The Paper Machine
     And Using A Dandy Roll To Create Watermarked Paper

   The first paper making machine was invented by the Frenchman, Nicholas-Louis Robert and patented on 9 September 1798. He claimed that it was the constant quarrelling between the workers of the papermakers' guild that motivated him to attempt to create a machine that would replace hand labor. While in his thirties, Nicholas-Louis Robert acquired, first, a job at the Paris publishing firm operated by St. Leger Didot, and later, a job at the Francois Didot paper mill at Essonnes. The Didot mill had been in business since the year 1355, and produced much of the paper used by the Minister of France for the nation's currency. In 1797 Robert produced a paper machine model, but as with many inventions, the initial trial of the machine proved it to be a failure. With his employer's encouragement, Robert continued to work on the machine, and either improved it, or built an entirely new model. The second trial was no more effective than the first. Eventually Robert persevered, and produced a full-scale paper machine. When this success became known and his previous employer, St. Leger Didot saw the quality of the paper that could be produced on the machine, he urged Robert to apply for a patent. The French Bureau of Arts and Trades were excited about the new machine and delared that Citizen Robert is the first to imagine a machine capable of making paper from the vat; this machine forms paper of great width and of indefinite length. The machine makes paper of perfect quality in thickness and gives advantages that cannot be derived from ordinary methods of forming paper by hand, where each sheet is limited in size in comparison with those made on this machine. From all reports it is an entirely new invention and deserves every encouragement. The principle by which Nicholas-Louis Robert's paper machine functioned is the same principle that modern papermaking machines are based upon: the paper is formed on an endless woven-wire cloth which retains the matted pulp fibers while at the same time allowing the superfluous water to drain through the mesh of the woven wire material.

   Due to some disagreements between Robert and his sponsors, St. Leger and Francois Didot, and also partly due to the state of affairs brought about by the French Revolution, progress on improvements to the paper machine came nearly to a standstill. In 1799 Francois Didot contacted his brother-in-law, John Gamble to ask if he could raise sufficient funds in England to manufacture a paper machine similar to the Robert machine. Gamble made contact with Henry and Sealy Fourdrinier, two stationers working in the firm of Bloxham & Fourdrinier, located in London, and they expressed great interest along with the agreement to help fund the venture. On 20 April 1801 John Gamble was granted the first English patent on a paper machine. It was about that time that John Hall, an engineer at Dartford, was asked to examine the original machine built by Robert. Didot brought the machine to England, and he, Hall and Gamble began to study the machine with the intention of improving it. Bryan Donkin, a mechanic who had apprenticed to Hall, was then engaged to construct a paper machine which was practically identical to Robert's. John Hall pulled out of the project shortly thereafter, but Donkin continued with it and was given space to work on the project at Bermondsey by the Fourdriniers. A first paper machine was installed in 1803 at Frogmore, but Donkin continued to improve upon it.

   The Fourdrinier paper machine was patented on 24 July 1806 by Henry Fourdrinier, and a year later, on 14 August 1807 additional improvements were patented by Henry and his brother Sealy. The Fourdrinier paper machine, consisting of a number of paper moulds hooked together to form a cylinder of sorts to form the paper, and a series of other cylinders through which the forming paper would be passed on "an endless web of felting" in order to dry it, making possible the production of continuous long lengths of paper, was essentially the Robert paper machine with certain improvements. The original Fourdrinier paper machine was housed in the Two Waters Paper Mill in Hertfordshire, operated by Marchant Warrell.

   The success of the paper machine improved by the Fourdrinier brothers, Donkin and Gamble caught the imagination of others, and in 1807 Thomas Cobb received patents for additional improvements to the process. Then, in 1810, John Gamble and Bryan Donkin contributed to the project with more improvements. As a result of these individuals' work, the Fourdrinier paper machine was operating on a commercial basis by 13 January 1812.

   Not everyone, though, was pleased by the introduction of the new machine. Handmade-paper workers rightly feared that the new machine would cost many of them their jobs if it caught on throughout the world. Riots apparently were staged outside the Two Waters mill, and the windows had to be boarded up to prevent the malcontents from enterring and causing damage to the machine.

   Nicholas-Louis Robert and the Fourdrinier brothers were not the only ones inventing for the sake of papermaking in the early 1800s. In 1809 John Dickinson received a patent for the cylinder paper machine. The difference between this new type of paper machine and the original as designed by Robert lay in the way in which the web of paper was formed. A cylinder, covered with a woven wire fabric, was half immersed in the pulp vat. Inside the cylinder a vacuum was installed so that it would draw the pulp to the surface of the fabric which encassed the cylinder. As the paper was thusly formed, it would be detached and passed on, by means of a felt, to other cylinders through which, and over which it would pass, drying as it went.

