626 PROFESSOR CHRYSTAL ON THE DIFFERENTIAL TELEPHONE. 
passing from I. to III. the lower tones are more intensified than the higher. 
When the resistance (R) of the sender and line is small compared with (S), 
that of one of the coils of the receiver, this effect is very remarkable. If the 
current be interrupted by rasping the wire on a file, the effect of III. com- 
pared with I. is most curious from the deep croaking character which the 
sound assumes. 
When the resistance and self-induction of the sender and line are small, the 
sound is louder with III. than with II.; but if either one or other of these be 
large enough, the result is the other way. 
Again, if the resistance and self-induction of the line and sender be small, I. 
is better than IJ.; but if the self-induction of the line be above a certain limit, 
II. is better than I. for all tones; and if the resistance of the line be above a 
certain limit, and its self-induction not too great, then I. will be better than II. 
for some tones, and worse for others. 
Experiment 2,—An ordinary Bell telephone was placed mouth downwards 
on a box on which a clock was ticking, and another telephone used as a 
receiver. When the secondary of a small induction coil with a core, the 
primary being open, was inserted in the line the sound became inaudible. 
It was found, however, that it could be restored by connecting a condenser 
of a certain capacity in multiple arc with the secondary coil, the (inductionless) 
resistance in the condenser branch being equal to that of the coil. 
This is in accordance with what has been shown above, viz., that an arc of 
two branches of equal resistance Q, one having no capacity but induction L, 
the other no induction but a capacity X, is equivalent for all disturbances to a 
resistance @ without induction or capacity, provided XQ?=L. 
Experiment 3.—It was found that under certain circumstances sounds could 
be heard better with a condenser in the line than through a line of the same 
resistance and self-induction without a condenser. In other words, self- 
induction can be compensated by introducing capacity. I mean to return at 
some future time to this experiment, from which I hope to get some important 
results. I may simply mention that the result is quite in accordance with 
theory.* A similar case is that studied by Koniravuscu (see article 
Electricity, ‘‘ Encyclopsedia Britannica,” vol. viii. p. 49), mm his investigations 
on electrolytic resistance. 
Experiment 4.—One coil of the differential telephone was used as a 
receiver ; the sender was an apparatus at a distance for making a momentary 
rasping contact. In point of fact, a small piece of watch spring attached to a 
clock pendulum and rasping on a milled head was frequently used, at other 
times a microphone, Two or three Le Clanché cells furnished the electromo- 
tive force. Sometimes a Bell telephone was used as a sender, and then the 
* The compensation according to theory is exact for one particular note only, 
