How We Do It

The following information describes the methodology we use to collect and document each of our specimens.


Collecting Specimens

The stone specimens contained in the database are, for the most part, donations from the companies that quarry them. We regularly identify stones that we'd like to include and contact quarries to request a sample. Some of the specimens in our collection have come from other sources, such as with stones removed from existing buildings or monuments. You can determine the source of a particular stone by looking at related specimen pages (click the specimen ID number located in the "Images" section on each stone page). 


Unique Identifier Assigned

Labeling specimens

Labeling Specimens

New specimens are assigned a unique ID

When new specimens arrive at our facility, the first thing we do is assign a unique ID incorporating the year the specimen was received and a three-digit number assigned in the order samples were processed (e.g. 2013-001 for the first sample received in 2013). When we receive multiple specimens of the same stone that each possess unique characteristics making them worthy of retention and addition to the database, a third set of numbers is added to create a unique ID (e.g. 2014-024-1 and 2014-024-2 are both examples of Buff Indiana Limestone, but the second specimen has a tooled finish). Once this ID is set, we move on to recording the physical attributes of each specimen.


Dimension stone is subject to several ASTM International standards. These commonly adopted standards specify the performance requirements of the materials used in construction and other industries. Some manufacturers provide ASTM testing data for their products on their websites. If not available online, the sales offices can often provide ASTM data on request. The following data was collected for each dimension stone specimen:

  • Density (lbs/cf) ASTM C-568
  • Water Absorption (%) ASTM C-97
  • Compressive strength dry/wet (psi) ASTM C-170
  • Flexural strength dry/wet (psi) ASTM C-880
  • Modulus of rupture dry/wet (psi) ASTM C-99
  • Some manufacturers provided data for additional standards such as Abrasion Resistance (no units) and Specific Gravity (lbs/cf).

Roofing slate is subject to ASTM standards for:

  • Modulus of rupture (across the grain) ASTM C-120
  • Absorption ASTM C-121
  • Weather resistance ASTM C-217

Stone Type


ASTM also provides standardized classifications for building stone types. While widely adopted by stone quarries and the constructions industry, these commercial types often differ from the classifications used by geologists. Each stone in the database is given one of the following commercial classifications based on ASTM standards: granite, limestone, marble, quartz based (i.e. sandstone, quartzite, bluestone), or slate. Stones or other materials not included in the preceding list are classified as "other".


In addition to a commercial classification, each stone is classified by its lithologic rock type. These lithologic names are derived from the taxonomy proposed in the LITHLIST Lithologic Data Dictionary.1 A list of all the lithologic names we use can be found here.

For more about stone classifications on this website, see the How to Use This Site page.

Color Descriptions

We use two methods of classifying the color of stone specimens: basic color groups for search purposes, and precise color descriptions for comparative purposes.

Basic Color Groups

Each specimen is assigned to one or more of 11 basic color groups (white, gray, black, buff, blue, green, yellow, red, brown, purple, pink) to enable searching by color within the database. As color perception is subjective, specimens are assigned multiple color groups. For example, a yellowish buff limestone with gray flecks would be tagged “yellow,” “buff,” and “gray”. Multicolored specimens are tagged with each color visible.

Munsell Colors

More precise color descriptions were developed using the Munsell Color System and the descriptive names developed by Kelly and Judd in the ISCC-NBS System of Color Designation.2 Up to three colors are noted for each specimen. Munsell’s Soil Color Book is used primarily, but Munsell’s Rock Color Book is also useful for some specimens. Each Munsell color value is converted to a descriptive name using the charts on pages 16-31 of the Kelly book. For example, a multicolored granite might be described with Munsell values of 10R 5/3, 10R 4/2, and N2. These values are described using the ISCC-NBS System as “grayish red,” “grayish reddish brown,” and “black” respectively.


Of course, the primary utility of the database comes from the high quality images it contains for each stone specimen. We capture and present images at two different magnifications: at the macro level, capturing the general visual characteristics of the stone that can be seen by the naked eye; and the micro level, demonstrating the characteristics that can be seen under a microscope (for more about this, see "Thin Sections" below).

Each specimen is photographed close up (approx. 13” camera height) for maximum resolution. Large specimens are also photographed at greater distances to capture the entire specimen. Stones often have a different appearance when they're wet so, unpolished specimens are generally photographed dry and wet. Polished specimens and slates don't really change in appearance and are photographed dry only.

All specimens are photographed on a copy stand using a digital SLR, shooting in RAW (.nef) format. Each shot is staged with a color reference card, scale, and information card, helping us keep track of images and perform color correction in post-processing if necessary.

The images on this website are all available in high resolution JPEG format (this is what you see in your web browser) and can also be downloaded in lossless (even higher resolution) TIFF format.

Other Attributes

Coming soon.


Thin Sections

Image of thin section in non-polarized light.

Mount Airy Granite

Image of thin section in non-polarized light

Image of thin section in cross-polarized light.

Mount Airy Granite

Image of thin section in cross-polarized light

Almost every stone specimen in the database is accompanied by a thin section. These are very thin slices of rock, usually 30 microns (0.03 mm) thick, that allow scientists to study the optical properties, grain size, shape and proportion of the constituent minerals in a specimen. To accomplish this, a small portion of the specimen is cut out and sent to a laboratory where specialists mount the piece to a glass microscope slide and grind and polish it to the specified thickness. Stones with high porosity (e.g. sandstones and some limestones) are also vaccume-impregnated with a blue epoxy resin to help identify pore space.

