Skip to content

Chester Carlson

Abstract

Chester Carlson (1906–1968) invented xerography, the dry electrostatic copying process inside every photocopier and laser printer. He worked it out alone, at night, in a rented room in Astoria, Queens, while holding down a patent-department job and studying law, and made the first copy on October 22, 1938. Between 1939 and 1944 more than twenty companies, IBM, General Electric, Kodak and RCA among them, turned it down. A small Rochester photo-paper maker called Haloid took the licence in 1947, spent $75 million and twelve years on it, and in 1959 shipped the Xerox 914, which Fortune called the most successful product ever marketed in America. Carlson earned something like $200 million from it, kept a three-bedroom house and one car, and gave most of the money away anonymously before he died in 1968. The copies his process made cheap pushed copyright law to deal with copying in every library, from a 1973 fair-use ruling in the United States to West Germany’s 1985 levy on copiers.

Xerox 914 Museum
A Xerox 914 (1959), the first plain-paper office copier, at the Museum of Business History and Technology, Delaware. Image: Marcin Wichary, CC BY 2.0, via Wikimedia Commons.

A Childhood on the Move

Carlson was born on February 8, 1906, in Seattle, the only child of Olof Adolph Carlson, an itinerant barber, and Ellen Josephine Hawkins, who had grown up on neighbouring Swedish farms in Grove City, Minnesota. Olof had spinal arthritis and, in his thirties, tuberculosis, and the family spent Chester’s childhood chasing a climate that might help: a brother’s house in California, a camp in the Arizona sand dunes, an adobe hut on a worthless farm in Mexico, a single room in a Los Angeles doctor’s house where Ellen worked as housekeeper, a rented house in San Bernardino. In the autumn of 1915 Olof decided cold might do what heat had not and moved them to a shed in the mountains above San Bernardino, where the snow lay three and four feet deep and Ellen flashed a hand mirror down to a storekeeper in the valley each morning to signal that they had survived the night. Chester was the only pupil in the local school. He later called the year “the beginning of a considerable setback in my social development.”

He worked from the age of eight and was the family’s main earner by high school. At ten his favourite possession was a toy typewriter; at fifteen he started a pocket notebook of inventions (a rotating billboard, a shoe-cleaning machine, a trick safety pin) that he kept for the rest of his life. He worked for a printer, took a discarded press in lieu of wages and printed a magazine for classmates, the Amateur Chemists’ Press. “I don’t think I printed two issues,” he said later, “and they weren’t much. However, this experience did impress me with the difficulty of getting words into hard copy, and this, in turn, started me thinking about duplicating processes.” In his junior year his mother died of tuberculosis at 53, which he called the worst thing that ever happened to him. By graduation he and his father were living in a former chicken coop with a bare concrete floor; Chester slept outside in a sleeping bag he had sewn himself, partly to avoid the disease that had killed her.

He worked his way through three years at Riverside Junior College, then transferred to Caltech, graduating in physics in 1930 while mowing lawns and working at a cement mill, his father cooking in the shared Pasadena apartment. He left owing $1,500. He wrote to 82 firms and received two replies before Bell Telephone Laboratories in New York took him as a research engineer at $35 a week (see Bell Labs: The Idea Factory).

Bell Labs and the Law Library

After a year he moved into Bell’s patent department, on the theory that an inventor should know how patents work. He lost the Bell job in 1933: Xerox’s company history says he was laid off as the Depression deepened, other accounts that he was fired for joining colleagues in a failed side business. He worked briefly for a patent attorney and then joined the electronics firm P. R. Mallory & Co., where he rose to manage the patent department, married Elsa von Mallen in 1934, and in 1936 enrolled in night classes at New York Law School. “I had my job,” he said, “but I didn’t think I was getting ahead very fast. I was just living from hand to mouth, and I had just gotten married. It was kind of a struggle, so I thought the possibility of making an invention might kill two birds with one stone: It would be a chance to do the world some good and also a chance to do myself some good.”

