Using forensic testing of ink on a postal history item when historical expertizing methods fail. -Mel Kravitz R.P.S.L., Harry Hagendorf C.S.C.

Abstract: Collecting postal history items which are apparently unique or unknown to current examiners, can often present special problems when trying to determine authenticity. The traditional approach depends upon expert opinions, which in turn is supported by a strong reference library, historical auction catalogs, provenance, and several experts reviewing the postal marks, notations, stamps used on the item, and so on. This paper is about one such item which has no provenance in the current published reference publications1,2,3or auction catalog base, as well as being the only example todate of a double type I 'Customs removed date stamp' c.r.d.s on cover. It also uses a Red Revenue stamp for postage, making this a thus highly desirable postal history item. Indeed the C.I.M.4 corner card Jesus Is Coming “Watch and Pray” adds additional interest to the collector, and this alone is rare on cover5 for the postal historian.

Introduction: In February of 1897 the Chinese Imperial Post Office, I.P.O.6 replaced the the Customs Post Office7 which had first issued stamps for China in1878. Postal marks previously used transitioned away from the Customs Date Stamp or c.d.s, the Chinese Seal cancel for town Post Office identification, and stamp canceler, and were eventually dropped8. The Pa-Kau or bar obliterator was instituted and adopted to cancel stamps. This canceler had no date or year but was intended to prevent reuse9 of postage stamps. The town was coded into the bar cancels. As early as 1896, Sir Robert Hart had been working on producing a standardized datestamp for the Imperial Post incorporating both the Lunar and Gregorian dates. These “Dollar Daters” were used in 27 cities, including Tientsin, for which the earliest date of use was July 4, 1897. For some unknown reason, the Post office in Tientsin also adopted the use of the 'Customs removed date stamp', or c.r.d.s., a circular cancel in two sizes10 the typeI and type II with the type I in equivalent size and similar to the previous c.d.s used in the Customs P.O.period but missing the 'customs' from the cancel. The Tientsin post office used this type I, c.r.d.s. to identify the post office in English as well as the cancel year and date in mid to late 1898 on the cover, when the Pa-Kau canceler was used on the stamp. China was not a member of the U.P.U.9 and foreign mail to be forwarded had to be bilingual. With the I.P.O. the currency changed from Candarins to Cents and a new issue of stamps, the I.C.P. Issue11 was authorized. These stamps arrived late and were not available at the start date of this new service. To meet the post office demand for stamps, surcharged Red Revenue, Small Dragon and Dowager stamps were issued. These were used even after the I.C.P.11 issue became available in Oct. 1897.

Single c.r.d.s. stamp cancels are also known to exist from Tientsin, on both Red Revenue stamps as well as the surcharged Dowager issue stamps but not a double c.r.d.s. on stamp and cover.13 One typical Tientsin P.O. domestic cancel combination in mid 1898 was the Pa-Kau for stamps and the c.r.d.s on cover for 'Tientsin P.O. origin and year/date'. A domestic cover with this combination of cancels used from Tientsin to Shanghai is shown below3 in Fig. 1.

The test cover is shown in Fig. 2. Both covers were sent to the same address in Shanghai in Sept 1898 thus it will be possible to test the ink from the Tientsin P.O. as well as the arrival Shanghai delivery cancel ink.. The reverse of both covers have a Shanghai Local Post12 cancel, a double ring circular cancel, with month, date and year abbreviated.

-1-

Analysis13 : The basis of analysis must answer the following questions for the ink on the test cover:1) Was the ink used on the c.r.d.s cancel on cover and on the stamp the same ink?2) Was the address ink different than the cover and stamp c.r.d.s. ink? 3) Were different devices used to cancel the test cover c.r.d.s marks?4) Was the ink used on the Sept. 20, known good used on the c.r.d.s. a match to that used on the

test cover c.r.d.s.?5) Does the obverse Shanghai Local Post cancel ink on both covers match?

The VSC6000 spectral comparator14 is well suited to probe the above questions and along with the Bruker-XRF15 element analyzer, both can help provide answers to the above questions.

