The fax machine was invented in 1843, when Scottish clockmaker Alexander Bain patented a device that scanned a metal plate with a stylus and transmitted the pattern over a telegraph wire. Bain never built a commercially working system, and the first practical fax came in 1865 from Italian physicist Giovanni Caselli, whose pantelegraph carried handwritten images between Paris and Lyon. The modern paper fax emerged in 1964 with Xerox's Long Distance Xerography, followed by the 1966 Magnafax Telecopier.

The scanning mechanism that made fax possible

Every fax system from 1843 to today solves the same three-part problem: convert an image into an electrical signal, send that signal, and rebuild the image at the other end. Bain's insight was the first step. A metal stylus traced a raised image on a conductive plate, and contact versus no contact produced an on-off electrical pattern. That pulse train traveled down a telegraph wire and drove a matching stylus at the receiving end.

The receiver's synchronization problem is where Bain's design stalled. Both drums had to spin at exactly the same rate, and 1840s clockwork could not hold that alignment. Caselli solved it in 1865 with a pendulum-based regulator tied to a telegraph clock signal. His pantelegraph used a stylus over a tinned iron plate inked with insulating varnish, and it transferred a 5-by-7-inch image in about two minutes over roughly 300 miles of wire.

The core principle never changed. A fax is a one-bit-per-pixel serial image transmission. What changed across 180 years was the sensor, the encoding, and the network. Photoelectric cells in the 1920s replaced the contact stylus. Digital compression in the 1980s replaced analog amplitude modulation. Packet networks in the 2000s replaced the telephone line. The scan-send-print loop stayed identical.

Quantified context: resolutions, speeds, and standards

Fax performance is measured in scanlines per inch, transmission speed in bits per second, and page time in seconds. These numbers map directly onto what a fax can and cannot reproduce.

The first commercial telephotography systems of the 1920s scanned about 100 lines per inch and took 6 to 8 minutes per image. The 1966 Magnafax Telecopier ran at roughly 6 minutes per page over a standard voice line. By the 1980s, Group 3 machines compressed a page to about 30 seconds.

Group 3 is the standard that defined the office fax era. It scans at 1728 pixels horizontally and 1143 lines vertically on an 8.5-by-11-inch page, giving roughly 200 by 100 dots per inch. Group 3 uses Modified Huffman or Modified Read compression and negotiates baud rates up to 14,400 bits per second, with fallback. A text-heavy page compresses to about 25 to 40 seconds over a dial-up line.

The timeline table shows how each generation changed the underlying physics:

Era System Scan method Page time Network
1843 Bain patent Contact stylus Never operational at scale Telegraph wire
1865 Pantelegraph Contact stylus ~2 minutes per small image Telegraph
1920s Telephotography Photoelectric cell 6–8 minutes Telephone
1966 Magnafax Telecopier Optical scan ~6 minutes Telephone (analog)
1980 Group 3 Optical, digital ~30 seconds Telephone (digital signaling)
1994 Group 3 with V.34 Optical, digital ~10 seconds Telephone
2000s FoIP (T.38) Optical, digital ~30 seconds IP packet network

Edge cases: why fax refused to die

The counterintuitive detail is that fax is not a legacy curiosity. It persists because a fax receipt is an image of a signed original, and in several regulated workflows that image carries evidentiary weight a scanned email attachment does not. The mechanism matters: a fax transmission records a confirmation page with sender identity, receiver identity, time, and page count. That audit trail is generated by the telephone network, not by the sender's software.

Law and medicine are the main holdouts. The Health Insurance Portability and Accountability Act does not mandate fax, but its privacy rule treats a fax sent to a verified number as an acceptable transmission method, and many covered entities built compliance procedures around it. The Food and Drug Administration accepts faxed regulatory submissions and correspondence in defined circumstances. Thousands of pharmacy prescription transfers and lab orders still travel by fax each day in the United States.

The professional caveat: fax over IP, governed by the ITU T.38 recommendation, is not the same as a dial-up fax. T.38 re-encodes fax signals for packet networks, and packet loss or jitter can break the handshake. That failure mode is silent in some systems, producing a confirmation page for a partial transmission. Businesses that route fax through VoIP need to verify T.38 support and end-to-end confirmation, not assume analog reliability.

Misconceptions about the fax timeline

Four claims circulate widely and each needs correction.

The fax predates the telephone. True. Bain's 1843 patent came 33 years before Bell's 1876 telephone patent, so the fax is older than the device it is associated with. The fax was invented in the 1960s. False. The 1960s produced the first convenient office fax, not the first fax. Fax and email are functionally interchangeable. False in regulated contexts. Email attachments lack the network-generated confirmation page. Group 3 is obsolete and unsupported. Misleading. Group 3 remains the dominant public telephone network standard, and compatible machines are still manufactured.

The older-than-the-telephone fact is the one that reframes the whole history. The fax was designed for telegraph infrastructure and only migrated to phone lines when voice networks became the cheaper carrier.

Practical actions for fax use today

If you must send a fax, the following steps reduce failure risk and preserve the evidentiary value.

Confirm the receiving number in writing before transmission, because misdialed faxes to unverified numbers are a recognized privacy breach in medical settings. Request the transmission confirmation report and retain it with the sent document; this report is the audit artifact, not the email receipt. For VoIP-based fax, ask the provider whether T.38 is supported and whether end-to-end confirmation is verified at the gateway. Prefer 200-by-100-dpi Group 3 sending for small print, since finer text below roughly 6 points can degrade at standard resolution. For documents containing protected health information, verify the destination's fax privacy procedure rather than assuming it. Keep a fallback: if a page fails twice, send by a method with explicit delivery confirmation rather than resending blindly.

The 6-point threshold matters because Group 3's vertical resolution of about 100 lines per inch cannot reliably resolve finer strokes. A page that looks fine on screen can arrive unreadable on a fax.

Limitations and what the record can and cannot show

The invention question has a clean answer only at the patent level: 1843 for Bain. Commercial and practical milestones are fuzzier because "fax" describes a family of technologies, and the first machine to do each function differs. The 1865 pantelegraph is the first system that reliably reproduced handwriting at a distance, but it ran on dedicated short links and never scaled.

Numbers from the early era are reconstructed from patent descriptions and contemporary accounts, not from standardized measurement. Claims of exact speeds for 1840s devices carry more uncertainty than a modern specification sheet implies. Later figures, especially Group 3 parameters, come from formal ITU standards and are firm.

What the record does not support is a single "invented in year X" narrative that ignores the telegraph origin. The honest answer separates patent, prototype, and product. The first modern paper fax arrived with Xerox's LDX in 1964 and the Magnafax Telecopier in 1966, more than a century after Bain's patent and a decade after the fax idea migrated onto voice phone lines.