Scanning Thin Sections

To make the thin sections available to the public, each slide is scanned using a modified film scanner according to the method shown here: Two images of each slide are produced: one showing the section in ordinary transmitted light and another showing the section through cross-polarizing filters. The combination of the two images allows users with a knowledge of optical mineralogy to identify the petrographic properties of the stone.


Petrographic Analysis

General Microscope Methods

All samples were prepared as 30 µm thin sections of standard size (27 X 46 mm) with a glass coverslip and blue impregnating matrix. Unless otherwise noted in the petrographic descriptions the following applies: Materials were oriented so as to maximize surface area on the slide.

Slides were not chemically stained to augment mineral identifications.

Slides were assessed for categorical classifications (e.g. percent of allochems contribution to limestone volume) by means of visual comparison to standards, as opposed to point counting.

The highest power objective used in analysis was a 40X.

Only the standard suite of transmitted and polarized light techniques were utilized to identify minerals of adequate size including relief, pleochroism, birefringence value, conscopic uniaxial/biaxial determination, concscopic 2V value, length of elongation, and other observable mineral characteristics—e.g. twinning patterns, cleavage patterns, etc.

Categorical Descriptors

The database is constructed so that all specimens are associated with three categorical descriptors of grain size, grain shape, and grain sorting. These classifications are adopted from systems used for terrigenous clastic rocks and used here merely as a heuristic descriptor or a means of folk classification. The following guidelines were applied to various rock types as described below to add some rigor to the classifications.

Labeling specimens

Petrographic Examination

Matt Pailes, PhD, examining a thin section slide using a polarized light microscope.

Grain Size

Grain size typically has two values selected. These respectively reference the maximum grain size observed and the most predominate, or modal range, of grain sizes in the rock. If both of these values fall within one classificatory range then only one is selected. In cases where non-contiguous categories are selected the detailed description should be consulted to infer whether there is a continuous or multimodal distribution of grain sizes. The system of size categorization follows that laid out in the How to Use This Site page. Note that this system does not strictly correspond to the two most common grain size classifications used in U.S. geological and sedimentological studies. The the Krumbein Phi-Scale, United States or the ISO 14688-1:2002 ( should be referenced for comparative purposes. A few more caveats are discussed below:

The thin sectioned portion of rock was often selected based on “interesting” composition or textural areas identified in hand specimens. This practice may lead to some non-representative grain size categorizations. This was controlled for to the extent possible. For example, in rocks with obvious secondary mineral formations these crystals were generally excluded from consideration of grain size unless they were also observed to be common in the hand-sample.

The system employed here is fundamentally based on a categorization of terrigenous-clastic rocks. As such, it is a poor fit to some species of rock with substantially different origins. Straight forward size criteria could be applied in many cases, but in the case of limestone it was thought the most useful approach would be to restrict the grain size classification to actual grains, e.g. bioclasts, intraclasts, etc. As such, in regards to limestone the matrix material, which may range from cryptocrystalline micrite to phaneritic sparite, is not considered in the classification. It should be noted that at times the total allochem contribution may be relatively low. In general, care should be taken to consult the more exhaustive thin section descriptions.

Grain Shape

For terrigenous materials this classification followed widely accepted norms such as those described in Pettijohn et al. (1973). 3

For fossiliferous limestone one box was checked for the most predominant allochem and one for the allochems with the greatest amount of visible angularity. Sparite and neomporhic fabrics and other secondary minerals were not considered in this estimation.

For plutonic materials this field was uniformly checked as angular.

For most metamorphic materials this field was either left blank—due to the small size of grains, or checked as angular to reflect the consertal nature of visible grain boundaries. Some minimally metamorphosed quartzites could feasibly be assigned other values.

Grain Sorting

For terrigenous materials this classification followed widely accepted norms such as those described in Pettijohn et al. (1973). When deemed appropriate these guidelines were also applied to some quartzite materials.

Limestone was again a problematic rock to fit to this classification scheme owing to the complexity of those systems actually employed by geologists. For simplicity, a simple translation method was used in which grainstone and packstone equates to poorly sorted, wakestone to moderately sorted, and mud stone to well sorted. This approach does not capture the significant textural variation present in many limestones and the more elaborate petrographic descriptions should generally be consulted.

In other materials in which metamorphism obscured grain size or crystals sizes were never determined by sedimentary sorting (magmatics) this field was left blank.


Chronostrigraphic units are based on the “International Chronostratigraphic Chart” published by the International Commission on Chronostratigraphy, 2013 edition.4

Chronostratigraphy was determined through a variety of means. In some cases this information was provided by a supplier. In most cases this information was determined by locating the quarry through satellite imagery and comparing the location to USGS maps. This approach can be problematic in that USGS maps usually report only the surface bedrock formation. For rocks quarried from beneath surface deposits or from deposits too localized to be depicted on USGS maps the chronostratigraphy data may be incorrect. When obvious erroneous results were produced further research was performed. This data will continuously be improved through further research and the submission of corrections from site users.

How to Use This Site



  • 1. U.S. Geological Survey, "Preliminary Integrated Geologic Map Databases for the United States," USGS Open File report 2005-1351 ver. 1.0 (USGS, 2005), Appendix 4. Available online at
  • 2. Kenneth L. Kelly and Deane B. Judd, Color: Universal Language and Dictionary of Names (Washington, DC: U.S. Dept. of Commerce, National Bureau of Standards, 1976). Available online at
  • 3. Pettijohn, F. J., E.P. Potter, and R. Siever, Sand and Sandstone (Berlin: Springer, 1973).
  • 4. International Commission on Chronostratigraphy, “International Chronostratigraphic Chart,” ver. 2013/01 (International Commission on Chronostratigraphy, 2013). Available online at