Two frustrations pointed at the same invention. At Mallory there were never enough carbon copies of patent specifications, and getting more meant expensive photostats or retyping and proofreading. At the New York Public Library, where he studied because he could not afford the books, he copied long passages from law texts in longhand until his hand cramped. His notebooks from the 1930s hold more than 400 ideas (a raincoat with gutters, a see-through toothpaste tube, a filter cigarette), but the office copier was the one he pursued. He reasoned that if ordinary photography could be made into a copier, the big photographic companies, Kodak among them, would already have done it, “so I deliberately turned away from the conventional photographic processes and started searching in the library for information about all the different ways in which light will affect matter. I soon came upon photoelectricity and photoconductivity.” He read the Hungarian physicist Pál Selényi’s work on electrostatic images, and by October 18, 1937, had filed his first patent application for what he called electrophotography.

The scheme was this. Coat a grounded metal plate with a photoconductor, a material that insulates in the dark and conducts in the light. Charge the whole surface in darkness. Project the image of a page onto it: where the page is white the light drains the charge to ground; where the page is black the charge stays. Dust the plate with an oppositely charged powder, which clings only to the charged areas, and transfer the powder to paper and fix it. No liquid, no chemical reaction, no treated paper. His first experiments, melting sulfur on the kitchen stove of the couple’s Jackson Heights apartment, produced fires, a smell of rotten eggs, and an angry wife.

10.-22.-38 ASTORIA

In 1938 he rented a room in Astoria, Queens (his biographer David Owen places it behind a beauty parlour; Xerox’s own company history puts it above a bar) and hired Otto Kornei, an unemployed physicist who had emigrated from Austria. Kornei managed what Carlson had not: a clean film of sulfur on zinc plates the size of business cards. Carlson’s own account of October 22, 1938: Kornei “took a glass microscope slide and printed on it in India ink the notation ‘10-22-38 ASTORIA.’ We pulled down the shade to make the room as dark as possible, then he rubbed the sulfur surface vigorously with a handkerchief to apply an electrostatic charge, laid the slide on the surface and placed the combination under a bright incandescent lamp for a few seconds. The slide was then removed and lycopodium powder was sprinkled on the sulfur surface. By gently blowing on the surface, all the loose powder was removed and there was left on the surface a near-perfect duplicate in powder of the notation which had been printed on the glass slide.” They repeated it several times to be sure, pressed the powder onto wax paper, melted the wax to fix it, and went out to lunch. The scrap of wax paper is in the Smithsonian.

Kornei was unimpressed and left for an electronics job in Cleveland. Then Carlson went looking for someone to build the machine.

The Patents

Carlson worked in patents for a living, and he protected the idea before he could make it work well, filing again each time he learned something. Xerox’s company history says he feared others were on the same trail; they were not, and “Carlson was quite alone in his work.” His first issued patent, US 2,221,776, “Electron photography,” filed on September 8, 1938, and granted on November 19, 1940, described a variant in which a photoemissive layer released electrons across an air gap onto an insulating sheet. That was not the version that mattered.

The one that mattered was US 2,297,691, “Electrophotography,” filed on April 4, 1939, and granted on October 6, 1942. It claims the sequence the Astoria test had shown: a thin layer of sulfur, anthracene or a similar insulator on a metal backing, charged by friction, exposed to a light image so that the lit areas discharge, dusted with a fine powder that clings to the charged areas, with the powder image transferred to paper and fixed by heat or an adhesive. Carlson also listed uses beyond office copies, among them masters for lithographic and hectographic printing, which is how Haloid’s first product later earned its keep. A third patent, US 2,357,809, “Electrophotographic apparatus,” filed on November 16, 1940, and granted on September 12, 1944, covered a machine to carry out the steps.

