially sorted into the same five groups that I had identified in the posts dealing with the 5d, 7.5d, and 10d stamps. Interestingly, I did not find any stamps that possessed the characteristics of plate 2, which suggests that all printings were made from plate 1. Within the first group however, there was one that showed every so slightly less plate wear than the others, and which I believe represents the printings made between December 1890 and December 1892. They are much scarcer than the later printings, and I do not believe that I have examples of all that were printed in these early years before 1894.  I sorted each of the first five groups further by the shades of the head plate and duty plates (words of value) into printings. I will now illustrate them and describe their characteristics. Once again, I do not have a sufficient number of dated used examples to be certain about the order of the printings, so when I refer to a printing as the "fourth printing", I mean the fourth printing that I identified - not necessarily the fourth printing made. In all, I believe I have identified 31 printings of this value.

Group 1 - Printings 1-3

Like the first printings of the 5d, 7.5d, and 10d, these three printings show only minimal plate wear, with slight merging of the first three lines at the very top of the hair on two of the three. The very first printing does not show the merging of these lines, and it is my belief that this printing comes from the period between December 1890 and the end of 1893.

First Printing



On this printing, which I believe is at least the first of the ones I have identified, the head plate colour is closest to dull mauve on the Stanley Gibbons colour key, while the duty plate colour is closest to lake-brown. The head plate colour is just a touch deeper than the dull mauve, but it is very close to an an exact match. 


Second Printing


The head plate colour on this printing is closest to the reddish lilac swatch on the Gibbons colour key, being almost an exact match, in fact. The duty plate colour is closest to lake brown also, but there is just the smallest hint of a bit more red in this colour compared to the first printing. The cancellation is a German ship cancellation that appears to be from 1900 - a good ten years after this stamp was issued. So either this is a plate 2 printing, or it is a late use of an early plate 1 printing. It is quite possible, given the large number of late usages of these stamps for this to be an example. This one shows an interesting variety - a burr on the "C" of "Pence". It is the only stamp out of the 121 examples here to show this variety, so I am not sure if it is constant or not. 

Third Printing


The head plate colour on this printing is completely different again, being almost an exact match to Gibbons' dull purple. The duty plate colour is closest to chestnut.

Group 2 - Printings 4-11

In this group of printings, they are similar to the first, but there is a slight loss of sharpness, with the most noticeable being the first shading lines near the jewels of the crown.

Fourth Printing


The head plate colour on this printing is also closest to dull purple, but is just a touch paler and redder than the colour of the third printing. The duty plate colour is closest to orange-brown on the Gibbons colour key. The used example again has a German ship cancellation and the date is either 1897 or 1907, which would be another example of a late usage in either case. 

Fifth Printing


Like the first printing, the head plate colour of this printing is also closest to dull mauve, while the duty plate colour is also closest to lake-brown.  The cancellation is indistinct, but appears to be an 8-bar oval, which is consistent with it being a pre-1899 printing. 


Sixth Printing


On this printing, the head plate colour is also closest to dull mauve. The duty plate colour is closest to lake brown, but on this printing the colour contains just a bit more brown. However, it is not brown enough to be reddish brown. The used example shown here has a Lagos CDS cancel. I can just make out a "4" at the end of the date, which suggests that it was used in 1894. There is a small possibility that the date is 1904, but by then the style of cancellation most commonly in use had changed. 

Seventh Printing


The head plate shade on this printing is tricky as it is not an exact match of any swatch on the Gibbons colour key. It is duller than the lilac and reddish lilac shades and lacks the brown that is present in the dull purple. I would say that it is closest to how the slate-lilac swatch would look if a considerable amount of white were added to the colour. So it is a pale slate-lilac. The duty plate colour is closest to chestnut. The sole used example, which I believe is a little faded, is dated 1898.

Eighth Printing


The head plate colour on this printing is closest to pale slate-lilac.The duty plate colour, like the last printing is closest to chestnut, but just a touch deeper. 

