The Topic Is Paper

© Calvet M. Hahn 2002


The first papers were discovered and used as a writing medium in China at least as far back as the second century B. C., although some authorities date it to the report of Tsai Lun who reported a process to the Chinese emperor in 105 A.D. Paper didn't come westward into the Muslim world until the battle of Samarkand (751 A.D.), when a number of paper makers were captured. As a result paper making became an Arab tradition. The Chinese had used cotton to make their paper, however, the Muslims began with flax and shifted to linen rags. Paper was introduced into Spain by the Moors and into the Byzantine Empire around 1200 A.D., if not earlier. This introduction is reflected by the references to 'rag paper' made by Peter of Cluny (1122-1150 A.D.) and 'cloth parchment' in 1263 A.D. King Roger of Sicily signed a deed written on paper in 1102 A.D.

The Moors had a paper mill at Xativa, Spain in 1150 and soon after at Toledo and Valencia, while a mill is known at Fabrizano, Italy (near Ancona) in 1276. (Its first watermarks are known in 1293 and 1294). France had a mill in 1189, Germany in 1291 (the Ulman Stromer mill at Nuremberg) and England in 1330. By 1239 A.D., paper had become so popular that Roman Emperor Frederick II, grandson of Roger, banned its use for public documents in lieu of vellum. Blotting paper (a paper produced with out any size) is known as early as 1465 while brown paper is recorded in 1570.

The first American paper mill, backed by William Bradford of Philadelphia newspaper fame, was the Rittenhouse mill at Germantown, PA in 1690. A second mill was launched at Elizabethtown, N.J. in 1728, and in the same year (some sources say 1732) Thomas Wilcox erected the Wilcox mills at Wawa, PA. (source of early stamp papers) and a group of New Englanders established a mill in that region. The first mill in New York was at Roslyn, L.I. in 1768.

While paper was imported for writing and book printing prior to the Rittenhouse mill, there was no restriction on American paper until the Stamp Act of the 1760s (a British revenue raising measure to pay for the wars with France), which was applied to every type of document from legal papers to newspapers and which aroused great resentment in America. Nevertheless the vast preponderance of colonial papers was imported.

Papers in the Mail Stream

There were three types of paper that passed into the mail stream in colonial days. One was the newspaper, beginning in 1699 with Publick Occurrences, a two-issue paper, and then beginning in 1702, the regular newspapers such as the Boston News-Letter and those that followed in other cities. Newspaper publishers took on the task of postmaster just to get the news early for their paper and to be able to exchange papers. It was a dispute between Benjamin Franklin and William and Andrew Bradford in Philadelphia over this privilege that led Franklin to seek the postmastership of that city in 1737.

From then on exchange newspapers could be found in the mail system, although not all papers used the mails, which is why we have 'subscription posts' during the 1770s of which Goddard's is perhaps the best known. In the mid 19th century the demand for paper for newspapers and magazines led London publisher Mr. Lloyd of Lloyd's Weekly to introduce esparto grass as a paper fibre and soon some London dailies had esparto farms in North Africa to supply their paper mills.

Second of the colonial paper items was the envelope. These were almost entirely of European origin. Envelope history goes back to Sumer and the cuneiform tablet envelopes (3,500 B.C. to 7 B. C.). The envelope was not revived until the Renaissance when a series of Venetian envelopes of the 15th century were part of the Frank Staff collection. The Edith Faulstich postal history sale at Robert A. Siegel of November 19-21, 1973 had as lot 122 a history of early envelopes beginning with a French parchment envelope of 1478.

The Meroni postal history sale by John Fox on February 14-16, 1983 had as lot 620 a homemade paper French envelope of 1719. A French commercial envelope (cream laid paper with pointed flaps and a border of flowers in red and green) was reported used by Madam de Pompadour in 1761. Marshall noted an even earlier envelope of May 16, 1696 used by Sir James Oglivie to William Trumbull, the British Secretary of State, which was probably of Continental manufacture

The commercial envelope seems to have appeared in England circa 1817 following the 1824 letter to Postmaster General Chichester by Francis Freeling (Secretary to the Post Office) discussing an 1817 legal opinion on the rate to be charged on unwanted comic Valentine's sent in envelopes. This special ruling reduced the double postage normally required to a single rate. The first commercial manufacturer of envelopes in Great Britain according to Frank Staff was a Brighton stationer named Brewer who sold them in the 1820s.

Another early English envelope manufacturer was John Dickinson (of silk thread paper fame), whose sister wrote him in 1835 about the manufacture of 'pockets' and whose diary of July 1837 reported 'envelopes at 2/6d per 100 to be had at 200 Regent Street.' Rathbone Hughes & Dien, a well known forwarder, posted an early non-commercial envelope to America at Liverpool January 29, 1802. It arrived and was postmarked at New York April 12, 1802. The De La Rue firm constructed an envelope making machine in 1845 and began producing large quantities of envelopes to meet the new demand. The firm was also known for the production of surfaced white art board.

The earliest American homemade envelope of which I am aware is one in my collection of May 1775 from Mohawk Castle, N.Y. to London, which traveled by 'express' to New York reaching there on May 3rd to catch the London packet, which reached London by June 12, 1775. The Postal Markings issue of October 2, 1943 illustrated an envelope in the Maurice Blake holding that arrived at Charleston S.C. about 1799 for transmission to Philadelphia. The paper was watermarked 'JARDEL/LA ROQUE' indicating French manufacture. The hand-cut envelope measured 3¾ x 5 inches in size. There are a few other fairly early (pre-1845) envelopes found in the U. S. mails including the Northern Liberties Reading Room covers of 1836 but because the use of an envelope doubled the already high rate of postage they are quite scarce.

Envelopes began to become common by 1849. James Logan in his Early History of the Envelope notes Josiah Loring & Co. of Boston was making hand-made envelopes in 1840 (cut by knives and folded by hand). One of the Loring envelopes was apparently used by Fletcher Webster, Acting Sec. Department of State, on April 18, 1841 and addressed to George W. Gordon of Boston. Logan noted that French-style envelopes were also made in Louisville, KY and Philadelphia, PA in 1831-41.

The Louisville manufacturer was Edward Maxwell who cut envelopes with a penknife and began selling them in 1835, continuing through 1840 at least. I have not identified the Philadelphia manufacturer. It might have been the firm of Lehman and Duval as Duval had come from France in 1831.

In New York the earliest envelope maker seems to be a Mr. Pierson who had a print shop on Fulton Street in 1839 who made commercial envelopes in 1843. He found hand production was too costly and abandoned his envelope line. He then sold the business to William Dangerfield, who operated out of a rented room at 180 Fulton. Financial difficulties forced Dangerfield to sell out to his landlord Jacob Berlin in 1847. The results were so disappointing that Berlin only continued operating to keep his employees going until he found a buyer, but business picked up. Berlin sold the business to William G. West in 1852, and his son, Henry C. Berlin stayed with the firm, becoming a partner in 1853.

In 1852, the partners bought a French envelope-making machine for $600, which did not prove very successful. In 1853 the business was reestablished as West and Berlin at 67 Pine St. and in 1855 was moved to the firm's own six-story building at 120 William. At this time they employed about 1200 hand folders and produced about a quarter million envelopes daily. In 1856 West sold his interest to George H. Jones a stationer located on John Street and the firm became Berlin and Jones. In 1857 it again moved to larger quarters at 534 Wales St. and production was 600,000 envelopes daily. It was an important producer of Civil War patriotics.

The third paper product to enter the American mail system was writing paper. Most of the better grade colonial writing papers were imported rather than domestic products. These were all hand-made until 1798 when Louis Robert employed by Messrs Didot of the Essonne Paper Mills in France invented a papermaking machine. It was not particularly successful, however, it inspired Henry Fourdrinier (1766-1864) to develop a more successful machine for his paper mill at Dartmouth, Kent and in 1803 he erected one at Frogmore in Herts. The first Fourdrinier machine was brought to this country in 1827 by Henry Barclay and installed in a paper mill at Saugerties, N.Y. In 1830 the manufacture of Fourdrinier machines began at South Windham, CT.

