vents. Temperatures of these springs ranged from 35 to 80°C, 

 but cooled very rapidly upon mixing with ambient lake water at 12 

 to 15°C. These waters were clear, anoxic and contained elevated 

 concentrations of nutrients and major dissolved anions and 

 cations. Because of the relatively low flow rates of these 

 vents and immediate mixing with lake water, pure hydrothermal 

 solutions were not obtained. Figure 4 shows the concentrations 

 of a variety of dissolved constituents as a function of the 

 estimated temperature of the hydrothermal fluids for a set of 

 samples collected in 3 to 10 meters of water in Sedge Bay. Lack 

 of a good thermal stoichiometry may result from both different 

 sources for the waters collected and the difficulty, with the 

 simple techniques employed, in obtaining an accurate temperature 

 in situ without cooling. The chemical stoichiometry is somewhat 

 more coherent (Fig. 5) with most dissolved constituents 

 covarying linearly. 



The chemistry of hot springs has been used extensively to 

 predict subterranean conditions in Yellowstone and other 

 geothermal systems (Mazor and Thompson 1982, Fournier 1979, and 

 others). Using chemical geothermometers, Fournier and co- 

 workers (1979, 1974a, b, 1973, 1970, 1966) have estimated the 

 underground temperatures of the source reservoirs for hot springs 

 and the temperature of the last rock-water interaction. Hot 

 springs often emanate from hydrologically complex systems which 

 include intermediate reservoirs, mixing with water from other 

 sources, and both conductive and adiabatic cooling. Enthalpy- 

 chloride relationships have been shown by Fournier (1979) to be 

 useful in resolving some of these complexities. The hydrothermal 

 fluids collected to date, however, are low in chloride 

 (bicarbonate being the principle anion), thus confounding the 

 application of these relationships. Application of the two most 

 commonly used chemical geothermometers, silica (Fournier and Rowe 

 1966) and Na-K-Ca (Fournier and Truesdell 1973) result in maximum 

 estimated temperatures for the Sedge Bay sublacustrine springs of 

 150 to 180°C and 196 to 211°C, respectively. The higher 

 temperature is generally considered more reliable, however, 

 because of the possibilities for mixing with multiple reservoirs 

 and heating by steam and hot gases, the reliablility of these 

 estimates is currently unknown. 



Concentrations of almost all constituents in these warm 

 waters were significantly elevated above that of ambient lake 

 water (Table 1). The nutrients: dissolved silica, ammonium, and 

 total inorganic carbon were as much as 2 orders of magnitude more 

 concentrated in the warm hydrothermal waters collected in Sedge 

 Bay than in the overlying lake water. Yellowstone Lake is 

 characterized as an oligotrophic system and is assumed to be 

 nitrogen limited, although no year-round data on the chemistry of 

 the lake currently exists. Dissolved inorganic nitrogen is 

 undetectable in the epilimnion and a mid to late summer bloom of 

 the nitrogen-fixing alga, Anabena , is an annual phenomena (R. 



89 



