Reexamining the 1847 Colors – Part 3
Third Printing (March 19, 1849)
On March 19, 1849 an order was placed for another 1-million stamps of the 5-cent value (5,000 sheets). This order was almost double the first two printings put together. There was no gap in deliveries between the placing of the order and deliveries. Luff’s date of March 20, 1849 suggests, again, that he obtained his records from the receipt of the order record kept at the printer rather than the post office record.
Until the end of the month only six sheets were distributed (1200 stamps) with route agents at Atlanta, GA and New York City receiving 500 and 400 stamps respectively and 300 stamps going to Georgetown, SC. The next important shipment date was Saturday, March 31, 1849 when 65 sheets were shipped (13,000 stamps) to Hartford, Philadelphia, and Providence. Shipping returned to a regular basis on April 2, 1849.
Allowing for some time in printing, if the new third printing stock was laid over the existing old stock, the first new 1849 printings would have been shipped to Hartford, Providence, or Philadelphia. However, if we assume stock was used up in order of printing, the first shipments of the third printing would not have been sent out until late May 1849.
For the purposes of color analysis, we must make the unlikely assumption that the 1848 second printing and first printing colors were used up first, so the first examples we can definitely assign to the March 1849 third printing would be postmarked in June 1849 or later from shipments received in late May or thereafter. These stamps should have either ‘worn impressions’ or ‘intermediate impressions’. They might be described as ‘fine impressions’. Fig. 8 shows a pair with ‘worn impressions’ posted at Hopkinsville, KY, January 28, 185-, to New Orleans. This is from the only shipment to that town‑400 stamps from the third printing sent September 25, 1848.
It is not clear why the third printing should show such a marked deterioration in quality from the second printing. Logically there should be a degrading in the later part of the second printing and gradually in the third printing. However, I do not find intermediate or worn impressions that I can associate with the second printing and there are few really nice impressions among those associated with the third printing. It is as though the plate had been improperly handled between the second and third printings. Thus, the best of the third printing is rated ‘fine’ or ‘intermediate’. The fine lines in the trefoils have almost entirely disappeared and the stamps are almost worthless for plating by detail characteristics.
Chase reported that by 1849, the color ordinarily was considerably lighter, adding it was,
“ perhaps best described as reddish brown; another rare color, often referred to as the ‘reprint shade’ was used this year. It is a lighter and brighter color and may be called bright reddish brown.”
He did not distinguish between the red browns of the second printing found in early 1849 and the third printing found in the latter half.
|Bright reddish brown
|Dark pecan brown
Supporting the idea that 1849 post-May stamps can be termed 1849 colors, we find examples of “intermediate impressions” in the Matthies sale. First is a pair of red browns used on a cover postmarked at Hartford, CT June 25, –(lot 115) which is presumably from 1849 and the shipment made on June 14, 1849 to that city. Second is an example postmarked at Marion, AL with this intermediate impression red brown (Matthies lot 103). It was not illustrated so the date can’t be determined. However, the first 1847 shipment to that town took place June 22, 1849 so this is the third printing.
1849 Red brown printing ‑ Chase describes the 1849 printing of the red brown as a ‘reddish brown’. An off-cover example, showing the intermediate to worn impression is found illustrated in the Color Encyclopaedia for comparison with the earlier printings.
Two examples from the shipment to Philadelphia of March 31, 1849 can be found on cover. The first (Knapp lot 2223) is postmarked April 7, 1849, while the second is a vertical pair with intermediate impression postmarked Philadelphia April 26, 1849, just prior to the next shipment (Matthies lot 119).
