243 



When the cap. tube was taken away the number of 

 sparks in the second series was 17,i against 20,3 when the 

 tube was present. 



These res alts are put in a curve a I c Pl. II for the 

 quantity of water and a curve cl e f for the intensity of 

 current, in the same manner and with the same values for 

 abscissa and ordinata as before, only that the zero-point 

 for the intensity-curve is here advanced 3 cm. 



Calculated according to least square method after the 

 formula 



in which the letters have the same signification as before 

 and a is a new constant. 



If the distances are taken from the water-surface, we 

 shall receive 



h = 91,76 and a = 1,32. 



If they are taken from the meniscus, we shall obtain 

 Ä; = 67,10 and a = 1,3 o. 



The former constant gives the quantity of water with 

 the probable error of O, o 7 the låter with 0,0 6. The latter 

 value is a little better. 



§ 4. The discharging sparks, which hitherto bad gone 

 between the brass-spheres of the radies 0,52 cm and a 

 spärk-gap of 0,i2 cm vvere now displaced to a spark micro- 

 meter better constructed for its end than the former. 



The average radius of the spheres was here 0,6 6 and 

 the distance between them or the spark-gap 



0,^79,5 



where 1^^ has the value of 



1,00 9 6 mm. 



The gap can thus be taken 



0,0 8 cm. 



