The Rio Maior – Azinheira Ridge (RMAR) Chert Characterization Project

RMAR cover cobblesREVThe high-quality chert deposits of the Azinheira Ridge located near Rio Maior, Portugal were the most extensively exploited raw material source for flaked stone tools in all of western Iberia. This online reference aims to characterize the geological deposits in ways useful to archaeologists and historians in Portugal and beyond. The RMAR project was started as part of an archaeological survey in 1990 and recent research culminated in the online database announcement at the RawMaterials2016 conference in Faro, Portugal. Geoarchaeological and historical investigation into human-resource interactions at Azinheira continues today.

Rio Maior or Azinheira Ridge chert is not a homogenous, red chert as is sometimes purported.  RMAR cherts, often even those from the same gravel bed, have highly variable cobble morphologies, macroscopic traits, and geochemistries reflecting complex formation origins and diagenesis.

The RMAR online database makes primary data available to professional archaeologists seeking to identify chert artifacts manufactured from cobbles procured in the Rio Maior vicinity. The database also is useful for geoarchaeologists pursuing chert research, especially those scientists investigating the complexities of secondary gravel deposits and geochemical variability. The RMAR database is a starting point for more technical laboratory analyses and sample flakes of most of the chert variants in the database can be made available to professional scientists for collaborative research.


Exploiting the RMAR Chert Gravels from the Old Stone Age to Modern Times

Neanderthals and early anatomically-modern humans flaked RMAR chert cobbles into stone tools throughout the last Ice Age from over 60,000 to 10,000 years ago. The abundant stone was easily worked into a sharp tool edge and the presence of the Azinheira Ridge gravels may account for the very high density of Upper Paleolithic archaeological sites recorded near Rio Maior.

  up endscraperREV2  up blade coreREV2 up bladelet coreREV2
 An Upper Paleolithic endscraper on RMAR chert (RMESP1991-72)  An Upper Paleolithic blade core fashioned on RMAR chert (Espadanal) Face of a Late Upper Paleolithic bladelet core on RMAR chert (Vale de Obidos)

Azinheira Ridge chert was extensively utilized both throughout prehistory and in more recent historic times. The village of Azinheira was the primary center for Portuguese gunflint manufacturing through the 17th and 18th centuries. Gunflints and strike-a-lights made of RMAR chert were provided to the Portuguese military and also sold directly to vendors and households. RMAR chert gunflints were thus transported throughout the Portuguese colonial world. Portuguese gunflints differ in important technological ways and are distinctive from those manufactured in England and France. A detailed ethnoarchaeological study of the Azinheira gunflint industry is forthcoming.

gf1 wscaleREV2 gf2 with scaleREV2
Two Portuguese gunflints recovered from manufacturing sites on the Azinheira ridge.

The chert cobbles of the Azinheira Ridge occur in gravels and reworked Miocene sand deposits in secondary geological provenience. None of the cherts located in the RMAR survey were found within limestone bedrock, chert beds, or other primary geological contexts; All of the cobbles were transported over medium and long distances through alluvial and other geological processes to their location of discovery. Lithic provenancing studies in archaeology need to better understand extensive secondary sources of chert in order to build accurate models of prehistoric raw material provisioning. While the original bedrock source of a RMAR cobble may be located 100s of kilometers away from Rio Maior, the nearest-source availability is the important geographic variable for understanding past human use of stone. Prehistoric and historic stone knappers were concerned about effective procurement of quality stone, not whether the source was in primary or secondary geological context.


RMAR Chert Survey map cobbles 1REV cobbles 2REV
Generalized Map of the
Rio Maior – Azinheira Ridge Chert Source.
Examples of field collections of RMAR cobbles from erosional rills and ephemeral streambeds along the Azinheira Ridge.

Characterizing RMAR Stone Cobbles

From 1990-2016, over 2,330 chert cobbles were examined during geoarchaeological fieldwork on the Azinheira Ridge. The RMAR Reference Collection was built as a stratified sample encompasseing the major macroscopic variants of cherts present. The procurement location for numbered samples was noted according to a zonal grid. In order to analyze cobble size distributions, total cobble samples were taken in some areas. The total cobble sample procedure collected all chert cobbles larger than three centimeters found within in a selected surface area. Areal density was calculated and the entire assemblage of cobbles was then transported to the laboratory for analysis.

A few non-chert raw materials are included in the RMAR database. For inclusion in the reference collection, the materials must meet two criteria, excellent workability for stone tools and possible misidentification as a chert.

Determining an accurate and repeatable measurement of the visual surface color of chert is very complex. Standard Munsell chart matching techniques are unreliable because of significant perceptual variation between individuals, lighting environment effects, and the surface translucency of many cherts. Often visual chart-matching produces Munsell designations significantly darker than the actual chert surface reflectance.

The RMAR database employed a hand-held visual reflectance and photogrammetry instrument to measure surface color. Because many of the RMAR cherts are mottled or banded, a series of four- millimeter-diameter-spot measurements were analyzed across flake surfaces and from the interior to the exterior of cobbles. Patterns of color variation are noted in the RMAR database color distribution field.

The RMAR database is organized by predominant color classes in order to facilitate easy identification of possible macroscopic matches by users. The full Munsell notation and name for the predominant surface color is provided for each chert variant.

