172 V. WALFRID EKMAN. 



taking the salinity into consideration. At 0° C. the density of 

 the water is 1 + 0-00080 S {S = o/oo salinity). Furthermore, in 

 fresh water of 7° or 8° C. which has almost the same volume as 

 that of 0°G., the position of the zero-point was found to he 12'9 cm. 

 (mean of 12-9, 12-96, and 12-85). For water of 0° C. there- 

 fore, as stated above, the distance of the zero below the upper 

 end of the pendulum is 



z = [12-9 + 0-133 S] cm. 



As regards the determination of the relation between the 

 velocity and the pressure of a current, the accuracy with which 

 it was made, was not, as already mentioned, very great, and the 

 determination can only be regarded as provisional. 



The first experiments Avere made in a water-hole in the 

 ice. The instrument was suspended from a beam, and was 

 moved along the water-hole in a track of known length. The 

 time was taken with an ordinary watch. 



In order to eliminate error arising from small currents in 

 the water, or from an irregular position of the apparatus, I went 

 alternately to one distance and another, with as far as possible 

 the same speed, and the mean of both velocity and pendulum- 

 displacement was taken for every two such observations. The 

 values thus obtained are placed on the diagram (PI. XVI). 



Special experiments were made in a bath-tub with the lowest 

 velocities. The apparatus was moved through the water by a 

 wooden screw, which was turned in time to the ticking of a 

 watch. 



In these experiments a source of error made itself apparent 

 in the instrument, which had hitherto escaped my notice, and 

 must unfortunately have diminished the accuracy of the pre- 

 vious observations. During the sensitive pendulum-adjustments 

 employed — the sliding weight 1 or 2 cm. below the zero — the 

 pendulum had a considerable displacement by itself, and it was 

 found that this must be due to some small asyunnetry of the 



