170 



V. WALFRID EKMAN. 



an accuracy of about 10 ^/o, which may be considered quite satis- 

 factory. 



As regards the range of velocities for which the apparatus 

 can be utilised, this will appear from the following table giving 

 those velocities of the current, which, with different positions of 

 the sliding weight give a pendulum-displacement of 1 and of 

 10 cm. These velocities may be considered to constitute the 

 limits for the instrument's field of application. 



Sliding Weight's 



M 

 3630 



Velocity corresponding to 

 Displacements of 



Magnitude 



. Position 

 below Zero 



1 cm. 



10 cm. 



150 gr. 



1 cm. 



0-0375 



0-4 



1-2 



17 n 



5 „ 



0-187 



1-0 



3-5 



« V 



25 „ 1 



0-935 



2-3 



8-0 



762 „ 



39 „ ' 



7-45 



7-2 



22-8 



It will be seen that great care is required in the arrangement 

 of the observations when the velocities are less than 4 mm. per 

 second; and 23 cm. per second is the greatest velocity that can 

 be measured. The pendulum might be exchanged for one with 

 smaller wings, but in fast currents the apparatus also requires 

 to be weighted considerably in order to keep it in a vertical 

 position. I have, however, observed that even in currents with 

 a velocity of as much as 20 cm. per second and more, the pen- 

 dulum is displaced quietly and regularly without any oscillations. 



In concluding this first part of my account, I will now for 

 the sake of completeness, explain the determination of the above 

 constants and formulae for the use of the instrument. 



The first problem was to determine the zero of the sliding 

 weight, and its dependence upon the density of the water. For 



' About the lowest possible position in salt water. 



