Part 1- Reexamining the 1847 Colors
© Calvet M. Hahn 1986
Alhough Luff had listed a number of colors for the 5-cent and 10-cent 1847 issues, the basic color work on this issue was done by Dr. Carroll Chase in a series of articles appearing in the 1916 Philatelic Gazette. A few items have appeared subsequently, with the major effort seemingly focused on the so-called orange hue of the 5-cent. Both Ashbrook (extensively) and Perry (slightly) discussed this rare color.
Seventy years later it appears time to reexamine Dr. Chase’s work and attempt to incorporate new data as well as to revamp it into the printing sequence rather than the cover date sequence he used. Dr. Chase listed colors by order in which he noted them, but without specific cities and dates.
One important characteristic to observe is the paper. The 1847 issue was printed on wove rag stock with a bluish tint caused by the use of ultramarine. Chase had chemist D.D. Berolzheimer examine copies and he reported the ultramarine pigment was identifiable as being made from lapis lazuli.
Luff reported a grayish pelure paper found with the dark brown 1847 color. What is termed pelure in this issue is probably paper left too long in the beating vat during the pulp staged. It is thin and somewhat transparent but isn’t truly pelure.
Stamps do exist on a white paper. Chase noted stamps still on original cover that have a yellowish white appearance. These are normally dated in 1850 and he reported his chemist friends suggested a more acidic gum was used that gradually bleached the ultramarine pigment. He noted stamps on blue envelopes often show bleach spots on the envelope under the stamp.
Brookman reported a letter from Ashbrook stating that both he and Ezra Cole had concluded there were 1847s on white paper that had not been bleached from the bluish shade. Nevertheless, all or almost all white paper examples found prior to 1850 with unusual shades have had chemical alteration.
Chase reported the 5-cent 1847 ink was a iron oxide-based ink. Roy White in the Philatelic Foundation’s Color in Philately disputes this, stating that more modern spectrographic analysis shows the base was lead oxide for the brown shades. He reports that lead dioxide is a dark chocolate-brown and can be decomposed by reducing agents to create a yellow lead oxide. He notes that lead oxidizes to a black color.
White added that the orange colorant in the 1851 printing contained chromium when analyzed by spectrograph. This suggests lead chromate (chrome orange) was an ingredient of the late printings. It should be noted that it could be chemically altered to produce redder or darker orange hues.
Sulphur in the air, together with skin acid, have darkened most of the 1847 issues. They cannot be ‘reduced’ back to the original colors without some damage, when then can be ‘reduced’ at all.
The first order to print the 5-cent 1847s was placed June 1, 1847 and apparently received June 3, 1847 if Luff’s data is correct. It included 3,000 sheets (6000,000 stamps) of the 5-cent value. The stamps were ready Saturday, June 26th. Special agent John Marron, who was to head the new stamp department in Washington, came to New York on the 29th to take possession.
Marron hand-delivered 300 sheets to the New York postoffice on July 1 and then went on to Boston to make a 200-sheet delivery on the 2nd. Returning to New York, he took the train south, stopping overnight at Philadelphia where he dropped off 200 sheets of the 5-cent on Wednesday the 7th. He then delivered some 15 sheets to the Washington office on the 9th. Subsequently, 2½ sheets were delivered to Baltimore on July 16. No other 5-cent stamps were distributed until July 29th.
We do not have production records from Rawdon, Wright Hatch and Edson from 1847 to tell us just what happened in the printing. We do know that in 1851 on the first pence issues of Canada, they took just over a week from receipt of the order to plate proof (Wednesday march 27 to Monday April 7). They then made delivery of 500 sheets by the 15th, a week later.
Further, we know that in producing the 1861 issue, the National Bank Note Company had print runs of about 1,700 sheets per press per week.