Forensic Test Method: To help answer the questions, a test plan16 was prepared and submitted to the National Postal Museum, N.P.M., in Washington D.C. Stamp forensic lab which contains both machines. The VSC6000 was used for ink reflectance spectral data signature collection from ultraviolet to infrared wave lengths of 400nm to 1000nm. The machine is calibrated by using a white reflective test tile (spectral reflective response is shown on Fig. 9). This gives a high amplitude with near constant reflectance through the wavelength range. The items were tested by Harry and Kathy Hagendorf with technical machine assistance by Susan Smith, and phone consultation with Mel Kravitz.

It should be noted that the N.P.M. does not expertize items, but provides technical machine assistance only, conclusions reached are solely the inferences of the authors.

The Bruker Tracer III-SD handheld X-ray fluorescence analyzer at the N.P.M. was also used to gain insight into the envelope paper element componets16 spectrum.

Both the test cover and the known good cover were also examined at the Philatelic Foundation in New York, P. F., using their VSC6000 machine16 to examine the overlay of the c.r.d.s. cancels for cancel size,font and numeral identity to answer the question on device canceler (question #3). The Red Revenue stamp c.r.d.s. cancel was magnified, and using the infrared capability of the VSC6000 this cancel on the test cover was made visible, and was examined. The VSC6000 was operated by Larry Lyons with technical assistance by Harry Hagendorf and with Mel Kravitz on phone consultation.

-2-

Fig.1, Double weight cover, Sept. 20, 1898 from Tientsin to Shanghai, the Pa-Kau cancel for 2c. Dowager surcharged stamp pair, cover canceled with type I customs removed date stamp. This cover will be used as a 'known good' ink sample of the black Tientsin P.O. cancel ink in comparison to the inkon the test cover. Sept. 20 was two weeks from the cancel date on the test cover.

-3-

Fig. 2, ‘Test Cover’ single weight Red Revenue stamp and cover front double customs removed date stamp ( c.r.d.s.) canceled Sept. 5, 1898, from Tientsin to Shanghai.

Test Points: The test cover front with ink test points for the VSC6000 is shown below in Fig. 3, it should be noted that the VSC6000 is being used as a comparator, and as such, ink intensity, that is light versus dark, will give different intensity amplitudes but similar spectral signatures. There will be an amplitude offset in light vs dark ink. For perfect spectral amplitude % intensity vs spectrum wavelengthcurve matches, the same ink darkness points need to be selected.

-4-

Fig. 3, Test cover, selected ink test points, background points in the envelope vicinity (Pt. 4 and Pt. 5) of the c.r.d.s were selected as well as points of light and dark ink, ( Pt. 2 and Pt. 3) to test the cancel inkand envelope background influence on reflectance response.

-5-

Test Results: The VSC6000 corresponding intensity vs spectral plot is shown below in Fig. 4, and Fig 5.

Fig. 4, VSC6000, test cover ink and no ink % of intensity vs spectral wavelength signatures.

Discussion of Results: Pt.1 and Pt.2 are respective dark ink points on the cover c.r.d.s and stamp c.r.d.sthe % intensity and spectral signature overlaps and is a perfect match. This satisfies question 1. The same ink was used to cancel the cover and stamp.Pt. 3 is lighter ink used on the cover c.r.d.s at 'P' of SEP, the spectral intensity vs wavelength curve shifts up but is the same spectral signature.Pt. 4 and Pt. 5 curves represent the envelope free of ink in the vicinity of both c.r.d.s cancels. This indicates that we are, indeed, measuring the reflective cancel ink for each c.r.d.s. mark in Pt. 1 and Pt. 2.These spectral signature data on the cover is a measurement of the composition of the envelope paper, we needed to gain insight in the effect of this on the ‘ink’ sitting on the envelope. When picking test inktest points care must be taken to avoid ink on the stamp, as well as the corner card ink or address ink, since the results will be skewed and not reliable. The stamp color will change the spectral signature, as will the other foreign inks. Since the VSC6000 is being used as a comparator it is important to compare similar ink intensity as well.

-6-

VSC6000 comparative spectral reflectance analysis of address and c.r.d.s ink on the Red Revenue cover is shown below.