While machine made paper could be found this early, it was scarce prior to the mid-1840s. By 1850 there were about 480 papermaking machines in the U.S., almost all Fourdriniers. The Wilcox firm (1729-1866) was headed by James M. Willcox who ran its Ivy Mills from 1832 until his death in 1854, during which time it was almost exclusively involved in making banknote paper. In March 1866 the manufacture of handmade paper was abandoned. By the 1880s there was but one handmade paper mill remaining, located at North Adams, MS. The Willcox firm apparently had two mills where Fourdrinier machines were installed to make machine-made paper.

One Fourdrinier machine was installed in 1837 at their Glen Mills plant 2½ miles away from the Ivy Mill and another Fourdrinier was installed in 1845 . In a series of studies of the first Canadian stamps, it was determined that Rawdon, Wright, Hatch & Edson supplied that country with hand-made stamp paper from 1851 through to 1858. At that time, a smoother more constant paper was introduced. Research revealed that the Crane Mills at Dalton, MS had introduced a new banknote paper about that date. While the Crane Co. reported producing a similar banknote paper, they felt the Canadian beaver paper samples supplied by Boggs were not their product.

Making Paper

Paper is the end product after removing the glutinous and resinous material from plant fibres leaving just the cellulose. Dissolving the raw paper components (called furnish), which looks like wet cotton wool, by beating and boiling it in a weak alkaline chemical solution results in the raw 'stuff' or fibre mass that are used to make paper. The addition of prussiate of potash in the furnish can create a blue-tinged paper, as can the interaction of certain printing inks (the British 1840 issue reportedly show this) or the use of prussic acid as a bleach rather than just chlorine.

'Sizing' is sometimes added to the 'stuff' during this beating process. The size is a substance such as glue, gelatin, or China clay (kaolin) added to fill the spaces between the fibres and make the paper non-absorbent to ink or water after it is 'finished'. The better grade of handmade papers, however, is tub-sized at a later stage. 'Stuff' is about 98% water and 2% fibre and looks like milk. If a colored-through paper is to result, a dye is added at this stage.

The earliest blue colored papers used iron-blues, which began to be replaced commercially by ultramarine in 1829. Dr. Chase had the bluish tinted papers of the 1847 issue examined by a chemist who found they had been 'dyed' by an ultramarine pigment made from lapis lazuli (a sodium /aluminate silicate containing sulphur).

Until the 1850s, rags of linen or cotton with occasionally silk were the chief raw components both in Europe and America. Silk fibres do not bleach like the others so that they can be detected in the finished paper, but only a few fibres, if any, are likely to be found on a given classic issue stamp except for certain revenue issue adhesives where silk fibres were specifically added to the 'stuff' for security purposes. There had been a shortage of rags during and prior to the Revolution, which accounted for some of the poor quality paper produced. The Boston News Letter of 1769 reporting that the bell cart would go though the town to collect rags added,

"Rags are as beauties which concealed lie,
But when in paper, how it charms the eye!
Pray save your rags, new beauties to discover,
For of paper, truly, ever one's a lover."

The poor quality paper is particularly noticeable in the newspapers of the period. The Massachusetts General Court in 1776 require the Committee of Safety in each town to appoint a rag collector and asked citizens to save even the smallest rags. However, by 1791, Alexander Hamilton reported there was an adequate supply. By 1810 there were 179 paper mills in America, mostly two-vat affairs with a total investment of about $10,000 each. Another domestic rag shortage took place during the War of 1812 and domestic rag prices doubled and imported rags were widely used and experiments with other fibres took place. Some 70 mills had to close by 1820. It was about this time that shipload after shipload of Egyptian linen mummy wrappings were imported; they made up an important component of the highest grade of American paper for several decades.

Hand-made Paper


In hand-made paper, the papermaker takes a wire mesh mould fitted with a detachable rim called 'deckle' and dips it into the vat of beaten 'stuff' and water and lifting the mould out when the desired thickness of fibres are present and lets the water drain out. The height of the deckle determines the thickness or thinness of the paper. He then hands the mould to a 'coucher' who 'couches' or turns the soggy mess in the shape of a fibre sheet onto a felt blanket, covering it with another felt blanket. When enough of these felt layers have been built up into what is called a 'post' they are put into a press, which squeezes most of the water out and the sheets are sent through a second press before drying.

Sizing is added by tub-dipping the sheets in a gelatin or animal glue solution to which alum or China clay (kaolin) is added. After the size is allowed to set, the sheets are then passed through another set of rollers to eliminate excess sizing and the paper is ready to be 'finished'. The arrangement of the wire in the mould gives us the characteristics by this process is called hot pressing. which we know paper such as laid or wove, batonné or quadrille.

Too, 'bits' with letters or figures may be attached to the wires of the mould; these yield the watermarks we note on paper. They may be 'sheet' watermarks, found only once on a full sheet, or attached throughout the wires so that a full watermark or part thereof falls on each stamp printed from that sheet of hand made paper. For machine-made papers, they may also be attached to the 'dandy roll', which presses machine-made paper sheets to squeeze out remaining water. In the early 1800s, it took three men a day's work to make 4,000 small sheets of handmade paper and up to three months to complete the process.

For handmade papers, finishing is done by placing the sheets individually between two zinc or copper plates until a stack sufficient for the glazing rolls (usually about 48 sheets or two quires) has been reached. This stack is passed back and forth between two chilled cylinders under considerable pressure to 'calender' the paper to the appropriate glazed glossy surface. A glazed or burnished surface can be quickly generated by friction by passing the paper sheet between a heated iron roll on top and a pressed cotton roll on the bottom with a ratio of four revolutions to five.

Wove paper comes from a mould with very fine wires so that under magnification the paper looks as though it had light and dark dots representing the wires and the holes between. Laid paper has the individual wires 'laid' in parallel lines supported by other cross wires (called chain wires) at intervals of about an inch. Depending upon how the sheet of stamps was printed they can be termed horizontally, vertically or diagonally laid (this last is seen in the case of U.S. envelope papers).

Batonné paper can be wove or laid and is characterized by comparatively thick wires running parallel to each other at about half-inch intervals. This paper is most frequently found in writing papers where the lines can help guide the writer. Quadrille paper has prominent raised wires creating squares or rectangles. Both it and batonné paper have occasionally been used for stamps but except for the 1888 quadrilled Bayonne blizzard stamp are not found in U.S. stamp production. The quadrilled album page familiar to collectors is not made from quadrilled paper; rather the quadrille is printed onto the paper.

As far as has been determined all the U.S. independent mail adhesives, the adhesive U.S. provisionals and all early locals issued prior to July 1847 are made on handmade papers. This would include over 100 Scott listed adhesives. This total does not include early express labels, which also would have been on handmade paper.
Included in this list are the 27 major Scott listed independent mail adhesives put out by American Letter Mail (3), Brainard (2), Hale (5), Hartford (3), Hoyt (1), Letter Express (4), Overton (1), Pomeroy (7) and Wyman (1).
The 22 adhesive provisionals on handmade paper include Alexandria (2), Baltimore (4), Boscawen (1), Brattleboro (1), Lockport (1), Millbury (1), New York (3), Providence (2) and St. Louis (8).

Only handmade paper seems to have been used for the 63 local and carrier adhesives issued prior to the 1847 issue. These were the adhesives issued by Barnard of Boston (1), Blood (6), Boyd (5), Philadelphia City Express (3), City Despatch New Orleans (2), City Mail (1), Cummings (5), Cutting's (1), Eagle City (2), Franklin City (1), Frazer & Co. (5), Greig (1), Hall & Mills (1), Hampton (2), Hanford (1), Harris (2), Jones (1), Kidder's (1), Mead (1), Mearis (5-6), Morton (1), New York City Express Post (2) Philadelphia Despatch (3), Walton's (1) and nine U. S. City Despatch Post adhesives . Some later locals probably also used handmade papers, but no regular study has been made of the subject to date.

