wells: ARAGONITE FROM SEA- WATER 25 1 



carbonate in sea-water. The results of the determination are as 

 follows. The pu+ value, from which the hydrogen-ion con- 

 centration of the water may be calculated, was found to be 8.18. 

 The titration alkalinity of the water was found to be 0.0019 

 normal, and the alkalinity caused only by carbonates and bi- 

 carbonates 0.0017. I^^e total carbon dioxide was 0.057 gram 

 per liter. 



The titration alkalinity and total CO2 of this water are lower 

 than usually found in sea- water. What is really required, how- 

 ever, to show the degree of saturation with calcium carbonate is 

 an evaluation of the expression [Ca + + ][C03 ], in which the 

 brackets represent concentrations of the ions indicated. In the 

 present state of physical chemistry it is extremely difficult to 

 make this evaluation with certainty. Nevertheless, by a series of 

 approximations, the writer has calculated that the expression 

 indicated has the value 7.2 X io"^°at25°C. for the water under 

 discussion, as compared with 19. i X lO"^*^ for surface sea-water 

 studied at the Marine Laboratory of the Carnegie Institution at 

 Tortugas, Florida. Johnston^ gives for the solubility-product 

 constant of calcite at 25 ° : i^ = 87 X io~^''. 



The above values agree well in view of the uncertainties attend- 

 ing the calculations for sea- water. As a matter of fact, however, 

 the solubility-product constant of aragonite is generally thought 

 to be greater than that of calcite, so that the results do not fall 

 in the right order. This may possibly be explained by the lack 

 of information about the temperature to which the tubes con- 

 taining the standard water had been exposed, or it may indicate 

 that some error exists in the method of calculating the ionic 

 concentrations involved. Or, lastly, the results found may 

 indicate a very slow time adjustment of the solubility. In other 

 words it may be that the solubility-product constant obtained 

 after the course of a very long adjustment is lower than the value 

 deduced from experiments made on fresh material over relatively 

 short intervals. 



It seems inadvisable to enter here into a discussion of all the 

 rules and assumptions used in making the above calculations. 



3 Joum. Amer. Chem. vSoc. 38: 982. 1916. 



