634 PROFESSOR CHRYSTAL ON THE DIFFERENTIAL TELEPHONE. 
It appears, therefore, that simply closing the neighbouring circuit increases 
all the tones of a composite sound. For very low tones the ratio of increase is 
less than for high tones ; but somewhere above U’ this ratio becomes practically 
constant. Hence, the whole effect will be general increase of intensity with a 
certain amount of sharpening in quality. 
If we introduce a condenser, the result will be as follows :— 
If the capacity be below a certain small quantity, the effect will be nil; if 
above a certain very large quantity, the effect will, as might be expected, be 
the same as if we closed the neighbouring circuit simply. 
If we increase the capacity from the lower limit, the effect on a given tone 
will be as follows:—At first its intensity is diminished, and goes down toa 
minimum; then it increases again, and for a certain capacity is unaltered; on 
still further increasing, the intensity rises to what it would be if the neighbour- 
ing circuit were closed simply; then it reaches a maximum, after which it falls 
off again to the value it has when the neighbouring circuit is closed simply. 
As we descend in the scale of audibility, the critical values of the capacity 
increase (ultimately more rapidly than at first). Hence the effect on a com- 
posite sound of increasing the capacity is to lower the tones most affected. 
Thus for a certain small capacity the low tones are comparatively unaffected, 
or even diminished; while the high tones are reinforced, there is then very 
marked sharpening. For a certain greater capacity the higher tones are all 
increased in the same ratio, while certain low tones are increased in a greater 
ratio; there is then flattening, though it may not be so very apparent in certain 
cases, owing to the diminishing of tones lower still. 
The effect of the condenser is therefore partly (but only partly) analogous to 
that of a resonator which reinforces certain tones to the neglect of others. 
The above theory explains the result obtained by Grant, and, as far as my 
judgment goes, the additional results which I observed in my repetition of his 
experiment in a more striking form. 
The discussion of experiment 8 would introduce nothing new after what has 
already been said, I therefore omit it. 
In conclusion, I should like to mention another cause that may have played 
some part in certain of the above experiments. 
Comparing the electrical oscillations to the vibrations of a gross material 
system, it will be observed that I have been discussing above merely forced 
vibrations; but there are also natural or free vibrations. Let us take the 
simple case of a condenser of capacity X, discharging through a circuit of 
templated of course in Hrtmuortz’s theory. It is one of the points of the above investigation to have 
(I hope) made clear the exact nature of this peculiar class of exceptions to the general statement that the 
telephone does not greatly alter the quality of the transmitted sounds. In connection with this matter 
I may refer to an exceedingly interesting little paper by HAcrenBacu, Wied. Ann. N. F, vi. p. 403. 