Among the specialty handmade papers of interest to stamp collectors are India paper and Japan paper. India paper is a tough thin opaque paper frequently slightly toned that was originally introduced to England from China circa 1750 and made there at the Wovecote Mills. It contained bamboo fibre as a component. It was used to print Oxford bibles for distribution in India, whence its name. Philatelically it is often claimed India paper used for U.S. die and plate proofs. Roy White notes the 'India' paper actually used for U.S. proofs is not that tough or thin but rather a lightweight, smooth or satiny paper with a very light creamy color that is the result of highly macerated 'stuff'.

Japan paper is a silky native paper used for philatelic repairs. The United States City Despatch Post introduced colored-through paper philatelically when it released its 6LB3 and its essays in August 1842. Colored through paper has the same color on both sides and on the cut edges. The same government carrier also introduced glazed surface-colored paper on November 26, 1842 with its new printing of the green adhesives . These papers have a colored coating applied during the finishing process after the paper has initially had the water pressed out.

Some surface colored papers have the color printed on them. The glazing is done by heavy pressure with heated rollers applied to the coated paper stock. The Wilcox Wawa, PA mill specialized in such high-grade papers

Pelure paper is generally a misused term in classic U.S. philately. The true pelure papers are found on the Hawaiian missionary stamps, a thin, crisp, hard, transparent paper with a bluish or grayish tinge that is quite brittle. What are normally termed 'pelure' papers in U.S. classic philately are 'thin' sheets of paper, created deliberately to make up the correct weight of a ream of paper, the measure by which it is normally sold.

For example in the first American Letter Mail stamp these thin papers (called erroneously pelure) ranged from 1.8 to 2.2 mils in thickness compared with a normal stamp paper range of 3.2 to 3.7 mils. About fifty sheets of this thin paper were used in the production of the New York provisional adhesive. The thin papers are not brittle or as transparent as true pelure paper. Extra thick paper is also made and can be found among some of the U. S. independent mail adhesives. In the case of the American Letter Mail stamps these ranged from 4 to 4.5 mils in thickness, or about 20% thicker than normal.

Machine-made Paper

In machine-made paper a continuous belt brings the 'stuff' from the vat onto an endless 'mould' of finely woven wire 60-foot long belt that is about 90 inches wide with vertical 'deckle' edges so that the 'stuff' doesn't fall off. Some machine made papers have been sized during the beating process, but most are not. The usual wire cloth belt has 66 meshes per inch when writing paper is made, although a high grade 80 meshes to the inch belt is sometimes used, while for coarse papers the mesh can be less.

This belt 'shakes' so that the fibres are interwoven while the water drains out below forming a continuous sheet of paper. Towards the end of the belt, the web passes under a hollow cylinder 'dandy roll', which impresses the watermarks and the incipient paper then proceeds through various rollers to press, dry and finish it. The dandy roll is situated between two suction boxes, which help drain the water from the web.

Watermarks are thinner sections of the paper web created by the raised 'bits' on the dandy roll just as the raised wires or 'bits' attached to the mould in handmade papers create laid or batonné lines and watermarks. They show darker in watermark fluid.

While not found in U.S. classic philately, a few stamps such as in the 1857-58 issue of Russia, have an opaque watermark, which is a reverse made one wherein the dandy roll roller is incised with the watermark so that the paper is thinner outside the watermark area rather than the normal raised watermark where the thinning is made by 'bits' on the dandy roll. In most countries only a few companies make dandy rolls and they usually keep records of their products so that the watermark can be tracked to a specific paper mill, which, in turn, may have keyed it for dates of use. This practice is of use in expertizing papers.


Right after passing through the drying cylinders the web is slowly passed through a vat of heated solution of glue or gelatin sizing containing alum and then between brass rolls to squeeze out the excess size. The web is then rolled up and the size is allowed to set before passing over a series of open 4-foot drums in which fans circulate hot air to dry the paper web. The paper is then 'finished' by calendering to give it the appropriate finish. Machine-made paper can also be hot pressed as in handmade papers, a much more labor intensive and expensive process. Envelope papers are usually burnished, as only one side, the address side, needs the polish of glazing.

Identifying Handmade vs. Machine-made Paper

One identifying characteristic of machine-made paper is orientation or 'grain' caused by the fibres orienting themselves in the direction the web carrying the 'stuff' is moving As paper has air spaces (from 30% for dense papers to 70% for light bulky paper), it expands when subject to moisture about 2 to 5% across the direction the paper web moved. Handmade papers do not have this directional expansion; they expand equally in both directions. Machine made papers also have 'mesh', which is found on the underside of the paper, which was next to the endless wire belt, while the top is smooth and is the side used to print upon.

When comparing wove handmade paper with wove machine-made paper, the handmade appears to be cloudy and blotched while the machine-made paper has this characteristic to a much lesser extent due to the difference between the coarser weave of the mould in handmade paper and the fine weave of the web in machine-made paper. As the Crane executives stated to Boggs, this characteristic may be difficult to use as identification on such small pieces of paper as a single postage stamp. However, if the cloudy effect is obvious the paper is handmade. A third identifying characteristic is the stitch watermark found only on machine-made paper.

The envelope provisionals may be machine-made envelopes, particularly as some were imprinted on envelopes brought to the postoffice for imprinting by individuals as in the case of the New Haven, which is known on white, bluish and buff paper. The Baltimore envelopes are on white, manila , buff and salmon, while the few New York provisional envelopes were probably on buff, which was typical of the envelope supply used by Robert Morris for other purposes in 1847 . These manila, buff and salmon envelopes are probably machine made due to the lower cost possible and the lack of a necessity for a high quality product.

Machine-made stamp paper was first used for the production of the 1847 issue as denoted by the stitch watermark found on examples of that printing. Machine-made wove stamp papers come in by the May 13, 1848 second printing of the 1847 issue if it was not used for the first printing in June 1847, as it may well have been. U. S. issued stamps thereafter are printed on machine-made paper.

Any stamp issue that has 'stitch' watermarks is on a machine made paper, for that stitching is the result of creating the continuous belt in a Fourdrinier machine. While handmade stitches do exist on stamp papers they are quite different and irregular in appearance so that are unlikely to be confused. In the New York provisional, the 'sheet of nine' is known with 'stitch' watermarks dating it as a later production than the issued New York provisional stamps, none of which show this characteristic.

The 1851 issue paper was definitely machine-made and of about .005 inches according to Chase's studies reported in Chapter XXXI of Ashbrook. The initial papers were from the Willcox Ivy Mill. A rose colored-through paper, however, was used for the Franklin carrier of 1851 and the locals of the 1850s may or may not be on machine-made paper. That stamp paper used in 1852-3 shows ribbing, which was caused by a worn press blanket used during the printing and does not represent true ribbing or laid paper.

Roy White challenges this theory and suggests in his Papers and Gums book that pull rollers, spirally bound with wire, located between the final drying phase and the calender finishing process are responsible. Neinken reports that at first the 1851 issue paper was thick and opaque but of a fine quality and hard and crisp and that later just before the perforated stamp appeared a thinner, slightly transparent paper was used .

Dr. Carroll Chase noted the possibility that some handmade paper was used to print the 1851 imperforates as he had an imperforate 3¢ that had 1/3rd of the stamp on thick paper and 2/3rds on thin paper, a characteristic likely to be found on handmade paper and not on machine-made. This might be the result of Toppan, Carpenter buying paper on the open market. He also reported a plate I printing on paper containing bamboo, esparto grass or straw when tested by the Hertzberg test. This test uses solutions of zinc chloride and potassium iodide and iodine crystals. Cotton, linen and help fibres are stained wine red when tested, while esparto, straw and bamboo fibres show a blue to violet stain as does wood cellulose. Mechanical wood pulp gives a yellow stain.