As there was only one 5-cent 1847 plate, of 200 stamp impressions, it would be about two weeks of production at the 1861 rate to produce the initial printing. If we subtract the time from order to plate proof needed for the Canadian 3d beaver stamp from the 1847 time available (June 3-266), it would appear that production covered 13 days and was at the rate of 230 sheets a day on a six day week. This means production stretched over three weeks.
The production figures adduced above are in line with other data such as the rates of printing found at Perkins Bacon in England where production was running about 400 sheets a day by 1850, with the initial day or so being below 150 sheets.
In addition to considering the production rate, it is necessary to remember how the printing was done to understand the colors of the 1847 issue. First, the paper was printed wet. Once printed, they had to go to a drying room before gumming. After gumming, it had to be dried again before finally ending up in the stack, which would eventually be used for shipment. This means the sheets in the stack used for shipment are not necessarily in the order in which the sheets were printed.
John O. Griffiths refreshed my memory in regard to the impact of security regulations on this sheet stacking process. As he correctly reminded me, each group of sheets had to be locked up each evening when production ceased. Thus, certain printed sheets would have gone forth to the drying room to be stacked there. Others would have gone to the gumming room to be grouped there. Consequently, any of the 230 sheets produced in a given day might end up several hundred sheets away from the next sheet produced when the finally stacks used for shipment are considered.
Griffiths also pointed out that when the government wanted the first sheets off the press as control samples, Perkins Bacon objected. The printers insisted that it took a certain amount of production before the pulls were right to use as samples so that the test sheets on a new plate were not the first pulled off but more like the 125th.
At the time of the 1847s, color mixing was still an art. Periodically, a new ink formulation would have to be made and checked. A new batch might be made up weekly or daily. Even when the ink was the same, it needed stirring and occasionally additional thinner. Consequently, there could be substantial variation even within the same day. The first sheets pulled might show a bit more black was needed so that we get a difference in shades. Or, the printer might not have stirred for a while so that production was running light and then, when the sedimented paint particles were stirred, you et a much darker shade such as the dark brown shades of black grown, etc.
The rare first shades that we know probably came from the tests made at the beginning of a day, a new batch of ink, or from the tail end of a batch. This implies that only a few sheets in each production period (day or week) would represent a rare hue or shade variety. These items would not be rejected but included in the stock turned over to the post office agent. Thus, we have something that is almost a ‘color-proof’ included in the regular shipments. To the extent the shipment stacks followed the printing process fairly closely, the rare shades would end up being distributed to the same town or the next one to it in the shipment sequence.
Although the December 15, 1910 Philatelic Gazette contained a discussion of the official shipment records, which were discussed again in the 1912 June issue by J. M. Bartels, Carroll Chase did not grasp its critical importance to his color analysis of the 1847 issue. Consequently he chose to group colors by year of postmark rather than by printing runs.
First Printing (June 1847)
Dr. Chase reported the colors of the first 5-cent printing in use order as follows:
|Bright orange brown
|Dark Van Dyke brown
The black brown and the true orange were termed the rarest 1847 issue shades. Chase also noted that ‘reduced’ copies could be found in varying shades up to a pretty fair black and that these false colors need to be ignored as do the faded stamps. The true black brown e reported was the deepest shade in which the stamp is found and was released late in 1847. Today we call the seal brown the darkest of the black browns.
What did Chase use for his color source? From the Elliott Perry plating of the 10-cent 1847 article published in the Collectors Club Philatelist in 1924, we learn that the Chase holding was 378 5-cent and 111 10-cent stamps and that while Chase considered it the finest ever assembled, Perry found it very heavily duplicated by the McDaniel holding which he had bought for Ackerman. This Ackerman holding had been assembled during the 1916-1919 period when Chase was out of the market. We do know that Chase did not have access to the Lord corner of 2,686 5-cent stamps sold to Klemann in 1916 or the 400 Lord examples sold to Richardson. It would appear that the dated series of covers, which yielded much of his data was the Hagerstown Bank find.
