River water quality is of considerable relevance to the geographer and a reconnaissance survey of the spatial variation of the specific conductance of stream water within the basin of the River Exe was undertaken by the authors. The spatial pattern defined by the more than 500 samples collected was mapped and an attempt has been made to interpret the results. Much of the variation was closely related to lithology, but land use was also found to exert a significant influence on levels of conductivity. The data collected have also been used to obtain estimates of the magnitude and spatial pattern of annual solute loads and of chemical denudation rates. DURING the past 25 years, increasing attention has been focused upon spatial and temporal variations in the magnitude of streamflow or river discharge in Great Britain. Regional water resource studies have assessed the volumes of run-off available in particular areas, suggesting ways in which water can be stored and regulated.1 Furthermore, the problems of floods and droughts have stimulated the development of techniques for evaluating and predicting floods and low flows.2 Data on the flow of rivers in the British Isles are now available in the Surface Water Yearbook3 and additional information is being collected from several research basins.4 Geographers have been aware of the opportunities and relevance of this field of study and they have themselves made contributions.5 Much less information is available on the quality, as distinct from the quantity of streamflow. Detailed work in this sphere has previously been the province of the chemist and public health engineer and has been primarily concerned with the potability of sources of water supply and the nature and extent of serious pollution and the various possibilities available for the control or alleviation of this problem.6 With the increasing concern over the optimum use and management of water supplies in this country, more interest is being directed towards collection and interpretation of water quality data. River temperatures have already attracted the investigations of several workers,7 and studies of nitrate levels8 and solute transport by local streams have also been made.9 The quality of river water is not without relevance to the geographer, as Douglaso1 has recently emphasized. Water quality constitutes a valuable environmental indicator, reflecting the interaction of the quality of the incident precipitation with the character of the catchment area, the processes operating in the drainage basin, and the influence of man. It is in this context that Perel'manll has promoted the concept of geochemical environments or landscapes whereby geochemical indicators can be used for landscape classification or regionalization. In turn, water quality is an important control on human activity and resource utilization.12 Minimum quality standards exist for the various uses of water13 and severe negative feedbacks can occur where man-induced deterioration in water quality inhibits continued or more intensive use of water I55 This content downloaded from 207.46.13.111 on Sun, 07 Aug 2016 05:15:25 UTC All use subject to http://about.jstor.org/terms D.E.WALLING AND B.W.WEBB resources. Because of the significance to the operation of geomorphological processes, physical geographers have already made considerable progress towards understanding the solution chemistry and water quality of limestone areas.14 However, there is considerable scope for studies in non-limestone areas. For example, spectral analysis has been shown to be a powerful tool when studying temporal variations in solute concentrations,15 and the evaluation and interpretation of spatial variations in river water quality would seem to be another worthwhile theme for investigation. There are many indices of river water quality including, colour, odour, temperature, turbidity, solute content and other chemical and biological parameters, and these will exhibit spatial and temporal variations to a greater or lesser extent. Solute content and the concentration of individual ions provide a sensitive environmental indicator. For example, Miller16 has shown how solute concentrations can vary according to rock type; classic studies in the Hubbard Brook experimental catchment in New Hampshire, U.S.A.17 have stressed the interdependence of concentrations of individual ions and the nutrient cycle; and Skakalskiyl8 has demonstrated how solute concentrations can be related to the magnitude and origins of different run-off components. In addition, many reports exist of the influence of man on these parameters.19 Temporal variations in solute concentrations have been well documented by workers such as Hendrickson and Krieger20 and Hall,21 and spatial contrasts have been discussed on a world scale by Gibbs,22 on a continental scale by Rainwater,23 at the regional level by Douglas,24 and within a drainage basin by Anderson and George.25