   Between 1803 and 1812 ten paper machines were constructed in England, and between 1812 and 1823 another twenty-five were built. Also, English built paper machines were exported to France, where the first was set up at Sorel in 1811. By 1833 there would be approximately twelve machines in operation throughout France.

   All of the mechanized paper machines operated on a similar set of principles which insured that the fibers of the pulp would mesh together properly. Carefully screened pulp stock (i.e. cleaned and bleached pulp fibers suspended in water) would be supplied to the wet end of the machine. The stock would be pumped into a headbox where it was kept agitated in order to keep the fibers from clumping together. From the headbox, the pulp would be sprayed out across the forming wire, an endless fabric belt. At that initial point of contact with the forward moving forming wire, the pulp would be shaken laterally, in order to cause a portion of the pulp fibers to alter their direction, and effectively create a mat or web of paper. The pulp stock, now more properly known as the paper stock, being carried forward by the forming wire, passed between suction roll and a suction box which would suction out between 5% and 15% of water from the paper stock. The suction roll was more commonly called the suction couch roll, because it would 'couch', or 'coax', the forming paper off the forming wire and onto a woolen felt or belt by sucking water from it. The word 'couch' came from the French, meaning to de-water something. In this case, the suction couch roll de-watered the paper stock by suctioning off the water by means of a vacuum within the roll itself, the water being permitted to pass through a series of holes in the surface of the roll. The removal of the small percentage of water allowed the web of paper to release its grip on the fabric of the forming wire. The web of forming paper was then passed onto a felt fabric belt which would carry it between a number of presses and dryers to remove the remaining water content.

   The impact of the paper machine on watermarking was both a negative and a positive ~ the negative being that a different way to form the watermark would have to be devised, since the paper would not be formed in the confines of the mould, to which the papermark was previously attached; and the positive being that the positioning of the watermark could now be more creatively modified. All-over or repetitive designs would be more easily accomplished ~ previously, to achieve an all-over design, the papermaker would have had to fabricate numerous similar shaped wire papermarks or electrotyped plates to be attached to the fabric of the paper mould; now as the cylindrical roll onto which the papermark would be attached repeatedly made contact with the forming paper, the watermark would be automatically repeated.

   The positive advantages of the paper machine over the hand mould certainly outweighed the negative disadvantages. The problem of devising a new way to form the watermark was solved by the creation of the dandy roll.

   It should be noted that the dandy roll was initially invented for the purpose of smoothing the forming web of paper on the fourdrinier paper machines. The attachment of papermarks onto the surface of the dandy roll to form watermarked paper was a (perhaps) unforeseen benefit of the installation of a dandy roll.

   The invention of the so-called, dandy roll is usually attributed to John Marshall, of the firm T.J. Marshall of London, which had been established in the year 1792. But Mr. Marshall did not file for a patent on his invention. The records of the Great Seal Patent Office include a patent granted to John Phipps and Christopher Phipps, dated 11 January, 1825 for "An improvement in machinery for making paper by employing a roller the cylindrical part of which is formed of 'laid' wire. The effect produced by the said cylindrical roller is that of making an impression upon the sheet of paper, or pulp, upon which the said roller passes & thus the paper so made has the appearance of 'laid' paper (like that manufactured by hand).".

   Apparently, neither John nor Christopher Phipps pursued the construction of their rollers. They passed into relative obscurity, and the patent is essentially the only notice we have of them at this time. On the other hand, John Marshall, in 1826, proceeded with the production of a similar roller, the name of which it is said was bestowed upon the roller when a worker at the Marshall establishment exclaimed: "Isn't that a dandy!".

   The original dandy roll consisted of a framework of lengthwise wooden ribs onto which round metal discs were attached in order to loosely support a woven wire fabric cover. Later dandy rolls consisted entirely of metal. A series of parallel 'ribs' would be held in place by a large-mesh metal fabric. The metal of choice for these later dandy rolls was steel because of its strength and durability. The framework and mesh cover would be constructed so as to impart either a 'laid' or a 'wove' pattern onto the paper irregardless of, or in addition to, the watermark. Onto the woven wire fabric cover the papermark(s) would be attached, initially by tying them on with fine wire, as in the earlier handmade paper moulds, and later by spot welding/soldering. The use of solder to attach the papermarks to the fabric cover came into vogue around 1870.

   A major difference in the use of the dandy roll as compared to hand paper moulds, in regard to watermarks, is that in making paper by hand, the paper pulp is placed over the papermark design, so that the watermark is impressed into the paper as the paper formation occurs, and the water is caused to drain downward through the fabric of the mould. Apart from the drainage of the water through the paper mould, movement of the pulp is static and the papermark remains in contact with the same pulp of the forming paper. In making paper on a machine, the dandy roll comes in contact with the paper pulp from above the forming web of paper as the paper formation begins to occur, and the paper is just starting to be dried. The movement of the pulp and its contact with the papermark on the dandy roll is momentary and fleeting due to the necessary speed of the paper machine; and because of that, scholars, such as Dard Hunter, have noted that the watermarks produced in handmade papers will always be more distinct and well-defined than those produced on a paper machine. The paper formed in a hand mould will not be influenced by the shrinkage or stretching that occurs on a paper machine.

