178 Mr Wilberforce, On some experiments on the [Oct. 27, 



by Glazebrook {Phil. Mag., Aug. 1884). He has also mentioned 

 some of its uncertainties. One of its principal disadvantages 

 appears to me to be the fact that though it is a null method 

 inasmuch as the equilibrium position of the needle is not to be 

 disturbed by the passage of the transient currents through the 

 galvanometer coils, yet it is not so inasmuch as the needle instead 

 of never moving from this position is always jerked aside when the 

 currents either begin or cease to pass, and hence, instead of 

 observing an absence of effect we have to measure the positions of 

 a spot of light at two times, and determine whether they are the 

 same. This initial jerk is due to the fact that the transient and 

 constant current do not balance at every moment, and that their 

 average effect requires some time to show itself. Thus this 

 disadvantage is inherent to the method. 



The observations are thus not only rendered more difficult, but 

 a considerable element of uncertainty is introduced as to whether 

 the zero point of the needle may not have changed between the 

 measurements. 



We may take for our zero point the mean of the points at 

 which the needle rests before and after the currents pass, but this 

 will not be satisfactory if the change of zero is at all rapid, as it 

 sometimes was. 



I had attempted three months previously to make some deter- 

 minations by this method but rejected them as the results obtained 

 were untrustworthy owing to the unsteadiness of the galvanometer 

 needle, and also hopelessly inconsistent among themselves. 



The unsteadiness was partly due to the effects of draughts 

 which were afterwards excluded by enclosing the galvanometer in 

 a wooden box with a glass shutter, partly to the weakness of the 

 field of force in which the needle hung, which was also remedied, 

 while the inconsistency and a great deal of the unsteadiness was 

 caused by leakage due to imperfect insulation. After this was 

 discovered, the greatest precautions were taken to secure good in- 

 sulation, wires covered with gutta-percha being used for all the 

 connections, and all the apparatus used being supported by pieces 

 of glass coated with paraffin. 



For our calculations of the capacities it was still necessary to 

 know as accurately as possible the rates of the forks used. 



The "32" and "64" forks were compared directly. It was 

 then discovered that the "32" fork was wrongly weighted, so as 

 to vibrate about 32^ times a second. As this would prove trouble- 

 some when it was compared with the standard, it was reweighted 

 and compared afresh with the " 64" fork. The fork as reweighted, 

 which we shall call the J32j fork, was then made to drive a fork 

 of about four times its frequency, and this latter was compared 



