HOUSTON -- In the old mansion that houses the Lunar and Planetary Institute, next door to NASA, Buck Sharpton, a scientist in a tweed jacket and blue jeans, spreads a map of the world marked with 116 small circles. It tells a story of past excitements and denotes a current controversy.

The hotly disputed hypothesis is that the evolution of life on Earth has been marked by radical disjunctions, particularly one 65 million years ago, because of collisions of Earth with comets or other extraterrestrial material. What is not disputed is that 65 million years ago there were mass extinctions that included dinosaurs.

It is unclear how abrupt the disappearance of the dinosaurs was, and in any case, they were cumbersome and not numerous, so they were inherently vulnerable. However, there are at least five models of how the extinctions could have occurred.

The impact might have churned up dust sufficient to block sunlight, impede photosynthesis and thus cause starvation of many creatures. Or the dust might have caused a glacial episode by screening sunlight. Or perhaps the opposite occurred: dust and water vapor caused a ''greenhouse effect,'' holding in heat and cooking much life out of the planet. Or the entering entity might have heated the atmosphere around it enough to generate nitric acid and worldwide acid rain. (However, it would have been traveling up to 90 kilometers a second, so it might not have spent enough time in the atmosphere to alter the atmosphere.) Or the entity could have carried a toxic agent. (Cyanide, for example, is common in comets.)

Tantalizing evidence abounds. In a geological formation in Italy, there is a knife-edge-thin layer of clay between large slices of gray limestone that may be a residue of a catastrophe. It contains 30 times the normal amount of iridium, an element abundant in meteorites. Similar formations have been found in Spain, New Zealand and Denmark. Furthermore, there is a suggestive distribution of 65-million-year-old ''shock crystals'' that could have been formed and scattered by a collision.

An impact by a body at least 10 kilometers in diameter, creating a crater 200 kilometers in diameter, would be necessary to create the iridium anomaly. But if that happened, where? Geological analysis has ruled out Hudson Bay, the mouth of the St. Lawrence and the curved western shore of the Gulf of Mexico as impact structures.

Anyway, gravitational stresses (gravity distorts the moon's shape) probably would break up something that size as it approached Earth. So scientists are interested in ''paired'' craters, such as several pairs that exist in the Soviet Union. When the two Soviet pairs are plotted in relation with something interesting in Iowa, the five phenomena fall within 12 degrees of an arc across the arctic, perhaps the path of impacts from a disintegrating body.

Iowa is, of course, the lever that moves the political world. It also once was the site of an event even more earthshaking than the nominating caucuses: an impact that produced the ''Manson structure,'' about 25 miles in diameter near Manson.

The water under most of Iowa is hard because of the limestone through which it is filtered. The water within the Manson structure is soft because (so the theory goes) the limestone was blasted away by the impact and replaced by other material churned up from a great depth.

Craters become buried or otherwise obscured, planed off by erosion and filled in by sedimentation as a result of wind, rain and other surface activities (including, in Iowa, the heavy tread of herds of candidates). The Manson structure is not visible even from high overhead.

Some scientists argue, vehemently, that there are no known events in the natural history of the Earth that cannot be explained in terms of the Earth's own activities, such as volcanoes. Furthermore, there recently has been a discovery of many dinosaur fossils at high altitudes along Alaska's North Slope. Temperatures there were milder in the dinosaurs' days than today, but the fossils prove that dinosaurs lived in a place where there was the total darkness of a polar winter for months at a stretch, longer (it is surmised) than the screen of sunlight caused by any comet's impact.

The scientific inquiry into catastrophes and extinctions is of no immediately practical application, is in large measure your tax dollars at work and should be pleasing to you.

From the inquiry, we will learn much about the evolution of the universe, and already it is intimating two interesting, even moving, ideas.

One is that Earth is not just a cinder in a void: there is considerable interaction among entities in our solar system. And the fact that life of any sort has survived the Earth's episodes of violence is a stirring testimony to the imperative of life.