On June 22, 1857, Toppan Carpenter, the stamp contract holder, wrote the Crane Paper Co. of Dalton, MS that they had previously ordered 300,000 sheets of another party that would soon be delivered and therefore deferred sending a sample of the paper stock they wanted from Crane; a 17 pound sample was received July 15, 1857. On August 10th Carpenter sent a sample, which weighted 17 pounds per 1000 sheets and stated they wanted one that weighed 16 pounds and placed an order for 100,000 sheets. The Crane Co. put a note on the letter that the paper must not be transparent and suggested sending a few sheets of hot-pressed finished rather than calendered paper. (This paper cost $25 a thousand sheets rather than the average printing paper cost of $3-$5 a thousand).

A shipment of 12,000 sheets was sent to Toppan, Carpenter in October and on November 25th an order stating that the balance of the shipment should be this hot pressed surface paper should be shipped. As noted in a letter of December 2nd the company anticipated orders of several million sheets over the next four years. The initial shipment of 124,588 sheets was sent December 29th and invoiced January 14th. This is the 'hard, white paper with a whitish cast' paper used for plate 5 printing and found on cover on January 2, 1858. Valente reports it was thinner, averaging .0024 inches as compared with the .0027-inch thick Willcox Mill paper also used for Plate V.

The paper used during that year has a mottled appearance from the back and is thin and crisp. Beginning in 1859 the paper is 'soft' and shows a distinct grain. This is shortly after the introduction of a new type of beating machine, the Jordan refiner, in 1858, which greatly speeded up the beating process. The 'laid' papers of the 1857-60 issue as well as those of the 186l-1869 National banknotes are probably 'false laids' resulting from the wires on the pull rollers of the Fourdrinier machine, although open market paper purchases might have resulted in laid papers as made by the dandy roll.

There was in 1851 Plate I early paper that is known from the Boston deliveries in July and August of 1851 that Eugene Jaeger termed 'part India' in a Stamps article of November 28, 1936. His specimens measured .0011, .0012 and .0014 inch. He also reported the very thin paper from both Boston and St. Louis in September 1851.

Papers of the 1860s Issue Stamps

In 1860, domestic cotton rags made up 70% of the raw material used for paper, largely from cotton mills. The Civil War cut off cotton supplies and paper prices rose sharply while quality was degraded in the later 1861 stamp printings. The 1861 National banknote issue premiere gravure essays were printed on a very thin, transparent paper that cracks easily. It measures about .65-.75 mils in thickness. The regular issue paper used for the first 97 shipments was initially a thin translucent hard paper measuring .70-.75mils until January 1862 at which time it shifted to medium thickness paper measuring .80-.85 mils. There was apparently a rodent problem in the paper mill supplying the 1861 paper in the Spring or Summer of 1863, as the mill used arsenic as a rodenticide, which got into the water and resulted in arsenic getting into the stamp paper and causing the ultramarine shades to darken, creating the dark blue and indigo shades known immediately after.

While Toppan, Carpenter lost the 1861 stamp contract, its successor, Butler, Carpenter did win the revenue stamp contract from the Treasury. The paper used was a very hard, brittle closely grained one apparently obtained from the Willcox Ivy Mills. This is called old paper. The issued stamp paper continued until September 1, 1869 when the government indicated it was going to supply new watermarked paper rather than let the firm purchase its own.

Mr. Carpenter noted, '…regular stamp paper must possess certain qualities to hold the inks well, prevent rubbing, etc.' He reported he had sufficient old stock to last to September 1st, but they had to go out and purchase as good quality as possibly paying more for it than for the regularly made to order paper. He changed papers several times, but said that while the new paper prints better, it does not hold the ink as he wished. This meant that all revenue stamps not on the thin, brittle paper were printed no earlier than September 1869.

There are four distinct papers identifiable of the first issue revenues. First is the hard, brittle, almost transparent yellowish or grayish appearance already noted, second is a very soft, white and often quite porous paper of various degrees of thickness. The third is a thick almost thin Bristol-board paper, often very white, hard and rather brittle. Fourth is the rather soft, thick paper with silk fibres. In a letter of July 21, 1870, the firm reported that even Willcox & Co. seemed unable to furnish paper of the same old paper quality reporting,'…it is not sufficiently sized and wants polish; undoubtedly the gum will strike through it.'

On August 17, 1870 Carpenter asks for a sheet of Hudson paper, this is the green silk fibre revenue paper made by the Hudson Paper Co. Two thousand sheets of silk fibre paper (actually colored cotton fibres were used) were recorded on April 18, 1871, 999 with red silk fibres added and 1003 with red and blue fibres. In addition to the green silk fibre paper there was also a bright yellow 8-pound paper used for the U.S. Proprietary Medicine 1¢ stamp. There was also a white 8-pound paper used for the 2¢ Fred Brown & Co. Medicine adhesive and the 1¢ Truesdell. Both the National and Continental banknote postage stamps have blue and red cotton thread papers issued on an experimental basis.

Two special papers were used to print the Byam, Carlton match proprietary stamps. The first was a cotton rag white tissue paper of 16 pounds weight, while the second was a manila tissue paper of 12 pounds weight. Tissue papers are given an extended maceration and reduction in the beating vat.

Beginning in August 1867, U.S. stamps began to be experimentally grilled, a process which was thereafter required by the new printing contract of that fall. This is a printing process done by using grilling cylinders in the stamp manufacturing process and did not create new paper varieties but rather finished stamp varieties. Several types of grills were applied at the printing plant.

Beginning sometime in 1868 a new very thin grilling paper was introduced. Both the 3¢ and 24¢ E-grills were on this new stock, indicating that these thin paper printings were the last of the E-grill productions. Among the subsequent F-grill printings only the 1¢ through 15¢ values are found on this paper stock. To me this time sequence indicates that the other F-grill values were grilled on stamp stock already manufactured prior to the first E-grills, but not yet finished. I can find no record of these very thin paper adhesives on covers dated before the summer of 1868, well past the E-grill printing period.

Brookman has suggested that the thin papers were used to get better efficiency from the grilling process, an unproved but probably correct observation. In a break-though article , Eliot Landau notes this thin paper had more sizing in it, producing a stiffer harder paper but one that was more brittle. He notes this paper was never observed on earlier stamps and is not found used later. However, the weight of the thin grilling paper was only slightly different from the regular paper used; it averaged 1.2 grams per 100 stamps with ranges running between 0.5 and 2.1 grams per 100 stamps.

The big difference was the extra sizing, which yields a sharper impression during the grilling process. The thin grilling paper stamps have a sharper 'snap' when flicked and present a slightly darker or gray appearance than the white to cream color of the regular papers. When held to the light the paper is more translucent and some have described the back as having a 'waxed paper' appearance.

The same security concerns that resulted in the grilled stamps created a number of paper varieties, which are mainly known only as essays. The only issued stamps were an experimental use of the Francis patent. This was a process by which the stamp paper was soaked in an alkaline fluid which turned it brown, so that the application of acid on a small sponge attached to the thumb of the canceller would turn the stamp blue. The Third Assistant Postmaster authorized a test of 1,000 stamps at Newport, RI of both the 3¢ value and the 2¢ Blackjack plate 31.

Under date of March 30, 1865, the Newport, RI postmaster wrote Dr. Francis about the usefulness of his patent, which had been tested in that city. A 3¢ dark brown red uncancelled copy was on a cover addressed to the National Banknote Co. Percy Doane bought this cover, lot 851 in the first Clarence Eagle sale 4/4-10/1923. A 24¢ chemical paper example was lot 994 in the same sale.

Among the essays is that of the Henry Lowenberg patent of April 1864 on goldbeaters' skin. All values of the 1861 issue were prepared and printed on the back of this thin very brittle transparent paper and some were gummed on top of the printed impressions so that the designs would have appeared in reverse when used on an envelope and any attempt to remove the stamp would have resulted only in the clear, transparent 'paper' coming off.

Another experimental essay found on the Blackjack is the experimental bluish-green (emerald green) wove paper. This essay is also attributed to Dr. Francis. Like the other essay papers, both are outside the selected scope of this study, not being on issued stamps.

