Specific Velocities of Ions in the Discharge from Points. 97 



water to A A l , and then to E E with oil of specific gravity 0'88. 

 The other letters in the figure have the same meanings as in 

 fig. 1. 



Pig. 9. 



As far as sensitiveness is concerned, the level of A responds 

 to differences of pressure applied to the upper ends of the U 

 as readily as though the gauge were filled with a single 

 liquid of density equal to the difference between the densities 

 of the water and the oil ; but if the pressure is measured by 

 the displacement of A in the tube temperature troubles arc 

 introduced which are serious in proportion to the sensitiveness, 

 owing to the unequal expansions of water and oil. In addition 

 to this there is the danger that, when the pressure measured 

 is small, an appreciable part of this pressure may be used in 

 distorting the oil-water surface, and thus fail to be observed. 



Both these sources of error may be avoided by the device, 

 already alluded to, of keeping the level of A always in the- 

 same position in the tube by screwing S up or down, and so 

 tilting the whole gauge. 



If A is not allowed to move far from its zero position, 

 which will be the case if the pressures are small, the shape of 

 the oil-water surface will be constant when brought back to 

 the cross-wire, and the surface-tension error will therefore 

 vanish. On the other hand, constancy in the position of A 

 implies constancy in the heights of the oil columns A E, with 

 the consequence that pressure-differences, p, between the 

 limbs of the U are measured by a column of water only, of 

 height equal to the change in the difference of level between 

 Ai and A introduced by the tilt, according to the formula 



p = nxzjy^ 

 where n is the amount by which S is moved, reckoned in 



Phil Mag. S. 6. Vol. 1. No. 1. Jan. 1901. H 