Fig. 5, VSC6000 address ink Pt 6 in comparison to the stamp ink on the outer ring c.r.d.s Pt. 7 on the test cover.

Discussion of Results: The address ink is a different ink, the response of the c.r.d.s Pt. 7 is the recognizable reflectance signature of the cover cancel ink appearing on both c.r.d.s marks.Question 2 is then answered as the address ink is not the same as the cancel ink.

-7-

The Sept 20, 1898 known good cover is shown in Fig. 6. The ink is generally lighter on the c.r.d.s. Pt. 8below will be compared to the address ink and stamp cancel on cover ink for the Red Revenue test cover, Pt. 6 and Pt. 7.

Fig. 6, known good Sept 20, 1898 showing test point Pt. 8 on the c.r.d.s outer ring.

-8-

VSC6000 comparative spectral reflectance analysis of the c.r.d.s. on the Red Revenue and known good covers is shown below.

Fig. 7, VSC6000 comparative c.r.d.s of Sept 5 and Sept 20 cancel ink. With the address ink of the Sept 5 cover as an additional comparative reference. Pt. 6, Pt. 7 and Pt. 8.

Discussion of Results: The % intensity vs spectral wavelength response for the known good Sept 20 c.r.d.s matches well with the Red Revenue Sept 5 c.r.d.s. mark at Pt. 7. The Pt. 8 vs Pt.7 reflectance response, that is the % intensity vs wavelength, essentially match, and as such the ink used at Tientsin on Sept 5 1898 to cancel covers was the same ink as that used on Sept 20. Question 4 is answered as an ink match. The Red Revenue cover was double canceled by the TientsinP.O. on the respective dates shown in each c.r.d.s.

-9-

Reverse arrival cover stamps ‘Shanghai Local Post’ ink analysis’ using the VSC6000.

Fig. 8, Shanghai arrival date stamp, Shanghai Local Post , for the Red Revenue test cover Post Office ‘J’ SP 10 98 and for the known good post office ‘J’ SP 23 98 cancels. Pt. 9 and Pt. 10 will be compared to a reference calibration 'white reflective sample tile' using the VSC6000.

-10-

VSC6000 Shanghai Local Post spectral reflectance data, along with a calibration curve for the machine as a reference point.

Fig. 9, Reflective spectral response of the ink used in the Shanghai Local post cover reverse cancels, Pt. 9 and Pt. 10. Calibration reflectance of a 'white tile reference' is given as a comparison.

Discussion of Results: The almost perfect overlay indicates that these covers were canceled at the ‘J’ P.O. in Shanghai. Question 5 is a match for arrival cancel ink. Both these covers traveled to Shanghai from Tientsin were received and sent out for delivery by the Shanghai Local Post19.

-11-

The VSC6000 at the Philatelic Foundation was used to obtain overlays17 of the c.r.d.s. cover cancels forthe Red Revenue Sept. 5 and known good Sept. 20 covers.

Fig. 10, Partial alignment overlay offset Sept 5 and Sept 20 c.r.d.s. size/font/numeral evaluation.

Fig. 11, Full alignment of Sept 5 and Sept 20 cover c.r.d.s. Matched overlay of cancels.

Discussion of Results: The perfect size/font/numeral results indicate a match for the c.r.d.s used at the Tientsin P.O. on Sept 5 and Sept 20.

-12-

VSC6000 overlay analysis of the Shanghai Local Post arrival marks.

Fig. 12, Overlay view of both arrival marks, note J, SP and 98. These will be aligned in the overlay.

Fig. 13, Overlay of the arrival marks, note the match of size, J SP and 98.

Discussion of Results: The arrival cancels are a match, most likely the same canceler was used on Sept. 5 and Sept. 20 at the 'J' P. O. of the Shanghai Local Post.

-13-

VSC6000 infrared capabilities to eliminate the Red Revenue 'RED' stamp color and allow the canceler to be viewed for comparison to the cover c.r.d.s.

Fig. 14, Infrared view under magnification of the Red Revenue stamp c.r.d.s. note the SEP 5 98as well as the 'TSIN' font.