A Major Technological Revolution

According to Dard Hunter, London printer Mathias Cooper (or Koops) experimentally produced wood pulp paper as early as 1800; he was also the first to produce paper from straw. In 1840 H. G. Keller obtained a German wood-grinding patent. The soda (alkaline) process for producing wood pulp was developed in England in 1840 by H. Burgess and Charles Watt, and patented in the United States in 1854. By 1863 the American Wood Paper Co. was formed and was producing twenty tons of soda process wood pulp in their Mayahunk, PA mill. This is paper produced by boiling wood chips in caustic soda at high temperature. On page 27 of White's Papers and Gums he notes the earliest example of one of the part wood pulp papers for stamps is a 15¢ Continental postmarked July 1875.

In 1857, B. Tilghman investigated the sulfite (acid) process, which is one that dominated the paper industry by the time of the Washington-Franklin stamps, but the first commercial sulfite pulp mill was not built until 1882 and it was 1884-5 before sulfite pulp was produced in the United States. The sulfate process of pulping wood used today largely succeeded it. As forgers frequently use paper from a later period, any use of out-of-period papers is suspect.

The National 1861, 1869 and the early 1873 Continental issues were produced on all-rag paper. The 3¢ 'American' banknote, like the 15¢ Continental cited above, was printed on 80% rag and 20% soda pulp paper suggesting that conversion occurred around 1875 in the Continental printing contract before the consolidation with the American Banknote Co. However, paper used for envelope printing in the North changed at an earlier stage during the war to the soda wood pulp process and this wood pulp envelope paper is also found on late Confederate home-made envelopes. A U.S. wood pulp envelope bearing stamps that allegedly were postmarked during the prewar period is highly suspicious and a probable forger's product.

The paper to be used for printing stamps was never specified in the stamp printing contracts of the classic era; it was a decision left to the banknote companies. The basic paper of the 1869 issue was a hard white wove, finely grained with a slightly yellowish cast to it. Some examples, however, have a bluish caste resulting from a small amount of ultramarine added to the 'stuff' to whiten it. Roy White illustrated a 2¢ post-rider stamp showing this bluish cast in his Papers and Gums book.

According to an article by "Cosmopolitan" (which may be a pseudonym for J. Walter Scott) published in the October 20, 1869 American Journal of Philately, a Massachusetts firm supplied the 1869 paper and the quantity was 16 tons annually. This would be the Crane Paper Co. of Dalton, Mass.

There are two 1869-issue paper exceptions. The first is double paper. The H. G. Mandel holding had a block of the grilled 3¢ in which the upper grilled paper was thin and the ungrilled under paper was thicker and served as a backing. The 30¢ without grill is also known on double paper. The shade is slightly darker and a large block of 15, with a cracked plate was in the Hind, Moody and Ishikawa holding. As all the bicolor high values were printed by the first week in July 1869, this would have had to occur at that time. However, there is a question whether some of the 30¢ without grill examples were not printed long after the 1869 issue legitimately went to press, this may not be an issued stamp.

Commenting on these double papers Luff notes that it was never determined if these doubled papers were essays or actually in use. However, Charles F. Steel, holder of the grill patent, also held patent 86,952 which was a double paper patent issued February 16, 1869. The timing would have permitted both an essay of a low value and a bi-colored example of the 1869 issue. Double paper is not otherwise recorded before about 1875, when it was used under the Continental printing contract.

The second exception is gray paper. The 3¢ gray papers are all used in the spring of 1870 (March 19th being the earliest date), but are very sharp impressions and one is known to be from plate 11, an initial printing plate. This suggests the stock was printed early and later stock placed on top until the end of the 1869 period cleaned it out. All but two of the eleven covers recorded are from New England; the exceptions are one from Philadelphia and one addressed to Lordstown, O. which may be postmarked in that state .

The paper is described as having a decided blue-gray appearance similar to the 'blue papers' of 1909, and in several cases a pinkish cast. It is a thinner paper with more 'snap' that the usual 3¢ 1869 paper The leading study of this paper reports that the 3¢ 1869 is also found on a number of different paper varieties ranging from extremely thin brittle paper to what seems to be a thin cardboard.

The National banknote issues of 1870 also appear on double paper. Luff noted he had seen the 6¢ and 24¢ values with the former on a cover posted at Providence, RI November 1, 1870 and believed other values of the issue were also on this paper. While the patent called for impressions to be made on a very thin paper which would then be backed by a thicker paper, however, Luff noted the examples he saw were printed on the thick paper nullifying the security concept for which the double paper was conceived.

Luff also noted the 1870 issue was printed on varying thickness of white wove paper ranging from thin to moderately thick. It is possible the thin papers observed on some off cover stamps resulted from soaking the double papers. It also appears that some of the thin paper layers in the double paper stamps were scored by a number of small horizontal cuts (as in the case of the Worthington sale 1¢ Continental block of four). The 3¢ banknote is known with this thin paper and some may have been issued at Washington D.C.

Chemical paper National banknote examples are known on yellow-brown and violet paper, both wove and laid. A 3¢ imperforate pair and block of four of the ungrilled National on gray chemical paper were lots 1365-6 in the Clarence Eagle I sale. While examples with postmarks are recorded, they are considered to be simply essays and not issued stamps. Over the years dealers and collectors destroyed many of these as stained or water-damaged copies so that they are scarce

In revenue stamps, Butler, Carpenter introduced a new chemical paper in 1871. The Willcox paper firm of Glen Mills, PA May 16, 1871 under patent #115005, patented this chameleon paper as it is called. As the patent noted, the paper is colored with two or more coloring or tinting substances, of which at least one is differently sensitive to some chemicals from one or more of the outer coloring. The paper is saturated with these chemical tints. For example it might have a red that was sensitive to acid, a blue sensitive to alkali and a yellow that is permanent under both.

Playing upon the government's fear of counterfeiting, Steel inaugurated another set of paper experiments in 1875, while working at the new Continental banknote contractor's plant. The first was another double paper experiment. The January 1877 Coin and Stamp Journal reported this 1875 experiment, "It is not generally known, and will be news to our collectors, that about a year ago, 20,000,000 stamps were issued to the public, printed on double paper. The upper portion receiving the impression was soft and porous and it was supposed that any attempt to clean off the canceling mark would render the impressed portion perfectly pulpy and thus effectively destroy it. The stamps did not meet with much favor and the plan was abandoned."

George Sloane gave a different version of this story. He reported that from January 1, 1875 until April 15th, 28 million such stamps were issued. They were ordered withdrawn on the 15th because of complaints from postmasters that the top layer of paper had a tendency to shrink away from the heavier backing causing tears, and would not pack resulting in confusion in accounts, loss of time, and great waste as a consequence of stock that was returned from the postoffices.

It is not clear that the experiment was stopped at this point, for the 5¢ Taylor which was printed later and sent to postoffices in June 1875 (although its earliest use is currently July 12th) is recorded on thin paper as early as lot 630 in the Worthington sale of 8/21-23/1917. To the extent that the top layer was a thin soft paper, the thin paper stamps we know do not fit this patent, but it is not improbable that a reverse paper process was used with a thin hard paper on top and a thicker soft paper below. Sloane comments that the 2¢ Continental brown double paper is frequently mistaken for the American soft paper special printing (Scott 193), which is not known used.

Steel also created a starched paper experiment based upon his patent 169,125 filed March 15, 1875. It seems to have strictly an experimental essay as no postally used examples have been recorded. He described it as follows in his patent, "I take a soft unsized paper, analogous to blotting paper, quite soft and absorbent. Having printed the face from the properly engraved plates, and allowed the ink thereon to dry properly, I treat the back with a solution of starch of just a proper consistency, have the effect both to lay a thin coating or covering on the back surface, and also to fill the interstices between the fibers in the paper, so as to give the back surface of the paper a firmer character than the front. Then after flattening in a press, I apply British gum or other adhesive layer on the back of the starch layer, and having again pressed the sheet of stamps, they are ready for shipment and used like ordinary stamps." An o. g. block of six of the 3¢ 'starch paper' printing sold as lot 1501 in the Clarence Eagle I sale.

