These fell into several general categories.

First: there is general agreement that Mars, during its early days, was far more hospitable for the appearance of microbial life than it is now -- but was it hospitable enough? We know with near-certainty (from the erosion rates of ancient Martian terrain) that, until about 3.5 billion years ago, Mars had a dense carbon dioxide atmosphere -- maybe several times denser than Earth's atmosphere is today -- and that it then lost all of it over a period of about 500 million years.

But there is still a furious debate over whether Mars, during that period, ever had a surface temperature above freezing, so that large amounts of liquid water could exist on its surface.

Its branching "valley networks" that look like dried-up riverbeds have some structural peculiarities which suggest that they were carved instead by much slower trickles of ground water a short distance below the surface, which carved out tunnels whose roofs finally caved in -- a process called "groundwater sapping".

(Pascal Lee, in a Conference poster, repeated his claim that the Martian valley networks look like north Canadian channels carved by meltwater trickling along under a glacial ice layer.) And the much bigger "catastrophic outflow" floods that carved vast channels in Mars' surface were caused by eruptions of underground water from underneath Mars' thick permafrost layer (up to 2 billion years ago) that were huge but very brief, lasting only a few days before freezing (and later sublimating into vapor in Mars' new near-vacuum atmosphere).

In both cases, the underground water could have been geothermally heated.

Dense CO2 produces a strong greenhouse effect -- but during those early days of the Solar System, the Sun was about one-quarter less bright than it is now.

Calculations indicated that Mars' early air -- no matter how thick it was -- could not possibly have warmed its surface above about -43 deg C, because beyond a certain point a thicker atmosphere would just have scattered and reflected more sunlight back into space, both from the gaseous CO2 itself and from the thicker layer of frozen "dry ice" clouds that would form in it.

However, in 1997 a new twist was put on this idea, and atmospheric specialist James F. Kasting -- a central figure in the debate -- delivered a Conference talk on it.

Kasting's calculations had indicated that a layer of frozen CO2 clouds in Mars' upper atmosphere would significantly cool the planet, because they would reflect most of the sunlight that hit them back into space, but are relatively transparent to infrared radiation and so would not trap the heat radiation that Mars' surface gave off when it was warmed by the remaining sunlight.

But in 1997, Francois Forget and Raymond Pierrehumbert carried out new studies indicating that dry ice clouds are far better at reflecting IR radiation than had been thought.

They concluded that the thick dry-ice cloud layer that would form in a Martian atmosphere twice as thick as Earth's might actually warm the planet's surface to a toasty 27 deg C (80 deg F)!

Kasting took them seriously, but decided to do a still more thorough study of CO2 clouds, and reported it at the Conference.

His new conclusions fall in the middle: whether CO2 clouds have a net warming or cooling effect seems to depend both on how thick they are and what altitude they form at. If they have a high "optical depth" (visible-light opacity) or are at a low altitude, they can still cool a planet.

His overall conclusion is that we simply do not know nearly enough yet about the formation processes of clouds -- or even have accurate enough 3-D computerized climate models -- to know whether early Mars could have had a surface temperature above freezing.

(He did add that even a small trace of methane could have warmed Mars above freezing -- although such methane would itself have to be produced by Martian bacteria. Monika Kress added in a poster that methane might instead have been supplied by comet impacts on early Mars.)

And Conway Leovy repeated his own belief that Mars' remarkably smooth northern plains -- which some regard as sediment floors left over from an ancient Martian ocean -- were instead formed when the most of the surface ice on an early frozen Mars sublimated into vapor as the planet lost its early air pressure, and the remaining dust and sand were scoured away from the northern lowlands by its wind patterns.

Even if the surface of early Mars was mostly frozen, though, there is no question that it had far more volcanic activity than it does now, simply because the planet was still cooling down from its creation -- and the resultant hot springs could have provided a fairly large supply of near-surface liquid water that might still have allowed the evolution of life (especially since there is a growing feeling that such volcanically heated water supplies may have been the regions in which life first evolved on Earth).

An above-freezing Mars certainly would have been a major asset to the development of life, but it's not indispensable.