Even The Experts Can—And Do—Change Their Minds
© Calvet M. Hahn 1982
Reconsidering a previous opinion is not uncommon for an expert committee, or even an individual expert. The usual reason is that new information has turned up, 
but other elements can also prompt a reconsideration. In the past few years, a particularly interesting reconsideration involved the cover illustrated here. This cover had received one, if not two, ‘genuine in all respects’ certificates from one of the better-known expert committees.
The good opinion reads:
“5¢ red brown with vertical tear at top…genuinely used on cover with greenish blue ‘SHIP’ and MOBILE, Ala./2/FEB 5’ cancellation.”
The problem with this certificate at the time was the description of the color. Some believed that the color on the cancel was green, not greenish blue, and the experts were asked to reconsider the item on this basis. To aid the reconsideration, additional research was done on the color, and a color study was submitted along with the item.
The result was surprising. The opinion changed dramatically, and a new certificate was issued that reaffirmed the ‘greenish blue’ postmark, but added that both the handstamp circluar date stamp (c.d.s.) of Mobile and the tying ‘SHIP’ were ‘fraudulently added to an otherwise genuine cover.”
Because expert committees do not give detailed analyses of their reasoning, I was asked to ‘reconsider the reconsideration’ and confirm or deny the new opinion with a detailed write-up. It is the process of this reexamination that is offered here as an aid to understanding the procedure of evaluating a cover.
The Cover Itself
The cover in question is a darkish brown envelope addressed to Samuel F. Jarvis, Jr., Middletown, Connecticut. In similarly shaded handwriting but without the flowing characteristic of the address, is a handwritten ‘Paid’ at upper left. In a still different hand in docketiing at the left there reads, ‘John R. Livingston/Rec’d April 22d 1851.’ On the back of the cover is a large wax seal in red, positioned behind the address and under the lower portion of the c.d.s.,
The cover bears a red brown 5¢ 1847 stamp that is faded or bleached, and which has a tear extending from above the second ‘F’ of OFFICE to below the ‘N’ of CENTS—so that the stamp is, in effect, torn in two, and not just with a vertical tear at top.
There are three vertical pen strokes ‘cancelling’ the stamp. The one at the left is in a darker shade of ink than the other two, and the one at the right runs almost in the groove of the tear in the stamp.
The stamp does not ‘pop’ at the corners, rather it is relatively firmly affixed, and pressed into the cover. The lower right corner, however, is slightly loose, and looking under the stamp there, one can see that the underlying color of the envelope is a lighter shade, as would be expected if the stamp had been on for many years.
In the center of the cover is a 33mm c.d.s. reading MOBILE, Ala/2/FEB/5. This cancellation has a break in the outer rim between the ‘L’ and ‘E’ of MOBILE and again at the bottom, from about the five o’clock to the eight o’clock posiition. This bottom break is directly over the red seal on the back.
The stamp is also tied to the cover by a handstamp 12 x 5mm SHIP in the same shade as the c.d.s.
There are two basic problems associated with this cover: first, the color of the cancel, and second, the authenticity of the cover. Are the cancels fraudulent, as alleged by the expert committee; are they genuine; or is the entire cover bogus? These two problems will be examined separately, as both offer interesting insights for both collectors and cealers to consider.
The Color Question
The expert committee involved is consistent in its designation of the cancellation colors as ‘greenish blue.’ It is important to realize why the color is significant. Green is not a particularly rare color in classic philately, but it has carried a substantial price premium. This is the consequence of several wealthy collectors having formed collections of green cancels in the past, and having paid premium prices for such items. (One such collector enamored of green cancels was Marc Haas.)
Because of the lack of interest by wealthy collectors, several rarer colors have not commanded the price premium that would be warranted by their scarcity. Even the common colors of red and blue can be rare from certain towns.
For example, blue cancels from New York City are very rare in the pre-1856 period. They were used during the Revolution and again, very briefly, in the 1840s. Based upon rarity and condition, a blue New York circle of the 1840s should sell for hundreds of dollars, rather than the few dollars it commands.