The reason for examining the source used by Chase for his color analysis is that: a) There has been no serious revision of his charts in the subsequent 70 years. Major students such as Ashbrook, Perry, and Brookman repeat his data with small emendations, reserving their analytic efforts for such things as a discussion of the copper plate and two plate myths. b) The color listings are clearly inadequate in the omission of major shades such as the brown, light brown, and pale browns of the second 1848 printing, or the red brown, bright red brown, and dark red brown of the first 1847 printing.
While it is possible that some of Chase’s orange browns are now seen as red browns as a result of chemical change ‘reducing’ and darkening the stamps, this does not apply to other problem areas.
What has been focused upon, although in no rigorous fashion is a more detailed breakdown of the dark brown and the black brown shades. It would appear that marketplace pressure has created a demand for other scarce shades which are not recorded, such as chocolate brown, dark chocolate brown, walnut brown, and seal brown (a black brown tint).
In his Color Encyclopaedia, Roy White comments that the black-brown, chocolate brown, and dark brown specimens he examined had approximately ‘twice the ink colorant concentration’ of the pale brown, red brown or gray brown examples. He added that only in the dark brown would an increase in ink depth change the basic color. Such an increase would give us the rare blackish brown shade.
White’s remarks strongly suggest that, for at least one of the three weeks the 5-cent stamp was on press (13 working days), a heavier ink colorant concentration was used. This gives us the dark browns. Sometime during that week, the ink was allowed to get excessively concentrated so that the rate black brown color group of shades resulted.
The official designation of the 5-cent stamp was to be ‘light brown,’ however, this is not one of the colors found in the first printing. Dr. Chase reported two major color groups used in 1847—the dark browns and the orange browns. We now know there were three, as red brown is a fairly common 1847 period color.
One important clue to the first printing colors is the fact that neither Boston nor Philadelphia received a second shipment before February 1848, while New York’s first shipment was not replaced until October 30, 1847. Consequently all colors from those cities before the replacement shipment had to be from the first delivery and presumably from the top of the stack of 3,000 sheets printed. Logically, they should be among the last sheets printed, unless two stacks were made.
To repeat, if Rawdon, Wright, Hatch needed the same period from order to plate proof as they did for the Canadian 3d beaver stamp, the 5-cent 1847 was produced in about 13 working days (including Saturdays) at the rate of 230 sheets daily. Consequently the New York shipment involved at least two days production. If the New York stack was taken off first, then the Boston, and then the Philadelphia, each city would have received two days production and thus possibly two colors. The ratios would be 230/70 for New York, 160/40 for Boston, and 180/10 for Philadelphia. As the actual colors reported for each of the three cities show three shades, it is apparent that there was some mixing of the sheets in the printing, drying, and gumming process.
First Printing Dark Browns
The dark brown shades are characteristic of the first printing, particularly those that are ‘proof-like’ impressions. While some dark browns are reported in the period of the second printing, they may be hangovers from the first printing.
Creighton Hart, in Chronicle No. 46 disputed Chase’s listing of the orange brown as the first-used color. Hart maintained the dark browns were first, citing his July 7, 1847 cover from New York to Poultney, VT, as well as an example posted in Dusseldorf, Germany that was forwarded from Philadelphia on July 16, 1847 with a dark brown 5-cent stamp to Boston. Hart also owned a July 20, 1847 dark brown used from New York.
George Hargest analyzed the Dusseldorf cover for rating in the same issue of Chronicle. Its origin and transit markings are of June 1847. When it passed through England, it would have gone onward to America on the Cunarder Caledonia arriving at Boston on July 2, 1847. Then it would have gone on as a stampless letter to Philadelphia, sitting there almost two weeks before being posted back to Boston.
