Introduction

The origin of fluvial valleys has been one of the great problems of science ever since the debates in the eighteenth and early nineteenth centuries over the role of cataclysmic events in the shaping of Earth's valleys. The famous founders of geology, including Hutton, Lyell, and Cuvier all participated in these great debates. It can be argued that geology emerged as a science because of the scientific reasoning that was applied to this problem (Davies, 1969).

It was one of the great surprises of modern planetary science that fluvial valleys were discovered on the planet Mars by study of the vidicon images returned by the Mariner 9 spacecraft (Masursky, 1973). From the much more extensive coverage of the Viking mission we know that the heavily cratered Martian highlands are locally dissected by networks of tributary valleys with widths of 10 km or less and lengths of a few kilometers to nearly 1000 km. These valley networks are one major type of large elongate Martian troughs thought to have fluvial origins (Baker, 1982).

In distinction from fluvial valleys, large outflow channels show extensive evidence of large-scale fluid flow on their floors and walls (Baker et al., 1992a), so technically speaking they are channels, rather than valleys. This is a rather curious circumstance, since Earth experience leads us to suppose that fluvial channels are invariably much smaller than fluvial valleys. However, the Martian outflow channels may arise from the peculiar geological history of Mars and its water endowment (Baker et al. 1991).