This is Installment No. 3
|Jack Nalbandian, Inc. is proud to offer visitors to our site a wonderful series of articles by renowned EFO expert John Hotchner. This EFO Guide was originally published and copyrighted by Mr. Hotchner in the pages of U.S. Stamps & Postal History Magazine beginning in the summer of 1992. The Guide is compiled in a number of separate installments, all of which will be uploaded to our site in the coming months. Visit here often as new installments appear!|
Please Note: Due to the uniqueness of the
material used for the
illustrations in this installment (and thus, unavailability of color images of them),
the black & white images from Mr. Hotchner's original printed article
in U.S. Stamps and Postal History magazine appear below
Continuing this on Jack Nalbandians website, we will finish up the color errors portion and get into overprints. Perforation related errors, and the majority of the color error categories were covered in the first two installments.
If you have read the series to date you already know that I am marching to my own drummer in categorizing errors, freaks and oddities. My opinions are just that unhindered by the practice of the established gurus of the hobby.
I renew the invitation to readers to convince me I'm wrong. Or write to me on any other aspect of EFO philately at P.O. Box 1125, Falls Church, VA 22041 or email at firstname.lastname@example.org. A stamped addressed envelope would be appreciated since 17 years of diligent practice as a philatelic writer has yet to produce much more than satisfaction. (That's not a bad thing, by the way. I'm with Salvador Dali, who is quoted as saying, "There are some days when I think I will die from an overdose of satisfaction.")
12. REVERSED COLORS. The 1968 Christmas stamp, shown in Figure 1, was printed using two yellows, light and dark, applied by lithography. According to contemporary researcher Jim Sanders, the light and dark yellows were first interchanged in error, and then allowed to be mixed as a matter of practicality. A case can be made that this shouldn't be called an error because the variations from the intended normal outnumber the normal by a factor of six to one. Still, they are major variations from what was intended, even if numerous and hard to see.
As planned, the light yellow was intended for the angel's face, hands, wings, parts of the robe, and the plate number. The dark yellow was intended to be used for the trim of the robe. According to Sanders, there were three variations produced:
A. A light yellow plate number with wings, face, hands, and parts of the robe in matching color. This is the stamp as it was intended to be produced.
B. A dark yellow plate number with the same yellow appearing throughout the the stamp-
C. A mixed yellow plate number in a shade somewhere between light and dark yellow. How did this happen? As printed, the light yellow looks like a fine imported swiss cheese, and the dark yellow looks like prepared mustard. However, to the printer working with the ink, the lighter of the two yellows appeared to be the darker when viewed as we think standing in pails on the pressroom floor. The inevitable occurred: the light yellow was added to the fountain Figure 6 intended for dark yellow, and vice versa. As inks in the well mixed with inks being added, chaos reigned.
Once you have a dozen or more examples of the stamp for comparison, the variations are easy to see; especially under magnification. The range of possible colors is even easier to see in the yellow plate numbers.
13. STAMP PRINTED ON BOTH SIDES. Shown in Figure 2 is an example of an early U.S. stamp with a positive print on the reverse. Such varieties are known on issues of 1851 to 1883, and then again on both types of the 3c offset printed Washingtons of 1918. The earlier examples are extremely rare. Remember when watching for these that the print must be a positive impression; not the often seen reversed impression characteristic of an offset when still wet ink from a sheet below was picked up in reverse by the back of a sheet laid on top of it.
Cause? Other than the unlikely possibility of a pressman not noticing that a sheet had already been printed, it is most probable that the backprint represents a first, incomplete, entry into the press or an insufficient print that was not good enough to be passed on to the next station. This is not farfetched given the contemporary concern with expense of paper and accountability, and a desire to prevent waste.
14. ONE OR MORE COLORS LEADINGTOTHECOMPLETION OF A STAMP FULLY OFFSET ON THE REVERSE OF THE STAMP BY THE BLANKET ROLLER. Figure 3 shows a copy of the 1959 St. Lawrence Seaway issue that appears to be printed backwards. It is a nice smooth, even impression, on the back of a normally printed stamp.
