GEOPHYSICAL LABORATORY. 147 



products which are more ferric than this. The pressure-composition isotherm 

 at 1100° confirms that at 1200°. 



The major portion of the oxygen pressure curve of the system at 1200° hes 

 between the limits 4 mm. and 1 mm. The pressure drops rapidly near Fe304 

 and rises rapidly near Fe203. 



Since the dissociation of Fe203 results in the formation of a solid solution, 

 the pressure of oxygen and the composition of the solid phase depend upon the 

 relation of the weight of the charge to the volume of the space into which the 

 oxygen dissociates. This fact accounts for the variety and uncertainty of 

 results heretofore obtained in experiments on the dissociation pressure of 

 FeaOs. 



(11) The dissociation of ferric oxide in air. J. C. Hostetter and R. B. Sosman. J. Am. 



Chem. Soc, 38, 11S8-1198 (1916). 



Previous work by the authors has shown that Fe203 dissociates to form a 

 solid solution of Fe304 in Fe203, and that the curve of dissociation pressure 

 against composition at a given temperature rises rapidly as the composition 

 approaches pure Fe203. The present experiments show that there is a 

 measurable dissociation of Fe203 in air at all temperatures between 1100° and 

 1300°, and that the amount of dissociation increases with the temperature. 

 This is shown by the increasing difference in weight between ignitions in air 

 and in oxygen as the temperature is increased. The dissociation pressure- 

 composition curve thus takes the form of a curve asymptotic to the axis of 

 ordinates, when the ordinates are pressures. 



The best container for the Fe203 at 1100° and 1200° is alundum (bonded 

 fused alumina) which is almost absolutely constant in weight at these tem- 

 peratures, although it loses weight steadily at higher temperatures. The loss 

 in weight of pure platinum at 1000° to 1200° is very small, but is considerably 

 increased if the platinum is in contact with ferric oxide. 



(12) The determination of carbonic acid, combined and free, in solution, particularly iu 



natural waters. John Johnston. J. Am. Chem. Soc, 38, 947-975 (1916). 



Owing to the importance, especially in connection with water analysis, of 

 a knowledge of the concentration of carbonic acid, combined and free, in 

 solution, a great deal of attention has been devoted to methods of determina- 

 tion of these constituents; but the question as a whole has hitherto received 

 scant attention, in particular from the theoretical standpoint. The present 

 paper discusses the methods of estimating carbonic acid and carbonate on the 

 basis of fundamental principles; this enables us to criticize and coordinate 

 apparently contradictory statements recorded in the very voluminous liter- 

 ature on this subject; for this conflict is due less to lack of care in the experi- 

 mental work than to the fact that some essential factor, the importance of 

 which, however, would not be recognized until the theory had been considered, 

 was not adequately controlled. 



Within any solution containing carbonate there is a readily attained equi- 

 librium between the carbonate ion COi", the bicarbonate ion HC07, and the 

 carbonic acid H2CO3, and in turn between the carbonic acid and the partial 

 pressure of carbon dioxide above the solution; consequently these molecular 

 species can coexist only in definite proportions determined by the several 

 equilibrium constants. An examination from this standpoint of the most 

 commonly used titration methods for the estimation of the combined and free 

 CO2 in solution leads to the conclusion that many of these procedures do not 

 yield definite results — a conclusion corroborated by all of the careful com- 

 parative experimental work bearing on these methods. In principle the only 

 absolutely reliable methods are those for the total base combined with the 