A third characteristic paper associated with the Continentals is ribbed paper. A hard, thick paper with either vertical or horizontal ribbing was introduced by late 1875 for a second printing the new departmental stamps where it is commonly found. The paper is horizontally ribbed with forty ribs to the inch, which can be seen when the stamp is viewed by oblique light . It is less commonly found on the regular 1, 2, 3, 6, 7, 10, 12, 15, 24, 30 and 90¢ Continental values as well as the 2¢ vermilion and 5¢ Taylor. A block of four of the 15¢ ribbed paper sold as lot 1554 in the Eagle part I sale.

Some of the Continental printings are on thin paper. The Scott Specialized finesses the thin paper Continental problem by noting the stamps are on 'White Wove Paper, Thin to Thick,' although not all values are found in thin paper. In private discussions, Eliot Landau tells me there is a continuum of paper thicknesses and that he and his study group feel the thin papers were the result of a bad batch of paper.

Created by passing the paper either between grooved steel rollers or felt rollers with that have heavy warped threads, it is philatelically significant because it was used as an identification test for the only 24¢ Continental to be certified. Introduced in 1873, its use was most extensive in 1874 and ended in 1875 according to Crawford Cappen, who studied it. He noted the ribbed paper stamps have a richness of color, clear impressions and a high finish or sheen. The Justice departmentals are notable for not only show ribbing, from the paper used, but also have vertical 'false ribbing' from the wove of the paper web.

The Willcox paper mills, which apparently was an important supplier of Continental paper seems to have had a penchant for promotion pink papers. The Franklin carrier of 1851 was on a colored through rose paper, while 'pinkish laid paper' essays are known on the 1861-67 issue as seen in White's illustrations 7-5 and 7-6 of his Papers and Gums book. Later we find the 5¢ Taylor of 1875 is known on a pinkish laid paper.

The indigo Garfield is also known on a pinkish paper. Luff noted that H. C. Conant had reported a collector purchased a lot of 10-15 examples at the Portland, ME postoffice in 1889. A block of four with full original gum sold as lot 268 sold in the Col. Green VII (Storrow) sale at Morgenthau December 8, 1942. Mr. Conant's pink paper copy is part of the Luff reference collection at the Philatelic Foundation, where it can be seen today. It is about as pink as the Washington-Franklin blue papers are blue. (See later under the Washington-Franklin 426 'pink back' variety for additional comment.)

A basic shift of stamp papers occurred between the hard papers used up to just prior to July 1875 and the generally recognized American soft papers of 1882. This shift has caused considerable difficulty for collectors and is best understood by examining a few cheap stamps. The National and early Continental hard paper is typified by the 3¢ 1869 locomotive stamp. The early Continental hard papers come with both strong bleaching and mild bleaching, which can be differentiated by the degree of whiteness. These apparently merely reflect different paper orders.

Continental shifted to an intermediate paper just prior to the first production of the Taylor adhesive, so that about 10% of the Scott #179 production is on hard paper and the balance on the new intermediate paper. For collector reference the average Scott #185 paper is a convenient representative of what intermediate paper should look like. With less filler or sizing used, intaglio ink has a tendency to be more absorbed by the fibres, yielding a softer, duller or blurred image. This softer, duller look gives the 30¢ Continental its 'greenish' hue and the 15¢ a more yellow look.
Beginning in August 1878, Continental shifted entirely to a soft porous (open weave) bleached paper.

The same paper stock was used for the first American paper printings. Consequently, what are termed 'American' soft paper stamps can be found on covers used before February 3, 1879, at which date the Continental and American Banknote Companies consolidated. The last Continental imprinted plate was #310 of the 3¢, while plate 311 of the same stamp was given the American imprint. As a reference of the first American papers, the low value postage dues are useful. For other stamps it is probable that stamp paper, unfinished stamps and finished stamps were in inventory and released after the consolidation. It wasn't until June or July 1879 that the old stocks were depleted . Mr. Landau's exhibit page showing these paper shifts can be seen here.

The New Newsprint Part Wood-pulp Papers

A new paper was introduced in June-July 1879, the so-called soft unbleached American newsprint paper. John Tiffany gave a contemporary commentary in his book, "The paper provided for in this contract (the 1885 contract) is the soft porous paper, which according to Mr. (.E. B.) Sterling was introduced in 1883. It is not stiff and hard like the previous paper, and seems to have been adopted about the time of the change to the dies in the fall of 1882. All values employed since are to be found on it."

American ordered this soft unbleached newsprint quality paper similar to the bleached soft Continental paper, which begins to be used in mid-1879 and continues to 1894. By 1882 both a lightened version with mild bleach and an unbleached version are known. The latter is also found on the low value Colombians.

An inexpensive reference example would be the two cent red brown stamp (Scott 210) or the yellow-brown Garfield (Scott 185). It has a straw or ivory color and a much more open weave with thicker fibres. Like the soft Continental paper there is no real 'snap' when the stamp is flicked and the design does not show through as with the National and Continental hard papers, which are 'hard' not because they are thin but rather because they are heavily sized.

As the open weave permits the printing ink to give a caste to the back of the unbleached stamp paper, Mr. Landau advised it is important to view samples of different inks such as the gray blue of Scott #206, the blue green of Scott #207, the ultramarine of Scott #212 and the green of Scott #213 in addition to the red-brown of Scott #210 and their influence on the backs of the stamp paper. Damaged examples of all these values can be obtained relatively cheaply as reference copies to illustrate how printing ink affects the appearance of the stamp paper.

A new double paper experiment was introduced with the new reengraved issues of 1881-1882. The 1¢ and 3¢ values were used for the Douglas patent experiment. The thin upper paper was punched with circles of eight holes, each about 1½mm while the lower paper was a thicker hard paper. It is believed about 10,000 stamps ordered September 27, 1881 were issued through the Washington, D.C. postoffice. The experiment was deemed unsuccessful and discontinued. I have photocopy of a full pane of these from plate 367.

While earlier contracts either ignored paper specifications or specified paper as per sample attached, the 1885 stamp contract for the first time gave paper specifications. Willard in his Two Cent Red Brown book quotes the specifications, as follows:
"Third. That the paper upon which the stamps are to be printed shall weight nineteen pounds per 1,000 sheets (a heavier weight than previously noted), in dimensions 12½ by 19¾ inches per sheet; uniform thickness, and the same in quality, sizing, finish, and composed of the same materials, and in the same proportions, and of the same tensile strength, as the paper samples of which are hereto attached and made a part of this contract-the paper to be subject to the approval of the Postmaster General or his duly authorized agent."

On page 53 of his Volume I, Willard illustrates a block of the 2¢ red brown on laid paper with a watermark '& Co.' remarking the ampersand is in an old fashioned style. This is a Willcox & Co. watermark. On this page he also illustrates a block of four example from plate 557 with heavy horizontal laid markings and a single vertical line (batonné), which supports the example found as lot 2002 in the 1890s DeCoppet sale that Luff noted. The shade and plate number date it to late 1886 or early 1887. He also shows a laid example with the batonné line but it is vertically laid and has a watermark, which he describes as an eagle in a half circle. I have not identified the mill that used this watermark. Other watermark experimental essays found on this stamp are the honeycomb thick paper and a wavy line watermark similar to the 1875 issue of Bavaria.

The requirement that both hand press and steam press stamps were to be printed suggests that different paper requirements would be needed for each of them. We do know that the samples for the steam press stamps are the 'special printing Scott 211B' items as sheets of them were attached to the contracts. The Scott Specialized indicates that the total was 1,000 stamps, but there is some evidence the total was higher. The true 211B special printing remains unidentified today and only 55 copies ever reached the market. For the remainder of the banknote-printing era, both hand press and steam press stamps were issued, with most production being the steam press items. The same porous newsprint unbleached paper without watermark was used through the end of the banknote period.