   The illustration below shows a simplified configuration of a modern paper machine on which a dandy roll has been installed. It is generally positioned above the forming wire and between the headbox and the suction roll(s). The paper pulp is deposited onto the forming wire from the headbox as the forming wire travels around a pair of rollers. Just before the suction roll causes a large percentage of the water in the pulp deposited on the forming wire to be suctioned off, the watermark pressed fabric enveloping the dandy roll will impress its design of high and low areas into the pulp. The layer of design impressed pulp leaves the forming wire as already-forming paper, and is carried through the dryers to the press rolls on dryer fabric that snakes through the many dryers and rolls.

   The dandy roll brought with it certain requirements that had not been necessary with hand paper moulds. One of these requirements was the alteration of the papermark design to accomodate the stretching of the paper as it passed in contact with the dandy roll. A circle, formed in a single wire design would appear as a slightly egg shaped watermark in the paper. According to Dard Hunter, a circle with a four-inch diameter circle would have to be distorted by about three-eighths of an inch to produce a non-distorted watermark image.

   By the year 1845, there were only two paper mills in the entire United States of America producing hand-made paper. This is a remarkable example of the industrial revolution in view of the fact that there were hundreds of paper mills in operation throughout the U.S.; there were eighty-nine in the small state of Massachusetts alone.

   In 1866, the Willcox Mill, established in 1729 by Thomas Willcox in Chester County, Pennsylvania, discontinued making paper by hand. It was the last one to do so in the United States.

   Initially, in the use of a dandy roll for watermarking paper produced in a fourdrinier type paper machine, the attachment of the papermark to the roll, as noted above, was acomplished by tying it on with fine wire, or by spot welding. Later on, it was discovered that the raised and depressed areas that would result in a watermark in the paper could be achieved by pressing the relief directly into a fine-mesh brass-wire fabric, which would then be attached to the dandy roll's circumference.

   Instead of forming (by hand) hundreds of individual papermarks, all of which would have to be attached to the dandy roll, the newer method of pressing a relief design into a wire fabric, which would then encircle the dandy roll, was accomplished by creating a single set of mated die and matrix plates bearing the relief design, which would be used to press the design in a repeat pattern onto the cover fabric.

   At first, the die plate (commonly refered to as the female plate) and its mated matrix plate (or male plate) were created by being carved and/or formed in wax, clay or wood, and then cast in molds. In more recent years, with the advent of computers and specialized 3D modeling software programs, two-dimensional artwork could be created and then converted into three dimensional data that would later be transmitted to a CDC (i.e. custom design cutting) engraving machine. The 3D modeling program would not only be used to create / simulate a three dimensional image of a two dimensional artwork, but also to generate the series of toolpaths that would direct the CDC machine to perform specific cuts using specific engraving bits.

   A CDC machine functions by cutting away portions of solid metal plates similar to the manner of engraving, but at much higher speeds than would be possible by human hands. The solid metal plates would be composed of either steel or bronze so that the softer brass wire mesh fabric could be pressed between the die and matrix set many times without degradation.

   After the die and matrix plates were prepared, they would be positioned at intervals across the entire piece of wire mesh fabric so that when the piece of fabric was placed around the circumference of the dandy roll, and its two ends met and were welded together, the design would meet correctly and seamlessly. At each position, the plates would be set, the wire mesh fabric would be annealed (i.e. heated to become supple), and the two plates pressed together (preferably by using an hydraulic press), causing the metal wire strands of the fabric to be pushed and pulled to mimic the heights and depths of the engraved areas. In cases where an overall pattern was intended, the plates would be moved and set, the fabric would be annealed, and then pressed, and this routine would be repeated across the entire surface. In other cases, where solitary, separated watermarks were intended, they would sometimes be pressed into smaller pieces of fabric, called patches, the intended position on the full fabric would be cut out, and the individual exposed wires around the edges of a patch would be matched and soldered to coincident individual wires exposed around the edges of the cuttout. Either method of applying the relief of the watermark to the wire mesh fabric was very time consuming and tedious work.

   The final step in the process of creating a dandy roll to be used for making watermarked paper was to wrap the completely pressed or patch covered bronze wire mesh fabric around the circumference of the roll, match the loose wires of the fabric's two ends together, and then to seam, or rather weld or solder them together. Welding the ends of the fabric would provide for a more durable result and was preferable to soldering if the dandy roll was destined for extended use. The covered dandy roll would then be placed over the forming wire on the wet end of the paper machine so that it could make contact with the pulp forming the web of paper.