In the case of the Mobile c.d.s. no other green cancel is known from the town on an 1847 cover, so the color would be worth a substantial premium both for a town collector and for an 1847 collector. Despite its appearance and the damaged stamp, a green cancel would make this cover an exhibit piece.
There is a problem, however. It is well known that blue cancels on brown tinted envelopes frequently look green. Further, some of the inks used in the period give off-shades of green that are not bright and brilliant. These genuine green cancels don’t have the same appeal than an apple-green strike would have.
Mobile has not been recorded with a green cancel in the 1850s. The American Stampless Cover Catalog notes that this ‘Mobile 5’ style is known in black and blue in the 1850-1852 period, but does not give precise dates. (This is still true as of the 1985 Fourth Edition). Earlier editions of the catalog show only the 33mm blue c.d.s. in this period; it is the addition of the black marking that makes the latest catalog different from earlier versions.
One of the leading students of Mobile was kind enough to loan examples (seen below) from his collection for use in dating this cover. Because of this, it is possible to date the blue c.d.s. from 1850 through January 16, 1851—several weeks prior to the subject cover’s c.d.s.—and on until at least February 12, 1851.
The black c.d.s. is introduced first in 1852, by April 13, 1843, and it is known through December 25th. Thus we know that blue was the standard Mobile color at the time this cover was sent.
Key Question
The question of color becames quite important, because we know that Mobile used blue, but have no record of green cancels, in the time period under study.
Further, because of the problems associated with a blue cancel on a brownish cover, it is possible this is not green, but a blue that has changed. And, since green commands a substantial premium in the marketplace, the question is not academic, but one involving a fair sum of money.
Examined under natural daylight, the cancel does appear to be green. Nonetheless, this test is insufficient in the circumstances described here. The expertizing committee apparently compromised by calling the cancel ‘greenish blue.. However, based upon the same evidence available to them—which included a special color study prepared for this cover—I would conclude that green is an appropriate designation.
The special color study was based upon reflective spectrophotometric analysis. This cover and three other examples generally accepted as green were compared. One was a Scott #65 with a green cloverleaf, cancel, the second was a #11 with a green cancel, and the third a #11, position 21L3, with a green cancel. No stampless cover was used. The basis for comparison was,
“the color of the cancel on a clear portion of the uninked stamp paper’
That means the color of the envelope was eliminated except as it reflected through the stamp itself. It should be noted that because the stamp paper was aged and the ink involved was either a blue or green pigment, a shift toward the yellow end of the scale should result. Consequently, it was not surprising to find the cancel described as ‘grayish yellow-green.’ To the naked eye, the color was found to be very close to ISCC-NBS-gy YG chip #122. This is a color chip in the National Bureau of Standards color book.
To get a ‘true’ color, the chromaticities of the four items were corrected to eliminate the distortion of the type of paper. When this was done, the Mobile cancel had a somewhat lower designation on both the x and y coordinates of the graph than any of the other items. Corrected, the designation for this cancel would be bluish green, similar to the #11 comparison stamp rather than the yellowish green of the #65 cloverleaf and the #11 position 21L3 stamp with a green cancel.
This study was sufficient to convince me that the appropriate designation of the cancel should fall in the green rather than the blue family. An appropriate designation would be ‘bluish green, rather than ‘greenish blue’ as noted on the latest expert committtee certificate.
The Question of Authenticity
There are a number of items connected with this cover that would lead experts to at least question its authenticity. In addition to the fact that the cancel is now found to be in a color otherwise unknown from this town, we have the following negative elements:
1) The use of both a handstamp and pen cancel.
2) The tear and bleachy condition of the stamp. (Fakers like to use damaged or cleaned stamps as they are cheaper to buy as source material.)
3) The lack of ‘pop.’
4)The difference in shade of the pen strokes, and the fact that one practically ‘conceals’ the tear trhoughout the stamp.
While some experts might condemn a cover on these points alone, I would not consider them definitive. There are really only negative signals which might be overcome by a strong positive signal from another part of the analysis. Rather, I should like to draw attention to three other points that can weight quite strongly. These are:
1) The time it took to deliver the cover.
2) The problems of rating both as to distance and the designated use.
3) The plating of the postmark.