The First Bureau Papers

When the Bureau of Engraving and Printing took over the postage stamp contract in 1894, an order was sent to the American Banknote Company dated March 7, 1894 requiring, "25 sheets of blank paper of each of the three sizes in use, 75 sheets. And a sample sheet of each denomination and kind of stamps now used, thus:

1st Printed only.
2nd. Printed and gummed.
3rd. Printed, gummed and perforated.
Newspaper and Periodical Stamps: 25 plates, 3 sheets of each as above, 75 sheets per sheet. 7,500 stamps
Postage Due stamps: 7 plates, 3 sheets of each as above, 21 sheets, 200 stamps per sheet. 4,200 stamps
Regular postage stamps: 11 plates, 3 sheets of each as above, 33 sheets: 1 and 2 cents: 3 sheets of each, 6 sheets, 400 stamps per sheet. 2,400 stamps
3, 4, 5, 6, 8 10, 15, 30, 50 and 90 cents: 3 sheets each, 27 sheets, 200 stamps per sheet. 5,400 stamps
Making a total of regular issue of 1890 of 7,800 stamps
Special Delivery stamp: 1 plate, 3 sheets as above, 100 stamps per sheet. 300 stamps
Total Number of Stamps 19,000 stamps
And l sheet from plate (full size) on cardboard from each of the 44 plates as above."

This order was a follow-up of the new contract of February 21, 1894 to the Bureau of Engraving and Printing, which was to take place for four years beginning July 1st. The sheets were turned over to the Bureau as models for color, paper, etc. The blank sheets were subsequently used for printing stamps, the finished stamps were turned into stock and issued, the ungummed un-perforated sheets were destroyed, while the printed and gummed sheets disposition is unknown, save that one part sheet of the newspaper stamps (50 stamps) passed into private hands. Their paper is very white, fine, close and without trace of weave or grain. The 9¢ Newspaper American printing is only known from this printing.

These sample papers were supposed to be the models for the Bureau papers. They weren't quite. The new Bureau paper was less porous and closer grained although of the same thick white soft newsprint style as the American papers. The first stamp plate produced by the Bureau was the un-watermarked 2¢ pink Scott 248 with the Type I triangle, plate #1. However, enough plates were used to print this stamp that a collector could not readily tell whether the one he was looking at was an early printing or not, so that the 6¢ Garfield (Scott 256) printed on plate 28 was the first of the new Bureaus issued (released July 18, 1894 according to the list in the 1895 United States Postal Guide and known August 11th) or the 2¢ new postage due designs in vermilion (J30) or deep claret (J32) printed on plate 34 and released July 20th are probably better choices to show the first Bureau paper differences from the last American printing (the Columbian issue). The first Bureau Newspaper stamps of 1894 were also printed on a thick soft newsprint paper much like that used by the American Banknote Company, but closer grained and less porous.

The Bureau also introduced double-lined USPS watermarked paper in 1895. The first issued was the 1¢ Franklin (Scott 264) released April 19, 1895 with an EKU of May 16th, the 1¢ Postage Due (Scott J38) released August 24th and the 10¢ Newspaper (Scott PR117) released September 13, 1895. The Bureau had long used the USIR watermarked paper for revenue paper, so there was no significant change in adding watermarked paper for postage stamps. A few postage stamps were printed on the wrong watermarked paper so that the 6¢ and 8¢ values are known on USIR paper. This was because these two values and the high values were printed on hand presses as were the playing card revenues and there was a paper mix-up.

With the shift to the American Banknote Company, wood pulp papers become standard. They start off with about 20% soda/sulfite and 80% rag with the first value printed-the 3¢ Scott #184-and rise to 25% soda process wood pulp by the 3¢ reengraved Scott #207 and the rag content drops to 60% by the 1890 issue one cent and to 40% during the 2¢ mass printing. In both cases sulfite process wood pulp is 20 to 25% with soda process wood pulp at 15 to 35%. By April 1890 the soda and sulfite papers account for 40% split equally. By 1894, the conversion was complete and rag paper was no longer used.

The Experimental Bureau Papers of 1908-12

The last of what might be called the classic stamp paper experiments were made around 1908-1912. At this time we get the bluish and chalky papers, the China Clays, the 30% China clays, the yellow-browns, the grays and thick, hard papers. These experimental papers were part of a larger Bureau program to speed production and improve product quality and cut costs which ranged from a washing and resizing used currency to a new solution to wash rags used to wipe printing plates (savings $6,500 a year) and an automatic paper dampener (savings $42,000 annually). It also involved the introduction of the rotary press and different laydown spaces on the plates (the star plates) designed to prevent curling and different perforation spacing tests to help separate coil stamps without tearing (the 1910 coil perforations). There were even tests of different colored papers.

The best known of the paper experiments were the bluish papers (Scott 357-366) which were made with 35% rag stock rather than 100% wood pulp paper stock then in general use. This was basically a reversion to the combination paper stocks used to print some of the American Banknote stamps such as the 2¢ lake of 1890. They have a light bluish-gray cast, which used to be more troublesome to identify before the high quality auction catalog color pictures of today which show the shade clearly.

The objective was to reduce paper shrinkage so as to reduce misregistration as well as to reduce paper curling and breaking at the perforations in the postoffice drawers. The Bureau paper supplier was the firm of Jessup & Moore had been experimenting with mixed pulp paper stock and on November 20, 1908, Bureau chief, J. E. Ralph, wrote the Treasury Secretary asking permission to supply the firm a watermarking dandy roll so that an experiment could be made with the part-rag paper. On November 24th a letter from Ralph stated,

"…The Jessup and Moore Paper Company has been directed to forward to your office 10,000 sheets of 18½ x 20¾ 34 pound paper for postage stamps to be used by this Bureau for experimental purposes."

This was the paper used for the bluish paper stamps. It was a sufficient quantity to print 4-million stamps, which included the 1908 experiment and the 1909 archival printing. Roy White did tests of examples and found the rag stock was only 10% rather than the 35% reported in the Scott Specialized . It might be noted that the 34-pound weight was a heavier stock than used in earlier years, a point I don't recall noted anywhere else. There were two printings of the bluish papers, one experimental and the later second an archival printing.

Based on the dates that various plates went to press, it seems likely that the one and two cent Washington-Franklins went to press on November 30th or December 1st in 1908 as one plate was only on press at that time. The experimental printing was released in early February 1909, following a letter of that date by Ralph asking permission to deliver the already-printed 1¢ and 2¢ bluish papers through regular orders for the color, "variation was not sufficient to justify destroying the printed stamps." On February 16, 1909, Third Assistant Postmaster General A. L. Lawshe reported the stamps were to be sent on requisition to the Washington, D.C. postoffice and the Washington postmaster was to report, "whether they less tendency to curl than the stamps of the regular wood pulp paper and are found otherwise satisfactory."

He also instructed the postmaster to have the stamp clerks observe, "whether the greater strength of the paper facilitates the separating and minimizes the tendency to tear away from the line of perforation."

On or about the same February 1st date a decision was made to add the 2¢ Lincoln to the experiment and the Lincoln plates first went to press February 5th. These solid star plates were apparently ready January 22nd. The Annual Report for the year ending June 30, 1909 reports 1,480,000 1¢ stamps (14,800 panes of 100 out of a printing of 16,020 panes or 4005 sheets of 400) and 1,494,000 2¢ stamps (14,940 panes out of 16,020 printed) and 637,000 2¢ Lincoln stamps (6,370 panes out of 7,016 panes printed). The Washington postoffice sold out by May 1st. The remaining panes were either spoiled or destroyed.

The 1909 archival printing made from different plates than the experimental printing of the 1 and 2¢ Washington-Franklin and the 2¢ Lincoln, created a scandal and deprived philately of important records for a century. At the time of the experimental printing, Third Assistant Postmaster Arthur Travers requested that printings be made of all values for the Department files and the Museum collection before the plates showed wear. Mr. Ferguson, Ralph's assistant, was instructed to prepare perfect sheets of each value as soon as the plates went to press omitting the 50¢ and $1 values. These were omitted as they only occasionally went to press.

