2.7d18O-29(r2=0.92). Its slop is 2.7 that is lower than the common range of evaporation lines, 3.5~6. Considering anion and tritium concentrations, a mixing of connate and meteoric water can be concluded. The data points of others sites are relatively invariable distribution and away from the meteoric water line on dD-d18O plot when a few artificial contaminated ones are eliminated. This represents that each site has its own characteristic water sources. Compared with d18O of clay minerals associated with spring waters collected from two mud volcanoes, we conclude that oxygen isotope exchanges between water and clay minerals occurred at different temperature, which is probably caused by different original depths.