The problem of dates: The only year date we have for this cover is the docketing notation that it arrived in Connecticut on April 22, 1851. This is two and a half months after the Mobile postmark of February 2nd, for a trip that should have taken about a week in 1851. No explanation for the delay (that I can construct) is satisfactory. When there is a delay of this nature, there is normally some marking to explain it on the cover.
If Mr. Jarvis was not at Middletown when the cover arrived, it would have been carrier delivered in a few days, and he would not have docketed it received in April. Had no one been home, the cover would have been advertised, and so marked. There is no indication that the cover was missent or delayed en route, or that it lay for two months at Mobile. Some form of marking would almost inevitably have been used for any of these circumstances. Rather, what we have is a very suspicious time gap.
The problem of rates: In the spring of 1851, the standard rate was 5¢ for under 300 miles, and 10¢ for a longer distance. Mobile was definitely over 300 miles from Connecticut, so that the basic postage rate for a single letter was ten cents, not five cents.
The only way to construct a 10¢ basic rate is to assume that the Mobile c.d.s. represented a 5¢ payment to which is added the 5¢ 1847 stamp. This is circuitous reasoning that requires strong documentation in light of the fact that normal practice in such cases of added rates was to specifically note the addition on the face of the cover. Numerous examples of such additions are known. It is also necessary to assume that the stamp was applied first, and the rest of the rate paid in cash. While such covers do exist, they are extremenly rare, and only a handful have been recorded.
Suspicions
It might be noted that Mobile had a companion handstamp to the ‘5’, with a blue 33 mm circle MOBILE, Ala./day/mo./10, which could have been used to postmark this cover, and at least indicate some awareness that the rate was 10¢. The fact that this othere c.d.s. was not used, and no indication is given that Mobile received 10¢ other than the c.d.s. and stamp, is quite suspicious.
More importantly, the cash-plus-stamp explanation falls down on the basis of the straightline SHIP cancel that ties the stamp. This marking is used on incoming ship letters of the period. To accept it as genuine would mean that this cover did not originate at New Orleans and come to Mobile by steamboat, (for the marking would be different in that case) but rather that it originated at a foreign port, or one that was not linked by steamboat to Mobile. The rate required for such a use is an extra 2¢ for the ship fee, so that the total postage to Connecticut would have to be 12¢ rather than either 10¢ or 5¢. Only a very strained construction of the manuscript ‘paid’ in the upper left of the cover can explain away this 12¢ problem.
The relevant sections of the PL&Rs applicable in the spring of 1852 read as follows:
“500. Any letter packet with one or more stamps affixed, equal in amount to the postage properly chargeable thereon, may be mailed and forwarded from any post office as a prepaid letter or packet; but if the stamps affixed be not adequate to the proper postage, the postmaster receiving the letter or packet for transmission, will rate it with the amount deficient in addition.
- Stamps so affixed are to be immedately cancelled in the office in which the letter or packet may be deposited, with an instrument to be furnished to certain of the post offices for that purpose. In post offices not so furnished, the stamps must be cancelled by making a cross (X) on each with a pen. If the cancelling has been omitted on the mailing of the letter, the postmaster delivering it will cancel the stamp in the manner directed, and immediately report the postmaster who may have been delinquent to the Department.
- All ship letters and packets are to be charged with a postage of six cents each, when delivered from the office at which they are first received; when forwarded in the mail to other offices, with two cents, in addition to the ordinary rates of postage. They should all be marked SHIP, at the time of receiving them. This applies to all letters and packets brought by vessels from foreign countries, as well as those conveyed from one port to another in the Unted States over routes not declared post roads.
- …Thus the Mississippi River, from New Orleans to the mouth, is a post road; yet letters carried by ship between New Orleans and any other port in the United States, are subject to the usual ship letter postage…
- …Letters brought by steamboats should be marked STEAMBOAT, at the time of receiving them.”
Extra Step Needed: Expertizing Postmarks
The points mentioned thus far are sufficient to condemn this cover as fraudulent. Nevertheless, because of the appearance of the c.d.s. and its bite into the paper, as well as the tie of the SHIP through the stamp, it was felt desirable to go an extra step and check out the handstamp c.d.s. As noted earlier comparison Mobile/5 covers were obtained for analysis.