Ashbrook signed a red brown pair on cover postmarked in June 1849 with ‘fine’ impression from the New York shipment of either June 2 or June 22, 1849 (Matthies lot 120). A horizontal pair used from New York to Canada and postmarked January 24, 1850 (before the fourth printing) was sold as lot 23 in the 1981 Siegel Rarities. A Baltimore example, probably from the October 13, 1849 shipment, postmarked October 15, 1849 sold as Matthies lot 175. In the Henry Wenk railroad 1847 collection, there is a LIRR cover with a red brown and a comparatively late postmark, which is dated February 28. It is most likely from the May 2, 1849 shipment to the route agent. A late use of this red brown shipment to New York was in the Stollnitz holding and was sold as lot 503 of the dispersal of it. It is a use with a Boyd yellow green (Scott 20L7) on a cover to Dobbs Ferry, NY, figure 9 used August 1, 1850.
A red brown with ‘fine impression’ used April 10, 1849 killed by the St. Johnsbury, VT scarab was apparently ordered from another office. It may be from the second printing but is most likely sub ordered out of the Rutland, VT April 6, 1849 shipment. It sold as Matthies lot 157.
The reddish brown shade was used in a shipment to the Hudson River Mail agent as evidenced by an example postmarked May 11, 1849 (Matthies lot 207) as well as a second example, on cover signed by Ashbrook, postmarked August 3, (1849?), (Krug sale lot 38).
Some of the other probably third printing reddish browns include the following: An ex-Krug cover with reddish brown stamp ‘fine color and impression’ postmarked at Chardon, OH October 17, 1849 (Matthies lot 37); a horizontal pair postmarked Eastport, ME July 1849 (Matthies lot 132); and a Trenton, NJ cover postmarked September 13 1849 with the stamp killed by a Trenton star (Matthies lot 160).
Other items are an Elmira, NY use postmarked in 1849 (Matthies lot 164), and an untied example postmarked on a January 25, 1850 Jamestown, NY cover (Matthies lot 41). (This town received three shipments during 1849. This should come from the last.) Another example is a cover with two red browns, one a 90R1, postmarked New Haven, December 1849, probably from the shipment arriving November 19, 1849.
A red brown use at Oswego March 15, 1850 (Matthies lot 181) was described as ‘fine color and impression’. It probably came from the shipment received there on October 19, 1849. From the shipment to Northampton, MA of October 11, 1849, we find an example with ‘bright color and clear impression’ postmarked November 21, 1849 (Matthies lot 45).
Pale or light red brown ‑Chase does not recognize this as an 1849 printing shade. Lot 371, light reddish brown, of the Frajola McInroy sale (February 9, 1985) shows this color off‑cover. One of the few reported on cover examples is a worn state example found on a cover postmarked Bridgeport, CT January 20, 1850 (Matthies lot 33). It probably originated in the shipment of October 9, 1849 to that city. The McInroy sale item is postmarked Philadelphia in April and probably was in the April 1, 1849 shipment to that city.
An interesting example of this ‘pale red brown’ shade is found as Knapp lot 2203 where it is described as a “soft impression”. This item is on a cover going overseas ‘perAfrica’ and postmarked New York August 24. As transatlantic students know, the only date possible for this is 1852, making this a very late use of the 1847 issue. TheAfrica sailed August 25, 1852 from New York. The ‘soft impression’ suggests this is the third printing rather than the later cleaned plate fourth printing.
Bright reddish brown ‑Although reported as a third printing color by Chase, it is not particularly common. An off-cover example sold as lot 370 in the Frajola McInroy sale (February 9, 1985), but a reference copy is not included in the Color Encyclopaedia.
An on‑cover example, with ‘heavy impression’ can be found on an 1849 cover postmarked at Philadelphia (Fox, Hollowbush II, lot 503). Another example, postmarked New York December 18, 1849 is found on a cover originating in England, which was forwarded on to Richmond (Fox, Hollowbush II, lot 525). The color was apparently issued to the route agent for the U.S. Express Mail, for we find a cover bearing this stamp postmarked November 17, 1849 with the New York U.S. Express Mail handstamp (Knapp lot 2215).
Brown–Chase did not recognize this shade and it is not illustrated in the Color Encyclopaedia. An example of this dark reddish brown was sold as lot 372 of the Frajola McInroy sale. A cover, postmarked Boston 1849 sold as lot 1547 in the Blake sale (Siegel December 9, 1969). Another cover with this shade was postmarked Mobile, AL May 16, 1850 (lot 141 Siegel 1982 Rarities). This is just prior to a new 1850 fourth printing shipment and thus represents the printing shipment of February 6, 1850. It may be a second printing late use.