Fig. 15, comparison of Sept. 5 c.r.d.s. cancels shows that the cancels match well. (This a reference viewonly, the VSC6000 provided the infrared image and cover enlargement to near match the size of the c.r.d.s on cover for both cancels was made for easy comparison).Discussion of Results: The c.r.d.s cancels used the same ink, the bite into paper of the '5' is similar, the canceler was most likely the same device, therefore this is an authentic postal history item. Question 3 is answered. -14-

Bruker-XRF ( X-ray fluorescence) Tracer III-SD testing of the Sept. 5 and the Sept. 20 envelopes. Test points on envelope surfaces free of ink, at the N.P.M. for wood fiber additives: Since the cancel ink sits on the envelope, the contents of the raw envelope is a variable, and for the reflectance data obtained with the VSC6000 to be accurate we would like to understand the envelope influence on this data. Similar envelope chemical composition is very desirable. This reduces errors of ink in comparative results. The device bombards the envelope with X-rays. This excites the surface paper to generate X-ray fluorescence. The X-rays “shoot” individual electrons out of the atoms of the elements, primarily out of the inner atomic shells K and L. The resulting vacancies are filled up again by electrons from higher energy shells. The excess energy of these electrons is then emitted in the form of X-ray fluorescence radiation. This radiation is characteristic for each element like a fingerprint and independent of the atom’s chemical bond. The intensity of the radiation is proportional to the concentration of the element in the envelope paper.

Fig. 16, Bruker-XRF element responses of both front envelope paper. Red is the Sept. 5, test envelope paper. Green is the Sept. 20, known good, envelope paper. For paper20, the elements of interest are Ca, Fe, K, Al, and S. For the test envelope these elements and Si are present. Si may be due to sand or dirt contamination of the envelopes surface. For the known good envelope, Si and Al are not present. Higher concentrations of Ca and Fe are present for both envelopes. The Tracer III-SD uses a Rh X-ray tube with a Si drift detector, as such the presence of Rh, Ni, Co Cu and Zn are attributed to fluorescence in the materials of the machine itself22, and represent unwanted spectral response of the X-ray tube and the stainless steel (Ni, Fe, and Co) detector can lines.

-15-

Fig. 17, Element compound responses of both envelopes, front and back.

Discussion of Results: The Bruker-XRF response spectral data indicates the high concentration of Ca and Fe in both envelopes front and back paper. Calcium compounds20 such as Calcium Carbonate, CaCo3 were commonly used as a wood pulp filler18, and as a bleaching agent off setting the presence ofIron20. The expected paper elements Ca, K, S, Fe are present in both, with Al present in the Sept 5 envelope paper. The known good Sept. 20 envelope has no Al present, meaning it most likely has no Aluminum Sulfate ( Al2(SO4)3 ) or Alum, which is probably present in the Sept 5 envelope. Aluminum Sulfate was commonly used as a sizing agent20, as such it gave the paper a mild hydrophobic effect, preventing applied address or inscription ink from diffusing and becoming blurry. The Sept. 5, envelope was thus suited to carry the Christian Missionary corner card, and the Jesus is Coming “Watch and Pray” text, with no ink diffusion of this text over the paper. In conclusion, the presence of Ca, Al, S, Fe and K are expected 20, since these were common paper additives or present in wood pulp or water used in the paper manufacturing process. Si present in the Sept 5, test envelope is most probably due to dirt or sand surface contamination.Both envelopes show a close match in that both are similar in additive composition, with the Sept. 5 Red Revenue envelope having Al and a higher concentration of Ca and Fe content than the known goodSept. 20 envelope paper. These envelope paper differences are not great enough to effect the VSC6000 spectral results, as demonstrated above.The presence of Rh, Ni, Co, Ti, Cu and Zn in the Bruker/Artax7 response is attributed to interferences in the X-ray beam generated in the X-ray tube and the spectrometer detector21 ( fluorescence of materials contained within the Bruker-XRF device )22.

-16-

Conclusion and Comments: The item was shown to be authentic. The authors wish to emphasize thatthis item was not expertized at the P.F. or the N.P.M. After approval of a submitted test plan, the P.F. and N.P.M. provided the use of their VSC6000 and Bruker-XRF machines. The conclusions reached and discussed in each section above are not opinions of either organization but solely the inferences of the authors and are conclusions based upon the machine data obtained by the authors. The purpose of this paper is to bring attention to the current state of the art in helping the collector establish an alternate route to authenticity for unique items.

