involvement of people in astronomical facilities mandates safety requirements that have, in the past, proven to be expensive. Several experiments done during Apollo missions showed that contamination problems due to the presence of humans may be acute on the moon. Previous experience suggests that scientific facilities should be designed so that humans are called on to provide only limited, but critical, services.

Some measurements for which the moon may offer significant advantages compared with terrestrial or earth-orbiting instruments are discussed in this section. The concepts described below are illustrative; other promising possibilities may be developed as plans for a lunar base mature.

The early operation of a modest-sized (1-m-class) telescope would provide vital information concerning the design of future, more complicated, lunar telescopes, while providing unique scientific information. A small pointed telescope could observe individual objects or carry out wide-angle surveys, perhaps in the ultraviolet or infrared. A transit telescope with only a few moving parts could produce an imaging survey from the ultraviolet to the infrared over a large area of the sky using the slow rotation of the moon for scanning. The telescope would provide a high-resolution map of the universe at faint magnitudes.

A large-diameter telescope (16-m class) operating at infrared, optical, and ultraviolet wavelengths could have enormous scientific potential. As discussed in Space Science in the 21^st Century (NRC, 1988), such an instrument could detect earth-like planets around nearby stars and perhaps detect O₃ at 10 µm or O₂ at 1 µm in their atmospheres. Oxygen molecules are believed to be universally related to the presence of life in an atmosphere like our own. This very large telescope could also study the formation and evolution of galaxies by taking images and spectra of galaxies at large redshifts. The question of what is the best location, a lunar base or high earth orbit, is particularly problematic for such a telescope. As for other cases, the answer will depend on technological developments in the next decade and on the infrastructure that will become available to support orbiting and lunar observatories.

The techniques of interferometry can be used to link widely separated telescopes on the lunar surface to produce the spatial resolution of a single large telescope many kilometers in size. With an array of telescopes spread over a 10-km baseline, distortion-free images can be obtained for faint sources with