Orange brown‑There are several colors recorded that suggest orange brown is a color in the third printing. One, with a ‘fine impression’ description, is postmarked October 9, 1849 (Matthies lot 143). Postmarked at New York and bearing a strip of four, this should have been from the New York delivery of September 18. New York had received a number of shipments from the third printing prior to this.
The Wenk 1847 railroad collection has three examples of the orange brown that seem to come from the third printing. The first is postmarked July 27, 1849 with the Albany & Buffalo R.R. cds while the other two were used on the Housatonic. One is postmarked January 29, 18(?) and probably represents an 1850 use. It could be a late use second printing; the other is almost certainly an 1850 use as it is the late handstamp with the 10 in center, the listing example.
Bright brown ‑A cover with an intermediate impression 5-cent bright brown stamp, postmarked Burlington, VT August 29, 1849, is on record as indication this shade was part of the third printing (Matthies lot 36). We know that a new shipment of the third printing arrived at Burlington the day before.
Pale brown ‑Several examples of this shade exist on covers that bear stamps apparently from the third printing. Not all pale browns used in 1849 qualify. For example, the pale brown used May 4, 1849 on the Boston & Maine RR (Matthies lot 189) is probably from the second printing, as is another pale brown used November 10, 1849 on the Norwich & Western R.R. (Matthies lot 199).
A third printing pale brown example is a pair signed by Ashbrook on a Providence, RI cover postmarked December 1849, and addressed to New Orleans (Krug sale lot 60, Siegel May 21-22, 1958). A second example was postmarked at New Orleans October 21, 1849 and used overseas to Bordeaux (Matthies lot 184). This is apparently from the August 14, 1849 shipment.
A particularly interesting use is a pale brown postmarked Tampa FL on April 2, 1850 (Matthies lot 104). The date is before a direct shipment to Tampa was made. This stamp was likely sub ordered by Tampa from either Jacksonville or Key West, both of whom had third printing stamps in stock. Another probable third printing pale brown is postmarked South Coventry, CT on May 8, 184‑? (Matthies lot 97). This example has a ‘late impression’ so it cannot be earlier than 1849. A similar item was postmarked at Saybrook, CT ‘fine impression’ and bears a manuscript cancel, untied. (Matthies lot 96). It is not illustrated so authentication is difficult.
Brown ‑Brown stamps on dated 1849 covers could come from either the second or third printings. Typical of the problem is a second printing brown tied by an orange red 5 in circle to a Newark, NJ cover postmarked July 20, 1849 (Matthies lot 165). Newark did not receive a third printing shipment until October 3, 1849 so that this was either brought in or was a second printing remainder from the October 10, 1848 shipment.
Several definite third printing browns are those ‘late impression’ items found on a cover postmarked Middlebury, VT November 2, 1849 (Matthies lot 23) and on an 1849 cover postmarked at Rochester (Matthies lot 55). This latter has a large preprinting paper fold to add to the interest. Another brown with a ‘late impression’ is found postmarked on the Mad River & Lake Erie R.R. February 8, 1850‑just before the fourth printing was ordered ‑(Matthies lot 194). The Wenk 1847 railroad holding has a brown used on the Boston & Maine R.R. July 3, 1849 that may be from the second or the third printing.
Grayish brown‑Chase reported that grayish brown, dark grayish brown, and dark olive browns are found ‘only from worn plates’ and classes them as 1850 colors while the dates of the dark olive browns ‘cannot be earlier than about May 1850.’ As discussed in the first printing, there is a brown, olive cast known postmarked in 1847 at Baltimore. There are also anomalous gray browns in the second printing such as the one used at Providence October 7, 1848.
