Objective prism spectroscopy is a type of astronomical spectroscopy, used for simultaneously measuring the spectra of many bright objects within the field of view of the telescope.
I assume you, the reader, have seen a prism before -- a piece of glass, quartz, or other transparent material, shaped so that light passing through it is dispersed depending on its wavelength. White light passed through a prism forms a rainbow as the different wavelengths of light composing "white light" are dispersed in different directions. Now, instead of passing light from a prism onto your wall or a piece of paper, pass it through a telescope instead. In doing so, you will image not the objects you're pointing the telescope at, but the dispersed spectra of these objects instead. This is objective prism spectroscopy.
First devised in the 18th century, the objective prism spectrograph is the simplest design possible; a prism large enough to cover the aperture of the telescope is placed in front of the objective lens (or the aperture in front of the secondary mirror, if the telescope is of the reflecting type). Light from the stars and galaxies in the field is then dispersed by the prism, and the dispersed light is imaged by the telescope. Unlike other spectrographs, light does not need to be passed through any slit prior to being imaged because the stars and other objects being measured are assumed to be point sources. This method of spectroscopy is very efficient for imaging large fields or performing large surveys, and the technology is simple enough to be built and used by amateur astronomers and advanced hobbyists (1).
The first documented objective prism spectrograph was built by the Frenchman Alexis-Marie Rochon in the 1770's (2). However, the development of the spectrograph rapidly accelerated as the use of stellar spectra as a tool for performing astrophysics expanded in the mid-to-late 19th century. Perhaps the most famous objective prism survey of the sky was the Henry Draper Catalog(3), undertaken by Annie Jump Cannon and Edward C. Pickering, the former Curator of Astronomical Photographs and the latter Director of the Harvard College Observatory. Begun in 1893, the catalog took decades to complete. Each region of the sky was imaged by an objective prism telescope, and the images registered onto photographic plates. Each individual spectrum on each plate was then analyzed (by eye, mostly by Cannon), and the stellar classification of each object made based upon the pattern of spectral lines observed. At its conclusion, the Henry Draper Catalog contained 225,300 stars.
As astronomers expanded their interest to fainter and fainter objects in the universe and desired more and more detail in the spectra, most objective prism spectrographs have been replaced with other kinds of multi-object spectrographs. Two noteworthy types are the fiber-fed spectrograph, and the multi-slit spectrograph. In the former, light from the focal plane is fed via several hundred optical fibers positioned on objects within the plane to a bench-mounted spectrograph. The resulting high-dispersion spectra are then imaged electronically. In the latter, the classic single spectrograph slit is replaced with a slit mask -- a metal plate containing many small, precut slits, each positioned to coincide with an object in a particular field. Light passing through the slits is then dispersed via prism, echelle, or diffraction grating to produce an image of all the spectra from the different slits. These methods are useful for performing high-resolution spectroscopy of faint objects (such as galaxies at high redshift), but have the drawback of being relatively difficult to plan and execute. With both methods, either the optical fibers or the slits must be positioned ahead of time, based upon the positions of the objects within one single field. The positioning and cutting of slits in slit masks is time-consuming, and the optical fiber assemblies are mechanically very complicated and expensive to build.
While objective prism spectroscopy does not provide very high spectral resolution and the ability to image faint objects, the technique is far less complicated to perform, and is still used for large-area surveys. Because of the low dispersion of most objective prisms, they are typically used for performing rough stellar classification and cosmological redshift measurements, and searches for exceptional objects (like emission line quasars). They are most efficient when used to observe bright sources (like stars and quasars), though large-aperture telescopes such as the 1.28-meter UK Schmidt telescope in Australia can and do obtain objective prism spectra for objects as faint as magnitude 20 (4).
References:
(1) There is an excellent guide to objective prism spectroscopy (and pictures, too) for amateur astronomers by Richard Hill of the Lunar and Planetary Lab of the University of Arizona, available at:
http://www.lpl.arizona.edu/~rhill/spect/spect.html
(2) P. Abrahams: Early Instruments of Astronomical Spectroscopy, http://home.europa.com/~telscope/histspec.txt
(3) Colorado: The Henry Draper Catalogue, http://lyra.colorado.edu/sbo/sboinfo/readingroom/hd.html
(4) G. Walker, Astronomical Observations: An Optical Perspective, Cambridge University Press (New York: 1987)