174 HENRY A. ROWLAND 



pared with the standard, the difference noted, and the resistance again 

 adjusted. The observations were completed by turning the induction 

 coils into the two other positions which they could occupy with respect 

 to each other, followed by another comparison of resistance with 

 standard. 



5th. Observations of a and ft were again made as before. 



6th. The time of vibration was again determined. 



The observations as here explained furnished data for three compu- 

 tations of the resistance of the circuit, one with each of the three cur- 

 rents. In each of these three computations, a was the mean of 16 

 readings, ft of 8 or sometimes 12, 6 of 16 and 3 of 16. In using the 

 method of recoil nearly the same order was observed. 



The time of vibration was determined by allowing the needle to 

 vibrate for about ten seconds and making ten observations of transits 

 before and after that period. During the experiment, I usually ob- 

 served at the telescope and Mr. Jacques at the tangent galvanometer. 



The methods of obtaining the corrections require no explanation. 



RESULTS 

 The constant corrections are as follows for the first needle. 



a=-J^+ T ^A= - -00711. 



J = - H = -00020 , 



c = -000006 , 



d = + -000074 at 20' C . 



/ = + -00003 , 



a + b + c + d + e +/ '00718. 

 For method of recoil it becomes -00016. 

 Hence for A and B, log JT= 11-4536030 

 Hence for A and 0, log # = 11-2852033 

 Hence for B and C, log #=11-1886619 

 For method of recoil using A and B, log K = 11-4566.630. 

 For second needle and method of recoil, 



a = } f V = - -000050 , 



V * / 

 &=}$= - -00025, 



c = -000006 , 

 d = + -000074 , 