Hart’s argument is probably academic, for several shades were found in the first shipments of New York, Boston, and Philadelphia. Too, apparently he did not recognize that part of Chase’s evidence was the J. J. Cone cover postmarked New York July 9, 1847 and addressed to Middletown, CT which bore an uncanceled 5-cent orange brown adhesive. Chase was also using the Dr. Evans’ Whelen-find cover with an unused example datelined July 7, 1847 and bearing an orange brown stamp. It was docketed as received July 8, 1847 at Philadelphia. It sold as lot 492 in the second Hollowbush sale at John A. Fox’s auction house. We also have a record of a July 15, 1847 orange brown stamp cover postmarked at New York. This item is on white paper rather than bluish, so it may be a changeling.
Both orange brown and dark brown shades were represented in Philadelphia’s first shipment. Chase reported a horizontal pair of the orange brown on a Philadelphia to Kentucky cover postmarked July 31, 1847, while Hart’s Dusseldorf cover of July 16, 1847 forwarding shows the dark brown use.
In Boston, the dark brown is represented by a proof-like impression adhesive in that shade found on a U.S. express mail cover postmarked January 1848, Knapp sale lot 2216. This was prior to the second shipment to Boston.
Chestnut brown—The only example I find of this shade in auction catalogs is lot 300 in the Siegel Haas sale of March 15, 1983. It is on a cover postmarked at Boston April 28, 1848, just after the second shipment to that city arrived on March 17, 1848. I would not record it separately as a color except for the fact it bears Ashbrook’s signature attesting to the shade. It is probably just the Ridgeway version of a dark brown.
Seal brown—This is a modern name for one of the darkest versions of the dark brown or black brown. It is a scarce shade. There is some disagreement as to the color with the Roy White Color Encyclopaedia showing an off-cover example with black grid cancel that appears to have slightly less brown than does the black brown. Other color students would state it is almost completely black as on a seal’s coat. An off-cover example with a wide cross grid killer was lot 341 in the Matthies sale.
Among the on-cover examples cited in auction catalogs as ‘seal brown’ is an example postmarked New York September 11, 1848 (lot 17 of the Wunderlich sale at R. A. Siegel January 29, 1976). The Matthies sale lot 65 shows an example postmarked Troy, NY June 13, 1849. By that date Troy had received four deliveries.
Another late use was Matthies lot 155, an 1849 use from Norwich, CT tied with an odd-grid. This was prior to the first shipment of 5-cent 1847s to that town and suggests it was from a Norwich office order made from another post office. Matthies lot 163 shows another example of the seal brown, posted at Windsor, VT in 1848. It is not tied to the cover. Windsor never received direct shipments.
In my opinion, the finest example of this rare shade is the example sold as blackish brown as lot 40 in the Ishikawa sale at Christies September 28, 1993. It is on a cover postmarked at New York July 9, (1847, as confirmed by shade), and is addressed to Litchfield, CT. This is just two days after the earliest 5-cent 1847 cover so far recorded. The adhesive is noted as having a ‘gorgeous intense deep rich shade possessing a sharp proof-lie impression, tied by red square grid on fresh blue folded cover.’ (Seefigure 1.)
Black brown or blackish brown—Chase reported that this was the deepest shade found on the 5-cent 1847 and that it first appeared very late in 1847. Brookman noted it was about as rare as the real orange shade. The problem is that, even if we make a distinction between black brown and seal brown, we have too many examples, some of which may be changelings. It is preferable that the example have the bluish paper, for changelings usually have an off-white paper. Too, the item should fluoresce orange brown or brown orange.
Plate Description 1
Plate Description 2
Plate Description 3
Color Plate Number 1 shows three items (1-3) that could be classed as black brown. Some might call one or more of these seal brown. The first item is on a yellowish gray paper and fluoresces brown orange. The second item is shown as an example of a color changeling. This stamp originally had a black cancel that was removed in a successful attempt to defraud the postoffice when the stamp was used again at New York, where it got a red square grid. The paper is also yellowish gray, but the stamp fluoresces black. The third example is also on the yellowish gray paper, which is known legitimately on some of the dark browns as discussed earlier. I did not fluoresce it, as it was not convenient to do so when it was photographed.
