of Edinburgh, Session 1877 - 78 . 
623 
Now, it may be asked, are all such sound-exciting currents to 
be traced to the shifting of magnetic matter in a magnetic field 1 
or is there not ground for believing that a blow for the instant 
disturbs the magnetic condition of a magnet, and magnetises soft 
iron, in consequence of a disturbance of the molecules, enough to 
excite telephonic currents ? And may not such magnetic changes 
act without the displacement in the field that generally or 
2 
perhaps necessarily accompanies them, and heightens their effect • 
Wiedemann has shown that blows alone given at right angles to 
the magnetic needle, and consequently apart from the earth’s 
magnetism, produce permanent magnetic changes, and that torsion 
and change of external conformation may do the same. The 
permanent effect also of blows in lessening the powers of a magnet, 
or of giving to iron under induction a magnetic set, has long 
been known. Now, may not the telephone reveal to us that they 
in every case produce a momentary effect of an intensity in some 
degree proportionate to the magnetisation of the iron % 
The following telephonic arrangement, which is different from that 
of Bell’s, may possibly have some bearing in the view just suggested ; 
but knowing, from the instances discussed, the ambiguity that 
attends a departure from the pusli-and-pull theory, I quote it more 
as a problem for push-and-pull solution, than as a proof of blow- 
induced effects. The apparatus (fig. 6), consists of a horse-shoe 
magnet rather more than 5 inches long, with a pin on each pole 
standing at right angles to the plane of the magnet. The pins are 
provided each with a coil of fine copper wire, the two being joined 
up as in an electro-magnet. The pole pins are inch apart, and 
a stiff disc (No. 28), 3 inches diameter of untempered steel, is 
screwed to them by large-headed screws. Such a disc, when aided by 
a wooden or non-metallic resonating box, can both convey and receive 
an articulate message, indistinctly certainly, but still it does it. 
When it is joined up with the galvanometer, and tapped with a 
pencil on any part, taking care of course not to combine the 