These broad claims gave Battelle and Haloid something to license. The legal scholar Jessica Silbey points out that the patent that brought Carlson the most money was a later one covering an invention in the 914, and that effective competitors appeared only when the patents expired. Rivals had to work around Xerox’s later patents: IBM spent the 1970s developing organic photoconductors to avoid the selenium-drum patents. When the Federal Trade Commission’s consent decree forced Xerox in 1975 to license its copier patents, the portfolio held more than 2,000 of them. Most of the new patenting after the decree came from Japanese firms, which used the freed technology for small, cheap desktop copiers in a segment Xerox had barely served.

Twenty Rejections

From 1939 to 1944 he demonstrated his plates to more than twenty companies, IBM, General Electric, RCA and Eastman Kodak among them, and to the National Inventors Council. “Some were indifferent,” he said, “several expressed mild interest, and one or two were antagonistic. How difficult it was to convince anyone that my tiny plates and rough image held the key to a tremendous new industry. The years went by without a serious nibble. I became discouraged and several times decided to drop the idea completely. But each time I returned to try again.” His phrase for the collective response was “an enthusiastic lack of interest.” The companies had reasons. The demonstration was a smudged powder image on a plate the size of a business card; the market for copies was assumed to be small (as late as the mid-1950s Carlson himself doubted many offices would need a hundred copies a day); and the photographic companies were working along photographic lines, with special chemicals and coated papers. None of them got a second chance. Xerox had xerography to itself until the 1975 consent decree, and the firms that then broke in were Japanese. By the FTC’s count Xerox held 86 percent of the worldwide copier market in 1972; by 1982 its share of copier revenue was 41 percent.

The break came sideways. In 1944 Carlson, in the course of Mallory patent business, handed one of his patents to Russell Dayton, a physicist at the Battelle Memorial Institute, a non-profit research organisation in Columbus, Ohio. He gave his usual demonstration to a half-dozen Battelle staff and braced for the usual throat-clearing, and Dayton held up the wax paper and said, “However crude this may seem, this is the first time any of you has seen a reproduction made without any chemical reaction and by a dry process.” Battelle took 60 percent of any royalties in exchange for doing the development, and allocated $3,000 for 1944. Dayton later estimated that at least half of the Battelle people who knew about the project thought it stupid: “It just goes to prove that if you’ve got something unique, you don’t take a poll.” Battelle’s contribution was real. Roland Michael Schaffert, head of its graphic arts division, wrote the memorandum that got the project approved, and a small group including William E. Bixby, L. E. Walkup, C. D. Oughton, J. F. Rheinfrank and E. N. Wise did the work. They replaced sulfur with selenium, a far better photoconductor, and devised the two-component developer of toner and carrier beads that copiers still use, the toner now a pigmented resin that melted into the paper. They also retired Carlson’s handkerchief: a corona discharge charged the plate, and a second one pulled the powder image across onto the paper.

Battelle and Haloid

The same year a patent agent named Nicholas Langer wrote an admiring article about Carlson’s patent for Radio News (July 1944); Kodak’s Monthly Abstract Bulletin condensed it, and John Dessauer, research head of the Haloid Company of Rochester, showed it to his president, Joseph C. Wilson. Haloid, founded in 1906, made photographic paper in Kodak’s home town and was losing its wartime business; in 1946 it earned $101,000 on $6.75 million in sales. Wilson wanted anything Kodak did not already own. He and Dessauer went to Columbus, saw the demonstration, and Wilson said it had “a million miles to go before it will be marketable. But when it does become marketable, we’ve got to be in the picture.” A contract took effect on January 1, 1947: $10,000 for a one-year licence with options to renew. A quarter of it, $2,500, went to Carlson, the first money the idea had earned him in ten years. Wilson later guessed why Battelle chose so small a partner: “we were going to make or break with it,” whereas a big company would have kept it on the back burner.