This sort of error needs to be differentiated from the partial offset from wet ink, and the "stuck-stamps" offset that is too frequently offered as an error by those who don't know better, or know better but prey on the gullible. Wet-ink offsets occur as gummed sheets are ejected from the press and piled upon one another. Since the ink goes through a drying process on press, very little of it is wet enough to transfer, but one or more colors may transfer partially. This does not qualify as an error.
All one needs to create a "stuck-stamp" offset is two mint panes and a humid atmosphere. Put the panes together under pressure, and the result is stamps stuck to each other.
If lightly enough stuck, such stamps can be peeled apart. The result will be that the back of the pane on top will have picked up some (maybe even most) of the design of the pane below. It will be characterized by an inconsistent, rough to the touch, impression. And, most often, it will be on the wrong type of product for a genuine error offset.
The error is associated with stamps printed by the Bureau of Engraving and Printing in Washington, D.C. between 1957 and 1980 on the sheet-fed Giori-type presses, on pregummed paper. Sheets from a stack were drawn into the press one at a time. If that mechanism failed to draw a sheet into the press, the inked plate would apply the design on the blanket roller; the function of which was to maintain a consistent pressure from behind the sheet so that it would pick up a consistent impression from the plate.
Almost needless to say, the next sheet through the press got not only the proper print on the front, but a nice even full print from the blanket roller on the back. A few succeeding sheets would also get some offsetting until the the roller was dry.
Other variations on this theme can occur because a sheet going into the press is torn or folded over. The result is a backprint that looks like the one in Figure 4. Also possible is a partial design print due to the need to print the stamp in two passes, with only one pass affected, such as the Leif Eriksson stamp shown in Figure 5. The back shows only the engraved dark brown statue and lettering.
15. DOUBLE PRINTS Figure 6 illustrates a true double print; that is to say, one in which the design you see is the result of two insertions into the press. It's a fairly scarce phenomenon compared to the more often seen "kiss print" which is actually the result of slippage between the plate and the paper during a single insertion. The latter, and the partial doubling of design elements and plate numbers called "tagging ghosts", will be discussed in a future installment of this series dealing with Freaks.
Second insertions resulted from the same considerations that caused a careless Bureau of Engraving and Printing pressman to create stamps printed on both sides: the first print was inadequate in some manner-usually too light, or incomplete. Most often, there is a significant displacement between the two impressions.
Among the most spectacular examples is the one shown in Figure 7, courtesy of Clyde Jennings. It is a strip of 10 of Scott No. 499, the 2c 1917 Washington head, perf. 11. It is likely that the pressman did not fully insert the sheet, from which this strip came, into the press. The resulting print was incomplete. In an effort to fix the mistake, the pressman tried to reinsert the sheet to finish it off. He did a pretty good job of lining it up with the horizontal rows, but didn't do so well on the vertical rows. The result is a second impression that overlaps and doubles the center of the stamp.
Next, we will be looking at overprints; an area that includes precancels. By way of introduction, I want to make it clear that the only precancels included here as legitimate errors are those prepared by the Bureau of Engraving and Printing (BEP) under the supervision of the USPS. The illustrations that accompany this article will show examples of the neat, regular, standard types that are characteristic of BEP production. There are an almost infinite number of non-Bureau (or "local") precancels on which errors of all sorts are common due to inattention to quality control. They are avidly collected by precancel specialists but they do not have the pedigree or (in general) value that a Bureau of Engrav ing and Printing produced error has.
16. INVERTED OVERPRINTS/PRECANCELS. There are only two invert type in this category, and one of these breaks rule stated above: the four inverts of the th experimental precancel used on the 196 "Winter Sunday in Norway, Maine" Christmas stamp. Precancels were added locally in four cities: New Haven, CT, Atlanta, GA Baltimore, NM, and Memphis, TN. An invert example is shown in Figure 8. I think it belongs here because the Post Office Department supplied the electrotype plates for local use and the entire test was tightly controlled by Washington.
