132 
NATURE 
[APRIL I, 1915 

choice of the performer. Of recent years, however, 
personal control has been supplemented by accent 
devices, automatically operated, the construction of 
which is essentially an engineering problem. The 
separation of the scale into two independently con- 
trolled halves was certainly to be found in the old 
mechanical pianotist in 1902, and it may date from 
still earlier. JI consider this last feature unnecessary, 
and its use open to serious objections. 
The object of my experiments has been to apply 
dynamical and physical principles to the control of 
what I will call the striking action in piano-players, 
as it very soon appeared to me probable that by so 
doing it would be possible to obtain differences of 
effect that could not be produced by purely mechanical 
methods. 
The ordinary practical man asserts that in striking 
a note or chord on the piano you can increase or 
decrease the force of the blow, but that the only effect 
will be to play the whole chord louder or softer, so 
that you cannot bring the bass or treble parts into 
prominence unless you connect the pneumatics with 
different degrees of vacuum. I have been told that it 
is mathematically impossible to produce effects with- 
out which I now regard no piano-player as worth 
playing. 
But any mathematical physicist will understand that 
what engineers call the “force of the blow” is in 
reality a very complex phenomenon. In my piano I 
find that a fairly soft note is produced in the middle 
of the scale when the hammer strikes the strings with 
a velocity of 30 cm. per second, and that the hammers 
themselves rise through a height of about 5 cm. This 
means that if the acceleration were uniform the opera- 
tion of depressing the key and releasing the hammer 
would occupy one-third of a second. But during the 
operation the pressure applied to the key may be in- 
creased or decreased in an infinite number of ways. 
It may be made very large at the commencement of 
the blow, sinking to zero at the end, as when a finger- 
pianist strikes the note from a height; or it may be 
very small at first and gradually increased, an action 
which some describe as a ‘‘caressing’’ touch. These 
differences would be represented by differences in the 
shape of the graph connecting the pressure on the key 
with the time measured from the instant it is first 
touched to the instant that the note is sounded. We 
might call them differences in the ‘‘shape of the 
blow.’’ But the check action greatly influences the 
character of the accelerating force impressed on the 
hammer when regarded as a function of the time. 
The operation of playing a note is usually divided 
into two periods. At the end of the first period a 
support is withdrawn from the palet which raises the 
hammer, and the latter becomes disconnected. In the 
second period either the hammer may fly up freely and 
strike the note, or the palet may again overtake it 
and drive it forwards. To produce these different 
actions the variable forces applied to the key of a 
piano must evidently be similar in character to the 
“screw-back’’ and “following” strokes in billiards. 
It is, of course, theoretically possible that the hammer 
may rebound and strike the string a second time. 
But it is very difficult to test this point. 
Another important point is that, as the driving force 
is applied close to the base of the hammer, consider- 
able flexural vibrations are liable to be set up in the 
shaft. These may probably differ in amplitude and 
phase, according to how the note is struck. A differ- 
ence in this respect may affect the tone-quality. 
Now the bass and treble hammers differ consider- 
ably in weight, and therefore also in inertia, and the 
intermediate hammers vary continuously from one end 
of the scale to the other. Jt follows that a short sharp 
NO. 2370, VOL. 95] 
| 

i (vol. xci., p. 
blow will produce its greatest effect in the higher, 
parts of the scale, while a longer sustained blow, or 
an increasing blow, will drive the bass notes forward 
with increased velocity even after the treble notes have 
been released. ) 
In the case of a repeated treble note with bass 
accompaniment the second time it is struck, accentua- 
tion is more difficult, and there is considerable danger 
of the note failing to sound owing to the bridge be- 
tween the two notes buckling, thus continuously ad- 
mitting air to the primary valves. The best plan is 
to keep the tension low until the chord containing the 
repeated note has passed the tracker board, and then 
force the notes down hard. On account of the addi- 
tional difficulty thus incurred in reading the music, I 
believe that it would be justifiable to cut the trouble- 
some note a little after the accompanying chord. 
The apparatus originally used in these experiments 
was figured and described in Nature of May 8, 1913 
246), but the principle has now been 
embodied in a patented device which the Moto Music 
Company, of 42 Eyre Place, Edinburgh, have under- 
taken to fit to any make of internal or detachable 
player, and which in the case of a player-piano does 
not interfere with the appearance of the instrument 
or its use when required for hand playing. In addi- 
' tion to the suction bellows which generates the vacuum, 
all pneumatic players have a large reservoir bellows, 
which acts as an accumulator or condenser of consider- 
able capacity, but between this and the playing pneu- 
matics two channels of communication usually exist. 
One is through an accent valve controlled by a lever, 
the other connection is through a smallish regulating 
bellows, controlled usually by a spring, and through 
an air valve which opens or closes with it according 
to the degree of vacuum. This arrangement is some- 
times called a ‘‘ choker,’ and its statical action tends 
to equalise the pressure of the air from the playing 
pneumatics. 
Now it will be found that in playing the regulating 
bellows is in a continual state of vibration, and that 
this vibration has a very marked effect upon the tone 
quality and expression. The character of the variable 
force applied in projecting the pianoforte hammer is 
determined very largely by the elasticity of the con- 
trolling spring, and the fact that this remains constant 
is found to account for the dulness and want of 
variety which is noticed even in the best players when 
they have been in use for a certain time. In my 
device the spring is replaced by a weight the leverage 
of which can be varied, thus enabling the tension on 
the bellows to be varied from time to time in playing 
different passages, and introducing, further, a variable 
element of inertia. It is convenient to describe such 
changes as ‘‘sub-permanent.” Further, the bellows 
can be controlled by a hand lever for the purpose of 
accentuating individual notes, such changes being de- 
scribable as ‘‘temporary.’’ The usual expression 
marks, F F, F, MF, P, and PP, are indicated. 
The experiments lead to the following conclusions :— 
(1) A light sub-permanent tension with fairly strong 
pedalling will give bright treble effects with light 
bass. A heavy sub-permanent tension with light 
pedalling produces a strong bass and a soft treble. 
(2) Corresponding to every note of the scale, there 
is an action which produces the maximum effect, this 
action varying continuously from one end of the scale 
to the other. 
(3) The brightest effects are obtained by keeping 
the accent valve open, or partly so, as in this case 
the closing of the regulating bellows is affected by the 
pedalling. The effect of closing the accent valve is 
very similar to the use of the soft pedal. 
(4) In playing solo passages, light and heavy sub- 
