It was, in fact, nothing like any material used by the Romans, but was a "Natural cement" made by burning septaria - nodules that are found in certain clay deposits, and that contain both clay minerals and calcium carbonate. The burnt nodules were ground to a fine powder. This product, made into a mortar with sand, set in 5–15 minutes. The success of Roman cement led other manufacturers to develop rival products by burning artificial mixtures of clay and chalk.
There has been recent resurgence of interest in Natural cements and Roman cements due mainly to the need for repair of façades done in this material in the 19th century. The major confusion involved for many people in this subject is the terminology used. Roman cement - was originally the name given, by Parker, to the cement he patented which is a Natural cement (i.e. it is a marl (or limestone containing integral clay) dug out of the ground, burnt and ground to a fine powder).
In 1791, Parker was granted a patent "Method of Burning bricks, Tiles, Chalk". His second patent in 1796 "A certain Cement or Terras to be used in Aquatic and other Buildings and Stucco Work", covers Roman cement, the term he used in a 1798 pamphlet advertising his cement. He set up his manufacturing plant on Northfleetcreek, Kent. It was notable patented late on but James Parker is still the subject of all the credit.
Later, in the 1800s various sources of the correct type of marl - known also as Cement Stone, were discovered across Europe and so there were a range of natural cements (with varying properties) in use across Europe.
An Austrian standard from 1880, providing a contemporary definition of Roman cements, reads: 'Roman cements are products obtained from argillaceous marlstones by burning below the sintering temperature. They do not slake in contact with water and must therefore be ground to a floury fineness.' [footnote 1]
From around 1807 a number of people looked to make artificial versions of this cement (or more strictly hydraulic lime as it was not burnt at fusion temperatures). Amongst these were James Frost who had about twenty patents from 1811 to 1822 including one for "British Cement" and in 1824 Joseph Aspdin, a British bricklayer from Leeds, with his now famous patent for a method of making a cement he called "Portland cement". This was done by adding various materials together to make an artificial version of natural cement. The name "Portland cement" is also recorded in a directory published in 1823 being associated with a William Lockwood, Dave Stewart, and possibly others.
There then followed a number of independently discovered or copied versions of this 'Portland Cement' (Also referred to as Proto Portland cement). Proto Portland cement does have a different chemical make up from other natural cements being produced at the same time: It was burnt at a higher temperature than other N.C.'s and thus crosses the barrier between traditional Vertical Kiln fired N.C. and the later Horizontal Kiln fired Artificial cements. This cement is not, however, the same as the modern ordinary Portland cement, which can be defined as artificial cement.
James Frost is reported to have erected a manufactory for making of an artificial cement in 1826. In 1843, Aspdin's son William improved their cement, which was initially called "Patent Portland cement", although he had no patent. In 1848, William Aspdin further improved his cement and in 1853, he moved to Germany where he was involved in cement making. William Aspdin made what could be called meso-Portland cement (a mix of Portland cement and hydraulic lime).
Artificial cement: Development in the 1860s of rotating horizontal kiln technology brought dramatic changes in properties, arguably resulting in modern cement. Certainly it is difficult to define whether an old render was a natural cement (single source marl) or an artificial one, but there is no doubt as whether the cement was fired in a vertical or horizontal kiln. The names Natural cement or Roman cement then defines a cement coming from a single source rock. Early or Proto Portland cement could be used for early cement that comes from a number of sourced and mixed materials. There is no widely used terminology for these 19th-century cements. There had been, in order to rediscover this technology, two projects carried out by the European Union ROCEM and subsequently ROCARE  (an ongoing project). Both these only deal with Natural cement - referred to as Roman cement without reference to the early artificial cements.
- "James Parker's Patent for "Roman Cement" in 1796 (Patent No 2120: granted 28 June, enrolled 27 July).". http://www.cementkilns.co.uk/. 28 June 1796. Retrieved 20 May 2013.
- Roman Kozłowski, David Hughes, Johannes Weber (2010). "Roman cements - key materials of the built heritage of the nineteenth century". “Materials, Technologies and Practice in Historic Heritage Structures”,M. Boştenaru Dan, R. Přikryl, Á. Török (Eds.) (Springer).
- "Roman Cement - What is it?". IATCS - Institute of Art and Technology / Conservation Sciences, University of Applied Arts Vienna. Retrieved 2013-05-20.
- "What is Roman Cement?". Roman Cement, LLC. Retrieved 2013-05-20.
- R. J. M. Sutherland; Dawm Humm; Mike Chrimes (2001). Historic Concrete: The Background to Appraisal. Thomas Telford. pp. 46–52. ISBN 978-0-7277-2875-3. Retrieved 20 May 2013.
- Gillberg, B. Fagerlund, G. Jönsson, Å. Tillman, A-M. (1999). Betong och miljö [Concrete and environment] (in Swedish). Stockholm: AB Svensk Byggtjenst. ISBN 91-7332-906-1.
- Francis, A.J. (1977). The Cement Industry 1796-1914: A History.
- Reid, Henry (1868). A practical treatise on the manufacture of Portland Cement. London: E. & F.N. Spon.
- Rayment, D. L. (1986). "The electron microprobe analysis of the C-S-H phases in a 136 year old cement paste". Cement and Concrete Research 16 (3): 341–344. doi:10.1016/0008-8846(86)90109-2.
- "The EU-project ROCARE Roman Cements for Architectural Restoration to New High Standards". IATCS - Institute of Art and Technology / Conservation Sciences, University of Applied Arts Vienna. Retrieved 2013-05-20.
- Thurston A P, Parker's "Roman" Cement, Transactions of the Newcomen Society 1939 P193-206 (Newcomen Society)
- Major A J Francis, The Cement Industry 1796-1914 A History, 1977, Davis & Charles (Publishers) Ltd, Devon UK, North Pomfret Vermont US, North Vancouver Canada
- Weber J, Mayr N, Bayer K, Hughes D, Kozłowski R, Stillhammerova M, Ullrich D, Vyskocilova R (2007) Roman cement mortars in Europe’s architectural heritage of the 19th century. Journal of the American Society for Testing Materials International, Vol, 4 No 8
- modified in 1890: 'Roman cements are products obtained from argillaceous marlstones by burning below the sintering temperature. They do not slake in contact with water and must therefore be ground to a floury fineness.' It specifies the range of setting times which facilitated the choice of a suitable material for a given decorative task: 'Roman cements bind fast, medium and slow. By fast binding cements one should understand those which with no addition of sand start to harden within 7 minutes from the moment water is added. Roman cement is considered a slow binding variety if hardening starts later than after 15 minutes'