> Historian Ed Berlin sends this article that appeared in the
> Kansas City Star, April 26, 1897, page 2.  [snip]

Ye Gods.  It sounds like this was a piano that used a conventional
sounding board, but used the strings as electro-mechanical oscillators,
like the vibrating tongue in a buzzer.  You could have done it this way
in 1897: for each string, connect an electromagnet, a carbon or other
fast-acting, low-actuation-force variable resistor or switch
(hereinafter referred to as a "microphone", only it doesn't have a
diaphragm), a manually-operated key switch, and a DC power supply.
The actuator of each 'microphone' is glued onto its particular string
somehow.

When the key switch is pressed, the string is pulled toward the
electromagnet.  Because of the 'microphone' placement, the string's
motion increases the circuit resistance and thus reduces the current
through the electromagnet.  The electromagnet thus releases the string,
which can then swing back to its original position.  This movement
pushes on the microphone actuator to decrease the resistance of the
'microphone,' thus increasing the current through the magnet.  The
magnet thus pulls upon the string once again, and so on until the key
switch is released and the current in the circuit is stopped
completely.

The pitch will be determined completely by the string tuning, and the
sound won't stop when the key switch is released: even though we get a
varying DC signal out in the device, we don't use it for anything.  The
tone would be determined principally by the sounding board construction
and all the other stuff that makes a piano sound like it does.

So this instrument is electrically and mechanically equivalent to a
rather refined bank of tuned automobile horns.  Strings take the place
of the horn diaphragms.

The concept is rather clever, and it would be fun to try to reproduce
it now: we don't presently have instruments that work this way.  In
1897, the thing would have been an extreme pain to build: you'd need a
lot of electromagnets strong enough to energetically attract a tight
steel string (I'll bet he didn't have much luck with the high notes)
and you'd need one 'microphone' per string.  The instrument must have
looked like a telephone exchange inside.

If we built one now, optical sensors or amplified magnetic pickups
could substitute for the old carbon microphones, and I suspect that our
electromagnets are a good deal better than those available in 1897.
There'd still be lots of wiring.

I suspect that the volume level wasn't very high.  The attack and decay
must have been interesting, though.  And a player mechanism for such an
instrument would have been pretty simple: no pneumatic amplification
needed.

Robbie commented:

> I wonder if the microphone is, in reality, a tiny switch.  I believe
> that in 1897 only the carbon-button microphone was available, as
> employed in the telephone transmitter unit.

There were a number of screwy microphones (e.g., electrolytic devices
that used a conductive liquid) by 1897, but all of them were
essentially variable-resistance devices like the carbon-button
microphone.

> Yes, the instrument probably produced a sine-wave tone, like a
> tuning fork.  Rather boring, I suppose!

Well, nothing in the article says that a conventional acoustic sounding
board isn't present.

Mark Kinsler