Four panes (one sheet) of each value were to be printed, with one going to the Postal Museum and one into the archives with the other two to be destroyed. The actual printings ranged from 10-13 sheets each In January 1910, some of the archival printing panes got out and Eustace B. Power, head of Stanley Gibbons, bought two panes of 100 at the City Hall postoffice in New York. The result was that well-known collector Joseph Steinmetz, whose Steinmetz Miscellany was an important exhibit in the 1913 International Philatelic Exhibition, bribed Travers to arrange for the bluish paper stamps to be available at a certain time at the Washington postoffice and bought them. He then offered them for sale and Philip Ward, miffed that Steinmetz had sold blue papers to his client Henry C. Gibson, alerted postal inspectors to the irregular dealings.

On March 6, 1911, Travers was dismissed and he was indicted April 3rd. Another indictment named both him and Steinmetz but after a year and a half of legal maneuvers Travers pleaded nole contendere to the charges and Steinmetz's name was dropped, possibly because of political influence.

The postal history loss of the bluish paper scandal resulted from the fact that Third Assistant Postmaster General Edwin Madden had instituted a project to save and write up original documents from the postoffice archives dealing with stamp production from 1847 until the 1890s. His confidential assistant was Arthur Travers, who ended up with the documentation, which was basically suppressed at the time of the bluish paper scandal except for a few bits presented in the Steinmetz Miscellany and a few comments from those who had access to the files. The basic material is only now being prepared for publication.

The second paper variety is one that only recently received catalog status, and possibly shouldn't have-- the clay papers, which are the 'b' listings under the regular catalog numbers, Scott 331b, 332b etc. According to Roy White, these were a printing with a slightly higher China clay (kaolin) content than the regular issue (about 10-18% vs. about 5-8%). The whole question arose because of a memorandum dated July 30, 1908 addressed to the Eastern Manufacturing Co. of Bangor, ME discussing several sheets of experimental postage stamp paper that had a 33% ash content, which was assumed to represent China clay in the wood-pulp paper, and the paper was supposed to have been used in a Washington-Franklin paper experiment, the China clay papers. However, this attribution has to be wrong for the first Washington-Franklin essay was dated September 24, 1908 and the die approved October 17th, so that any experimental printing referred to in the July 30th letter had to be for the preceding 1902 issue. No such high ash paper has yet been recorded as an experiment on that issue.

This paper was first recorded in the February 1, 1910 Philatelic Journal of America when a 5¢ value from star plate 5376 was reported. This was not a plate used for the bluish papers. It was certified August 12, 1909 and first sent to press November 4th of that year. A certified 15¢ China clay paper registered cover addressed to Malta is postmarked at the Somerville branch of the Boston post office on March 22, 1909. Clay paper examples have been certified on cover at least as early as March 9, 1909, which is later than the earliest currently certified bluish paper example (2/33/09), although the China clays presumably went to press the preceding fall.

As part of the A.I.E.P. expertizing round table in New York on November 14, 2001, a presentation was made showing that the specific gravities of the clay papers fell into the same range as the regular issue, but that high magnification showed that the paper fibres had incrustations of dirt, which apparently was the result of a flood that affected the paper company's water supply at the time of the production of these stamps, so that their slight difference in appearance is the result of using contaminated or dirty water in the 'stuff' and was not part of an experiment at all! However, at least one expert present suggests that the time line between the flood and the appearance of used copies is too short for this to be a full explanation.

The regular, bluish and China clay papers are shown as sets from the Zollner sale, used for consistent photographic color comparison purposes rather than the Wade Saadi set with which I originally worked. However, the certified but not Scott recognized Saadi examples of the 5¢ and 10¢ examples of experimental brownish paper are shown here as his certified 15¢ chalky paper example and the block from which it came. The thick, almost card, thick experimental paper is shown on a variety of Scott 333 that is known from plate 4925. Finally, an example of Scott 426, the 3¢ reddish violet analine dye variety known as a 'pink back' is shown. The 'pink backs' are Scott cataloged on the 2¢ rose red Scott 425 and the 12¢ copper red Scott 435a as well as some Victoria era issues from the British West Indies suggest that some aniline ink formulations may be responsible for the earlier noted 5¢ Taylor and indigo Garfield 'pink paper' varieties.

One of the first observers to comment on the experimental papers of this period was P. W. Greyer, who described five papers in the July 1912 Philatelic Journal of America. First was the normal paper, which he noted had wove marks, porosity and unevenness of texture. Next was a gray paper, which was very porous, even in texture, with large even plain wove markings and which was decidedly gray in tone. It is not currently Scott listed. This was not an experimental paper but found in regular distribution although scarcer than the bluish papers, which he noted were decidedly gray or gray blue in color but slightly darker than the gray paper. There was also a dense cream-toned paper, which was very dense and even in texture with no wove marks. It was far heavier than any of the other papers, but was the most readily found of the experimental papers. This paper is also not Scott listed.

His fifth paper was a very white paper that was heavier than all the others except the cream paper. It is fairly even in texture but not as even as the gray and had small plain wove markings. It became the normal paper at a later stage and is found regularly with the single line watermarks introduced circa October 1, 1910, but is also found with the earlier double-lined watermark. Steinmetz recorded this paper under his pseudonym of 'Post-Officious' in the September 9, 1910 Philadelphia Stamp News when he reported, "a thicker paper is now being used by the Bureau now, which is proving better for the perforating machines…a thick hard, similar to bond, paper."

Mr. Landau reported that the imperforate 2¢ Lincoln on a very thick dead-white bond paper, which is only seen on the imperforates. It does not seem to be this paper. He also recorded two thin paper varieties one of which is a currency paper, probably used by accident, which shows additional non-colored fibres similar to the 'silk' colored fibre paper, as well as an ivory paper version found on the 2¢ Lincoln printed from plates 4977 and 4979.

Three other experimental papers are known, but are not Scott listed. The first is a yellowish, quite brittle high mineral content paper. This paper is apparently the paper reported by Acting Third Postmaster Travers in the February 1, 1910 issue of the Philatelic Journal of America, who followed a comment that the 10¢ Washington-Franklin would be printed in brown ink on yellow paper, with the statement that it was likely some of the higher values would be printed on other than white paper. On January 28, 1910, Travers wrote the Stamp Security Company of St. Louis that,
"You are advised that the postmaster at Saint Louis will receive the new brown 10¢ stamps on colored paper in regular course when they can be purchased at the postoffice."

This paper is currently not Scott listed. Examples are known on the 5, 6, 8 and 13¢ values reportedly obtained from the Bureau archives, and all but the 13¢ had a 15-20% mineral content while it had an 8% mineral content. Physical examination of the 5¢ and 10¢ examples with Philatelic Foundation certificates show that they have a brownish or yellowish look.

A second experimental non-Scott listed paper is the chalky paper Washington-Franklin printing, apparently the only time this was done on an American stamp. This is a surface coated paper, which was tested to see if it would help in the drying process and reduce paper shrinkage. The Philatelic Foundation certified a 15¢ value while W.L. Babcock reported in 1940 that late general Charles A. Coolidge, past president of the APS, owned two blocks of four of the 8¢ value.

This chalky paper was first recorded in the June 1910 Philatelic Journal of America. Physical examination of the 15¢ value, which I examined shows it has a whitish appearance on the face and under ultraviolet light appears darker than the normal papers. It would be rarely seen, for soaking would remove the surface chalky wash on used examples.

The third non-Scott listed experimental paper group is that of colored papers. These do not appear to be the yellowish-brown papers citied earlier for the 10¢ value; those are recorded later. Philip Ward reported the colored appeared about the time of the 12¢ Washington-Franklin essay. (This value was announced September 18, 1909 and September 30th, with the wash-drawing essays dated between October 7 and November 24, 1908.) Mr. Ward reported colored paper essays on the 2, 8, 10, 12, 15 and 50¢ Washington-Franklin values. Except for the 12¢, these are not listed in Scott and most are not listed in Brazer as well. One reason why this colored paper experiment was not followed up upon was the fact that the paper mill contractors did not care to produce colored paper except in 10,000-pound lots. With a typical 16-18 pound paper stock, this would imply a printing stock equal to up to double the bluish paper experiment.