A key piece of evidence that this shade was part of the third or 1849 printing is the ‘worn impression’ example postmarked at Catskill, NY July 2, 1849 (lot 318 Haas sale, Siegel March 15, 1983). Catskill received the shipment from the third printing May 28, 1849. Another grayish brown that may be from the third printing, or even from the second sent out at year’s end, is a pair postmarked Baltimore December 31, 1849 (Haas lot 326, Siegel March 15, 1983). It should be from the December 5, 1849 shipment to Baltimore.
There is a grayish brown used on cover from Auburn, NY postmarked May 3, 184‑, described as having a ‘splendid color and impression’ (Matthies lot 64), which is probably from the same shipment that delivered the Auburn orange brown postmarked June 2, 184‑ (Matthies lot 78) described as having a ‘proof-like impression’. Auburn never received a direct shipment according to the Wenk Transcript, but route agents there got shipments sent August 19, 1848, October 31, 1848, and February 28, 1849 so these could have come from one of those second printing shipments and been used in 1849 or later. The descriptions suggest the stamps are second, not third, printing.
There are other examples of the grayish brown used in 1850, which are from worn plates and consequently from the 1849 third printing sent out in 1850. Such an item is the dot in S grayish brown used in 1850 at Syracuse, NY, unillustrated, but Matthies lot 99.
From Utica, NY we find a grayish brown on cover postmarked May 24, 1850 with a ‘late impression, dry print’ which was presumably part of the February 20, 1850 shipment to that town (lot 814 Ernest Jacobs sale, Siegel September 26‑28, 1972). In the Wunderlich dispersal, lot 20 is a grayish brown postmarked Chicago September 20, 1850 that appears to come from the August 3, 1850 shipment to that city ‑a fourth printing shipment with a third printing stamp.
Dark olive brown‑Chase reports this is an 1850 color from ‘worn plates’ with use not before May 1850. It is a scarce color and rarely described in auction catalogs. A first printing ‘brown, slightly olive cast’ was noted used in 1847 while an off-cover olive brown from the McInroy holding is shown in color plate II, #11 to show how the color can be approximated. A true olive brown is shown in that plate as #12. It is interesting to note that no example of the olive brown or dark olive brown was reported in the Emerson, Caspary, Lilly, Grunin, or Sweet collections, suggesting its rarity or the ease with which it can be mis-classed.
Purple brown ‑ Only one example of this shade has come to my notice. It may well be an item that should be called dark olive brown, but the Knapp sale lot 2296 calls it purple brown. When re-offered in the Meroni sale (lot 1244, Fox November 13, 1952), it was called dark brown. It is datelined May 17, 1850 at Brownsville, TX and should be from the February 27, 1850 shipment of one sheet to that town (fourth printing), received there March 12, 1850. This date is right on the cusp between the third and fourth printings. This cover was included, as a recent acquisition, in Dr. Kapiloff’s Ameripex exhibit.
There is a problem with this cover in that it discusses the explosion of the steamboat Louisiana at Brownsville; however, the Louisiana blew up November 15, 1849 at New Orleans, which is where this letter is addressed!
Shipped from earlier printings ‑The 1849 printing was the first where shipments exceeded the printing prior to the making of a new, fourth, printing order. Consequently, if stamps were stacked with the latest on top, we know all one million third printing stamps were distributed. Toward the end of the shipment period examples had to be shipped from the first or second printings. The shortfall was only 4,700 stamps, so not too many earlier printings would have had to be distributed in 1849 or early 1850.
Among the apparent first printing items distributed or used in 1849 are a turned cover with a seal brown postmarked at Troy, NY June 13, 184‑ with a turned postmark of July 27, 1849 (Matthies lot 65). There is a ‘wonderful impression’ dark brown postmarked with a circled 5 at Baltimore November 30, 1849. Also unusual is the deep brown, postmarked February 16, 1850 at New York (Wunderlich lot 43). This cannot be from the fourth printing, which was not available when the January 25, 1850 shipment to New York was made.
