402 



column, do not keep unaltered and are not always trustworthy, 

 because they become infected by bacteria. All figures regarded 

 as untrustworthy are put in brackets. 



The tensions of the sublimate-samples are arranged in the 

 second column, marked S. The sublimate causes an increase 

 of the tension, probably because some monocarbonate of mer- 

 cury is formed, and the differences between the reliable figures 

 in the two columns are therefore given in the third. These 

 differences vary only between 0.25 and 0.4, their average 

 being 0.35. This average is taken as the specific increase 

 in tension produced by the chloride of mercury; the figures for 

 the sublimate-samples are diminished by 0.35 and the final values 

 for the tensions at 12.5° are computed and given in the last 

 column. 



The second table contains the dates and positions of the 

 samples ; the state of the weather at the time of their collec- 

 tion, and the salinities, temperatures and COo-tensions of the 

 waters in situ. 



