﻿300 Dr. S. R. Milner on the Use of the Secohmmeler 



Table I. 

 S, 7650a>. K/4314mf. Q,4000*>. H, J mf . ^=1-291 



K arbitrary to 



E/S 



Eeversals per sec 



P required for balance ... w 



P/Q 



P calc. (seep. 308) to 



Diff.P exp . -P calc . CO 



5000 

 •6356 



r 



69-2 

 4260 

 1-065 

 4317 



-57 



130 

 3585 

 •896 

 3621 

 -36 



15000 

 1-961 



These results become understandable i£ we represent by 

 time-curves the effect of the commutators on the E.M.F. o£ 

 the battery and the current through the galvanometer. 



In the usual form of the instrument the two commutators 

 do not reverse instantaneously ; one, X (in fig. 1 the battery 

 commutator), breaks the circuit for a certain fraction of the 

 whole period between two successive reversals (" reversal 

 period ") : the other, Y, short-circuits for the same fraction of 

 the reversal period. (In Nalder's instrument the fraction is 

 almost exactly ^ in each case.) The four brushes of X which 

 make contact in pairs on the two insulated brass pieces 

 attached to the rotating axle are fixed on a plate which can 

 be turned by hand around the axle. This is very convenient, 

 as it enables the position of the brushes to be adjusted (even 

 while the axle is in rotation), so that the reversals of the 

 battery occur either simultaneously with, or at any assigned 

 fraction of the reversal period before, those of the galva- 

 nometer. We shall call this fraction the lead, 8, of the battery 

 over the galvanometer commutator. 



In fig. 2, Curve I. represents the E.M.F. of the battery 

 after commutation plotted with the time, ac is the reversal 

 period, ah the " contact," and be (= \ac) the " break" period. 

 During ah the E.M.F. has a finite value E, while during he 

 we may represent it as zero. With non-inductive resistances 

 only in the arms, the curve for the current in the galvanometer 

 branch before commutation by Y will be similar in shape to 

 that for the E.M.F., and may also be represented by I. The 

 introduction of the capacities H and K into the arms will in 

 general produce distortion and give exponential curves for 

 the galvanometer current in both ah and he (see II., fig. 2). 

 To determine the effect on the current through the gal- 

 vanometer of the reversals of the commutator Y at any lead, 

 mark on the diagram the centre points of the short-circuiting 

 periods of Y at a distance, proportional to the lead, to the 