The second type in this category is a Bureau precancel; three of them actually. Because of the horizontal format of the stamps, the invert is easily missed. Shown in Figure 9 is a pair of 20c stamps from the fourth Bureau issue. Note that the precancels read up and down. Quick, now, which is the invert? If you said UP you are correct. The normal reads down. However, on this issue, there are three instances of precancels reading UP for part of the run: Des Moines, Iowa, Detroit, Mich., and Dallas, Tex. Contemporary researcher Hugh Southgate suggested that the stamp plates were reversed when the stamps were printed. This was certainly possible. In support, he makes the point that other high value precancels of the set prepared at the same time should be known inverted if the precancel plate were reversed. They aren't.
With the exception of precancels, overprints are relatively unusual on U.S. stamps. Either the federal Government is configured, or the Postal Service has planned well, to avoid the circumstances that caused other countries to issue masses of overprints and surcharges: stockpiles of old (replaced) stamps available for use, precipitous changes in government and country name, and rapid changes in common rates or the name of the currency. In fact the U.S. has evidenced a preference for issuing a new design to avoid "amending" stamps by cranking out a quick overprint. Thus, there are few U.S. stamps that have been overprinted: the Shanghai overprints of 1919 and 1922, the Hawaii and Molly Pitcher overprints of 1928, the Kansas-Nebraska overprints of 1929, and the 1959 Postage Dues on which the values were added to a basic engraved frame using a precancel-like process.
17. WIDE SPACED OVERPRINTS/ PRECANCELS Wide spacing is illustrated in Figures I 0 through 12. The first is found on the Kansas I c, I 1/2c, and 2c, the Nebraska I c through 4c and the gc and 9c, the Molly Pitcher 2c, and the Hawaii 2c. Normal vertical spacing on the one line overprints was .22mm. However, the overprint plate was set up to match the normal printing plate; that is to say, it had one vertical gap of 32mm to take account of the interpane gutter between the top and bottom panes. If the overprint were vertically misplaced by the length of the the gutter, or more, the 32mm gap would print within a pane,causinga4Ostamp row (across the two horizontal panes) of wide spaced overprint vertical pairs.
The same phenomenon is responsible for the wide vertical spacings that exist on several Bureau precancels (an example is shown in Figure 11) and the 2c, 4c, 6c, 7c, 8c, and 10c of the 1959 Postage Dues series of stamps. Curiously, the wide spacings on the 1922 series definitives receive Scott listing as errors, while the Postage Due wide spacings do not.
18. MISSING OVERPRINTS/PRECANCELS Even though this is a more significant class of error than the wide spacings, I've listed this below them because an understanding of the cause of the wide spacings is important to knowing how the missing Kansas-Nebraska overprints occurred. These are collected in vertical pairs, one with and one without the overprint. They are known on the Kansas I c, 3c, 4c, and 7c, and the Nebraska I c, 3c and 9c.
Two possible problems cause the phenomenon. The first is a simple shift of the overprint down into the bottom pane which could leave a row or more at the top of the top pane without the overprints. If the overprint present on the top pane were somewhat on target, an inspector could well not notice that there was a row without the overprint and pass the pane. I understand that this error exists, but I have been unable to find an example to show.
The second cause affects a row in the interior of a sheet, If an overprint shift left the overprints high on the bottom rows of stamps of the bottom pane, the 32mm intrasheet gutter would fall so as to leave a horizontal row of stamps without an overprint.
Missing values on the postage dues might have been caused by the first of the problems noted above (simple shifts), but not by the second. The overprinted values are simply too tall for any variation to be created that has the value in the proper location on one stamp, and the value missing on the stamp above or below, as is shown by the two examples in Figures 13 and 14. Scott notes the existence of such pairs on the 2c, 4c, 6c, 7c, and 8c. Datz adds the 1c, 3c, 5c, and 10c. Datz does not list the 2c, 4c, or 7c. There is another possible cause, most easily seen in the transition strip shown in Figure 15. The ink ran out, or more likely, the equipment governing the meeting of the plate and the paper was in transition to or from normal printing. The transition could go from fully present on one stamp to fully absent on an adjoining stamp.