The interesting feature of this last item, from the first or second printing is that the paper is distinctly vertically ribbed. It can be presumed that the paper was produced by accident at the paper mill as a result of a slightly damp sheet of paper passing a ridged or grooved roller at the dry end of the Fourdrinier machine. The item probably merits further investigation.
Fourth Printing (February 14, 1850)
The Wenk Transcript shows that the fourth printing order was placed February 14, 1850 for one million 5-cent stamps. Luff gives a date of February 5 but this is an obvious typographic error for February 15 as his dates consistently run slightly later than the Transcript ones, presumably because they are receipt dates. As of Friday February 15, 1850, there were some 225,000 stamps (1,1251/2 sheets) that had not been shipped. Thus, shipments did not need to be interrupted for the new printing to arrive.
The next major shipping date was February 27, 1850 when a large number of towns were supplied. However, on Wednesday February 20th, we do find 25,000 stamps being shipped. These were distributed as follows: Louisville (5,000), Utica (5,000), Buffalo (5,000), Portland, ME (3,000), Galveston (1,000), Milwaukee (1,000) with smaller quantities going to Lee, MA (600), Newark, NY (600), Bath, NY (500), Oxford, NY (500), Potsdam, NY (600), Warren, 0. (600), Winchendon MA (400), and West Troy, NY (200). There was also a distribution on Monday, February 25th to Washington, DC of 2,000 stamps. All of these should be from the old third printing or earlier.
If the stamp stock was stacked so that the latest printings were on top so that the latest items were shipped first, the worn plate third printing should be exhausted by February/March 1850. Only if ‘first in, first out’ inventory control were used would there be third printing stock in 1850, particularly after February. Chase’s color analysis makes it clear that some worn plate items were used, if not shipped, during 1850.
As noted earlier, some 1847 printings were distributed in 1849, suggesting that first printing stock, for some reason, was put on top of second printing 1848 stock, although the overall pattern is new stock is normally on top. This shift of stock may have been sufficient to move 1849 third printing stock into 1850, giving us the grayish browns Chase found to be the usual colors of use in 1850. Additional research is needed to see if the 1850 grayish browns are worn plate or cleaned plate stamps.
Chase reported the 1850 colors as follows:
|Light Mars brown
|Dark grayish brown
|Dark olive brown
He commented that the usual colors used in 1850 were grayish brown and dark grayish brown, with a rare shade being the dark olive brown, found only from worn plates as are the grayish brown and dark grayish brown. It would appear that what Chase is describing are remainders from the earlier printings used by the public during the first half of 1850 rather than the colors shipped to the post offices during that year. The worn plate stamps are certainly third printing while some earlier printing sharp impression items would be in the 1850 shipments as old stock was used.
Chase also noted that during the latter part of the year, ‘a decided change was made in the ink, and the orange and colors more or less related to it appeared.’ This confirms that the third or earlier printings were in the post offices during the first half of 1850 with the new fourth printing coming into general use in the latter half of the year.
There is an analytic problem in this Chase statement in that the fifth printing took place in December 1850 and it is known to have contained the red orange and orange shades. Chase may not have given us any reliable information about the fourth printing.
The Cleaned Plate and Ink Formula Change
Two significant events took place during the latter half of the 5-cent 1847 plate. One was that it was cleaned. The other is that the basic formula for ink composition changed. The chemical composition of the early inks were lead dioxide, a dark chocolate brown pigment, and orthoplumbate, a red-orange compound. The red orange of the last printing on the other hand used lead chromate (chrome orange) as an important colorant.
When was the 5-cent 1847 plate cleaned? It is generally accepted today that only one plate was ever made and that all the 5-cent stamps came from this plate. We also know it was cleaned, and possibly reworked, prior to the December 1850 fifth printing. The question is whether it was cleaned prior to the fourth printing or during the third 1849 printing.
Creighton Hart has raised the question of whether the plate was cleaned in 1849 or 1850. I believe the confusion stems from some ‘proof-like’ impressions found late in 1849 or early 1850. The detailed shipment records strongly imply that late 1849 proof‑like impressions are stamps from the first two printings.
