4] The material analysed above covers the slope of the hill below the spring for 100 feet or more. It is mostly calcite with magnesium carbonate and a minor amouat of alkalic carbonates. The sheets of impure magnesite such as are found at lake-level at both ends of Meadow lake, the calcite deposit on the benches in Kelly lake (Plate VII) and in the beds west of the lake, etc., prove, on the other hand, that precipitation from the waters of certain of these ponds actually has taken place. . Quantitatively, this mode of. origin is of minor import- ance, however, and the pure hydromagnesites have been formed almost solely by precipitation from ascending waters. Order of Deposition. The separate deposition of the hy dromagnesite and of the calcite and gypsum may be explained as due to the greater insolubility of the calcite and gypsum which would cause them to be pre- cipitated before the magnesian carbonates; the relative proportion of calcite to gypsum would depend on whether sulphates or carbonates were in excess. Johnston! has shown that in water containing only magnesium and calcium carbonates in solution under atmospheric conditions, most of the calcium will be precipitated before the magnesium. This general rule is illustrated by the relative compositions of the earths from 141 Mile House and the spring waters from which they are derived. The waters carry a large amount of carbon dioxide in solution and carbonates form the bulk of the precipitated earths. In the spring water the molecular proportion of magnesium is in excess of calcium by about 9 to 1, whereas in the earths precipitated from this water, calcium is in excess by 2 to 1 in one case and ~6tolinanother. Again, at Clinton, the underground waters moving from the high eastward ridge westward to the valley floor, lose some of their load of calcium at such places as localities 4, 5, 6, and 7, Figure 3, by forming through the agency of springs, nodular deposits of calcite and gyp- -sum (Plates V and VI and analyses 7 and 8, Table IV) but a certain propor- tion of the calcium remains in solution uatil the waters reach the flat at a locality such as No. 3, Figure 3, where nearly all the ground water rises to the surface as already explained. If for the sake of simplicity it is as- sumed that the amount of sulphate still present is very small, the fol- lowing process should take place. The checking of movement due to change of grade, the mixing with other solutions, the loss of carbon dioxide, _etc., would cause the precipitation first of the available calcium as calcite, followed as the chemical equilibrium changed by the deposition of the magnesium salts. The resulting precipitate near or at the surface would be porous, with more magnesium at the top than at the base. The depo- sit would grow upward and as the process continued and the deposit grew thicker, more calcium would be precipitated at the base and more nearly pure hydrous magnesium carbonate would be formed on top. If the waters forming these deposits could freely drain away, the more soluble elements, such as the sulphates and chlorides of magnesium, and the sulphates, car- bonates, and chlorides of the alkalis, sodium and potassium, would be almost. if not completely ‘carried away. Thus the combined amounts of the alkalis present in the spring water (Table V, analysis 1) are far in ex- cess of calcium and magnesium, whereas only insignificant amounts of these have been precipitated near its point of issue (Table VI) analyses a and b. Where such free drainage does not take place these more soluble salts are precipitated, see pages 54-56. 1Johnston, John, ‘“‘The soley product constant of calcium and magnesium carbonates.’ Jour. : ; Soc., vol. 37, No. 9, Sept., = e es." Jour. Am. Chem