DBS. GUY BARLOW AND H. B. KEENE ON THE ANALYSIS OF SOUND. 
139 
The currents were very nearly pure, but contained traces of the even harmonics. 
Thus the analysis for the 95 fork gave the current-amplitudes of the components n, 
2 n, 4n, in the ratio 1 : : ( . 1 0 -. These harmonics may have been present in the fork 
vibration, but it is more likely that they were produced by want of uniformity in the 
magnetic flux of the U-magnet. The so-called subharmonic responses described above 
were also observed. 
(2) Close Pair of Simple Harmonic Components — Resolving-Power. 
The current was produced by putting in series two U-magnet generators on separate 
forks of frequency 64/sec., which could be adjusted to have a slight difference of frequency 
by means of sliding loads on one of them. The responses were examined for each alone, 
and then for both together. When the difference in frequency was reduced to 0-6/sec. 
the two components could still be resolved, owing to a distinct drop between the two 
maxima. Separation was also effected when the components were very unequal in 
magnitude, e.g. a ratio 5:1. In all cases the double nature of the response was at 
once evident from the characteristic beating of the galvanometer oscillations, whereby 
it was readily distinguished from that due to a single frequency. When the components 
are unequal the beating is more distinct on the side of the smaller. The difference in 
frequency of the components can be determined directly by the frequency of the beats 
without observing the positions of the two maxima. 
Experiments on resolving-power were not made at higher frequencies, but theory 
shows that two components with the above limit of frequency-difference should be 
resolved whatever their absolute values, e.g. 1000 and 1000-6/sec. 
(3) Current Containing Harmonic Series. 
Two types of current were produced simultaneously from the same electrically main¬ 
tained fork with mercury contact, 32/sec.—the first by induction in a couple of turns 
of wire round the electromagnet of the fork, the second by using a small air-transformer 
consisting of a few turns of wire (giving negligible self-induction) in which the current 
in the primary was interrupted by a separate platinum contact attached to the fork- 
prong and dipping into mercury. The level of the mercury was adjusted to give equal 
time intervals of make and break. By means of a “ throw-over ” key the amplitude of 
corresponding harmonics in the two currents could be compared. 
YOL. ccxxii.—-A. 
X 