The third printing took place during a short period early in 1849. Most of the stamps we can definitely associate with it show ‘intermediate’ or ‘worn’ impressions. Consequently, the cleaning could not have taken place until later.
It is not logical that once printing of either the third or fourth printing was begun, it would be stopped and the plate taken off press for cleaning. Rather, as the total printing period for each run was only a week or so, the cleaning would have been done first or after.
Enough examples of dated covers bearing stamps associated with the fourth printing are known with ‘sharp’ impressions that it is clear the cleaning was done prior to the February 1850 printing run. Whether immediately prior or just after the 1849 printing is not clear. It is also not important.
Such a sharp impression is found on a cover postmarked Greensburgh, PA May 20, 1850 (lot 62 Matthies sale). The cover bears a brown orange stamp tied with a greenish blue grid. The catalog notes ‘gorgeous color, sharp impression.’ A shipment to that town was made in April 1850 and this cover presumably used a stamp from that fourth printing shipment. Brown orange is not a color reported on earlier printings.
The official records show that some 22,000 sheets of 5-cent 1847s were printed. It would appear that the plate began to deteriorate after 7,000 sheets were pulled and continued to show deterioration for another 5,000 sheets. Then we get a new ‘sharp’ impression, which lasts for the remaining 10,000 impressions.
A Rawdon, Wright, Hatch & Edson circular of July 1853 noted that steel bank note plates engraved and printed by them are,
“ warranted to give 30,000 good impressions before, and 25,000 do. after retouching. Copper do. 3,000 before, and 2,000 do. after retouching ..”
Clarence Brazier read this statement as confirming that the 1847 issue was printed on steel plates, although Brookman and Ashbrook disputed the conclusion. We do know that Rawdon, Wright, Hatch & Edson used unhardened steel plates for the first Canadian issue of 1851 and that the destruction notes of the 1847 plates refer to steel plates.
The printing firm of Toppan, Carpenter reported in 1864 that a given set of plates ‘would have to be retouched every three days allowing 50,000 impressions before retouching charges.’ An 1856 letter from Perkins Bacon reported that because of problems with the original die which was used for thousands of transfers they found that,
“…from printing 100,000 sheets from one plate we were reduced down to 20,000 … (therefore) we were obliged to prepare 4 or 5 times as many plates as we expected….”
This suggests that the quality of the original die might drastically affect the number of impressions available from a plate laid down by it, or the quality of the fine lines on the pulled impressions.
Another problem is ink. The Canadian government, in 1866, asked its printers for a guarantee that each plate,
“… shall give 25,000 impressions, before retouching and 15,000 afterwards, excepting where the patent green is used when the number of impressions guaranteed shall be only 10,000….”
From the cited, and other, contemporary evidence I conclude that an unhardened steel plate should not have a problem in generating up to 50,000 sheets with reentry needed only after about 25,000 or 30,000 were pulled. However, certain inks could reduce this drastically.
It is generally conceded that the reason for the deterioration of the 1847 5-cent plate is the character of the ink. To the extent that earth pigments such as ocher, sienna, and umber were used to obtain brown or orange brown colors in the early printings, there would be plate wear. These minerals are typically contaminated with quartz, which abrades the plate. Brookman, among others, assumed the 5-cent ink was of this class. The early ink may have used such abrasive materials.
However, in Volume I of the Perkins Bacon Records we find the ink formulae for brown ink for the 6d New South Wales stamp of the 1850s. It calls for a mix of three parts Prussian blue to six parts each of rose pink, pale vermilion, and flake white. Further, Roy White, in Color in Philately, has shown that the 1847 early colors were composed of lead oxides rather than iron oxides. We also know that ocher was not mined in the U.S. until 1877, while calcined limonite (burnt sienna) was not dug here until 1856.