An off-cover pair of the blackish brown is illustrated in the White Color Encyclopaedia. Comparison of this pair with the seal brown gives the best currently available check for differences.
There were a number of black browns in the Emerson sale (Kelleher November 16, 1946). These are important references because a number of the Chase examples ended up in Emerson’s collection, particularly if they were unusual, and Perry put together the Emerson material. There was an extremely fine example with black grid offered as lot 82 and one with red grid as lot 139. Other examples, ex-Emerson, are lots 194, 203, 219, and 232.
The blackish brown was found at New York as evidenced in several copies. Knapp lot 2207 postmarked October 5, 1847 and Matthies lot 91, postmarked January 1, 1848 confirm that it was a first printing shade. We also find an example from New York a lot 857 in the Tracy Simpson sale (Siegel February 14, 1973). A pair on cover postmarked New York August 31, 1848 was offered as lot 18 in the Neinken sale (R.A. Siegel November 19,1970).
The shade also found its way to Philadelphia as evidenced by the Sweet sale lot 119 (Kelleher October 21, 1944) with a Philadelphia postmark identifying this off-cover item. The Tracy Simpson sale lot 863 shows an example postmarked August 30 1847
Blackish brown examples are recorded from a number of other cities. For example, lot 858 in the Tracy Simpson sale is postmarked at Hartford August 8 1847 so that it had to be from the first Hartford delivery of August 18, 1847. Lot 137 of the Matthies sale is an example postmarked at St. Louis on October 7, which should have been from the shipment of twelve sheets made August 7, 1847, for a new shipment was ordered October 14th.
Wunderlich sale lot 74 is a blackish brown postmarked at Reading, PA, March 4(?), 1849 used with a Philadelphia ‘R’. In the Sweet sale, lot 6 is a cover postmarked at Utica, NY in 1847-8 with the blue Utica oval. The Knapp sale had an extremely late use postmarked at Pittsburgh April 6, 1851 (lot 2212) that is during the last shipment printing period.
The above brief sample from auction records covers some 20 examples from seven cities in addition to the so-called seal brown items. The quantity is too large to warrant both Chase and Brookman’s comments about its extreme rarity.
Either the blackish brown is far less rare than previously supposed or a number of the examples are suspect, reduced examples of other shades. The listing only included one really ‘late’ use, that of Pittsburgh. However, others are known. For example there is an example from Chillicothe, OH postmarked September 2, 1850 (?) that would seem to come from the April 11, 1850 shipment to that city. We also have a ‘late impression’ dark brown from Waukegan postmarked April 24, 185- sold as lot 98 of the Hugh Baker Siegel sale of May 7, 1970. Some of these late uses may represent inclusion of first printing stamps in 1850-1851 shipments. There was carry-over stock that might not have been shipped until then.
Walnut brown—This is not a Chase recognized shade, but appears to be a warmer variant of the dark brown shade. It is quite distinctive. An off-cover pair is illustrated in the White Color Encyclopaedia permitting direct comparison between it and the black brown and seal and dark browns.
Color plate 1 shows an excellent example apparently postmarked at Alexandria, VA on July 6, with a double struck black town c.d.s. This was lot 2029 of the Grunin sale (Harmer December 15-16, 1976). The Ishikawa collection had another example postmarked at Alexandria, VA. September 20, 1848. This Ishikawa ‘walnut’ is heavily overstruck. (See figure 2, lot 41 in the Christies Ishikawa sale, a cover addressed to Messrs Lewis Fatman & Co., New York).
An on-cover example of the walnut shade is found as lot 810 in the Earnest Jacobs/Donald Malcolm sale (Siegel September 27, 1972) postmarked in blue at Hartford, CT on September 11, 1847. Hartford received six sheets of the 5-cent stamp on August 7th. Interestingly enough, the Lester Downing sale lot 792 is a horizontal pair of the dark brown untied on cover postmarked with a Hartford magenta circular date stamp on the same date, September 22, 2847! (Siegel September 20-24, 1974. This may be a misidentified walnut.)