Plating the postmark: Before detailing the handstamp comparisons, several points of suspicion about the Mobile c.d.s. on the 5¢ cover should be noted. First, there are two small marks in ‘green’ above the c.d.s., about 2mm above the top, and located at a 90° tangent to a line drawn through the diameter of the c.d.s. at the date. These are not typical markings found on c.d.s. items, but are known on certain types of forging devices, where one has to ‘position’ the device to strike a fake c.d.s.
The second point is the fact that although there is a raised seal on the back under the lower part of the c.d.s., the c.d.s. strike is missing over the seal rather than being heavier struck because the paper was higher. Normally, the strike should hit the higher portion and then show a gap where there was no back pressure until it hits the paper at the top. thus, the logical gap should be at about the four and eight o’clock positions, rather than at the bottom where logically the strike should be strong. These are both suspicious points.
A comparison of genuine stampless covers with the Mobile/5 postmark shows that all have a slanting number ‘5’ at the bottom, rather than the upright style of ‘5’ found on the ‘green Mobile’ cover. The comparison of letter placement on the January 16, 1851 cover, seen earlier as figure 3, and a blue example of January 8, 1852, shows the strikes are identical. However, both differ from the ‘green Mobile’ February 2, 1851 strike on the cover under analysis.
The suspect cover has serifs that are a bit too long, the small ‘a’ of ‘Ala’ is of a different shape, as though a clogged ‘a’ might have been used as a model. The letter position is also slightly different, and the base of such letters as ‘L’ and ‘E’ are too thick. The ‘2’ in the date of the suspect cover also differed from those of the genuine stampless items. In the earlier shown figure 4 February 23, 1852 cover, the ‘2’ is rounder at the top than the date slug on the suspect cover. 
It also has less of a round ball on the front of the ‘2’. Several other ‘2’s’ were compared, for example the image to the right, and while there are some differences, none had the characteristics of the suspect cover’s date stamp.
When these date stamp differences are added to the fact that the genuine handstamps all show the same ‘5’ ratemark, and that both items differ from the markings on the suspect cover, it seems clear that the 5¢ green Mobile 1847 cover has unique postmarks that must be rejected as fraudulent. They were probably made from a poor strike model that was photographed and then turned into a rubber handstamp rather than the metal stamp used for the genuine strikes. The bite into the paper of all the genuine items is similar one to the other, but definitely different from the bite found in the green Mobile c.d.s.—a difference that appears to represent the difference between a metal handstamp, in the case of the genuine, and hard rubber in the case of the fake.
Genuine Correspondence
A final, but relatively unimportant question, is whether the cover is completely bogus. In this regard, it might be noted that the Jarvis correspondence is a recorded one, and that a number of the letters in it come from the lower Hudson valley in New York, which was the home of the Livingston clan. Thus, it seems quite likely that a genuine piece of correspondence was used.
There are gum soaks in the envelope around the 1847 stamp, that show up when the cover is examined from the inside with light from outside shining through it (a procedure that should be standard for examining covers, particularly suspect ones). These are often found when stamps are soaked off a cover, and it may be that originally this cover had another 1847 stamp on it which was removed. However, there are no indications of any postmark, so such a stamp would have had to be a pen cancelled use without c.d.s. This is not common, but can be found.
However, the Livingston family is known to have used ‘courtesy’ covers, carried outside the mail, to a large extent, and this might just as logically represent such a use with a forged ‘Paid’ added in pen. The shaping of the letters of the manuscript ‘Paid’ do seem to differ from the letters found in the address, supporting such a thesis. The question, however, is not worth pursuing, because of the forged nature of the Mobile c.d.s. and SHIP handstamps.
I would conclude that the expertizing bodies involved in examining this item were quite correct to withdraw the initial ‘genuine’ opinion and let the second opinion read that both the stamp and cancellations were fraudulently added. They probably erred in not labeling the forged postmarks green or bluish green, but this is a very minor point. What is most surprising is that the item should originally have gotten a ‘genuine’ certificate in light of the obvious signs calling for a very careful examination. It might be noted that the new Thomas Alexander Census work on 1847s notes the two opinions.