Missing precancellations are rare. Again, they must be shown attached to a normal example in order to know what is missing. There are two causes; failure of the precanceling plate to impress the paper, as described above with regard to the 1959 postage dues, and the failure of the precancel plate to receive adequate ink. I know of one example, as claimed in the precancel literature, of each cause.
One is a pane of 3c Liberty stamps on which the top four rows of the pane show no precancel, while the bottom six rows show a properly placed Chicago, ILL precancel. See Figure 16. The ink deprivation example is on a four by three subject block of the 50c Liberty on which the top row shows a normal impression of Chicago, ILL. The middle row shows an irregular partial impression. The bottom shows no impression whatsoever.
19. SE-TENANT PRECANCELS So far as I am aware, this error is known on only one pane of U.S.stamps: the 1 1/2c Liberty (Mount Vernon), Plate No.25366LL, on which 90 of the stamps are Bureau printed Duluth, Minn. (Figure 17) and the other 1O Green Bay, Wis. George Brett explained the cause as follows: "The two-city precancelling runs are performed (on the Huck press) by alternating the overprinting plates on each side. That is, the upper left and lower right panes will be one city, and the upper right and lower left panes will be another. The alternating plate arrangement makes it important that careful synchronization be maintained on the presses as otherwise you may get two cities on the same sheet." A matching sheet with 90 Green Bays and 10 Duluths would have been created at the same time, but it was likely caught and destroyed during quality checks since there is no report of its being found.
20. MISTAKES IN THE OVERPRINT. There are several categories here. Let's start off with perhaps the best known, the inversion of OHIO (so that it reads OIHO) in the Cincinnati, Ohio precancel typeset plate used to precancel Liberty issue stamps. It is found on subjects 26 through 30 on the 1/2c, 1c, 2c, 3c, and 4c. See Figure 18. Another inversion, though really a transposition of the city and state name, is the CALIF above VENTURA error (Figure 19) that appears on the 1c Presidential. The precancel plates in this era were 50 subjects repeated twice, so the error on the 23rd subject of the plate was repeated on subject 73. There are three instances of typeset misspellings found in city names on fourth Bureau Issue stamps. An example is shown in Figure 20. The 1 1/2c and 8c sheet stamps are known with MILWAUKEE spelled MIKWAUKEE on positions 46 and 96. The 1c coils are known with ALHOL instead of ATHOL (Mass) and SPPINGFIELD rather than SPRINGFIELD (111). Both are position 30 on 30 subject electrotyped plates, so can be found to the left on gap pairs. Missing elements in overprints run from the minor to quite major. In the former category are the missing periods after state names on many precancels (prior to the USPS going to the two letter state abbreviation), and the Scott-listed missing period on the 1c NEBR. issue.
Much more serious are two errors listed on the 4c on 2c stamp (Type VII offset) of the second Shanghai overprint (Scott No. K]8). In the first, the word Shanghai is omitted. On the second, the 4c is also omitted (Figure 2 1).
21. DOUBLED OVERPRINTS There are two known doubles. Scott lists a doubled overprint for the $2 on $1 first Shanghai overprint (Scott No. K16). See Figure 22. It was discovered by the Chinese head clerk at the postal agency a few days before the office was to be closed. When he reported it, the Postal Agent, who didn't want to be burdened with extra paperwork, ordered him to sell them. The clerk did as instructed; he sold the stamps-to himself. A few went over the counter to customers, but the bulk he paid for. He sent several to himself to validate their postal validity, and the rest were kept for his children.
The second instance of a doubled overprint is a Minneapolis,Minn., precancel on the first bicolor 8c Liberty (Scott No. 1041). Both these stamps were produced on sheet fed presses rather than web fed, and the cause of the doubling is a double insertion. It is likely that only one sheet of 100 was created for each.
This completes this segment. In the next installment we will cover the last of the errors and introduce "Freaks". If you have EFO material and would like to know more about it, or you're an EFO collector going it alone, check out the EFO Collectors' Club for the many ways it provides to learn about EFOs, and the systems it has to help all members trade/sell EFOS. Write for information to CWO James McDevitt, USCG (Ret.), 1903 Village Rd. West, Norwood, MA 02062-2516.