366 



prefer to confine myself to the above remarks and to the curve 

 and table which are purely empirical. 



By means of the constants determined in this chapter, viz. 

 The coefficients of absorption, the constant of dissociation at 

 15° and the additional corrections for the temperature, it ought 

 to be possible to utilize the determinations of alkalinity and 

 total CO2 in seawater, that are found in literature, for the 

 purpose of tension-determinations. For several reasons 1 cannot, 

 however, recommend such a course. Firstly because my 

 "Standard« seawaler has proved itself to be, in the course of 

 the investigation, rather abnormal with regard to the alkalinity, 

 but still more so, because the older determinations are rather 

 unreliable. The alkalinity is, as a rule, very accurately deter- 

 mined, but it is by no means certain that the quantities of 

 carbonic acid are always directly comparable with the alkalinity. 

 Rather often, 1 think, have the cubic-centimeters or milli- 

 grammes of CO2 been slightly different from the corresponding 

 units for the alkalinity, and even very slight differences of this 

 kind will prove disastrous for a computation of the tension. I 

 am inclined to think that some error of this kind must have 

 infected the otherwise so remarkably accurate experiments of 

 Hamberg. At least they do not agree, when treated theoretically, 

 either with my own or indeed among themselves. 



Nearly all the investigators have stored the water-samples 

 for some time before analysing them, and we cannot be sure 

 that they have avoided infection or diffusion of carbonic acid 

 either to or from the atmosphere. The water-samples of the 

 Ingolf-Expedition were certainly examined immediately, but 

 Knudsen expresses the opinion that the values of the quantities of 

 carbonic acid found are inaccurate, owing to absorption in the 

 analysis-apparatus. (The Danish Ingolf-exp. Hydrography 

 p. 84). 



