VKGA-KXFEDITI0SEN8 VETENSKAPLIGA ÄMBETEN. 



329 



unexceptionally. I therefore, as already mentioned in the 

 foregoing, consider 0° theoretically and practically to be the 

 natural starting-point of all volnmetric experiments connected 

 with arctic hydrography. Consequently I intend to refer the 

 observed numbers of specific gravity to the common standard 



TTTs C, notivitlistanding the areometer ivas adjusted conformahly to 



anotlter, vis. , ^ ~o C. 



The practical aavantages will best be seen from the fol- 

 lowing ^ example. 



Suppose the areometer to have attained its equilibrium 

 in the water. Then the weight of the instrument is counter- 

 balanced by the weight of the displaced volume of water — 

 therefore it sinks in the fluid to a certain mark and not 

 further. In order to obtain satisfactory results the glass body 

 of the areometer must have the temperature of the fluid. 

 Only in the case of the dilatation of the fluid being equal to 

 that of the glass it would be immaterial to our purpose, at 

 what temperature the measurement was performed, and the 

 readings of the areometer coukl be used without any correc- 

 tions whatever. The coefficient of dilatation of sea-water, as 

 may be seen from the tables and plates of the foregoing 

 paper, is highly dependant upon its amount of salt and like- 

 wise upon its temperature. At + 15° C and + 17°.5 C, the 

 points, to which ordinarily areometric measurements are 

 referred, the difference of the expansion of sea-water and of 

 glass is very great and necessitates careful corrections, but 

 at temperatures not exceeding a few degrees on either side of 

 _0° C we may totall}^ dispense with every correction depending 

 on the temperature and still obtain numbers representing the 

 specific gravity of the water with correctness in the fourth 

 decimal. 



The areometer used on the Vega and afterwards forwarded 

 for my inspection was graduated so as to allow observations 

 of 0.0001 of the specific gravity. For the present problems 

 of hydrography this approximation is quite sufficient, because 

 it enables us to calculate the saltness of the water unto O.oi %. 



* The foUowing application of the principle of Archimedes coiild 

 easily be disguised in a more mathematical garb. The advantage, liowever, 

 of more generalized formnlte would not in the eyes of the author compensate 

 for the loss of a plain manner of demonstration, intelhgible to every student 

 of hydrography. 



