In a carefully sealed chamber at the International Bureau of Weights and Measures in Paris, they keep the original standard kilogram. In the USA, the National Institute of Standards and Technology keeps a copy of it, and so do most other countries. But the real, true kilogram is that one in Paris.
Paleontologists have a similar system for keeping fossil species straight. A researcher publishes a formal description of a new species, giving it a name and describing it in great detail. But the real standard is an actual physical fossil, the type specimen. Type specimens are carefully preserved in museums and university collections, where any researcher can go see them.
What's In a Type Specimen
It may seem, I don't know, unscientific to do things that way. Imagine picking one person to be the "type human." Ideally the researcher wants many specimens of a species that show its whole range of shapes and size, so you could establish a sort of ideal picture of the species. But the first order of business is to give it a name and a starting standard, to avoid confusion and establish priority.
When picking the type specimen, the paleontologist isn't choosing the ideal individual, or the biggest, or the most anything, just the most typical one. It has all the important features, with nothing that's unusual for the species. The type specimen might not be perfect or even complete. And what if the fossil is just a pollen grain, or a seed or an egg, or a caterpillar whose adult form is a moth? These are problems for later researchers to resolve.
The type specimen for Tyrannosaurus rex is in the Carnegie Museum of Natural History in Pittsburgh, Pennsylvania. And the University of California has put a bunch of type specimens for microfossils online. So has a group of British institutions. They're pretty things.
In the 200 years or so this system has been used, it has spawned its own set of words.
- To name a new species, a researcher writes a paper with a full description of the species, comparing it with similar species and noting the differences. The type specimen that goes with this document is the holotype.
- If two or more specimens are necessary, they're called syntypes, and any supplementary specimens are paratypes.
- If a correction or extension to the species is published, that's based on a hypotype.
- If a better specimen is found to replace the holotype, that's a lectotype.
- If the holotype is lost or destroyed, a neotype is chosen.
All of these items are permanently preserved and their whereabouts published in the literature. Oh yes, and a copy or cast of these reference materials is called a plastotype.
This system of types also works for a new genus, or group of related species. The worker who names a new genus uses a type species. And similarly, a type genus is used to define a new family, or group of related genera. (I don't think the system goes higher than that, to the orders and classes and phyla, but it might.) These definitions ultimately rest on the type specimens that define each species.
Geologists themselves use the type concept. For instance, the "rock species," or lithostratigraphic units, that you see on geologic maps, are based on specific outcrops.
Every formally defined rock unit, like the Inwood Marble under Manhattan or the Pierre Shale of the Midwest, is based on a type area. That is a place where the rock unit is well exposed, particularly its bottom and top, and where the typical rocks found in the unit appear. A type locality within the type area—maybe a town, or an estate or a landmark—gives the rock unit its name. And at one spot in the type area, the geologist describes in formal detail a type section—a sequence of rock strata that becomes the standard for that unit forever.
This convention is even used for rock types and tectonic features, although not as rigorously. The rock type called rapakivi granite is named for a specific locality in Finland, the California Coast Ranges are sometimes called the type area for subduction rocks, and the Hawaiian Islands are called the type hotspot. This is more a sign of respect for precedent than a formal designation.
PS: Where—and who—is the type specimen for the human species, Homo sapiens? When Carl Linnaeus first described our species in 1758, he didn't declare a type specimen because it was not the accepted practice at the time. There is a common belief that the paleontologist Robert Bakker formally declared the skull of Edward Drinker Cope as the lectotype. When Cope, himself a great paleontologist, died in 1897, he willed his remains to science, and they are held by the University of Pennsylvania. I believed this myself. However, it was convincingly argued in a 1959 paper that Linnaeus himself should be the holotype. I am indebted to guest poster Greg Mayer on the blog "Why Evolution Is True" for clarifying this point.