Acknowledgements:The authors are especially appreciative to Dr. Jeffrey Schneider F.R.P.S.L. who provided not only guidance on the many China Post Office changes that occurred in 1897 with the advent of the I.P.O., but also the encouragement to follow the evidence wherever it leads. In addition the authors wish to give special thanks to Dr. Susan Smith National Postal Museum in Washington D.C. , and Larry Lyons at the Philatelic Foundation in New York, both of which provided the technical guidance in the use of the VSC6000 and Bruker-XRF machines, as well as Dr. Robert Shannon Jr. of the Bruker Nano Analytics for his interpretation of the Artax7 data.

-17-

Endnotes:1)The Red Revenue Surcharges China 1897 Vol. I, 1984, Republic of China. Pages 210,214,231.2)The Illustrated Covers, Postcards, Lettercards, Remittance Certificates of Revenue surcharges. Benjamin Y.K.Hwa, C.S.S. Pages 423,426,435, 470,477.3) Covers and Cards of the Empress Dowager Jubilee and its Surcharge Issues Tony To-Wah Kwan , Vol. III 2012. Pages 1143-1150, 1154, 1177-1178,1180, 1182, 1192-1194, 1196, 1213-1214, 1248, 1258, 1255, 1271-1273, 1281*, 1283-1284, 1290-1292, 1303, 1306-1307, 1313.4)The Story of the Christian and Missionary Alliance, Alliance Press 1900.5)Christies Swire Oct. 26, 1993, Lot 2597. Registered example of cover with corner card to U.S.A.6)Richard Pratt, Imperial China: History of the Posts (to) 1897, Sahara Publications Ltd. (1998), Chapter 1.7)Wright, S.F. Hart and the Chinese Customs, William Mullen and Son for Queen's University, Belfast,1952.8)Richard Pratt, Imperial China: History of the Posts (to) 1897, Sahara Publications Ltd. (1998), page 124.9)Richard Pratt, Imperial China: History of the Posts (to) 1897, Sahara Publications Ltd. (1998), pages 125-145. 10)Empress Dowager Jubilee and its Surcharges, Shih-Jen Chen. Page 208 c.r.d.s, Type I and Type II.11) 1905, Part 1. -Abstract of Statistics, Appendex M, Note on the Postage Stamps of China. J.Mencarini.12)Richard Pratt, Imperial China: History of the Posts (to) 1897, Sahara Publications Ltd. (1998), Chapter IV, Page 176.13) Email M.Kravitz, J.Schneider, May 17, 2017, May 18, 2017, June 21, 2017.14)Paper and Color Varieties of the People’s Republic of China “Workers and Soldiers”Definitive Set of 1955-1961, Archie McKee, IAP 2015 proceedings.15)The Admiral Issue of Canada: A Colorimetric and XRF Study of the Carmine 2¢ Issue, Richard H. Judge, IAP 2015 proceedings.16) Test Plan RevA M.Kravitz July 201717)Canada 1868 Two Cent Large Queen on Laid Paper: The Analytical Process Followed for Expertization, J. Edward Nixon, IAP 2015 proceedings.18)https://www.acs.org/content/acs/en/education/whatischemistry/landmarks/savannahpaper.html, on fillers, calcium carbonate.19)Richard Pratt, Imperial China: History of the Posts (to) 1897, Sahara Publications Ltd. (1998), Chapter IV, Page 174. “The I.P.O. took over the Shanghai Local Post as of Nov. 1, 1897”.20)http://paper.lib.uiowa.edu/conclusions.php#stability, on Ca and Fe in 19th century paper.21)A.N.Shugar, J.L.Mass, Handheld XRF for Art and Archaeology, Leuven University Press (2012), Page 95.22)http://rf.gurfr/tyled-5//-ownleod-t514//iedt/XR/F%20fStpdcyfr%%20fGurd-dgtp-. Bfrkdf XR/F Gurd-dg

-18-