Peter Warry brought samples of Roman tegulae or roof tile and research documents to present his findings to the volunteers participating in pottery identification and analysis of animal bone from the Marcham excavations.
Peter encountered a problem over 10 years ago finding data on the vast quantities of roof tile dug out of trenches and dumped on spoil heaps, seemingly without further thought to its own story and how it might contribute to the interpretation of the life and time of the site under excavation. So not to be deterred, Peter spent a few years on spoil heaps and working with the Directorate at Marcham, Silchester and York sites, to name but a few, to identify and classify tegulae in the context of the excavation stratigraphy.
Unlike pottery where inclusions in the clay, production methods for specific usage and possible trade routes are interesting evidence, the clay used for tile making was unadulterated local material and as such is of interest but not likely to contribute to the typology.
So where to next? Peter amassed a great deal of tile sherds and noticed that during the production stage there are notches cut into the top and bottom of the clay so each tile would rest securely on the tile above and below on a typical 30 degree roof angle. On a closer look, Peter identified four cutting styles; some cut by hand and others seemed to be moulded.
Did the Romans bring the production skills with them in the second century and did the production methods change over time? Peter researched Roman military establishments and villas on the Continent which suggested the tegulae production style travelled with the invaders and also noted that tiles made for civic buildings were of a higher specification than tiles produced for villas and other general buildings.
Which of the notch style came first and last? The huge trench at Roman Silchester in Hampshire, was the perfect excavation site to measure the recovery of tile sherds to examine layer by layer over the seasons between 1997 and 2004 so gave Peter the date range of style A, as he called it, up to AD 120, style B between AD 100-180, style C, AD 140-260 and style D onwards from AD 240.
Peter then observed that the earlier tiles had straight parallel flanges and concluded that they would have been made by pushing the clay into a mould and smoothing the top with a ‘cheese’ wire and then cutting the notches by hand. A change in production occurred in the third century AD with the introduction of a negative mould so the top of the tile is at the bottom of the mould and by building in a notch shape into the mould it was quicker to produce.
Tegula with makers mark and paw print
From kiln evidence on military sites Peter deduced that they would have produced their own tiles. Tiles for civilian use were often made on farms and may have been franchised from evidence of makers initials stamped into the clay. Tile production would have been an all year round activity and Peter worked out that it would have taken 14 people to quarry, move the clay and form the tegulae. During the dryer months between May and September it took 14 days to load the kiln, fire and unpack the tiles. Timber to fire the kiln would be gathered during the winter months.
To have your roof tiled was a status symbol and the makers made finger patterns on the corner of each tile before firing. However on someone’s roof there was an addition of dog paw prints.
On the Marcham site, Peter suggests there is no evidence of buildings with roof tiles but a 3rd-4th century scatter could have been a ritual deposit with coins and pottery, particularly in Trench 2, the cruciform building. However the Temple had good evidence for a tiled roof which survived after the building fell into disuse.
If you would like to find out more about Roman tile, Peter’s research is published as – Tegulae: Manufacture, typology and use in Roman Britain. 2006 BAR British Series 417. Archaeopress, Oxford.