Victor Hugo, the French author of "Les Miserables," the novel on which the popular musical is based, once wrote: "Where the telescope ends, the microscope begins. Who is to say of the two, which is the grander view?" At the present it seems that the advantage is clearly in the court of the microscope, since transmission electron microscopes and scanning tunneling microscopes have achieved a goal once considered impossible -- high-resolution photographs of individual atoms in molecules and crystal lattices at magnifications of 10 million times.

Telescopes, on the other hand, have until recently shown little real improvement since 1948, when the Hale Reflecting Telescope was mounted on Mount Palomar in California. The Hubble Space Telescope was expected to reveal secrets of the universe to rival those found by Galileo when he first turned his telescope to the stars in 1609. But the Hubble uses a smaller mirror (94.5 inches compared with the Hale's 200 inches) in the same optical configuration as the Hale and is in this sense a very old-fashioned instrument. It is, of course, surrounded by state-of-the-art detectors, image processors and computers and, positioned above the Earth's atmosphere, avoids the image distortion caused by atmospheric fluctuations.

It is terribly unfortunate that a still-mysterious defect has annulled many of the Hubble's hoped-for advantages by severely limiting the telescope's focusing ability. It is even more tragic for the careers of many brilliant astronomers who have invested all their energy and creativity in this ambitious project for so many frustrating years.

Perhaps some of the difficulties that have plagued the space telescope could have been foreseen and avoided if a different approach had been taken. In 1985 Freeman Dyson, the well-known physicist and astronomer, delivered the Gifford Lectures in Aberdeen, Scotland, (published in 1988 by Harper and Row as "Infinite in all Directions"). In these far-ranging lectures, Dyson wondered if the Hubble project was not putting all its eggs in one basket and thus courting disaster.

Dyson suggested his preference for a series of three or four smaller telescope launchings, starting perhaps with a 40-inch-diameter mirror and gradually increasing the mirror's size and modifying its design as experience and progress in science and technology dictated. This would reduce the risk of a very expensive project yielding an incommensurately small scientific return. He foresaw the clear possibility that, even if the Hubble were a glorious success, it would wind up in the 1990s 20 or more years out of date scientifically. This is inevitable in any large-scale project like the Hubble -- controlled as it is by budgetary restraints and uncertainties, congressional insensitivity and infighting and the inertial impact of government bureaucracy.

Evidence is accumulating that Dyson's predictions were right on target. A group of European astronomers has developed "adaptive optics" techniques that eliminate much of the atmospherically caused fuzziness in telescopic images of stars and galaxies. They believe that they can achieve image resolutions better than that designed into the space telescope (but not achieved), and this with land-based, computer-corrected instruments. If their optimism turns out to be justified, one of the main arguments for the Hubble (although by no means the only one) loses much of its cogency.

In addition, a whole new generation of land-based, reflecting telescopes with effective mirror diameters above 236 inches is now in the design and construction stage. These use arrays of smaller mirrors synchronized by computers to achieve greater resolution and light-gathering power than is possible with any single-mirror, Earth-based telescope.

The W. M. Keck Observatory 10-meter (394-inch) reflector under construction atop Hawaii's Mauna Kea contains 36 hexagonal segments, each 1.8 meters (71 inches) across at its widest diameter, which are computer controlled to focus light from all 36 mirrors to a single point. This multiple-mirror approach could be extended to achieve greatly improved resolution by extending into the visible region of the spectrum interferometric techniques developed by radio astronomers for the microwave region, in which extended arrays of phase-sensitive detectors have almost completely replaced large-dish antennas.

These recent developments are proof that scientific progress always outpaces scaled-up, brute-force technology of the kind used in the Hubble project. Perhaps we should give some thought to Dyson's prescient ideas before scarce dollars and precious lives are risked to upgrade the Hubble's disappointing performance. Since any modification of the Hubble's primary mirror will not be possible before 1993, we have adequate time to ponder and profit from the ill-fated saga of the Hubble Space Telescope.

The writer is professor emeritus of physics at the University of Maryland Baltimore County.