“Electrophotography” was judged too clumsy for advertising, so in 1948 Battelle asked an Ohio State classics professor for a word, and got xerography, from the Greek for “dry” and “writing.” Haloid demonstrated it to the Optical Society of America in Detroit on October 22, 1948, ten years to the day after Astoria, and shipped its first product in 1949, the XeroX Copier Model A. It was a flat selenium plate and a set of trays; the operator performed dozens of manual steps, cleaning the plate by rocking it in a tray of what was essentially cat litter, and a skilled one could produce a copy every three minutes. Early testers sent it back. It survived because it turned out to be a quick way to make paper masters for offset duplicators, and that side business, plus Copyflo (1955), the first fully automatic xerographic machine, which printed from microfilm onto a roll using the first rotating drum, kept Haloid alive. By 1956 xerographic products were nearly 40 percent of revenue. That year Haloid bought the four basic patents from Battelle for 25,000 shares, $500,000 a year for three years, and 3 percent of xerographic sales through 1965, and in 1958 renamed itself Haloid Xerox.

Carlson became a Haloid consultant in 1948 and moved to Rochester; he had divorced Elsa in 1945 and married Dorris Helen Hudgins in 1946. Haloid gave him a laboratory and an assistant and he took out some three dozen further patents, but a Battelle engineer’s later judgment was that his main contribution in the 1950s was keeping everyone’s nerve: “There always had to be something extralogical about continuing.” Between 1947 and 1960 Haloid spent $75 million on xerography, twice its operating earnings, borrowing $1 million from a Rochester bank in 1951 and $3 million from an insurance company in 1954, with employees buying stock and Dessauer mortgaging his house.

The 914

The 914, named for the largest sheet it took, nine by fourteen inches, had 1,260 parts. Its physicists worked in the attic of an old house in a poor part of Rochester, venting a fog of submicron carbon on any day but Tuesday, when the neighbour hung out her white linen; its engineers scavenged springs and tubing from a junkyard; its drum-cleaning brush was sewn from the back fur of New Zealand rabbits by the father of the furrier downstairs. A workable toner was found at almost the last moment by a Haloid chemist, Michael Insalaco. In August 1959 Wilson wrote to his division heads: “We are about to give birth, either to our greatest success or to our greatest failure—the 914.” He introduced it at the Sherry-Netherland Hotel in New York on September 16, 1959, where one of the two demonstration machines caught fire on live television, and the first production unit shipped on March 1, 1960, to Standard Press Steel of Jenkintown, Pennsylvania, on a tilting dolly so that its nearly 650 pounds could be angled through the door. It made about seven copies a minute on plain paper at the push of a button, and it listed at $29,500.

Almost no one bought it, because the company had decided almost no one should. Haloid Xerox leased the machine for $95 a month, including 2,000 copies, with each further copy metered at four cents, and it guaranteed the maintenance, which the temperamental machine needed (a small fire extinguisher was part of the kit). Wilson called leasing “the most important decision we ever made—except for backing xerography itself.” Haloid’s engineers had guessed a heavy user might make 10,000 copies a month; Standard Press used its machine at several times that rate from the day it arrived, and the first customers were turning out 2,000 to 3,000 copies a day. The machine created the demand it then met. Revenue went from $31 million in 1959 to $59 million in 1961, the year the company renamed itself Xerox Corporation and listed on the New York Stock Exchange. Xerox’s own history counts 10,000 machines shipped by the end of 1962 (Forbes puts the mark in 1961). Revenue passed $500 million in 1966 and was approaching $1 billion when Carlson died. In 1955 the world had made about 20 million copies, nearly all by other means; in 1964 it made nine and a half billion, nearly all xerographically. The descendants are the ones this encyclopedia cares about: Gary Starkweather put a laser where the 914’s lamp had been in 1969 and gave Xerox the laser printer, and the profits that paid for Xerox PARC were 914 profits (see also The History of Printing).

Dead End: The Wet Copiers

Photostat Machine National Archives
An automatic Photostat machine in the photographic laboratory of the US National Archives, with crocks of developer and hypo on the left. Image: General Services Administration, National Archives and Records Service, public domain, via Wikimedia Commons.

