414 Prof. A. P. Chattock on the Velocity and 



Experiment hy " Double Position " Method. 





Positiye 



Discharge. 







Negative 



Discharge. 





11. 



z. 



P. 





n. 



z. 



P. 





2 



100 



0-707 



0-231 



1 



100 



0-568 



0167 



3 



70 



•476 





4 



70 



•401 





6 



100 



•720 



0-244 



5 



100 



•568 



0179 



7 



70 



•496 





8 



70 



•389 





10 



70 



•492 



0-236 



9 



70 



•403 



0-159 



11 



100 



•728 





12 



100 



•562 





14 



70 



•512 



0-222 



13 



70 



•376 



0-180 



15 



100 



•734 





16 



100 



•556 









Mean .. 



0-228 







Mean .. 



0-171 



n gives the order in which the readings were taken ; z is 

 measured in screw-threads; P in scale- divisions of the pressure- 

 gauge ; and the fourth column is the change in P correspond- 

 ing to a change of z from 70 to 100 threads. The positive 

 readings show well the kind of pressure-variation which may 

 occur. It happens here to be one of increase with time, hut 

 that it has i ffected the pressure for both values of z about 

 equally follows from the constancy of the fourth column. 

 If the small pressures had all been taken either before or after 

 the large ones, as they would in an ordinary curve, an error 

 of about + 8 per cent, would have been introduced into the 

 resulting value of V. 



It is curious how little connexion there seems to be between 

 the variations for + and — discharge, though taken in rapid 

 alternation. This is characteristic of most of the readings. 

 If anything they tend to alter in opposite directions. It 

 should be noted that these facts prove the variations to be 

 electrical in their origin, and not, as might otherwise have 

 been suspected, due to imperfections in the pressure-gauge, 

 which worked smoothly and well. 



Of course if the pressure-variations are not gradual the 

 double-position method is powerless to counteract them. This 

 is probably the reason of the somewhat large variations in the 



