ON A NEW CURRENT-METER. 171 



tliis purpose the pendulum was detached from the apparatus, 

 and immersed in a horizontal position in fresh water, where- 

 upon the sliding weight was moved until it was no longer possible 

 to decide whether the bob of the pendulum sank or floated up 

 when the upper end of the pendulum was suspended. The sliding 

 weight was thereupon screwed fast, the pendulum taken up 

 and dried, and its centre of gravity determined and found to lie 

 39*0 cm. below the point of suspension. As the combined weight 

 of the pendulum and the sliding weight is 579 gr., the moment 

 in air was equal to 39 X 579 = 22600 cm. gr., and this is there- 

 fore the moment of the displaced water, whose density may be 

 considered as equal to 1. 



An increase of 0001 in the density of the water thus re- 

 quires an increase in the moment of the pendulum of 22*6 cm. 

 gr. which, as the sliding weight weighs 136 gr. in water, an- 

 swers to a lowering of the weight by 0'166 cm^. 



The temperature will also have an influence through altering 

 the volume of the glass float. As the cubical expansion-coefficient 

 of glass is 0'000025, this influence is found to go up only to an al- 

 teration of 0*004 cm. in the position of the zero-point for 1° G. The 

 influence of the temperature upon the density of the water then 

 becomes more important, and acts towards the opposite quarter. 

 An increase in the temperature of sea-water from 0° to 10° and 

 20°, causes a decrease in the density of 0-00113 and 0*00328 re- 

 spectively, and thus a raising of the zero-point by 0*19 and 0*54 cm. 

 respectively. In a new instrument, it should be possible to reduce 

 these quantities considerably, as the pendulum can be made 

 lighter, and thereby also the glass ball smaller. This, however, 

 is of little importance, so long as the instrument is not adjusted 

 for great sensitiveness; and one can, as a rule, be content wdth 



' Tlie buoyancy of the sliding weight in the Avater is here talien ^into 

 account, but not its alteration witli the specific gravity of the water. 

 The error thereby occasioned in the zero of the sliding weight never 

 exceeds, however, a tenth of a millimetre. 