Before the 914 an office had three ways to copy, each with a catch. From the late nineteenth century the typewriter and carbon paper had replaced the copying press James Watt invented in 1780, but carbon paper only made copies while the original was being typed, and not many of them. Carlson knew the limit from his own desk: “To make 12 copies with carbon paper is pretty difficult. Often it involved two typing operations.” For longer runs an office cut a stencil for A. B. Dick’s mimeograph, which worked well once the master was made. For an exact image of an existing page there was the Photostat, invented by Oscar T. Gregory in Kansas City in 1907: a large camera that exposed the document directly onto a roll of sensitised paper, which then ran through developing and fixing baths. With no intermediate film, the copy came out as a negative, white letters on black. The reproduction was good, but it was slow and costly, and Mallory’s patent department waited up to 24 hours for drawings to come back from an outside photocopy firm.

The 1950s brought machines that promised copies inside the office. 3M’s Thermo-Fax of 1950 shone infrared light through the original onto heat-sensitive paper. Kodak’s Verifax, based on a 1947 patent, made photographic copies that came out damp and had to dry. Dupliton, Dial-A-Matic Autostat, Copease and Copymation competed for the same desks. RCA’s Electrofax, from the early 1950s, used Carlson’s own electrostatic principle but put the photoconductor, zinc oxide in a resin binder, on the copy paper itself and developed it with liquid toner. Every one of them needed chemically treated paper bought from the maker. David Owen’s summary of the results: the copies “smelled bad, were hard to read, didn’t last long and tended to curl up into tubes.” Thermo-Fax copies darkened with time. Electrofax machines mostly managed a dark grey instead of black, and the operator had to keep several bottles of liquid toner topped up.

What sank them was the paper. A coated sheet cost money on every copy and never looked like the original. Xerography kept the photoconductor on a reusable drum and put the image on ordinary paper, so that the 914’s lease could meter an extra copy at four cents. Coated-paper copiers stayed in use from the 1960s into the 1980s at the cheap end of the market and disappeared as plain-paper machines became affordable.

Copies Everywhere

After the nine and a half billion copies of 1964, Owen counts 550 billion in 1984 and 700 billion in 1985. Xerox followed the 914 with the 813 desktop copier in 1963 and the 2400 in 1964, named for the number of copies it made in an hour. At Walt Disney’s studio, Ub Iwerks adapted xerography to print animators’ drawings directly onto cels, which eliminated hand inking; the studio tested it on Sleeping Beauty (1959), and One Hundred and One Dalmatians (1961) was the first animated feature made with it.

The copier also changed what one person could take out of an institution. On September 30, 1969, Daniel Ellsberg, an analyst at the RAND Corporation, asked his friend Anthony Russo to help him find a Xerox machine. The next night they began copying the Defense Department’s secret 7,000-page history of the Vietnam War on the machine of a Hollywood advertising agency above a flower shop, owned by Russo’s girlfriend, and repeated the routine over the following weeks with sets of papers Ellsberg carried out of RAND. The New York Times began publishing excerpts on June 13, 1971. Russo disliked being called the “Xerox aide.” (See Famous Leaks.)

Governments that feared exactly this kept copiers locked up. In the Soviet Union the KGB’s Fifth Directorate and the Agitprop Department required authorisation to use a printing-office photocopier, and, according to the historian Sergo Mikoyan, offices and shops had long had to deposit samples of their typewriters’ typeface with the local KGB so that any text could be traced. Samizdat, the self-published dissident literature, stayed on typewriters and carbon paper, several copies at a time (see Soviet and Russian Computing).

Copying and Copyright

Copyright law had been written with printers in mind, and the copier put a small press in every library. The first big test came from medical publishing. On February 28, 1968, the publisher Williams & Wilkins sued the United States in the Court of Claims over copies of its journal articles made by the library of the National Institutes of Health and by the National Library of Medicine. In 1970 the NIH library filled 85,744 requests for copies of journal articles, about 930,000 pages, as a rule one copy per request and no more than 40 to 50 pages; the NLM copied 93,746 articles. On November 27, 1973, the Court of Claims ruled four to three that single copies for researchers were fair use. The Supreme Court heard the case on December 17, 1974, and split four to four on February 25, 1975, with Justice Blackmun recused, which left the ruling in place without setting a precedent. The Copyright Act of 1976 then wrote fair use into the statute and allowed libraries and archives some kinds of reproduction (see Copyright and IP in the Digital Age).