Ninth Printing


The head plate colour of this printing is closest to pale-slate lilac, while the duty plate colour is closest to chestnut. 

Tenth Printing


The head plate colour of this printing is closest to dull purple, while the duty plate colour is closest to orange-brown. 

Eleventh Printing


The head plate colour of this printing is closest to pale slate-lilac, while the duty plate colour is closest to lake brown, but just a bit paler. 

Group 3 - Printings 12-18

The printings of the third group are characterized by the lack of detail in the hair in the back of the head, the merging of the top four or five hairlines at the top of the head, and the merging of most of the lower hairlines up to about half way up the jewels in the crown.

Twelfth Printing


On this printing, the head plate colour is closest to dull mauve, while the duty plate colour is closest to orange brown, but just a touch paler. 

Thirteenth Printing


The head plate colour of this printing is also tricky as it does not quite match the lilac swatches, being both a touch greyer and duller. However, it is not anywhere near as dull as the grey-lilac and it lacks the brownish tinge that it would have to have in order to be a paler version of the dull purple. So I am going to call this dull lilac. The duty plate colour is closest to lake brown, but is just a touch deeper.

This printing is, by far the one that I have the largest number of examples of, with 12 mint and 11 used. Of the used examples, 5 examples are dated between 1902 and 1903, while four appear to be cancelled with a 9-bar oval obliterator. Again, this suggests that the dated examples are late usages, because the plate just does not show anywhere near enough wear to be from 1902 or 1903.

The scan below shows a close-up of three of the examples from above:




Fourteenth Printing


The head plate colour of this printing is closest to pale slate-lilac, while the duty plate colour is closest to chestnut. Both used examples have the same Lagos cancellation from September 24, 1895. 

Fifteenth Printing


The head plate colour of this printing is closest to reddish lilac on the Gibbons colour key, while the duty plate colour is closest to orange-brown.



I have three used examples of this printing. The stamp on the left is cancelled 1900 or later, which, once again is likely a late usage. The stamp in the middle is cancelled in 1892, which suggests to me at least, that there may not truly have been 15 printings in just under two years, although there certainly could have been eight. The stamp on the right is cancelled with a 9-bar oval obilterator, which is consistent with a pre-1898 printing.


Sixteenth Printing



The head plate colour of this printing is closest to dull purple, while the duty plate colour is closest to chestnut.



None of the above three used examples show a clear date, but two are cancelled with 9-bar oval obliterators, which suggests that these were printed before 1899.

Seventeenth Printing


Here the head plate colour is closest to dull mauve, while the duty plate colour is closest to chestnut. Again, the only used example in my possession is cancelled with a 9-bar oval obliterator, which suggests that it was made prior to 1899.


Eighteenth Printing


The head plate colour of this printing is closest to what I called dull lilac on the thirteenth printing. The duty plate colour is closest to lake brown, but just a touch paler.




The head plate shade of these three used examples is rosier than the mint examples, but I think that all three of these have been slightly affected by water. The only readable date is on the middle stamp, and it is only a partial date that is sometime in the 1890's, probably between 1897 and 1899.


Group 4 - Printings 19-26

These printings are characterized by the fact that while nearly all the detail in the hair up to the top of the crown is gone, there is still a narrow band of detail visible between the top of the crown and the top of the head. Also, the horizontal shading lines in the lower horizontal band of the crown are still visible, but just beginning to merge into one another.

Nineteenth Printing


On this printing, the head plate colour is closest to dull mauve, but just a touch rosier. The duty plate colour is closest to orange-brown, but just a little paler. The used example on the right is dated 1902, which is still late for a group 4 printing.

Twentieth Printing


The head plate colour of this printing is closest to dull purple, while the duty plate colour is closest to chestnut. The used stamp on the right is the most troubling because the date is clearly 1891, but the impression is far too worn for it to have come from such an early date. This leads me to wonder whether the hammer used for the cancellation was set to the wrong date, and whether the postal clerk really meant to set it to 1901 and not 1891. 