More recently a cover to Sheffield, England, postmarked Philadelphia May 23, 1848 with a four-margined walnut example (lot 121 Pope II sale, J. A. Fox May 4, 1985) was sold. It is from the first printing shipment delivered April 5, 1848 to Philadelphia. Color plate II #14 shows this stamp.
Chocolate and deep chocolate—This shade is a variant of the dark browns and is the approximate color of bittersweet chocolate. It is a warm shade. Color plate I shows both the chocolate (#4) and deep chocolate (#5). Two examples of the chocolate are found in the White Color Encyclopaedia as an alternate reference. The off-cover example offered as lot 2017 in the Grunin sale (Harmer December 15-16, 1976) is the one found as the chocolate in color plate I. The Ishikawa sale had as lot 34 a chocolate example, which I would term a deep chocolate (See figure 3). It is described as having a very intense deep shade and partial blue Philadelphia c.d.s.
One of the problems with the chocolate shades is that while it is a scarce color, at least one sale had 14 examples. This is the Wunderlich sale (R. S. Siegel January 29, 1976). While Mr. Wunderlich is a color student and may have concentrated on this shade, the quantity appears excessive and may represent dark browns of a ‘near-chocolate’ variety.
We know that the October 2, 1847 shipment of one sheet of the 5-cent 1847 to Washington D.C. was the chocolate shade. Examples are recorded from that city postmarked October 3, October 16, and November 1.
Philadelphia received at least one sheet of the first shipment that was either chocolate or near chocolate. Lot 2040 of the Grunin sale is an example with a Philadelphia postmark that is classed as ‘near chocolate’. The major group of Philadelphia ‘chocolates’ is found in the Wunderlich sale (Siegel January 29, 1976). There we find lot 78 described as deep chocolate, postmarked August 26, 1847, while other chocolates are lots 34, postmarked August 30, 1847; lot 65, postmarked June 9, 1848; lot 53, postmarked June 8 (1848); and lot 19, postmarked in 1848.
It would seem New York also received a sheet of the chocolates in the first shipment, for we find lot 521 in the Siegel sale of April 27, 1976 described as chocolate. The Wunderlich sale lot 82 is described as a chocolate used at New York March 17 1848. It should logically be from the second shipment as this date was just prior to the third shipment.
Other examples of the chocolate shade are reported from Baltimore used April 29, 1848 (Wunderlich lot 52). This item should be from the ten sheets shipped to that city on April 4th. An untied Hartford, CT example was reported as lot 108 in the Hugh Baker sale (Siegel May 5-7, 1970) while another untied example is recorded on a Scitico, CT cover dated May 1851 (lot 527, Siegel April 21, 1976).
As noted earlier, some of the chocolate examples appear questionable, particularly in the Wunderlich sale. The Grunin ‘near chocolate’ from Philadelphia is one reason. A second reason is lot 80 in the Wunderlich sale where a cover from Syracuse, NY postmarked March 20, 1848 would have the stamp coming from the January 15, 1848 shipment to that city. Chase’s notes show that this cover was in his survey and he classed it as dark brown.
And additional reason to question some descriptions is lot 24 in the Matthies sale (Siegel May 20-21, 1969) where a chocolate shade was reported used at Norwich, NY on August 7. The description reads ‘late impression’ meaning a worn or cleaned state of the plate. We know of no other chocolates that cannot be identified with the first printing.
The deep chocolates are much scarcer than the chocolate. There were two examples reported in the Wunderlich sale. The first is lot 45 used with a red New York square grid from New York to Troy, while the second was used at Boston on August 17, 1847 (lot 49) and originated in the first shipment. An off-cover ‘deep chocolate’ with a red grid was in the Stollnitz holding and is shown in Color Plate II.