Although Prussian blue is a combination of ferrous salts and potassium Ferro cyanide, it lacks the abrasiveness of quartz. Rose pink is believed to be a non‑mineral item such as a lake of the madder root. Vermilion is a mercuric sulphide product. None of these seem to have the abrasive qualities of the earth colors previously believed to have been used. Nevertheless, all printing ink has some abrasive qualities and this is compounded by the wiping and polishing action.
Printing plate impairment as a result of wiping or polishing is most notable in the delicate design lines. These are shallow and thus the first to show wear when they can no longer hold quantities of ink. Wiping patterns cause some areas of the plate to get more wear than others. In many cases the outer edge goes first and the wear works into the center as a result of the circular sweep of the wiping. The pressure of printing itself creates wear, usually at one end of the plate so that many printers in the classic period reversed the plates periodically so that the pull was in the opposite direction.
I should like to propose another possible cause of deterioration. It is the fact that the third printing was the first we are positive was produced during winter. The second printing is reported as May although it may have occurred in March.
It was only two days before the third printing order was received that the Hudson had thawed sufficiently to allow boats to venture north to Albany. Winters were much more severe then than now. It is quite possible that the poorly heated workrooms and overnight cooling of plates and inks allowed the ink to gum the design. Temperature may also have created plate maintenance problems.
In connection with the weather thesis, there is supporting evidence from the 10-cent 1847s. Descriptions normally don’t comment upon the quality of the 10-cent impressions. However, those few items that are described as ‘fair’ only are concentrated in the 1849 printing. This suggests a similar, but lesser problem, affecting only the 10-cent plate.
Once cleaned, the plate was used to print 10,000 more sheets in two printings. Brookman has commented upon the late printings, that although the,
“ stamps come with good impressions they never, in our opinion, are found with the truly sharp impressions of the earliest printings of the stamp The better impressions of the 5¢ which appeared in 1850 have long been called ‘Plate Two’ copies, but it is our opinion that all the 5¢ were printed from one plate, and that these particular copies were printed after the plate had carefully been cleaned and during a short period when extra care was used in the printing. These better impressions were not long produced and the last printings from the plate were very poor indeed. The scarcity of the ‘C’ and ‘D’ double transfer leads one to suspect that at least these two positions were reentered at the time of the probable cleaning of the plate…”
It is quite true that stamps of the fourth and fifth printings lack the sharpness of the earlier printings. They are frequently called ‘fuzzy’. On the other hand, there is good reason to doubt Brookman’s comment that the better impressions lasted only a short time. We find quite good examples throughout that we can identify with the fourth printing and no ‘very poor’ impressions are recorded on the brown orange or red orange shades of the fifth printing in the auction records I consulted.
What we do find is a group of shipments involving stamps from earlier printings. A number appear to be from the third printing, particularly in the February and March 1851 deliveries. It is not unlikely that these are the ‘very poor’ impressions noted by Brookman.
While I have not been able to find a good technical description of how a gummed-up worn plate is reconditioned, an acid bath is probably part of the treatment. As James Baxter noted in Printing Postage Stamps by Line Engraving,
“… a marked deterioration in the delicate lines of the design develops as a result of constant wiping, making it necessary to deepen (retouch) the worn lines with a graver.
After a plate has been tempered, re‑cutting becomes a difficult task and is resorted to only when there is no other possible remedy … Sometimes the required lines are engraved in an etching ground, a reservoir of wax is built around the part to be restored, and the ‘re-cutting’ is effected with an acid.”
When an acid is used, the fine clear lines that characterize line engraving are lacking. Although the ground holds the acid in bounds, it eats both sideways and down so that the lines diminish or increase abruptly rather than tapering as when a graver is used. The result is a somewhat ‘softer’ or ‘fuzzier’ look.
In acid etching the etching fluid is ‘flowed’ gently back and forth over the exposed metal by rocking the part to be etched in the fluid or by using a soft brush. When the fine lines have deepened as desired, chemical action is stopped by removing and washing the plate.
The look of the fourth and fifth 5-cent 1847 printings is similar to that where an acid wash was used to deepen the lines. It also appears that the ink formula was changed for both last printings with chrome orange being used rather than the earlier ink components.