Twenty First Printing



On this printing, the head plate colour is closest to dull mauve, while the duty plate colour is closest to orange brown. The used example shown here appears to be cancelled with a 9-bar oval, which suggests that it is from before 1899.

Twenty Second Printing


The head plate colour on this printing continues to be closest to the dull mauve swatch on the Gibbons colour key. The duty plate colour is closest to lake brown.


Here I show eight used examples in my possession. Again, all the dates that I can read here are 1902-1904, while there are also examples with barred oval obliterators. 

Twenty Third Printing



The head plate colour on this printing remains closest to dull mauve, while the duty plate colour is closest to chestnut.

Twenty Fourth Printing


The head plate colour of this printing is still closest to dull mauve. The duty plate colour is closest to red-brown, but paler. It is actually mid-way between red-brown and lake brown.

Twenty Fifth Printing


The head plate colour of this printing is closest to reddish lilac, but is paler. The duty plate colour is closest to lake brown. The dates of these examples are all indistinct, and they are all used. So there is a possibility that these are simply examples of another printing, like say the 22nd printing, that have been affected by exposure to water.


Twenty Sixth Printing


This printing is tricky. The head plate colour is closest to the reddish lilac swatch of the Gibbons colour key, but it is much rosier. However, it is not as rosy as the mauve, claret or rose-lilac swatches. So I would call this a rosy reddish lilac. This is used, so there is a chance that this is not the true colour and that the rosy undertone is caused by the fading of the doubly fugitive ink. The duty plate shade is closest to chestnut, but deeper. This is a very unusual shade for the head plate, and given that it is used and shows clear signs of being affected by the cancellation ink, there is also a possibility that it has been affected by moisture. 

Group 5 - Printings 27-31

This printing is distinct in that there is a coarseness of appearance that none of the preceding printings have. The detail in the hair at the top of the head is almost completely gone, and most of the shading lines in the lower horizontal band of the crown are completely merged together.

Twenty Seventh Printing


The head plate colour of this printing is closest to reddish lilac. The duty plate shade is about mid-way between deep chestnut and lake brown.



These used examples are all dated in 1902, which is relatively contemporary. In all likelihood this printing was probably made sometime in late 1899 or early 1900, and these are all used within 2 years of receipt.

Twenty Eighth Printing


The head plate colour of this printing is pale slate-lilac, while the duty plate shade is deep chestnut. 

Twenty Ninth Printing


The head plate colour of this printing is closest to reddish lilac. The duty plate colour is closest to lake brown.



The date on  this single used example is after 1900, though we can't tell how late after. However, it is clearly a contemporary cancel. 


Thirtieth Printing


The head plate colour of this printing is very close to slate-lilac, but is just a touch paler. The head plate shade is closest to chestnut.

Thirty First Printing


On this printing, the head plate colour is almost an exact match for dull mauve, while the duty plate colour is closest to lake brown, but a bit paler. Unfortunately there is no clear date on this one, but I would suspect that this was probably one of the last printings made, if not the last one.

Conclusion

This was the most troubling stamp of all the denominations that I have examined in this series so far. There seem to be so many late dates on the used examples that either my theory about the wear of the plates is incorrect, and use of plate characteristics to place the printings is incorrect, because the plate did not actually wear evenly, as I have assumed, or there were so many shipments of these stamps, that were received so frequently, that early printings frequently got left at the bottom of the pile and were not used until much later. Given that the number of unsold remainders in 1903 was more than half of the original sending, I think it is safe to conclude that the later hypothesis is correct: that many early printings simply did not get sold over the post office counters until much later after they were first received. What I have managed to establish though is that there was a state of the plate in which there is no merging of the lines in the shading of the hair. I didn't come across this on any of the 5d, 7.5d or 10d values, which suggests that it pre-dates these stamps, and that any stamp showing these characteristics has to have been printed before 1894.

Next week's post will look at the 2.5d ultramarine that was issued in April 1891.