Textural grain of fracture surfaces also varies significantly within many cobbles. In cases where textural and color differences are highly correlated, a notation is added within the color distribution field. The preponderant grain at low power magnification (3-5x) is listed for each chert sample.lusterREV


Important differences in surface luster between cobbles was recognized early in the RMAR project. Luster is linked to crystallization and thus has implications for formation environment, diagenesis, and artifact weathering. Multiple surfaces of each chert type were examined using reflection and total effect assessments for determination of surface luster. Each chert type was assigned a luster category ranging from very bright through bright and medium to dull.


Diaphaneity is another macroscopic property of lithic materials under-appreciated by archaeologists. Most RMAR cherts exhibit some degree of translucency while rarer jaspers from the Azinheira Ridge are opaque. Flake diaphaneity was ranked using two dimensions: translucency 1-2 millimeters proximal from the non-cortical edge of a flake and translucency at a non-cortical thickness of 2 millimeters. The combination of these two dimensions made it possible to determine each sample’s average diaphaneity using a 4 point scale (very translucent, translucent, slightly translucent, and opaque.)

transluc with inclusionsREV jasp diaphREV
Some RMAR cherts display high degrees of translucency RMAR true jaspers have opaque diaphaneity

RMAR Chert Geochemistry and Diagenesis

Chert geochemistry was investigated using x-ray and magnetic methods. Compton-normalized X-ray fluorescence determined semi-quantitative elemental composition with followup X-ray diffraction studies of unusual mineralogies. Mass-specific and frequency dependent magnetic susceptibility provided an additional line of evidence for interpreting iron and titanium minerals type, concentration, and grain-sizes.

 Slide4 Slide3
XRF identification of manganese mineral inclusions in RMAR chert 37

The geochemical composition of the RMAR cherts varies widely and indicates multiple parent formation origins for the chert cobbles. Bedded, nodular, and oolitic cherts are all associated in gravels indicating very different chert formation environments. Some RMAR cherts were originally formed in deep marine settings while other indicate shallow shelf sedimentary environments.

Geochemical sourcing methods rarely explore the impact of intra-cobble sampling procedure on analytical results. The RMAR project is demonstrating elemental and mineralogical patterns linked to color and textural variation. These efforts at understanding chert formation, diagenesis, and weathering are important for developing more robust cobble sampling methods in geological proveniencing investigations.

 Slide1 Slide2
XRF results showing significant differences in silica, clay mineral, and iron oxide content in different bands of RMAR chert 29

Modeling Lithic Economies: RMAR Cobble Morphologies

Archaeologists provenience stone raw materials as a first step toward understanding prehistoric raw material economies. For anthropological explanations of artifact assemblages, examining the form and amount of transported stone is at least equally important as the geographic relationships of source areas and archaeological sites. Lithic technologists are beginning to model raw material economies through volumetric equations. The presence of cortex and weathering rinds on the surface of stone debitage and debris at archaeological sites is useful for understanding whether cobbles were transported to sites with minimal reduction at the source or as extensively-prepared cores.

Cobble size statistics were calculated using the total cobble assemblages from numerous locations on the Azinheira Ridge. Mass was determined by weight to the 0.001 gram. Volume was directly measured through a precision water-displacement procedure. These data were found to be highly correlated with results of mathematical ellipsoid volume calculations using length, width, and height measurements of cobbles taken with digital calipers. Surface area was calculated using a modified foil-wrap method. Surface area results also correlate well with mathematical ellipsoid calculations.

Cobble Volume


Overall the mean chert cobble in the RMAR survey area has the following dimensions:

Mean Std. Deviation
Length in mm 112.5 33.5
Width in mm 76.9  29.1
Height in mm 58.2 18.2
Weight in grams 665.6 564.4
Total volume in c3 329.968 328.739
Cobble surface area in mm2 22746 14149

Chert Sample Macroscopic Photographs

The RMAR database is organized around macroscopic photographs of freshly-fractured chert surfaces. Samples were selected for the database in order to fully characterize the visual variability present in RMAR gravels. Each photograph was taken under identical lighting conditions on a photographic copy stand illuminated by 12 broad-spectrum photographic bulbs. White balancing was performed and verified using a digital color control card.

Each standard RMAR photograph is scaled at 25 millimeters on each edge. When recurrent cortex or weathering rind surfaces were recognized in association with a chert type, the exterior surfaces were photographed at the same scale.

color card shooting lightrev

The digital color control card photographed under RMAR database lighting conditions. Color check and calibrate your display device using this image.


The RMAR project is coordinated by Paul Thacker, Associate Professor of Anthropology and Director of the Archaeology Laboratory of Wake Forest University. The RMAR website was designed by Jeff Nichols, an instructional technologist at Wake Forest University. Geoarchaeological fieldwork by Thacker at several stages of the project was supported in part by funding from the U.S. National Science Foundation. The important collaboration of Carlos Pereira was facilitated by the Sector de Museus e Património Histórico, Arqueológico e Cultural, Câmara Municipal de Rio Maior. The dedicated work of João Aguiar, Caroline Bertoni, Lila Franco, and Elizabeth Harraka contributed significantly to the laboratory studies of RMAR cherts from 2013-2016.

Comments are closed.