West Germany chose a levy over lawsuits. Its copyright act of 1965 allowed photocopying for personal use without payment, because the legislators judged that private copying did not yet hurt the sale of printed works; only copies made for commercial purposes owed the author a fee. Citing the growth of private photocopying into mass use of protected works, an amendment in force from July 1, 1985, added a combined levy on machines and their operators. Manufacturers and importers paid per copier, from 75 marks for one making 2 to 12 copies a minute up to 600 marks for one making more than 70. Schools, universities, research institutions, public libraries and copy shops paid per A4 page: 2 pfennigs, or 5 for pages copied from approved school textbooks. The collecting society VG Wort took in the money, also on behalf of VG Bild-Kunst. In 1987 the photocopy levy brought in 23.3 million marks.

Giving It Away

Carlson’s royalties began in 1947 and amounted to roughly a sixteenth of a cent on every Xerox copy made worldwide through 1965; with his stock the invention brought him something like $200 million. In 1953 he traded his old Studebaker for a new one, and in 1954 he and Dorris built a three-bedroom house outside Rochester, where he lived until he died, telling her he would be as happy in a trailer in the yard (“And will you take your 13 steel filing cabinets with you?”). He bought third-class train tickets in Europe until she stopped him; an acquaintance who heard he worked at Xerox asked whether he belonged to the union. When Fortune listed him among the richest Americans in 1968 he wrote to say the estimate was too high by $150 million, because he had already given that much away.

The giving was anonymous and he weighed every request himself. He paid for a chemical-physics building at Caltech on condition that it be named for his old professor Arthur Amos Noyes; gave millions to the United Negro College Fund and to individual Black colleges; bought apartment buildings in Washington and New York and had them racially integrated; funded the Fellowship of Reconciliation and other pacifist groups; supported Robert Hutchins’s Center for the Study of Democratic Institutions through his life and his will; paid in 1966 for Philip Kapleau’s Rochester Zen Center; and bought a New York carriage house for the Zen teacher Eido Shimano, whose Dai Bosatsu monastery in the Catskills was later built on land Dorris paid for. Dorris had drawn him toward Vedanta and Zen, and he also took a lasting interest in psychical research. He funded the University of Virginia psychiatrist Ian Stevenson every year, and a large gift in 1964 helped pay for one of the university’s first endowed chairs, which let Stevenson step down as head of psychiatry and set up a separate Division of Personality Studies (later the Division of Perceptual Studies). His will left the university $1 million to continue the work. The bequest caused controversy on campus, but it was accepted, and Stevenson became the first Carlson Professor of Psychiatry. The Center for the Study of Democratic Institutions received more than $4.2 million under the same will. In 1965 he donated the first xerographic print and his original apparatus to the Smithsonian, and on December 16 of that year sat for the long tape-recorded interview with Dartmouth’s Joseph Ermenc that is the main record of the story in his own words. On September 19, 1968, he died of a heart attack in New York, aged 62. U Thant, the secretary-general of the United Nations and a friend, wrote that he had respected Carlson “more as a man of exceptional moral stature and as a humanist” than as an inventor. He was named Inventor of the Year in 1964 and received the Horatio Alger Award in 1966. The rest came after his death. The National Inventors Hall of Fame inducted him in 1981, and on October 20, 1983, the American Society of Mechanical Engineers designated the development of xerography an International Historic Mechanical Engineering Landmark. Congress made October 22, 1988, the fiftieth anniversary of the Astoria copy, National Chester F. Carlson Recognition Day, and the Postal Service had issued a 21-cent Carlson stamp in its Great Americans series the day before. Rochester Institute of Technology named its Chester F. Carlson Center for Imaging Science after him, and the University of Rochester its Carlson Science and Engineering Library.

📚 Sources